Handbook of
HUMAN FACTORS in MEDICAL DEVICE DESIGN
TAF-ER562X_WINGR-09-0401-C000.indd i
11/9/10 1:53:35 AM
TAF-ER5...
414 downloads
2989 Views
25MB Size
Report
This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form
Handbook of
HUMAN FACTORS in MEDICAL DEVICE DESIGN
TAF-ER562X_WINGR-09-0401-C000.indd i
11/9/10 1:53:35 AM
TAF-ER562X_WINGR-09-0401-C000.indd ii
11/9/10 1:53:35 AM
Handbook of
HUMAN FACTORS in MEDICAL DEVICE DESIGN Edited by
Matthew B. Weinger Michael E. Wiklund Daryle J. Gardner-Bonneau Assistant Editor
Lori M. Kelly
Boca Raton London New York
CRC Press is an imprint of the Taylor & Francis Group, an informa business
TAF-ER562X_WINGR-09-0401-C000.indd iii
11/9/10 1:53:35 AM
CRC Press Taylor & Francis Group 6000 Broken Sound Parkway NW, Suite 300 Boca Raton, FL 33487-2742 © 2011 by Taylor and Francis Group, LLC CRC Press is an imprint of Taylor & Francis Group, an Informa business No claim to original U.S. Government works Printed in the United States of America on acid-free paper 10 9 8 7 6 5 4 3 2 1 International Standard Book Number-13: 978-1-4200-6351-6 (Ebook-PDF) This book contains information obtained from authentic and highly regarded sources. Reasonable efforts have been made to publish reliable data and information, but the author and publisher cannot assume responsibility for the validity of all materials or the consequences of their use. The authors and publishers have attempted to trace the copyright holders of all material reproduced in this publication and apologize to copyright holders if permission to publish in this form has not been obtained. If any copyright material has not been acknowledged please write and let us know so we may rectify in any future reprint. Except as permitted under U.S. Copyright Law, no part of this book may be reprinted, reproduced, transmitted, or utilized in any form by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying, microfilming, and recording, or in any information storage or retrieval system, without written permission from the publishers. For permission to photocopy or use material electronically from this work, please access www.copyright.com (http://www.copyright.com/) or contact the Copyright Clearance Center, Inc. (CCC), 222 Rosewood Drive, Danvers, MA 01923, 978-750-8400. CCC is a not-for-profit organization that provides licenses and registration for a variety of users. For organizations that have been granted a photocopy license by the CCC, a separate system of payment has been arranged. Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe. Visit the Taylor & Francis Web site at http://www.taylorandfrancis.com and the CRC Press Web site at http://www.crcpress.com
Dedication The editors express their gratitude to Peter Carstensen for making this Handbook possible.
After a distinguished career in aerospace, Peter spent over 30 years with the U.S. Food and Drug Administration (FDA), where he introduced the agency to human factors engineering and ultimately served as its human factors team leader. His intelligence, patience, and wit served him well in his role as a human factors advocate within the government and as a convener of and participant in international human factors standards committees. Having retired from the FDA in 2008, Peter is still improving medical device safety and usability through his efforts as a human factors engineering and regulatory compliance consultant. On joining the FDA in the mid-1970s, Peter and his colleagues recognized that medical device mishaps were frequently related to user-interface design shortcomings. Beginning with initiatives to make anesthesia equipment safer through the application of human factors design principles and more recently leading the international community to establish standards for the application of human factors in medical device design, Peter’s tireless work helped give human factors a “seat at the table” in settings ranging from engineering meetings to regulatory reviews. As a result, with support from his FDA colleagues (notably Dick Sawyer, Robert Cangelosi, Ron Kaye, and Michael Mendelson), Peter made a substantial difference in the world, improving the safety of medical devices and likely saving an untold number of patient lives.
TAF-ER562X_WINGR-09-0401-C000.indd v
11/9/10 1:53:36 AM
TAF-ER562X_WINGR-09-0401-C000.indd vi
11/9/10 1:53:36 AM
Contents Introduction ..................................................................................................................... ix Michael E. Wiklund Chapter 1 General Principles ...........................................................................................1 Michael E. Wiklund and Matthew B. Weinger Chapter 2 Basic Human Abilities...................................................................................23 Edmond W. Israelski Chapter 3 Environment of Use .......................................................................................63 Pascale Carayon, Ben-Tzion Karsh, Carla J. Alvarado, Matthew B. Weinger, and Michael Wiklund Chapter 4 Anthropometry and Biomechanics ................................................................97 W. Gary Allread and Edmond W. Israelski Chapter 5 Documentation ............................................................................................153 John W. Gwynne III and David A. Kobus Chapter 6 Testing and Evaluation ................................................................................201 Edmond W. Israelski Chapter 7 Controls ......................................................................................................251 Stephen B. Wilcox Chapter 8 Visual Displays ............................................................................................297 William H. Muto and Michael E. Maddox Chapter 9 Connections and Connectors ....................................................................... 351 Joseph F. Dyro Chapter 10 Alarm Design ..............................................................................................397 Stephen B. Wilcox Chapter 11 Software User Interfaces ............................................................................425 Michael E. Wiklund vii
TAF-ER562X_WINGR-09-0401-C000.indd vii
11/9/10 1:53:36 AM
viii
Contents
Chapter 12 Workstations ................................................................................................471 Michael E. Wiklund Chapter 13 Signs, Symbols, and Markings ....................................................................543 Michael J. Kalsher and Michael S. Wogalter Chapter 14 Packaging ....................................................................................................595 Michael E. Maddox and Larry W. Avery Chapter 15 Device Life Cycle ........................................................................................623 Michael E. Maddox and Larry W. Avery Chapter 16 Hand Tool Design ........................................................................................645 Richard Botney, Mary Beth Privitera, Ramon Berguer, and Robert G. Radwin Chapter 17 Mobile Medical Devices..............................................................................715 Richard Stein and Michael E. Wiklund Chapter 18 Home Health Care .......................................................................................747 Daryle J. Gardner-Bonneau Chapter 19 Cross-National and Cross-Cultural Design of Medical Devices .................771 Uvo Hoelscher, Long Liu, Torsten Gruchmann, and Carl Pantiskas Editors ............................................................................................................................795 Contributors ..................................................................................................................797 Index .............................................................................................................................. 805
TAF-ER562X_WINGR-09-0401-C000.indd viii
11/9/10 1:53:36 AM
Introduction Michael E. Wiklund The editors and contributing authors developed this Handbook to promote the design of safe, effective, and usable medical devices. Broadly speaking, it is a response to the historically high rate of injuries and deaths that have been directly or indirectly caused by medical devices with user-interface design shortcomings*. Following the handbook’s guidance should help medical device developers create user interfaces that are less prone to use error. The resulting devices will be more likely to support the highest quality of medical care while enhancing user productivity.
ABOUT HUMAN FACTORS Readers who are new to human factors should be excited to learn how much the discipline has to offer the medical device industry, health care providers, and patients. By applying knowledge about human characteristics to produce effective interfaces, human factors practitioners help to create medical devices that are easier to learn and use. Specifically, a well-designed user interface enables device users, such as physicians, nurses, therapists, and technicians, to draw on their past experience to use a new device, avoid errors, quickly detect and recover from errors when they occur, feel confident about their decisions and actions, and be physically comfortable. As such, a good user interface can improve safety and the quality of patient care. Conversely, a poorly designed user interface can respond in unexpected ways to user actions, increase workload, induce critical errors that jeopardize patient and clinician safety, and cause fatigue and discomfort. Medical device developers who invest in human factors find that increased user interface quality gives them a competitive advantage in the marketplace because customers prefer their products to other offerings. The investment also meets a standard of care that has been established over several decades as more companies incorporate human factors processes and design practices into their device development process. Moreover, medical device manufacturers are now compelled to invest in human factors by international standards† and the expectations set forth by medical device regulators, such as the U.S. Food and Drug Administration, Health Canada, and Germany’s Federal Institute for Drugs and Medical Devices.
HANDBOOK CONTENT As a prelude to detailed design guidance, Chapters 1 through 6 provide advice on fundamental topics, such as aligning the interactive nature of medical devices to the expected use * Institute of Medicine. To err is human – Building a safer health system (Washington, DC: National Academies Press, 2000) p. 1. † International Electrotechnical Commission. IEC 60601-1-6: Medical Electrical Equipment – Part 1-6: General Requirements for Safety – Collateral Standard: Usability. 2004. ix
TAF-ER562X_WINGR-09-0401-C000.indd ix
11/9/10 1:53:36 AM
x
Introduction
environments ranging from hospitals to ambulances to patients’ homes, drawing on anthropometric and biometric data to ensure that designs match the intended users’ bodies and physical abilities, and conducting usability tests and other evaluations to ensure that devices perform as intended. Chapters 7 through 19 have an applied design focus, offering guidance on the design of specific types of devices (e.g., Chapter 16, “Hand Tool Design”) as well as designing devices for particular use environments (e.g., Chapter 18, “Home Health Care”). Adapted in part from established design standards and conventions, the design guidance also reflects professional judgment extracted from many years of applied analysis and design. Therefore, while much of the guidance is founded on empirical data, it is also based on the contributing authors’ applied design and evaluation experience. Accordingly, readers should exercise their own judgment when applying the design guidance and note that published standards should always take precedence. Although this Handbook’s content is the most current available as of early 2007, other resources should be consulted, as needed, to ensure that the design guidelines remain valid.
RELATIONSHIP TO AAMI STANDARDS The genesis for this Handbook was the work of the Association for the Advancement of Medical Instrumentation’s (AAMI’s) Human Factors Engineering Committee, which produced ANSI/AAMI HE-74-2001, Design Process for the Human Factors Engineering of Medical Devices. This standard provides design process guidance but no specific device design guidelines. The committee recognized the need for a comprehensive resource for human factors design guidance that was specifically tailored to medical devices. One of the committee’s first national standards, ANSI/AAMI HE-48-1993, Human Factors Engineering Guidelines and Preferred Ppractices for the Design of Medical Devices, contained a modest amount of design guidance, much of which was adopted from an existing standard for military equipment (MIL-STD-1472—Human Factor Engineering). The Committee concluded that a subgroup of its members, augmented by outside human factors specialists, should independently produce a handbook that the committee could then draw on to produce an updated national standard. As planned, AAMI’s Human Factors Engineering Committee drew content from this handbook to produce its own design standard for medical device user interfaces (ANSI/ AAMI HE-75-2007, a complement to ANSI/AAMI HE-74-2001). However, there are notable differences between the AAMI’s design standard and this Handbook: • This Handbook includes expanded discussions of design issues, product design case studies, and supporting illustrations. • AAMI’s standard has a terse writing style and explicit organizational structure befitting a standard. • Readers can expect minor differences in the guidance found in the two documents. • This Handbook’s content was not vetted by the AAMI committee, which includes representatives from numerous medical device manufacturers as well as independent medical product design professionals.
TAF-ER562X_WINGR-09-0401-C000.indd x
11/9/10 1:53:36 AM
Introduction
xi
INTENDED AUDIENCE This Handbook should be a valuable resource for the following people involved in the myriad aspects of medical device development, marketing, and support: • • • • • • • • • •
Clinical engineers Design strategists/managers Electrical engineers Graphic designers Human factors engineers/ ergonomists Industrial designers Instructional media developers/ writers Marketing specialists Mechanical engineers Medical practitioners
• Patient safety and medical device researchers • Product liability litigators • Product planners/managers • Regulatory affairs specialists • Quality assurance specialists • Risk analysts/managers • Safety engineers • Software developers • User-interface designers
HOW TO USE THIS HANDBOOK The Handbook can serve several purposes: 1. Medical device developers can convert pertinent design guidelines into product requirements. 2. Medical device evaluators, including test and evaluation personnel and regulators, can set performance criteria for design evaluations (e.g., inspections and usability tests) based on the design guidelines. 3. Human factors advocates can cite the Handbook and specific content to raise their organizations’ user-interface quality standards and secure resources for human factors programs. 4. Students and interested professionals can learn about good human factors design practices. Each chapter is written to stand alone, assuming that people will refer to the subject of interest rather than read the Handbook from start to finish. Therefore, certain topics are intentionally addressed by multiple chapters, ensuring the completeness of technical discussion but leading to some inconsistencies because of varying author perspectives and opinions on design. While each chapter presents its own approach to promoting user-centered design, most share a common organization scheme. Most begin with a general introduction to the selected topic, followed by the presentation of general and special design considerations and then specific, numbered design guidelines. Some chapters include one or more cases studies to instantiate the guidance, and conclude with a listing of resources, literature, and Web site references. Guidelines are numbered sequentially and prefixed with the number of the chapter in which they appear (e.g., the guidelines presented in Chapter 11, “Software User Interface,” are numbered from 11.1 to 11.105). Most guidelines employ the term “should” to promote a desirable design characteristic or performance level. In contrast, design standards usually
TAF-ER562X_WINGR-09-0401-C000.indd xi
11/9/10 1:53:36 AM
xii
Introduction
employ the term “shall.” By deliberately using the term “should,” the authors leave room for designer judgment and acknowledge that this Handbook serves an advisory purpose rather than establishing mandates. Chapters include numerous cross references to related guidance found in other chapters. This is especially true of closely related chapters, such as Chapter 7, “Controls,” and Chapter 8, “Visual Displays.” Chapters also include numerous tables and illustrations to further elucidate technical discussions and improve readability.
CLOSING REMARKS Undoubtedly, this Handbook could cover many more topics in the manner of a multivolume encyclopedia. Also, each chapter could expand to cover a wider range of design issues. However, the editors are satisfied that the 19 chapters focus on some of the most important human factors issues facing medical device developers while providing numerous references to other valuable resources. We hope that our readers will appreciate the importance of designing medical devices compatible with human needs and preferences. Enhanced medical device user interfaces are certain to make a positive difference in the lives of caregivers and patients alike.
TAF-ER562X_WINGR-09-0401-C000.indd xii
11/9/10 1:53:36 AM
1 General Principles Michael E. Wiklund, MS, CHFP; Matthew B. Weinger, MD CONTENTS 1.1 Seek User Input............................................................................................................2 1.1.1 Involve Users Early and Often ..........................................................................2 1.1.2 Refine Designs through Usability Testing ........................................................3 1.2 Establish Design Priorities...........................................................................................4 1.2.1 Err on the Side of Design Simplicity ................................................................4 1.2.2 Ensure Safe Use ................................................................................................4 1.2.3 Ensure Essential Communication .....................................................................5 1.2.4 Anticipate Device Failures................................................................................5 1.2.5 Facilitate Workflow ...........................................................................................5 1.3 Accommodate User Characteristics and Capabilities ..................................................6 1.3.1 Do Not Expect Users to Become Masters.........................................................6 1.3.2 Expect Use Errors .............................................................................................6 1.3.3 Accommodate Diverse Users ............................................................................7 1.3.4 Maximize Accessibility ....................................................................................7 1.3.5 Consider External Factors That Influence Task Performance...........................8 1.4 Accommodate Users’ Needs and Preferences .............................................................8 1.4.1 Accommodate User Preferences up to a Point ..................................................8 1.4.2 Do Not Rely Exclusively on “Thought Leaders” ..............................................9 1.4.3 Enable Users to Set the Pace .............................................................................9 1.5 Establish Realistic Expectations of Users ..................................................................10 1.5.1 Do Not Rely on Training ................................................................................10 1.5.2 Do Not Rely on Instructions for Use...............................................................10 1.5.3 Do Not Rely on Warnings ...............................................................................11 1.5.4 Do Not Rely on Memory ................................................................................11 1.5.5 Avoid Information Overload ...........................................................................11 1.5.6 Do Not Assign Users Tasks That Are Better Suited to Other Users or Devices .......................................................................................................12 1.6 Consider Real-World Demands..................................................................................12 1.6.1 Consider the Context of Use ...........................................................................12 1.6.2 Consider Worst-Case Scenarios ......................................................................13 1.6.3 Make Devices as Rugged as Necessary ..........................................................13 1.6.4 Limit User Workload ......................................................................................14 1.6.5 Consider the Potential for Device Migration into Additional Uses or Use Environments ...........................................................................................15 1
TAF-ER562X_WINGR-09-0401-C001.indd 1
11/8/10 11:39:53 PM
2
Handbook of Human Factors in Medical Device Design
1.7 Develop Compatible Designs .....................................................................................15 1.7.1 Accommodate Mental Models ......................................................................15 1.7.2 Establish Natural or Conventional Mappings ...............................................15 1.7.3 Follow Industry Conventions and Consensus Standards ..............................17 1.8 Optimize User Interactions ........................................................................................17 1.8.1 Make Devices Error Tolerant ........................................................................17 1.8.2 Fail in a Safe Manner ....................................................................................17 1.8.3 Avoid Physical Strain, Repetitive Motions, and Cumulative Trauma ...........18 1.8.4 Enable Users to Anticipate Future Events .....................................................18 1.8.5 Confirm Important Actions ...........................................................................18 1.8.6 Make Critical Controls Robust and Guard Them .........................................19 1.8.7 Clarify Operational Modes ...........................................................................19 1.8.8 Employ Redundant Coding ...........................................................................20 1.8.9 Design to Prevent User Confusion ................................................................20 1.8.10 Don’t Neglect Device Appeal .......................................................................21 1.9 Summary ...................................................................................................................22 Resources ...........................................................................................................................22 References ..........................................................................................................................22 Although an understanding of detailed human factors guidelines is helpful when designing a medical device, a command of the general design principles of human factors engineering is critical. After all, clinicians and caregivers are usually able to cope with devices that have specific design shortcomings as long as the flaws do not lead to serious use errors or pose insurmountable obstacles to accomplishing a task. In fact, few medical devices are perfect from a user-interface design standpoint. They usually violate one specific guideline or another. It’s another story altogether if a medical device violates a general human factors design principle. Serious violations, such as presenting information at an excessive pace or expecting the user to carefully read an instruction manual before using a device, can render a medical device unsafe or unusable. Accordingly, designers should focus on meeting the high-level design principles before they perfect the details. After all, there is no sense in refining a fundamentally flawed device. In contrast, great products arise from fundamentally correct solutions that are subsequently honed to a state of excellence. This chapter provides an overview of these human factors. Many of these principles will be echoed or built on in subsequent chapters. This chapter presents several high-level design principles intended to help designers produce fundamentally correct user interfaces. For those readers unfamiliar with the human factors design process, reference to ANSI/AAMI HE-74 (2001) may be of value.
1.1 SEEK USER INPUT 1.1.1 INVOLVE USERS EARLY AND OFTEN Medical device users can offer invaluable guidance at each stage of user-interface development. Early in the design process, they can critique existing devices, explain contextual factors that must be accommodated in the design, offer a vision of user interactions with
TAF-ER562X_WINGR-09-0401-C001.indd 2
11/8/10 11:39:54 PM
General Principles
3
the device, and help set usability objectives. As a user-interface design evolves, users can comment on features they like and dislike, describe design attributes that give them trouble, and participate in more formal usability tests (see Chapter 6, “Testing and Evaluation”). Toward the end of the design process, users can help to verify the quality of a near-final design by participating in a usability test of a working prototype. A high level of diverse user involvement throughout the design process helps to ensure that the final device is well suited to the intended users. It also avoids last-minute design modifications to resolve usability problems.
1.1.2 REFINE DESIGNS THROUGH USABILITY TESTING Usability testing is a critical component of the human factors engineering process. Usability testing is a reliable method by which to discern user-interface design issues that could affect safety, efficacy, and satisfaction. Progressive manufacturers often extend testing beyond primary tasks (e.g., using a defibrillator to shock a patient in cardiac arrest) to include setup, storage, maintenance, and even repair tasks. In a typical test session, representative device users perform typical or critical tasks in an appropriate environment, which may range from a conference room to a sophisticated, high-fidelity simulation of the intended clinical care environment (Figure 1.1). The level of test fidelity usually increases as the device progresses from a concept to a refined prototype. Testing early in the design process and then several more times as the design evolves is an effective way to prevent user interaction problems from persisting into the later stages of the design process, a time when effective solutions to problems are more limited and expensive to implement. It is important to find a sample of test participants who accurately reflect the range of user characteristics rather than choosing “thought leaders” who bring special knowledge and motivation to the test. User characteristics that should be considered include physical attributes (i.e., traditional ergonomics), abilities and skills, needs, and psychological attributes. When a device will be used by several distinctly different user groups (e.g., both physicians and patients), tests should be conducted separately with representative participants from each group.
FIGURE 1.1 Clinicians participating in a medical simulation that incorporates several medical devices and a sophisticated mannequin. (From the Center for Experimental Learning and Assessment, Vanderbilt University, Nashville, TN. With permission.)
TAF-ER562X_WINGR-09-0401-C001.indd 3
11/8/10 11:39:54 PM
4
Handbook of Human Factors in Medical Device Design
1.2 ESTABLISH DESIGN PRIORITIES 1.2.1 ERR ON THE SIDE OF DESIGN SIMPLICITY In medical device design, simpler is usually better. Most medical device users dislike devices equipped with all the “bells and whistles,” especially if the “extras” get in the way of performing basic tasks. Indeed, some medical devices are loaded with features intended to give them a competitive advantage over competing devices. Yet features aimed at enhancing sales can cost a company in terms of customer goodwill if they complicate device operation. Accordingly, developers are well advised to produce devices that focus on the basics and exclude features offering little day-to-day value. The added complexity of “bells and whistles” can interfere with initial ease of use and is usually not worth it. That said, designers have to be careful about eliminating advanced features that offer real value to sophisticated users even if such users represent a small percentage of the overall user population. In such cases, faced with divergent market needs, a manufacturer should consider developing two devices rather than a single, compromised version (Figure 1.2). Similarly, for maintenance tasks, simple is again better. Designers should seek ways to limit the level of skill required to maintain and repair a device as well as the number of steps and the need for special tools.
1.2.2 ENSURE SAFE USE Medical devices should minimize the risk of injury to both users and patients, including physical and psychological injury, during normal and emergency device operation. Applying this principle to a CT scanner, designers would promote design solutions that reduce users’ risk of trauma due to moving parts and of patients feeling claustrophobic as a result of being enclosed in a tight space. Applying the principle to a portable patient
FIGURE 1.2 Two ultrasound scanners (GE Logiq 9 and SonoSite 180Plus System), developed by separate manufacturers, are targeted toward different user populations and use environments. (From http://www.stormoff.com and http://images.google.com. With permission.)
TAF-ER562X_WINGR-09-0401-C001.indd 4
11/8/10 11:39:54 PM
General Principles
5
monitor, designers would avoid placing a heavy instrument on a wheeled pole that is vulnerable to tipping over, which could cause injury to both clinicians and patients, cause property damage, and disrupt the care delivery process. On the other hand, user-interface designers must consider the consequences of dynamic user interactions. For example, in designing a portable glucose monitor for home use, the control–display relationships must be designed to minimize the risk of the diabetic user (who may have a preexisting visual impairment and an acutely abnormal blood sugar) collecting the sample incorrectly or misreading the resulting displayed value. Thus, potential device-induced user harm can either be due to static design characteristics (e.g., mechanical consequences of the physical design) or to use errors during device interaction.
1.2.3 ENSURE ESSENTIAL COMMUNICATION During busy and stressful times, people often must work harder to communicate with each other, and this might lead them to raise their voices, repeat themselves to make sure they were heard, or even grasp someone’s arm to get their attention. Similarly, a well-designed medical device should be capable of reliably communicating critical information, such as a change in a patient’s vital signs that could be life threatening, during busy and stressful moments. Accordingly, designers should employ redundant methods to communicate vital information. Also, where possible, they should provide users with a clear and concise explanation of any problem (including the source) and how to correct it. For example, the device may employ both a sufficiently loud audible alarm and a complementary visual alarm, thereby using two sensory channels to increase the chance of detection. Moreover, the visual alarm might flash to draw attention, and the audible alarm might be set at an attention-getting frequency that some would regard as noxious—a means to motivate users to attend to it. (Note: Standards have taken such needs into account, leading to alarms that some users do find annoying when they are not contextually appropriate; see Chapter 10, “Alarm Design.”) Finally, all designs should be evaluated in the context of the overall use environment, including other devices commonly in use, to ensure that the design solution does not result in unintended consequences, including impaired clinician–patient, clinician– clinician, or clinician–device communication.
1.2.4 ANTICIPATE DEVICE FAILURES Devices will fail. When they do, it is important to communicate the failure to users and, where possible, indicate the cause and recommend appropriate remedial action. This is especially important when a device failure, such as the failure of an air-in-blood detector, places a patient at immediate risk. Ideally, devices will fail safely, but sometimes user intervention is needed to ensure a safe outcome. Therefore, designers should consider the full range of failure modes and develop strategies and detailed user-interface or other solutions for coping with them. As with communicating critical information (see above), it is helpful to communicate both the cause of the device failure and proper remedial or coping actions in clear and concise terms.
1.2.5 FACILITATE WORKFLOW Humans are loathe to change to accommodate new medical devices unless there is a clear benefit in terms of work efficiency or effectiveness. Therefore, designers should take care to
TAF-ER562X_WINGR-09-0401-C001.indd 5
11/8/10 11:39:54 PM
6
Handbook of Human Factors in Medical Device Design
understand the impact that their device might have on task flow, obviously avoiding negative effects. Additionally, designers should analyze how people will use their device, making sure that the user interface is organized to facilitate urgent, frequent, and critical tasks. For example, in a software interface of a physiological monitor, a dedicated, surface-level control for recording a patient parameter waveform would be a better design than relegating this control to a lower-level screen. Potential device uses should be analyzed formally using techniques such as contextual inquiry, task analysis, or usability testing.
1.3 ACCOMMODATE USER CHARACTERISTICS AND CAPABILITIES 1.3.1 DO NOT EXPECT USERS TO BECOME MASTERS While well-designed devices should have high learnability, do not overestimate the ability of users to master a device’s functions. In reality, most users master just the critical (from their perspective) portions of a device’s functions even if the device is only modestly complex (see Table 1.1). In other words, being practical-minded and often pressed for time, users master just the functions they use frequently. They tend to disregard other device functions until they are forced to deal with them, expecting at that time to draw on their intuition and peer support to operate the device correctly. Accordingly, designers should make infrequently used tasks maximally intuitive (especially if the task is life-critical) because most users will approach them in the same manner as novices.
1.3.2 EXPECT USE ERRORS Many factors can contribute to device use error. Therefore, while maintaining a respectful view of device users, designers should assume that users will err frequently (Beydon et al., 2001; Samore et al., 2004). They should not assume that all users will operate a device with equivalent levels of preparation, attitude, vigilance, and motivation. Rather, a disconcerting proportion of users may have insufficient training, have forgotten their training since the last time they used the device, be fatigued from working long hours (Gaba and Howard, 2002; Weinger and Ancoli-Israel, 2002), have insufficient technical aptitude to interact with complex devices, or be rushed, overworked, or distracted. If anything, designers should overestimate rather than underestimate the chances of a use error. Thus, designers should make the errors obvious, provide the means for rapid recovery, and guide users through the recovery process. TABLE 1.1 Comparison of the Varying Levels of Mastery of Infusion Pump Setup and Operation Tasks Level of Mastery at Performing Specific Tasks Sample Users Nurse X Physician Y Biomedical engineer Z
TAF-ER562X_WINGR-09-0401-C001.indd 6
Determine the Total Volume of IV Fluid
Set up a “Piggyback” Infusion
Change the Battery
High Medium Medium
Medium Low Low
Low Low High
11/8/10 11:39:54 PM
General Principles
7
1.3.3 ACCOMMODATE DIVERSE USERS It is perilous to assume homogeneity in the user population. It is even more perilous for designers to assume that users are just like them. This is why many human factors researchers choose to conduct fieldwork leading to the formulation of personas (also called user profiles) of typical users to guide a design effort. Depending on the device, users may indeed be a comparatively small and specialized population, such as highly trained interventional cardiologists who operate catheterization laboratory equipment. However, over-the-counter devices, such as glucose meters, blood pressure monitors, metered dose inhalers, and infant apnea monitors, will be used by diverse individuals, including the young and old and people with disabilities. In either case, designers should accommodate the needs of users who have different sizes, shapes, physical abilities, intellectual capabilities, experiences, and so on (see Chapter 4, “Anthropometry and Biomechanics,” and Chapter 2, “Basic Human Abilities”). A simple example of accommodating user diversity is designing a surgical tool that can be used comfortably by individuals with either small or large hands (see Chapter 16, “Hand Tool Design”). Other examples include designing a computerized patient data entry screen for people who have extensive computer experience as well as those with relatively little experience or a mammography machine that is accessible to both ambulatory individuals and those in wheelchairs.
1.3.4 MAXIMIZE ACCESSIBILITY The term accessibility has traditionally been associated with architectural features, such as sidewalks, building entrances, and restrooms. Features such as curb cuts, automatic doors, and large restroom stalls equipped with assist bars are products of regulations and political activism that have improved accessibility to public spaces. In recent years, consumer electronic and information technology products have incorporated features to make them more accessible to users with physical or sensory impairments. For example, every federal Web site now describes all figures in the text to accommodate people with visual impairments who use screen readers. Similar improvements can be made to medical devices, making them more usable by people with disabilities. As cited above, imaging devices can be made more accessible to people in wheelchairs and those with limited range of motion or sensory-motor control by addressing their needs during device design (Figure 1.3). The
FIGURE 1.3 Mammography machines can accommodate a seated patient. (From http://qap.sdsu. edu/education/breastcancerreview/Bc_diag/Bccore/photo4d.gif. With permission.)
TAF-ER562X_WINGR-09-0401-C001.indd 7
11/8/10 11:39:54 PM
8
Handbook of Human Factors in Medical Device Design
FIGURE 1.4 First responders might have difficulty reading a defibrillator display if space constraints on a commuter train require them to place the unit at an acute viewing angle. (From http:// www.sanarena.ch/icons/fotos/Nothilfetag2004/reanimation1.jpg. With permission.)
same may be said of many other types of medical devices, including examination tables and all sorts of diagnostic and therapeutic devices. An example of the latter is a glucose meter that speaks user instructions and meter readings to facilitate use by individuals with visual impairments.
1.3.5 CONSIDER EXTERNAL FACTORS THAT INFLUENCE TASK PERFORMANCE Sometimes, people use medical devices in a relatively isolated manner, such as reprogramming an insulin pump while seated at a desk in a quiet room. However, medical devices are often used in a more dynamic and potentially distracting setting (see Chapter 3, “Environment of Use”). It might be quite noisy. It may be particularly hot (e.g., using devices in the course of a rescue conducted outdoors on a 100°F day). Other people or devices may be vying for the user’s attention. Users may be wearing protective gear (e.g., glasses and gloves) to prevent injury or contamination. When considering the many potential external factors, a given design might be found to be incompatible with some uses, such as a paramedic who is wearing thick gloves and trying to press the right button or to read a display at an acute angle (see Figure 1.4). Analyzing the resulting design trade-offs (e.g., different control element choices might be better under different use conditions) is a core part of effective human factors engineering.
1.4 ACCOMMODATE USERS’ NEEDS AND PREFERENCES 1.4.1 ACCOMMODATE USER PREFERENCES UP TO A POINT Medical device consumers—particularly large hospitals that purchase large lots— often ask manufacturers to customize devices according to their institutions’ needs. Such requests often motivate designers to add configuration options into devices. This
TAF-ER562X_WINGR-09-0401-C001.indd 8
11/8/10 11:39:55 PM
General Principles
9
supports the appropriate goal of making devices adapt to the users rather than the other way around. However, such adaptability brings with it the risk of user interfaces becoming less stable or less predictable or even compromising performance. Someone accustomed to one device’s setup may be confused when encountering what appears to be the same type of device but with a completely different setup. For example, a nurse who works shifts at several different hospitals each week may encounter infusion pumps that look the same but work quite differently, setting the stage for a use error due to negative transference. One solution is to limit user-interface variability where possible by developing optimal solutions for all users and then make the remaining setup differences obvious. Consistency across device models regardless of manufacturer (e.g., all parenteral infusion pumps would have the same controls just like all cars have a steering wheel) could substantially reduce use errors. This approach may be resisted, however, because it asks manufacturers to sacrifice brand identity and potential competitive advantages. Also, manufacturer compliance with de facto industry standards may be compromised by existing patents and licensing agreements. An alternative is to design a device that can be readily set to a particular institution’s or individual’s preferences, thereby accommodating market niches while ensuring that such preferences do not impede typical users. Designers should keep in mind that users are not designers. Thus, although users might express specific needs and preferences or even suggest a detailed design, this input may prove unreliable, undesirable, or unworkable. Accordingly, designers should view themselves as interpreters, taking and prioritizing user input while also applying their own expertise and creativity to produce the best possible designs. Selective user input can be particularly dangerous when care is not taken to rigorously study and evaluate potential design alternatives.
1.4.2 DO NOT RELY EXCLUSIVELY ON “THOUGHT LEADERS” Manufacturers are prone to rely on guidance from accomplished and interested clinicians, individuals also referred to as “thought leaders.” Manufacturers also tend to rely more on clinicians who represent large accounts, the goal being to give extra emphasis to the particular institution’s needs in order to keep their business. Indeed, such individuals can be an excellent source of design guidance, particularly with regard to identifying use needs and preferences. However, their relatively sophisticated viewpoint and capabilities as well as their expanded knowledge of specific design issues and trade-offs lead them to develop biases. Also, such individuals might push a particular design solution harder than is appropriate because of their emotional investment in it or even for the sake of ego gratification. Accordingly, design teams should seek users who are more representative of the typical user to define user needs and preferences as well as to get reactions to designs in progress.
1.4.3 ENABLE USERS TO SET THE PACE Human beings become annoyed when machines set the work pace. Often the pace will be too slow or too fast because of individual performance differences. Moreover, machinepaced tasks do not readily accommodate work stoppages due to interruptions, including
TAF-ER562X_WINGR-09-0401-C001.indd 9
11/8/10 11:39:55 PM
10
Handbook of Human Factors in Medical Device Design
FIGURE 1.5 These software user interfaces give users control over the pace of work by presenting controls to continue and start tasks. (From http://www.visionchips.com/images/s3.gif and http:// alaris.choc.org/graphics/display_wsk_gdl_lido_new_1b.jpg. With permission.)
emergency situations. Thus, designers should let device users set the work pace by, for example, requiring them to provide input before proceeding to the next step in a procedure (Figure 1.5).
1.5 ESTABLISH REALISTIC EXPECTATIONS OF USERS 1.5.1 DO NOT RELY ON TRAINING Caregivers do not always receive proper training before using a given device. As discussed earlier, clinicians’ work demands leave little time for training and for reading instruction manuals. Also, a new employee or a traveling nurse who is filling a temporary gap in staffing may not have received training before he or she uses a device, institutional policies prohibiting such untrained use notwithstanding. Even when users receive proper training, they may forget what they learned by the time they use the device, especially if the device is used infrequently (e.g., a few times per year). Or caregivers who are familiar with a device might simply forget or be confused about how to perform an uncommon task. Accordingly, medical devices should be designed for intuitive operation even when users are expected to be highly trained. Furthermore, if designers anticipate that a medical device may migrate from clinical use to off-label use, for example, from hospitals to use in people’s homes, it should be designed for intuitive use by laypersons even if it is not initially intended for use by such individuals.
1.5.2 DO NOT RELY ON INSTRUCTIONS FOR USE While some manufacturers produce excellent instructions for use, users might still disregard them in favor of a hands-on demonstration (i.e., an in-service) from a manufacturer’s representative, staff educator, or knowledgeable peer. The instruction manual will often be difficult or impossible for users to access while also using the given device (Figure 1.6). Therefore, designers should not count on users reviewing and absorbing information found only in the instructions for use. While the instructions may describe a device’s theory of operation in a helpful manner or may even be essential to understanding the device’s performance, it is unlikely that many users will read them. In most cases, the caregiver’s workday is too hectic to spend the time to thoroughly read or even skim device instructions. Users may even neglect or overlook instructions for use placed on the device, perhaps in the form of a label or online help system. Many users simply ignore instructions in favor of
TAF-ER562X_WINGR-09-0401-C001.indd 10
11/8/10 11:39:55 PM
General Principles
FIGURE 1.6 device.
11
User manuals are often stored in a central location rather than with the medical
other learning methods. Therefore, designers should account for the reduced opportunity to inform device users through documentation by making devices as intuitive to operate as possible.
1.5.3 DO NOT RELY ON WARNINGS The presence of many warnings and on-device instructions often indicates user-interface shortcomings (see also Chapter 10, “Alarm Design,” and Chapter 13, “Signs, Symbols, and Markings”). The best way to address a hazard is to eliminate it or design a user interface to guard against it. As such, warnings should be regarded as the last but nonetheless useful step toward preventing a problem, especially problems that can lead to property damage or personal injury. Unfortunately, devices become “papered” with warnings when development teams fail to implement the design changes necessary to correct fundamental design problems. This solution is problematic not only because it leaves fundamental flaws intact but also because the presence of multiple warnings may lead to “warning fatigue,” causing users to be less attentive to any warnings on the device.
1.5.4 DO NOT RELY ON MEMORY People can be forgetful and are often easily distracted. Also, daily life can place excessive burdens on short-term memory. Therefore, designers should not rely on users to remember information, such as a test result or numerical code, to perform a task. Moreover, operational sequences should not require users to remember next steps. It is far better to present users with the crucial information they need to perform a task correctly. It is also helpful to bring required tasks to the user’s attention.
1.5.5 AVOID INFORMATION OVERLOAD Medical devices often flood caregivers with more information than they could possibly use to accomplish associated tasks. The result can be information overload, a condition in which the caregiver cannot receive and process the information fast enough for it to be useful (see Chapter 2, “Basic Human Abilities”). One solution is to provide users with
TAF-ER562X_WINGR-09-0401-C001.indd 11
11/8/10 11:39:55 PM
12
Handbook of Human Factors in Medical Device Design
information of primary interest at the moment it is needed while allowing them to easily obtain secondary information at a later time. For example, an at-home dialysis patient may usually want to know that “everything is okay” and determine how long before his or her treatment is complete but occasionally opt to check his or her blood pressure, the amount of fluid “taken off,” and other dialysis parameters. Another solution is to preprocess information, relieving the caregiver of the task. For example, a device might graph the relationship between two parameters and indicate minimum and maximum values, thereby performing the work that users otherwise must do in their heads. The key is to simplify information while still ensuring that important contextual information and subtle nuances are retained. Another example is emphasizing values in a list that exceed set limits rather than making users recall the limits and search the entire list for excursions from those limits.
1.5.6 DO NOT ASSIGN USERS TASKS THAT ARE BETTER SUITED TO OTHER USERS OR DEVICES Assign users and machines the tasks they are best qualified to perform and make sure that the distribution ensures safety and user satisfaction. One effective way to reduce workload is to let devices perform the functions they do best rather than give people the extra work (see Chapter 2, “Basic Human Abilities”). For example, a device is usually better at monitoring a steady-state process for unusual events, a task that is eventually fatiguing for most people and leads to reduced vigilance. Computer-based devices easily retain information that users may forget. A robotic device can hold an instrument steadily in a precise position for long periods of time. On the other hand, some functions (e.g., complex problem solving or pattern recognition) might be better performed by people and should not be quickly delegated to technology. Designers should automate tasks skillfully to avoid unintended consequences on overall user performance. Even if devices might better perform some tasks, designers should be cautious about shifting so many functions to the device that the users lose their awareness of the current situation as well as their ability to respond to emergencies. People generally prefer to be actors rather than observers in a process, except when the required actions are tedious or fatiguing, divert their attention from more important tasks, or are clearly performed better by machines.
1.6 CONSIDER REAL-WORLD DEMANDS 1.6.1 CONSIDER THE CONTEXT OF USE There is a tendency to create designs that work well for trained users who focus their full attention on operating the device in a quiet environment, such as the previously cited example of reprogramming an insulin pump at a desk in a quiet room. But medical devices are commonly used by diverse users in several different use environments. In addition, medical devices are sometimes used by distracted, fatigued, and/or marginally trained individuals. Clinicians often work in chaotic environments and need to focus more attention on their patients than their equipment. These real-world conditions can have a significant effect on a user’s interactions with a medical device, masking audible signals and interfering with concentration, for example. Designers need to consider realistic use conditions. The first step is to learn about users through field research conducted in a manner that does not appreciably influence the way people are working. Researchers often discover that medical
TAF-ER562X_WINGR-09-0401-C001.indd 12
11/8/10 11:39:56 PM
General Principles
13
devices are in the periphery of the users’ minds, particularly devices that function in a relatively autonomous manner, requiring only occasional checking. Later during device design, the impact of a new device on the ability of users to deliver quality care as well as its ability to work well with other devices should be tested. Well-designed medical devices tend to fit naturally into the home or workplace, garnering favorable reviews from users. Flawed devices tend to draw end users’ dissatisfaction, generating complaints about how they interfere with the normal workflow, and require too much attention.
1.6.2 CONSIDER WORST-CASE SCENARIOS In the normal course of device development, mechanical engineers purposely drop devices to the floor or shake them for hours in a test chamber to see if and how they break. Userinterface designers need to perform equivalent tests of their designs. Thus, user interfaces must be subjected to worst-case scenarios to see if and how they fail. The goal is to see how well an untrained or minimally trained user performs when asked to operate an unfamiliar device; to observe people using a device under harsh environmental conditions, such as at night in a moving ambulance (Figure 1.7) or helicopter; and to see what happens when people are under time pressure to perform a critical and difficult task. By stressing the user interface, designers can learn how to make the device more effective in the real world. Note that minimizing the risk of faulty design or use errors causing human injury requires an integrated approach that includes rigorous design practices, usability testing, and risk analysis.
1.6.3 MAKE DEVICES AS RUGGED AS NECESSARY Some medical devices take a beating during their useful life, which may span 10 to 20 years of daily use, and, just like mechanical components, user-interface components should
FIGURE 1.7 Paramedics use myriad medical devices in a moving vehicle that may complicate device-related tasks, such as reading a display and operating controls. (Shutterstock image 31634293.)
TAF-ER562X_WINGR-09-0401-C001.indd 13
11/8/10 11:39:56 PM
14
Handbook of Human Factors in Medical Device Design
FIGURE 1.8 Patient transport device has rugged design to be specifically used on stairs over many years. (From http://www.ems.stryker.com/detail.jsp?id=6 and Strykker 2005 Catalog, ems_2005_ catalog.pdf.)
be designed to last this long (see Figure 1.8). Heavily used devices that are subject to jostling and impacts require heavier-duty user-interface components. This means selecting switches that are unlikely to break even if struck with considerable force, screens that will not crack if elbowed or bumped, and labels that will not become unreadable after years of scrubbing with antiseptic solutions. The designer will need to balance the need for ruggedness against other design goals, such as ease of switch actuation and avoiding display parallax.
1.6.4 LIMIT USER WORKLOAD Caregivers are often overworked, enduring vigorous 12-hour or longer shifts, sometimes several days in a row. This explains why some caregivers actively or passively reject medical devices that create too much mental or physical work, as discussed earlier. Clinicians want to focus on their patients, not on distracting technology. Caregivers will seek shortcuts and work-arounds if a device distracts them from more important tasks even if the work-saving strategies are strongly discouraged by their institution. Patients who use medical devices are similarly disinterested in investing undue time or effort into device use given the many other demands and interests in their lives. Therefore, designers should pursue opportunities to reduce the time required to learn how to use and operate devices. For example, a high-quality video complemented by a quick reference card might shorten the time required to learn to operate a ventilator. As another example, a point-of-care blood gas analyzer might allow a clinician to commence with data entry tasks while the device completes a calibration check or prepares to analyze a sample.
TAF-ER562X_WINGR-09-0401-C001.indd 14
11/8/10 11:39:57 PM
General Principles
15
1.6.5 CONSIDER THE POTENTIAL FOR DEVICE MIGRATION INTO ADDITIONAL USES OR USE ENVIRONMENTS It is common for medical devices to migrate from sophisticated medical settings, such as an intensive care unit, to less sophisticated settings, such as inpatient wards, outpatient clinics, and patients’ homes (see Chapter 18, “Home Health Care”). Or a device intended for use on pediatric patients might ultimately be used on adults, such as small women. In some cases, the device manufacturer might not have anticipated such migration or considered the needs of the new user population in the original design efforts. This can lead to problems for both the manufacturer and new users. Consider the case of the infusion pump designed for use in the hospital but used by the parent of a sick child at home, a so-called off-label use. Lacking sufficient medical knowledge or training on how to use the device properly, the parent might experience a use error, leading to a tragic loss of life. In turn, this might lead to a costly lawsuit against the manufacturer. Or consider the case in which a pulse oximeter is used as a respiratory monitor in patients receiving narcotic pain medications at home, but in the presence of supplemental oxygen, the monitor does not alarm until the patient has stopped breathing, leading to an adverse outcome. Such scenarios underscore the value of anticipating alternative uses of medical devices. After all, manufacturers are probably better off protecting users against hazards associated with predictable, unintended device uses rather than reacting to a product liability or personal injury claim.
1.7 DEVELOP COMPATIBLE DESIGNS 1.7.1 ACCOMMODATE MENTAL MODELS People frequently have an established mental model or “big picture” in mind when they use a new device. Usually, their mental model is based on previous experience using similar devices. For example, users may expect certain controls to function in a particular manner and be surprised by a control that functions differently. Or they may be accustomed to configuring the device the way they were taught in nursing school, only to find that a new device requires an alternative approach. Such incompatibilities can make a device harder to learn to use and can induce errors even among experienced users who may unconsciously fall into an old use pattern (sometimes called negative transfer of training). Therefore, designers need to take care to identify established mental models and accommodate them where possible. When the need for change exists, designers are often better off making major rather than minor changes so that the difference is more readily apparent. Also, designers should provide “affordances” that help users form an accurate mental model of how a device works. Affordances might include organizing the user interface according to a simple “metaphor,” effective use of labels, clear and redundant feedback in response to user inputs, helpful warning messages, and a quick reference card that emphasizes how a device is different from other similar devices.
1.7.2 ESTABLISH NATURAL OR CONVENTIONAL MAPPINGS When people associate an action with a design element, such as turning a knob clockwise, they are mapping. An example of natural mapping is squeezing the bag on an anesthesia machine to fill a patient’s lungs with air; the action and outcome are largely self-evident
TAF-ER562X_WINGR-09-0401-C001.indd 15
11/8/10 11:39:57 PM
16
Handbook of Human Factors in Medical Device Design
FIGURE 1.9 The axial and left–right alignment of this power injector’s piston controls (top of photo) maps naturally to the movement of the pistons (bottom of photo).
(see Figures 1.9 and 1.10 for other examples of mapping). Turning a knob on an anesthesia machine clockwise to increase the rate of gas flow is an example of a conventional mapping; while it may not be as self-evident to a naive individual, experienced anesthesia providers will perform the task in an automatic, subconscious manner. When mappings are natural or conventional, users will find the associated devices more intuitive to operate. Incorrect or
Head Neck Chest Abdomen
Arm
Hand Thigh Calf Misc
Foot
FIGURE 1.10 This power injector’s software screen (unrelated to the power injector shown in Figure 1.9) provides users with an intuitive means of configuring the device for scanning specific body parts by mapping control elements to corresponding body parts. (From EZEM [design] and Wiklund, 1995 [photo].)
TAF-ER562X_WINGR-09-0401-C001.indd 16
11/8/10 11:39:57 PM
General Principles
17
unconventional mappings may cause users to take longer to learn to use a device or increase use errors. The challenge for designers is to establish effective mappings when the natural or conventional one that has emerged over time departs from a given company’s established practice. Or the designer might learn that conventional mappings differ among user populations (e.g., France, Japan, and the United States). In such cases, it might be appropriate to customize devices to the given market or develop optimal mapping through a rigorous process of design iteration and usability testing. More information on how to design interface elements to be more intuitive is provided in other sections of this handbook (e.g., Chapter 7, “Controls,” Chapter 8, “Visual Displays,” and Chapter 12, “Workstations”).
1.7.3 FOLLOW INDUSTRY CONVENTIONS AND CONSENSUS STANDARDS Manufacturers often seek ways to make their devices stand out from competitors’ devices, thereby fortifying their brand identity and competitiveness. On the other hand, device users value consistency, particularly with regard to operational characteristics. When a new device works like similar devices, including consumer products such as cellular phones and word processing software, users have less to learn about the new device and can apply past experience more readily. Human factors specialists call this positive transfer. Therefore, designers should not diverge substantially from conventional design practice or industry standards unless there is a compelling reason to do so. There can be very good reasons to deviate, such as a demonstrable increase in design intuitiveness, task efficiency, or error prevention. In fact, to foster innovation and continual design improvements, designers are encouraged to challenge de facto conventions and incorporate new designs if there is good evidence that the alternative will lead to better user performance. On the other hand, making a device seem different just for the sake of being different is a poor practice. For example, designers should not diverge from the standard color codes for alarm signals simply because they think that a magenta-colored alarm indicator is more conspicuous than a red one or the company’s branding scheme dictates use of the nonstandard color. American users have learned that high-priority visual alarms are red, for example, and expect all devices to adopt this convention.
1.8 OPTIMIZE USER INTERACTIONS 1.8.1 MAKE DEVICES ERROR TOLERANT As stated previously, devices fail, and users make errors. Consistent with modern principles of resilience engineering, device designs should be tolerant of error to minimize harm to users or patients. Design approaches to accomplish error tolerance include failing safe (see below), considering the device in the overall context of use, providing more information about the implications of specific use actions, making errors or unwanted deviations more visible to users, making potential risks more visible to users, and facilitating error recovery.
1.8.2 FAIL IN A SAFE MANNER A basic engineering design principle is to fail safely. Consider the household example of an iron that shuts itself off if the user fails to do so, possibly preventing a fire. For medical
TAF-ER562X_WINGR-09-0401-C001.indd 17
11/8/10 11:39:58 PM
18
Handbook of Human Factors in Medical Device Design
devices, failing in a safe manner is important because patient lives are at stake. The concept of failing safely can be extended from electromechanical failures to use errors. For example, an infusion device might be designed to prevent users from setting an unsafe level of drug delivery. A laser treatment device should not be able to be activated if the emergency stop control is inoperative.
1.8.3 AVOID PHYSICAL STRAIN, REPETITIVE MOTIONS, AND CUMULATIVE TRAUMA The repetitious nature of many medical procedures, such as firmly squeezing and releasing a surgical stapler, puts caregivers at risk for repetitive motion or cumulative trauma disorders (see Chapter 16, “Hand Tool Design”). Therefore, designers should seek opportunities to reduce the number of repetitive actions required to operate a given device (see Chapter 4, “Anthropometry and Biomechanics”). They should also keep manually applied forces to a minimum, eliminate pressure points between devices and users, and enable users to maintain neutral joint positions. Moreover, they should limit the amount of time users are required to apply a constant force (e.g., continuously squeezing the handles on a grasping tool) even if the force is relatively small. Often, a relatively simple design change will achieve these goals, protecting users from injury.
1.8.4 ENABLE USERS TO ANTICIPATE FUTURE EVENTS To provide the best patient care, caregivers sometimes need to predict the most likely future course of disease manifestations and therapeutic interventions. In other words, caregivers try to figure out what is about to happen rather than simply reacting to what has happened or is happening at the moment. This is especially true in situations in which a caregiver is delivering a therapy that can have a dramatic effect on the patient’s physiological state, such as the intravenous delivery of a blood pressure medication. Where possible, designs should enable caregivers to “see ahead.” For example, a monitoring device might present historical parametric values as well as values forecasted for the next 5, 10, and 30 minutes.
1.8.5 CONFIRM IMPORTANT ACTIONS In the medical arena, confirmation messages may serve an important or even critical purpose, considering that some actions are irreversible and could lead to patient injury (Figure 1.11).
FIGURE 1.11
Sample confirmation message.
TAF-ER562X_WINGR-09-0401-C001.indd 18
11/8/10 11:39:58 PM
General Principles
19
Therefore, even though some users may regard confirmation messages as a hindrance—a wasted extra step—the benefit of such messages often outweighs the annoyance they cause. However, the need for users to perform tasks efficiently should not be underestimated when weighing decisions about which actions require confirmation. Thus, the benefits of confirmation messages should be demonstrated through user testing, ensuring that they do not replace one problem with another, such as users confirming their actions without thinking about it (i.e., performing tasks in a rote manner).
1.8.6 MAKE CRITICAL CONTROLS ROBUST AND GUARD THEM Certain medical devices may be exposed to rough handling, particularly when they are used outdoors or during emergency responses (e.g., patient resuscitation procedures or “codes”). Medical devices are subject to being dropped or bumped. Therefore, the user interface needs to be designed to prevent accidental actuation of critical controls. Incidental contact with a device’s front panel should not, for example, deactivate the device or alter a critical control setting. This is why many medical devices have physically guarded power buttons (Figure 1.12) and require the user to confirm critical adjustments (see Chapter 7, “Controls”). If a critical control fails, an alternative means of control should be provided. For example, in the case of a damaged pump stop switch, a mechanical means of ceasing pump action should be provided (and be readily identifiable by the user).
1.8.7 CLARIFY OPERATIONAL MODES One way that designers seek to simplify medical devices is to incorporate multiple operational “modes.” In principle, this approach is a sensible way to facilitate context-specific tasks and to limit users’ exposure to non-relevant device features. However, problems can arise when the user enters the wrong mode and does not realize it (often called mode error). It would be problematic, for example, if an anesthesiologist were monitoring an adult patient
FIGURE 1.12 (See color insert following page 564.) An emergency stop button on a scanner is recessed to prevent inadvertent actuation. Large size, red color, symbolic label, shape, and recessing also provide redundant means of differentiating this control from others.
TAF-ER562X_WINGR-09-0401-C001.indd 19
11/8/10 11:39:58 PM
20
Handbook of Human Factors in Medical Device Design
using a monitor placed in “demonstration” mode (in which the numbers displayed are normal and unchanging) or had the anesthesia workstation’s software user interface set for a pediatric case. Accordingly, designers should make operational modes and their characteristics readily apparent.
1.8.8 EMPLOY REDUNDANT CODING Redundant coding of displays and controls can be a powerful way to ensure reliable device operation (see Chapter 7, “Controls,” and Chapter 8, “Visual Displays”). The concern is that a user who may be fatigued or distracted might actuate the wrong control or mistake one display value for another (e.g., a “1” for a “7”). These kinds of use errors are less likely if displays and controls employ more than one means of coding. Coding options include varying the user-interface element’s size, shape, color, texture, or placement. For example, anesthesia machines use redundant coding (knob color, shape, and texture) to ensure that caregivers turn the correct knob to increase the flow of 100% oxygen rather than air (see Chapter 12, “Workstations”).
1.8.9 DESIGN TO PREVENT USER CONFUSION While considering the need for making devices compatible, as discussed elsewhere, also consider when it is appropriate to make devices distinct. For example, it is critical to distinguish power cable receptacles from sensor cable receptacles, thereby avoiding circumstances in which a user might plug a patient sensor lead into an AC outlet and shock the patient (Figure 1.13). Devices and elements thereof may be distinguished using the coding methods described above. In the case of plugs and receptacles, size and shape coding is particularly appropriate, making it impossible to fit a given lead into the wrong port. This principle of using a physical constraint to dissuade undesirable user actions applies to controls, connectors, and other design elements.
FIGURE 1.13 Monitor’s sensor leads are incompatible with AC power receptacles. (From Wiklund, M., Medical Device and Equipment Design: Usability Engineering and Ergonomics, Interpharm Press, Boca Raton, FL, 1995.)
TAF-ER562X_WINGR-09-0401-C001.indd 20
11/8/10 11:39:58 PM
General Principles
21
1.8.10 DON’T NEGLECT DEVICE APPEAL It is important to recognize that human factors in medical device design are not just about achieving safe and effective task performance. They are also about satisfying user needs, which includes making medical devices pleasing to use. Devices that are easy to use as well as appealing to view and touch will engender greater user satisfaction. One payoff from making devices appealing is that patients—and particularly children—may find them less frightening. Moreover, users may be more motivated to learn how to use appealing devices properly. Added appeal may also lead to increased user vigilance and job satisfaction. For example, a user may pay closer attention to a display with a pleasing appearance that also draws attention to important information as opposed to one that has a garish appearance that draws attention to less important information (see Chapter 11, “Software User Interfaces”). Devices designed to provide an impression of quality can inspire greater user confidence (and even pride of ownership). A user may be drawn to a portable patient monitor because of design qualities (e.g., an enclosure that looks attractive and easy to handle) that extend beyond functionality to boost appeal. The same can be said of tools that look comfortable to hold (Figure 1.14). However, medical devices intended for use in the home should not look like toys; otherwise, they might attract children to play with them.
FIGURE 1.14 Close attention to visual and tactile design considerations, such as rounded surfaces, distinguishable controls, and other styling cues, contribute to a device’s usability and appeal and earned these sinus cavity surgery and eye examination devices Medical Design Excellence Awards. (From http://www.gyrus-ent.com/health/diego/rhinologyDiego.htm and http://www.welchallyn. com/medical/products/catalog/detail.asp?ID=29365. With permission.)
TAF-ER562X_WINGR-09-0401-C001.indd 21
11/8/10 11:39:59 PM
22
Handbook of Human Factors in Medical Device Design
1.9 SUMMARY The general considerations (i.e., principles) discussed above are just a fraction of the broadlevel factors, albeit some of the more important ones, to consider when designing a medical device’s user interface. Accordingly, readers should regard these as a starting point and supplement them with additional considerations presented in the following sections of this handbook as well as other reference documents.
RESOURCES Jacko, J. A. and Sears, A. (2003). The Human-Computer Interaction Handbook: Fundamentals, Evolving Technologies and Emerging Applications. Mahwah, NJ: Lawrence Erlbaum Associates. National Aeronautics and Space Administration. (1989). Man-Systems Integration Standards. NASA-STD-3000a. Houston: Lyndon B. Johnson Space Center. Nielson, J. (1993). Usability Engineering. Boston: Academic Press. Norman, D. (1988). The Design of Everyday Things. New York: Basic Books. Rouse, W. B. (1991). Design for Success: A Human-Centered Approach to Designing Successful Products and Systems. New York: Wiley-Interscience. Salvendy, G. (Ed.). (2006). Handbook of Human Factors and Ergonomics (3rd ed.). New York: John Wiley & Sons. Sanders, M. S. and McCormick, E. J. (1993). Human Factors in Engineering and Design (8th ed.). New York: McGraw-Hill. Sawyer, D. (1996). Do it by Design: An Introduction to Human Factors in Medical Devices. Washington, DC: U.S. Department of Health and Human Services, Food and Drug Administration. U.S. Department of Defense. (1996). Human Engineering Design Criteria for Military Systems, Equipment, and Facilities. MIL-STD-1472F. Washington, DC: U.S. Department of Defense. Wickens, C. and Holland, J. (2000). Engineering Psychology and Human Performance (3rd ed.). New York: Prentice Hall. Wiklund, M. (Ed.). (1995). Medical Device and Equipment Design: Usability Engineering and Ergonomics. Boca Raton, FL: Interpharm Press. Wiklund, M. and Wilcox, S. (2005). Designing Usability into Medical Devices. Boca Raton, FL: Interpharm Press. Woodson, W. E., Tilman, B., and Tilman, P. (1992). Human Factors Design Handbook: Information and Guidelines for the Design of Systems, Facilities, Equipment, and Products for Human Use (2nd ed.). New York: McGraw-Hill.
REFERENCES American National Standards Institute/Association for the Advancement of Medical Instrumentation (ANSI/AAMI). (2001). Human Factors Design Process for Medical Devices. ANSI/AAMI HE-74-2001. Arlington, VA: Association for the Advancement of Medical Instrumentation. Beydon, L., Conreux, F., Le Gall, R., Safran, D., Cazalaa, J. B., et al. (2001). Analysis of the French health ministry’s national register of incidents involving medical devices in anaesthesia and intensive care. British Journal of Anaesthesia 86:382–87. Gaba, D. M. and Howard, S. K. (2002). Patient safety: Fatigue among clinicians and the safety of patients. New England Journal of Medicine 347:1249–55. Samore, M. H., Evans, R. S., Lassen, A., Gould, P., Lloyd J., Gardner, R. M., et al. (2004). Surveillance of medical device-related hazards and adverse events in hospitalized patients. Journal of the American Medical Association 291:325–34. Weinger, M. B. and Ancoli-Israel, S. (2002). Sleep deprivation and clinical performance. Journal of the American Medical Association 287:955–57.
TAF-ER562X_WINGR-09-0401-C001.indd 22
11/8/10 11:39:59 PM
2 Basic Human Abilities Edmond W. Israelski, PhD CONTENTS 2.1 Sensory and Perceptual Abilities ...............................................................................25 2.1.1 Vision ..............................................................................................................25 2.1.1.1 Threshold of Seeing ........................................................................25 2.1.1.2 Visual Acuity ..................................................................................26 2.1.1.3 Visual Angle....................................................................................27 2.1.1.4 Minimum Visual Angle ..................................................................27 2.1.1.5 Dynamic versus Static Acuity .........................................................29 2.1.1.6 Accommodation (Focusing Abilities) .............................................29 2.1.1.7 Visual Field .....................................................................................29 2.1.1.8 Color Vision ....................................................................................29 2.1.1.9 Color Vision Deficiencies ................................................................32 2.1.1.10 Color Discrimination Recommendations ........................................33 2.1.1.11 Recommendations for Printed Colors .............................................34 2.1.1.12 Recommendations for Colored Lights.............................................34 2.1.1.13 Recommendations for Color Combinations (Legibility and Visibility) ..................................................................................35 2.1.1.14 Dark Adaptation ..............................................................................36 2.1.2 Visual Perception ............................................................................................37 2.1.2.1 Distance and Perceived Size ...........................................................37 2.1.2.2 True Object Size ..............................................................................38 2.1.2.3 Common Visual Illusions ................................................................38 2.1.2.4 Perception of Motion .......................................................................39 2.1.2.5 Flickering Lights .............................................................................39 2.1.2.6 Photosensitive Epilepsy ...................................................................39 2.1.3 Auditory Perception ........................................................................................40 2.1.3.1 Loudness Measurements .................................................................40 2.1.3.2 Relationship of Phones to Sones .....................................................41 2.1.3.3 Loudness (Sones) Calculation for Complex Sounds ........................42 2.1.3.4 Pitch Measurement ..........................................................................42 2.1.3.5 Differential Hearing Thresholds .....................................................42 2.1.3.6 Effects of Aging on Hearing Sensitivity .........................................43 2.1.4 Other Sensory Modalities ...............................................................................45 2.1.4.1 Skin (Somesthetic) Senses ...............................................................45 2.1.4.2 Muscle Sense ...................................................................................46 23
TAF-ER562X_WINGR-09-0401-C002.indd 23
11/8/10 11:40:31 PM
24
Handbook of Human Factors in Medical Device Design
2.1.4.3 Sense of Balance .............................................................................46 2.1.4.4 Chemical Senses .............................................................................46 2.1.5 Other Perceptual Abilities...............................................................................46 2.1.5.1 Estimation of Time Intervals...........................................................46 2.1.5.2 Estimation of Other Physical Quantities .........................................46 2.2 Human Information Processing .................................................................................48 2.2.1 Limitations on Information Processing Abilities ............................................48 2.2.1.1 Channel Capacity ............................................................................48 2.2.1.2 Attention..........................................................................................49 2.2.1.3 Vigilance (Sustained Attention) ......................................................49 2.2.2 Speed of Information Processing ....................................................................49 2.2.2.1 Reaction Time .................................................................................49 2.2.2.2 Speed versus Accuracy....................................................................51 2.2.2.3 Human Memory ..............................................................................51 2.2.2.4 Working Memory ............................................................................51 2.2.2.5 Long-Term Memory ........................................................................53 2.2.2.6 Estimation and Decision-Making Abilities .....................................53 2.3 Human Response Capabilities ...................................................................................54 2.3.1 Speed of Movement ........................................................................................54 2.3.2 Principles of Motion Economy .......................................................................56 2.3.3 Speech Attributes ............................................................................................57 2.3.3.1 Loudness Levels of Speech .............................................................57 2.3.3.2 Frequency Characteristics of Speech ..............................................58 2.4 Human versus Machine Capabilities .........................................................................59 Resources ...........................................................................................................................60 References ..........................................................................................................................60 This chapter presents a brief overview of basic human skills and abilities that will aid the designer of medical devices to better understand the guidance provided in subsequent chapters in this book. These basic human skills and abilities and their interrelationships are shown in Figure 2.1. The organization of this chapter follows the flow of how humans sense, perceive, process, and respond to the world around them, as shown in the figure that portrays the basic inputs and outputs of the human information processing system. Specifically, the following areas of basic human capabilities and corresponding limitations are covered: vision, visual perception, audition (or hearing), sensation, information processing, human response capabilities, and human versus machine trade-offs. Many specific design recommendations result from knowledge of basic human skills and abilities, and these recommendations are covered in specific chapters in this book (see Chapter 8, Stimuli
Sensory processing
Input
FIGURE 2.1
Response
Perception
Information processing
Response processing
Output
Organization of basic human skills and abilities.
TAF-ER562X_WINGR-09-0401-C002.indd 24
11/8/10 11:40:32 PM
Basic Human Abilities
25
“Visual Displays”; Chapter 10, “Alarm Design”; and Chapter 16, “Hand Tool Design”). Design guidance from this chapter is noted by numbered guidelines.
2.1 SENSORY AND PERCEPTUAL ABILITIES This section covers basic human sensory and perceptual abilities in the areas of vision, hearing, touch, balance, and perceptual estimation. Individual differences in perceptual abilities are quite large and are due not only to inherent physiological factors but also differences in experience, motivation, and preconceived ideas about incoming sensory information, sometimes called one’s “psychological set” or “expectations.” It should be noted that unless specifically mentioned, much of the following data are based on relatively young humans without major disabilities. Designers must be careful when applying data from this chapter if their intended users include special populations, such as the elderly or disabled. In that case, information provided in Chapter 18, “Home Health Care,” may be more useful.
2.1.1 VISION The human visual sensory system is quite complex. In this section, information is presented on visual thresholds for seeing, visual acuity (both static and dynamic), visual angles, accommodation or focusing, visual field, color vision, and dark adaptation. 2.1.1.1 Threshold of Seeing The sensitivity of the human visual system covers a wide range, as does the visual threshold (i.e., the minimum light level in which an object can be visually identified) under various ambient lighting conditions (Figure 2.2; Van Cott and Kinkade, 1972). Rod vision comes from visual receptors found on the back of the eye, the retina, that are most sensitive under 100,000 Upper limit of visual tolerance 10,000 1,000
Approximate luminance level (brightness) in ft. lamberts
100
Cone vision only
Average earth on a clear day Average earth on a cloudy day
10
White paper in good reading light
1
White paper 1 ft. from std. candle
0.1 0.01 0.001 0.0001 0.00001
0.000001
FIGURE 2.2
Fresh snow on a clear day
Snow in full moon
Rod and cone vision
Average earth in full moon
Snow in starlight Grass in starlight
Rod vision only
Absolute threshold of seeing
Threshold of seeing and luminance levels of various objects.
TAF-ER562X_WINGR-09-0401-C002.indd 25
11/8/10 11:40:32 PM
26
Handbook of Human Factors in Medical Device Design
low light conditions to shades of black and white (e.g., a patient room during sleeping hours). Cone vision comes from visual receptors that are sensitive to color and operate best under higher light levels (e.g., in an operating room). 2.1.1.2 Visual Acuity A number of measures of visual acuity exist: • • • • •
Minimum distinguishable (detection of detail in an arbitrary test target) Minimum perceptible (detection of a spot, e.g., on a radiograph) Minimum separable (detection of a gap between parts of a target) Stereoscopic acuity (detection of depth for a three-dimensional target) Vernier acuity (detection of lateral displacement of one line from another)
These measurements of acuity all apply to static or stationary objects. In addition, there is dynamic visual acuity of the smallest detail that can be detected for a moving target. Visual acuity can be affected by a variety of factors (Table 2.1). TABLE 2.1 Factors That Affect Visual Acuity Factor
Positive Example
Negative Example
Amount and kind of illumination
Bright operating room light
Viewing time
Momentary occlusion message on IV Long time period for viewing pulse pump screen oximeter readings Light-colored numbers showing pulse Yellow trace lines of respiratory rate rate on a dark-background patient on a white-background patient monitor monitor Large font indicating on/off for a Small print on a catheter package ventilator power switch label indicating French size
Object contrast with background Object size (visual angle subtended by the object at the eye) Object color Direction of viewing (position of the image on the retina)
Glare from outside sunlight on IV pump screen
Bright red flashing alarm on an Lettering in pale pastel colors enteral pump indicating it is empty indicating length of a nasogastric tube Patient monitor screen placed at a PCA pump placed on the bottom of 45-degree angle above a patient’s bed an IV pole at knee level
Movement of the object or viewer
Stationary IV pump screen
Heart monitor vibrating from the motion in an ambulance
Accommodation or focusing abilities of the viewer’s visual system
Large-screen monitor (over 17 inches) for an ultrasound machine
Fatigued operator of a small-screen portable patient monitor
Optical alignment of both eyes or convergence abilities
Surgeon being able to accurately judge the depth of cutting with a scalpel
Surgeon attempting to judge distance from a laparoscopic pincer by looking at a video monitor
Dark adaptation
Ambulance driver reading red gauges while driving at night
Trying to find a central line port in a darkened patient room after entering from a brightly lit hallway
TAF-ER562X_WINGR-09-0401-C002.indd 26
11/8/10 11:40:32 PM
Basic Human Abilities
27
7 mm Visual angle A
Size of the object So
tan A
Distance of object do or A (minutes of arc)
(57.3) 60 So do
if A
So do 10°
FIGURE 2.3 Calculation of visual angle A in minutes of arc when object size and distance are known.
2.1.1.3 Visual Angle The visual angle subtended on the retina at the back of the eye can be calculated using the formula and related diagram in Figure 2.3 (Cornsweet, 1970). In the diagram, the apex of the triangle is assumed to be 7 mm behind the foremost point of the cornea. Examples of the visual angle A cast by some common objects at a given distance do are shown in Table 2.2 (Cornsweet, 1970). 2.1.1.4 Minimum Visual Angle Minimum visual angle is the value of the visual angle cast on the retina for the following types of limiting conditions (Dreyfuss, 1966): • Minimum perceptible visual angle is approximately 1 second of a degree for a thin wire against bright sky. (Visible stars may subtend an angle as low as 0.056 second.) • Preferred angle for reading English text is 20 to 22 minutes of arc. Marginally acceptable angles range from 16 to 18 minutes of arc, with 12 minutes considered the threshold of readability. Guideline 2.1: Minimum Type Size Type size should not be less than 3 points when read at 14 inches under the most favorable lighting conditions (1 point = 1/72, or 0.01384 inch). See Table 2.3 for details.
TABLE 2.2 Visual Angle for Common Objects Object Sun Moon Quarter Quarter Quarter Lowercase pica-type letter
Distance (do)
Visual Angle (A)
93,000,000 miles 240,000 miles Arm’s length (70 cm) 90 yards 3 miles Reading distance (40 cm)
30 minutes 30 minutes 2 degrees 1 minute 1 second 13 minutes
1 degree = 60 minutes of arc; 1 minute = 60 seconds of arc; tan 1 second = 0.0000048; tan 1 minute = 0.00029. For small angles, the tangent of an angle varies linearly with the size of the angle (e.g., tan 10 minutes = 10 × tan 1 minute).
TAF-ER562X_WINGR-09-0401-C002.indd 27
11/8/10 11:40:32 PM
28
Handbook of Human Factors in Medical Device Design
TABLE 2.3 Recommended Character Sizes and Corresponding Font Sizes for Various Reading Distances Character Height vs. Reading Distance Character Heighta (in.)
Reading Distance (in.)
Visual Angle (minutes of arc)
Font Sizeb
Preferred (Upper Bound) 0.102
16
22
Actual font size 7.5
0.154
24
22
Actual font size 11
0.230 0.768 1.152
36 120 180
22 22 22
Font 55 Font 72
0.093 0.140
16 24
0.209 0.698 1.047
36 120 180
0.084 0.126
16 24
0.188 0.628 0.942
36 120 180
0.074 0.112
16 24
0.168 0.558 0.838
36 120 180
16 16 16
0.056 0.084 0.126 0.419
16 24 36 120
Minimum Threshold 12 12 12 12
0.628
180
12
a b
Actual size 16
Preferred (Lower Bound) Actual font size 7 20 20 Actual font size 10 20 20 20
Actual font size 15 Font 50 Font 75
Adequate (Upper Bound) 18 Actual font size 6 18 Actual font size 9 18 18 18
Actual font size 13.5 Font 45 Font 68
Adequate (Lower Bound) Actual font size 5.5 16 16 Actual font size 8
Actual font size 12 Font 40 Font 60
Actual font size 4
Actual font size 6
Actual font size 9 Font 30 Font 45
Smallest lowercase letter height. Font size is the distance from the highest ascender to the lowest descender of any character in the font set. Assumptions: Contrast ratio >7:1; luminance >35 cd/m2.
TAF-ER562X_WINGR-09-0401-C002.indd 28
11/8/10 11:40:32 PM
Basic Human Abilities
29
2.1.1.5 Dynamic versus Static Acuity An individual’s dynamic visual acuity of moving targets is not strongly related to static acuity. Thresholds for dynamic acuity increase rapidly for rates of motion exceeding 60 inches of visual field per second and are considerably higher with shorter viewing times or longer target travel distance. Both dynamic and static acuity decrease with age (Burg, 1966). 2.1.1.6 Accommodation (Focusing Abilities) Guideline 2.2: Typical Visual Deficiencies Designers need to be aware of typical visual deficiencies of their users, including nearsightedness, farsightedness, astigmatism, and aging eyes. This would require making text and images larger and with higher contrast on labels, displays, and documentation. Accommodation is the adjustment of the lens of the eye to focus light rays properly on the receptor cells of the retina. Normal accommodation and common types of inadequate accommodation abilities are shown in Figure 2.4. The figure shows what causes the common visual impairments of nearsightedness (myopia) and farsightedness (hyperopia) and describes the effect of age on the ability to focus on near objects (presbyopia). Presbyopia is the inability of the eye to focus sharply on nearby objects, resulting from loss of elasticity of the crystalline lens with advancing age. The average age of onset is 40 years. Astigmatism, which also decreases the ability to focus, is a defect in which the unequal curvature of one or more refractive surfaces of the eye, usually the cornea, prevents light rays from focusing clearly at one point on the retina, resulting in blurred vision.
2.1.1.7 Visual Field Figure 2.5 shows the binocular (for both eyes) visual field measured in degrees. The normal line of site can be as small as 10 degrees from the horizontal plane as shown in the figure. Figure 2.6 shows the monocular visual field (for the right eye). The field for the left eye would be the reverse, or mirror image, of the right eye (Woodson and Conover, 1964). 2.1.1.8 Color Vision There are two types of receptor cells in the retina: rods and cones. The rods function under dim light and do not respond to color. They are located on the periphery of the retina away from the fovea, which is in the center of the retina, where vision is the sharpest. The cones Retina
Normal far vision – Light focused on retina
Lens normal Object
Normal near vision – Light focused on retina Lens curved
d
Age
20
40
60
d = nearest point of focus (average inches)
4.0
8.8
40.0
Nearsightedness (elongated eyeball and/or excess curvature of lens) – Light focused in front of retina for far objects Farsightedness (shortened eyeball and/or inability to increase lens curvature sufficiently as in aging) – Light focused in back of retina for near objects
FIGURE 2.4
Common visual impairments and focus point distances for different ages.
TAF-ER562X_WINGR-09-0401-C002.indd 29
11/8/10 11:40:33 PM
30
Handbook of Human Factors in Medical Device Design
50° 15°
Color vision limit 30°
94°
30°
e 70°
al lin Norm t h of sig 94°
0°
FIGURE 2.5
Normal visual fields in vertical and horizontal directions for both eyes.
function under relatively higher-intensity light and respond to color. They are located primarily in the area of the fovea. Factors that influence color vision are brightness (i.e., light intensity), hue (dominant wavelength), and saturation (pureness of color). As illumination decreases, so does color vision sensitivity. Not all zones of the retina are equally sensitive to color. Toward the periphery, objects can still be distinguished even though their color cannot. Some colors are recognized at greater angles away from the fovea than others. Figure 2.7 shows the limits of the retina zones in which the various colors can, under normal illumination, be correctly recognized (Woodson and Conover, 1964). Figure 2.8 shows the sensitivity of the human visual system to different colors (wavelengths of light). One curve shows the sensitivity of the cone receptors, and the second Nasal
Temporal 55° 30°
60°
Cornea Pupil opening
Eye
65°
34°
Occluded by nose
Occluded by eyebrow No color (under normal levels of illumination) Blind spot
70°
40°
Color area (cones) Occluded by cheek
Lens Retina
Fovea
(Sharpest vision at fovea with rapidly decreasing acuity toward retina periphery)
FIGURE 2.6 Monocular visual field of the right eye showing occlusions and color areas. (From Woodson, W. E. and Conover, D. W., Human Engineering Guide, University of California Press, Berkeley, CA, 1964. With permission.)
TAF-ER562X_WINGR-09-0401-C002.indd 30
11/8/10 11:40:33 PM
Basic Human Abilities
31 Temporal Nasal White Blue Yellow Red Green
90° 65°
60°
50° 33° 25° 15° 10° Fovea
FIGURE 2.7
Retinal zones that are primarily sensitive to different colors.
curve shows the sensitivity of the rod receptors. The rod system takes over primarily after the eye is adapted to very low light levels, which is also known as dark adaptation (Glazer and Hammell, 1970). Guideline 2.3: Use of Color in Low Light Conditions Data from the human visual system have led to the following design recommendations: 1. Red and orange are poorly visible under low light conditions and should be avoided. 2. Blue, green, and yellow are equally visible under both low and higher light conditions and are good color choices, 3. Under low light conditions, blues and cyan colors are more visible. There is a shift in color sensitivity under these lower light conditions toward the blue end of the color spectrum. 48
100
Cone vision (color)
Rod vision (shades of gray)
Red
Yellow
Blue
20
Green
40
Orange
60
Violet
Relative visibility
80
0 400
450
500 550 Wavelength (mμ)
600
650
700
FIGURE 2.8 Relative visibility of different colors (wavelengths) for rod vision (dim light) and cone vision (brighter light). Curves are individually normalized and not normalized to each other.
TAF-ER562X_WINGR-09-0401-C002.indd 31
11/8/10 11:40:33 PM
32
Handbook of Human Factors in Medical Device Design
TABLE 2.4 Types of Deficient Color Vision and Their Frequencies Percentage of Population Affected Type
Effect
Trichromacy (three colors)
Normal color vision
Dichromacy (two colors) Protanopia (red blindness)
(Sees only two colors plus shades of gray) Inability to see hues other than yellow and blue with red and bluish-green seen as same shade of gray Inability to see hues other than yellow and blue with green and bluish-red seen as same shade of gray Inability to see hues other than red and green with yellow-green and purplish-blue seen as same shade of gray Inability to see hues other than red and green where color spectrum appears red at long wavelengths, green in the center, and red again at short wavelengths
Deuteranopia (green blindness) Tritanopia
Tetartanopia
Anomalous trichromacy
Protanomaly (red weak)
Deuteranomaly (green weak)
Tritanomaly Monochromatism
(Sees all colors but mismatches them, especially under dim light or small light sources) There is a foreshortened red end of the spectrum that requires more red than normal to match pure yellow; the spectrum is shifted, and normal yellow is seen as greenish There is no foreshortening of the red end of the spectrum, and more green is required to match pure yellow; the spectrum is shifted, and normal yellow is seen as orange More blue than normal is required to match cyan or blue-green Complete loss of color discrimination and poor acuity
Males 92.0
Females 98.0
1.0
0.02
1.1
0.01
0.0001
Very rare
Very rare
Very rare
1.0
0.02
4.9
0.38
Very rare 0.003
Very rare 0.002
2.1.1.9 Color Vision Deficiencies Approximately 8% of males and 2% of females have some degree of deficient color vision. Table 2.4 describes the various types of deficient color vision and their relative incidence in the U.S. population (Israelski, 1978). An example of the effects of color vision impairments is shown in Figure 2.9, which shows a patient monitor as seen by users with normal color vision, tritanopia, red blindness (protanopia), and green blindness (deuteranopia).
TAF-ER562X_WINGR-09-0401-C002.indd 32
11/8/10 11:40:33 PM
Basic Human Abilities
33
Normal color vision
Tritanopia
Red blind (protanopia)
Green blind (deuteranopia)
FIGURE 2.9 (See color insert following page 564.) Comparison of colors seen by various users of a patient monitor with normal and deficient color vision, including tritanopia, protanopia, and deuteranopia.
2.1.1.10 Color Discrimination Recommendations Up to eight saturated surface colors (excluding black and white) can be used for color coding with practically error-free discrimination for color normal people. Color coding with more than eight colors produces higher error rates. Fewer colors would place less demand on memory. Under strictly controlled conditions, with a high level of training and the use of various combinations of hue and saturation, up to 50 colors can be identified with high accuracy. In any color-coding scheme, colors should subtend a visual angle of at least 15 minutes of arc. Guideline 2.4: Color Only One Form of Coding In any design, color coding should be a redundant information source and never stand as the only means of coding. If the population for which equipment is being designed is known to include significant numbers of color-deficient users, color coding should be avoided.
However, if a significant number of users may be color-deficient and color coding is still desired, only three colors may be safely used according to military aviation standards for the Army and Navy (Army-Navy Aeronautical Specification AN-C-56). These colors are not pure colors and therefore allow better discrimination for color-deficient users: • Aviation red (MIL-C-2505A Red) • Aviation green (MIL-C-2505A Green) • Aviation blue (MIL-C-2505A Blue) Other red, green, and blue hues may cause confusion. Colors are not recommended for use at larger distances because blue and green are more likely to be confused. White or yellow should not be added to the code because of probable red–yellow and green–white confusion for color-deficient individuals.
TAF-ER562X_WINGR-09-0401-C002.indd 33
11/8/10 11:40:33 PM
34
Handbook of Human Factors in Medical Device Design
TABLE 2.5 Recommended Printed Color Codes Using the Munsell Color System 8-Color Code n 1R 9R 1Y 7GY 9G 5B 1P 3RP
7-Color Code
6-Color Code
5-Color Code
4-Color Code
p
n
p
n
p
n
p
n
p
999 892 946 960 1099 1087 1135 1003
5R 3YR 5Y 1G 7BG 7PB 3RP
1008 890 1128 1103 1095 1133 1003
1R 3YR 9Y 5G 5B 9P
999 890 1131 1101 1087 1005
1R 7YR 7GY 1B 5P
999 884 960 1093 1007
1R 1Y 9G 1P
999 946 1099 1135
n, book notation of Munsell color system; p, Munsell production number; R, red; Y, yellow; G, green; B, blue; P, purple.
2.1.1.11 Recommendations for Printed Colors Guideline 2.5: Printed Color Recommendations For printed materials, colors should be chosen from the Munsell color systems (Table 2.5) (Cleland, 2004; Conover and Kraft, 1958; McCormick, 1970).
The Munsell color system is the system of color notation developed by A. H. Munsell in 1905 and identifies color in terms of three attributes—HUE, VALUE, and CHROMA— which are described symbolically. The HUE (H) notation of a color indicates its relation to a visually equally spaced scale of 100 hues. There are five principal and five intermediate positioned hue steps within this scale. The hue notation in general use is based on the 10 major hue names: Red (5R), Yellow-Red (5YR), Yellow (5Y), Green-Yellow (5GY), Green (5G), Blue-Green (5BG), Blue (5B), Purple-Blue (5PB), Purple (5P), and Red-Purple (5RP). The VALUE (V) notation indicates the lightness or darkness of a color in relation to a neutral gray scale, which extends from absolute black (value symbol 0/) to absolute white (value symbol 10/). The symbol 5/ is used for the middle gray and for all chromatic colors that appear halfway in value between absolute black and absolute white. The CHROMA (C) notation indicates the degree of divergence of a given hue from a neutral gray of the same value. The scale of chroma extends from /0 for a neutral gray to /10, /12, /14, or further, depending on the strength (saturation) of the sample to be evaluated. 2.1.1.12 Recommendations for Colored Lights Guideline 2.6: Recommendations for Colored Lights Colored lights need to be chosen using available data to have good recognizability and the least amount of confusion. Table 2.6 recommends colors and describes their effect on color recognition of small-point light sources near the threshold of visibility (Dreyfuss, 1966). Table 2.7 shows 10 colored light choices that were shown to reduce confusion error; that is, these 10 wavelengths had a less than 2% misidentification error in experimental studies
TAF-ER562X_WINGR-09-0401-C002.indd 34
11/8/10 11:40:34 PM
Basic Human Abilities
35
TABLE 2.6 Recommendations for Colored Lights Color Discrimination a
Easiest Easier Difficult More difficult More difficult
Colors Recommended Red and green lights are easiest to recognize for color-normal individuals. White light is the next easiest to recognize. Yellow (or orange) is more difficult to recognize. Blue and green lights are very difficult to differentiate at a distance greater than10 feet. Yellow, white, and orange lights are difficult to differentiate at distances greater than 10 feet.
Source: Based on Dreyfuss, H., Measure of Man (2nd ed.), Watson Guptill, New York, NY, 1966. The best set of three colored lights is red, green, and white.
a
(Chapanis and Halsey, 1956). Use of the values in Table 2.7 increases the number of color choices that are easily recognizable and less likely to be misidentified.
2.1.1.13 Recommendations for Color Combinations (Legibility and Visibility) Guideline 2.7: Recommendations for Color Combinations Research has shown that the following lists are the best combinations of colors for providing good legibility, discrimination, and visibility.
The most legible color combinations for text are listed in order of legibility (Dreyfuss, 1966): 1. Black on white (most legible). 2. Black on yellow (most attention gained). 3. Green on white. TABLE 2.7 Ten Colored Light Choices That Have Less Than a 2% Misidentification Error Rate RGB Values Color Description Red Burnt orange Orange Yellow Yellow-green Green Light green Cyan Pale blue Medium blue
Wavelength (nm)
R
G
B
642 610 596 582 556 515 504 494 476 430
255 255 255 255 168 18 0 0 0 28
12 137 192 247 255 255 255 255 184 0
0 0 0 0 0 0 76 204 255 255
Source: Based on Chapanis, A. and Halsey, R., J Psychol, 42, 99, 1956. RGB, red, green, blue.
TAF-ER562X_WINGR-09-0401-C002.indd 35
11/8/10 11:40:34 PM
36
Handbook of Human Factors in Medical Device Design
4. Red on white. 5. White on blue. 6. Combinations of pure red and green or red and blue should not be used. 7. White on black may cause problems of smearing or irradiation of white on the black background if printing or electronic displays are not carefully controlled. For visibility of opaque colors under typical light conditions, the following colors and combinations are recommended: • • • •
Yellow is the most luminous and visible. Orange and red-orange hold maximum attention value. Blue is likely to be out of focus and indistinct. Red on blue or blue on red should not be used since each focuses differently on the retina and creates an induced three-dimensional effect called chromosteriopsis.
2.1.1.14 Dark Adaptation The rod system takes time to function efficiently (becoming dark adapted) after the eyes experience a change from bright to dark illumination conditions. There are both physiological and neurochemical changes that occur on dark adaptation. The pupils dilate, and the rod receptors become more sensitive. Guideline 2.8: Red Maintains Dark Adaptation The use of red goggles or red lighting does not affect rod vision and is useful in maintaining dark adaptation.
Figure 2.10 shows the lowering of visual thresholds as time in the dark progresses (Glazer and Hammel, 1970; Nutting, 1916). Full dark adaptation can take as long as 30 minutes, 5
Log of threshold brightness (μml)
Centrally fixated fields
Visual angle 2°
4
3° 3
5° 2
10° 20°
1 0
10 20 Minutes in the dark
30
FIGURE 2.10 Dark adaptation. Threshold of seeing decreases as a function of time in darkness and width of visual field or angle.
TAF-ER562X_WINGR-09-0401-C002.indd 36
11/8/10 11:40:34 PM
Basic Human Abilities
37
B
A
FIGURE 2.11 Distance and perceived size of objects. Although arrows A and B subtend the same visual angle at the observer’s eye and therefore produce images of the same size on his retina, arrow A is seen to be farther away and hence appears to be larger.
but significant adaptation can take place in 8 to 10 minutes. The human eye also adapts to bright light chiefly by constricting the pupil. On subsequent exposure to low light levels, the pupils will normally dilate, after which the rod system starts the chemical process of increasing sensitivity to these low light levels.
2.1.2 VISUAL PERCEPTION There are known limits to human visual perception and its processing. Some of these visual limitations and resulting common errors are summarized in this section. Awareness of these common limitations will reduce the display of ambiguous information. Chapter 8, “Visual Displays,” provides practical recommendations for the design of visual displays. 2.1.2.1 Distance and Perceived Size The principle of perspective states that objects appear to be smaller when they are farther away. Distance and perceived size are related in the manner shown in Figure 2.11. We learn in childhood the distance–size relationship, and therefore it is automatically taken into account when we observe distant objects. If two objects that are actually the same size are perceived as being at different distances, the one that seems to be farther away will look larger. However, designers may unintentionally create misperceptions in size perception, as in Figure 2.12.
FIGURE 2.12 Perceived size of distant objects. The cone on the right is perceived as being larger than the identical cone at the left only because it seems to be farther away. (From Kaufman, L. and Rock, I., Scientific American, July 1972. With permission.)
TAF-ER562X_WINGR-09-0401-C002.indd 37
11/8/10 11:40:34 PM
38
Handbook of Human Factors in Medical Device Design
FIGURE 2.13 Object size misperception. The horizon moon appears to be farther away, although it is not. The viewer automatically takes the apparent distance into account. The viewer then unconsciously applies the rule that, of two objects forming images of equal size, the more distant must be the larger. (From Kaufman, L. and Rock, I., Scientific American, July 1972. With permission)
2.1.2.2 True Object Size The same concept applies to perceived object size in the presence of misperceived visual reference information. The apparent-distance theory holds that this is what happens in the case of the moon illusion and is illustrated in Figure 2.13. Guideline 2.9: Object Size and Distance Misperceptions Designers should not create objects that may be misperceived by users because of well-known visual processing limitations and illusions.
2.1.2.3 Common Visual Illusions The human visual system is easily fooled by visual illusions. Designers need to be aware of these visual illusions and of course avoid any graphic treatments that might trigger these problems of unreliable visual interpretations. Parallax error is one form of visual illusion that is commonly encountered and must be avoided. The error is seen as an apparent change in the position of an object, such as a medical device meter reading caused by a change of the observer’s line of sight, as illustrated in Figures 2.14 and 2.15.
Viewing from the left
Viewing from the right
FIGURE 2.14 Parallax errors. Note that the apparent reading of the meter needle position changes depending on the observer’s position from either the left or the right viewing angle. The image of the needle on the background display mirror is the true reading. Mirrors such as these are an aid to reduce parallax error.
TAF-ER562X_WINGR-09-0401-C002.indd 38
11/8/10 11:40:34 PM
Basic Human Abilities
39
FIGURE 2.15 Straight-line illusions. Note how the observer sees lines that are not straight even when they are.
2.1.2.4 Perception of Motion Humans are often confused by slow relative motion between us and another object when there is no dependable frame of reference. Apparent motion can be induced, as is commonly observed in motion pictures (24 still frames per second with each shown twice) as well as television and cathode ray tube (CRT) displays (typically 50 or 60 frames per second). Another example of induced apparent motion is the phi-phenomenon, in which rapid successive flashes of individual lights arranged in a row or circle give the appearance of individual light-source motion. Guideline 2.10: Perception of Motion The designer should be aware of methods to create apparent motion in visual displays either to avoid confusion or to take advantage of them as potential information sources.
2.1.2.5 Flickering Lights Prolonged perception of flicker (over 20 minutes) in a light being flashed on and off causes visual fatigue and annoyance. A related concept is when flickering lights are perceived as being steady. The frequency at which a flashing light is perceived as having a continuous intensity level is called the critical fusion frequency (CFF). The CFF increases with increasing average light intensity and with decreasing proportion of the light–dark cycle occupied by the flash (percent modulation or duty cycle). CFF varies from 2 Hz up to 50 to 60 Hz for high-intensity light sources. 2.1.2.6 Photosensitive Epilepsy Flashing lights or the flicker of a computer monitor at certain speeds can trigger a seizure in susceptible individuals. This problem is called photosensitive epilepsy, photic epilepsy, or photogenic epilepsy (Harding and Jeavons, 1995): • Approximately 1 in 200 people have epilepsy, and of these, only 3% to 5% have seizures induced by flashing lights. • Photosensitivity is more common in children and adolescents and becomes less common after the early 20s.
TAF-ER562X_WINGR-09-0401-C002.indd 39
11/8/10 11:40:35 PM
40
Handbook of Human Factors in Medical Device Design
• The flicker frequencies that can trigger seizures vary greatly by individual and are in the range of 5 to 60 Hz. • Only 50% of photosensitive people are sensitive to 50 Hz, but 75% are sensitive to 25 Hz. Rates below 5 Hz are considered relatively safe. Guideline 2.11: Display Flicker Liquid crystal displays have no flicker and are preferable to CRTs in situations where flicker should be avoided, such as in medical display devices for use in the hospital rooms of seizure patients.
2.1.3 AUDITORY PERCEPTION This section describes the basics of human hearing abilities (see Chapter 10, “Alarm Design,” and Chapter 3, “Environment of Use,” for more detail and recommendations on environmental effects of sound). Loudness or sound volume is the subjective measure equivalent to sound intensity. The pitch or tone of a sound is the subjective measure equivalent to sound frequency. The loudness of a sound is perceived differently at various frequencies. For example, infusion pump alarm signals will be perceived as louder at higher frequencies. Thresholds for hearing sounds and feeling pain at different sound intensity levels are also a function of frequency (see Figure 2.16). 2.1.3.1 Loudness Measurements Two commonly used subjective loudness measures of sound intensity are phones and sones: • Loudness level (phones). Phones are the subjective measure of any tone intensity that is numerically equal to the sound pressure level (SPL) in decibels (dB) of a 140 Threshold of pain 120 Threshold of feeling
Prolonged exposure causes hearing loss
Intensity level (dB)
100
80
60
40 Words combine to make meaning 20
Threshold of hearing tones
Sounds become words Sounds can be heard
0
–20
FIGURE 2.16 Thresholds of hearing and pain. The curves show the intensity levels of sound as a function of frequency for two sets of thresholds: minimal threshold to hear pure tones and the threshold of pain. (From Woodson, W. E. and Conover, D. W., Human Engineering Guide, University of California Press, Berkeley, CA, 1964. With permission.)
TAF-ER562X_WINGR-09-0401-C002.indd 40
11/8/10 11:40:35 PM
Basic Human Abilities
41
TABLE 2.8 Loudness for Common Noise Sources, Including Health Care Settings Noise Source Patient room at night (EPA recommended) Residential inside, quiet Patient room during day (EPA recommended) Household ventilating fan Automobile at 50 feet Typical patient room—peaks Anesthesia equipment—peaks ICU—peak sound levels IV pump alarm Hospital beeper Operating room—peaks Inside MRI machine Punch press, 3 feet Nail-making machine, 6 feet Pneumatic riveter
SPL (dB)
Loudness (sones)
35 42 45 56 68 70 76 80 85 89 90 95 103 111 128
<0.5 1 1.3 7 14 24 54 74 99 127 164 222 350 800 3,000
standard 1,000-Hz tone. This measure indicates subjective equality of any tone compared to a 1,000-Hz standard. • Loudness (sones). Sones are another measure of relative subjective sound intensity. One sone is defined as the loudness of a 1,000-Hz tone at 40-dB SPL. A sound twice as loud is 2 sones, a sound half as loud is half a sone, and so on. Common noise sources and typical loudness values in sones are listed in Table 2.8 (McCormick, 1970; Weinger and Englund, 1990). 2.1.3.2 Relationship of Phones to Sones Sones are measured on a logarithmic scale, while phones are measured on a linear scale. This relationship is subject to many influences, but it can be shown reasonably well by the line in Figure 2.17. Basically, loudness in sones doubles for every 10-phone increase in loudness level (Stevens, 1955). 200
Loudness (sones)
100 50 20 10 5 2 1 40
50
60 70 80 90 Loudness level (phones)
100
110
FIGURE 2.17 The relationship between two common measures of subjective loudness. Sones (loudness) and phones (loudness level).
TAF-ER562X_WINGR-09-0401-C002.indd 41
11/8/10 11:40:35 PM
42
Handbook of Human Factors in Medical Device Design
TABLE 2.9 Loudness (Sones) Calculation for Complex Sounds (SPL) Band (dB) 20 30 40 50 60 70 80 90 100 110 120
Band Loudness Index (S) for Octave Bands Midpoint of Octave Band (Hz) 31.5
63
125
250
500
0.16 0.49 0.07 0.37 0.77 1.18 0.26 0.62 1.13 1.82 2.24 0.94 1.56 2.44 3.4 4.1 2.11 3.2 5.0 6.2 7.4 4.3 6.7 9.3 11.1 13.5 8.8 13.6 17.5 21.4 26.5 18.7 28.5 35.3 44.0 56.0 44.0 61.0 77.0 97.0 121.0 105.0 130.0 160.0 197.0 242.0
Loudness Level
1,000
2,000
4,000
8,000
(sones)
(phones)
0.18 0.67 1.44 2.68 4.9 8.8 16.4 32.9 71.0 149.0 298.0
0.30 0.87 1.75 3.2 5.8 10.5 20.0 41.0 90.0 184.0 367.0
0.45 1.10 2.11 3.8 7.0 12.6 24.7 52.0 113.0 226.0
0.61 1.35 2.53 4.6 8.3 15.3 30.5 66.0 139.0 278.0
0.25 0.50 1.00 2.00 4.00 8.00 16.0 32.0 64.0 128.0 256.0
20 30 40 50 60 70 80 90 100 110 120
2.1.3.3 Loudness (Sones) Calculation for Complex Sounds Complex sounds, such as the drone of a laboratory diagnostic system for blood testing, have many combinations of sounds at various frequencies. Sones are calculated for complex sounds by using the following procedure. The SPL is measured for each of the nine octave bands with center frequencies listed in Table 2.9 (McCormick, 1970; Peterson and Gross, 1967). After obtaining these values, the procedure is as follows: 1. From Table 2.9, find the proper loudness index for each band level (S). 2. Add all the loudness indexes (∑ S). 3. Multiply this sum by 0.3. 4. Add this product to 0.7 times the index that has the largest value (Smax). The total loudness in sones is (0.3 ∑ S + 0.7 Smax). 5. This total loudness (sones) can be converted to loudness level (phones) by using the two columns at the right of the Table 2.9. 2.1.3.4 Pitch Measurement Pitch is a subjective attribute of sound quality and is determined primarily by frequency but also by intensity and the complexity of a sound’s spectrum (see Figure 2.18). The scale unit is the mel, which is defined as the pitch of a 1,000-Hz tone at 40-dB SPL. 2.1.3.5 Differential Hearing Thresholds Humans can discriminate sounds on the basis of sound frequency and intensity (Figures 2.19 and 2.20). The term just noticeable difference (JND) is used by psychophysicists to mean a difference in some stimulus attributes that is detectable 75% of the time on the average (Foley and Moray, 1987).
TAF-ER562X_WINGR-09-0401-C002.indd 42
11/8/10 11:40:35 PM
Basic Human Abilities
43 4,000
Pitch (mels)
3,000
2,000
1,000
0 20
50
200 500 2,000 5,000 20,000 Frequency (Hz)
FIGURE 2.18 Pitch in the subjective units of mels as a function of frequency. (From Stevens, S. S. and Vollunan, J., Am J Psychol, 53, 329, 1940.)
2.1.3.6 Effects of Aging on Hearing Sensitivity Guideline 2.12: Adjustable Volume Sounds for medical devices should have volume adjustments to control for ambient noise levels and to accommodate older users with hearing loss.
Just noticeable difference in frequency (Hz)
As we age, our hearing sensitivity decreases, particularly for higher frequencies. Figure 2.21 shows the decrease in hearing sensitivity with increasing age across the normal 50
20
5 dB
10
10 dB 5
15 dB 20 dB 40 dB 70 dB
2 50
100
200
500
1,000
2,000
5,000
10,000
Frequency (Hz)
FIGURE 2.19 Just noticeable differences in frequency changes to pure tones of various frequencies as a function of sound pressure levels. Smaller-frequency differences are more noticeable at higher sound intensity levels. (From Shower, E. G. and Biddulph, R., J Acoust Soc Am, 3, 275, 1931. With permission.)
TAF-ER562X_WINGR-09-0401-C002.indd 43
11/8/10 11:40:35 PM
44
Handbook of Human Factors in Medical Device Design
Just noticeable difference in sound pressure level (dB)
10
5 2
1 1,000 Hz 70 Hz White
0.5
noise 1,000 Hz
4,000 Hz 0.2
0
20
40 60 80 100 Level above threshold (dB)
120
FIGURE 2.20 Just noticeable differences in sound pressure level for pure tones of various frequencies and white noise. Differences in sound intensity are easier to hear at higher frequencies. (From Miller, G. A., J Acoust Soc Am, 19, 609, 1947 and Riesz, R. R., Physiol Rev, 31, 867, 1928. With permission.)
spectrum of human audition with separate curves for males and females. There are many causes for age-related hearing loss (also called presbycusis), including the following: • • • •
Physiological changes to the inner ear Cumulative effects of exposure to loud noises Effects of stress Indirectly due to some slowing of the cognitive processing of auditory information
Commonly, as hearing sensitivity decreases for frequencies above 2,000 Hz, it becomes increasingly more difficult to hear speech. Speech consonants, such as “b,” “c,” “f,” and “t” become harder to discern. Women
–5 0
500 and 1,000 Hz
5 Hearing loss (dB)
Men
500 Hz
10
1,000 Hz
15
2,000 Hz 3,000 Hz 4,000 Hz
20
2,000 Hz
25 30
3,000 Hz 4,000 Hz
35 40
20
30
40
50
60
70 20 Age (yr)
30
40
50
60
70
FIGURE 2.21 The loss of hearing with age for males and females as a function of frequency. (From American Standards Association, The Relations of Hearing Loss to Noise Exposure, American Standards Association, New York, 1954. With permission.)
TAF-ER562X_WINGR-09-0401-C002.indd 44
11/8/10 11:40:36 PM
Basic Human Abilities
45
2.1.4 OTHER SENSORY MODALITIES Other sensory modalities are generally poorer information input systems relative to the visual and auditory systems. Touch, vibration, temperature, pain, and other skin sensations are not as fully developed neurologically in humans as our sight and hearing. A brief description of these other sensory modalities follows. 2.1.4.1 Skin (Somesthetic) Senses Senses related to the skin are multifaceted (Woodson and Conover, 1964; Corso, 1967).
Threshold amplitudes (thousandths of an inch)
• Touch—The stimulus for this sense is deformation of the skin due to pressure being applied. Sensitivity varies with location on the body to which pressure is applied; for example, JND thresholds range from pressures of 3 g/mm2 on the fingertips to 12 g/mm2 on the back of the hand and from 0.2 g/mm2 on the cornea of the eye to 250 g/mm2 on the thick parts of the sole of the foot. More rapid dynamic application of pressure results in lower thresholds. Once a constant pressure is applied, the sense will adapt, and awareness of contact will cease, as in the case of wearing surgical gloves. Laparoscopic surgical instruments include touch as one source of feedback (i.e., tactile feedback) to the surgeon. • Vibration—Sensitivity to local skin vibrations is another dimension of touch and varies with location on the body and the frequency of the vibrations (Figure 2.22). Medical devices such as implantable heart rhythm monitors use vibrations as warnings to patients of an incipient electrical shock. • Temperature—Sensitivity to hot and cold stimuli also varies with body location. Thinner, softer skin areas, such as the inner thighs, are more sensitive than rough, thick skin areas, such as the soles of the feet. Subjective judgments of absolute temperature of objects or fluids in contact with the skin are unreliable. Even relative temperature judgments are often erroneous. Normal skin temperature is in the range 90.5°F to 92.3°F. Skin temperatures below 32°F and above 125°F become painful. • Pain—This important sense provides a clear signal of a hazardous condition. Common stimuli include thermal, mechanical, chemical, and electrical sources. Pain localization can be precise but is sometimes referred to distant locations. • Other skin sensations—Other skin sensations are generally combinations or variations of the primary sensory modalities, for example, the sensation of moisture is a combination of pressure and thermal stimuli.
FIGURE 2.22
0.3 0.2 0.1 0.03 0
10
100 250 Frequency (Hz)
1,000
4,000
Vibration sensitivity thresholds as a function of frequency.
TAF-ER562X_WINGR-09-0401-C002.indd 45
11/8/10 11:40:36 PM
46
Handbook of Human Factors in Medical Device Design
2.1.4.2 Muscle Sense Proprioception and kinesthesis are the sensory feedback mechanisms for motor muscle control and body posture. Together these senses provide important information for the coordination of complex physical acts. Proprioception is the usually unconscious perception of spatial orientation arising from stimuli within the body itself and would include knowledge of where a joint is positioned. Kinesthesis is the sensation that informs our brains about joint motion and acceleration. For example, a medical device user can discriminate differences in control knob shape, position, and degree of rotation through a combination of these senses (Carlson, 2000). Sensors in the muscles provide information about muscle stretch (tension) and tone. 2.1.4.3 Sense of Balance Organs located in the inner ear provide sensory information necessary to maintain balance or orientation and to detect motion of the body. These organs are part of the vestibular system and include the semicircular canals. The semicircular canals are a three-part fluid-filled network, detecting acceleration in the three perpendicular planes. The absolute threshold for detection of angular acceleration of the body is 0.1°/s2 (Carlson, 2000). 2.1.4.4 Chemical Senses The taste sense receptors (gustation) are located on the tongue and provide four basic taste qualities: salty, sour, bitter, and sweet. Other taste sensations are combinations of these four. The smell sense receptors (olfaction) are located in the upper part of the nasal cavity and are estimated to be 10,000 times more sensitive than taste for absolute detection thresholds but much poorer for difference detection thresholds. The fundamental odors are considered to be spicy, fragrant, resinous, burnt, and putrid. Stimulus thresholds vary; 5.83 mg/L of air for detection of ethyl ether is required, whereas the threshold for musk oil is 1.00004 mg/L of air. The commonly accepted conversion to parts per million (ppm) is 1 mg/L = 1 ppm. Both chemical senses are unsuitable for reliable information transmission primarily because they quickly adapt to incoming stimuli so that awareness of these stimuli ceases. Neither of these sources should ever be depended on as a primary information source even though they often do have some practical information value, as in the detection of hazardous gases or burning insulation (Carlson, 2000).
2.1.5 OTHER PERCEPTUAL ABILITIES There are a number of other human perceptual attributes that may be of relevance to device designers. This section briefly describes the extent and limitations of these human abilities. 2.1.5.1 Estimation of Time Intervals Even with extensive training, human judgment of time passage is generally inaccurate and unreliable. This statement applies whether a person is asked to produce a stated interval of time or to estimate time that has passed. Research has shown that we tend to underestimate elapsed time when we are active and overestimate elapsed time when we are passive. Table 2.10 illustrates the large range of error in these judgments (Fraise, 1963). 2.1.5.2 Estimation of Other Physical Quantities Humans have inherent biases in their ability to estimate many other physical quantities (Table 2.11). Designers need to take into account the imperfect estimation skills of device users.
TAF-ER562X_WINGR-09-0401-C002.indd 46
11/8/10 11:40:36 PM
Basic Human Abilities
47
TABLE 2.10 Errors in Judgments of Elapsed Time Estimation Activity During a Standard 200-Second Interval
Average Estimation (in seconds)
Average Percentage Error
Range of Error (standard deviation)
Passive Resting and trying to sleep
241.7
12
107.8
Holding the arms outstretched
228.4
14
96.2
Listening to a metronome 66 beats per minute 184 beats per minute
223.7 214.1
12 7
92.4 85.2
Pressing a point on the skin
210.2
5
78.4
Active Reading a passage in a mirror
181.8
14
77.6
Taking dictation
174.6
13
77.6
Doing numerical division
168.9
11
70.2
TABLE 2.11 Human Estimation of Physical Quantities Physical Quantity
Human Estimation Tendency
Comments and References
Horizontal distance
Underestimate
Vertical height
Overestimate when looking up Tendency to overestimate is greatest Underestimate when looking down for pilots at night (Weintrab and Virsu, 1975)
Speed of an object
Overestimate if object is accelerating
(Hatayama and Tada, 1972)
Constancy of speed
Speed perceived as fluctuating
(Runeson, 1974)
Geometric angle
Underestimate acute angles Overestimate obtuse angles
(Weintrab and Virsu, 1975)
Temperature (ambient)
Overestimate when hot Underestimate when cold
Depends greatly on adaptation level, humidity, and air movement (Geldard, 1972)
Weight of object
Overestimate if bulky Underestimate if compact
(Geldard, 1972)
Number of items (without counting)
Consistently underestimate
(Bevan et al., 1963)
Volume, area, object temperature, Unreliable estimates with no acceleration, and compass bearing general tendency
TAF-ER562X_WINGR-09-0401-C002.indd 47
Thirty percent of the population has some depth perception deficiencies (Richards, 1973)
(Geldard, 1972)
11/8/10 11:40:36 PM
48
Handbook of Human Factors in Medical Device Design
2.2 HUMAN INFORMATION PROCESSING This section provides an overview of basic human abilities in processing sensory and perceptual input information. The point at which the perceptual process ends and higher-level information processing begins cannot be clearly defined. A bit is defined by the field of information theory as the amount of information obtained when one of two equally likely alternatives is specified.
2.2.1 LIMITATIONS ON INFORMATION PROCESSING ABILITIES 2.2.1.1 Channel Capacity Channel capacity is remarkably similar among the sensory modalities in terms of the maximum number of levels of input stimuli that can be reliably discriminated on an absolute basis. The maximum number of levels for a single dimension of a sensory channel clusters around 7.0, which represents 2.8 bits of information for each single presentation of the input stimulus. Humans are better at making relative judgments than absolute judgments of input stimuli (Tables 2.12 and 2.13). For example, the average person can reliably identify about five different tone frequencies when each is presented in isolation from the others, while the number of reliable discriminations increases to over 1,800 if only a relative discrimination between two tones presented in sequence is required (adapted from Miller, 1956; Van Cott and Kinkade, 1972; Woodson and Conover, 1964).
TABLE 2.12 Channel Capacities for Various Human Senses with One Stimulus at a Time Number of Discriminable Levels Sense
Stimulus Dimensions
On an Absolute Basis (H = bits/stimulus)
Vision
Pointer position on a linear scale Size of squares Areas Hue
15 (3.9) 5 (2.2) 6 (2.6) 9 (3.1)
Brightness Flashing rate of white light Line length Direction of line inclination Line curvature Loudness Pitch Intensity Duration Location on chest Frequency Intensity (salty or sweet) Intensity
5 (2.2) 3 (1.7) 8 (3.0) 11 (3.3) 5 (2.2) 5 (2.2) 5 (2.2) 4 (2.0) 5 (2.2) 7 (2.8) 5 (2.2) 5 (2.2) 4 (2.0)
Audition Touch (vibration)
Taste Smell
TAF-ER562X_WINGR-09-0401-C002.indd 48
On a Relative Basis
128 (medium brightness) 570 (white light) 460 (duty cycle = 0.5)
325 (at 2,000 Hz) 1,800 (at 60-dB SPL) 1.5
180
11/8/10 11:40:36 PM
Basic Human Abilities
49
TABLE 2.13 Channel Capacities for Various Human Senses with Multiple Stimuli Being Presented at the Same Time Number of Discriminable Levels Sense
Stimulus Dimensions
Vision
Size, brightness, and hue (varied together) Position of points in a square (no grid) Hue and saturation Loudness and pitch Loudness, pitch, cadence, duty cycle, total duration, and specific location
Audition
On an Absolute Basis (H = bits/stimulus) 18 (4.1) 24 (4.6) 13 (3.6) 9 (3.1) 150 (7.4)
On a Relative Basis >570 >570 >1,800
Guideline 2.13: Improving Information Discrimination Information discrimination can be increased by allowing relative judgments to be made, presenting multidimensional stimuli, and increasing the rate of sequential presentation of stimulus material.
2.2.1.2 Attention People usually attend to only one source of sensory information at a time. This is true because we are basically single-channel processors, although we may do some time sharing of attention similar to the way a single-processor computer multiplexes or multitasks. Information from other unattended input channels is not totally blocked but rather is processed only incompletely with the result that sometimes certain kinds of information do get through. For example, individuals at a cocktail party can attend to a single voice amidst a babble of competing voices (the so-called cocktail party effect) yet still hear and recognize their own name being spoken in another conversation. A possible design implication is that a simple alternating tone or flashing light could precede important information presented in a channel likely to be otherwise unattended. 2.2.1.3 Vigilance (Sustained Attention) Psychologists are beginning to understand more about human performance during long periods of sustained attention (usually called vigilance activity) such as during a search for infrequent trouble signals on a display (e.g., missed heartbeats on a patient monitor). Table 2.14 lists various task conditions that affect human performance during prolonged vigilance (Van Cott and Kincaid, 1972; Wickens and Carswell, 1997).
2.2.2 SPEED OF INFORMATION PROCESSING 2.2.2.1 Reaction Time The time it takes for a person to react to an input stimulus and initiate a response is called reaction time. Simple reaction time (RT) involves only one response to a single stimulus.
TAF-ER562X_WINGR-09-0401-C002.indd 49
11/8/10 11:40:36 PM
50
Handbook of Human Factors in Medical Device Design
TABLE 2.14 Task Conditions That Affect Vigilance Performance Improved Probability of Signal Detection • Simultaneous presentation of signals to dual channels • Observers monitoring display in pairs; members of pairs permitted to speak with one another; 10 minutes of rest each 30 minutes of work; random schedule inspection by supervisor • Introduction of artificial signals during vigilance period to which a response is required • Introduction of knowledge of results of artificial signals Decreased Probability of Correct Detection • Introduction of artificial signals for which a response is not required • Higher or lower task load • Introduction of a secondary display monitoring task • Users report only signals of which they are sure Change in Probability of Detection with Time • A short pretest followed by infrequently appearing signals during vigilance • High initial probability of detection, decreasing rapidly • A few pretest signals before vigilance period • No CR reduces decrement in probability of detection with time • Prolonged continuous vigilance • No CR decreases probability of correct signal detection
Simple RTs for the different senses are provided in Table 2.15 (Brebner and Welford, 1980; Pierce and Karlin, 1957). Response time, a related concept, is the sum of reaction time, the cognitive processing time to an input stimulus, and the time to generate a response. The processing time is sometimes called think time. Choice RTs require a choice to be made among a number of stimuli and responses and typically are longer than simple RTs. The rate at which information is processed, such as during choice RT, is dependent on many factors (e.g., sensory channel, the kind and intensity of stimulus) (Pierce and Karlin, 1957). RT can decrease with training (up to 10%), the use of an alerting signal, an increase in intensity or duration of the stimulus, and optimized compatibility between delivered stimulus and expected response. RT increases (deteriorates) generally with age, fatigue, and the use of central nervous system– depressing drugs, such as alcohol. TABLE 2.15 Simple Reaction Times for Various Sensory Stimuli Stimulus Type Visual Auditory Tactual (haptic) Pain Cold Warm Movement (body rotation)
TAF-ER562X_WINGR-09-0401-C002.indd 50
Reaction Time (ms) 150–225 120–185 115–190 400–1,000 150 180 520
11/8/10 11:40:36 PM
Basic Human Abilities
51 50
Errors per 100 moves
40
Fast speeds (8 decisions per minute)
30
20
Slow speeds (3 decisions per minute)
10
0
0
5 10 20 30 40 Load (number of columns displayed)
50
FIGURE 2.23 Speed versus accuracy trade-off (error rate versus speed of decisions). These data come from a study using a panel of varying numbers of columns of comparison numbers. The mental load increased with the number of columns displayed. The fast and slow speeds were, respectively, six and three decisions per minute. (From Mackworth, J. F. and Mackworth, N. H., J Opt Soc Am, 52, 713–716, 1958 and McCormick, E. J., Human Factors Engineering, McGraw Hill, New York, 1970. With permission.)
2.2.2.2 Speed versus Accuracy The so-called “speed-versus-accuracy trade-off” describes an approximately linear relationship with accuracy decreasing with either increased demands/faster response or additional workload. This relationship is shown in Figure 2.23 as the effects of increasing mental load on errors. 2.2.2.3 Human Memory Experimental cognitive psychologists refer to three basic types of human memory: sensory information memory, working short-term memory (STM), and long-term memory (LTM). Sensory memory has less bearing on device design and thus is not discussed further. Longterm memory is further categorized into two main areas: procedural (memory for processes and how to do things) and declarative (memory for facts and what to do). Table 2.16 summarizes the most important and distinguishing characteristics of working, declarative, and procedural memory (Kyllonen and Alluisi, 1987). 2.2.2.4 Working Memory This kind of STM transiently holds new information from the senses or other mental processes as might a computer data buffer. STM can be characterized by having fast access and retrieval time, limited capacity, and rapid loss of content unless actively attended to. The fast-access advantage of STM is countered by its limited capacity. The capacity is limited to five to nine “chunks” of information. This concept of working memory size limitation is also known as the “magic number seven plus or minus two” (Miller, 1956). The unit “chunk” is not defined precisely but can be considered a psychologically meaningful unit of information for material to be placed in STM. A large but limited number of bits of information may be contained in a chunk. This depends on the schemes used to recode bits
TAF-ER562X_WINGR-09-0401-C002.indd 51
11/8/10 11:40:37 PM
52
Handbook of Human Factors in Medical Device Design
TABLE 2.16 Main Characteristics of Memory Systems Characteristic Primary function
Working Memory (Short Term)
Long-Term Memory Declarative
Center of all thought/learning Stores meaning of inputs
Subset of declarative
Procedural Permanent storage of how-to knowledge
Storage of facts
Temporary storage in flux Capacity
Highly limited, 7 ± 2 chunks
Unlimited
Unlimited
Contents
Primarily acoustical codes
Semantic codes (primary)
Same as declarative
Secondarily visual/spatial
Spatial codes Acoustic codes Motor codes (physical movement skills) Temporal codes
Information units
Same as declarative
Concepts Schemata/frames/scripts
Production rules from very specific to general (if–then rules)
Organization
Same as declarative
Hierarchical with multiple levels of complexity
Flat
Learn/forget processes
Decays with time (73 seconds Learning by being told for one item, 7 seconds for (passive advice taking) three items)
Generalizations (inductive and deductive)
Increased decay time with rehearsal
Encoding
Learn by doing (active practice)
Interference from similar stimuli
Limited by retrieval paths and associations
Strengthening and reinforcement
Displacement (3–7 slots)
Very slow decay
Discrimination Analogies Problem solving
of input information into psychologically meaningful units. For example, if an individual had to memorize randomly generated alphabet letters, each letter would be a chunk. But if the task involved the memorization of randomly chosen words, then each word would be a chunk. A chunk of medical information could be patient name, room, and bed position. Information can be held in STM or working memory for as long as it is actively given attention, which usually involves rehearsing or reviewing the material over and over again. There is a human tendency to store information in working memory in an acoustical form whenever possible. The practical implication is that devices should be designed to minimize listening and talking when information is being held in working memory because these activities interfere more with STM.
TAF-ER562X_WINGR-09-0401-C002.indd 52
11/8/10 11:40:37 PM
Basic Human Abilities
53
2.2.2.5 Long-Term Memory LTM has capacity ranging from 109 to 1015 or more bits of information. Retrieval from declarative LTM (facts) is slower than from working memory. Information gets transferred into LTM (either from STM or directly from the senses) only if appropriate links or associations can be established with psychologically meaningful material already in the long-term store LTM. Procedural memory (how to do things) is best learned by actively practicing a skill and appears to be very slow to decay (e.g., riding a bicycle, programming a computer, or remembering how to perform a laryngoscopy). This is because declarative LTM is organized much like a thesaurus with information having close meaningful associations being grouped near each other in the brain. People see and hear verbal messages more quickly and accurately if words with associated meanings are grouped close together, especially when competing signals are present. 2.2.2.6 Estimation and Decision-Making Abilities Humans are limited in both estimation and decision-making abilities. Among the highestlevel mathematical operations that people can perform “in their heads” are first-order integration and differentiation; these are performed crudely at best. Even simple arithmetic operations are performed poorly as soon as a person is stressed by demands for higher speed or accuracy. Our guessing behavior or probabilistic estimation skills show the following human tendencies (Kahneman and Tversky, 2000): • Overestimation of true probability for low-probability events and underestimation of true probability for high-probability events (especially overestimation of the probability of chains of unlikely events and underestimation of cumulative risks of events over a long period of time, such as the relationship between smoking and cancer). • Overestimation of true probability for events viewed as favorable and underestimation of true probability for events viewed as unfavorable. An example of overestimation is playing the lottery with enormously high odds. An example of underestimation is feeling threatened by a neurosurgery with an 80% success rate. • Unwillingness to believe constant probabilities for outcomes of successive independent events (also called the gambler’s fallacy, e.g., disbelief that p = .50 for a tail on the next toss after a fair coin has been tossed 10 times, each time coming up tails). • Human decision making is often not logical and, depending on circumstances or on how a problem is framed, we can be risk aversive or risk seeking. • Humans want to avoid false alarms when it comes to safety-related events. Signal detection theory (Swets, 1964) makes certain predictions about our decision-making criteria levels depending on the relative costs for false alarms versus the benefits of hits or true correct decisions. These trade-offs between decision criteria level and outcome consequences are dynamic and sometimes rapidly change. For example, when we are answering yes/no questions as part of a dangerous disease-screening questionnaire, we answer “yes” more frequently with more lenient criteria for decision making if the benefits of early valid detection of the disease are very positive and the therapy is simple, and we answer “yes” less frequently with stricter criteria if the costs of a false alarm are high (e.g., a very painful therapy for the disease). • Humans rationalize and rethink decisions in ways that are not always obvious or predictable. At the root of these problems is the concept of cognitive dissonance as
TAF-ER562X_WINGR-09-0401-C002.indd 53
11/8/10 11:40:37 PM
54
Handbook of Human Factors in Medical Device Design
studied by Festinger (1957). When faced with two options, we need to resolve any conflict that each choice presents in terms of differences in positive and negative consequences. Dissonance or conflict may exist both before and after we make decisions. Among common mechanisms for resolving dissonance are the following: • Rationalization—After making a highly conflicted choice, we reduce dissonance by altering our beliefs about something for which we formerly had strong convictions. For example, we choose to spend time participating in a long, boring memory experiment for very little reward or pay. We later rationalize our choice by saying that the experiment really was not that boring but was in fact very interesting. • Selective exposure or filtering—We begin to notice only positive attributes or features of our conflicted choice and ignore or filter any subsequent negative attributes that reveal themselves. For example, after being prescribed a certain drug, you begin to notice with increased frequency many other people taking it and also notice the same drug appearing in television ads. • Commitment—We increase our confidence about the quality of our conflicted choice and become steadfast in this belief. This behavioral tendency is related to sunk-cost bias, whereby we are not willing to give up or cut our losses on a bad investment that keeps falling in value. • Defensiveness—We resist any challenges to the soundness of our conflicted choices and categorize all feedback on our decision as wrong or biased. • Regret—We admit that we made the wrong conflicted decision and wish we had made a different choice. An example is the so-called buyer’s remorse we experience after making a large expenditure for an item of questionable quality (Malle, 2001). Regret has been shown to have different long-term and shortterm strengths. − Short-term regret is stronger for actions or commitments taken, such as buying a house. − Long-term regret is stronger for inactions, such as the trip you never took.
2.3 HUMAN RESPONSE CAPABILITIES In addition to limits in the ability to sense, perceive, and process information, humans have significant limitations in response capabilities after initial cognitive processing is finished. Only highlights of this broad topic are covered in this section, and the reader is referred to more in-depth treatments (see Eastman Kodak Company, 1989; Salvendy, 1997). Physical response data, such as strength, reach, and endurance, are covered in Chapter 4, “Anthropometry and Biomechanics.”
2.3.1 SPEED OF MOVEMENT Speed of hand and arm movements are dependent on a number of factors (Brown, 1949; Fitts, 1954). Hand movement time increases nonlinearly as a function of distance moved (Figure 2.24): • Maximum hand velocity for distances less than 3.2 feet is about 10 feet/sec. • Maximum hand velocity is about 20 feet/sec for longer distances.
TAF-ER562X_WINGR-09-0401-C002.indd 54
11/8/10 11:40:37 PM
Basic Human Abilities
55 1.0 0.8 0.6
Time (sec)
0.4
Total time
Primary movement time Reaction time
0.2
Secondary movement time
0.1 0.08 2
4
6 8 10 20 Distance moved (cm)
40
FIGURE 2.24 Hand movement time as a function of distance moved. Movement time is for nonrepetitive movements of an object by the right hand from one position to another with complete visual feedback. Primary movement time is the time taken to make the major movement toward a target after the reaction time delay. Secondary movement time is the time taken to make small final adjustments while reaching the target.
Figure 2.25 shows the relationship of hand movement distance as a function of time. Fitts’s law describes the relationship between the speed of a control movement versus its difficulty (Fitts, 1954). The law can be used to predict a wide variety of user movements, including those involving surgical tools, computer mouse movements, and foot controls. Movement time increases proportionately with distance to a target and decreases with
Movement of right hand (inches)
16 15”
Pick up object
12
10”
8 5”
Return
4
0
0
0.20
0.40 0.60 Time (sec)
0.80
1.00
FIGURE 2.25 The times shown are for the movement of one hand from an initial position to some other position to pick up an object and to return to the original position. Reaction time delay is not included. The operation of a control panel switch involves this type of movement. The curves illustrate that long movements take proportionately less time in relation to length than short movements. (From Barnes, R. M., Motion and Time Study, John Wiley & Sons, New York, 1963 and Glazer, S. and Hammell, R., Physical Design of Electronic Systems, Prentice Hall, Englewood Cliffs, NJ, 1970. With permission.)
TAF-ER562X_WINGR-09-0401-C002.indd 55
11/8/10 11:40:37 PM
56
Handbook of Human Factors in Medical Device Design
larger target sizes. Fitts’s law implies that constant ratios of movement accuracy and movement distances result in constant movement times. It is beyond the scope of this chapter to describe the details. An excellent description of Fitts’s law and its practical applications for product design is found in Knight (1987).
2.3.2 PRINCIPLES OF MOTION ECONOMY Industrial engineers have completed large numbers of time and motion studies that suggest that the following principles may be used to increase the speed, accuracy, and ease of manual operations. Such tasks are not satisfying to most workers and may lead to errors as well as musculoskeletal disorders. See Chapter 16, “Hand Tool Design,” for more detailed guidance on design to accommodate human limitations in body movement and Chapter 12, “Workstations,” for guidance on the design of workspaces to include manual tasks. Guideline 2.14: Principles to Achieve Motion Economy The following principles, adapted from Barnes (1963), should be considered: 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16.
17. 18.
Avoid repetitive manual tasks. Normally, machines are better for repetitive tasks. Both hands should begin and complete their motion at the same instant. The hands should not be idle except during rest periods. Motions of the arms should be made simultaneously and in opposite and symmetrical directions. The motion sequence that uses the fewest steps is the best for performing a given task. Horizontal hand movements are faster than vertical. Hands should be relieved of all work that can be performed more advantageously by the feet or other parts of the body. Where possible, work should be held by jigs or vises so that hands are more free to operate. Tools, materials, and controls should be located in an arc around the workplace and as near the worker as possible. Tools and materials should be pre-positioned to eliminate searching and selecting. Two or more tools should be combined whenever possible. The height of the workplace and the chair should preferably be arranged so that alternate sitting and standing at work is easily possible. Continuous, curved motions are preferable to straight-line motions involving sudden and sharp changes in direction. Ballistic movements are faster, easier, and more accurate than restricted or controlled movements. Rhythm is essential to the smooth and automatic performance of an operation, and the work should be arranged to permit easy and natural rhythm whenever possible. Successive movements should be so interrelated that one movement passes easily into the next, each ending in a position favorable for the beginning of the next movement. A movement is less fatiguing if it occurs in the direction that takes the greatest possible advantage of gravity. When a forcible stroke is required, the movements and the material of the worker must be arranged so that when the stroke is delivered, it has reached its greatest momentum.
TAF-ER562X_WINGR-09-0401-C002.indd 56
11/8/10 11:40:37 PM
Basic Human Abilities
57
19. Momentum should be reduced to a minimum if it must be overcome by muscular effort. 20. Hesitation or the temporary and often minute cessation of motion should be analyzed; its cause should be accounted for and, if possible, eliminated. 21. If a specific combination of movements has been determined as most suitable, emphasize form rather than accuracy even if this results in poor performance at the beginning of the learning period. 22. Arm movements that are mostly a pivoting of the elbow with small shoulder and upperarm action are faster and more accurate than those with a greater amount of shoulder and upper-arm action. 23. Manual limb movements terminated by mechanical devices take shorter periods of time compared to movements terminated solely by visual cues. 24. Single-hand visual positioning movements are faster and more accurate for short distances on a line 60 degrees from straight ahead on the same side of the body. 25. Two-handed visual positioning movements are most accurate straight ahead and fastest 30 degrees right or left of straight ahead. 26. In blind positioning movements, humans tend to undershoot long distances and overshoot short distances. Straight-ahead movements tend to be the most accurate. 27. Continuous movements in a horizontal plane are more accurate in certain angular directions from the midline of the body. For example, if 0 degrees is straight ahead, the most accurate movements for right-handed people would be 45 and 225 degrees and for lefthanded people 135 and 315 degrees. 28. Tremor or small vibrations of parts of the body degrade precision work and can be controlled by providing a visual reference, providing support of the body in general and any body part in particular, providing placement of the hand within 8 inches (20.3 cm) above or below the heart, or providing a limited amount of mechanical friction to absorb vibration energy.
2.3.3 SPEECH ATTRIBUTES 2.3.3.1 Loudness Levels of Speech This section describes some quantitative aspects of the human speech response and production capabilities. These attributes will become more important in device design as technology advances to allow more reliable and robust speech recognition systems. Table 2.17 shows average talker speech output in terms of sound pressure levels in dB (Morgan, 1963). Table 2.18 shows the distribution of talker loudness levels (Fletcher, 1953).
TABLE 2.17 Sound Pressure Levels of Speech 1 Meter from the Talker Normal Level (dB) Measure of Sound Pressure
Whisper (dB)
Minimum
Average
Maximum
Shout
Peak instantaneous pressure Speech peaks Long-time RMS pressures Speech minimum
70 58 46 30
79 67 55 39
89 79 65 49
99 87 75 59
110 98 86 70
TAF-ER562X_WINGR-09-0401-C002.indd 57
11/8/10 11:40:37 PM
58
Handbook of Human Factors in Medical Device Design
TABLE 2.18 Distributions of Speaking Volume for Persons Using the Telephone Volume Level Range (dB)a
Percentage of Speakers
<54 54–57 57–60
7 9 1 4 18 22 17 9 4 ~0
60–63 63–66 66–69 69–72 72–75 >75 a
Above sound pressure of 0.0002 millibars at a point 1 m from the talker’s lips.
2.3.3.2 Frequency Characteristics of Speech Because speech is a complex time varying quantity, its measurement is complex. Usually, speech is divided into a number of frequency bands and into a number of time segments (Figure 2.26). The average frequency is 128 Hz for males and 256 Hz for females. Most of the energy is below 1,000 Hz, with very little above 5 KHz (Morgan, 1963). Overall level (dB*) 90 Peak instantaneous 80 pressures
60
Speech-spectrum level (dB*)
50
A. Instantaneous pressure exceeded 1% of the B. RMS pressure time exceeded 10% of the time
40 30
Speech peaks 70 60
Long time RMS pressures Speech minima
20
50 C. RMS pressure exceeded 80% of the time
10 0 –10 100
200
*Re 0.0002 μ bar Male speech
500
1,000
2,000
5,000
10,000
Frequency (Hz)
FIGURE 2.26 The curves show how the intensity of male speech varies as a function of frequency and by different measurement criteria. At the upper right of the figure are corresponding overall levels for unanalyzed, unfiltered speech.
TAF-ER562X_WINGR-09-0401-C002.indd 58
11/8/10 11:40:37 PM
Basic Human Abilities
59
2.4 HUMAN VERSUS MACHINE CAPABILITIES One way to summarize the basic elements of this introduction to basic human skills and abilities is to compare the relative advantages and disadvantages of humans versus machines or mechanical systems (Table 2.19). For example, we know that some human limitations when compared to machines include limited strength, low endurance, slower processing speed, less accuracy, emotionally impaired decision making, and severe STM and LTM limitations. On the other hand, machines do not fatigue, are faster, are more accurate, can more easily do parallel processing, and are much better at repetitive tasks. Many trade-offs
TABLE 2.19 Human versus Machine Capabilities Humans Limitations
Advantages
Force—Limited strength Endurance—Fatigues easily Speed—Significant time needed for decision making and movement time. Accuracy—Unreliable; makes constant and variable errors. Computing—Slow and error prone. Decision making—Best strategy not always adapted; emotions interfere. Information processing—Basically a single-channel processor, which is easily overloaded; performance greatly dependent on motivation. Limited short-term working memory; long-term memory, although large, has unreliable and slow access.
Visual acuity and range are very good. Visual information processing system extremely logical and flexible. Range of detection is extremely wide with good sensitivity for audition and vision. Perception—Ability to make order out of complex situations; detection possible under high noise. Can reason inductively; can follow up intuition. Very flexible; can easily change rules of operation with changes in situation. Attention is easily shifted; only essential information is selected for processing. When highly motivated can perform under adverse conditions with parts out of order (injuries).
Machines Limitations Decision making is limited. Inductive reasoning not possible. Must be monitored. All activities must be thoroughly planned and preprogrammed thoroughly. Must get careful maintenance. May not operate at all if some parts are broken.
TAF-ER562X_WINGR-09-0401-C002.indd 59
Advantages Great forces are possible. Does not fatigue easily. High speed. Great accuracy attainable. Large short-term working memory. For narrow applications long-term memory is superior. Complex problems can be handled deductively. Excellent for repetitive work; unaffected by emotions and motivational needs. Can perform simultaneous operations easily.
11/8/10 11:40:38 PM
60
Handbook of Human Factors in Medical Device Design
are required in assigning tasks to humans versus machines. The designer must fully understand the intended tasks, users, and use environment to be able to effectively select and design device functional attributes. The remainder of this book provides detailed guidelines on specific design topics.
RESOURCES Cobb, P. W. and Moss, F. K. (1928). The four variables in the visual threshold. Franklin Institute Journal, 205, 831. Hill, N. E. G. (1947). The recognition of colored light signals which are near the limit of visibility. Proceedings of the Physiological Society, 59, 560. Kahneman, D., Slovic, P., and Tversky, A. (1982). Judgment Under Uncertainty: Heuristics and Biases. Cambridge, MA: Cambridge University Press. Kling, J. W. and Riggs, L. A. (1971). Experimental Psychology. New York: Holt, Rinehart and Winston. McFarling, L. H. and Heimstra, N. W. (1975). Pacing, product complexity, and task perception in simulated inspection. Human Factors, 17(4), 361–67. Moray, N. (1972). Listening and Attention. Baltimore: Penguin. Stevens, S. S. (1947). Sensitivity to changes in the intensity of white noise and its relations to masking and loudness. Journal of the Acoustical Society of America, 19, 609. Stevens, S. S. (1951). Handbook of Experimental Psychology. New York: John Wiley & Sons. Stroh, C. M. (1971). Vigilance: The Problem of Sustained Attention. Oxford: Pergamon.
REFERENCES American Standards Association. (1954). The Relations of Hearing Loss to Noise Exposure. New York: American Standards Association. Barnes, R. M. (1960). Motion and Time Study. New York: John Wiley & Sons. Bevan, W., Maier, R. A., and Helson, H. (1963). The influence of context upon the estimation of number. American Journal of Psychology, 76, 464–69. Brebner, J. T. and Welford, A. T. (1980). Introduction: An historical background sketch. In A. T. Welford (Ed.), Reaction Times (pp. 1–23). New York: Academic Press. Brown, J. S. and Slater-Hammel, A. T. (1949). Discrete movements in the horizontal plane as a function of their length and direction. Journal of Experimental Psychology, 39, 84–95. Burg, A. (1966). Visual acuity as measured by static and dynamic tests: A comparative evaluation. Journal of Applied Psychology, 50, 460–66. Carlson, N. R. (2000). Foundations of Physiological Psychology (5th ed.). New York: Allyn & Bacon. Chapanis, A. and Halsey, R. (1956). Absolute judgments of spectrum colors. Journal of Psychology, 42, 99. Cleland, T. M. (2004). A Practical Description of the Munsell Color System and Suggestions for Its Use 1937. Whitefish, MT: Kessinger Publishing. Conover, D. W. and Kraft, C. L. (1958, October). The Use of Color in Coding Displays (USAF WADC-TR-55-471). Dayton, OH: Wright-Patterson Air Force Base. Cornsweet, T. N. (1970). Visual Perception. New York: Academic Press. Corso, J. F. (1967). The Experimental Psychology of Sensory Behavior. New York: Holt, Rinehart and Winston. Dreyfuss, H. (1966). Measure of Man (2nd ed.). New York: Watson Cuptill. Drury, C. G. (1975). Inspection of sheet materials—Model and data. Human Factors, 17(3), 257–65.
TAF-ER562X_WINGR-09-0401-C002.indd 60
11/8/10 11:40:38 PM
Basic Human Abilities
61
Eastman Kodak Company. (1989). Ergonomic Design for People at Work: Vol. II. The Design of Jobs, Including Work Patterns, Hours of Work, Manual Materials Handling Tasks, Methods to Evaluate Job Demands, and the Physiological Basis of Work. New York: Van Nostrand Reinhold. Festinger, L. (1957). Theories of Cognitive Dissonance. Palo Alto, CA: Stanford University Press. Fitts, P. M. (1954). The information capacity of the human motor system in controlling the amplitude of movement. Journal of Experimental Psychology 47:381–91. Fletcher, H. and Munson, W. A. (1933). Loudness: Its definition, measurement, and calculation. Journal of the Acoustical Society of America, 5, 82–108. Foley, P. and Moray, N. (1987). Sensation, perception and systems design. In G. Salvendy (Ed.), Handbook of Human Factors. New York: John Wiley & Sons. Fraisse, P. (1963). The Psychology of Time. New York: Harper & Row. Geldard, F. A. (1972). The Human Senses. New York: John Wiley & Sons. Glazer, S. and Hammell, R. (1970). Man-machine interaction. Physical Design of Electronic Systems: Vol. 1. Englewood Cliffs, NJ: Prentice Hall. Harding G. F. and Jeavons, P. M. (1995). Photosensitive Epilepsy. Cambridge, UK: Mac Keith Press. Hatayama, T. and Tada, H. (1972). The experimental study of speed perception of the car on the road III: Speed judgements of drivers in the night time. Tohoku Psychologia Folia, 31(1), 4. Israelski, E. W. (1978). Commonplace human factors problems experienced by the colorblind: A pilot questionnaire survey. Proceedings of the 22nd Annual Meeting of the Human Factors Society, 347–51. Human Factors and Ergonomics Society, Santa Monica, CA. Kahneman, D. and Tversky, A. (Eds.) (2000). Choices, Values and Frames. New York: Cambridge University Press. Kaufman, L. and Rock, I. (1972). The moon illusion. Scientific American, July. Knight, J. L. (1987). Manual control and tracking. In G. Salvendy (Ed.), Handbook of Human Factors. New York: John Wiley & Sons. Kyllonen, P. C. and Alluisi, E. H. (1987). Learning and forgetting: Facts and skills. In G. Salvendy (Ed.), Handbook of Human Factors. New York: John Wiley & Sons. Lindsay, P., II. and Norman, D. A. (1973). Human Information Processing. New York: Academic Press. Mackworth, J. F. and Mackworth, N. H. (1958). Eye fixations recorded on changing visual scenes by television eye-marker. Journal of Optical Society of America, 52, 713–16. Malle B. F. (2001). Intentions and intentionality. In L. J. Moses and D. A. Baldwin (Eds.), Foundations of Social Cognition. Cambridge, MA: MIT Press. McCormick, E. J. (1970). Human Factors Engineering. New York: McGraw-Hill. Miller, G. A. (1956). The magic number seven, plus or minus two: Some limits on our capacity for processing information. Psychological Review, 63, 81–97. Miller, G. A. (1947). Sensitivity to changes in the intensity of white noise and its relations to masking and loudness. Journal of the Acoustical Society of America, 19, 609. Morgan, C. T. (Ed.) (1963). Human Engineering Guide to Equipment Design. New York: McGraw Hill. Norman, D. A. (1969). Memory and Attention. New York: John Wiley & Sons. Nutting, P. G. (1916). Effects of brightness and contrast in vision. Transcriptions of the Illuminating Engineering Society, 11, 939. Peterson, A. P. G. and Gross, E. E., Jr. (1967). Handbook of Noise Measurement. New Concord, MA: General Radio Corp. Pierce, J. R. and Karlin, J. E. (1957). Reading rates and the information rate of a human channel. Bell System Technical Journal, 36, 497–516. Richards, W. (1973). Factors Affecting Depth Perception: Final Report (Government Report AD-75926 1). Washington, DC: U.S. Government Printing Office. Riesz, R. R. (1928). Differential intensity sensitivity of the ear for pure tones. Physiological Review, 31, 867.
TAF-ER562X_WINGR-09-0401-C002.indd 61
11/8/10 11:40:38 PM
62
Handbook of Human Factors in Medical Device Design
Runeson, S. (1974). Constant velocity: Not perceived as such. Psychological Research, 37, 3–23. Salvendy, G. (1997). Handbook of Human Factors and Ergonomics. New York: John Wiley & Sons. Shower, E. G. and Biddulph, R. (1931). Differential pitch sensitivity of the ear. Journal of the Acoustical Society of America, 3, 275. Stevens, S. S. (1955). The Measurement of loudness. Journal of the Acoustical Society of America, 27, 815. Stevens, S. S. and Vollunan, J. (1940). The relation of pitch to frequency: A revised scale. American Journal of Psychology, 53, 329. Swets, J. A. (1964). Signal Detection and Recognition by Human Observers. New York: John Wiley & Sons. Van Cott, H. and Kinkaide, R. (1972). Human Engineering Guide to Equipment Design. Washington DC: U.S. Government Printing Office. Weinger, M. B. and Englund, C. E. (1990). Ergonomic and human-factors affecting anesthetic vigilance and monitoring performance in the operating-room environment. Anesthesiology, 73, 995–1021. Weintrab, D. J. and Virsu, V. (1972). Estimating the vertex of converging lines: Angle misperception. Perception and Psychophysics, 11, 277–83. Wickens, D. C. and Carswell, C. M. (1997). Information processing. In G. Salvendy (Ed.), Handbook of Human Factors and Ergonomics. New York: John Wiley & Sons. Woodson, W. E. and Conover D. W. (1964). Human Engineering Guide. Berkeley, CA: University of California Press.
TAF-ER562X_WINGR-09-0401-C002.indd 62
11/8/10 11:40:38 PM
3 Environment of Use Pascale Carayon, PhD; Ben-Tzion Karsh, PhD; Carla J. Alvarado, PhD; Matthew B. Weinger, MD; Michael Wiklund, MS, CHFP CONTENTS 3.1 Space and Physical Constraints .................................................................................65 3.1.1 General Principles...........................................................................................65 3.1.2 Design Guidelines ...........................................................................................66 3.1.2.1 Device Form and Configuration .......................................................67 3.1.2.2 Security and Privacy .........................................................................69 3.1.2.3 Device Labeling................................................................................69 3.2 Lighting ....................................................................................................................70 3.2.1 General Principles...........................................................................................70 3.2.2 Design Guidelines ...........................................................................................71 3.2.2.1 Illumination ......................................................................................71 3.2.2.2 Special Medical Environments and Applications .............................73 3.3 Noise ..........................................................................................................................74 3.3.1 General Principles...........................................................................................74 3.3.2 Design Guidelines ...........................................................................................76 3.3.2.1 Critical Communications ..................................................................79 3.4 Climate: Thermal Environment .................................................................................80 3.4.1 General Principles...........................................................................................80 3.4.2 Design Guidelines ...........................................................................................82 3.5 Climate: Humidity .....................................................................................................82 3.5.1 General Principles...........................................................................................82 3.5.2 Design Guidelines ...........................................................................................83 3.6 Climate: Airflow and Pressure...................................................................................83 3.6.1 General Principles...........................................................................................83 3.6.2 Design Guidelines ...........................................................................................84 3.7 Vibration ....................................................................................................................85 3.7.1 General Principles...........................................................................................85 3.7.2 Design Guidelines ...........................................................................................86 3.8 Energy Sources ..........................................................................................................87 3.8.1 General Principles...........................................................................................88 3.8.2 Design Guidelines ...........................................................................................88
63
TAF-ER562X_WINGR-09-0401-C003.indd 63
11/8/10 11:41:07 PM
64
Handbook of Human Factors in Medical Device Design
3.9 Design for Infection Control ......................................................................................89 3.9.1 General Principles...........................................................................................89 3.9.2 Design Guidelines ...........................................................................................90 3.10 Case Studies ...............................................................................................................91 3.10.1 Cleaning of Flexible Gastroscopes .................................................................91 3.10.2 Patient Use Insulin Infusion Pump .................................................................92 Resources ...........................................................................................................................93 References ..........................................................................................................................93 This chapter describes the environmental issues that need to be considered in the process of designing effective and safe medical devices. It is important to recognize that the physical environment is only one element of the work system in which devices are used (Vincent, Taylor-Adams, and Stanhope, 1998; Weinger and Englund, 1990; Wiklund, 1995). According to Smith and Carayon (Carayon and Smith, 2000; Smith and Carayon-Sainfort, 1989), the work system is comprised of five elements: (1) the individual (end user), (2) tasks, (3) tools and technologies (including the medical devices), (4) physical environment, and (5) organizational conditions. In this chapter, we focus on the effects of the physical environment on end users and the interaction between the physical environment and a particular type of technology: medical devices. The ergonomic aspects of the environment include factors such as space and physical constraints, lighting, noise, climate, vibration, and electromagnetic radiation (Parsons, 2000). In this chapter, we examine each of these factors individually, addressing each environmental characteristic in the context of medical device use. People interact continuously with their environment in a dynamic manner, experiencing the environment as a whole (Parsons, 2000), and, therefore, typically do not respond to a single environmental factor in isolation. Moreover, various environmental factors can interact to produce synergistic or compound effects on human performance. Thus, medical device designers must consider the entire integrated physical environment. Both the main effects of environmental factors and their interactions can affect end users and their ability to use medical devices safely and effectively (Figure 3.1). Medical devices should be designed to be both easy to use and effective (i.e., do what they are supposed to do) in the context of expected use. This means that medical devices must be designed with consideration of the environment of use, the people who will use them, the other devices likely to be used at the same time, and the way in which the device will be used. Medical device designers must study the anticipated environments of use and test devices within them if they are to design devices that will succeed in those environments (Bruckart, Licina, and Quattlebaum, 1993; International Electrotechnical Commission [IEC], 2004; Wickens et al., 2004). This chapter covers medical devices and medical device use environments in the United States. Unusual or special environments, such as mobile (e.g., ambulance, helicopter; see Chapter 17, “Mobile Medical Devices”), public or outdoor (e.g., playgrounds, roadways), or home (see Chapter 17 and Chapter 18, “Home Health Care”) environments, require special design considerations. However, recommendations herein may be applicable to many health care environments and should be used as appropriate. This guidance is not intended to restrict medical device innovation or improvements in current design or use techniques, noting that device design is regulated and advised by many federal agencies (e.g., the Food and Drug Administration [FDA]) and national and international standards).
TAF-ER562X_WINGR-09-0401-C003.indd 64
11/8/10 11:41:07 PM
Environment of Use
65
FIGURE 3.1 The complex environment of hospital operating rooms (shown) or intensive care units provides substantial challenges for the medical device designer who must consider myriad factors, including lighting, temperature, noise levels, and the impact of other devices. (Photo courtesy of Frank Painter.)
3.1 SPACE AND PHYSICAL CONSTRAINTS The goal of human factors is to design systems that reduce human error, increase productivity, and enhance safety and comfort (Wickens et al., 2004). Designing ergonomic workspaces is one of the major ways to improve the fit between humans, medical devices, and the patient care physical space. Health care presents many challenging venues and physical design constraints for the device designer. Hospitals and clinics are forever adding patient care equipment, and on-site storage is often limited or not considered at the time of purchase. The designer needs to consider not only how the device will be used in the primary use environment but also how the device will interact with people and other devices in the available space and where the device will be stored when not in use. Additionally, devices designed for clinical arenas are often difficult to modify for mobile or home care, where use and storage space may be far more limited.
3.1.1 GENERAL PRINCIPLES The physical space in which a device will be used must be sufficiently large to accommodate the functions, people, and devices for which it is intended. In the case of health care, if the space is too constrained, not only could users be uncomfortable and device use impaired, but there might also be a greater risk of use errors resulting in harm to patients and caregivers. Devices that cannot be placed in a convenient location within a work area are likely to be used less effectively (and may even be used less often). Therefore, designing devices for use in a specific environment must consider the space available and the associated reach, line of sight, and related physical use requirements. Line of sight (i.e., ability to see a critical device attribute from a typical use position), access (i.e., ability to reach and manipulate the device), and clearance (i.e., space between use elements such as adjacent
TAF-ER562X_WINGR-09-0401-C003.indd 65
11/8/10 11:41:07 PM
66
Handbook of Human Factors in Medical Device Design
FIGURE 3.2 Helicopter interior showing substantial physical space limitations. (Photo courtesy of Wiklund Research and Design.)
controls) must consider the entire use environment. For example, if a device is likely to be placed atop an existing workstation (e.g., an anesthesia monitor placed on top of an anesthesia gas machine), then its design should accommodate the visual field of view and reach of both sitting and standing users. In some use environments, the presence of caregivers, the patient, and medical devices leaves little room for the addition of new devices without compromising clinical care. For example, in the transport care environment (e.g., ambulances and helicopters; see Figure 3.2), space is extremely limited, and clearance challenges can be substantial. Designers also need to consider typical task workflow and other environmental factors. For example, devices that require power cords or other connections may present a significant tripping hazard in a congested use environment occupied by fast-moving workers (e.g., a busy emergency room). Trip and fall hazards are relatively common in health care for both patients and workers. People could, for example, fall and sustain injuries by slipping on wet surfaces created by the device, tripping over cords or tubing associated with the device, or tripping over the device itself. If not disposable, the medical device may stay in the use environment for prolonged periods. Power supplies, device cleaning, and maintenance should be considered during design if the device is likely to reside in a patient’s room, for example, for many days.
3.1.2 DESIGN GUIDELINES The following specific design guidelines apply to physical and space considerations. Guideline 3.1: Device Visibility Medical devices should be visible in their expected use environments, which might range from outdoor locations to operating rooms. For example, a device commonly placed on top of a patient covered by medium-blue sheets should be a contrasting color to make the device easier to find. Similarly, devices frequently exposed to blood should not be colored red.
TAF-ER562X_WINGR-09-0401-C003.indd 66
11/8/10 11:41:08 PM
Environment of Use
67
Guideline 3.2: Adequate Space for Device Use The physical space in which a device will be used should be sufficiently large to accommodate the functions, people, and devices for which it is intended.
Guideline 3.3: Space Considerations in Device Design Device designers should consider the space available and the associated reach, line of sight, and related physical use requirements.
Guideline 3.4: Reach Requirements for Smallest Users Reach requirements of the device and environment should not exceed the reach of the smallest users as they operate a hand-operated device and/or activate a foot control (Wickens et al., 2004) (see Chapter 4, “Anthropometry and Biomechanics”).
Guideline 3.5: Line of Sight and Device Positioning Device displays should be readily seen and easily read by the users (see Chapter 8, “Visual Displays”) when placed in the expected range of locations and positions in the intended use environment. This requires proper positioning and a clear line of sight with respect to the device and other equipment in the area.
Guideline 3.6: Critical Controls Accessible Critical controls should be accessible and not placed in tandem with or too close to other critical controls or devices that may be activated/deactivated inadvertently (see Chapter 7, “Controls”).
Guideline 3.7: Device Clearances Device clearances should accommodate not only routine and nonroutine use but also cleaning and maintenance requirements. When designing clearances, the expected positioning of the device in the use environment and associated equipment and devices should be considered.
Guideline 3.8: Slipping and Tripping Hazards Designers should identify and minimize or eliminate slipping and tripping hazards associated with device use.
3.1.2.1 Device Form and Configuration A medical device’s form in a particular use environment can have a strong influence on its safety and usability. For example, an ultrasound scanner that works well in a hospital’s surgical unit might not be able to be on a helicopter ambulance because of its contextually large size and weight. In contrast, a new-generation handheld ultrasound scanner, which is light and compact, could easily fit in a helicopter. Similarly, medical devices that have a single, inflexible configuration may function well in some use environments but not in others, where configurable devices would have a distinct advantage. Guideline 3.9: Compactness for Use in Congested Areas Medical devices intended for use in congested care environments, such as intensive care units, should be as compact as possible to enable effective personnel movement and provide room for other essential equipment (see Chapter 12, “Workstations”).
TAF-ER562X_WINGR-09-0401-C003.indd 67
11/8/10 11:41:09 PM
68
Handbook of Human Factors in Medical Device Design
FIGURE 3.3 Anesthesia workstations may be used on either the patient’s right or left side to accommodate clinician preference, room configurations, and surgical procedures. (From Music Therapy World, mmmagazine and University of Michigan Anesthesia.)
Guideline 3.10: Self-Contained Devices Medical devices should be as self-contained as possible rather than spreading out unnecessarily in ways that interfere with effective personnel movement.
Guideline 3.11: Flexible Placement Within Use Environment Medical devices should enable various placements within the care environment. For example, anesthesia workstations should be equally usable when they are placed on the right or left side of a caregiver positioned in front of the patient’s head (see Figure 3.3).
Guideline 3.12: Cable, Tube, and Wire Management Medical devices should incorporate the means to organize cables, tubes, and wires so they do not become tangled or confused with one another or with those associated with other devices.
Guideline 3.13: Device Use In Small Workspaces When necessary, medical devices should facilitate use in small workspaces. Environments such as a helicopter ambulance might limit users to reaching controls placed on a front panel, precluding access to a back panel (see Figure 3.4). In such cases, placing a device’s power switch on the back panel would pose a major usability problem.
FIGURE 3.4 Ventilator mounting scheme (see right-hand wall) provides access only to the device’s front panel in this helicopter ambulance. (Photos courtesy of Wiklund Research and Design.)
TAF-ER562X_WINGR-09-0401-C003.indd 68
11/8/10 11:41:09 PM
Environment of Use
69
Guideline 3.14: Damage Resistance Medical devices used in rescue operations should be able to withstand frequent impacts and rough handling by personnel focused on a given emergency rather than handling equipment carefully. Therefore, a patient monitor used in an ambulance or during patient transfers might be equipped with an oversized handle (encouraging proper carrying) and shock-absorbing bumpers.
Guideline 3.15: Stable Platform to Prevent Tipping Medical devices should be sufficiently stable to withstand the likely types of impacts in the intended use environments. For example, a cart-mounted electrocardiograph should not be subject to tipping over when quickly pushed out of the way in an emergency or struck by a hospital bed being rolled down the hallway.
3.1.2.2 Security and Privacy Medical devices should be protected against accidental actuations such as might occur when bumping into a control panel. Also, some medical devices, such as infusion pumps used to deliver narcotic drugs, might be subject to unauthorized use or even malicious tampering (i.e., for drug diversion), suggesting the need for protective mechanisms. Finally, the details of medical care are not a matter of public disclosure. Federal laws strictly regulate the access and release of medical information. Guideline 3.16: Protection Against Inadvertent Actuation Controls on medical devices used in congested environments should be protected against inadvertent actuation by moving equipment and personnel that might bump against them (see Chapter 17, “Mobile Medical Devices”).
Guideline 3.17: Tamper Resistance Medical devices used in uncontrolled or unsupervised environments should be protected against tampering, including unauthorized use. For example, electronic controls might require the user to enter a password to unlock them or use a physical key to open a cover.
Guideline 3.18: Protecting Patient Privacy Where practicable, medical devices should be designed to prevent bystanders (e.g., hospital visitors and other patients) from viewing confidential patient information appearing on displays. For example, a computer display might be equipped with an overlay (e.g., polarizing film) that prevents viewing from the side. However, such overlays should not impede reading by intended users from likely viewing angles.
3.1.2.3 Device Labeling It is important for clinicians to rapidly locate, identify, and take note of any necessary precautions associated with a given medical device. Guideline 3.19: Indicate Prohibited Uses Medical devices should indicate pertinent, prohibited uses. For example, a medical device containing ferrous metals that is likely to be used near an MRI scanner should warn users about exposing it to an intense magnetic field.
TAF-ER562X_WINGR-09-0401-C003.indd 69
11/8/10 11:41:11 PM
70
Handbook of Human Factors in Medical Device Design
Guideline 3.20: Bold Labeling Medical devices should be labeled boldly to facilitate rapid identification, particularly when the device will be used or stored among with many other devices and supplies (see Chapter 13, “Signs, Symbols, and Markings”).
Guideline 3.21: Warning Label Placement Warning labels should be placed where intended users are most likely to see them. In some cases, multiple labels with the same warning may be necessary to ensure reliable viewing in various use environments and use positions.
Guideline 3.22: Consistent Color Coding Color coding should be consistent with local and national conventions. For example, controls on medical air valves used in the United States should be colored yellow, while those used in several European countries should be checkered black and white.
3.2 LIGHTING Characteristics of lighting in the use environment that are relevant to device design include illumination, luminance, contrast, glare, and shadow (see also Chapter 2, “Basic Human Abilities,” and Chapter 8, “Visual Displays”). Illumination is the amount of light falling onto a surface, commonly measured in units of lux, an SI unit (one foot-candle equals 10.76 lux). Luminance is the light generated by a surface and is commonly measured in units of candelas per square meter (cd/m2), an SI unit (one foot-lambert equals 3.426 cd/m2). The ratio of the amount of light reflected by a surface (luminance) to the amount of light striking the surface (illuminance) is called the reflectance. The formula for reflectance is Reflectance = π × luminance (cd/m2)/illuminance (lux) The contrast or luminance ratio is the ratio of luminance of any two surfaces or areas in the visual field. If there are large differences in luminance in the environment, the eye must adapt to these different levels as it moves from one area of the environment to another. Large differences in luminance in the environment can be sources of glare and cause visual discomfort, annoyance, or decreased visual performance (Sanders and McCormick, 1993). Glare is the reflectance of bright light off a surface that reduces visual contrast and impairs visibility. Glare can be direct if it is caused by a bright area in the visual field or indirect if it is caused by light being reflected by a surface in the visual field. Glare depends on the range of luminances in the visual field, on the reflectance of surfaces in the visual field, and on the position of the light source relative to the line of sight of the user. As the glare source becomes closer to the line of sight, discomfort increases (Sanders and McCormick, 1993).
3.2.1 GENERAL PRINCIPLES A poorly designed lighting environment will provide excessive or insufficient light for the intended tasks or activities. The effects of lighting on performance depend on the visual stimulus, individual characteristics (e.g., age), visual demands of the task, and physical environment (Boyce, 1997). Lighting parameters mainly affect the visual and perceptual
TAF-ER562X_WINGR-09-0401-C003.indd 70
11/8/10 11:41:11 PM
Environment of Use
71
TABLE 3.1 Recommended Illumination Levels by Location Location Bed, observation Bed, examination Operating table (directed locally) Operating room where other work is being performed X-ray processing room
Illumination (lux) 400 1,000 50,000–100,000 400–500 50
Source: From The Chartered Institution of Building Services Engineers, Lighting Guide—Hospitals and Health Care Buildings (LG2: 1989 ed.), The Chartered Institution of Building Services Engineers, London, 1989. With permission.
aspects of a task. For instance, light produced by medical devices and other equipment (e.g., bright lighting in an operating room) can create direct glare. Other lighting attributes affecting visual performance include illumination, luminance contrast (e.g., contrast between an object and its background), and reflectance. Display luminance can affect visual search performance when viewing radiographs (Krupinski, Roehrig, and Furukawa, 1999). Lighting levels vary widely depending on the health care use environment and the associated medical procedure or therapy. The visual requirements of clinical tasks performed with medical devices tend to be very high because of the precision required. For example, illumination levels of 10,000 to 20,000 lux are recommended for surgical procedures (Sanders and McCormick, 1993). Brighter illumination can improve the depth of field and therefore increase visual acuity. On the other hand, high illumination increases the likelihood of glare and shadows and may overburden the visual system. Even in the same use environment, lighting levels may vary considerably depending on the specific task being performed. For example, while operating rooms are normally brightly lit by general room lights as well as special spotlights (i.e., operating room lights), they might be dimly lit to facilitate a minimally invasive surgical procedure, such as arthroscopy, during which procedural guidance is derived from images on large displays. Therefore, devices used in the operating room, such as a ventilator, must be designed for use in variable lighting conditions. Lighting conditions will be even more variable for medical devices designed for patients to use in homes, public spaces, and outdoors. Recommendations for illumination levels in various areas of hospitals and health care buildings can be found for the United Kingdom in the Lighting Guide of The Chartered Institution of Building Services Engineers (1989). For North America, illumination recommendations can be found in the IES Lighting Handbook (Illuminating Engineering Society of North America, 1981) (see Tables 3.1 and 3.2).
3.2.2 DESIGN GUIDELINES 3.2.2.1 Illumination Designers should determine the visual demands of the tasks performed with the medical devices, such as the need for precision work and the characteristics of the visual environment. This information should be used to determine the level of illumination that
TAF-ER562X_WINGR-09-0401-C003.indd 71
11/8/10 11:41:11 PM
72
Handbook of Human Factors in Medical Device Design
TABLE 3.2 Recommended Illumination Levels by Visual Task Demands Visual Demands
Illumination (lux)
Visual demands are not high Performance of visual tasks—high-contrast items or large size Performance of visual tasks—medium-contrast items or small size Performance of visual tasks—low-contrast items or very small size Performance of exacting visual tasks (e.g., surgery)
100 300 500 1,000 3,000
Source: From S. N. Chengalur, S. J. Rodgers, and T. E. Bernard, Kodak’s Ergonomic Design for People at Work (2nd ed.), John Wiley & Sons, Hoboken, NJ, 2004. With permission.
should be provided for the tasks and thereby dictate the design of displays, controls, and device labeling. Guideline 3.23: Intended Use Environment Illumination The illumination levels necessary to safely and effectively use a device should be determined by knowledge of the intended use environments and user tasks.
Guideline 3.24: Adjustable Lighting Levels Light sources (e.g., a display backlight) should be adjustable (including the option to turn them off) to facilitate use in a variety of expected use environments.
Guideline 3.25: Luminance Ratios Visual performance is clearly affected not only by the type of display equipment (e.g., viewing box versus cathode ray tube monitor) but also by its luminance characteristics (Krupinski et al., 1999). The luminance ratios of device displays should be as follows: about 1 between tasks and adjacent darker surroundings, 0.33 between tasks and adjacent lighter surroundings, and 10 between tasks and more remote surfaces (Chengalur, Rodgers, and Bernard, 2004).
Guideline 3.26: Display Reflectance to Prevent Glare The reflectance of the device display should be such that it produces as little glare as possible. In general, surfaces should diffuse light (e.g., matte finishes) (Chengalur et al., 2004). If the medical device produces light, the light source should be shielded and not produce direct glare.
Guideline 3.27: Illuminated Displays and Controls Medical devices used in low lighting conditions (e.g., outdoors at night or in a diagnostic procedure room with its lights dimmed to facilitate minimally invasive procedures) should be illuminated as required to ensure proper device operation (see Figure 3.5). For example, keyboards and controls can be spotlighted, while displays can be backlighted.
Guideline 3.28: Attentuation of Transmitted Light If a device incorporates see-through materials (glass/Plexiglas between the user and what he or she is to view, then the design should consider the attenuation of transmitted light.
TAF-ER562X_WINGR-09-0401-C003.indd 72
11/8/10 11:41:11 PM
Environment of Use
73
FIGURE 3.5 Medic takes a soldier’s vital signs while traveling in a vibrating and dimly lit military transport aircraft. (Courtesy of Air Force Photo Gallery, image 081107-F-062E-477.) For instance, under normal lighting conditions, a reduction of 6% to 16% of the illumination was observed during the illumination’s passage through the walls or roof of infant incubators (Sjors et al., 1992). Incubators with double walls produce a greater decrease in illumination than incubators with single walls.
Guideline 3.29: Display Legibility in Bright Light As required, displays should remain legible when exposed to direct sunlight and intense artificial light sources (e.g., operating room lights).
Guideline 3.30: Display Cleanability Device displays should be designed so that they are easy to clean, thereby maintaining intended brightness (McCarthy and Brennan, 2003).
3.2.2.2 Special Medical Environments and Applications 3.2.2.2.1 Radiographic Imaging Environments Interpretation of medical/clinical images is dependent on image viewing conditions, including the lighting characteristics of the display equipment (e.g., viewing boxes and visual display). A range of guidelines has been developed that define appropriate radiological viewing conditions (see Table 3.3). 3.2.2.2.2 Endoscopic Devices The presence or absence of shadow can affect visual performance. The presence of shadows can provide depth cues that facilitate the reconstruction of three-dimensional images. Hanna, Cresswell, and Cuschieri (2002) demonstrated that the use of a shadow-inducing system facilitated mental processing of an endoscopic image reconstructed as a threedimensional picture of the operative field. On the other hand, the presence of shadows may affect visual performance as well as working postures (e.g., the surgeon has to adopt
TAF-ER562X_WINGR-09-0401-C003.indd 73
11/8/10 11:41:11 PM
74
Handbook of Human Factors in Medical Device Design
TABLE 3.3 Guidelines on Radiological Image Viewing Conditions Source World Health Organization (1982) CEC (1996) CEC (1997)
Brightness of Viewing Box (cd/m2)
Uniformity of Viewing Box (%)a
Illumination (lux)
1,500–3,000 2,000–4,000 >1,700
<15 Not available <30
<100 Not available <50
Source: Adapted from E. McCarthy and P. C. Brennan, British Journal of Radiology, 76, 94–97, 2003. a Uniformity of viewing box is measured as the maximum deviation of brightness as a percentage of the sum of the maximum and minimum brightness levels (McCarthy and Brennan, 2003).
awkward postures to avoid shadows in the visual field). Instruments equipped with coaxial illumination provide illumination that is parallel to the line of sight and facilitate shadowfree images, and can significantly improve working postures (Chang, 2002). 3.2.2.2.3 Use of Goggles or Other Eye Gear Because of concerns about infection exposure, clinical personnel often wear face shields, protective glasses, or goggles. Some situations (e.g., laser procedures) require use of colored or other types of special glasses. When indicated, device displays should accommodate the resulting glare, parallax, and altered vision. The use of night-vision goggles allows the performance of various medical procedures, such as intubations at night under emergency field conditions (Gillis and Miles, 2001). However, providing a small light source that projects a focused beam of light may provide safer care than the use of night-vision goggles, for instance, in obtaining intravenous access (Schwartz and Charity, 2001). Guideline 3.31: Device Use with Goggles If the expectation is that a device will be used by people wearing goggles or protective eye gear, the design should accommodate such use without decreases in user performance.
Guideline 3.32: Displays Viewed through Goggles Displays that might be viewed by users wearing night-vision goggles (e.g., paramedics serving in helicopters at night) or laser protection goggles (during laser procedures) should be optimized for such use and provide readily accessible brightness controls.
3.3 NOISE 3.3.1 GENERAL PRINCIPLES Noise is a common attribute of medical device use environments and must be considered during device design (see also Chapter 2, “Basic Human Abilities,” and Chapter 10, “Alarm Design”). Noise can contribute to a number of undesirable outcomes in health care delivery settings. For example, noise can increase length of hospital stay (Fife and Rappaport, 1976) and contribute to patients’ sleep–wake abnormalities in the intensive care unit
TAF-ER562X_WINGR-09-0401-C003.indd 74
11/8/10 11:41:12 PM
Environment of Use
75
(Freedman et al., 2001). Noise can increase heart rate, subjective stress, and annoyance (Morrison et al., 2003), and possibly impair task performance, concentration, and complex problem-solving (Morrison et al., 2003). Depending on the task conditions and the type of sounds, noise may impair performances, have minimal impact, or can even improve performance (Sanders and McCormick, 1993). Difficult tasks that place high demands on perceptual-motor skills and/or information processing (e.g., team communication during high-risk emergency procedures) may be most affected by noise (Sanders and McCormick, 1993). Loud noise can impair reading comprehension. Even noise as low as 68 dBA (where dBA is decibels on the A-weighted logarithmic scale for sound loudness)—about the same volume as a person speaking normally from 3 to 5 feet away—can impair the retrieval of information from memory (Crocker, 1997). Noise also influences response bias; people tend to think an auditory signal is present only when they are absolutely sure they heard the signal. Loud noises can cause a startle response that disrupts perceptual-motor performance. However, with repeated exposure to noise, the magnitude of this startle response diminishes. Noise appears to have little impact on motor performance or visual functions such as acuity, contrast discrimination, dark vision, and accommodation (Sanders and McCormick, 1993). There are two important ways that noise can affect medical device use. First, noise can prevent health care personnel from hearing device-generated communications, information signals, or alarm sounds. That is, noise (from speech, devices, procedures, and so on) can mask sounds emitted from medical devices. Masking occurs when one aspect of the sound environment reduces the sensitivity of the ear to other aspects of the sound environment. For example, paramedics, emergency medical technicians, emergency physicians, and transport nurses could correctly identify 96% of breath sounds in a quiet environment but only 54% when inside an ambulance (Brown et al., 1997). Masking of device sounds by noise increases when device sounds are near the frequency of the masking tone and its harmonic overtones and as noise intensity increases (Sanders and McCormick, 1993). Complex sounds are more difficult to mask than simple sounds. For sound levels in clinical facilities, the World Health Organization and the U.S. Environmental Protection Agency recommend that hospital rooms should not be louder than 30 to 45 dBA (Shankar et al., 2001; World Health Organization, 1999). Table 3.4 shows noise levels (dBA) of common sound-producing events. The noise guidelines for newly constructed or renovated hospital nurseries recommended that continuous sound in occupied TABLE 3.4 Noise Levels Quality Just audible Very quiet Quiet Moderately loud Loud Very loud Uncomfortably loud
dBA
Source
Effect
10 20–30 50 70 90 100 120
Whisper Light traffic Vacuum cleaner Pneumatic drill Power mower Discotheque
<35 dbA desired for sleep <50 dbA desired for work Annoyance Hearing loss with chronic exposure Pain and distress
Source: Reproduced from W. E. Morrison, E. C. Haas, D. H. Shaffner, E. S. Garrett, and J. C. Fackler, Critical Care Medicine, 31, 113–119, 2003. With permission.
TAF-ER562X_WINGR-09-0401-C003.indd 75
11/8/10 11:41:12 PM
76
Handbook of Human Factors in Medical Device Design
TABLE 3.5 Noise Levels Measured in Hospital Environments Study
Setting
Bentley et al. (1977)
ICU
Redding et al. (1977) Meyer et al. (1994) Meyer-Falcke et al. (1994) McLaughlin et al. (1996)
ICU ICU SICU
Tsiou et al. (1998)
Six-bed ICU
Keipert (1985) Couper et al. (1994)
Peds ward Peds ward
CSICU
Noise Levels (dBA) Average 42–53 Peaks >70 Average 74 Peaks 83.6 Average >65 Peaks >110 Average 58–77 Peak 101 Average 60–67 Peak 83–90 Peaks 80–88 65
Source: Adapted from W. E. Morrison, E. C. Haas, D. H. Shaffner, E. S. Garrett, and J. C. Fackler, Critical Care Medicine, 31, 113–119, 2003. With permission. ICU, intensive care unit; SICU, surgical ICU; CSICU, cardiac surgical ICU; Peds, pediatrics.
bed spaces or patient care areas should not exceed an hourly equivalent continuous sound level (Leq) of 50 dBA (Philbin, Robertson, and Hall, 1999). But, as shown in Table 3.5, adapted from Morrison et al. (2003), noise averages from a variety of hospital studies regularly exceed these guidelines, and peaks can exceed 100 dBA, a level considered dangerous for hearing (Wickens et al., 2004). The second way that noise can affect medical device use is through its effects on performance and stress. Acute noise exposure causes a stress response, while chronic exposure can adversely affect health (Weinger and Englund 1990; Wickens et al., 2004). People have more difficulty hearing an auditory signal when they are simultaneously attending to or performing another activity (Sust and Lazarus, 2003); this is typical in health care, as clinicians are typically conducting a variety of patient care tasks while at the same time attending to medical devices’ auditory signals. Blood pressure values obtained by medical personnel in a quiet environment were significantly more accurate than those obtained in an ambulance, although the results may have been affected by the motion of the ambulance (Prasad et al., 1994). Table 3.6 provides examples of a variety of noise sources in hospitals. Medical devices are also used in transport vehicles such as ambulances, fixed-wing aircraft, and rotary-wing aircraft where noise levels can range from 65 to more than 99 dBA (Campbell et al., 1984; Macnab et al., 1995).
3.3.2 DESIGN GUIDELINES* A key design principle is that if a medical device auditory signal occurs in the presence of noise, the threshold of detectability of the signal will be increased; this auditory threshold *
See Chapter 10, “Alarm Design.”
TAF-ER562X_WINGR-09-0401-C003.indd 76
11/8/10 11:41:12 PM
Environment of Use
77
TABLE 3.6 Noise Levels Associated with Common Events and Activities in Hospitals Noise Sources Adult ICU item Connecting/disconnecting gas supplies Oximeter alarm Nebulizer Telephone ringing Ventilator alarm Monitor alarm Television (loud) Ventilator IV alarm Air conditioner Open suctions Open oxygen sources Dialysis alarm Patient transfer Moving bed Respirators
Sound Level (dBA)
88 81 80 80 79 79 79 78 77 75 70–82 70–77 63 60–66 58 49–72
Study
Tsiou et al. (1998) Kahn et al. (1998) Kahn et al. (1998) Kahn et al. (1998) Kahn et al. (1998) Kahn et al. (1998) Kahn et al. (1998) Kahn et al. (1998) Kahn et al. (1998) Kahn et al. (1998) Tsiou et al. (1998) Tsiou et al. (1998) Kam et al. (1994) Kam et al. (1994) Kam et al. (1994) Tsiou et al. (1998)
Pediatric ICU item Cleaning crew/equipment Ventilator alarms Infants crying Conversations Trauma phone Monitor alarms Overhead pages Neonatal incubators Medication pump alarms
≤96 ≤79 peak 78 ≤73 peak 73 62–74 59–84 58 55–56
PACU Patient crying ECG alarms Patient coughing Changing bed linen Staff normal conversation Opening plastic packets Cardiac monitors
80–86 75–78 70 56–66 56–60 55–72 44–78
Kam et al. (1994) Kam et al. (1994) Kam et al. (1994) Kam et al. (1994) Kam et al. (1994) Kam et al. (1994) Kam et al. (1994)
Operating room Suction in use Suction not in use Surgeon instructions Ventilator Ward areas Operating theaters Intensive care
75–80 73 66–72 65 58 57 57
Kam et al. (1994) Kam et al. (1994) Kam et al. (1994) Kam et al. (1994) Kam et al. (1994) Kam et al. (1994) Kam et al. (1994)
Morrison et al. (2003) Morrison et al. (2003) Morrison et al. (2003) Morrison et al. (2003) Morrison et al. (2003) Morrison et al. (2003) Morrison et al. (2003) Kam et al. (1994) Morrison et al. (2003)
continued
TAF-ER562X_WINGR-09-0401-C003.indd 77
11/8/10 11:41:12 PM
78
Handbook of Human Factors in Medical Device Design
TABLE 3.6 (CONTINUED) Noise Levels Associated with Common Events and Activities in Hospitals Noise Sources Other hospital areas Nonmedical care Corridors Clinics
Sound Level (dBA) 65 61 59
Study Bayo et al. (1995) Bayo et al. (1995) Bayo et al. (1995)
ICU, intensive care unit; IV, intravenous; PACU, postanesthesia care unit; ECG, electrocardiogram.
needs to be exceeded if the signal is to be detected (Sanders and McCormick, 1993). Effective design thus requires a thorough understanding of the noise levels in the environment of use. For example, one of the recommendations is for auditory signals to be at least 15 dB above the masking threshold. Guideline 3.33: Quiet Device Except when Signaling Medical devices should be as quiet as possible except when a certain amount of noise provides useful audible feedback to the user or patient. For example, portable oxygen generators should operate quietly so that they do not disturb patients, particularly when they are trying to sleep. Conversely, it might be best for a hospital bed to make a whirring noise when moving as a means to help protect against harm that could come from inadvertent bed motion.
Guideline 3.34: Auditory Signal Intensity above Masking Threshold The sound level of an auditory signal should generally be at least 15 dBA but not more than 25 dBA above the threshold. However, Sust and Lazarus (2003) suggest that levels as high as 30 dBA may sometimes be required to attract attention during other patient care activities. Avoid auditory alarms that are in the dangerous range for hearing (above 85 to 90 dBA) unless that is the only intensity that the auditory signal can be heard. In this latter case, the signal duration should be brief.
Guideline 3.35: Auditory Signal Frequency Auditory signals should have frequencies between 500 and 3,000 Hz because the ear is most sensitive to the middle ranges.
Guideline 3.36: Auditory Signal Duration Auditory signals should be at least 500 ms in duration (preferably between 500 and 1,000 ms). A shorter duration signal will not sound as loud, and the ear does not respond immediately to sound. If auditory signals must be shorter, their intensity should be increased.
Guideline 3.37: Get Users’ Attention, then Provide Useful Information When presenting complex information with auditory signals, use an attention-demanding signal, followed by a designation (or information-emitting) signal that provides the meaning of the signal.
TAF-ER562X_WINGR-09-0401-C003.indd 78
11/8/10 11:41:12 PM
Environment of Use
79
Guideline 3.38: Auditory Signal Consistency The same signal should always be used to convey the same information. Similarly, avoid conflict with signals that are already being used to convey other types of information. This may be especially important in noisy environments where some degree of masking is occurring and some signals may mask others.
Guideline 3.39: Consistency of Auditory Signals with Learned Meanings When appropriate, auditory signals should be compatible with natural relationships (i.e., consistent with human associations with other environmental and man-made sounds). For example, higher-frequency wailing signals are generally interpreted as indicating an emergency.
Guideline 3.40: Auditory Signal Usability Testing All auditory signals should undergo usability testing with representative users for detection and discrimination in the intended use environments.
3.3.2.1 Critical Communications When a medical device requires attention, perhaps because there is a problem or a consumable has run out, it must draw the user’s attention in a reliable manner. The most appropriate means of communicating a critical message will depend on the use environment. If task and environmental analyses indicate that the device will be used in a noisy environment, nonauditory signals should be incorporated to convey critical device information. Auditory displays are preferred over visual displays when the message is simple and short, will not be referred to later, deals with events in time, and calls for immediate action (Sanders and McCormick, 1993). Auditory displays may also be preferred when the visual system is overburdened, illumination limits vision, the user moves from place to place, or a verbal response is required. Overall, auditory alarms produce greater compliance than visual alarms (Wickens et al., 2004). But, considering the noise in most health care environments, especially transport vehicles, visual signals and displays should usually be incorporated into the device as a complement to the auditory signals. Fromm, Campbell, and Schlieter (1995) studied the ability of health care personnel to react to visual alarms in medical helicopters. Reaction times to visual alarms were very long and variable, with mean reaction times of 81 ± 78 seconds, with a range of 3 seconds to more than 5 minutes. Therefore, careful analysis followed by rigorous testing in potential use environments should guide design decisions in noisy environments. Guideline 3.41: Reliable Detection of Alarm Signals Where appropriate, alarms should be designed for reliable detection in visually congested and/or noisy environments. In some cases, this might call for redundant annunciation in multiple locations, such as in a patient’s room, adjacent corridor, and central workstation.
Guideline 3.42: Audible Alarms In visually congested environments, alarms should be annunciated by means of an audible signal (see Chapter 10, “Alarm Design”).
TAF-ER562X_WINGR-09-0401-C003.indd 79
11/8/10 11:41:12 PM
80
Handbook of Human Factors in Medical Device Design
Guideline 3.43: Visual Alarms In noisy environments, alarms should be annunciated by means of a visual signal, such as a flashing light (see Chapter 10, “Alarm Design”).
Guideline 3.44: Tactile Alarms In visually complex and noisy environments, alarms should be annunciated by means of a tactile signal, such as making a device worn on the body (e.g., pager) vibrate (see Chapter 10, “Alarm Design”).
Guideline 3.45: Voice Prompts Voice prompting should only be used to communicate critical information (i.e., instructions on how to operate an automated external defibrillator) when other information channels (e.g., vision and touch) are already devoted to sensing other stimuli or when a voice is most likely to produce the desired behavior (see Chapter 10, “Alarm Design”).
3.4 CLIMATE: THERMAL ENVIRONMENT Many medical devices, such as stretchers and portable ventilators, venture outside of environmentally controlled settings like hospitals and doctors offices. Clearly, such devices will be exposed to a wide range of climatic conditions, placing special demands on the equipment and users. The devices and users might also be subjected to physical stresses, such as intense vibration or jarring. Also, the device might be used in a potentially contagious environment. Although most clinical settings in the United States are highly controlled thermal/humidity environments, medical devices are often used in much less controlled environments, including outdoors. Even the most highly controlled environment, the operating room, can have temperatures ranging from 12°C (cardiac surgery) to 30°C (burn or neonatal surgery).
3.4.1 GENERAL PRINCIPLES Personal comfort varies with time of day, season, diet, health status, and clothing choices as well as job or task stress and cultural variables and expectations (Chengalur et al., 2004). Even under optimal comfort conditions, a small percentage (<3%) of the population is too warm or too cold (Fanger, 1970). Six main factors should be considered in device design and user responses to thermal environments: • • • • • •
Air temperature Radiant temperature Air velocity (affected by air-handling systems and fans) Humidity Level of user activity Users’ clothing
If the user is subjected to thermal stress, then the thermoregulatory system responds by changing its state to maintain core body temperature. This response has consequences for the user’s health, comfort, and working efficiency (Parsons, 2000). It is difficult to
TAF-ER562X_WINGR-09-0401-C003.indd 80
11/8/10 11:41:12 PM
Environment of Use
81
accurately predict the effects of thermal environments on specific patient care task performance because of the many variables involved. There are, however, numerous nonmedical studies on the effects of thermal stresses on human performance (Parsons, 1993). When humans are exposed to a hot environment, the effects on task performance will depend on individual levels of arousal and motivation as well as individual differences, such as degree of environmental acclimatization (Parsons, 2000). As heat stress increases, mental and physical performance will be affected adversely. Performance on sustained vigilance tasks, often required during patient care, can be impaired in slightly warm environments that produce soporific effects. For example, a nurse who is heavily garbed for patient isolation may be more likely to miss early signs of an evolving patient crisis if performing monotonous, repetitive monitoring tasks in a room warmed to a higher-than-normal temperature for the health of a burn patient who loses heat more readily. Cold stress is generally a less common health care occupational hazard than heat stress. However, providing emergency care outdoors can expose the device user to the cold and elements, including snow and ice (Figure 3.6). Additionally, because of the requirements of the Occupational Safety and Health Administration’s Bloodborne Pathogens Rule to wear moisture-resistant, personal protective barriers for surgical procedures, many operating rooms have lowered ambient room temperature to accommodate medical staff wearing this extra protective gear. Operating room nurses and anesthesia providers, who do not typically wear as much protective clothing, must cope with these cold temperatures. The two primary physiological reactions to cold stress, vasoconstriction and shivering, reduce net heat loss, thus maintaining core temperature (Sanders and McCormick, 1993). However, vasoconstriction can cause decreased tactile sensitivity, increasing the risk of use error when interacting with medical devices. Shivering can be severe enough to impair vision and the execution of even simple fine motor tasks, such as activating a control or holding a portable device. Additionally, clothing required to reduce heat loss, such as heavy gloves, can interfere with tactile sensation or the ability to grasp. Hats can cause hearing loss. Jackets can
FIGURE 3.6 Ski patrollers prepare to attach a splint in cold weather. The use of medical devices in such outdoor conditions will be affected not only by the temperature but also by the required protective gloves and clothing. (From http://www.nsw-ski-patrol.org.au/perisher/pics/treat.jpg. With permission.)
TAF-ER562X_WINGR-09-0401-C003.indd 81
11/8/10 11:41:12 PM
82
Handbook of Human Factors in Medical Device Design
reduce mobility and range of motion. Thus, user-interface designs need to account for users wearing protective clothing and gear as well as any physical or mental impairments that a user might be under in harsh conditions.
3.4.2 DESIGN GUIDELINES Guideline 3.46: Minimize Effect on Ambient Temperature Medical devices should not unduly affect the ambient temperature of likely use areas.
Guideline 3.47: Consider Users’ Clothing Requirements If the temperature in the use environment will necessitate unusual clothing requirements, the implications for device use should be incorporated into design and evaluation activities.
Guideline 3.48: Device Use Wearing Gloves Controls and handles should be designed to provide sufficient tactile feedback and control to persons wearing protective gloves in cold weather conditions.
Guideline 3.49: Use in Inclement Weather Medical devices intended for use outdoors should be usable in all possible weather conditions, such as extreme heat or cold, extremely low or high humidity, high wind, blowing sand, rain or snowfall, and so on.
Guideline 3.50: Protective Eyewear Information on displays that might be viewed by users wearing protective eyewear (e.g., goggles protecting the eyes from extreme cold) should be oversized to facilitate reading, particularly when the eyewear is frosted, fogged, smudged, or scratched.
Guideline 3.51: Handles and Grips Handles and equivalent gripping surfaces should continue to provide a secure grip when wet and icy. This goal might be achieved by allowing a closed grip around a handle (with the thumb crossing over the fingers), making it less likely to slip out of the hand.
3.5 CLIMATE: HUMIDITY 3.5.1 GENERAL PRINCIPLES Humidity is the amount of moisture in the air. Without active intervention, indoor humidity tracks the prevailing outdoor humidity. Outdoor humidity is usually lower during the winter months and higher in the summer. In the summer, air-conditioning systems reduce indoor humidity. Home care environments have wider fluctuations in humidity than clinical care environments. Home humidifiers are used in conjunction with home heating units or as stand-alone humidifiers. Hospitals and clinics no longer use such devices because they can introduce potential infectious agents (e.g., fungi and bacteria) into the environment. Low humidity can dry user’s mucous membranes, skin, and lips (Sanders and McCormick, 1993). Eye irritation occurs when the relative humidity is at or below 30%, particularly for
TAF-ER562X_WINGR-09-0401-C003.indd 82
11/8/10 11:41:13 PM
Environment of Use
83
those who wear contact lenses. This effect becomes pronounced after 4 hours (Rohles and Konz, 1987). Areas used for device cleaning and reprocessing are often warm and humid because of instrument reprocessing equipment; the effects on cleaning personnel’s performance are amplified by the required personal protective equipment (PPE) (i.e., gloves, goggles, and gowns). In addition to their thermal burden, PPE can interfere with job performance and jeopardize worker safety. Medical devices should be designed so that maintenance/reprocessing personnel can do their jobs safely and efficiently. An ASTM F23 Committee for Protective Clothing Human Factors Subcommittee is developing new standards for measuring the overall thermal comfort and breathability (i.e., ability to release moisture, perhaps while also preventing its ingress) of protective clothing. The standard will also specify consistent methods to evaluate the effectiveness of cooling devices used with protective clothing.
3.5.2 DESIGN GUIDELINES Guideline 3.52: Minimize Effect on Ambient Humidity As humidity increases, discomfort will be felt at the higher end of the thermal comfort zone (79°F/26°C). Device use should not increase the work area’s humidity above 55% for a person performing light work (Chengalur et al., 2004). If at all possible, the device should not contribute to ambient humidity. Humidity levels above the desired range can cause condensation, damaged equipment, or microbial contamination.
Guideline 3.53: Device Storage in High Humidity In areas of high humidity and/or temperature, attention should be given to device storage requirements and degradation of device components, such as adhesives and chemical reagents.
3.6 CLIMATE: AIRFLOW AND PRESSURE 3.6.1 GENERAL PRINCIPLES Many medical devices incorporate fans to maintain component stability, evacuate generated heat, or dispose of toxic substances (e.g., smoke). Devices, especially large devices such as whole-body imaging systems, can significantly influence not only temperature but also airflow and pressure within patient care environments. Designers of medical devices that could affect or be affected by the climate and air quality within their use environments need to be familiar with existing health care facility environmental standards and guidelines. The construction and environmental guidelines of the American Institute of Architects (2001) and the standards of the American Society of Heating, Refrigerating and Air-Conditioning Engineers (2003) provides minimum standards for design and construction of health care facility ventilation systems. The environment of care standards of The Joint Commission (formerly the Joint Commission on Accreditation of Healthcare Organizations [JCAHO]) require hospitals to install systems that maintain appropriate pressure relationships, air exchange rates, and filtration efficiencies systems to control airborne contaminants such as biological agents, gases, fumes, and dust (JCAHO, 2004).
TAF-ER562X_WINGR-09-0401-C003.indd 83
11/8/10 11:41:13 PM
84
Handbook of Human Factors in Medical Device Design
3.6.2 DESIGN GUIDELINES Guideline 3.54: Device Effects on Air Handling Systems Medical device designs should minimize all potential interactions between the device and the air-handling systems in the expected environments of use.
Guideline 3.55: Device Exhaust Direction Fans, such as those used to cool electrically powered components (see Figure 3.7), should not blow directly on medical personnel and patients.
Guideline 3.56: Device Exhaust Filtration If a device generates an exhaust, filtration of the exhaust should be incorporated into the design.
Guideline 3.57: No Airborne Contaminants Medical devices should not generate any airborne contaminants that exceed prescribed containment standards and guidelines.
Guideline 3.58: Minimal Effect on Negative-Pressure Rooms The exhaust from medical devices used in isolation environments should not be so forceful that it appreciably changes the direction of air flow in the room, which presumably is set to remove contaminated air through a low pressure inlet.
Guideline 3.59: Use in High- and Low-Pressure Environments Medical devices may be used at high altitude (i.e., subatmospheric) or at greater than atmospheric pressures (i.e., decompression chambers or in submarines). Devices to be used in decompression chambers should function correctly and be usable at pressures up to 6 atmospheres (600 kilopascals [kPa], or ∼85 psi [lbf/in2]). Subatmospheric pressures are attained at high-altitude medical facilities (e.g., Everest base camp medical clinic) and in medical aircraft
FIGURE 3.7 Ultrasound scanner’s cooling vents point toward the side and back, avoiding directing heated air toward the technician or patient.
TAF-ER562X_WINGR-09-0401-C003.indd 84
11/8/10 11:41:13 PM
Environment of Use
85
(cabin pressurized typically at 62 kPA) or spacecraft. For devices intended for use at very high altitude, the reduced gravity during use must also be considered.
3.7 VIBRATION 3.7.1 GENERAL PRINCIPLES Design guidelines for vibration include how to design for device use in a vibrating environment and how to reduce the vibration generated by a device. This section covers only the former (see also Chapter 17, “Mobile Medical Devices”), while Chapter 16, “Hand Tool Design,” covers the latter. Common medical device use scenarios with vibration include moving vehicles (ambulance and helicopter) and device use while moving (e.g., use of a patient monitor on a gurney while transporting the patient from one place to another in a hospital). Nonmedical devices (e.g., heating, ventilating, and air-conditioning systems) can induce vibration in otherwise stationary use environments if improperly designed or installed. Exposure to vibration can cause pain, shortness of breath, anxiety, and changes in blood pressure (Macnab et al., 1995). Vibration can impair visual or motor performance (Griffin, 1997; Sanders and McCormick, 1993; Wickens et al., 2004). In other words, both input and output processes can be affected by vibration. Table 3.7 shows the perceived comfort reported by healthy adults experiencing different levels of vibration. Visual performance is impaired by vibration because the continual movement of the image on the retina causes the image to appear blurry. Vibration can disrupt eye–hand coordination in the absence of effective stabilization. Thus, control input devices, such as touch screens, might be unreliable (Wickens et al., 2004). Additionally, medical device display vibration can reduce the ability to see fine detail. Performance can be impaired when using isometric controls (depending on the control gain and control order) (Griffin, 1997). Control performance in vibrating environments will be affected by control shape and orientation, task difficulty, and the axis of vibration (Griffin, 1997). Patient transport by ambulance, fixed-wing, or rotary-wing aircraft can produce substantial vibration (see Table 3.8). Vibration acceleration magnitude in transport vehicles varies from 0.4 to 5.6 m/s2, depending on the axis and the type of vehicle (Campbell et al., 1984). Rotary-wing aircraft (i.e., helicopters) tend to produce the greatest vibration, while TABLE 3.7 Vibration Comfort Zones for Healthy Adults Comfort Level Still comfortable A little uncomfortable Fairly uncomfortable Uncomfortable Very uncomfortable Extremely uncomfortable
Acceleration (m/s2) <0.31 0.31–0.63 0.5–1.0 0.8–1.6 1.25–2.5 >2.0
Source: Adapted from A. Macnab, Y. Chen, F. Gagnon, B. Bora, B., and C. Laszlo, Aviation Space and Environmental Medicine, 66, 212–219, 1995. With permission.
TAF-ER562X_WINGR-09-0401-C003.indd 85
11/8/10 11:41:14 PM
86
Handbook of Human Factors in Medical Device Design
TABLE 3.8 Vibration Experienced in Flying Vehicles Vehicle Cruising fixed-wing aircraft Fixed-wing aircraft Fixed-wing aircraft Hovercraft Helicopter 1 Helicopter 2 Rotary-wing aircraft Rotary-wing aircraft
Vibration Acceleration (m/s2)
Study
0.04 (while cruising) 0.25–0.4 (average) 0.86 (maximum obtained) 0.5 (average) 0.75 (average) 1.3 (average) 2.35 (maximum obtained) 5.6 (while cruising)
Campbell et al. (1984) Macnab et al. (1995) Macnab et al. (1995) Macnab et al. (1995) Macnab et al. (1995) Macnab et al. (1995) Macnab et al. (1995) Campbell et al. (1984)
Source: Reproduced from A. Macnab, Y. Chen, F. Gagnon, B. Bora, B., and C. Laszlo, Aviation Space and Environmental Medicine, 66, 212–219, 1995. With permission. Measurements will vary depending on make and model of vehicle as well as many other factors.
fixed-wing aircraft produce less. In general, propeller-powered aircraft produce more vibration than jet-powered aircraft. Other factors affecting vibration include vehicle load, road surface and tires (for ambulances), altitude, and weather.
3.7.2 DESIGN GUIDELINES Guideline 3.60: Minimal Device Vibration Medical devices should produce as little vibration as possible except when vibration provides useful tactile feedback to the user or patient.
Guideline 3.61: Vibrating Controls Controls subject to vibration due to vehicular movement, for example, should allow adjustment by means of relatively gross motor hand and finger movements rather than fine ones. Alternatively, they should be isolated from sources of vibration.
Guideline 3.62: Vibrating Displays Information on displays subject to vibration should be oversized to facilitate reading (see Figure 3.8). Information should also contrast sharply against its background to counterbalance the effect of vibration. The space in between rows of displayed information should be similarly enlarged (Griffin, 1997; Wickens et al., 2004).
Guideline 3.63: Vibrating Quantitative Display Quantitative displays should distinguish between values that can be made reliably given the expected level of vibration. The length of scale units should be greater than the 1.3 to 1.8 mm recommended for use in nonvibrating environments.
Guideline 3.64: Usability Testing with Vibration In the face of known or expected vibration, the usability and effectiveness of controls and displays should be rigorously tested.
TAF-ER562X_WINGR-09-0401-C003.indd 86
11/8/10 11:41:14 PM
Environment of Use
87
Blood Blood pressure pressure
Blood pressure
Blood pressure FIGURE 3.8 Text should be made larger to account for reductions in legibility due to vibration (simulated in this graphic).
3.8 ENERGY SOURCES A range of energy sources exists in environments where medical devices are used, including ultrasound, x-ray, infrared, gamma, laser, and magnetic (e.g., MRI) radiation. Medical devices can generate energy sources themselves. These energy sources may interact with each other and affect the use of other medical devices (e.g., electromagnetic interference). Many energy sources can affect user or patient health and safety, necessitating the use of personal protective equipment. The presence of energy sources affects facility design (e.g., shielding) as well as the design of other devices used in the same environment. Table 3.9 provides a list of energy sources and examples of medical devices and implications for use and safety. The increasing use of laser light in medicine has implications for the design of other devices used in the same clinical environment. For example, because laser energy is an TABLE 3.9 Energy Sources, Associated Medical Devices, and Sample Design Implications Type of Energy Source Gamma X-ray Infrared Radio frequency
Examples of Medical Device Electromagnetic energy Gamma knife Fluoroscopy CT scan Infant warmer Electrocautery MRI
Laser
Surgical laser
Ultrasound
Other sources of energy Cardiac ultrasound
Audible sounda
Voice-activated robot control
a
Examples of Potential Use or Safety Implications for Design Need for shielding and PPE Need for shielding and PPE Heating of surroundings, burns Interference Other items in suite cannot be ferrormagnetic Eye damage or fires
Need for skin contact by probe— infection and sterility issues Affect hearing alarms of other devices
See Section 3.3 on noise.
TAF-ER562X_WINGR-09-0401-C003.indd 87
11/8/10 11:41:14 PM
88
Handbook of Human Factors in Medical Device Design
ignition source, other devices used in proximity (e.g., endotracheal tubes and drapes) should not be a ready fuel source that might lead to fire.
3.8.1 GENERAL PRINCIPLES Medical devices should function in all anticipated or reasonably foreseeable use environments, including environments containing electromagnetic energy sources. Bruckart et al. (1993) tested 34 medical devices under extreme conditions of temperature, humidity, altitude, and vibration as well as their electromagnetic compatibility. Ninety-one percent of the tested devices failed at least one of the electromagnetic interference tests, and many of the devices exceeded the electromagnetic emissions standard established for use in U.S. Army helicopters (U.S. Department of Defense, 1983). Electromagnetic interference can come from a variety of sources, including radio transmitters, aircraft, power lines and other power sources, computer timing and control circuits, electrical switching transients (e.g., lights turned on or off), fans or motors, and electrostatic discharges (e.g., lightning). Between 1979 and 1995, the FDA received more than 100 reports of malfunctioning electronic medical devices presumably due to electromagnetic interference (Silberberg, 1996). For example, electromagnetic interference often affects medical device displays (e.g., the ECG tracings shown in Figure 3.9). Myriad regulations, standards, and publications are available to guide the design of devices with improved electromagnetic compatibility. Medical devices may themselves produce harmful radiation and therefore require shielding and/or the wearing of protective equipment by users and patients. For instance, linear accelerators and cobalt emitters used in radiotherapy produce highly toxic gamma radiation.
3.8.2 DESIGN GUIDELINES Guideline 3.65: Electromagnetic Compatibility Testing Designers should test a medical device’s electromagnetic compatibility in the environment(s) in which the device will be used.
Normal ECG tracing
ECG tracing during FM transmission
FIGURE 3.9 Example of electromagnetic interference in an ECG monitor on board an operating U.S. Army Medevac helicopter. (From J. E. Bruckart, J. R. Licina, and M. Quattlebaum, Air Medical Journal, 1, 51–56, 1993. With permission.)
TAF-ER562X_WINGR-09-0401-C003.indd 88
11/8/10 11:41:14 PM
Environment of Use
89
Guideline 3.66: Display Reliability in Radiant Energy Environments Designs should minimize the display of erroneous or misleading information when a device is used as intended in radiant energy environments (e.g., near MRI machines).
Guideline 3.67: Labeled as Excluded Use As appropriate, devices should be clearly labeled as unsuitable for use in environments with radiant energy sources (e.g., banning ferromagnetic materials adjacent to MRI devices). The rationale for the exclusion should be provided.
Guideline 3.68: Shielding Provided Adequate shielding and other means should be provided to protect users and patients from harmful energy sources produced by a device.
Guideline 3.69: Protective Precautions If shielding and other designed protective measures are insufficient, the device instructions for use (including on-device labeling) should clearly specify all necessary protection precautions to be taken by users and patients.
Guideline 3.70: Accommodate Use of Protective Equipment When PPE or adjuvant protective equipment (e.g., vertical lead shield on wheels) is required, the device should accommodate its use through design. For example, a portable x-ray machine should be designed to allow users to operate the device from a safe distance. Bulky protective clothing may limit users’ range of motion and hamper work movements. Device shielding or the use of protective goggles may increase visual demands and affect visual performance. Note that other devices in the use environment with different operational requirements may also necessitate PPE use.
Guideline 3.71: Storage of Protective Equipment When protective equipment is required for use with a device, the device should provide means to store this equipment.
3.9 DESIGN FOR INFECTION CONTROL Adverse outcomes or illness among patients or health care workers can occur from medical device use due to, for example, inadvertent exposures to environmental pathogens (e.g., endoscopes and respiratory therapy equipment), improper ventilation and distribution of airborne pathogens (e.g., Mycobacterium tuberculosis and varicella-zoster virus), or toxic chemical exposure. Thus, the ability to sterilize, reliably clean, and maintain the integrity of medical devices is a critical design feature.
3.9.1 GENERAL PRINCIPLES A key to infection control is to design medical devices that are easy to clean or sterilize. Sterile devices are those that must be inserted into the body, typically via invasive
TAF-ER562X_WINGR-09-0401-C003.indd 89
11/8/10 11:41:14 PM
90
Handbook of Human Factors in Medical Device Design
procedures. Surgical procedures are almost always performed under sterile conditions. Thus, surgical devices must be easily used by care personnel wearing surgical gowns, gloves, hair coverings, and protective eyewear. While one type of medical device might be cleaned multiple times a day, another might be cleaned only rarely. Sterile medical devices may be used only once (single use or disposable device) or be reusable, in which case they must be resterilized after each use. For example, a stretcher might need a scrub down with a bleach disinfectant after each use, while a tonometer, used to test for glaucoma, might receive infrequent cleaning (except for the part that touches the eye, which is cleaned after each use). Accordingly, the use environment and the device’s use profile establish cleaning requirements. In an attempt to reduce cleaning efforts and difficulties, some devices might be outfitted with protective, disposable covers. It is important for a user to be able to correctly and quickly ascertain whether a device is clean (or sterile). Devices should not be a potential source of infections. For example, a device should not generate high humidity or have a moisture reservoir that can become contaminated and serve as a nidus for infectious agents. For PPE for infection control, medical devices should be accessible and usable when the user is in full PPE (i.e., mask, isolation gown, gloves, and protective eyewear). For example, a patient in isolation for respiratory tuberculosis, a wound infection, or a bloodborne pathogen may require the health care provider to wear a respirator, impervious gown, face shield, and gloves. This provider’s range of motion, tactile senses, and vision will be affected by PPE use and this should be considered in device design. Additionally, a powered air-purifying respirator may be used (e.g., during total joint replacement surgery). These systems include helmets, hoods, head covers, and face pieces and generate significant noise.
3.9.2 DESIGN GUIDELINES Guideline 3.72: Obvious Evidence of Contamination Medical devices that must be clean (though not necessarily sterile) should reveal contamination. For instance, a light-colored finish will show grime better than a dark one.
Guideline 3.73: Impervious to Fluid Contamination As appropriate, medical devices should be impervious to fluids that might be wiped on, spilled over, or projected toward them, such as cleaning solutions, urine from a detached catheter, saline from a burst IV bag, or blood spurting from a lacerated artery.
Guideline 3.74: Device Cleanability Medical devices intended for use in clean or sterile environments should be designed for easy, thorough cleaning.
Guideline 3.75: Instructions for Cleaning Users should be provided with clear instructions for device cleaning. Appropriate (and inappropriate) cleaners and disinfectants should be identified. Devices to be used in environments where particular cleaning methods are typically used should support the same cleaning methods.
TAF-ER562X_WINGR-09-0401-C003.indd 90
11/8/10 11:41:14 PM
Environment of Use
91
Guideline 3.76: Hand Hygiene Medical device use should not interfere with proper hand hygiene. If the device itself can become contaminated during patient care and use, the user must be able to properly decontaminate his or her hands.
Guideline 3.77: Sterility Support As appropriate, devices used in sterile environments should accommodate protective covers, such as those placed over operating room light handles. Displays and handles should be positioned for ready visual and physical access underneath such covers.
Guideline 3.78: Usable Wearing Protective Equipment Medical devices should be accessible and usable when the user is in full personal protective equipment (i.e., mask, isolation gown, gloves, and protective eyewear). For example, a patient in isolation for respiratory tuberculosis, a wound infection, or a blood-borne pathogen might require the health care provider to wear a respirator, imperious gown, face shield, and gloves.
3.10 CASE STUDIES 3.10.1 CLEANING OF FLEXIBLE GASTROSCOPES To date, all published episodes of pathogen transmission related to gastrointestinal (GI) endoscopy have been associated either with a failure to follow established cleaning and disinfection/sterilization guidelines or with the use of defective equipment (Nelson et al., 2004). All flexible GI endoscopes, used for diagnosis and intervention procedures in the human GI system, require initial decontamination, thorough cleaning, and final high-level disinfection or sterilization. On removal from the patient, the endoscope is initially cleaned externally with a manufacturer-approved detergent (Figure 3.10a) and then transported to another area for further cleaning and reprocessing. All internal channels of the device must be physically cleaned (Figure 3.10b) prior to terminal reprocessing. Maintainers must wear
(a)
(b)
FIGURE 3.10 Initial cleaning of a flexible endoscope. (a) External cleaning. (b) Internal cleaning. (Photos courtesy of Carla J. Alvarado.)
TAF-ER562X_WINGR-09-0401-C003.indd 91
11/8/10 11:41:14 PM
92
Handbook of Human Factors in Medical Device Design
gear that fully protects them from both potentially infective human material (PPE as per OSHA’s Bloodborne Pathogens Standard) and toxic chemicals used during cleaning and reprocessing. Flexible endoscopes are prone to damage and are extremely difficult to clean. Cleaning involves accessing near microscopic, difficult-to-navigate 90-degree angles in the device’s internal channels. The entire cleaning process must be efficient, given the need for rapid instrument turnaround time in busy endoscopy clinics or operating rooms. This may result in significant production pressure and the risk of rushed, incomplete cleaning. Improved device design that facilitates the cleaning process would be a major improvement in safety and efficiency.
3.10.2 PATIENT USE INSULIN INFUSION PUMP Insulin infusion minipumps deliver precise amounts of rapid-acting insulin throughout the day via a catheter into the diabetic patient’s body, usually via a needle inserted under the skin. The pump is worn by patients and gives them the option to deliver extra doses when needed. In principle, the technology gives people with diabetes better control over their disease, enabling them to lead longer and healthier lives. To deliver these benefits, the insulin pump needs to small, lightweight, and easy to use (Figure 3.11). In addition, these pumps must be designed for use in a wide range of use environments. People who use the pump typically clip it onto their belt, a close distance from the abdominal injection site (Figure 3.12). A disposable set, which includes a tube, adhesive patch, and catheter, conveys insulin from the pump to the user’s body. The whole system must work properly in a wide range of climate conditions and be able to withstand considerable physical abuse. For example, a user might wear the pump outdoors on an extremely cold winter morning to catch a bus to work. On boarding the bus, he or she might accidentally bang the pump against a stanchion. The device’s rugged construction and protective plastic case defend it against damage, while its internal workings are engineered for use in extremes of temperature.
FIGURE 3.11 Insulin infusion pump worn by a diabetic patient.
TAF-ER562X_WINGR-09-0401-C003.indd 92
FIGURE 3.12 The patient wears the pump on his belt and can carry on activities of daily living without impediment.
11/8/10 11:41:15 PM
Environment of Use
93
The pump is moisture sealed, preventing precipitation and other fluids from contaminating it and possibly leading to a malfunction. Specifically, the pump shown in this case (Figure 3.11) has elastomeric push buttons mounted on top of mechanical switches to keep moisture out while also providing good tactile feedback. Thus, users do not need to worry about using the device in inclement weather, such as administering a bolus dose while watching a child play soccer in the rain. Patients who want to keep their medical condition and course of treatment private can do so. The device is as compact as a cellular phone, making it easy to hide under an overshirt or jacket. Accounting for use in dim lighting or darkness, the pump has a built-in display backlight. A software-based control allows users to increase message tone volume in noisy environments (e.g., bus station) and minimize the volume or switch to vibratory alerts in quieter settings (e.g., library).
RESOURCES Salvendy, G. (Ed.). (2006). Handbook of Human Factors and Ergonomics (3rd. ed.). New York: John Wiley & Sons. Sanders, M. S. and McCormick, E. J. (1993). Human Factors in Engineering and Design. New York: McGraw-Hill.
REFERENCES American Institute of Architects. (2001). Guidelines for Design and Construction of Hospital and Health Care Facilities. Washington, DC: American Institute of Architects. American Society of Heating, Refrigerating and Air-Conditioning Engineers. (2003). HVAC Design Manual for Hospitals and Clinics. American Society of Heating, Refrigerating and AirConditioning Engineers. Atlanta, GA: ASHRAE. Bentley, S., Murphy, F., and Dudley, H. (1977). Perceived noise in surgical wards and an intensive care area: An objective analysis. British Medical Journal, 2, 1503–1506. Bayo, M. V., Garcia, A. M., and Garcia, A. (1995). Noise levels in an urban hospital and workers’ subjective responses. Archives of Environmental Health, 50(3), 247–251. Boyce, P. R. (1997). Illumination. In G. Salvendy (Ed.), Handbook of Human Factors and Ergonomics (pp. 858–890). New York: John Wiley & Sons. Brown, L. H., Gough, J. E., Bryan-Berg, D. M., and Hunt, R. C. (1997). Assessment of breath sounds during ambulance transport. Annals of Emergency Medicine, 29(2), 228–231. Bruckart, J. E., Licina, J. R., and Quattlebaum, M. (1993). Laboratory and flight tests of medical equipment for use in U.S. Army Medevac helicopters. Air Medical Journal, 1(3), 51–56. Campbell, A. N., Lightstone, A. D., Smith, J. M., Kirpalani, H., and Perlman, M. (1984). Mechanical vibration and sound levels experienced in neonatal transport. American Journal of Diseases of Children, 138(10), 967–970. Carayon, P. and Smith, M. J. (2000). Work organization and ergonomics. Applied Ergonomics, 31, 649–662. Chang, B. J. (2002). Ergonomic benefits of surgical telescope systems: Selection guidelines. Journal of the California Dental Association, 30(2), 161–169. Chengalur, S. N., Rodgers, S. J., and Bernard, T. E. (2004). Kodak’s Ergonomic Design for People at Work (2nd ed.). Hoboken, NJ: John Wiley & Sons. Commission of the European Communities. (1996). European Guidelines on Quality Criteria for Diagnostic Radiographic Images (Vol. EUR 16260). Brussels: Commission of the European Communities.
TAF-ER562X_WINGR-09-0401-C003.indd 93
11/8/10 11:41:17 PM
94
Handbook of Human Factors in Medical Device Design
Commission of the European Communities. (1997). Criteria for Acceptability of Radiological (Including Radiotherapy) and Nuclear Installations (Radiation Protection No. 91). Luxembourg: Commission of the European Communities. Couper, R. T., Hendy, K., Lloyd, N., Gray. N., Williams, S., Bates, D. J., et al. (1994). Traffic and noise in children’s wards. Medical Journal of Australia, 160, 338–341. Crocker, M. J. (1997). Noise. In G. Salvendy (Ed.), Handbook of Human Factors and Ergonomics (2nd ed., pp. 790–827). New York: John Wiley & Sons. Fanger, P. O. (1970). Thermal Comfort, Analysis and Applications in Environmental Engineering. Copenhagen: Danish Technical Press. Fife, D. and Rappaport, E. (1976). Noise and hospital stay. American Journal of Public Health, 66(7), 680–681. Freedman, N. S., Gazendam, J., Levan, L., Pack, A. I., and Schwab, R. J. (2001). Abnormal sleep/ wake cycles and the effect of environmental noise on sleep disruption in the intensive care unit. American Journal of Respiratory and Critical Care Medicine, 163(2), 451–457. Fromm, R. E., Jr., Campbell, E., and Schlieter, P. (1995). Inadequacy of visual alarms in helicopter air medical transport. Aviation Space and Environmental Medicine, 66(8), 784–786. Gillis, W. L. and Miles, R. J. (2001). Orotrachial intubation in darkness using night vision goggles. Military Medicine, 166, 984–986. Griffin, M. J. (1997). Vibration and motion. In G. Salvendy (Ed.), Handbook of Human Factors and Ergonomics (2nd ed., pp. 828–857). New York: John Wiley and Sons. Hanna, G. B., Cresswell, A. B., and Cuschieri, A. (2002). Shadow depth cues and endoscopic task performance. Archives of Surgery, 137, 1166–1169. Illuminating Engineering Society of North America. (1981). IES Lighting Handbook. New York: Illuminating Engineering Society of North America. International Electrotechnical Commission. (2004). IEC 60601-1-6. Medical Electrical Equipment— Part 1-6: General Requirements for Safety—Collateral Standard: Usability. Geneva: International Electrotechnical Commission. Joint Commission on Accreditation of Healthcare Organizations. (2004). Environment of Care Essentials for Health Care (4th ed.). Oak Brook, IL: Joint Commission Resources. Kahn, D. M., Cook, T. E., Carlisle, C. C., Nelson, D. L., Kramer, N. R., and Millman, R. P. (1998). Identification and modification of environmental noise in an ICU setting. Chest, 114(2), 535–540. Kam, P. C., Kam, A. C., and Thompson, J. F. (1994). Noise pollution in the anaesthetic and intensive care environment. Anaesthesia, 49(11), 982–986. Keipert, J. A. (1985). The harmful effects of noise in a children’s ward. Australian Paediatric Journal, 21, 101–103. Krupinski, E., Roehrig, H., and Furukawa, T. (1999). Influence of film and monitor display luminance on observer performance and visual search. Academic Radiology, 6(7), 411–418. Macnab, A., Chen, Y., Gagnon, F., Bora, B., and Laszlo, C. (1995). Vibration and noise in pediatric emergency transport vehicles: a potential cause of morbidity? Aviation Space and Environmental Medicine, 66(3), 212–219. McCarthy, E. and Brennan, P. C. (2003). Viewing conditions for diagnostic images in three major Dublin hospitals: A comparison with WHO and CEC recommendations. British Journal of Radiology, 76, 94–97. McLaughlin, A., McLaughlin, B., Elliott, J., Campalani, G, et al. (1996). Noise levels in a cardiac surgical intensive care unit: A preliminary study conducted in secret. Intensive Critical Care Nursing, 12, 226–230. Meyer, T. J., Eveloff, S. E., Bauer, M. S., Schwartz, W. A., Hill, N. S., Millman, R. P., et al. (1994). Adverse environmental conditions in the respiratory and medical ICU settings. Chest, 105, 1211–1216. Meyer-Falcke, A., Rack, R., Eichwede, F., Jansing, P. J., et al. (1994). How noisy are anaesthesia and intensive care medicine? Quantification of the patients’ stress. European Journal of Anaesthesiology, 11, 407–411.
TAF-ER562X_WINGR-09-0401-C003.indd 94
11/8/10 11:41:17 PM
Environment of Use
95
Morrison, W. E., Haas, E. C., Shaffner, D. H., Garrett, E. S., and Fackler, J. C. (2003). Noise, stress, and annoyance in a pediatric intensive care unit. Critical Care Medicine, 31(1), 113–119. Nelson, D. B., Jarvis, W. R., Rutala, W. A., Foxx-Orenstein, A. E., Isenberg, G., Dash, G. P., et al. (2004). Multi-society guideline for reprocessing flexible gastrointestinal endoscopes. Diseases of the Colon and Rectum, 47(4), 413–421. Parsons, K. C. (1993). Human Thermal Environments. London: Taylor & Francis. Parsons, K. C. (2000). Environmental ergonomics: A review of principles, methods and models. Applied Ergonomics, 31, 581–594. Philbin, M. K., Robertson, A., and Hall, J. W., III. (1999). Recommended permissible noise criteria for occupied, newly constructed or renovated hospital nurseries. Journal of Perinatology, 19(8, Pt. 1), 559–563. Prasad, N. H., Brown, L. H., Ausband, S. C., Cooper-Spruill, O., Carroll, R. G., and Whitley, T. W. (1994). Prehospital blood pressures: Inaccuracies caused by ambulance noise? American Journal of Emergency Medicine, 12(6), 617–620. Redding, J. S., Hargest, T. S., and Minksy, S.H. (1977). How noisy is intensive care? Critical Care Medicine, 5, 275–276. Rohles, F. H. and Konz, S. A. (1987). Climate. New York: John Wiley & Sons. Sanders, M. S. and McCormick, E. J. (1993). Human Factors in Engineering and Design (7th ed.). New York: McGraw-Hill. Schwartz, R. B. and Charity, B. M. (2001). Use of night vision goggles and low-level light source in obtaining intravenous access in tactical conditions of darkness. Military Medicine, 166, 982–983. Shankar, N., Malhotra, K. L., Ahuja, S., and Tandon, O. P. (2001). Noise pollution: A study of noise levels in the operation theatres of a general hospital during various surgical procedures. Journal of the Indian Medical Association, 99(5), 244, 246–247. Silberberg, J. L. (1996). What can/should we learn from reports of medical device electromagnetic interference. Compliance Engineering, 13(4), 41–57. Silberberg, J. L. (2001). Achieving medical device EMC: The role of regulations, standards, guidelines and publications. IEEE International Symposium on Electromagnetic Compatibility, 2, 1298–1303. Sjors, G., Hammarlund, K., Oberg, P.-A., and Sedin, G. (1992). An evaluation of environment and climate control in seven infant incubators. Biomedical Instrumentation and Technology, 26(4), 294–301. Smith, M. J. and Carayon-Sainfort, P. (1989). A balance theory of job design for stress reduction. International Journal of Industrial Ergonomics, 4, 67–79. Sust, C. A. and Lazarus, H. (2003). Signal perception during performance of an activity under the influence of noise. Noise and Health, 6(21), 51–62. The Chartered Institution of Building Services Engineers. (1989). Lighting Guide—Hospitals and Health Care Buildings (LG2: 1989 ed.). London: The Chartered Institution of Building Services Engineers. Tsiou, C., Eftymiatos, D., Theodossopoulou, E., Notis, P., and Kiriakou, K. (1998). Noise sources and levels in the Evgenidion Hospital intensive care unit. Intensive Care Medicine, 24(8), 845–847. U.S. Department of Defense. (1983). Military Standard: Environmental Test Methods and Engineering Guidelines (MIL-STD-810). Washington, DC: U.S. Department of Defense. Vincent, C., Taylor-Adams, S., and Stanhope, N. (1998). Framework for analysing risk and safety in clinical medicine. British Medical Journal, 316(7138), 1154–1157. Weinger, M. B. and Englund, C. E. (1990). Ergonomic and human factors affecting anesthetic vigilance and monitoring performance in the operating room environment. Anesthesiology, 73(5), 995–1021. Wickens, C. D., Lee, J. D., Liu, Y., and Becker, S. E. G. (2004). An Introduction to Human Factors Engineering (2nd ed.). Upper Saddle River, NJ: Prentice Hall.
TAF-ER562X_WINGR-09-0401-C003.indd 95
11/8/10 11:41:17 PM
96
Handbook of Human Factors in Medical Device Design
Wiklund, M. E. (1995). Medical Device and Equipment Design—Usability Engineering and Ergonomics. Buffalo Grove, IL: Interpharm Press. World Health Organization. (1982). Quality Assurance in Diagnostic Radiology. Geneva: World Health Organization. World Health Organization. (1999). Guidelines for Community Noise. Retrieved June 1, 2004, from http://www.who.int/docstore/peh/noise/guidelines2.html
TAF-ER562X_WINGR-09-0401-C003.indd 96
11/8/10 11:41:17 PM
and 4 Anthropometry Biomechanics W. Gary Allread, PhD, CPE; Edmond W. Israelski, PhD CONTENTS 4.1 General Principles of Good Anthropometric Design ................................................99 4.1.1 One-Dimensional Measurements ..................................................................100 4.1.1.1 Hand Data........................................................................................102 4.1.1.2 Foot Data .........................................................................................107 4.1.1.3 Data Regarding Children.................................................................107 4.1.1.4 Additional Data Sources.................................................................. 110 4.1.2 Mobility and Functional Measurements ....................................................... 111 4.1.2.1 Mobility ........................................................................................... 111 4.1.2.2 Functional Work .............................................................................. 111 4.1.2.3 Visual Work..................................................................................... 111 4.1.3 Strength ......................................................................................................... 112 4.1.3.1 Strength and Gender Differences .................................................... 116 4.1.3.2 Strength of the Upper Extremity ..................................................... 116 4.1.3.3 Strength of the Fingers and Hands .................................................. 117 4.1.3.4 Strength of the Feet ......................................................................... 117 4.1.4 Special Considerations ..................................................................................121 4.1.5 Design Guidelines for Body Dimension Data...............................................122 4.1.5.1 Designing for Population Extremes .................................................122 4.1.5.2 Designing for the Average User.......................................................124 4.1.5.3 Designing for Adjustability .............................................................124 4.1.5.4 Deriving Missing Data ....................................................................125 4.2 Case Studies in Anthropometric Design..................................................................127 4.2.1 Viewing Angle Determination for a Cardiac Output Monitor ......................127 4.2.1.1 Analysis ...........................................................................................127 4.2.1.2 High-Mount Position .......................................................................127 4.2.1.3 Calculated Viewing Angles for the High-Mount Position ...............127 4.2.1.4 Low-Mount Position ........................................................................128 4.2.1.5 Calculated Viewing Angle for the Low-Mount Position .................128 4.2.1.6 Viewing Angles, Summary, and Implications .................................129 4.2.2 Keyboard Height in a Diagnostic System Workstation .................................129 4.2.3 Finger Clearance Space Calculations ...........................................................130 4.2.3.1 Prototype Blood Analyzer—Sample Probe Hand Clearance .......... 131 97
TAF-ER562X_WINGR-09-0401-C004.indd 97
11/9/10 2:23:06 AM
98
Handbook of Human Factors in Medical Device Design
4.2.3.2 Portable Infusion Pump Access for Thumb Bolus Activation ......... 131 4.2.3.3 Clinical Chemistry Diagnostic System Access Opening ................132 4.3 General Principles of Good Biomechanical Design ................................................132 4.3.1 Critical Design Considerations .....................................................................133 4.3.1.1 Body Posture ...................................................................................133 4.3.1.2 Trade-Offs .......................................................................................133 4.3.1.3 Endurance........................................................................................134 4.3.1.4 Repetitive Motions ..........................................................................135 4.3.2 Special Populations .......................................................................................135 4.3.3 Design Guidelines for Tasks Involving Lifting .............................................136 4.3.3.1 NIOSH Revised Lifting Equation ...................................................136 4.3.3.2 ACGIH Lifting Threshold Limit Values .........................................137 4.3.3.3 Industrial Lumbar Motion Monitor Risk Assessment System ........138 4.3.3.4 Psychophysical Limits .....................................................................139 4.3.4 Design Guidelines for Tasks Involving Use of the Upper Extremity ............139 4.3.4.1 Strain Index .....................................................................................139 4.3.4.2 Rapid Upper Limb Assessment .......................................................140 4.3.5 Design Guidelines to Determine Strength Requirements .............................140 4.3.5.1 Three-Dimensional Static Strength Prediction Program.................140 4.4 Case Studies in Biomechanical Design ................................................................... 141 4.4.1 Knob Twisting Forces on Pole Clamps ......................................................... 141 4.4.1.1 Analysis ...........................................................................................142 4.4.2 Bending Forces on an Autoinjector Device .................................................. 143 4.4.2.1 Analysis ........................................................................................... 143 4.4.3 Snap-On Lid Removal Forces .......................................................................144 4.4.3.1 Analysis ...........................................................................................144 4.4.4 Pulling Forces on IV Tubing.........................................................................146 4.4.4.1 Analysis ...........................................................................................146 4.4.5 Pulling Force to Remove a Plunger from a Vial ...........................................146 4.4.5.1 Analysis ........................................................................................... 147 Anthropometry-Related Resources ..................................................................................148 Biomechanics-Related Resources ....................................................................................149 References ........................................................................................................................150 The understanding of human physical capabilities and limitations is fundamental to the design of effective medical devices. Properly designed medical tools, equipment, and workstations will help to reduce errors, decrease injury risk, increase productivity, and improve user satisfaction. To produce an effective medical device, one must have at least a basic understanding of the many scientific disciplines that are involved in proper design—engineering, psychology, anatomy, and physiology. Although this chapter does not provide details of all these areas, it acquaints the reader with the essential information and reference materials that designers commonly use to solve human-related design problems. The citations provided here are not meant to be an all-inclusive listing but were carefully chosen sources on topics that relate to many design concerns in the medical device realm. The two design issues addressed in this chapter are anthropometry and biomechanics. Anthropometry is the science of measuring and quantifying various human physical traits,
TAF-ER562X_WINGR-09-0401-C004.indd 98
11/9/10 2:23:06 AM
Anthropometry and Biomechanics
99
such as size, weight, proportion, mobility, and strength. Biomechanics is the use of laws of physics and engineering principles to study various body segments as they move and are acted on by internal and external forces. Most of the anthropometric and biomechanical information available today is not specific to health care professionals and patients or to medical equipment. However, much of the published body size data are just as useful to the medical device designer as they are to any other designer who must combine human physical dimensions with workplace tasks and equipment. Human biomechanical principles apply to the design process of a device as well. Therefore, this chapter contains many data tables, figures, and other resources that will facilitate the application of anthropometric and biomechanical data to the design of medical devices. This chapter presents a broad assortment of anthropometric data because of the inherent, wide range of human physical dimensions that exists in any population. This is particularly an issue in ethnically diverse workplaces, as individuals may differ drastically in their physical characteristics. In the United States, variability in the physical sizes of workers is expected to increase in the future. The U.S. Bureau of Labor Statistics projects that, in the upcoming decade, the percentage of newly hired African American, Hispanic, and Asian employees will far exceed new Caucasian workers. This affects the design of medical devices because these groups can differ in many physical dimensions. The resulting variability in size, strength, and mobility among device users requires that anthropometric and biomechanical factors be taken into account to ensure the proper design of medical devices, accessories, and work environments. This chapter includes data on body size distributions of males and females. This is important, as there are many medical devices for which knowledge of user anthropometrics is critical to design. Some examples include (1) the keyboard and monitor height for laboratory diagnostic (IVD) workstations, (2) the hand span for manipulating a suture closing device, and (3) the gaze height for the shortest and tallest users of an overhead vital signs monitor. This chapter also contains reference information for tools used to assess the biomechanical injury risk of work performed in medical environments. Anthropometric and biomechanical design principles will assist the medical device designer in producing tools, equipment, and workstations that will accommodate most of the individuals for whom they were intended. These principles also will improve safety and appropriate use as well as reduce the risk of musculoskeletal strain. However, they do not guarantee that the device will fit every user, and they do not guarantee that injuries or misuse will be eliminated among those interacting with these devices. Finally, it is impossible to include in this chapter information about every possible user group and work situation. Further, there are numerous types of physical limitations that can arise from a multitude of disabilities. However, this chapter directs the medical device designer to information not included here and illustrates how to determine needed data from existing sources.
4.1 GENERAL PRINCIPLES OF GOOD ANTHROPOMETRIC DESIGN In most cases, the design of medical tools, equipment, and workstations should accommodate adults ranging in size from a 5th-percentile female to a 95th-percentile male. A 5thpercentile value for any particular body dimension indicates that 5% of the population will be equal to or smaller than that value and that the sizes of the remaining 95% will be larger.
TAF-ER562X_WINGR-09-0401-C004.indd 99
11/9/10 2:23:06 AM
100
Handbook of Human Factors in Medical Device Design
Conversely, a 95th-percentile value indicates that 95% of the population will be equal to or smaller than that value on a particular dimension and that 5% will be larger. There is considerable internal variation among human body sizes. In other words, an individual who is measured at the 95th percentile on one particular body dimension likely will not fall in this same percentile for other dimensions. In addition, the relationship between two dimensions may vary as a function of gender. That is, body segment proportions for males can differ considerably from the proportions of the same segments in females. Medical device designers should take these facts into account. Designing a medical device in a range that will accommodate users from the 5th-percentile female to the 95th-percentile male will theoretically provide coverage for 90% of the user population for that dimension. While designers should accommodate the widest possible range of users, higher costs and infeasibility often prevent the designer from fully accommodating the entire user population. Certain medical equipment devices and tasks involve critical functions, when it is necessary to accommodate more than the central 90% of potential users. In these instances, data representing larger ranges of the population could be used, such as the first to 99th percentiles. However, there often are trade-offs involved in designing to these ranges. Cost and functionality must be considered when attempting to accommodate larger percentages of any given population. For example, designing an adjustable armrest for the 5th through 95th percentile of U.S. women, based on their seated elbow resting height, requires a range of 10.2 cm. However, this range increases nearly 50% (to 15.0 cm) when accommodating the 1st through 99th percentile of this population. This simple example shows that the inclusion of a greater segment of the population for a device’s design on any particular dimension may not be feasible or may significantly increase material cost. In summary, good anthropometric design of medical devices should do the following: • At a minimum, accommodate adults who range from a 5th-percentile female to a 95th-percentile male in size • Accommodate larger percentages of the population if (1) the device involves critical functions, (2) device usability and functionality are not compromised, or (3) there are not excessive costs involved in doing so
4.1.1 ONE-DIMENSIONAL MEASUREMENTS Individuals from ethnically diverse populations often operate the same medical devices. As a result, designers must understand the size ranges of these users. This chapter presents a representative sample of available data. Table 4.1 defines body segment terms, and Table 4.2 lists the measurements. These are based on values for a 40-year-old American male and a 40-year-old Asian female projected to the year 2000, which is the most recent international data available (National Aeronautics and Space Administration, 1989). Figure 4.1 graphically illustrates these dimensions. Device designers should use Tables 4.1 and 4.2 and Figure 4.1 as a starting point from which to assess the adequacy of their equipment designs. Historically, most anthropometric data have come from military populations, and some medical device designers may question the relevancy of this information. However, Kroemer, Kroemer, and Kroemer-Elbert (1997, p. 17) reported that “with proper caution and insight, we can use military anthropometric data to approximate size data for the general
TAF-ER562X_WINGR-09-0401-C004.indd 100
11/9/10 2:23:06 AM
Anthropometry and Biomechanics
101
TABLE 4.1 Descriptions of Male and Female Anthropometric Measurements Ref. No.a 64 100 178 194 200 215 236 249 308 312 330 378 381 411 416 420 459 506 529 612
Dimension
Ankle height The height of the level of minimum circumference of the leg Biacromial breadth The distance across the shoulders from the right to the left acromion Buttock The circumference of the body measured at the level of the maximum circumference posterior protuberance of the buttocks Buttock-knee length The horizontal distance from the rearmost surface of the buttocks to the front of the kneecaps Buttock-popliteal The horizontal distance from the rearmost surface of the buttock to the length back of the lower leg Calf height The height of the level of the maximum circumference of the lower leg Chest depth (male) The depth of the torso measured at nipple level Crotch height The height of the midpoint of the crotch Elbow height The height of the radiale Elbow rest height The height of the bottom of the tip of the elbow above the sitting surface Eye height, sitting The height of the inner corner of the eye above the sitting surface Forearm-forearm The distance across the tissue mass of the forearms measured with the breadth elbows flexed and resting lightly against the body Forearm-hand length The distance from the tip of the elbow to the tip of the longest finger Hand breadth The breadth of the hand as measured across the distal ends of the metacarpal bones Hand circumference The circumference of the hand measured around its knuckles Hand length The distance from the base of the hand to the top of the middle finger measured along the long axis of the hand Hip breadth, sitting The breadth of the body as measured across the widest portion of the hips Interscye The tape distance across the back between the posterior axillary folds at the lower level of the armpits Knee height, sitting The height, from the footrest surface, of the musculature just above the knee
639 678 735
Mid-shoulder height, sitting Neck circumference Popliteal height Scye circumference
754
Shoulder length
758 761
Sitting height Shoulder-elbow length Stature Thigh clearance Vertical trunk circumference
805 856 916
Definition
The height of the point on the shoulder halfway between the neck and the acromion above the sitting surface The maximum circumference of the neck, including the Adam’s apple (in males) The height of the underside of the upper leg above the footrest surface The circumference of the scye, measured in a vertical plane, as high as possible in the armpit and passing over the acromion The surface distance from the acromion, at the end of the shoulder blade, to the junction of the shoulder and the neck The height, from the sitting surface, to the top of the head The vertical distance from the acromion to the bottom of the elbow, measured with the elbow bent 90 degrees and the lower arm held horizontal The height of the top of the head The height of the highest point of the thigh above the sitting surface The circumference of the torso measured with the tape passing diagonally across the front of the body from the midpoint of the shoulder to the crotch, through the crotch, over the posterior protuberance of the buttock and along the small of the back continued
TAF-ER562X_WINGR-09-0401-C004.indd 101
11/9/10 2:23:06 AM
102
Handbook of Human Factors in Medical Device Design
TABLE 4.1 (CONTINUED) Descriptions of Male and Female Anthropometric Measurements Ref. No.a
Dimension
921
Waist back
946
Waist front
949 973
Waist height Wrist height
Definition The surface distance, along the spine, from waist level to the cervicale (the top of the seventh cervical vertebra) The surface distance in the midsagittal plane from waist level to the suprasternale The height of waist level The height of the stylion
Source: National Aeronautics and Space Administration, Man-Systems Integration Standards. NASA-STD3000A, Lyndon B. Johnson Space Center, Houston, TX, 1989. With permission. a Reference numbers refer to numbers shown in Figure 4.1, Table 4.2, and in the NASA standard document cited.
population. Dimensions of the head, hand, and foot are virtually the same in military and civilian populations.” Therefore, the data referred to in this chapter are arguably the best available. However, designers should learn about any unique physical characteristics of their medical device’s user population (e.g., elderly individuals, pregnant women) that may make it different from a military population so as to achieve a proper integration of these factors and a suitable design. A recent project has established a more comprehensive database of the civilian U.S. and European population: the Civilian American and European Surface Anthropometry Resource (CAESAR™) Project (Robinette, 2000). These anthropometric data are used in computer-aided design software applications to model human movement. The CAESAR™ data, from a sample of thousands of individuals, can be filtered by age (18–65 years), gender, weight, and ethnic group. These data are available for purchase from the Society of Automotive Engineers (Warrendale, PA). Data samples can be obtained from the U.S. Air Force’s Computerized Anthropometric Research and Design Laboratory (Wright-Patterson Air Force Base, OH). The choice of anthropometric database for use by medical device designers should do the following: • Be specific to the population of users of that device if such a database exists • Use civilian-based information when possible • Consider use of existing military data if it can be assumed that the dimensions of interest are not different between military and civilian populations Medical device designers often need even more specific anthropometric data. Obtaining these data may require measurements of specific body parts (such as the hands or feet), perhaps in specialized populations (such as children). Some of this information is presented in the following sections. 4.1.1.1 Hand Data Many medical devices are used specifically with the hands, making knowledge of measurement ranges for this body part critical. Devices that require hand dimension data would include any surgical instruments that are manipulated and need to accommodate specific
TAF-ER562X_WINGR-09-0401-C004.indd 102
11/9/10 2:23:07 AM
TAF-ER562X_WINGR-09-0401-C004.indd 103
64 100 178 194 200 215 236 249 308 312 330 378 381 411 416 420 459 506 529 612 639 678 735
Ref. No.
Ankle height Biacromial breadth Buttock circumference Buttock-knee length Buttock-popliteal length Calf height Chest depth (male) Crotch height Elbow height Elbow rest height Eye height, sitting Forearm-forearm breadth Forearm-hand length Hand breadth Hand circumference Hand length Hip breadth, sitting Interscye Knee height, sitting Midshoulder height, sitting Neck circumference Popliteal height Scye circumference
Dimension 12.0 (4.7) 37.9 (14.9) 91.0 (35.8) 56.8 (22.4) 46.9 (18.5) 32.5 (12.8) 21.8 (8.6) 79.4 (31.3) na 21.1 (8.3) 76.8 (30.3) 48.8 (19.2) na 8.2 (3.2) 20.3 (8.0) 17.9 (7.0) 34.6 (13.6) 32.9 (13.0) 52.6 (20.7) 60.8 (23.9) 35.5 (14.0) 40.6 (16.0) 44.4 (17.5)
5th 13.9 (5.5) 41.1 (16.2) 100.2 (39.4) 61.3 (24.1) 51.2 (20.2) 36.2 (14.3) 25.0 (9.8) 86.4 (34.0) na 25.4 (10.0) 81.9 (32.2) 55.1 (21.7) na 8.9 (3.5) 21.8 (8.6) 19.3 (7.6) 38.4 (15.1) 39.2 (15.4) 56.7 (22.3) 65.4 (25.7) 38.7 (15.2) 44.4 (17.5) 49.0 (19.3)
50th
Percentile
15.8 (6.2) 44.3 (17.5) 109.4 (43.1) 65.8 (25.9) 55.5 (21.9) 40.0 (15.7) 28.2 (11.1) 93.3 (36.7) na 29.7 (11.7) 86.9 (34.2) 61.5 (24.2) na 9.6 (3.8) 23.4 (9.2) 20.6 (8.1) 42.3 (16.6) 45.4 (17.9) 60.9 (24.0) 70.0 (27.5) 41.9 (16.5) 48.1 (19.0) 53.6 (21.1)
95th
40-Year-Old American Male, Projected for Year 2000
TABLE 4.2 Male and Female Anthropometric Measurements, in cm (in.)
5.2 (2.0) 32.4 (12.8) 79.9 (31.5) 48.9 (19.2) 37.9 (14.9) 25.5 (10.0) na 65.2 (25.7) 92.8 (36.5) 20.7 (8.2) 68.1 (26.8) na 37.3 (14.7) 6.9 (2.7) 16.5 (6.5) 15.8 (6.2) 30.4 (12.0) 32.4 (12.8) 41.6 (16.4) na 34.5 (13.6) 34.7 (13.6) 32.3 (12.7)
5th
6.1 (2.4) 35.7 (14.1) 87.1 (34.3) 53.3 (21.0) 41.7 (16.4) 28.9 (11.4) na 70.6 (27.8) 98.4 (38.8) 25.0 (9.9) 73.8 (29.1) na 41.7 (16.4) 7.8 (3.1) 17.9 (7.0) 17.2 (6.8) 33.7 (13.3) 35.7 (14.1) 45.6 (17.9) na 37.1 (14.6) 38.3 (15.1) 36.1 (14.2)
50th
Percentile 95th
continued
7.0 (2.8) 39.0 (15.4) 94.3 (37.1) 57.8 (22.7) 45.5 (17.9) 32.3 (12.7) na 76.1 (30.0) 104.1 (41.0) 29.3 (11.5) 79.6 (31.4) na 44.6 (17.6) 8.6 (3.4) 19.3 (7.6) 18.7 (7.3) 37.0 (14.6) 39.0 (15.4) 49.5 (19.5) na 39.7 (15.6) 41.9 (16.5) 39.8 (15.7)
40-Year-Old Japanese Female, Projected for Year 2000
Anthropometry and Biomechanics 103
11/9/10 2:23:07 AM
TAF-ER562X_WINGR-09-0401-C004.indd 104
Waist back Waist front Waist height Wrist height
921 946 949 973
43.7 (17.2) 37.2 (14.6) 100.4 (39.5) na
14.8 (5.8) 88.9 (35.0) 33.7 (13.3) 169.7 (66.8) 14.5 (5.7) 158.7 (62.5)
5th
47.6 (18.8) 40.9 (16.1) 108.3 (42.6) na
16.9 (6.7) 94.2 (37.1) 36.6 (14.4) 179.9 (70.8) 16.8 (6.6) 170.7 (67.2)
50th
51.6 (20.3) 44.6 (17.5) 116.2 (45.7) na
19.0 (7.5) 99.5 (39.2) 39.4 (15.5) 190.1 (74.8) 19.1 (7.5) 182.6 (71.9)
95th
35.2 (13.9) na 90.1 (35.5) 70.8 (27.9)
11.3 (4.4) 78.3 (30.8) 27.2 (10.7) 148.9 (58.6) 11.2 (4.4) 136.9 (53.9)
5th
38.1 (15.0) na 96.7 (38.1) 76.6 (30.2)
13.1 (5.1) 84.8 (33.4) 29.8 (11.7) 157.0 (61.8) 12.9 (5.1) 146.0 (57.5)
50th
Percentile 95th
41.0 (16.1) na 103.4 (40.7) 82.4 (32.4)
14.8 (5.8) 91.2 (35.9) 32.4 (12.8) 165.1 (65.0) 14.5 (5.7) 155.2 (61.1)
40-Year-Old Japanese Female, Projected for Year 2000
Source: National Aeronautics and Space Administration, Man-Systems Integration Standards. NASA-STD-3000A, Lyndon B. Johnson Space Center, Houston, TX, 1989. With permission. na, not available.
Shoulder length Sitting height Shoulder-elbow length Stature Thigh clearance Vertical trunk circumference
Dimension
754 758 761 805 856 916
Ref. No.
Percentile
40-Year-Old American Male, Projected for Year 2000
TABLE 4.2 (CONTINUED) Male and Female Anthropometric Measurements, in cm (in.)
104 Handbook of Human Factors in Medical Device Design
11/9/10 2:23:07 AM
Anthropometry and Biomechanics
(a)
105 378
805 754 506
236 308
639 612
921
916 459
973
64 (b)
758
411 330
420 761 416 529
194
678
(c) 735
100 946
312
178
949
381
856 200 249
215
FIGURE 4.1 Graphical illustrations of the body dimension data provided in Table 4.2. The reference numbers shown in these drawings relate to those listed in the first column of Table 4.2 (From National Aeronautics and Space Administration, Man-Systems Integration Standards, NASA-STD3000A, Lyndon B. Johnson Space Center, Houston, TX, 1989. With permission.)
ranges of hand dimensions, such as hand span. For example, an over-the-wire catheter control device needs to have a functional hand spread no larger than a 5th-percentile female’s dimension, so that a majority of intended users can operate the device properly. Table 4.3 presents (and Figure 4.2 illustrates) anthropometric hand data for both males and females. Pheasant (1996) compiled this information from a number of civilian and military sources
TAF-ER562X_WINGR-09-0401-C004.indd 105
11/9/10 2:23:07 AM
106
Handbook of Human Factors in Medical Device Design
TABLE 4.3 Male and Female Anthropometric Hand Measurements, in cm (in.)
Ref. No.a 1 2 3 4 5 6 7 8 9 10
11
12 13 15 16 17
18
19
20
Dimension Hand length Palm length Thumb length Index finger length Middle finger length Ring finger length Little finger length Thumb breadth (at interphalangeal joint) Thumb thickness (at interphalangeal joint) Index finger breadth (at the proximal interphalangeal joint) Index finger thickness (at the proximal interphalangeal joint) Hand breadth (metacarpal) Hand breadth (across thumb) Hand thickness (metacarpal) Hand thickness (including thumb) Maximum grip diameter (thumb and middle fingers just touching) Maximum spread (tip of thumb to tip of fifth finger) Maximum functional spread (tip end segments of the thumb and ring fingers) Minimum square access
5th
Male
Female
Percentile
Percentile
50th
95th
5th
50th
95th
17.30 (6.81) 18.90 (7.44) 20.50 (8.07) 15.90 (6.26) 17.40 (6.85) 18.90 (7.44) 9.80 (3.86) 10.70 (4.21) 11.60 (4.57) 8.90 (3.50) 9.70 (3.82) 10.50 (4.13) 4.40 (1.73) 5.10 (2.01) 5.80 (2.28) 4.00 (1.57) 4.70 (1.85) 5.30 (2.09) 6.40 (2.52) 7.20 (2.83) 7.90 (3.11) 6.00 (2.36) 6.70 (2.64) 7.40 (2.91) 7.60 (2.99) 8.30 (3.27) 9.00 (3.54) 6.90 (2.72) 7.70 (3.03) 8.40 (3.31) 6.50 (2.56) 7.20 (2.83) 8.00 (3.15) 5.90 (2.32) 6.60 (2.60) 7.30 (2.87) 4.80 (1.89) 5.50 (2.17) 6.30 (2.48) 4.30 (1.69) 5.00 (1.97) 5.70 (2.24) 2.00 (0.79) 2.30 (0.91) 2.60 (1.02) 1.70 (0.67) 1.90 (0.75) 2.10 (0.83) 1.90 (0.75)
2.20 (0.87)
2.40 (0.94)
1.50 (0.59)
1.80 (0.71)
2.00 (0.79)
1.90 (0.75)
2.10 (0.83)
2.30 (0.91)
1.60 (0.63)
1.80 (0.71)
2.00 (0.79)
1.70 (0.67)
1.90 (0.75)
2.10 (0.83)
1.40 (0.55)
1.60 (0.63)
1.80 (0.71)
7.80 (3.07)
8.70 (3.43)
9.50 (3.74)
6.90 (2.72)
7.60 (2.99)
8.30 (3.27)
9.70 (3.82) 10.50 (4.13) 11.40 (4.49)
8.40 (3.31)
9.20 (3.62)
9.90 (3.90)
2.70 (1.06)
3.30 (1.30)
3.80 (1.50)
2.40 (0.94)
2.80 (1.10)
3.30 (1.30)
4.40 (1.73)
5.10 (2.01)
5.80 (2.28)
4.00 (1.57)
4.50 (1.77)
5.00 (1.97)
4.50 (1.77)
5.20 (2.05)
5.90 (2.32)
4.30 (1.69)
4.80 (1.89)
5.30 (2.09)
17.80 (7.01) 20.60 (8.11) 23.40 (9.21) 16.50 (6.50) 19.00 (7.48) 21.50 (8.46)
12.20 (4.80) 14.20 (5.59) 16.20 (6.38) 10.90 (4.29) 12.70 (5.00) 14.50 (5.71)
5.60 (2.20)
6.60 (2.60)
7.60 (2.99)
5.00 (1.97)
5.80 (2.28)
6.70 (2.64)
Source: Pheasant, S., Bodyspace: Anthropometry, Ergonomics and the Design of Work, Taylor & Francis Ltd., London, 1996. a Reference numbers refer to numbers shown in Figure 4.2 and in Pheastant’s book.
TAF-ER562X_WINGR-09-0401-C004.indd 106
11/9/10 2:23:08 AM
Anthropometry and Biomechanics
107
6 7
17
4 10
1
12 2
5
3 13
9
18
11
15 16
19
8 20
FIGURE 4.2 Graphic illustrations of hand dimension data. The reference numbers shown in these drawings relate to those listed in the first column of Table 4.3 (From Pheasant, S., Bodyspace: Anthropometry, ergonomics and the design of work, Taylor & Francis Ltd., London, 1996.)
for use as a preliminary reference during tool and equipment design in which interaction with the hands is a primary requirement. 4.1.1.2 Foot Data Individuals use their feet to activate many types of medical devices. Table 4.4 provides anthropometric foot data for two relevant dimensions (illustrated in Figure 4.3) that were compiled from several sources for various national populations (Flugel, Greil, and Sommer, 1986; Gordon et al., 1989; Kagimoto, 1990; Pheasant, 1996). These data should be used as an initial source for designing foot-activated controls, workspace clearances, or other devices where knowledge of foot size ranges is important. Examples of medical devices that use the feet are activator switches for electrocauterization surgical instruments (that essentially burn human tissue to stop bleeding) and fluoroscopes (e.g., to image dye-stained blood as part of an angiogram to determine extent of coronary artery blockage). One important piece of data from Table 4.4 would be foot length of a 5th-percentile operator of these activation switches to be sure the pedal is not too long to be safely operated by someone with a small foot. Table 4.4 shows the 5th-percentile female dimension to be 21.5 cm. 4.1.1.3 Data Regarding Children Children also use a wide range of medical devices, such as handheld glucose meters, ambulatory enteral feeding pumps, external insulin pumps, and mobility aids, such as wheelchairs, crutches, and canes. Thus, designing devices for this population requires knowledge of their unique physical measurements. Anthropometric data for those younger than age 18 are less detailed than for adult populations. Table 4.5 presents some of the published data for children (Diffrient, Tilley, and Bardagjy, 1981), with reference measurements graphically
TAF-ER562X_WINGR-09-0401-C004.indd 107
11/9/10 2:23:08 AM
108
Handbook of Human Factors in Medical Device Design
TABLE 4.4 Male and Female Anthropometric Foot Measurements, in cm (in.)
Ref. No.a 1
2
Dimension Foot length (the maximal length of the right foot, when standing) U.S. adults, aged 19–60 British adults, aged 19–35 East German adults, aged 18–59 Japanese adults, aged 18–30 Foot breadth (the maximal breadth of the right foot, at right angle to the long axis of the foot, when standing) U.S. adults, aged 19–60 British adults, aged 19–35 East German adults, aged 18–59 Japanese adults, aged 18–30
Male
Female
Percentile
Percentile
5th
50th
95th
5th
50th
95th
24.9 (9.8) 24.0 (9.5) 24.3 (9.6) 23.4 (9.2)
27.0 (10.6) 26.5 (10.4) 26.4 (10.4) 25.1 (9.9)
29.2 (11.5) 28.5 (11.2) 28.5 (11.2) 26.9 (10.6)
22.4 (8.8) 21.5 (8.5) 22.2 (8.7) 21.7 (8.5)
24.4 (9.6) 23.5 (9.3) 24.1 (9.5) 23.2 (9.1)
26.5 (10.4) 25.5 (10.0) 26.0 (10.2) 24.6 (9.7)
9.2 (3.6) 10.1 (4.0) 11.0 (4.3) 8.5 (3.4) 9.5 (3.7) 11.0 (4.3) 9.1 (3.6) 10.2 (4.0) 11.3 (4.5) 9.7 (3.8) 10.4 (4.1) 11.1 (4.4)
8.2 (3.2) 8.0 (3.2) 8.3 (3.3) 8.9 (3.5)
9.0 (3.5) 9.8 (3.9) 9.0 (3.5) 10.0 (3.9) 9.3 (3.7) 10.4 (4.1) 9.6 (3.8) 10.3 (4.1)
Sources: Flugel, F., Greil, H., and Sommer, KI. Anthropologischer Atlas, Tribuene, Berlin, 1986; Gordon, C.C. et al., 1988 Anthropometric Survey of U.S. Army Personnel: Summary Statistics Interim Report, Natick-TR89/027. U.S. Army Natick Research, Development and Engineering Center, Natick, MA, 1989; Kagimoto, Y., Anthropometry of JASDF Personnel and its Applications for Human Engineering, Aeromedical Laboratory, Air Development and Test Wing, JASDF, Tokyo, 1990; Pheasant, S., Bodyspace: Anthropometry, Ergonomics and the Design of Work. Taylor & Francis Ltd., London, 1996. With permission. a Reference numbers refer to foot dimensions shown in Figure 4.3.
2
1
FIGURE 4.3 Graphic illustrations of foot dimension data. The reference numbers shown in these drawings relate to those listed in the first column of Table 4.4.
TAF-ER562X_WINGR-09-0401-C004.indd 108
11/9/10 2:23:09 AM
TAF-ER562X_WINGR-09-0401-C004.indd 109
50.5 (19.9) 12.7 (5.0) 9.7 (3.8) 35.3 (13.9) na 21.1 (8.3) 15.2 (6.0) 33.3 (13.1) na 8.1 (3.2) 19.3 (7.6) na na 16.8 (6.6) 3.8 (1.5) 3.4 (7.5)
Birth 66.0 (26.0) 15.2 (6.0) na 43.9 (17.3) 0.8 (0.3) 29.5 (11.6) 17.8 (7.0) 43.7 (17.2) 41.1 (16.2) 11.7 (4.6) 25.4 (10.0) 13.5 (5.3) na 20.8 (8.2) na 7.6 (16.7)
½ 74.9 (29.5) 17.5 (6.9) 13.2 (5.2) 47.2 (18.6) 1.3 (0.5) 32.0 (12.6) 20.3 (8.0) 47.5 (18.7) 44.5 (17.5) 13.0 (5.1) 30.5 (12.0) 15.2 (6.0) 9.7 (3.8) 24.4 (9.6) 6.4 (2.5) 10.1 (22.2)
1 80.8 (31.8) 18.3 (7.2) na 49.5 (19.5) 1.8 (0.7) 33.8 (13.3) 21.1 (8.3) 49.5 (19.5) 45.7 (18.0) 13.7 (5.4) 34.0 (13.4) na na 27.9 (11.0) na 11.4 (25.2)
1½ 86.4 (34.0) 19.1 (7.5) 14.0 (5.5) 49.8 (19.6) 2.5 (1.0) 34.5 (13.6) 22.4 (8.8) 50.8 (20.0) 46.2 (18.2) 14.5 (5.7) 37.1 (14.6) 16.0 (6.3) 10.7 (4.2) 31.2 (12.3) 6.6 (2.6) 12.6 (27.7)
2 91.2 (35.9) 19.3 (7.6) na 50.0 (19.7) 2.5 (1.0) 36.1 (14.2) 22.9 (9.0) 51.6 (20.3) 46.7 (18.4) 15.2 (6.0) 39.4 (15.5) na na 34.3 (13.5) na 13.6 (30.0)
2½
Source: Diffrient N., Tilley, A.R., and Bardagjy, J.C., Humanscale 1/2/3, MIT Press, Cambridge, MA, 1981. With permission. na, not available.
Body length Head length Head width Head circumference Neck length Trunk length Shoulder width Chest circumference Abdominal circumference Pelvic width Arm length Upper arm circumference Hand length Leg and thigh length Knee width Weight
Dimension
Age (years)
TABLE 4.5 Child Anthropometric Data, from Birth to Age 4, in cm (in.), and Weight, in kg (lbs.)
95.3 (37.5) 19.6 (7.7) 14.2 (5.6) 50.3 (19.8) 3.0 (1.2) 36.3 (14.3) 23.6 (9.3) 52.1 (20.5) 47.0 (18.5) 15.7 (6.2) 41.7 (16.4) 16.3 (6.4) 11.9 (4.7) 37.1 (14.6) 6.9 (2.7) 14.6 (32.2)
3
103.9 (40.9) 19.8 (7.8) 14.5 (5.7) 50.5 (19.9) 3.3 (1.3) 38.1 (15.0) 24.6 (9.7) 52.8 (20.8) 51.6 (20.3) 17.5 (6.9) 42.4 (16.7) 16.8 (6.6) 12.4 (4.9) 43.7 (17.2) 6.9 (2.7) 17.2 (38.0)
4
Anthropometry and Biomechanics 109
11/9/10 2:23:09 AM
110
Handbook of Human Factors in Medical Device Design Shoulder width
Arm circumference Chest circumference
Arm length
Head circumference
Neck length (chin to trunk)
Trunk length
Body length
Head length
Head width
Abdominal circumference Pelvic width
Knee width Knee pivot to floor
Leg and thigh length
Hand length
FIGURE 4.4 Reference dimensions for the child measurements presented in Table 4.5. (From Diffrient N., Tilley, A.R., and Bardagjy, J.C., Humanscale 1/2/3, MIT Press, Cambridge, MA, 1981. With permission.)
illustrated in Figure 4.4. Table 4.5 lists average anthropometric values for males, from birth to age four, at half-year increments. Within this age range, female dimensions differ little from males, making these data valid for both sexes. 4.1.1.4 Additional Data Sources The resources listed at the end of this chapter provide numerous sources of anthropometric data, both generally (for specific applications) and pertaining to specific populations. These should be searched for particular data not present in this chapter. PeopleSize 2000 (Open Ergonomics Ltd, Leicestershire, United Kingdom) is a device that provides anthropometric data in an electronic format. In summary, additional anthropometric data (beyond that contained in this chapter) should be consulted when designing devices: • For specific body parts, such the hands or feet • For use by special groups of individuals whose body measurements may differ from the adult population, such as infants, children, and adolescents
TAF-ER562X_WINGR-09-0401-C004.indd 110
11/9/10 2:23:09 AM
Anthropometry and Biomechanics
111
4.1.2 MOBILITY AND FUNCTIONAL MEASUREMENTS The unidimensional data (i.e., for single body parts) shown in Tables 4.2 through 4.5 are one source for designers of medical equipment and may cover many design applications. However, the designer must consider functional anthropometric data as well. The term “functional” involves considering movement of the body in two or more planes because people move, work, and use devices in three-dimensional space. 4.1.2.1 Mobility Human mobility (also called flexibility) is dependent on many factors, such as health, physical fitness, and age. This makes mobility data quite variable. However, one sample of 100 males and 100 females (Houy, 1983; Staff, 1983) provides a consistent basis for determining the ranges of motion for specific joints and body segments in the population. Table 4.6 presents these data, which are illustrated in Figure 4.5. An example of how to apply these mobility data to medical device designs involves the foot acivator switches mentioned previously. They must accommodate the smallest range of human motion. Table 4.5 shows that 5th-percentile males have the shortest ankle extension range (21 degrees). Thus, the switch should not require a movement range of more than 21 degrees. 4.1.2.2 Functional Work Body segment lengths and joint mobility during movement combine to determine functionality, particularly with regard to reaching. Figures 4.6 and 4.7 present reach envelopes recommended for seated and standing work tasks (Eastman Kodak Company, 1983). These reach envelopes are for individuals at the 5th percentile, thus representing the maximum horizontal and vertical distances at which medically related controls, equipment, and other objects should be placed from individuals to ensure use by a majority of the population. There are several software packages that use anthropometric data to produce virtual humans in three-dimensional space. The utility of these types of software is that they allow users to determine how humans of varying sizes will interact with equipment and workstations being designed. The more commonly used human modeling packages include Jack’s Task Analysis Toolkit (Siemens Product Lifecycle Management Software Inc., Plano, TX), Safework® Pro™ (Safework Inc., Montreal, Canada), ManneQuinPRO™ (NexGen Ergonomics, Montreal, Canada), and Ramsis and Anthropos (Human Solutions of North America, Inc., Troy, MI). The appropriate software package to use when designing a medical device depends on the specific application and working environment. 4.1.2.3 Visual Work Figure 4.8 integrates illustrations of anthropometric measurements needed for standing and seated visual work (Eastman Kodak Company, 1983). The medical device designer can use this information to accommodate the postural and visual needs of both large and small individuals. Designing medical devices to account for human mobility and function should do the following: • Integrate the known degrees of human joint flexibility. • Consider the user’s health, level of physical fitness, and age, as each can impact human mobility. • Understand that functional physical and visual work depends on knowledge of various body segment lengths operating together.
TAF-ER562X_WINGR-09-0401-C004.indd 111
11/9/10 2:23:10 AM
112
Handbook of Human Factors in Medical Device Design
TABLE 4.6 Male and Female Mobility Measurements, in degrees
Movement Neck
Shoulder
Elbow-forearm
Wrist
Hip
Knee
Ankle
Ventral flexion Dorsal flexion Right rotation Left rotation Flexion Extension Adduction Abduction Medial rotation Lateral rotation Flexion Supination Pronation Extension Flexion Adduction Abduction Flexion Adduction Abduction Medial rotation (prone) Lateral rotation (prone) Medial rotation (sitting) Lateral rotation (sitting) Flexion (standing) Flexion (prone) Medial rotation Lateral rotation Flexion Extension Adduction Abduction
Male
Female
Percentile
Percentile
5th
50th
95th
5th
50th
95th
25.0 38.0 56.0 67.5 161.0 41.5 36.0 106.0 68.5 16.0 122.5 86.0 42.5 47.0 50.5 14.0 22.0 95.0 15.5 38.0 30.5 21.5 18.0 18.0 87.0 99.5 14.5 21.0 18.0 21.0 15.0 11.0
43.0 56.5 74.0 77.0 178.0 57.5 50.5 123.5 95.0 31.5 138.0 107.5 65.0 62.0 67.5 22.0 30.5 109.5 26.0 59.0 46.0 33.0 28.0 26.5 103.5 117.0 23.0 33.5 29.0 35.5 25.0 19.0
60.0 74.0 85.0 85.0 193.5 76.0 63.0 140.0 114.0 46.0 150.0 135.0 86.5 76.0 85.0 30.0 40.0 130.0 39.0 81.0 62.5 46.0 43.0 37.0 122.0 130.0 35.0 48.0 34.0 51.5 38.0 30.0
34.0 47.5 67.0 64.0 169.5 47.0 37.5 106.0 94.0 19.5 135.5 87.0 63.0 56.5 53.5 16.5 19.0 103.0 27.0 47.0 30.5 29.0 20.5 20.5 99.5 116.0 18.5 28.5 13.0 30.5 13.0 11.5
51.5 70.5 81.0 77.0 184.5 66.0 52.5 122.5 110.5 37.0 148.0 108.5 81.0 72.0 71.5 26.5 28.0 125.0 38.5 66.0 44.5 45.5 32.0 33.0 113.5 130.0 31.5 43.5 23.0 41.0 23.5 24.0
69.0 93.5 95.0 90.0 199.5 85.0 67.5 139.0 127.0 54.5 160.5 130.0 99.0 87.5 89.5 36.5 37.0 147.0 50.0 85.0 58.5 62.0 43.5 45.5 127.5 144.0 44.5 58.5 33.0 51.5 34.0 36.5
Sources: Houy, D.A., Proceedings of the Human Factors Society 27th Annual Meeting, 377, 1983; Staff, K.R. Master’s thesis, Texas A&M University, 1983. With permission.
4.1.3 STRENGTH Designers of medical devices and work environments must also understand that body strength differs in several ways from one person to another and is dependent on many factors. When designing functions where user strength is an important issue, one should take into account those factors that may impact the user’s strength potential. These include (but
TAF-ER562X_WINGR-09-0401-C004.indd 112
11/9/10 2:23:10 AM
Anthropometry and Biomechanics
A
113
B B
A Ankle extension (A), abduction (B)
Ankle extension (A), flexion (B) Knee flexion, standing
Knee flexion (prone) Knee rotation medial (A), lateral (B)
Knee flexion, kneeling
A B
Hip flexion A A
B
B
Hip rotation, sitting, medial (A), lateral (B) Hip adduction (A), abduction (B)
B A
Hip rotation, prone, medial (A), lateral (B)
FIGURE 4.5 Reference illustrations for the mobility data presented in Table 4.6. (From Van Cott, H.P. and Kinkade, R.G., Human Engineering Guide to Equipment Design, U.S. Government Printing Office, Washington, DC, 1972. With permission.)
are not limited to) age, gender, health status, body part, body part position, direction of exertion, whether the exertion is applied statically or dynamically, and environmental issues. Because a multitude of factors determine an individual’s strength capacity, one should consult references included in this section for a summary of strength guidelines (Woodson, Tillman, and Tillman, 1992). This source includes information related to (1) ranges of hand strength for adults and children, (2) recommended upper limits for forces commonly used
TAF-ER562X_WINGR-09-0401-C004.indd 113
11/9/10 2:23:10 AM
114
Handbook of Human Factors in Medical Device Design cm Forward reach in.
20 5
30 10
40 15
20 30
75
25 20 50 15 10
25
Height above the work surface
5 in
cm
Centerline, 8–22 cm (3–9 in.) to right 38 cm (15 in.) to right of centerline 52 cm (2` in.) to right of centerline
FIGURE 4.6 Maximum forward reach capability for the right hand of a 5th-percentile female. Reach envelopes are given as a function of the hand moving away from the body’s centerline. (From Eastman Kodak Company, Ergonomic design for people at work, Volume I: Workplace, equipment, and environmental design and information transfer, Lifetime Learning Publications, Belmont, CA, 1983. With permission.) cm 25 Forward reach in. 10
50 20
30 84 79 200 74 69 175 64 59 150 54 49 125
Height above floor
44 39 100 34 29 75 24 in. cm Centerline, 15–30 cm (6–12 in.) to right 46 cm (18 in.) to right of centerline 61 cm (24 in.) to right of centerline 76 cm (30 in.) to right of centerline
FIGURE 4.7 Maximum one-handed and two-handed forward reach capability of a 5th-percentile person in a population of both males and females. Reach envelopes are given as a function of the hand moving away from the body’s centerline. (From Eastman Kodak Company, Ergonomic Design for People at Work, Volume I: Workplace, Equipment, and Environmental Design and Information Transfer, Lifetime Learning Publications, Belmont, CA, 1983. With permission.)
TAF-ER562X_WINGR-09-0401-C004.indd 114
11/9/10 2:23:11 AM
Anthropometry and Biomechanics
115
B Secondary visual displays E Viewing distance 50 cm (20 in.) Primary visual displays
A
Maximum height of workplace 165 cm (65 in.)
C 20º
D 10º
Top of primary displays 114 cm (45 in.)
Top of keyboard 76 cm (30 in.) Work surface 66 cm (26 in.)
(a) B Secondary visual displays
5
(20 mE c 0
in.
)
E 50 cm (20 in.)
C 20º
Maximum height of secondary displays 175 cm (69 in.)
Viewing Distance
A Primary visual displays
D 20º keyboard angle
Maximum height of primary displays 157 cm (62 in.) Top of keyboard 107 cm (42 in.) Work surface height 102 cm (40 in.)
(b)
FIGURE 4.8 Recommended dimensions for seated (a) and standing (b) workstations used for performing visual tasks. Primary displays refer to those that are most frequently monitored, while secondary displays are those less critical but still necessary for performing the task. (From Eastman Kodak Company, Ergonomic Design for People at Work, Volume I: Workplace, Equipment, and Environmental Design and Information Transfer, Lifetime Learning Publications, Belmont, CA, 1983. With permission.)
TAF-ER562X_WINGR-09-0401-C004.indd 115
11/9/10 2:23:12 AM
116
Handbook of Human Factors in Medical Device Design
on equipment and in control operations, and (3) lifting and carrying strengths. Strength data related to gender and the capabilities of the extremities are provided below. 4.1.3.1 Strength and Gender Differences U.S. Army data (U.S. Department of Defense, 1995) are some of the best sources of human strength information. When both males and females use medical equipment or when only male strength data are available, designers should consider the following general guidelines: • When the equipment requires exertions of the upper extremities, note that female arm and hand strength is slightly more than half (56.5%) that of males. • For force exertions required by the lower extremities (e.g., legs, feet), male strength capacity data should be reduced to nearly two-thirds (64.2%) to account for female strength abilities. • For devices involving trunk strength, female force requirements should be reduced to 66.0% of the limits used for males. 4.1.3.2 Strength of the Upper Extremity The shoulder and elbow joints together allow the upper extremity to move in many directions. However, strength capability is impacted by the positions of these joints and their direction of motion. Just as the proper design of a medical device’s physical characteristics often focuses on the smallest user (e.g., reach capability of a 5th-percentile female), the lowest strength capability of the user population should provide the basis for setting limits on the device’s force requirements. Thus, Table 4.7 presents maximum upper extremity muscle strength data for males at the 5th percentile, using Figure 4.9 as a graphical reference (adapted from U.S. Department of Defense, 1995). Here, strength data are listed for both the left and the right side, for pulling and pushing, and for upward, downward, inward, TABLE 4.7 Fifth-Percentile Male Muscle Strength of the Left and Right Arm, for Various Elbow Postures and Force Application Directions, in N (lbs.) 1 Degree of Elbow Flexion 180 degrees 150 degrees 120 degrees 90 degrees 60 degrees
2
3
4
5
6
7
Pull
Push
Up
Down
In
Out
Left
Right
Left
Right
Left
Right
Left
Right
Left
Right
222 (50) 187 (42) 151 (34) 142 (32) 116 (26)
231 (52) 249 (56) 187 (42) 165 (37) 107 (24)
187 (42) 133 (30) 116 (26) 98 (22) 98 (22)
222 (50) 187 (42) 160 (36) 160 (36) 151 (34)
40 (9) 67 (15) 76 (17) 76 (17) 67 (15)
62 (14) 80 (18) 107 (24) 89 (20) 89 (20)
58 (13) 80 (18) 93 (21) 93 (21) 80 (18)
76 (17) 89 (20) 116 (26) 116 (26) 89 (20)
58 (13) 67 (15) 89 (20) 71 (16) 76 (17)
89 (20) 89 (20) 98 (22) 80 (18) 89 (20)
Left Right 36 (8) 36 (8) 45 (10) 45 (10) 53 (12)
62 (14) 67 (15) 67 (15) 71 (16) 76 (17)
Source: Department of Defense, Handbook for Human Engineering Design Guidelines, MIL-HDBK-759C, 1995, Navy Publishing and Printing Office, Philadelphia, PA. With permission.) Note: Numbers in the first row of this table refer to the diagrams in Figure 4.9.
TAF-ER562X_WINGR-09-0401-C004.indd 116
11/9/10 2:23:14 AM
Anthropometry and Biomechanics
117
2
3
1 180°
150° 60°
90°
4
7
5
6
120°
FIGURE 4.9 Reference strength exertions for the data presented in Table 4.7. (From U.S. Department of Defense, Handbook for Human Engineering Design Guidelines, MIL-HDBK-759C, 1995, Navy Publishing and Printing Office, Philadelphia, PA. With permission.)
and outward motions. Data are presented for the force application directions at five different elbow flexion angles. Note that comparable female strength data are not available. Thus, the values in Table 4.7 should be reduced by 50% to 60% (as stated in the previous section) for equipment designed for use by females. As an example, consider the downward pull force required to lower a heart monitor attached to a moveable swing arm. Assuming that the elbow flexion angle of most users would be 150 degrees during this task, Table 4.7 shows that 5th-percentile males have a right-arm force capability of 89 N. Given that females also are likely to perform this task, a force reduction to 50.3 N (56.5% of 89 N) is reasonable to account for most possible users. In summary, medical devices that require force application of the upper extremity should account for the following: • • • •
The population using the device (e.g., males, females, or a combination) Human strength ability and differences due to elbow flexion Direction of the force application The hand used to apply the force
4.1.3.3 Strength of the Fingers and Hands Finger and hand strength data are relevant in the design and handling of particular medical equipment. Table 4.8 presents such data for two groups of males—students and industrial workers (Kroemer, Kroemer, and Kroemer-Elbert, 1994). 4.1.3.4 Strength of the Feet As with other body parts, foot strength is a function of age, gender, direction of motion, and leg position. Table 4.9 shows data related to the required strength for pressing and lifting either a bar or a pedal with the foot (Consumer and Competition Policy Directorate, Department of Trade and Industry, 2002b). Foot strength data are presented for both males and females and as a function of age.
TAF-ER562X_WINGR-09-0401-C004.indd 117
11/9/10 2:23:14 AM
118
Handbook of Human Factors in Medical Device Design
TABLE 4.8 Finger and Hand Forces That Two Groups of Males Can Exert, in a Variety of Hand/Handle Configurations Mean Forces (Standard Deviations) Exerted by Male Students (*) or Male Machinists, in N Digit 1 Digit 2 Digit 3 Hand/Handle Coupling and Description (Thumb) (Index) (Middle) Digit touch
Palm touch
Hook grip
One digit touches an object. – Touch perpendicular to 84 (33)* 43 (14)* extended digit 131 (42) 70 (17) – Digits 2–5 all pressed on one bar – Tip force, as in 30 (12)* typing 65 (12) Some part of the palm or hand touches the object.
Digit 4 (Ring)
Other
36 (13)* 30 (13)* 25 (10)* 76 (20) 57 (17) 55 (16) 162 (33) 29 (11)* 69 (22)
23 (9)* 50 (11)
19 (7)* 46 (14) 233 (65)
One finger or several fingers 61 (21)* 49 (17)* 48 (19)* 38 (13)* 34 (10)* hook(s) onto a ridge or 118 (24) 89 (29) 104 (26) 77 (21) 66 (17) handle (used where thumb counterforce is not needed). – All digits combined 108 (39)* 252 (63)
Tip pinch
– The thumb top opposes one fingertip.
Pad pinch
– Thumb pad opposes (On the palmer pad of one digits 2 finger (or the pads of and 3) 63 (12)* 61 (16)* 41 (12)* several fingers) near the 85 (16)* 34 (7) 70 (15) 54 (15) tops. 95 (19) – Thumb opposes the 98 (13)* (radial) side of the 112 (16) index finger.
Side pinch
Digit 5 (Little)
Power grasp The total inner hand surface grasps the (often cylindrical) handle, which runs parallel to the knuckles and generally protrudes on one or both sides from the hand.
50 (14)* 59 (15)
53 (14)* 38 (7)* 63 (16) 44 (12)
28 (7)* 30 (6)
31 (9)* 34 (7)
318 (61)* 366 (53)
Source: Kroemer, K.H.E., Kroemer, H.B., and Kroemer-Elbert, K.E. Ergonomics: How to Design for Ease and Efficiency, Prentice Hall, Englewood Cliffs, NJ, 1994. With permission.
TAF-ER562X_WINGR-09-0401-C004.indd 118
11/9/10 2:23:15 AM
TAF-ER562X_WINGR-09-0401-C004.indd 119
31–50
21–30
16–20
11–15
6–10
2–5
Age (years)
Female
Male
Female
Male
Female
Male
Female
Male
Female
Male
Female
Male
Gender
Lift Press Lift Press Lift Press Lift Press Lift Press Lift Press Lift Press Lift Press Lift Press Lift Press Lift Press Lift Press
Activity 46.4 (10.4) 93.5 (21.0) 48.7 (10.9) 84.8 (19.1) 90.0 (20.2) 169.9 (38.2) 133.4 (30.0) 232.9 (52.4) 223.9 (50.3) 466.4 (104.8) 178.7 (40.2) 428.4 (96.3) 327.1 (73.5) 657.7 (147.9) 180.6 (40.6) 403.6 (90.7) 320.0 (71.9) 632.7 (142.2) 228.3 (51.3) 469.0 (105.4) 309.4 (69.6) 673.0 (151.3) 210.7 (47.4) 519.2 (116.7)
Mean 15.6 (3.5) 29.5 (6.6) 16.9 (3.8) 33.4 (7.5) 35.4 (8.0) 40.5 (9.1) 41.6 (9.4) 124.7 (28.0) 69.0 (15.5) 126.0 (28.3) 50.5 (11.4) 26.7 (6.0) 56.5 (12.7) 129.0 (29.0) 80.2 (18.0) 116.7 (26.2) 99.4 (22.3) 178.4 (40.1) 49.6 (11.2) 64.6 (14.5) 49.2 (11.1) 137.9 (31.0) 59.3 (13.3) 189.7 (42.6)
Standard Deviation
Bar
18.2 (4.1) 44.3 (10.0) 18.2 (4.1) 44.3 (10.0) 58.1 (13.1) 100.9 (22.7) 73.0 (16.4) 95.8 (21.5) 117.9 (26.5) 250.2 (56.2) 147.8 (33.2) 394.4 (88.7) 242.4 (54.5) 476.8 (107.2) 102.9 (23.1) 296.6 (66.7) 147.8 (33.2) 399.5 (89.8) 142.8 (32.1) 389.2 (87.5) 247.4 (55.6) 461.3 (103.7) 127.8 (28.7) 209.0 (47.0)
Minimum 63.0 (14.2) 137.0 (30.8) 63.0 (14.2) 147.3 (33.1) 142.8 (32.1) 203.9 (45.8) 192.6 (43.3) 466.5 (104.9) 342.1 (76.9) 636.4 (143.1) 267.4 (60.1) 466.5 (104.9) 416.8 (93.7) 898.9 (202.1) 302.2 (67.9) 590.0 (132.6) 441.8 (99.3) 873.2 (196.3) 277.3 (62.3) 528.2 (118.7) 391.9 (88.1) 873.2 (196.3) 312.2 (70.2) 754.8 (169.7)
Maximum
TABLE 4.9 Foot Strength Capabilities for Pressing and Lifting a Bar and a Pedal, in N (lbs.)
42.6 (9.6) 79.2 (17.8) 33.2 (7.5) 80.3 (18.1) 72.0 (16.2) 151.4 (34.0) 98.0 (22.0) 162.7 (36.6) 142.8 (32.1) 280.4 (63.0) 118.9 (26.7) 197.7 (44.4) 172.7 (38.8) 407.6 (91.6) 98.9 (22.2) 216.3 (48.6) 155.6 (35.0) 341.4 (76.7) 136.1 (30.6) 227.1 (51.1) 200.4 (45.0) 329.2 (74.0) 129.7 (29.2) 238.0 (53.5)
Mean 12.3 (2.8) 35.4 (8.0) 13.0 (2.9) 33.6 (7.6) 24.3 (5.5) 38.9 (8.7) 36.3 (8.2) 64.1 (14.4) 45.9 (10.3) 78.4 (17.6) 44.2 (9.9) 107.6 (24.2) 52.0 (11.7) 177.7 (39.9) 35.1 (7.9) 128.6 (28.9) 33.7 (7.6) 137.9 (31.0) 31.5 (7.1) 66.3 (14.9) 39.4 (8.9) 130.0 (29.2) 24.4 (5.5) 106.8 (24.0)
Standard Deviation 23.2 (5.2) 18.5 (4.2) 23.2 (5.2) 39.1 (8.8) 48.1 (10.8) 106.1 (23.9) 43.1 (9.7) 75.2 (16.9) 48.1 (10.8) 147.3 (33.1) 83.0 (18.7) 111.2 (25.0) 117.9 (26.5) 167.8 (37.7) 63.0 (14.2) 85.5 (19.2) 107.9 (24.3) 147.3 (33.1) 83.0 (18.7) 167.8 (37.7) 157.7 (35.5) 173.0 (38.9) 92.9 (20.9) 121.5 (27.3)
Minimum
Pedal
continued
58.1 (13.1) 126.7 (28.5) 63.0 (14.2) 131.8 (29.6) 102.9 (23.1) 198.7 (44.7) 147.8 (33.2) 276.0 (62.0) 197.6 (44.4) 384.1 (86.3) 182.6 (41.0) 373.8 (84.0) 252.4 (56.7) 667.2 (150.0) 152.7 (34.3) 415.0 (93.3) 212.5 (47.8) 564.3 (126.9) 177.7 (39.9) 332.6 (74.8) 282.3 (63.5) 579.7 (130.3) 167.7 (37.7) 435.6 (97.9)
Maximum
Anthropometry and Biomechanics 119
11/9/10 2:23:16 AM
TAF-ER562X_WINGR-09-0401-C004.indd 120
Female
Male
Female
Male
Female
Male
Female
Male
Gender
Lift Press Lift Press Lift Press Lift Press Lift Press Lift Press Lift Press Lift Press
Activity 279.8 (62.9) 592.6 (133.2) 145.3 (32.7) 351.9 (79.1) 245.4 (55.2) 559.12 (125.7) 155.6 (35.0) 337.4 (75.8) 182.6 (41.0) 417.9 (93.9) 139.7 (31.4) 357.2 (80.3) 162.7 (36.6) 296.6 (66.7) 117.9 (26.5) 256.4 (57.6)
Mean 78.7 (17.7) 162.5 (36.5) 42.4 (9.5) 92.9 (20.9) 42.1 (9.5) 126.1 (28.3) 39.3 (8.8) 94.7 (21.3) 53.4 (12.0) 112.8 (25.4) 38.8 (8.7) 71.6 (16.1) 49.4 (11.1) 72.8 (16.4) 33.3 (7.5) 107.0 (24.1)
197.6 (44.4) 435.6 (97.9) 102.9 (23.1) 245.1 (55.1) 207.6 (46.7) 451.0 (101.4) 112.9 (25.4) 106.1 (23.9) 73.0 (16.4) 296.6 (66.7) 38.1 (8.6) 214.2 (48.2) 127.8 (28.7) 245.1 (55.1) 63.0 (14.2) 157.6 (35.4)
Minimum 386.9 (87.0) 811.4 (182.4) 202.6 (45.5) 466.5 (104.9) 317.2 (71.3) 770.2 (173.1) 252.4 (56.7) 512.8 (115.3) 242.4 (54.5) 605.5 (136.1) 182.6 (41.0) 481.9 (108.3) 197.6 (44.4) 348.0 (78.2) 147.8 (33.2) 430.4 (96.8)
Maximum 176.4 (39.7) 215.5 (48.4) 109.1 (24.5) 243.8 (54.8) 160.7 (36.1) 267.7 (60.2) 120.7 (27.1) 183.3 (41.2) 131.1 (29.5) 205.0 (46.1) 128.6 (28.9) 215.0 (48.3) 132.8 (29.9) 211.7 (47.6) 107.9 (24.3) 151.4 (34.0)
Mean 31.9 (7.2) 41.0 (9.2) 31.6 (7.1) 45.2 (10.2) 21.6 (4.9) 66.4 (14.9) 22.0 (4.9) 57.7 (13.0) 38.6 (8.7) 95.7 (21.5) 25.6 (5.8) 67.9 (15.3) 70.4 (15.8) 91.0 (20.5) 52.1 (11.7) 25.6 (5.8)
Standard Deviation 137.8 (31.0) 162.7 (36.6) 78.0 (17.5) 193.6 (43.5) 142.8 (32.1) 173.0 (38.9) 78.0 (17.5) 70.0 (15.7) 63.0 (14.2) 85.5 (19.2) 92.9 (20.9) 111.2 (25.0) 83.0 (18.7) 147.3 (33.1) 48.1 (10.8) 126.7 (28.5)
Minimum
Pedal
202.6 (45.5) 260.5 (58.6) 152.7 (34.3) 291.4 (65.5) 197.6 (44.4) 342.9 (77.1) 147.8 (33.2) 276.0 (62.0) 197.6 (44.4) 399.5 (89.8) 162.7 (36.6) 378.9 (85.2) 182.6 (41.0) 276.0 (62.0) 187.6 (42.2) 193.6 (43.5)
Maximum
Source: Consumer and Competition Policy Directorate, Department of Trade and Industry, Strength Data for Design Safety—Phase 2, London, United Kingdom, 2002. With permission.
81–90
71–80
61–70
51–60
Age (years)
Standard Deviation
Bar
TABLE 4.9 (CONTINUED) Foot Strength Capabilities for Pressing and Lifting a Bar and a Pedal, in N (lbs.)
120 Handbook of Human Factors in Medical Device Design
11/9/10 2:23:16 AM
Anthropometry and Biomechanics
121
4.1.4 SPECIAL CONSIDERATIONS Individuals with a particular temporary or permanent disability also use many medical devices. Such populations include children, older adults, wheelchair users, or persons with specific physical limitations. The reader is urged to consult the references listed in this chapter as well as Chapter 2, “Basic Human Abilities” for sources of anthropometric information for these and other special populations. Note that not all physical measurements of persons with disabilities are necessarily different from those for populations without disabilities. For instance, anthropometric data from the United Kingdom (Consumer and Competition Policy Directorate, Department of Trade and Industry, 2002a) found no significant differences in hand dimensions between those with disabilities (e.g., individuals whose impairments affected their reaching, dexterity, or manipulation capabilities) and those without disabilities. However, handgrip span (i.e., the distance between the thumb and little finger) differed significantly between groups. These measurements are shown in Table 4.10. TABLE 4.10 Handgrip Spans of Dexterity-Impaired and Unimpaired Males and Females, in cm (in.) Handgrip Span of Individuals with a Dexterity Impairment
Males Left hand Right hand Females Left hand Right hand Combined Left hand Right hand
Mean
Standard Deviation
Minimum
Maximum
7.2 (2.8) 6.7 (2.6)
2.3 (0.9) 2.2 (0.9)
2.3 (0.9) 1.6 (0.6)
14.1 (5.6) 11.2 (4.4)
6.1 (2.4) 6.2 (2.4)
2.1 (0.8) 2.0 (0.8)
0.5 (0.2) 0.8 (0.3)
12.9 (5.1) 11.1 (4.4)
6.4 (2.5) 6.3 (2.5)
2.2 (0.9) 2.1 (0.8)
0.5 (0.2) 0.8 (0.3)
14.1 (5.6) 11.2 (4.4)
Handgrip Span of Individuals without a Dexterity Impairment
Males Left hand Right hand Females Left hand Right hand Combined Left hand Right hand
Mean
Standard Deviation
Minimum
Maximum
8.5 (3.3) 8.4 (3.3)
2.4 (0.9) 2.0 (0.8)
4.4 (1.7) 4.6 (1.8)
14.2 (5.6) 13.1 (5.2)
7.1 (2.8) 7.3 (2.9)
1.9 (0.7) 1.8 (0.7)
2.8 (1.1) 3.2 (1.3)
12.1 (4.7) 11.6 (4.6)
7.6 (3.0) 7.7 (3.0)
2.2 (0.9) 2.0 (0.8)
2.8 (1.1) 3.2 (1.3)
14.2 (5.6) 13.1 (5.2)
Source: Consumer and Competition Policy Directorate, Department of Trade and Industry, Specific Anthropometric and Strength Data for People with Dexterity Disability, London, United Kingdom, 2002. With permission
TAF-ER562X_WINGR-09-0401-C004.indd 121
11/9/10 2:23:16 AM
122
Handbook of Human Factors in Medical Device Design 1200
270 Trunk extension
800
225
180
Trunk flexion Handgrip
600
135 Knee extension
400
Strength (lbs.-ft.)
Strength (N)
1000
Elbow flexion 90
31–35
51–55 Age (years)
71–75
FIGURE 4.10 Average strength for three age-groups of males for five different muscle groups. (From Viitasalo, J.T. et al., Ergonomics, 28, 1566, 1985. With permission.)
One should consider the physical limitations of individuals in special populations, particularly when designing equipment requiring physical force exertions. Aging affects muscle groups differently. An example of aging’s effects on isometric strength (Viitasalo et al., 1985) is shown in Figure 4.10. On average, males in their early 50s had about 80% of the strength of those in their early 30s. By the early 70s, average strength declines to about 60% of what it was four decades earlier. Although data were gathered only for males in this study, female strength capabilities also are known to decrease with age (see Chapter 18, “Home Health Care”). When a medical device is to be used by individuals in a special population, the designer should do the following: • Use anthropometric data that are reflective of any unique physical capabilities and limitations. • Recognize that users with a disability may not differ from those without a disability on every physical aspect.
4.1.5 DESIGN GUIDELINES FOR BODY DIMENSION DATA If anthropometric information is inappropriately applied during the design of medical devices, fewer individuals will be able to use the device safely and effectively. The following is a list of alternative approaches to assist in correctly applying anthropometric data. The approach used depends on the device being designed. 4.1.5.1 Designing for Population Extremes The first approach involves designing medical devices to accommodate users on both ends of the “size spectrum” with regard to their physical body dimensions. For example, those persons at the lower end of the spectrum (e.g., the 5th percentile) must be able to reach the necessary controls in a workstation or wrap their fingers around tool handles. These are
TAF-ER562X_WINGR-09-0401-C004.indd 122
11/9/10 2:23:17 AM
Anthropometry and Biomechanics
123
tasks that can be performed more easily by those having larger body part sizes (i.e., those above the 5th percentile). Similarly, persons at the higher end of the size spectrum (e.g., the 95th percentile) must be able to fit into a workstation, which can be affected by, for example, the legroom at seated workstations or the head clearance when walking underneath monitoring equipment. By creating medical devices to accommodate persons on each end of this anthropometry spectrum, the designer ensures that a majority of individuals (i.e., those falling between these two ranges of body size extremes) will be able to use these devices effectively and safely. For device or workstation dimensions that are critical to its operation, the designer should determine which of these two groups needs to be accommodated. Consider, for example, safety cabinet hoods in which pharmacy medications are prepared. Users often need to access devices at the rear of the cabinet. Thus, that dimension should accommodate individuals with a short arm reach. By default, those with longer arms also will be able to reach to this distance. Similarly, the foot cutout at the bottom of the safety cabinet should be designed to fit those with the largest feet. Those with smaller feet dimensions will naturally be accommodated. Guideline 4.1: Physical Accommodation of Most Users Medical devices should physically accommodate the widest practical range of users. A target range from the 5th-percentile female to the 95th percentile male is considered the nominally acceptable range. However, a target range from the 2.5th- or 1st-percentile female to the 97.5th- or 99th-percentile male is preferable, noting that the broader range of accommodation might require only slight adjustments in medical device proportioning. This is particularly true for hand-operated devices, such as surgical instruments for which an extra 0.1-inch-wide grip might increase handling comfort dramatically for many users.
Guideline 4.2: Avoid Use at Ability Extremes Although an anthropometric analysis might identify maximums for parameters such as reach and strength, medical devices should not require users to function at the extremes of their ability. Rather, medical devices should place only moderate physical demands on users to ensure comfortable interactions. Therefore, a medical device’s weight should be kept well below the limit of a small female’s lifting capabilities, for example.
Guideline 4.3: Accomodate Use with Protective Clothing Anthropometric analyses should account for users’ clothing, particularly protective gear (e.g., face shields, gloves, boots) (see Chapter 3, “Environment of Use”).
Guideline 4.4: Design for Both Genders Unless intended for exclusive use by either males or females, medical devices should be designed for equally effective use by both genders.
Guideline 4.5: Confirmation of Physical Accommodation While designs may be based on anthropometric analyses, physical accommodation should be confirmed by checking the design’s fit with users who have average and extreme physical characteristics (e.g., particularly tall, particularly short, and average height individuals). This approach recognizes the complexities of human movement in three dimensions that
TAF-ER562X_WINGR-09-0401-C004.indd 123
11/9/10 2:23:17 AM
124
Handbook of Human Factors in Medical Device Design
might not be fully or accurately addressed through anthropometric analyses, even those employing computer-based human models that permit sophisticated, three-dimensional analyses.
4.1.5.2 Designing for the Average User Designers sometimes use an approach to develop a device based on the average (i.e., mean, or 50th percentile) body size values on one or more body dimensions since anthropometric data tables commonly contain this information. In some cases this approach is appropriate, such as for noncritical design elements (e.g., the height of a latch to gain access to an X-ray machine). However, if particular body size elements are critical to the safe and proper design of a medical device, then designing for the “average person” is inappropriate. Only a very small percentage of individuals (male or female) are actually “average” on multiple anthropometric dimensions. For example, just because someone is “average” in terms of her standing height does not mean she also will have average arm length or average hand breadth. Thus, designing medical tools or equipment using “average” (or 50th percentile) data on several body dimensions likely will produce a device that cannot be used easily by most people. Guideline 4.6: Multiple Sizes of Devices Where practical, medical devices (e.g., certain surgical instruments) should be offered in varying sizes to accommodate diverse users. This approach can avoid undesirable compromises associated with producing a “one-size-fits-all” device that ultimately is sized optimally only for a subset of users. For example, it would be better to produce three different size clamps with variously spaced and sized finger openings instead of just one that is ill-suited to people with particularly small or large hands.
4.1.5.3 Designing for Adjustability As the data tables contained in this chapter show, human body sizes are inherently variable in a population. This is due to diversity, for example, in age, gender, ethnicity, and health status. Furthermore, personal preference often affects how individuals use equipment. A classic example of this is in chair design. Two individuals, even those having the same body size and shape, may prefer to sit at the same laboratory diagnostic workstation in very different ways. Thus, even when the designer integrates anthropometric data into a device’s design, users themselves may handle these devices differently. Factors related to body size and user preference interact in innumerable ways. This suggests that, where possible, one should design medical devices and workspaces so that they are adjustable. In other words, a device designed so that its physical components can be easily raised/lowered, widened/narrowed, or otherwise manipulated will accommodate not only the desired range of physical body sizes in the target population but also users’ personal preferences. Many medical devices need to be adjustable to accommodate the full range of users, whether they are patients, clinicians, or technicians. One example would be an adjustable arm on a television monitor used to display images as part of an endoscopic workstation, such as one used during a colonoscopy procedure. The gastroenterologist can move the arm such that he can easily see the images of the colon on the monitor as he manipulates the endoscope.
TAF-ER562X_WINGR-09-0401-C004.indd 124
11/9/10 2:23:17 AM
Anthropometry and Biomechanics
125
Thus, the approach for accommodating as large a segment of medical devices users as possible should follow these guidelines: Guideline 4.7: Don’t Use Population Averages Designers may make use of anthropometric data to guide conceptual development but should not use data on population averages in the detailed design, particularly the design of elements critical to operational effectiveness, safety, and comfort.
Guideline 4.8: Accomodate User Extremes Designers should size medical devices to accommodate users at the extremes. For example, workspaces should be spacious enough to accommodate large individuals, while controls should be placed sufficiently low on a control panel to be within the reach of small individuals.
Guideline 4.9: Incorporate Adjustability As necessary, medical devices should include adjustment mechanisms (e.g., a telescoping headrest) to accommodate users’ needs and preferences.
Guideline 4.10: Controls Placed Within Reach Controls should be positioned where users can reach them without overextending their arms (i.e., moving their shoulder joint forward from its normal position) (see Chapter 7, “Controls”). At a minimum, controls should be placed within the users’ functional reach, which may involve some stretching and twisting.
Guideline 4.11: Either Handed Use Only in specific cases should medical devices be designed for one-handed rather than twohanded use. Device design, for example, should facilitate use by people who might have an injured hand, joint disorder in one hand, or amputated arm/hand/fingers.
4.1.5.4 Deriving Missing Data Sometimes needed anthropometric dimensions for a medical device are not available for a particular user group. Thus, a designer will need to determine the appropriate values. There are several ways to accomplish this, as detailed below. Designers commonly derive needed body dimension estimates by conveniently measuring a few colleagues or individuals thought to be representative of the population of intended users. On the surface, this technique appears to be simple and cost-effective. However, one should avoid this approach especially if the measurement in question is critical to the use and safety of the device. For example, imagine that the engineer who is designing an MRI system decides to use his own body dimensions to specify the machine’s opening but that this particular design engineer was a 5th-percentile male. This design would be problematic since the opening would be too small for larger patients. An alternative method of determining a needed physical dimension is to measure the dimension in a representative sample of users, sufficient to account for human body size variability. This approach often is costly and time consuming, but it should be employed if the needed measurement is critical to perfecting the operation of the medical device or workstation. For example, spinal implants, such as rods and braces, are critical devices, and their designs demand accurate dimensions related to spinal column anatomy and the variability in size found in the population.
TAF-ER562X_WINGR-09-0401-C004.indd 125
11/9/10 2:23:17 AM
126
Handbook of Human Factors in Medical Device Design
Guideline 4.12: Obtaining Missing Data Where practical, derive missing data from a representative user population rather than a more convenient but less representative one. Remember that devices often initially intended for one user population (e.g., American surgeons) may ultimately be used widely by a broader population (e.g., international use by many types of clinicians).
A more common approach to determining the appropriate data involves estimating a needed dimension based on known measurements. There are several computational techniques used for this purpose. The assumption with the ratio scaling technique is that many individual body sizes (such as various segment lengths) are roughly proportional to one another. For example, determining the sleeve length for surgical gowns to be used in Europe requires arm length data. If these data are only available for a U.S. population, the ratio of 5th-, 50th-, and 95th-percentile arm lengths to standing heights for this population could be used to estimate European arm lengths, given their known standing heights (as this is a common measurement taken for many groups). The regression equation technique assumes that there is a linear relationship between two values. For example, suppose that arm length data (used to design surgical gown sleeve lengths) exist for the 5th, 50th, and 95th percentiles of a population but that they are needed for the 1st and 99th percentiles. Applying regression techniques to the known data would be used to extrapolate to the unknown (i.e., 1st and 99th percentile) values. The designer should evaluate the assumptions of this approach before using it extensively. Finally, in the probability statistics approach, a measurement for an entire population is estimated from data gathered on a small sample. Suppose that, to determine surgical gown sleeve length for Europeans, 15 such individuals are measured. Clearly, this number of people will not fully represent all Europeans. However, probability equations (from the average and variance data in the small sample) can be used to estimate population values from this sample. The specifics of these approaches can be found in any number of anthropometric texts, including Kroemer et al. (1997). A more in-depth review of these techniques will help the designer choose the appropriate method and develop an accurate estimate of a needed physical measurement. In order to derive needed missing data most accurately, follow these guidelines: Guideline 4.13: Larger Sample Size Preferred Use anthropometric data based on large sample sizes because such a data set is more likely to contain data whose ranges are reflective of that population.
Guideline 4.14: Largest Practical Study If the needed anthropometric data are not available, cannot be reliably derived from existing sources, and are essential to producing an effective design, conduct the largest practical anthropometric study to collect the necessary data and ensure its reliability.
Guideline 4.15: Most Pertinent User Population Designers should use anthropometric data pertaining to the most pertinent user population. For example, when designing a telemetry workstation, it is better to use data derived from measurements of nurses than military airline pilots or commercial air traffic controllers.
TAF-ER562X_WINGR-09-0401-C004.indd 126
11/9/10 2:23:17 AM
Anthropometry and Biomechanics
127
4.2 CASE STUDIES IN ANTHROPOMETRIC DESIGN 4.2.1 VIEWING ANGLE DETERMINATION FOR A CARDIAC OUTPUT MONITOR Designers seeking to improve a cardiac output monitor needed to select an LCD display that could accommodate an appropriate range of viewing angles. This range was based on the expected user population height extremes and the expected range of viewing positions in a typical hospital environment. Designers conducted field visits to hospitals as part of a contextual inquiry to collect the necessary data. They combined the field observational data with anthropometric data for 5th- and 95th-percentile users representing population extremes (for more details, refer to Chapter 9, “Visual Displays”). 4.2.1.1 Analysis One of the limitations of LCDs is that image quality typically is poor (because of inadequate luminance, contrast, or glare) when they are viewed at off-normal viewing angles in both horizontal and vertical directions. Horizontal viewing angle is that angle between the monitor and the user’s (left or right) position from it. Users generally stand directly in front of the display when interacting with the cardiac output monitor, except in certain situations (e.g., viewing readings while inserting a catheter). The worst-case horizontal angles appear to occur most often when viewing the cardiac output monitor from a distance. For example, at a patient’s bedside, the monitor often is set up so that the display faces the foot of the bed. The largest horizontal off-normal viewing angles (±45 degrees) occur when attempting to view the bedside monitor from the doorway of the patient’s room. The cardiac output monitor sometimes is located on a shelf or adjustable rack that places it at or near the user’s eye height. In this situation, the off-normal (upward or downward) angle does not pose a significant problem. Although there are a wide variety of other installation locations for these monitors, those affecting vertical viewing positions fall into two categories—high-mount position (above eye height) and low-mount position (below eye height). 4.2.1.2 High-Mount Position The top of the cardiac output monitor’s viewing area can be mounted above the user’s eye level. For example, in an operating room, the monitor may be stacked atop other equipment, such as an anesthesia workstation. The angle at which the user must look up (equivalent to the downward angle from the display normal line) will be determined by display height, user eye height, and distance from the user’s eye to the display. Figure 4.11 illustrates this type of installation. 4.2.1.3 Calculated Viewing Angles for the High-Mount Position A standard design approach for viewing angle is to accommodate the 5th- to 95thpercentile male and female U.S. population, or approximately 90% of the overall population. Figure 4.11 shows that the worst-case situation will be with the 5th-percentile user. Table 4.11 presents eye heights of 5th-percentile male and female U.S. adults (Pheasant, 1996) in standing and seated positions, where seating is assumed to be on 61.0 or 76.2 cm (24–30 inch) stools. “Difference” refers to the distance between the eye height of the specified population and the height of the display. Viewing angles, shown in the right-most
TAF-ER562X_WINGR-09-0401-C004.indd 127
11/9/10 2:23:17 AM
128
Handbook of Human Factors in Medical Device Design Viewing distance 0° Elevation Q2
Q2 height
Off-normal viewing angle
User’s eye position (standing or seated)
Eye height
Floor
FIGURE 4.11 Viewing angle with heart output monitor in a high vertical position. Q2 is the monitor display being evaluated.
column of Table 4.11, are calculated on the basis of the following assumptions (note that the height of the display was based on the highest location found in the field observations): • Display height (top of display): 200.7 cm (79 in.) • Assumed viewing distance: 45.7 cm (18 in.) 4.2.1.4 Low-Mount Position In many situations, such as in critical care units, the cardiac output monitor is placed where the user must look down, such as when the device is placed on a bedside stand. Again, viewing angle is determined by the height of the display, the user’s eye height, and the user’s distance from the display. In this situation (Figure 4.12), the largest viewing angle is required when the display is mounted at its lowest level and the user’s eye position is relatively high (e.g., a taller standing user). 4.2.1.5 Calculated Viewing Angle for the Low-Mount Position Table 4.12 shows calculated viewing angles based on the following assumptions, given that the tallest user in the population is represented by the 95th-percentile U.S. male (Pheasant, 1996): • Display height (bottom of display): 88.3 cm (34.8 in.) • Assumed viewing distance: 45.7 cm (18.0 in.) TABLE 4.11 Calculated (Off-Normal) Viewing Angles with High Vertical Position, in cm (in.) User 5th-percentile female—standing 5th-percentile female—stool (30 in.) 5th-percentile female—stool (24 in.) 5th-percentile male—standing 5th-percentile male—stool (30 in.) 5th-percentile male—stool (24 in.)
TAF-ER562X_WINGR-09-0401-C004.indd 128
User Eye Height
Difference
Viewing Angle (°)
142.0 (55.9) 145.2 (57.2) 130.0 (51.2) 159.5 (67.8) 150.2 (59.1) 135.0 (53.1)
58.7 (23.1) 55.5 (21.9) 70.7 (27.8) 41.2 (16.2) 50.5 (19.9) 65.7 (25.9)
52.1 50.5 57.1 42.0 47.8 55.2
11/9/10 2:23:18 AM
Anthropometry and Biomechanics
129 Viewing distance
Offnormal viewing angle
Q2
Eye height
Display height
0° Elevation
Floor
FIGURE 4.12 Viewing angle with heart output monitor in a low-mount position. Q2 is the monitor display being evaluated.
4.2.1.6 Viewing Angles, Summary, and Implications Measurements and derivations of horizontal and vertical viewing angles are based on the following assumptions: • The sample of observed users is representative of target customer users. • Typical and worst-case observations are representative of the target viewing scenarios. • User populations are represented by 5th- to 95th-percentile U.S. male and female populations. • The tasks being performed are not so critical that larger percentages of the population (i.e., greater than the central 90%) must be accommodated. With these assumptions, the viewing envelope for the cardiac output monitor and display (required maximum viewing angles) were calculated to be as follows: • Horizontal viewing angle: ±45 degrees • Viewing angle (viewed from below the display): –5 7degrees • Viewing angle (viewed from above the display): +64 degrees
4.2.2 KEYBOARD HEIGHT IN A DIAGNOSTIC SYSTEM WORKSTATION A PC-based workstation must be designed for use with a clinical chemistry analyzer diagnostic medical system. It is well known that laboratory technician operators range from small females to very tall males. The design required a determination of the keyboard TABLE 4.12 Calculated (Off-Normal) Viewing Angles with Low Vertical Position, in cm (in.) User 95th-percentile female—standing 95th-percentile male—standing
TAF-ER562X_WINGR-09-0401-C004.indd 129
User Eye Height
Difference
Viewing Angle (°)
163.0 (64.2) 182.5 (71.9)
74.7 (29.4) 94.2 (37.1)
58.5 64.1
11/9/10 2:23:18 AM
130
FIGURE 4.13
Handbook of Human Factors in Medical Device Design
Keyboard of a clinical chemistry analyzer diagnostic medical system.
height adjustment range. Figure 4.13 shows the keyboard workstation and the diagnostic system. The operators must stand since they need to interact with the system (e.g., place samples into the appropriate loading areas or load reagents and calibration samples). Thus, a standing operator occasionally would use the keyboard. The designers wanted to provide an adjustable keyboard tray to accommodate a reasonable range of users. Table 4.13 shows data for estimating keyboard height adjustability range. A practical adjustability range would be from 5th-percentile Asian female to a 95th-percentile Caucasian male. These values establish a keyboard height range from 92.7 to 114.0 cm. Note that if this operation were critical, the range should be extended by use of the 1st- to 99th- percentile values (90.7–117.1 cm).
4.2.3 FINGER CLEARANCE SPACE CALCULATIONS Clinicians and medical device operators use equipment with their hands placed in restricted spaces. This case study illustrates several examples using hand dimension data from Table 4.3 and Figure 4.2. These hand data are used to set design dimensions in critical hand clearance areas for several medical devices. TABLE 4.13 Keying Heights Required by 95th-Percentile Males and 5th-Percentile Females, in cm (in.) Percentile 5th
50th
95th
Reference
Male
99.6 (39.2)
106.9 (42.1)
114.0 (44.9)
Female (Japanese)
92.7 (36.5)
98.6 (38.8)
104.1 (41.0)
White and Churchill (1971) (computed) National Aeronautics and Space Administration (1978a,b,c)
Gender
TAF-ER562X_WINGR-09-0401-C004.indd 130
11/9/10 2:23:19 AM
Anthropometry and Biomechanics
FIGURE 4.14
131
Hand clearances for sample probe on blood analyzer prototype.
4.2.3.1 Prototype Blood Analyzer—Sample Probe Hand Clearance Figure 4.14 shows the front surface of a foam core mock-up for a blood analyzer prototype. Laboratory technicians sometimes need to insert a sample vial during a STAT (immediate) test of blood chemistry. They insert the vial of blood so that a pickup tube probe can vacuum-draw an aliquot from the sample. The technician’s hand must clear the spaces (labeled D1 and D2 in Figure 4.14). Because the sample vial typically is small, data taken from Table 4.3 can be used to show a first approximation of hand thickness for the 95thpercentile male (dimension 16). This hand thickness value, including the thumb, is 5.8 cm and is the minimum design value for D1 in Figure 4.14. Similarly, the minimum clearance for D2 would be set by dimension 15 (metacarpal hand thickness) for the 95th-percentile male, which is 3.8 cm. These values represent minimum bounds, and using a larger range is more likely to improve the design. 4.2.3.2 Portable Infusion Pump Access for Thumb Bolus Activation Lockboxes commonly enclose home-use, portable infusion pumps to secure patient-controlled administration of pain medication. The patient typically pushes a button at the end of a pendant to request a bolus (i.e., dose) of medication. A program controls dosage amount, the allowable lockout interval between doses, and the cumulative dose limit over a specified time. Sometimes the pendant is not used, requiring that the access opening be designed so patients can use their thumb or finger to push a bolus button incorporated on the portable pump itself. Figure 4.15 shows a pump, the clear plastic lockbox, and the opening for finger activation of the bolus button. Table 4.3 provides data for thumb breadth at the interphalangeal joint
FIGURE 4.15 Finger access opening in a lockbox for a portable infusion pump. (Photos courtesy of Abbott Laboratories.)
TAF-ER562X_WINGR-09-0401-C004.indd 131
11/9/10 2:23:20 AM
132
Handbook of Human Factors in Medical Device Design
FIGURE 4.16 Access opening for adding reaction vessels to clinical chemistry analyzer. (Photos courtesy of Abbott Laboratories.)
(dimension 8) for the largest anticipated thumb size (i.e., the 95th-percentile male). The minimum opening must be 2.6 cm plus a margin to avoid scraping the thumb against the plastic and to allow more flexibility in the pushing action. Dimension 15 (hand thickness at the metacarpal joint) would be appropriate for sizing the opening for the pointing finger if it were used to activate the bolus button. A person with this finger size (the 95th-percentile male) would require an even larger opening of 3.8 cm. Thus, this analysis shows the minimum opening to be 3.8 cm. 4.2.3.3 Clinical Chemistry Diagnostic System Access Opening Designers of a clinical chemistry diagnostic system that measures critical values of human samples for potassium, calcium, sodium, and so on needed to determine access opening size for the disposable reaction containers. The opening had to allow technicians to pour new, small, clear plastic reaction vessels into a sorter mechanism, while also preventing human access. It had been noted in previous risk analyses for use error that technicians were tempted to reach into the sorting area to push down on the vessels to create more space. This pushing action could damage the sorting mechanism, so designers considered placing a grate at the top of the access opening. The grate (shown in Figure 4.16) needed to be large enough to allow the free flow of new reaction containers but prevent the most foreseeably small hand from passing through it. Table 4.3, dimension 20 (minimum square access), shows 5.0 cm to be the smallest (i.e., 5th-percentile female) hand of a user of this system. Dimension 12 (hand breadth at metacarpal) of 6.9 cm, also for the 5th-percentile female, is another data point to consider in the design of section dimensions and shapes of grating. In this case study, these analyses led the designers to abandon the use of a grate that attempted to solve two problems simultaneously (i.e., easy ingress for the reaction containers while discouraging operator access).
4.3 GENERAL PRINCIPLES OF GOOD BIOMECHANICAL DESIGN Medical device design must account for human physical capabilities and limitations. Failure to consider biomechanical design issues can result in excessive muscle strength being required to use the device, muscle fatigue, reduced endurance and ability to work with the device for as long as is needed, increased numbers of errors, and longer task performance
TAF-ER562X_WINGR-09-0401-C004.indd 132
11/9/10 2:23:22 AM
Anthropometry and Biomechanics
133
times. Medical devices and workplaces designed using biomechanical principles can increase productivity, reduce user fatigue, and produce a safer working environment.
4.3.1 CRITICAL DESIGN CONSIDERATIONS 4.3.1.1 Body Posture Muscle strength in the trunk and extremities is greatest when the human body is in a nearly fetal-like position, that is, slightly crouched, with the arms and legs bent, the arms near the body, and the wrists in a nearly neutral posture. The strength capability of the body’s joints in these positions maximizes the tolerance to loads and forces imposed on them. Obviously, medical devices rarely are used when the body is in the fetal position. However, the following guidelines present two approaches to minimize the impact of physical activity and stress on the human body during work tasks: Guideline 4.16: Minimize Device and Load Weights Generally, designers should minimize a medical device’s weight and the force required to manipulate it.
Guideline 4.17: Seek Use in Neutral Postures Designers should design medical devices so that users can work in as close to neutral postures (i.e., relaxed joint positions) as possible while performing associated tasks.
These two approaches are ideal, but it is not possible to design every work situation to abide by these guidelines. However, reducing load weights and force requirements as well as focusing on correct body joint angles as much as is feasible will reduce stress on the body and increase efficiency when working with medical devices. 4.3.1.2 Trade-Offs The human body can be thought of as a system of links as when, for example, the hand/ wrist, elbow, and shoulder joints work together to operate a device. This often makes it difficult to design devices to minimize the force exerted by a joint and properly position that joint at the same time. In these situations, follow these guidelines: Guideline 4.18: Proper Position of Critical Body Part(s) Designers should determine the body part(s) most relevant for the task, most often used, or most likely to be stressed and design the given medical devices so that they can keep that body part or parts in proper (i.e., most comfortable) position.
Guideline 4.19: Stable and Preferred Use Positions Medical devices should enable users to assume comfortable, stable use positions. Often, it is advantageous to allow users to interact with a device from either a standing or seated position, whichever is most comfortable or preferable for them at a given time.
Guideline 4.20: Neutral Hand Position To avoid joint stress and fatigue, medical devices should enable users to maintain a hand position as close to neutral as possible, requiring only temporary excursions (if at all) to
TAF-ER562X_WINGR-09-0401-C004.indd 133
11/9/10 2:23:23 AM
134
Handbook of Human Factors in Medical Device Design
Strength
100%
50%
0
0
5 Endurance
10
min
FIGURE 4.17 The relationship between isometric muscle strength and muscle endurance. The curve illustrates that lower amounts of strength required when performing a physical activity result in an increased length of time that the muscle can sustain that activity. accomplish tasks. The hand is neutral when positioned as if one was shaking hands, with a straight (neither flexed nor extended) wrist and slightly closed hand with fingers curled.
Guideline 4.21: Comfortable Working Posture Medical devices should allow users to maintain a comfortable working posture. For example, a standing workstation should enable users to keep their back straight rather than require them to lean forward, which could cause back strain.
4.3.1.3 Endurance Muscular endurance is a function of the amount of strength exerted by a muscle or group of muscles. As shown in Figure 4.17, one can exert a muscle maximally for only a few seconds before fatigue sets in. However, the lower the percentage of one’s maximal strength that a task requires, the longer that strength level can be sustained before fatigue or physical discomfort occurs. Muscle endurance also is impacted by the length of time that muscle is used. Even if the body is in an “ideal” posture during device use and only low levels of muscle strength are required, the involved muscles still will fatigue if they must be exerted continually or without adequate rest. Therefore, designers should follow these guidelines: Guideline 4.22: Minimize Static Postures To avoid fatigue and strain, medical devices should not require users to maintain a specific posture for lengthy periods of time.
Guideline 4.23: Minimize Required Forces Medical devices should not require users to exert forces that could cause strain or rapid fatigue.
Guideline 4.24: Rest Periods between Uses When a medical device will require prolonged periods of muscle exertion, it should also allow for brief and intermittent periods of rest. Rest periods can delay the onset of fatigue that may degrade task performance or lead to the development of cumulative trauma.
TAF-ER562X_WINGR-09-0401-C004.indd 134
11/9/10 2:23:23 AM
Anthropometry and Biomechanics
135
4.3.1.4 Repetitive Motions Continual use of the same body part to perform a task is known to increase musculoskeletal injury risk. Examples of repetitive activities include: (1) continuous data entry on a computer keyboard, such as used in a computerized physician order entry system; (2) use of the same finger to activate a tool, such as a grasping device for endoscopic surgery; and (3) repeated rotation of the shoulder to reach into and out of a surgical field (e.g., handassisted laparoscopic nephrectomy). Guideline 4.25: Avoid Repetitive Motions To prevent cumulative trauma disorders (e.g., carpal tunnel syndrome), medical devices should not require users to perform a substantial number of repeated motions (e.g., twisting a tool) without appropriately timed rest periods.
Guideline 4.26: Flexible Body Positions Medical devices should not force users to assume precise body positions (e.g., sitting postures and hands grips). Rather, they should allow for flexible positioning so that users can vary how they perform a task to reduce the chances of physical stress.
There are several ways to reduce the physical impact from performing repetitive motions. Thus, designers should follow prior guidance regarding rest periods and strive to permit work to be done in as close to neutral body positions as possible. These topics are addressed in more detail in Chapter 16, “Hand Tools.”
4.3.2 SPECIAL POPULATIONS Physical capabilities, especially strength and endurance, differ among defi ned subgroups within the population. These include children, the aged, or those with certain physical impairments, such as muscular dystrophy. Although the previously mentioned biomechanical concepts (e.g., the impact of body posture, development of muscular fatigue) still are relevant for those groups, baseline differences may exist among these groups and with respect to the overall population (refer to Chapter 18, “Home Health Care”). It is not possible to discuss all situations in which physical limitations impact medical device design. However, as a first step when designing devices for use by a special population, consult the references at the end of this chapter. Also, consider the following guidelines: Guideline 4.27: Analyze and Understand Special Populations As appropriate, anthropometric analyses should address the requirements of special user populations, such as children, pregnant women, obese individuals, elders, persons with specific disabilities (e.g., paralysis, fused joints, arthritis), people with specific medical conditions (e.g., Parkinson's Disease), and people with physical anomalies (e.g., swollen body parts, growths, deformities).
Guideline 4.28: Accommodate Adaptive Equipment Medical devices should physically accommodate users’ adaptive equipment, including wheelchairs, crutches, splints, eyewear, and hearing aids.
TAF-ER562X_WINGR-09-0401-C004.indd 135
11/9/10 2:23:23 AM
136
Handbook of Human Factors in Medical Device Design
4.3.3 DESIGN GUIDELINES FOR TASKS INVOLVING LIFTING There are numerous situations in which designers may need to evaluate the risk of injury to individuals who must perform physically demanding tasks, such as lifting heavy, critical medical devices onto a mobile stretcher prior to transporting a trauma patient. There are a variety of ergonomic assessment tools available for use in conducting such evaluations. Each has its strengths and weaknesses, but it is important to use the appropriate tool to determine if there is injury risk associated with working in an environment or using a particular medical device. One category of assessment tools relates to the potential for injury to the low back when performing a lifting task, such as moving an obese patient from one bed to another. Several of the most frequently used tools for these types of evaluations are described below. 4.3.3.1 NIOSH Revised Lifting Equation The National Institute for Occupational Safety and Health (NIOSH) developed guidelines to determine the recommended (i.e., safe) weight limit for a lifting task (Waters, PutzAnderson, Garg, and Fine, 1993). These limits indicate load weights that nearly all healthy workers could lift without increasing their risk of developing low back pain. The guidance is based on the input of data for seven factors into a lifting equation: • Horizontal location of the load, measured as the distance of the hands away from the midpoint between the ankles • Vertical location of the load, which is the distance of the hands above the floor when holding the object • Vertical distance the load travels during the lift (i.e., the absolute value of the difference between the vertical heights at the origin and destination of the lift) • Amount of twisting required during the lift (also called the asymmetry), measured as the angle between the location of the load when it is lifted and the individual when standing in a “neutral” body posture • Lifting frequency of the task, measured as the average number of lifts required per minute • Length of time lifting is required, which takes into account both work time and rest periods • Quality of the hand-to-object coupling, which assesses the handling ease of the load The physical dimensions used as inputs for this equation are illustrated in Figure 4.18. As an example of its use, consider the task of stocking devices in a hospital supply room with the following working conditions. Boxes of supplies must be held 15 inches in front of the body and are initially located 18 inches above the floor. Employees lift these boxes onto a table 40 inches above the floor, and the lifting can be done directly in front of the body. The rate of lifting is (on average) once per minute, the process takes four hours to complete, and there are good handles on the supply boxes. Entering this information into the NIOSH lifting equation would produce a recommended weight limit of 9.5 kg (21 lbs.). This indicates that lifting boxes less than 9.5 kg (21 lbs.) is “safe,” but objects weighing more would increase the risk that individuals could develop low back pain.
TAF-ER562X_WINGR-09-0401-C004.indd 136
11/9/10 2:23:24 AM
Anthropometry and Biomechanics
137
Asymmetry
Vertical travel distance
Vertical location
Horizontal location
FIGURE 4.18 equation.
Illustration of the physical dimensions used as inputs for the NIOSH revised lifting
To ease the numerical complexity of these calculations, computer software packages have been designed to compute these weight limits. NIOSH has developed a list of software sources that are available on its Web site at http://www.cdc.gov/niosh. 4.3.3.2 ACGIH Lifting Threshold Limit Values The American Conference of Governmental Industrial Hygienists (ACGIH) has developed threshold limit values (TLVs) for recommended lifting conditions under which a majority of workers purportedly can be exposed repeatedly without developing low back and shoulder disorders (ACGIH, 2002). Information input into this assessment tool is similar to that of the NIOSH lifting equation. The TLVs are presented in three tables of weight limits, determined as a function of the following: • Task duration, or the total length of time the activity is performed in one day • Lifting frequency of the task, or the number of lifts performed per hour • Vertical height zone at the beginning of the lift, or the location of the hands at the beginning of the lift • Horizontal location of the load at the beginning of the lift, or the distance of the hands away from the midpoint between the ankles Using the previous hospital supply room lifting task example, the TLV under the stated working conditions is 7.0 kg (15.4 lbs.). This indicates that appropriate control measures should be implemented if supplies are lifted that exceed 7.0 kg (15.4 lbs.) in weight. Note that this weight limit is lower than that calculated previously using the NIOSH lifting equation. This is because different data and methodologies were used to generate the lifting
TAF-ER562X_WINGR-09-0401-C004.indd 137
11/9/10 2:23:24 AM
138
Handbook of Human Factors in Medical Device Design
FIGURE 4.19 The industrial lumbar motion monitor (iLMM). (Photo courtesy of Biodynamic Solutions, Inc.)
limits. The choice of assessments depends on how each tool’s assumptions compare to the lifting task being evaluated. 4.3.3.3 Industrial Lumbar Motion Monitor Risk Assessment System The industrial lumbar motion monitor (iLMM) and its associated risk model determine the probability that a job will produce a low back injury rate comparable to other lifting tasks having high numbers of back injuries (Marras et al., 1993). Thus, the physical design of a medical device may impact how a user interacts with it and, thus, the resulting low back injury rate. Figure 4.19 shows the iLMM device being worn to record trunk motions as a lifting task is performed. The iLMM records the instantaneous position, velocity, and acceleration of the trunk in the sagittal (forward bending), lateral (side bending), and transverse (twisting) planes of motion. The iLMM’s risk model incorporates five factors. Two factors are taken from the workplace: • Lift rate (the number of lifts required per hour across all job tasks) • External load moment, which is a function of the weight of the load and the horizontal distance it is handled from the spine The other three factors used in the risk model are data derived from the iLMM task measurements: • Sagittal flexion (i.e., the maximum amount of forward bending required of the job) • Twisting velocity (i.e., the average speed of axial rotation) • Lateral velocity (i.e., the maximum speed of lateral movement) This analytical approach could be beneficial in the design of some medical devices, such as patient lifting systems for hydrotherapy, or for the positioning and adjustment of mobile stretchers used in ambulances or evacuation helicopters. The iLMM system of hardware and software is commercially available, currently through NexGen Ergonomics (Montreal, Canada).
TAF-ER562X_WINGR-09-0401-C004.indd 138
11/9/10 2:23:24 AM
Anthropometry and Biomechanics
139
4.3.3.4 Psychophysical Limits Psychophysical limits refer to estimates of what individuals believe they can safely lift, lower, push, pull, and carry. These limits were derived from groups of males and females who were tested performing these types of tasks, and the subjects reported how much handling they felt was “safe.” From this, tables were created containing acceptable lifting/ force limits, as a function of (1) distance of the force exertion, (2) frequency of the activity, (3) vertical height at which the force is applied, and (4) percentage of an industrial population believed capable of safely performing the task (Snook and Ciriello, 1991). These psychophysical limits are some of the only available data regarding safe levels of pushing, pulling, or carrying. However, the definition of “safe” was derived subjectively rather than through objective testing. Thus, this information should be used as a starting point when designing medical devices requiring these types of activities. Numerous ergonomics reference books include these psychophysical tables, and the Web site of the Liberty Mutual Insurance Company (Boston, MA) offers these tables interactively. Guideline 4.29: Tools to Assess Physical Limits Designers should use one or more quantitative assessment tools to ensure that a task does not exceed safe lifting limits. For example, the NIOSH revised lifting equation or the ACGIH TLVs should be used for situations in which the lifting will be slow, smooth, and steady (e.g., no “jerky” motions). Consider using the Industrial Lumbar Motion Monitor Risk Assessment System when the lifting performed is highly dynamic and repetitive. Consider using psychophysical limits when the activity requires pushing, pulling, or carrying.
4.3.4 DESIGN GUIDELINES FOR TASKS INVOLVING USE OF THE UPPER EXTREMITY Many medical devices require activity by the shoulders, hands, and wrists. Thus, it is important to assess the injury risk to these body parts during device use. Guideline 4.30: Injury Risk Assessment Designers should use one or more quantitative assessment tools to ensure that the tasks involving the shoulders, hands, and wrists do not pose a risk of injury.
Several ergonomic tools are available to perform such assessments. As with lifting, upper extremity assessment tools also have their strengths and weaknesses. The most appropriate tool used will depend on the device being designed and how it is to be used. The more common tools to evaluate upper extremity injury risk are described below. 4.3.4.1 Strain Index The Strain Index (Moore and Garg, 1995) evaluates a job’s risk of producing a musculoskeletal disorder to the distal upper extremity (i.e., the hand/wrist joint). Six factors form the basis of the Strain Index: • • • •
Intensity of exertion, which relates to the job’s force requirements Duration of exertion (length of time the force is maintained) Efforts required per minute, or the repetitiveness of the job Hand/wrist posture, relative to the “neutral” position
TAF-ER562X_WINGR-09-0401-C004.indd 139
11/9/10 2:23:25 AM
140
Handbook of Human Factors in Medical Device Design
• Speed of work (i.e., the job pace) • Task duration, or the total length of time the job is performed The Strain Index is available as part of several commercially available computer-based packages, including ErgoIntelligence (NexGen Ergonomics, Montreal, Canada), ErgoTrack (ErgoTrack, Efland, NC), Job Evaluator Toolbox (Ergoweb, Inc., Park City, UT), and Job Hazard Pro (Production Technology, Tampa, FL). 4.3.4.2 Rapid Upper Limb Assessment The Rapid Upper Limb Assessment (RULA) tool can evaluate work with a medical device for which there is injury risk to the neck, shoulder, upper and lower arms, and hand/wrist (McAtamney and Corlett, 1993). It produces a score based on task repetition, posture, and force that can be compared with a recommended action level. An example of a medical device that could cause neck strain would be an optometric measurement device in which the head must be kept steady during peripheral optical field measurements. RULA-based software is available commercially in several computerized forms, including ErgoEASE (Ease, Inc., Mission Viejo, CA), ErgoIntelligence (NexGen Ergonomics, Montreal, Canada), ErgoSURE Pro (Magnitude, Branchburg, NJ), Job Evaluator Toolbox (Ergoweb, Inc., Park City, UT), Job Hazard Pro (Production Technology, Tampa, FL), Jack’s Task Analysis Toolkit (Siemens Product Lifecycle Management Software Inc., Plano, TX), and Osmond Ergonomic Workplace Solutions (Wimborne, UK). Guideline 4.31: Unacceptable Postures or Forces The designer should evaluate if medical device operation exposes users to unacceptable postures and forces to the upper extremity.
4.3.5 DESIGN GUIDELINES TO DETERMINE STRENGTH REQUIREMENTS 4.3.5.1 Three-Dimensional Static Strength Prediction Program The 3-Dimensional Static Strength Prediction Program is a software application available for purchase through the University of Michigan (Ann Arbor, MI). This program determines static strength requirements for activities such as lifting, pressing, pushing, and pulling (Chaffin and Andersson, 1999). The program provides estimates of the percentages of males and females in the population who have the strength capability (at the elbow, shoulder, torso, hip, knee, and ankle) for a given task. It also gives an estimate of the amount of spinal compression produced for the activity. Guideline 4.32: Strength Prediction Medical designers should evaluate whether acceptable levels of a user population have the static strength necessary to adequately use a device.
Guideline 4.33: Avoid Sustained Forces Medical devices should not require users to apply a force for extended periods of time (e.g., several minutes) because such exertion can cause fatigue and, in the case of repeated applications, lead to cumulative trauma.
TAF-ER562X_WINGR-09-0401-C004.indd 140
11/9/10 2:23:25 AM
Anthropometry and Biomechanics
141
Guideline 4.34: Avoid Unnecessarily Precise Forces Generally, medical devices should not require users to apply unnecessarily precise forces to exert proper control.
Guideline 4.35: Avoid Pressure Points Medical devices should not apply undue pressure (i.e., sharp or blunt force) to any body part. In addition to causing discomfort, excess pressure can reduce blood flow to a body part, injure soft tissue, and cause other physical ailments.
4.4 CASE STUDIES IN BIOMECHANICAL DESIGN 4.4.1 KNOB TWISTING FORCES ON POLE CLAMPS Often, critical care medical devices such as infusion pumps, internal feeding pumps, and continuous blood pressure monitors need to be clamped to a bedside pole. Knobs secure many pole-clamping devices. The main design consideration is to provide a mechanism that will tightly secure the device without requiring high torque on the knob so that a majority of health care workers have the ability to secure and release the clamp. The result of an unsecured device is that it can slide down the pole, causing personal injury or damage to the device. A pole clamp that is tightened by an especially strong user may preclude a weaker user from removing the device without mechanical assistance. There are three types of grips commonly used for twisting a pole clamp knob. Figure 4.20 shows a finger grip, which results in the least torque possible. Figure 4.21 shows a grip typical for grasping a jar, for which larger torque forces are possible. Figure 4.22 shows the
FIGURE 4.20
Knob turning with a finger grip. (Photo courtesy of Abbott Laboratories.)
FIGURE 4.21
Knob turning with a jar lid grip. (Photo courtesy of Abbott Laboratories.)
TAF-ER562X_WINGR-09-0401-C004.indd 141
11/9/10 2:23:25 AM
142
FIGURE 4.22 Laboratories.)
Handbook of Human Factors in Medical Device Design
Knob turning with a handlebar power grip. (Photo courtesy of Abbott
most powerful grip, such as that used to grasp a handlebar (Hanson and Israelski, 1977). The handlebar grip may not be possible for some pole clamps because of restricted space in the area where the clamp is mounted on the device. The design issue for this case involves determining the tightening torque force that a user with limited strength can apply to a pole clamp using a 6.9-cm-diameter round knob, as shown in these figures. 4.4.1.1 Analysis In this case study, the exact conditions for which a designer is seeking biomechanical data do not exist in precisely the desired form. Extrapolation often is needed to use data collected under similar (but not exact) circumstances to estimate the forces needed for design input. A good source of biomechanical data for people with varying amounts of dexterity comes from the Department of Trade and Industry (2002). These data were used to derive Figure 4.23, which shows clockwise tightening torque for 5th-percentile females with no impairment, as a function of knob diameter. Torque (in.-lbs.) vs. knob diameter (in.) 5th percentile, combined male and female, non-disabled, jar lid grip turning clockwise 18.00
Male
16.00 Female
14.00 12.00 10.00 8.00 6.00 4.00 2.00 0.00
0.00 0.50 1.00 1.50 2.00 2.50 3.00 3.50 4.00 4.50 5.00 5.50 Knob diameter (in.)
FIGURE 4.23
Torque generated versus knob diameter.
TAF-ER562X_WINGR-09-0401-C004.indd 142
11/9/10 2:23:26 AM
Anthropometry and Biomechanics
143
The texture and shape of a knob will affect the forces generated. For example, knurled knobs increase friction with the hand and allow one to apply larger forces, compared to smooth knobs. Triangular knobs allow the generation of even higher torque forces, as do knobs that have small extensible foldout levers (Kohl, 1983). To preclude stronger users from tightening something so much that it exceeds the loosening capabilities of weaker users, the design should employ a torque limiter. An example of a torque-limiting device would be a slipping clutch in a pole clamp design similar to automotive gas cap tightening mechanisms. See Chapter 7, “Controls,” for more information about choosing knob designs. A Department of Trade and Industry (2002) study showed that males and females at the 5th percentile can apply a force of 5.3 N on a knurled, 6.9-cm-diameter knob using the lid jar grip (Figure 4.22). Since the population sampled for this study was evenly split between genders, and female upper body strength is approximately 70% that of males, a gender-combined force of 5.3 N would be reduced by 0.82 for females. The resulting 5th-percentile female force capability would be 4.3 N of torque. If the clamp requires more than 4.3 N of torque to tighten, then less than 95% of female users will be able to place the device securely on the pole. Larger-diameter knobs are an option, although knobs with diameters greater than 10.7 cm no longer have an advantage, as users will begin to have trouble gripping them. The use of the more powerful handle grip also is more difficult in tighter spaces. Existing biomechanical data often will not be an exact match, so it may be necessary to conduct a separate strength measurement study using the parameters of the proposed design. Performing a special study also would be appropriate if risk levels from the analysis are estimated to be high enough that it does not seem reasonable to extrapolate from existing data.
4.4.2 BENDING FORCES ON AN AUTOINJECTOR DEVICE Designers of a penlike device for the autoinjection of a critical drug wanted to know the reasonably foreseeable bending force that might be exerted by a strong user. As users removed the device’s protective end caps, the concern was that they might exert excessive torque on the cylindrical body of the device, by holding each end and bending it (Figure 4.24). As noted in previous case studies, existing biomechanical data only provides an approximation of the information needed by the designers. 4.4.2.1 Analysis Table 4.8 shows values for the power grasp that are a reasonable approximation to the bending force at issue in this case study. The bottom row describing power grasp is relevant and shows a maximum mean force of 366 N, with a standard deviation of 53 N. Table 4.14 presents static forces exhibited by the strongest 50th- and 95th-percentile males when
FIGURE 4.24 Autoinjector drug delivery pen, 7 inches long by 1 inch in diameter (longitudinal bending/breakage with one hand at each end).
TAF-ER562X_WINGR-09-0401-C004.indd 143
11/9/10 2:23:27 AM
144
Handbook of Human Factors in Medical Device Design
TABLE 4.14 Bending Forces Generated by the Strongest Males, in N (lbs.) One-Hand Pulling Force Male (machinists)
Mean (50th percentile)
Standard Deviation
95th Percentile
365.6 (82.2)
52.9 (11.9)
452.8 (101.8)
pulling down a fixed cylindrical handle approximately 2.5 cm in diameter with one hand. (The original source does not specify the handle diameter.) The solution to this challenge is to design the device to have a maximum bending force of 46.2 kg, which can be generated on a cylindrical handle by a 95th-percentile male. There were no data specific to a two-handed maximum bending force on a cylinder. The approximation for one-handed operation still is reasonable if one assumes that the second hand holds the cylinder fixed and makes one end of the cylinder the fulcrum point for calculating a moment arm of 17.8 cm. If the user were to fix the center of the cylinder on a rigid surface as the fulcrum, then each hand theoretically could generate the 46.2 kg force over shorter lengths.
4.4.3 SNAP-ON LID REMOVAL FORCES This case study examined the forces that could be exerted to remove a lid from a vial. The vial contained medication for clinician or patient access and did not require a screw-on lid or a childproof container. The holding force of the lid needed to be sufficient to remain secure during transportation and storage but at the same time be able to withstand an attempt to open it by someone without proper tools. Thus, the two design questions were the following: • What is the foreseeable smallest force that should be exerted to reliably remove a snap-on lid from a medical vial? • What is the foreseeable largest force that can be exerted to remove a snap-on lid from a medical container that must be securely fastened until it is mechanically removed? 4.4.3.1 Analysis Both questions involved the same push forces with the thumb against a snap-on lid (Figure 4.25). However, quite different strength information was needed. A calculation was needed of both the 5th-percentile female force and the 95th-percentile male force. Thumb push against a lid was not found in the published biomechanical literature, but a reasonable approximation was found in Schoorlemmer and Kanis (1992). As shown in Figure 4.26, the presumed thumb pushing force positions were calculated for a plunger disk 2.0 cm in diameter, with disk clearance of 1.5 cm above the tabletop. Test subjects pushed from any free-posture position desired and in a seated position with elbows at a 90-degree angle. Values from the free-posture position were larger and seemed to be more appropriate to how snap-on lids might be handled in real-world conditions. Table 4.15 shows the values from the Schoorlemmer and Kanis study and the derived values for the 5th-percentile female force (0.9 kg) and the 95th-percentile male force (16.1 kg).
TAF-ER562X_WINGR-09-0401-C004.indd 144
11/9/10 2:23:27 AM
Anthropometry and Biomechanics
145
FIGURE 4.25
Snap-on lid vial (1.6 inches long by 0.8 inches in diameter).
FIGURE 4.26
Thumb-pushing-force positions.
TABLE 4.15 Calculations of Thumb-Pushing Forces for the 5th and 95th Percentiles Mean (50th percentile)
Standard Deviation (SD)
Force (Percentile)
19.5 pounds 15.3 pounds
9.7 pounds 8.1 pounds
35.4 pounds (95th) 2.0 pounds (5th)
Male Female
Note: Conversions used were 1 N = 0.2247 pound-force; 95th or 5th percentile = mean ± 1.645(SD), assuming normal distributions.
TAF-ER562X_WINGR-09-0401-C004.indd 145
11/9/10 2:23:27 AM
146
Handbook of Human Factors in Medical Device Design
This case study illustrates that, if there is any doubt about the applicability of existing biomechanical information, one should conduct a specific data gathering study. This is especially true if the tasks in question are known to be high risk.
4.4.4 PULLING FORCES ON IV TUBING Designers of intravenous (IV) tubing and accessories wanted to determine a reasonably foreseeable worst-case pulling force on IV tubing that might stem from patient or user abuse, whether purposeful or inadvertent. The expected pull forces would determine the mechanical breaking forces of the tube. The tube is attached to cassette pumping units that are typically inserted into volumetric infusion pumps. Some infusion pumps do not use proprietary cassettes for fluid pumping. Instead, they allow the IV tubing to be placed directly into the infusion pump while the peristaltic action of mechanical “fingers” pumps the IV fluids through the tubing. 4.4.4.1 Analysis Inadvertent pulling on tubes can arise from numerous situations. Use scenarios involving pull forces include the following: • The user moves while still accidentally holding on to the tubing. This likely could produce only a very small force (e.g., less than 27 N). • The user snags the tubing with an article of clothing or a hospital bracelet and moves with a large pull force before noticing the resistance and stopping. This force likely would be considerably less than the deliberate forces that have been published (e.g., less than 623 N standing, as described below). • The user pulls on tubing while attempting to remove the cassette because of a misunderstanding of the proper way to eject the cassette. This force could approach or exceed the measured biomechanical force of 623 N capable from a 95th-percentile or higher male. • The user loses his footing and starts to fall while holding on to the tubing. The user’s full body weight becomes the pulling force. For 95th-percentile males, this force could be greater than 956 N (White and Churchill, 1971). A reasonable maximum pulling force for 95th-percentile males (a conservative worst-case scenario for both patients and nurses) is 623 N. This is derived from the 95th-percentile static pulling force curve shown in Figure 4.27 (Chaffin, 1972). The maximum force is a function of the position of the hands (these curves are for standing, two-handed pulls). Note that the maximum, deliberate pull force is from a position 50.8 cm above the floor and 50.8 cm in front of the ankles. The 50th-percentile male force would be 556 N (Chaffin, 1972). The use of more conservative forces (at higher percentiles such as the 97th or 99th) may be overly restrictive. However, one would need to extrapolate from these curves to calculate the higher values.
4.4.5 PULLING FORCE TO REMOVE A PLUNGER FROM A VIAL This case study determines the reasonable force expected from the weakest clinical users to retract an injector plunger from a medication vial. Figure 4.28 shows a sketch of a medication vial that is activated by a plunger. The vial typically is threaded onto a medication container and, in turn, connected to tubing as part of a medication infusion system.
TAF-ER562X_WINGR-09-0401-C004.indd 146
11/9/10 2:23:28 AM
Anthropometry and Biomechanics
147
80
70
Vertical hand height above ankles (in.)
60
50
40
30 75 140
100
26 50
20
10
0 10
20
30
40
50
60
70
Horizontal hand position in front of ankles (in.)
FIGURE 4.27 Male static pulling forces (95th percentile while standing using both hands in different vertical and horizontal hand positions). Two-dimensional surfaces on the x-y plane show forces in pounds. (From Chaffin, D.B., Some Effects of Physical Exertion, Western Electric/University of Michigan Report, AT&T, New York, 1972. With permission.)
4.4.5.1 Analysis Because the goal is to determine the weakest forces likely to occur in the user population, the focus is on the 5th-percentile female. If the retraction force of the plunger exceeds the 5th-percentile female value, then more than 5% of female users of these vials will be unable to retract the plunger. Alternatively, the design could specify a retraction force
Plunger
FIGURE 4.28
Plunger-activated medication vial.
TAF-ER562X_WINGR-09-0401-C004.indd 147
11/9/10 2:23:28 AM
148
Handbook of Human Factors in Medical Device Design
that accommodates those below the 5th percentile (e.g., the 1st percentile). A mechanical lower bound would be established by the design parameters and the need to avoid inadvertent plunger movement during shipping or normal handling. The risk analysis would assess the consequences of a small number of users being unable to retract the plunger without a tool. Information about “hook grip” forces (Table 4.8) is the best match for plunger retraction force. The force data from use of all digits is a reasonable approximation for the retraction movement. Depending on the size of the plunger handle, not all fingers could be used. However, use of at least two fingers is likely, and therefore the “all-digits-combined” condition is a reasonable choice. The resulting pulling force for the weaker of the two male populations studied is 108 N, with a standard deviation (sd) of 39 N. Assuming a normal distribution, the calculated 5th-percentile male pulling force is 43.6 N (95th or 5th percentile = mean ± 1.645x[sd], assuming normal distributions). Since female force data were not published for this hand configuration, one may apply common conversion factors. Various gender conversion factors are described in the biomechanical literature, ranging from 0.55 to 0.70 (Kroemer et al., 1997; Pheasant, 1996). The lowest conversion factor would be the most conservative for this case study. Thus, the resulting 5th-percentile female plunger retraction force would be 24.0 N.
ANTHROPOMETRY-RELATED RESOURCES Brown, R., Rogers, N., Ward, J., Wright, D., and Jeffries, G. (1995). The application of an anthropometric database of elderly and disabled people. Biomedical Sciences Instrumentation, 31, 235–239. Chung, K. and Weimar, W. (1989). Anthropometric Studies for the Physically Disabled Population— Vol. II, Spinal Cord Injury (Report No. UVA-REC 102-89). Charlottesville: University of Virginia, Rehabilitation Engineering Center. Das, B. and Kozey, J. W. (1994). Structural anthropometry for wheelchair mobile adults. Proceedings of the 12th Triennial Congress of the International Ergonomics Association, Human Factors Association of Canada, Toronto, Ontario 3: 63–65. Diffrient, N., Tilley, A. R., and Bardagjy, J. C. (Henry Dreyfuss Associates). (1981). Humanscale 4/5/6. Cambridge, MA: MIT Press. Diffrient, N., Tilley, A. R., and Bardagjy, J. C. (Henry Dreyfuss Associates). (1981). Humanscale 7/8/9. Cambridge, MA: MIT Press. Gordon, C. C., Churchill, T., Clauser, C. E., Bradtmiller, B., McConville, J. T., Tebbetts, I., and Walker, R. A. (1989). 1988 Anthropometric Survey of U.S. Army Personnel: Summary Statistics Interim Report (Natick-TR-89/027). Natick, MA: U.S. Army Natick Research Development and Engineering Center. Greiner, T. M. (1990). Hand Anthropometry of U.S. Army Personnel (AD-A244-533). Natick, MA: U.S. Army Natick Research Development and Engineering Center. Hobson, D. A. and Molenbroek, J. F. M. (1990). Anthropometry and design for the disabled: Experiences with seating design for the cerebral palsy population. Applied Ergonomics, 21(1), 43–54. Hobson, D. A., Shaw, C. G., Monahan, L., and Mclaurin, C. (1987). Anthropometric Data for Design of Specialized Seating and Mobility Devices: A Preliminary Report (pp. 480–482). RESNA 10th Annual Conference: San Jose, CA: Rehabilitation Engineering Society of North America. Kagimoto, Y. (1990). Anthropometry of JASDF Personnel and Its Applications for Human Engineering. Tokyo: Aeromedical Laboratory, Air Development and Test Wing, JASDF.
TAF-ER562X_WINGR-09-0401-C004.indd 148
11/9/10 2:23:30 AM
Anthropometry and Biomechanics
149
Kouchi, M., Yokoyama, K., Yamashita, J., Yokoi, T., Ogi, H., Yoshioka, M., Atsumi, H., and Hotta, A. (1994). Human Body Dimensions Data for Ergonomic Design [Anthropometric data of Japanese adults]. Tsukuba, Japan: National Institute of Bioscience and Human Technology. Kozey, J. and Das, B. (1997). The determination of the maximum reach envelope for wheelchair mobile adults. In B. Das and W. Karwowski (Eds.), Advances in Occupational Ergonomics and Safety 1997 (pp. 315–318). Burke, VA: IOS Press. Kroemer, K. H. E. (2005). “Extra-Ordinary” Ergonomics: How to Accommodate Small and Big Persons, the Disabled and Elderly, Expectant Mothers, and Children. London: Taylor & Francis. Nowak, E. (1996). The role of anthropometry in design of work and life environments of the disabled population. International Journal of Industrial Ergonomics, 17, 113–121. Nowak, E. (1997). Anthropometry for the needs of disabled people. In S. Kumar (Ed.), Perspectives in Rehabilitation Ergonomics (pp. 302–338). Bristol, PA: Taylor & Francis. Peebles, L. and Norris, B. (1998). Adultdata: The Handbook of Adult Anthropometric and Strength Measurements—Data for Design Safety. London: Department of Trade and Industry. Pryor, H. B. and Thelander, H. E. (1967). Growth deviations in handicapped children: An anthropometric study. Clinical Pediatrics, 6(8), 501–512. Singh, J., Peng, C. M., Lim, M. K., and Ong, C. N. (1995). An anthropometric study of Singapore candidate aviators. Ergonomics, 38(4), 651–658. Skelton, D. A., Greig, C. A., Davies, J. M., and Young, A. (1994). Strength, power and related functional ability of healthy people aged 65–89 years. Age and Aging, 23, 371–377. Snyder, R. G. (1977). Anthropometry and Biomechanics of Selected Populations (Technical Report UM-HSRI-77-52). Morgantown, WV: National Institute for Occupational Safety and Health. Snyder, R. G. (1977). Anthropometry of Infants, Children, and Youths to Age 18 for Product Safety Design. Ann Arbor: Highway Safety Research Institute, University of Michigan; Warrendale, PA: Society for Automotive Engineers. Steenbekkers, L. P. A. and Molenbroek, J. F. M. (1990). Anthropometric data of children for nonspecialist users. Ergonomics, 33(4), 421–429. Stoudt, H. W. (1981). The anthropometry of the elderly. Human Factors, 23(1), 29–37. United States Department of Health and Human Services, National Center for Health Statistics. (1983). Health and Nutrition Examination Survey II, 1976–1980 (computer file). Ann Arbor, MI: Inter-University Consortium for Political and Social Research. White, R. (1982). Comparative Anthropometry of the Foot (TR-83/101). Natick, MA: U.S. Army Natick Research, Development and Engineering Center. Woodson, W. E., Tillman, B., and Tillman, P. (1992). Human Factors Design Handbook: Information for the Design of Systems, Facilities, Equipment, and Products for Human Use. New York: McGraw-Hill.
BIOMECHANICS-RELATED RESOURCES Brown, D., Knowlton, R. G., Hamill, J., Schneider, T. L., and Hetzler, R. K. (1990). Physiological and biomechanical differences between wheelchair-dependent and able-bodied subjects during wheelchair ergometry. European Journal of Applied Physiology and Occupational Physiology, 60, 179–182. Chaffin, D. B., Martin, B. J., and Andersson, G. B. (1999). Occupational Biomechanics (3rd ed.). New York: John Wiley & Sons. Eastman Kodak Company. (1989). Ergonomic Design for People at Work, Vol. II: The Design of Jobs, Including Work Patterns, Hours of Work, Manual Materials Handling Tasks, Methods to Evaluate Job Demands, and the Physiological Basis of Work. New York: Van Nostrand Reinhold.
TAF-ER562X_WINGR-09-0401-C004.indd 149
11/9/10 2:23:30 AM
150
Handbook of Human Factors in Medical Device Design
Jarosz, E. (1996). Determination of the workspace of wheelchair users. International Journal of Industrial Ergonomics, 17, 123–133. Karwowski, W. and Marras, W. S. (1999). The Occupational Ergonomics Handbook. Boca Raton, FL: CRC Press. Kenward, M. G. (1971). An approach to the design of wheelchairs for young users. Applied Ergonomics, 2(4), 221–225. National Research Council, Institute of Medicine. (2001). Musculoskeletal Disorders and the Workplace: Low Back and Upper Extremities. Washington, DC: National Academies Press. Nowak, E. (1989). Workspace for disabled people. Ergonomics, 32(9), 1077–1088. Salvendy, G. (2006). Handbook of Human Factors and Ergonomics (3rd ed.). New York: John Wiley & Sons.
REFERENCES American Conference of Governmental Industrial Hygienists. (2002). 2002 Threshold Limit Values for Chemical Substances and Physical Agents and Biological Exposure Indices. Cincinnati, OH: American Conference of Governmental Industrial Hygienists. Chaffin, D. B. (1972). Some Effects of Physical Exertion (Western Electric/University of Michigan Report). New York: AT&T. Chaffin, D. B. and Andersson, G. (1999). Occupational Biomechanics (3rd ed.). New York: John Wiley & Sons. Consumer and Competition Policy Directorate, Department of Trade and Industry. (2002a). Specific Anthropometric and Strength Data for People with Dexterity Disability. London, United Kingdom: Consumer and Competition Policy Directorate. Consumer and Competition Policy Directorate, Department of Trade and Industry. (2002b). Strength Data for Design Safety—Phase 2. London, United Kingdom: Consumer and Competition Policy Directorate. Department of Trade and Industry. (2002). Specific Anthropometric and Strength Data for People With Dexterity Disability. London, United Kingdom: Department of Trade and Industry. Diffrient, N., Tilley, A. R., and Bardagjy, J. C. (Henry Dreyfuss Associates). (1981). Humanscale 1/2/3. Cambridge, MA: MIT Press. Eastman Kodak Company. (1983). Ergonomic Design for People at Work, Volume I: Workplace, Equipment, and Environmental Design and Information Transfer. Belmont, CA: Lifetime Learning Publications. Flugel, F., Greil, H., and Sommer, K. I. (1986). Anthropologischer Atlas. Berlin: Tribuene. Gordon, C. C., Churchill, T., Clauser, C. E., Bradtmiller, B., McConville, J. T., Tebbetts, I., and Walker, R. A. (1989). 1988 Anthropometric Survey of U.S. Army Personnel: Summary Statistics Interim Report (Natick-TR-89/027). Natick, MA: U.S. Army Natick Research, Development and Engineering Center. Hanson, B. L. and Israelski, E. W. (1977). Human factors engineering in the outside plant: Bringing out the best. Bell Laboratories Record, February, 30–55. Houy, D. A. (1983). Range of joint motion in college males. Proceedings of the Human Factors Society 27th Annual Meeting, Human Factors Society, Santa Monica, CA, 374–378. Kagimoto, Y., ed. (1990). Anthropometry of JASDF Personnel and its Applications for Human Engineering. Tokyo: Aeromedical Laboratory, Air Development and Test Wing, JASDF. Kohl, G. A. (1983). Effects of shape and size of knobs on maximal hand-turning forces applied by females. Bell System Technical Journal 62(6): 1705–1712. Kroemer, K. H. E., Kroemer, H. B., and Kroemer-Elbert, K. E. (1994). Ergonomics: How to Design for Ease and Efficiency. Englewood Cliffs, NJ: Prentice Hall. Kroemer, K. H. E., Kroemer, H. J., and Kroemer-Elbert, K. E. (1997). Engineering Physiology: Bases of Human Factors/Ergonomics. New York: Van Nostrand Reinhold.
TAF-ER562X_WINGR-09-0401-C004.indd 150
11/9/10 2:23:30 AM
Anthropometry and Biomechanics
151
Marras, W. S., Lavender, S. A., Leurgans, S. E., Rajulu, S. L., Allread, W. G., Fathallah, F. A., et al. (1993). The role of dynamic three dimensional trunk motion in occupationally related low back disorders: The effects of workplace factors, trunk position and trunk motion characteristics on injury. Spine, 18(5), 617–628. McAtamney, L. and Corlett, E. N. (1993). RULA: A survey method for the investigation of workrelated upper limb disorders. Applied Ergonomics, 24(2), 91–99. Moore, J. S. and Garg, A. (1995). The strain index: A proposed method to analyze jobs for risk of distal upper extremity disorders. American Industrial Hygiene Association Journal, 56, 443–458. National Aeronautics and Space Administration. (1978a). Anthropometric Source Book, Vol. I: Anthropometry for Designers (NASA Reference Publication 1024). Washington, DC: National Aeronautics and Space Administration, Scientific and Technical Information Office. National Aeronautics and Space Administration. (1978b). Anthropometric Source Book, Vol. II: A Handbook of Anthropometric Data (NASA Reference Publication 1024). Washington, DC: National Aeronautics and Space Administration, Scientific and Technical Information Office. National Aeronautics and Space Administration. (1978c). Anthropometric Source Book, Vol. III: Annotated Bibliography of Anthropometry (NASA Reference Publication 1024). Washington, DC: National Aeronautics and Space Administration, Scientific and Technical Information Office. National Aeronautics and Space Administration. (1989). Man-Systems Integration Standards (NASA-STD-3000A). Houston: Lyndon B. Johnson Space Center. Pheasant, S. (1996). Bodyspace: Anthropometry, Ergonomics and the Design of Work. London: Taylor & Francis. Robinette, K. M. (2000). CAESAR measures up. Ergonomics in Design, 8(3), 17–23. Schoorlemmer, W. and Kanis, H. (1992). Operation of controls on everyday products. Proceedings of the Human Factors Society 36th Annual Meeting (pp. 509–513). Santa Monica, CA: Human Factors and Ergonomics Society. Snook, S. H. and Ciriello, V. M. (1991). The design of manual handling tasks: Revised tables of maximum acceptable weights and forces. Ergonomics, 34(9), 1197–1213. Staff, K. R. (1983). A Comparison of Range of Joint Mobility in College Females and Males. Master’s thesis, Industrial Engineering, Texas A&M University. U.S. Department of Defense. (1995). Handbook for Human Engineering Design Guidelines. MILHDBK-759C. Philadelphia, PA: Navy Publishing and Printing Office. Van Cott, H. P. and Kinkade, R. G. (1972). Human Engineering Guide to Equipment Design. Washington, DC: U.S. Government Printing Office. Viitasalo, J. T., Era, P., Leskinen, A. L., and Heikkinen, E. (1985). Muscular strength profiles and anthropometry in random samples of men aged 31–35, 51–55 and 71–75 years. Ergonomics, 28(11), 1563–1574. Waters, T. R., Putz-Anderson, V., Garg, A., and Fine, L. J. (1993). Revised NIOSH equation for the design and evaluation of manual lifting tasks. Ergonomics, 36(7), 749–776. White, R. M. and Churchill, E. (1971). The body size of soldiers: U.S. Army anthropometry—1966 (Technical Report 72-51-CE AD 743 465). Natick, MA: U.S. Army Natick Laboratories. Woodson, W. E., Tillman, B., and Tillman, P. (1992). Human Factors Design Handbook: Information for the Design of Systems, Facilities, Equipment, and Products for Human Use. New York: McGraw-Hill.
TAF-ER562X_WINGR-09-0401-C004.indd 151
11/9/10 2:23:30 AM
TAF-ER562X_WINGR-09-0401-C004.indd 152
11/9/10 2:23:30 AM
5 Documentation John W. Gwynne III, PhD; David A. Kobus, PhD, CPE CONTENTS 5.1 5.2 5.3 5.4 5.5
The Value of Well-Designed Documentation ..........................................................154 User Attitudes toward Documentation.....................................................................155 The Regulatory Imperative ......................................................................................156 Limitations of the Guidance in This Chapter ..........................................................157 General Principles ...................................................................................................157 5.5.1 Early Involvement Is Key..............................................................................157 5.5.2 Practice User-Centered Design .....................................................................158 5.5.3 Consider Environmental Factors...................................................................159 5.5.4 Conduct a Task Analysis ...............................................................................159 5.5.5 State Documentation Specifications .............................................................159 5.5.6 Produce Draft Documentation ......................................................................160 5.5.7 Iterate Usability Testing and Documentation Development .........................160 5.5.8 Documentation Design Checklist ................................................................. 161 5.6 Special Considerations............................................................................................. 161 5.6.1 Electronic Documentation ............................................................................ 161 5.6.1.1 Information Access .........................................................................163 5.6.1.2 Limitations ......................................................................................164 5.6.2 Documentation for Lay Users .......................................................................164 5.7 Design Guidelines....................................................................................................165 5.7.1 Guidelines for All Documentation ................................................................165 5.7.1.1 Content Guidelines ..........................................................................165 5.7.1.2 Presentation Guidelines ................................................................... 175 5.7.2 Guidelines for Electronic Documentation ....................................................187 5.7.2.1 Navigation .......................................................................................187 5.7.2.2 Hyperlinks .......................................................................................188 5.7.2.3 Language and Readability ...............................................................189 5.7.2.4 Organization and Layout .................................................................189 5.7.2.5 Illustrations ......................................................................................190 5.7.2.6 Web-Specific Graphics ....................................................................190 5.7.2.7 Highlighting .................................................................................... 191 5.7.2.8 Typography ......................................................................................192 5.7.2.9 Physical Characteristics ...................................................................193
153
TAF-ER562X_WINGR-09-0401-C005.indd 153
11/9/10 12:40:50 AM
154
Handbook of Human Factors in Medical Device Design
5.8 Case Studies .............................................................................................................193 5.8.1 Contact Lens Care ........................................................................................193 5.8.2 Blood Glucose Meter Operation ...................................................................195 Resources .........................................................................................................................196 References ........................................................................................................................197 Medical device documentation should give device users clear, accurate, and easy-to-follow instructions on how to set up, operate, maintain, and troubleshoot their devices. This chapter presents a human factors–based approach to designing documentation to achieve these goals and thereby promote the safe, reliable, and effective use of medical devices. Human factors principles for developing documentation are first presented to lay the groundwork for the documentation design process. Two emerging trends in medical device use—the increasing prevalence of electronic documentation and use of medical devices by laypersons—are considered in terms of their implications for documentation design. An extensive set of design guidelines is provided that applies to both paper and electronic documentation, followed by guidelines that take into consideration the unique characteristics of electronic documentation. Case studies illustrate the value of the guidelines and demonstrate the benefits of the user-centered approach advocated in this chapter. Finally, additional resources are provided for readers who desire to explore documentation design topics in more detail.
5.1 THE VALUE OF WELL-DESIGNED DOCUMENTATION Ideally, medical devices should be so well designed that no documentation is needed to operate them safely and reliably. Unfortunately, this design ideal has proved elusive. Device use can expose users and patients to many potential hazards (e.g., radiation, electric shock) that can be mitigated or reduced in several ways, including (1) improving the device user interface (controls, displays, on-device labeling, logic of operation), (2) providing initial and refresher training, and (3) creating documentation that gives users the information they need to safely operate and maintain a device (Kaye & Crowley, 2000). The increasing complexity and power of medical devices have placed greater demands on the persons who operate, maintain, and repair them. Complete, unambiguous, accurate, and easily understood documentation is important for all users, whether they are professionals in a health care setting or laypersons operating and maintaining a device in the home (see Chapter 18, “Home Health Care”). Documentation should make users aware of the potential hazards associated with device use and how to minimize them. Hazards associated with use errors represent a serious public health concern, as documented by the U.S. Food and Drug Administration’s (FDA’s) Medical Device Reporting program. The literature (Bogner, 1994; Donchin et al., 1995; Leape et al., 1991) indicates that the frequency and consequences of hazards arising from use errors exceed those arising from device failures. Numerous studies have shown that documentation for medical devices is often inadequate. For example, a study conducted by the FDA (Kingsley, 1995) found that nurses and doctors did not consider medical device information and labeling useful for several reasons, including (1) inadequate information content and formatting; (2) poor distribution methods; (3) the availability of other, more effective information sources; and (4) questions about the
TAF-ER562X_WINGR-09-0401-C005.indd 154
11/9/10 12:40:50 AM
Documentation
155
validity of some labeling. Clearer descriptions of device operating procedures and more specific troubleshooting information were the two improvements most desired by users. Unfortunately, documentation is often considered only as an afterthought to the device development process. This lack of integration between the design of the device and the development of associated documentation results in documentation that fails to provide users with information they can readily understand and use. This failure has contributed to poor user attitudes toward documentation, which we now consider in more depth.
5.2 USER ATTITUDES TOWARD DOCUMENTATION To write effective documentation for medical devices, it is necessary to understand device users’ attitudes toward documentation and how those attitudes influence their behavior. Most users typically prefer to avoid reading documentation and instead obtain information about device operation and troubleshooting from other sources, such as in-service training programs and training from health care professionals or the technical support staff (Kingsley, 1999). Professional caregivers generally prefer hands-on training to reading and following instructions in a user manual (Wiklund, 2004). When users do refer to documentation, they typically do not read it thoroughly, cover to cover. Instead, they scan it to locate specific topics of interest. Given the reluctance of many device users to refer to documentation, special efforts are needed not only to encourage them to use documentation but also to anticipate their needs and to create documentation that meets those needs. For instance, documentation should make it easy to locate information by (1) ordering topics in the same sequence they occur during device operation, (2) highlighting key words, and (3) organizing topics in a shallow hierarchy of clearly labeled headings and subheadings. Thus, the traditional format for medical device use documentation, the user manual, is not structured to facilitate user’s preferred style of interaction. Most user manuals are written with the goal of satisfying regulatory requirements for labeling rather than giving users a clear and usable information source for operating and maintaining a device. User manuals are consequently looked on with disdain by many users, who consider them to be a resource of last resort. A well-designed user manual can be a valuable user aid, but its development requires greater planning and effort than is typically expended. In addition to a user manual, it is often advisable to furnish a quick reference guide that describes basic and critical aspects of device operation and troubleshooting. Short, succinct, easily accessible, and highly relevant documentation will almost always be used more often than more comprehensive forms of documentation. Given that most medical device users resort to documentation only when they encounter a problem they cannot solve on their own or with the help of others, it is imperative to include troubleshooting information. A quick reference guide, for example, could present both device operating procedures and troubleshooting help. It is worth noting that regional differences exist in the use of device documentation. For instance, Europeans typically refer to user manuals more frequently than Americans (Wiklund, 2002). User manuals may also provide a preferred way of learning for some individuals despite a general tendency to the contrary as described above. Excellent device documentation can also serve as a starting point for in-house training programs. Therefore, even though documentation may not be widely used, it remains an important aspect of device design, development, and implementation.
TAF-ER562X_WINGR-09-0401-C005.indd 155
11/9/10 12:40:50 AM
156
Handbook of Human Factors in Medical Device Design
The manufacturer who waits until a medical device is at the final prototype stage before developing that device’s user documentation is missing out on a real opportunity. The creation and testing of user documentation often identifies shortcomings in a device’s user interface. Thus, manufacturers who integrate user documentation with other aspects of device design (including other types of labeling) will produce a more usable and probably safer and more effective device.
5.3 THE REGULATORY IMPERATIVE As alluded to above, governmental regulatory requirements constitute a major motivation for device manufacturers to create documentation. All medical devices manufactured in the United States are regulated by the FDA. The FDA requires device manufacturers to furnish printed documentation with their devices; such documentation is considered an extension of a device’s labeling, which details the conditions for its safe and effective use (Wiklund, 2002). The Federal Food, Drug, and Cosmetic Act (FFDCA) is the law under which the FDA regulates all medical devices. Section 201(m) of the FFDCA defines labeling as “all labels and other written, printed or graphic matter (a) upon any article or any of its containers or wrappers, or (b) accompanying such article” at any time while a device is held for sale after shipment or delivery for shipment in interstate commerce. The term accompanying is interpreted liberally to include posters, tags, pamphlets, circulars, booklets, brochures, instruction manuals, fillers, quick reference guides, troubleshooting guides, and so on. Accompanying also includes labeling that is brought together with the device after shipment or delivery for shipment in interstate commerce (see also Chapter 13, “Signs, Symbols, and Markings”). All medical device documentation must comply with the applicable law under the U.S. Code of Federal Regulations (2003a, 2003b), Title 21, Parts 801 (Labeling) and 809 (In Vitro Diagnostic Products for Human Use). An important consequence of FDA oversight is that documentation must meet the criteria set forth by the FDA before a medical device can receive FDA approval. Likewise, devices produced abroad must meet the regulatory requirements of those government organizations established to oversee medical device development and distribution. All medical devices and their accessories placed on the European Union market since June 15, 1998, must comply with the requirements of the Medical Devices Directive, which include labeling requirements. Device manufacturers that market their devices in Europe must comply with all essential requirements in Annex I to Directive 93/42/ EEC on medical devices as well as with the member states’ individual implementing legislation. The Medical Devices Directive provides that member states may require the accompanying information to a medical device to be in their national language or in another community language. All member states require that safety information be provided in their official language to ensure understanding by the end user. Device manufacturers thus face a heavy burden to provide documentation that not only is complete, correct, and understandable but also has been accurately translated from its language of origin to the national languages used in the countries where the device is marketed. The penalties for noncompliance with the requirements of the Medical Devices Directive vary from the removal of the device from the market to economic penalties and, in some cases, to criminal proceedings (Pilot, 1999).
TAF-ER562X_WINGR-09-0401-C005.indd 156
11/9/10 12:40:50 AM
Documentation
157
The human factors design guidance presented in this chapter should be used in conjunction with government regulatory guidance documents to create high-quality documentation that satisfies regulatory requirements while embodying sound human factors. In this way, documentation will incorporate the necessary regulatory content while still meeting users’ needs. In the United States, the FDA’s Office of Device Evaluation, Center for Devices and Radiological Health, has prepared a General Program Memorandum (#G91-1), Device Labeling Guidance, to assist designers in complying with FDA regulations for medical device labeling.
5.4 LIMITATIONS OF THE GUIDANCE IN THIS CHAPTER • The human factors design guidance in this chapter cannot be implemented in a cookbook fashion. Instead, it should be adaptively applied to each design project. The effectiveness with which this is done depends on the knowledge and expertise of the individual designer. There is no substitute for the skilled application of human factors principles and guidelines when developing documentation. • Understanding the rationale underlying human factors principles is indispensable for the proper application of the design guidelines in this chapter. No single approach or format is best for all documentation. The type of documentation being created, the procedural complexity of the device, the skills and knowledge needed to operate a device, the target user group, and the anticipated use environment are among the contextual factors that must be considered when developing documentation. • Good documentation cannot compensate for bad device design. If a device is poorly designed in terms of its human factors characteristics, good documentation will not improve the user interface. For each to be most effective, device design and documentation design should proceed in parallel, each process benefiting from the other through successive design–test iterations. Documentation design is best viewed as part of a larger effort to create devices that are easy, safe, and effective to use through the application of human factors principles and guidelines. This chapter furnishes general guidance for developers of medical device documentation. It is not intended to serve as a guide to satisfying governmental regulations for medical device labeling (although the information presented here is consistent with FDA publications and the literature). The designer should consult the governmental agencies that regulate medical devices in their intended market(s) for additional information on their requirements for device documentation.
5.5 GENERAL PRINCIPLES This section presents a general approach to creating documentation, based on human factors principles and guidelines for information presentation, training, documentation design, and human performance capabilities. This approach consists of a set of largely sequential activities, most of which apply equally to the development of paper and electronic documentation.
5.5.1 EARLY INVOLVEMENT IS KEY As discussed previously, the design and preparation of documentation should begin early in the device design and development process, preferably during the device specification phase.
TAF-ER562X_WINGR-09-0401-C005.indd 157
11/9/10 12:40:50 AM
158
Handbook of Human Factors in Medical Device Design
This will enable the documentation designer to gain a thorough understanding of the device and its documentation requirements. The designer should critically evaluate (1) the types of documentation needed for a device, (2) characteristics (educational, physical, sensory, and cognitive factors) of its intended end users, and (3) the effects of the anticipated use environment on how a device will be implemented and used. Not only can documentation designers furnish information useful for device designers, but a close working relationship between them will help ensure that draft documentation is available for the usability testing of device prototypes. Unfortunately, documentation development is often begun only after the device design and development effort is far along, at which point it is too late for the documentation designer to make meaningful contributions to the device design and refinement process (Callan, Gwynne, Sawyer, & Tolbert, 1993).
5.5.2 PRACTICE USER-CENTERED DESIGN Documentation designers must take into account the varied needs, skills, knowledge, and abilities of increasingly diverse user groups, along with the characteristics of the environments in which devices are used. User-centered design emphasizes user requirements, capabilities, needs, tasks, and goals and incorporates them as early as possible in the design process, when changes can still be made in a time and cost efficient way (Czaja & Nair, 2006; Meister & Enderwick, 2001; Nemeth, 2004). From the viewpoint of user-centered design, a key aspect of developing effective documentation is to characterize the anticipated users of a device in terms of the skills, knowledge, and abilities needed to use the device safely and effectively (see Chapter 2, “Basic Human Abilities”). Users vary along numerous dimensions relevant to documentation design, including the following: • Educational and reading levels, language comprehension, and familiarity with Web-based navigation • Physical abilities or disabilities related to visual and auditory perception, color vision, and motor functioning • Experience with similar technology or devices, including an evaluation of previous experience that could produce negative transfer (i.e., any user habits or preconceptions that could interfere with understanding the documentation or learning to use the device) • Understanding of operating principles and potential hazards associated with the technology The documentation must be designed to accommodate a wide range of skills, knowledge, and abilities. Horton (1994) identifies four types of medical device users: • Novice users are beginners. They will have little, if any, experience using medical devices and have at best only a limited knowledge of related topics. • Occasional users have previous experience with a device but are not familiar enough with it to remember details of its operation. • Transfer users have used similar types of medical devices before but not the one your documentation covers. These users understand general aspects of using medical devices but need help to gain proficiency on the new device.
TAF-ER562X_WINGR-09-0401-C005.indd 158
11/9/10 12:40:50 AM
Documentation
159
• Expert users have experience performing the tasks required to operate and maintain a device. If they refer to documentation, it will most likely be for troubleshooting help in response to an anomalous situation or to access a rarely used device capability. For many devices, it is likely that persons from several of these user categories will use the documentation. Identify the most common user categories for a given device and design the documentation to be appropriate for those categories.
5.5.3 CONSIDER ENVIRONMENTAL FACTORS Environmental factors should be considered during documentation design because they can impose important constraints on how devices are used. Environmental factors include lighting, noise levels, and the presence of other devices. These factors can act singly or in combination to affect the ability of device users to understand and follow documentation for device operation and maintenance. For example, a bedroom in a home may be poorly lit, making it harder to access and read device-based documentation. Refer to Chapter 3, “Environment of Use,” for an in-depth treatment of this important topic.
5.5.4 CONDUCT A TASK ANALYSIS To write effective documentation, it is vital to have a thorough understanding of how a medical device will be used to achieve its intended purpose. A task analysis is indispensable to gaining this understanding. Ideally, a task analysis should be conducted during the early stages of device design by human factor professionals skilled in task analysis techniques. A task analysis defines in detail all aspects of device operation and maintenance, including (1) the equipment, supplies, and materials needed; (2) the functions, tasks, and procedural steps that must be performed to properly set up, calibrate, operate, and maintain a device; (3) the time needed to perform these functions, tasks, and procedures; (4) the types of errors that can occur and the likelihood of those errors for the various procedures (Hollnagel, 2006); and (5) the procedures involved in identifying and troubleshooting device operational problems (Kirwan & Ainsworth, 1992). If operating and maintaining a device involves tasks that have a large cognitive aspect, the task analysis should identify the memory, attention, and other mental demands placed on device users. A task analysis furnishes the basis for writing step-by-step operating instructions, developing warning and caution statements, and providing complete and accurate device descriptions and other useful information to device users.
5.5.5 STATE DOCUMENTATION SPECIFICATIONS An important part of creating documentation is enumerating its content and format specifications. These specifications should take into account the characteristics of the intended users and the environment(s) in which a device will be used. Specifications will vary from one device to another; however, documentation specifications for most devices will include the following: • A list of the types of documentation required, such as user manuals, maintenance manuals, reminder sheets, checklists, and quick reference guides. These
TAF-ER562X_WINGR-09-0401-C005.indd 159
11/9/10 12:40:50 AM
160
• • • • •
Handbook of Human Factors in Medical Device Design
requirements will be based on the type of use that is intended for a device, but a user manual with complete device information and a quick reference guide that contains basic information on device operation and troubleshooting are likely to be needed. A specification for readability, including a list of requirements such as table of contents, indexing, and headings. A list of languages to be used and a list of special needs or restrictions on the documentation, such as limitations on color, font size, and document size. An outline for each type of documentation, including a list of expected illustrations. A schedule and plan for developing the documentation that is clearly linked to the device design schedule. A schedule for the test and evaluation of the documentation.
This information facilitates documentation design and development by providing both the necessary information and timetables for implementing those specifications that coincide with device design and development.
5.5.6 PRODUCE DRAFT DOCUMENTATION Once the specifications for documentation have been defined, the next step is to design draft versions of each type of documentation that will be created for a device. Produce the user manual in parallel with detailed device design or prototyping to facilitate usability testing of both the manual and the device. Depending on the nature of the device and its intended users, the user manual could include procedures for setup, installation, calibration, normal operation, emergency operation, maintenance, and troubleshooting. A quick reference guide, a valuable companion to the user manual, can reference specific pages in the user manual, making it easy for users to obtain additional information as needed. Development of maintenance and service manuals, if they are necessary, should start at the same time that prototype devices begin to resemble production models. Development of the user, maintenance, and service manuals should proceed in parallel with the device design process to help identify deficiencies in user-interface design. Apparent interface design shortcomings can be fed back to the device designers to improve the user interface.
5.5.7 ITERATE USABILITY TESTING AND DOCUMENTATION DEVELOPMENT Usability testing, in which target users refer to the documentation while using a device, should be done as part of the design process to assess how users perceive and respond to the documentation (Dumas & Redish, 1993; Nielsen, 1994; Rubin, 1994). The prototype documentation is tested on a representative sample of the anticipated end users of a device. Documentation evaluation should focus on content sufficiency, user comprehension, and documentation usability. Content sufficiency addresses the accuracy and completeness of the information contained in the documentation. User comprehension concerns the ability of device users to understand the documentation’s procedural instructions, warnings, troubleshooting advice, and supporting information. Testing should involve actually observing users while they attempt to operate the device using the documentation as a guide. Based on the usability test results, the documentation
TAF-ER562X_WINGR-09-0401-C005.indd 160
11/9/10 12:40:51 AM
Documentation
161
should be redesigned and then retested. Before initial testing, acceptance criteria should be established (See Chapter 6, “Testing and Evaluation”). This iterative design–test process is repeated until users can use the documentation effectively. The test and revision of the documentation should proceed in parallel with the test and development of the actual device. In this way, a close relationship is established between device documentation, operating characteristics, and interface features.
5.5.8 DOCUMENTATION DESIGN CHECKLIST The checklist in Table 5.1 provides a comprehensive summary of the characteristics of effective documentation and important activities in the documentation design process. Some items in the checklist may not apply to some types of documentation.
5.6 SPECIAL CONSIDERATIONS Documentation design efforts must be responsive to relevant trends in health care delivery. One such trend is the use of electronic means to acquire, store, and transmit health care information. Electronic documentation will become increasingly common in the future. The documentation designer must understand the unique properties of electronic documentation to make the most effective use of them. A second trend is the increasing prevalence of home-based health care. This trend has focused attention on the documentation needs of lay users, who often have little or no prior background in using medical devices, especially those that have migrated from professional settings into the home. Providing documentation that is easy for lay users to understand and follow is crucial to the safe and effective home use of medical devices.
5.6.1 ELECTRONIC DOCUMENTATION Computerization in health care delivery and care documentation has become ubiquitous in recent years. Hospitals are taking an increasingly paperless approach to vital signs recording, laboratory testing, staff communication, and clinical note taking (Wiklund, 2002). Following this trend, device documentation has also become increasingly computerized, although regulatory requirements for paper documentation still exist. It is thus important to be aware of how electronic documentation differs from paper documentation and to provide design guidelines that take these differences into account. Electronic documentation is any documentation information (1) stored on videotape, film, audiotape, or computer storage medium (hard disk, compact disk, floppy disk, DVD); (2) delivered online via a Web-based or e-mail user interface; or (3) delivered in real time via visual or audible modalities by the device itself. The challenge for the documentation designer is to exploit the unique features of these electronic media to create documentation that enhances the device user’s experience over and above what can be achieved with paper documentation. For example, electronic documentation can be developed so that it is adaptive and intelligent. Adaptive documentation keeps a record of a user’s interaction with it to identify procedures that have been problematic or topics that are of special interest to that user. When the user again seeks information on these topics, the system can provide the user with additional relevant information based on earlier interactions without requiring the user to submit detailed requests or queries for that information.
TAF-ER562X_WINGR-09-0401-C005.indd 161
11/9/10 12:40:51 AM
162
Handbook of Human Factors in Medical Device Design
TABLE 5.1 Documentation Design Checklist Our documentation: Complies with all labeling regulations Is written for the type of people who use our device Tells the user how to get help from the manufacturer Includes a table of contents Has general warnings and precautions near the beginning Briefly describes the purpose of the device Gives a physical description of the device with a labeled graphic Explains conditions under which the device should and should not be used Gives clear setup instructions Gives clear pre-use checkout procedures Gives clear and easy-to-follow operating instructions in a step-by-step format Provides cleaning instructions Describes maintenance that the user should do Explains how to store the device Has a clear, easy-to-use, and easy-to-find troubleshooting section Has a summary page with all the critical information on it Has an alphabetized index Has an easy-to-find date of issue Includes instructions on any accessories Is laid out to make sections easy to find Uses white space and other highlighting techniques to focus user attention on important information Is printed in at least 12-point type Has clear, well-labeled graphics in key places to help users understand the text Uses proper contrast so that text and graphics can be easily viewed We have: Done a task analysis for the procedures in our documentation Selected a suitable format (text, flowchart, list) Written and formatted procedures correctly Used appropriate sentence construction and word choice Tested our documentation to ensure a sixth- to seventh-grade reading level Properly written and placed specific warnings and cautions Avoided technical terms and jargon unfamiliar to our users Our manual: Has a durable distinctive cover Will stand up to the conditions in which it will be used Is constructed of nonshiny, durable paper Is laid out to make things easy to locate and update (e.g., tabs, color codes, binding) Lays flat on its own We have: Tested the documentation to make sure that its target users can read, understand, and follow it Taken steps to make sure that our documentation is available to our users Source: Backinger, C. L. and Kingsley, P. A. (1993, August). Write It Right: Recommendations for Developing User Instruction Manuals for Medical Devices Used in Home Health Care (HHS Publication FDA 93-4258). U.S. Department of Health and Human Services, Food and Drug Administration, Rockville, MD.
TAF-ER562X_WINGR-09-0401-C005.indd 162
11/9/10 12:40:51 AM
Documentation
163
Although electronic documentation shares many characteristics with paper documentation, it also possesses several unique features, such as sound, voice, animation, and ease of updating. These features have important implications for documentation design. While potentially useful, these features can also distract and confuse users if they are not implemented wisely. The design process for electronic documentation involves the same basic activities as does the development of paper documentation (e.g., task analysis, requirements definition, iterative testing and redesign). However, the unique features of electronic documentation require that specific questions be addressed in terms of user needs and the device operating environment. These design questions include the following: • Are the device procedures (operating, troubleshooting, and so on) complex enough to justify the costs of an electronic system? Paper documentation may be a better choice for simple procedures. • Is electronic documentation practical for the intended use environment (e.g., patient’s home, ambulance, emergency room)? • What form of information delivery would be most practical (e.g., device based, Web based, CD-ROM, DVD, audiotape)? For example, although device-based documentation might be more practical in some environments, what recourse would be available in the event of device failure? If troubleshooting procedures were part of the device-based documentation, they might not be accessible at the very time when they are needed most. • Could user limitations make electronic documentation hard to use? Limitations in vision, hearing, or manual dexterity may limit the value of electronic documentation for some user groups. • Will the intended users have ready access to computers or the other technologies needed to access electronic documentation? For instance, older persons may not own a computer or, for that matter, the home electronics needed to play DVDs and CD-ROMs. 5.6.1.1 Information Access The ability to access information quickly and efficiently is one of the potential advantages of electronic documentation. However, from the designer’s viewpoint, information access in electronic documentation is complicated by the fact that users often do not thoroughly read most forms of electronic documentation in a linear fashion. Instead, they skip around from one topic to another. As a consequence, readers rely on visual clues to the organization of documentation, especially when they are reading to follow procedures or make decisions (Wright, 1999). The documentation designer has little or no control over what parts of the documentation a user reads prior to reaching a particular topic or where a user will go next. Although this is also true for paper documentation, it is more pronounced for electronic documentation because of its enhanced navigational features, such as hyperlinks. An important design implication of the increased ease of navigating electronic documentation is that each topic should be self-contained and “stand alone” as much as possible from other topics. A given topic should answer a single question or address a single aspect of device operation. Insofar as possible, do not organize electronic documentation so that it cumulatively builds on topics presented earlier in the documentation. If a reference is made to earlier topics, make it easy for the reader to access them, for example, by using hyperlinks.
TAF-ER562X_WINGR-09-0401-C005.indd 163
11/9/10 12:40:51 AM
164
Handbook of Human Factors in Medical Device Design
Ease of access to important information is obviously critical, but the “how,” “when,” where,” and “why” require careful planning based on a thorough understanding of user information requirements and device operating characteristics. Given the linear, procedural nature typical of medical device operation, sets of procedural steps could be linked directly to illustrations of related controls and displays. Likewise, long, complicated procedures could be linked to overview information, place-keeping “roadmaps,” or warnings. But there are always trade-offs. For example, warnings critical to correct device operation and patient safety should always appear adjacent to the relevant step. In some cases, space limitations may dictate that less critical or general warnings be located in a separate section devoted to warnings. Make a strong effort to ensure that warnings can still be easily accessed, however. For example, several different steps could be linked to the same general warning. Users could then preview procedural steps from an earlier section on general warnings or vice versa. 5.6.1.2 Limitations Electronic documentation may not always be a practical solution. Likely drawbacks of electronic documentation include higher initial development and implementation costs. Depending on the form of electronic documentation that is used, other limitations include ongoing demands for computer storage space, display limitations associated with computer monitors and device screens, and the need for training for users who are not computer literate. In this last regard, overcoming initial user resistance to electronic documentation is a potential stumbling block. Some lay users may not have access to a computer or simply feel uncomfortable using electronic documentation. Another important drawback is that electronic documentation may not readily be accessible when it is needed most. For example, some clinical environments, such as emergency rooms or ambulances, may not be conducive to the use of electronic materials. In addition, if a device becomes inoperable, device-based documentation will be unavailable. Many devices with electronic documentation will also provide paper documentation. The choice to use paper or electronic documentation should be left to individual users who can then select the one that meets their particular needs. When both forms of documentation are to be provided, both should be designed on the basis of established human factors principles and guidelines. It is best to develop paper and electronic documentation concurrently, relying on a core set of content that can be expressed in either paper or electronic form, depending on individual users’ needs and preferences. Following this strategy, discrete units of text can be developed both as hypertext elements for incorporation into nonlinear, hyperlinked electronic documentation and as regular text for paper documentation. Importantly, because actual usage of the two forms of documentation may be different, they generally will not be identical in structure or format (e.g., a PDF of paper documentation will yield few of the benefits of full electronic documentation).
5.6.2 DOCUMENTATION FOR LAY USERS In recent years, changes in health care economics, combined with an aging population, have resulted in many patients being cared for at home. Traditionally, most medical devices have been used by health care practitioners in professional settings such as hospitals and doctors’ offices. However, as patient care has moved into the home, the medical devices needed for this care have followed, where they often are used by laypersons, including patients, family members, friends, or other nonprofessional caregivers. Lay users rely on
TAF-ER562X_WINGR-09-0401-C005.indd 164
11/9/10 12:40:51 AM
Documentation
165
whatever firsthand training and follow-up support they receive from health care professionals, along with device documentation. Lay users represent an extremely broad and diverse group in terms of the factors that influence how well they can operate a medical device. These factors include educational background, age, technical sophistication, physical capabilities, mental condition, and lack of prior relevant experience or training. In particular, the aging patient or lay caregiver presents challenges for designing both safe and usable devices and documentation (Rogers, 1997; Rogers & Fisk, 2001; Rogers, Fisk, & Walker, 1996; Vanderheiden, 2006). The home environment itself poses additional challenges to safe and effective device use, including the following (McCarthy et al., 1992): • • • •
Inadequate or inappropriate physical environments for device operation Inadequate infection control practices Inadequate communication channels Lack of quality control standards or procedures
For additional information about the challenges of designing documentation for lay users and the home environment, refer to Chapter 2, “Basic Human Abilities”; Chapter 3, “Environment of Use”; and Chapter 18, “Home Health Care.”
5.7 DESIGN GUIDELINES This section presents human factors design guidelines for creating effective user documentation. These guidelines are organized into two main parts. The first part presents guidelines that apply to both paper documentation and electronic documentation. The second part presents guidelines specifically for electronic documentation.
5.7.1 GUIDELINES FOR ALL DOCUMENTATION The guidelines in this section are organized into two major categories: content and presentation. Content refers to the type of information that should be included in device documentation, such as background information, procedures, risk communication, and supplementary information. Presentation refers to the manner and style used to convey that information, which includes language and readability, organization, illustrations, layout, highlighting, typography, and physical characteristics. With only a few noted exceptions, these guidelines apply equally to both paper and electronic documentation. They are appropriate for all forms of documentation, including user manuals, quick reference guides, online information, and device-based help systems. 5.7.1.1 Content Guidelines The content guidelines are organized into six categories: • • • • • •
Background information Procedures Troubleshooting information Warnings and precautions Supplementary information Translations
TAF-ER562X_WINGR-09-0401-C005.indd 165
11/9/10 12:40:51 AM
166
Handbook of Human Factors in Medical Device Design
Together, these topics address the types of information that device users need to know to be able to safely and reliably install, set up, calibrate, operate, clean, store, maintain, troubleshoot, and repair a device. The appropriate content for device documentation can be determined by several methods, including discussions with device designers, medical personnel, and end users. Additionally, applying human factors techniques, such as task analysis, assists in determining the cognitive, perceptual, and physical aspects of device use. The checklist in Table 5.2 shows the numerous topics that content information can encompass, depending on the device. 5.7.1.1.1 Background Information There is often information of which device users should be aware before using a device. This background information should be presented in the first part of the documentation, before the procedures on how to operate and maintain a device. Depending on the nature of the device, some or all of the following guidelines are appropriate for background information. Guideline 5.1: Describe the Purpose of the Device Documentation should describe the purpose of the device, the medical needs of persons who use it, and indications and contraindications for its use. Explain how any information that is supplied by the device should be used.
Guideline 5.2: Furnish Overview and Contact Information A statement at the beginning of the documentation should identify critical information and instructions to be read before device use. Give an overview of basic device operation and list manufacturer contact information (toll-free telephone number, e-mail address, Web address, and so on) to obtain help with device use or problems.
Guideline 5.3: Describe Device Components Documentation should describe each device component in enough detail so that the user can understand the role it plays in device operation. Include labeled illustrations that clearly show each component. Emphasize the need for users to familiarize themselves with all device components before using the device. If the device is electronic or mechanical, provide explanations for all major device features and operations, such as the following: • • • • • • • •
Operating modes Menus and navigation Display messages Control actuation Battery loading and testing Cleaning and maintenance Calibration Device-based help
Guideline 5.4: Provide General Warnings A warning is a statement that makes a user aware that an adverse health consequence may result from device use. A warning can also discuss hazards associated with noncompliant
TAF-ER562X_WINGR-09-0401-C005.indd 166
11/9/10 12:40:51 AM
Documentation
167
TABLE 5.2 Checklist of Content Information for Device Documentation —
User Assistance Information (e.g., toll-free phone number)
— Table of Contents — Glossary Descriptive Information — Name, other specific identifiers — Description of device features and characteristics — Purpose of device (indications for use) — When device should not be used (contraindications) — Potential risks/benefits — Expectations of device and procedure(s) associated with device — General and specific warnings and precautions — Importance of the need to adhere to a care regimen Operating Information (as applicable) — — — — — — — — — —
Setup instructions Calibration instructions Patient preparation instructions Operating instructions Graphics, tables, videos, animation, and pictures to accompany procedural steps Importance of the need to monitor the activity of the device Cleaning instructions Description of maintenance and who should do it Storage instructions Expected failure time and mode and its effect on the patient
Troubleshooting Information — — —
Instructions on how to safely dispose of the device Instructions on accessories Instructions on related, additional devices
Additional Information — — — — — — — —
Scientific information/clinical studies Disease and self-care information Adverse events Comparative information (e.g., success rates) Warranty Travel/international use Index Date of Printing
Source: Adapted from Food and Drug Administration (2001). Guidance on Medical Device Labeling: Final Guidance for Industry and FDA Reviewers. Food and Drug Administration, Center for Devices and Radiological Health, Rockville, MD.
device use or pitfalls in interpreting test results. There are two types of warnings: general and specific. General warnings consist of crucial information needed before operating a device, such as the need to stop using a device if certain symptoms arise. Present general warnings in the background section of the documentation. Avoid technical jargon so that a lay user can understand the warning. Specific warnings pertain to particular aspects of device operation.
TAF-ER562X_WINGR-09-0401-C005.indd 167
11/9/10 12:40:51 AM
168
Handbook of Human Factors in Medical Device Design
Present specific warnings in the appropriate procedure sections, adjacent to the step to which they apply.
Guideline 5.5: State Conditions of Device Use State the environmental conditions required for safe device operation. Identify all conditions under which a device should not be used. Cite any conditions that can adversely impact device operation, such as excessive temperature or humidity, and the consequences of operating the device under those conditions.
Guideline 5.6: Specify Supplies and Materials Describe and illustrate all the supplies and materials needed to operate and maintain the device. Specify the quantity, size, and type for all supplies and materials and whether they accompany the device. Provide information about any special handling required of the supplies, as needed.
Guideline 5.7: Clarify Differences when Multiple Models Exist When multiple models or versions of the same device exist and their user interface or operation differs, it is generally better to provide separate and unique documentation for each model/version. However, if documentation provides information about several different models or versions of the same device, provide a clear indication of how the models/versions differ and which aspects of the documentation apply to which model(s) or version(s).
Guideline 5.8: Describe Patient Preparation Patient preparation requirements should be clearly stated. For example, some medical devices require that patients ingest or do not ingest specific substances or foods prior to or during device use.
Guideline 5.9: Explain Device Storage Conditions Describe the proper preparation of the device for storage and desired storage conditions. Emphasize any conditions that could damage the device or cause it to malfunction in subsequent use. State the consequences of improper storage. Indicate how extended storage can adversely affect the device (Backinger & Kingsley, 1993).
5.7.1.1.2 Procedures A procedure is a set of instructions that tells a user how to operate, maintain, or troubleshoot a device. The clear and orderly presentation of procedures is vital to effective documentation. Understanding and following device operating instructions, which often contain both text and illustrations, requires a broad range of literacy skills. The following guidelines will help designers write easy-to-follow and effective operating procedures. Guideline 5.10: Include All Device Procedures Include all the procedures that a device user must perform when setting up, operating, maintaining, or troubleshooting a device. This information should derive from the task analyses performed during device design and development.
TAF-ER562X_WINGR-09-0401-C005.indd 168
11/9/10 12:40:51 AM
Documentation
169
Guideline 5.11: Provide Clear and Logical Procedure Sequences Organize procedures into clear and logical steps that can be followed accurately and reliably.
Guideline 5.12: Hierarchical Procedure Descriptions Procedures should be organized hierarchically so that related steps are grouped into sections. These sections should be given heading names that reflect their content (Wieringa, Moore, & Barnes, 1993).
Guideline 5.13: Write Procedures as Step-by-Step Instructions Procedures should be broken down into a series of short, discrete steps rather than a single long paragraph. Step-by-step instructions make it easier for users to understand and follow device operating and maintenance procedures. Each procedure usually includes the following information, as appropriate (Wieringa et al., 1993): • • • • •
Conditions under which the procedure should (or should not) be performed Supplies and equipment needed Timing requirements for time-critical steps Examples to clarify text and illustrations Factors that influence device use or output, including special circumstances
Guideline 5.14: Include All Steps and Conditions For each procedure, specify all the required steps and necessary conditions for safe and effective device operation and maintenance.
Guideline 5.15: Write Procedures That Match Users’ Skills and Knowledge Use a level of detail that that is appropriate for the intended users. The appropriate level of detail depends on the type of procedure, the frequency with which it is performed, and the skill and knowledge of the users (Wieringa et al., 1993).
Guideline 5.16: Write Procedural Steps as Positive Statements Write action instructions as positive statements, telling users in succinct, everyday language what actions they should perform.
Guideline 5.17: Selectively Use Negative Statements Concise, nonambiguous negative statements are appropriate when users need to be told what actions they should be sure not to perform (Wieringa et al., 1993).
Guideline 5.18: Include Rationales for Individual Steps Providing succinct rationale helps motivate users by explaining why a particular action is required. Keep rationales brief and include them with the applicable procedural steps.
Guideline 5.19: Avoid Nonessential Content Do not dilute procedures with lengthy justifications and rationales; they divert the reader’s attention from the procedural steps. Documentation for lay users should omit extraneous details or technical content that might draw attention away from important aspects of device operation.
TAF-ER562X_WINGR-09-0401-C005.indd 169
11/9/10 12:40:51 AM
170
Handbook of Human Factors in Medical Device Design
Guideline 5.20: Number the Individual Steps in a Procedure Procedural steps should be clearly broken out and numbered consecutively in Arabic numerals. Do not “bury” procedures in text. Figure 5.1 shows how a sequence of individual actions is specified to accomplish a procedure step (Callan & Gwynne, 1993).
Guideline 5.21: Use Illustrations to Clarify Text Clean and unambiguous illustrations will make the procedural steps described by text much easier to follow. Illustrations also help users who are not proficient in the given language to better understand how to operate a device. Refer to Section 5.7.2.5 on illustrations in this chapter for additional guidance.
Guideline 5.22: Link Procedural Text to Illustrations Explicitly link text and accompanying illustrations. Choose a convention for positioning text and illustrations in relation to each other and apply it consistently throughout the documentation.
Guideline 5.23: Use Explicit Cross-References Whenever possible, include cross-references explicitly in procedures (Wieringa et al., 1993). Cross-reference rather than include nonessential information (Horton, 1994).
Guideline 5.24: Minimize Cross-References Refrain from constant cross-referencing, especially if the information is essential. However, cross-references may be necessary to ensure accuracy and to simplify complex conditional information (Wieringa et al., 1993).
Guideline 5.25: Clearly Identify All Cross-References Specify any additional procedural step(s) or information to which the user should refer.
Guideline 5.26: Verify Cross-References When Revising Documentation Whenever documentation is revised, the accuracy of the identification information in its crossreferences must be checked. If inaccurate, cross-references can introduce errors in revised procedures (Wieringa et al., 1993).
STEP 5 – Place lens on your eye. (a) Place your lens on the tip of your index finger. (b) Hold down the lower eyelid with the middle finger of the same hand and look up. (c) Place the lens on the lower portion of the white part of the eye. (d) Look down and remove your finger from the lens. The lens should then center itself.
FIGURE 5.1
Example of a procedural step decomposed into specific actionable substeps.
TAF-ER562X_WINGR-09-0401-C005.indd 170
11/9/10 12:40:51 AM
Documentation
171
Guideline 5.27: Reading and Interpreting Test Results of Home Medical Devices Lay users of home medical devices need to be told what to do with any test results generated by the device. Include a description of how test results should be read and interpreted. Guidelines 5.28 to 5.30 describe elements essential to instructions on test results management.
Guideline 5.28: Cautionary Notes on Reading Test Results Explicitly state any conditions that can influence the reading of test results. Examples of these conditions include timing of reactions, temperature, and lighting conditions under which results are read.
Guideline 5.29: Interpreting Test Results State the range of acceptable results so that invalid results can be readily ascertained. Provide an interpretation of the meaning of all possible test results. The proper treatment of ambiguous results and any limitations of the test performed by the device are important to consider when interpreting results.
Guideline 5.30: Acting on Test Results Describe the action to follow for a given result, such as seeking advice from a health care professional or retesting to confirm a test result. Make the user aware of the possibility of obtaining invalid results, the hazards likely to be encountered from acting on them, and ways to detect and correct them.
5.7.1.1.3 Troubleshooting Information A major reason that device users refer to documentation is for troubleshooting advice. As long as a device is operating properly, many users will not consult documentation. Only when a problem arises that they cannot solve will they seek out device documentation. A complete and easy-to-use troubleshooting section is therefore extremely valuable for device users. Guideline 5.31: Provide a Thorough Troubleshooting Analysis Base troubleshooting information on an analysis that specifies (1) all reasonable types of device failure, (2) indicators for each type of failure, (3) the probable cause(s) for each type of failure, and (4) diagnostic procedures for identifying the failure (Inaba, Parsons, & Smillie, 2004).
Guideline 5.32: Anticipate Likely Problems Problems can occur during device setup, operation, or maintenance. Provide solutions for each of these classes of problems (FDA, 2001).
Guideline 5.33: Group Similar Types of Problems Group similar problems together and highlight each with a group heading. Put the most lifethreatening or critical problems first in each section (FDA, 2001).
Guideline 5.34: Clearly Describe Each Problem Clearly describe the specific problem, explain its likely cause(s), and provide a clear statement of the actions the device user should follow to diagnose and correct the problem. Use as few words as
TAF-ER562X_WINGR-09-0401-C005.indd 171
11/9/10 12:40:51 AM
172
Handbook of Human Factors in Medical Device Design
possible to help the user identify the observed problem. Do not confuse the user with the technical reasons for the problem unless the reasons are important to the corrective action (FDA, 2001).
Guideline 5.35: Write Clear Procedures for User-Fixable Problems If device users can fix a problem, provide them with clear, step-by-step instructions (refer to the “Procedures” section in this chapter for guidance on writing these instructions).
Guideline 5.36: Use Everyday Language Use everyday language descriptions, not technical terminology or error codes, to tell users what is wrong with a device. Make it easy for users to understand and diagnose a problem.
Guideline 5.37: List and Explain Error Messages If the device displays error messages, list each one separately, explain what it means, and describe the actions that are needed to resolve the error condition (FDA, 2001).
Guideline 5.38: Tell Users What to Do if They Cannot Fix the Problem If there are problems that users cannot or should not try to fix, provide clear and unambiguous information to that effect, including any hazards associated with user attempts to correct the problem. Provide contact information needed to obtain assistance (e.g., toll-free telephone numbers, e-mail addresses, Web addresses). Tell users how to report undesirable outcomes (adverse events), including device malfunctions and injury resulting from device use (FDA, 2001).
5.7.1.1.4 Warnings and Precautions Warnings and precautions convey information about device operating conditions that pose potential risks to device users, patients, the environment, or the device itself. The distinction between warnings and precautions is one of degree of likelihood and seriousness of the adverse event. Warnings alert users to situations that, if not avoided, could cause death or serious injury. In contrast, precautions alert users to potentially hazardous situations that, if not avoided, may result in minor or moderate injury to the patient or device user or damage to the device but that are unlikely to result in serious injury or death (American National Standards Institute [ANSI], 2002c). Refer to Chapter 13, “Signs, Symbols, and Markings,” for an in-depth discussion of warnings. For more information on warning label design, consult ANSI Z535 (ANSI, 2002a, 2002b, 2002c, 2002d), which contains information on safety color codes, icons, and safety symbols. Ryan (1991) and Wogalter (2006) provide extensive additional information on the design and application of warnings. Guideline 5.39: Label Warnings and Precautions The signal word “WARNING” is generally used to designate a warning hazard alert. The signal word “CAUTION” is generally used to designate a precaution hazard alert (FDA, 2001).
Guideline 5.40: Place Warnings and Cautions Immediately before the Step to Which They Apply Place warnings and cautions immediately before the step to which they apply to increase the probability that the user will read them (Wogalter, Godfrey, & Fontenelle, 1987).
TAF-ER562X_WINGR-09-0401-C005.indd 172
11/9/10 12:40:51 AM
Documentation
173
Guideline 5.41: Information Elements for Warnings and Precautions Warnings and precautions should include four key elements: • Signal word (WARNING or CAUTION) accompanied by the international standard safety alert symbol • Hazard statement: a brief description of the potential hazard • Consequences: what will likely happen if instructions are not followed • Description of the actions required to avoid or minimize the hazard (“do’s” and “don’ts”) The signal word must come first. The other three elements may be presented in any order that makes sense.
Guideline 5.42: Highlight the Signal Word The signal word should be highlighted by some combination of bolding, white space, or a contrasting color to make it stand out from adjacent text and graphics. Red is often used for warnings, while yellow is the common color for cautions.
Guideline 5.43: Use Standardized Symbols and Icons Whenever possible, use internationally recognized standards as a source of symbols and icons. ANSI Z535 contains standardized and widely followed guidance “for the design, application, and use of signs, colors and symbols intended to identify and warn against specific hazards and for other accident prevention purposes” (ANSI, 2002a, p. v).
Guideline 5.44: Use the Standard Safety Alert Symbol The standard safety alert symbol, shown in Figure 5.2, indicates a potential injury hazard. It is composed of an equilateral triangle surrounding an exclamation mark. This symbol should not be used to alert persons to property damage (ANSI, 2002b).
Guideline 5.45: Test New Symbols and Icons Prior to Use If you use nonstandard or custom symbols or icons, test them before including them in your documentation. Use testing procedures like those described in ISO 9186 (International Organization for Standardization, 2001) to help ensure that users will interpret your symbols in the way you intend.
Guideline 5.46: Hazard Statement Provide a realistic and accurate description of the harmful effects associated with device use.
FIGURE 5.2
Examples of safety alert symbols recommended by ANSI Z535.
TAF-ER562X_WINGR-09-0401-C005.indd 173
11/9/10 12:40:51 AM
174
Handbook of Human Factors in Medical Device Design
Guideline 5.47: Inform of any Unusual Hazards Be sure to inform users about unusual hazards (i.e., those not generally known by the user population) as well as any hazards specific to the particular model of the device being used.
Guideline 5.48: Use Warning and Precaution Statements Judiciously If documentation includes many warnings and precautions, especially those addressing risks already known to the user, the warnings and precautions will lose their desired effect.
Guideline 5.49: Make Content Appropriate for Intended Users Choose the information content for warnings and precautions carefully and with consideration of the intended users’ backgrounds.
Guideline 5.50: Include Any Relevant Environmental Factors Describe the environmental factors that can interfere with device use and result in harmful effects. This includes the care needed to avoid damaging a device as a result of use or misuse.
Guideline 5.51: Locate Warnings and Precautions Appropriately Warnings and precautions that cannot be taken out of context without losing their meaning (are closely tied to procedural steps) should be located immediately before the associated step (FDA, 2001). Warnings and precautions are most effective when integrated into the relevant task information at an appropriate location. Medical device documentation has traditionally presented warnings and precautions in a separate section set aside for them. Given the reluctance of device manufacturers to break with tradition, it is recommended that the warnings and precautions placed in separate sections be those that can be taken out of procedural context without losing their meaning.
5.7.1.1.5 Supplementary Information Supplementary information consists of information that device users may find helpful over the course of using a device over time, such as manufacturer contact information, device technical information, and patient information. Patient information may include a means for recording test results (especially for home medical devices) and a description of the treatment or monitoring regimen that a patient is following. Guideline 5.52: Manufacturer Contact Information Make manufacturer contact information easy to find, such as by placing it at the beginning or end of the documentation. Include the device manufacturer’s name and address, customer assistance and technical support telephone numbers, and e-mail and Web addresses. Corresponding information can also be furnished for the device distributor.
Guideline 5.53: Device Information Provide specific and complete information about the device. List the complete brand name, model number, and date of manufacture of the device. If software is included, list the version of the software used by the device and its date of release.
TAF-ER562X_WINGR-09-0401-C005.indd 174
11/9/10 12:40:52 AM
Documentation
175
Guideline 5.54: Date of Printing Provide the date the documentation was issued or revised where it can be found easily (e.g., cover or last page). The FDA requires dated labeling for prescription devices and recommends it for all other devices (FDA, 2001).
Guideline 5.55: Record of Patient Test Results Some medical devices, especially those used in the home by laypersons, measure some physiological or biochemical process in a series of tests over an extended time period. In these cases, include a form for recording the date, time, and results of each test, even if the device’s internal memory maintains a record. It may become corrupted and the data lost.
Guideline 5.56: Frequently Asked Questions (FAQs) It may be helpful to include a list of frequently asked questions with answers to them. This can help new or infrequent users of a device to better understand its operation and maintenance.
5.7.1.1.6 Translations Given the requirement of European Union communities for medical device labeling to be provided in the national languages of the countries in which a medical device is marketed, translations take on added importance (see also Chapter 19, “Cross-National and CrossCultural Design of Medical Devices”). Guideline 5.57: Provide Translations if Users Have Other Primary Languages Translate the documentation into other languages for any anticipated user populations.
Guideline 5.58: Take Regional Differences into Account Make sure translations take into account any regional/idiomatic differences within the same language (e.g., intra–Latin American variability in Spanish usage).
Guideline 5.59: Be Aware of Conversational Idiosyncrasies A “straight” translation done without regard to conversational idiosyncrasies of language usage (e.g., figures of speech) can yield inaccurate and misleading results. The translation should be done by someone who is thoroughly familiar with the conversational use of the language(s) involved.
Guideline 5.60: Back Translate to Verify Accuracy It is advisable to perform a back translation of the translated documentation. Compare the results with the original text to detect and correct misinterpretations and unintended nuances of meaning. The back translation should be done by someone other than the original translator.
5.7.1.2 Presentation Guidelines In contrast to content guidelines, which focus on what information should be provided to device users, presentation guidelines specify how that information should be presented. The way in which information is conveyed contributes to how well people understand,
TAF-ER562X_WINGR-09-0401-C005.indd 175
11/9/10 12:40:52 AM
176
Handbook of Human Factors in Medical Device Design
remember, and comply with it. This section organizes presentation guidelines into six categories: • • • • • •
Language and readability Organization and layout Illustrations Highlighting Typography and legibility Physical characteristics
5.7.1.2.1 Language and Readability Language is the use of spoken or written words to communicate information. The correct, succinct, and clear use of words minimizes reading effort and makes documentation understandable to as many readers as possible. Readability assesses the extent to which printed text can be read easily. It depends on writing style (e.g., word choice, sentence length, complexity) rather than information content. Together, language and readability are important contributors to the usability of documentation. This section presents guidelines for making documentation easy to read and understand. Guideline 5.61: Be Concise Use short (less than two lines), direct, positive sentences or phrases that tell the user what to do. This is especially important for steps in procedures. However, do not sacrifice reader comprehension for the sake of brevity. Poor:
Be aware of the fact that dampness may cause rust and adversely affect the operation of the device. Better: Dampness can rust the device and it may not operate properly.
Guideline 5.62: Active Voice Use the active voice rather than the passive voice for telling someone what to do. Poor: Information is presented by the user manual on how to use the device. Better: The user manual presents information on device use.
Guideline 5.63: Use an “Inverted Pyramid” Writing Style Present the most important idea at the beginning of a passage, followed by any necessary details and supporting information. Assume that the reader could lose interest or become distracted at any time.
Guideline 5.64: Use Simple, Familiar, Everyday Words Minimize the use of long words and complicated expressions. Otherwise, users may become confused and be more error prone.
Guideline 5.65: Describe Device Operating Procedures Concisely Divide procedures into several short ordered steps instead of fewer longer paragraphs; this improves comprehension and reading speed.
TAF-ER562X_WINGR-09-0401-C005.indd 176
11/9/10 12:40:52 AM
Documentation
177
Guideline 5.66: Keep Readability at or Below the Sixth- to Seventh-Grade Level Documentation written at the sixth- to seventh-grade level will be understood by most device users. Try to use words with fewer than four syllables and limit sentences to fewer than 25 words (Backinger & Kingsley, 1993). Ask yourself this question: “How should this text be written so that a bright youngster between 10 and 12 years old would understand it?” (Wiklund, 1998). Although this guideline may be less important for devices used exclusively by highly trained physicians, for example, remember that under heavy workload, stress, and time pressure, cognitive capabilities can be appreciably impaired.
Guideline 5.67: Use Specific and Unambiguous Words or Phrases Avoid vague, awkward, and ambiguous words or phrases—they may cause users to misunderstand the instructions and make errors. Poor: Better: Poor: Better:
Respond quickly. Respond within one minute. Device operates unreliably in a cool room. Device will not operate reliably below 60ºF.
Guideline 5.68: Avoid Medical Jargon Minimize technical terminology and acronyms (except in cases where they are more recognizable than the spelled-out words). This is especially important for documentation intended for lay users. Poor: The device should not be used by ESRD patients. Better: The device should not be used by patients with end-stage renal disease.
Guideline 5.69: Minimize the Use of Abbreviations If using abbreviations, define them when they are first used in each section instead of just the very first time they are used in the documentation. If more than five abbreviations are used throughout the documentation, provide a table of all abbreviations.
Guideline 5.70: Use Terminology Consistently Use the same terms to refer to the same device features and actions throughout the documentation, even if synonyms seem appropriate. Terminology changes can confuse users, especially those unfamiliar with the given language or technology. For example, if you initially call a component a “dial,” do not call it a “knob” later.
5.7.1.2.2 Organization and Layout Organization refers to how topics are arranged in relation to each other in documentation. Organization establishes the logical flow of information necessary for understanding and remembering how to operate a device. Layout refers to how information is formatted on a page or screen. A well-designed layout makes it easier to locate and use information. Guideline 5.71: Provide a Table of Contents The headings in the table of contents should correspond word for word with the headings and subheadings in the text. The headings should clearly convey the content of each section of the
TAF-ER562X_WINGR-09-0401-C005.indd 177
11/9/10 12:40:52 AM
178
Handbook of Human Factors in Medical Device Design
documentation and show the flow of actions involved in device use. Page numbers are designated for each topic to make them easy to locate. Figure 5.3 shows a table of contents from a contact lens care prototype manual that incorporates these features.
Guideline 5.72: Establish a Logical Sequence of Information Organize topics according to the sequence of information and actions needed to operate a device safely and effectively. The exact composition and sequence of information depends on the type of device, but the following organization is a sound general approach: A) An Opening section should contain background information on the purpose and functioning of the device. Any significant hazards associated with device use should be cited here, although they should be reiterated in a later section on warnings. Contact information can also be included here. B) A Procedures section should present step-by-step instructions involved in using a device. Procedures should be presented in the same order the user would typically follow when operating the device. C) A section on Warnings should help users make informed decisions about potential health hazards associated with device use. D) A Troubleshooting section should tell users how to deal with problems in device operation. A tabular format can be helpful for organizing troubleshooting information so that it is easy to use. Use one column for signs of trouble and another for user actions.
Guideline 5.73: Use Clear Titles, Headings, and Subheadings Titles, headings, and subheadings establish the documentation’s organization. Use labels for headings and subheadings that clearly describe what information is in a given section. This will speed location of information, improve user memory for device operation, and reinforce the order in which procedures should be performed.
Table of Contents Lens Care Chart ........................................................................................ 2 Introduction .............................................................................................. 3 General Precautions ................................................................................. 4 Putting On Lenses .................................................................................... 9 Taking Off Lenses ..................................................................................... 17 Cleaning and Disinfection ............................................................................ 21 Cold (Chemical) Disinfection ................................................................ 26 Heat (Thermal) Disinfection .................................................................. 28 Emergency Heat Disinfection ................................................................ 30 Enzyme Cleaning ...................................................................................... 32 Lens Case Cleaning and Disinfection ................................................... 33 What To Do When You Have Problems While Wearing Lenses.......... 35
FIGURE 5.3 Example of a table of contents with descriptive headings to convey the organization and flow of topics.
TAF-ER562X_WINGR-09-0401-C005.indd 178
11/9/10 12:40:52 AM
Documentation
179
Guideline 5.74: Use Indentations to Show Information Relationships Indentations can more clearly show relationships among the various parts of a section, such as device operating procedures and the step-by-step instructions needed to perform those procedures.
Guideline 5.75: Make it Easy to Locate Information Users should be able to quickly and easily find the information they need. Carefully design those aspects of documentation that help users locate information, such as page numbers, page and section headings, section outlines, and indexes.
Guideline 5.76: Include an Index An index, in alphabetical order at the end of documentation, should include all important topics covered by the documentation and the page numbers on which they are discussed (Backinger & Kingsley, 1993).
Figures 5.4 and 5.5 show how poorly laid-out procedures can be improved through the use of numbered step-by-step instructions and closely linked line drawing illustrations. 5.7.1.2.3 Illustrations Properly chosen and placed illustrations can greatly facilitate understanding of how to operate a device. Illustrations are usually remembered better than words (Kobus, Moses, & Bloom, 1994). Effective illustrations clarify text while reducing the amount of information that must be communicated via text. They also reduce readers’ reliance on text, a big advantage for poor readers or users whose native language is different from the language used by the documentation. (a) Pick up the two electrodes (with the wires inserted) so the electrode with the white wire is in your left hand while that with the black wire is in your right hand. Now turn the electrodes so the Velcro side is facing down.
White end to the right side
(b) 1. Attach the electrodes (they will be under the armpits when the belt is closed).
2. Check to be sure that the white wire is on baby’s right.
FIGURE 5.4
Whitewire on baby’s right.
Comparison of (a) poor mapping and (b) clear mapping of text to illustration.
TAF-ER562X_WINGR-09-0401-C005.indd 179
11/9/10 12:40:52 AM
180
Handbook of Human Factors in Medical Device Design (a) Prick your finger with a sterile lancet and massage gently if necessary to obtain a large drop of blood. Quickly turn your finger upside-down so that the drop is hanging from the finger. With practice, you will be able to make the drop quite large without it actually falling off your finger. Do not proceed to the next step until your blood sample is ready. (b) Obtain a drop of blood as follows: 1. Prick your finger. To lessen the discomfort, prick the side of your finger, not the finger pad.
2. Hold your finger downward and squeeze it firmly to form a large hanging drop of blood.
FIGURE 5.5
Original (a) versus improved (b) end-user instructions for medical device use.
Guideline 5.77: Make Illustrations Simple and Clear Illustrations should be clear, simple, and uncluttered. Represent only simple concepts, such as required user actions, features of the device, or aspects of its surroundings. Use a separate illustration for each distinct concept (FDA, 2001).
Guideline 5.78: Confine Action Illustrations to a Single Action Focus on a single action at a time instead of trying to depict multiple actions in the same illustration.
Guideline 5.79: Make Comparisons Explicit If two illustrations are compared, make the difference between them clear. Use a visual means, such as circles or arrows, to point out the difference (FDA, 2001).
Guideline 5.80: Make Illustrations Large Enough for Easy Viewing Illustrations should be large enough for users to see the key details and important words easily.
Guideline 5.81: Illustrations Should Clarify Text Use illustrations to clarify and reinforce the accompanying text. This will help readers to better understand the information conveyed by the documentation. Illustrations can also show aspects of device operation that are hard to express verbally.
Guideline 5.82: Link Illustrations to Text An illustration should always be accompanied by clear explanatory text. Set off text and its accompanying illustration by the use of lines, white space, or titles (FDA, 2001). Figure 5.6 shows how call-outs can be used to supplement an illustration of a device interface. The individual items called out in an illustration can be numbered so that they correspond with a numbered list in the accompanying text that describes each item.
TAF-ER562X_WINGR-09-0401-C005.indd 180
11/9/10 12:40:52 AM
Documentation
181
(12) Delivery rate arrow selectors
88
ml/h
88 ml
ml/h
Dose Delivery
Volume Delivery
START
(10) Start button
Dose Limit
Delivery Rate
88 (11) Volume delivered message display
(3) Dose limit display
(2) Center display
(1) Delivery rate display
88 ml
CLEAR
RESET
(5) Dose delivery display
LIMIT ON/OFF CLEAR
SILENCE
(9) Reset (8) Clear (7) Silence button button button
(4) Dose limit arrow selectors
(6) Limit on/off/clear button
FIGURE 5.6 Example of the use of call-outs to illustrate the features of a medical device (in this case, an infusion pump). Each call-out is keyed to a same-numbered description in the text.
Several types of illustrations can be used in device documentation; each type is best suited for a particular purpose. The alternatives are discussed in the following guidelines. Guideline 5.83: Eliminate Unnecessary Details To maximize comprehension, illustrations should focus on key user interface elements and omit unnecessary details.
Guideline 5.84: Use Line Drawings for Details Use line drawings to convey detailed aspects of device operation and device components. Line drawings emphasize specific details and object dimensions (Bailey, 1989). Line drawings should have dark, sharp lines for good contrast. Procedural steps are often more effectively illustrated with line drawings or cartoons rather than with photographs.
Guideline 5.85: Use Photographs for Realism Photographs convey the exact appearance of objects and show them in three dimensions. They are best for realistically depicting actual device operating conditions. However, photographs may have poor contrast and include unwanted details that can distract the user’s attention from critical aspects of device operation.
Guideline 5.86: Limit the Use of Exploded Views Use exploded views only for devices that the user should put together or take apart (FDA, 2001). Exploded views are more difficult to place in context. In addition, some users may be tempted to dismantle and/or tamper with a device if such a view is provided in the user manual, for example.
TAF-ER562X_WINGR-09-0401-C005.indd 181
11/9/10 12:40:53 AM
182
Handbook of Human Factors in Medical Device Design
Guideline 5.87: Use Tables for Comparisons Tables present text or numerical comparisons across more than one attribute or category. Include instructions on a table’s use and clearly label its sections (FDA, 2001). In general, quick reference guides do not require tables.
Guideline 5.88: Use Graphs for Trends Graphs show trends and how one factor varies with one or more other factors. Graphs might, for example, be used to depict changes in calibrating settings or operating conditions at different environments, temperatures, or altitudes.
Guideline 5.89: Color versus Black-and-White Illustrations Carefully evaluate the importance of color in understanding device operation when choosing between black-and-white and color illustrations. Color illustrations attract and hold readers’ attention better than black-and-white illustrations because of their lifelike character and greater conspicuity (Marcus, 1992). In addition, research has found that users have a strong preference for color illustrations (Callan et al., 1993). This preference results in more attention being given to instructions containing color illustrations. On the other hand, there is no evidence that color illustrations increase compliance with device operating procedures, although color used with other highlighting methods has been shown to improve memory for task features (Young & Wogalter, 1990).
5.7.1.2.4 Highlighting Highlighting provides visual relief, stresses important points, and sets off sections and subsections of text. The guidelines in this section can be used to bring important words and phrases to the reader’s attention. Guideline 5.90: Use a Small Number of Highlighting Techniques In most cases, two or three carefully selected highlighting techniques are better than a larger number. Readers may become confused or dismissive if presented with many forms of highlighting.
Guideline 5.91: Use Highlighting Consistently Use the same technique for the same type of information throughout the documentation.
Guideline 5.92: Use Highlighting Judiciously The effectiveness of highlighting will be reduced if it is overused. When overused, highlighting can distract users, especially novice, older, or infrequent users.
Guideline 5.93: White Space White space (i.e., blank space) is a simple yet effective highlighting technique. It can be used between sections or to set off important information. White space also improves the appearance of documentation and makes it easier to read. Figure 5.7 shows examples of white space that address the following points: A) Do not use excessive white space between lines of text. It impairs reading speed, comprehension, and legibility. It also lengthens documentation unnecessarily and increases printing costs.
TAF-ER562X_WINGR-09-0401-C005.indd 182
11/9/10 12:40:53 AM
Documentation
183
Too much white space between lines impairs reading speed, comprehension, and legibility. Excessive white space also increases printing costs and makes a set of instructions unnecessarily lengthy— undesirable consequences for both documentation developer and user.
On the other hand, documentation with insufficient white space often looks cramped and is hard to read. The lines will blur together for many readers, especially those with poor vision. This will quickly lead to eye fatigue and a tendency on the reader’s part to abandon the documentation. White space can also be used to delimit different sections in a set of instructions. It can draw the reader’s attention to parts of labeling the author wishes to emphasize. For example, white space can lead the reader to specific information that the author wishes to emphasize, like this. The eye is drawn to information that the author wants the reader to be sure to notice. Other types of highlighting techniques can be used in conjunction with white space to enhance its effect.
FIGURE 5.7
Examples of the use of white space.
B) On the other hand, if the amount of space between lines is too small, the lines will blur together for many readers, especially those with poor vision. Documentation with insufficient white space often looks cramped and is hard to read. C) White space also can be used to delimit different sections in a set of instructions. It can draw the reader’s attention to those parts of the documentation the author wishes to emphasize.
Guideline 5.94: Bolding Bolding is another simple yet effective highlighting technique. It can be used in conjunction with white space to emphasize critical information.
Guideline 5.95: Limit the Number of Colors The background and major page or screen components should be limited to three to five colors, including shades of gray. Be mindful that medical conventions are associated with some colors (e.g., red for alarms or arterial blood pressure values) (Wiklund, 1998).
Guideline 5.96: Provide Redundant Coding for Color Always use color highlighting with another highlighting technique (e.g., bolding). Otherwise, users with a color vision impairment may not be aware that some text is more important than other text (Nielsen, 2000).
Given the ease of manipulating documents using word processors, numerous other highlighting techniques are available. Carefully evaluate how your documentation will be used when deciding which, if any, of these other highlighting techniques to use.
TAF-ER562X_WINGR-09-0401-C005.indd 183
11/9/10 12:40:53 AM
184
Handbook of Human Factors in Medical Device Design
5.7.1.2.5 Typography and Legibility Typography is the arrangement, style, and general appearance of material printed from type. Typography contributes to a consistent “look and feel” for a document, which helps users gain familiarity with it and understand how it is organized. Three important factors for typographical variables are font size, font type, and line characteristics. Also discussed are color and contrast. 5.7.1.2.5.1 Font Size applications.
Figure 5.8 shows several different font sizes and their recommended
Guideline 5.97: Recommended Font Size for All-Around Use Use a large enough font size for text to be legible to the intended user audience under a wide range of use conditions. Twelve-point is an excellent compromise between legibility and the need to conserve space.
Guideline 5.98: Minimum Acceptable Font Size Ten-point font is an acceptable minimum size for general audiences. However, it is too small for older users or others with visual impairments, who will find small fonts hard, if not impossible, to read.
Guideline 5.99: Avoid 9-Point and Smaller Font Sizes Do not use 9-point and smaller fonts. Text in small fonts is read more slowly than text in larger fonts. Many readers will tend to skip over it or develop eyestrain.
Guideline 5.100: Larger Fonts have their Place Many home medical device users are older people who need larger fonts to read the documentation: • Fourteen-point font is good for readers with visual impairments and older users. Do not use 9-point and smaller font. It is too small to be legible to most readers.
10-point font is an acceptable minimum size for general audiences but not for older users and persons with visual impairments.
12-point font is an excellent compromise between the need to conserve space and to present legible instructions.
14-point font is good for visually impaired readers and older users. It is also useful for highlighting important information.
18-point font should be used sparingly, if at all. Reserve it for critical items you want to be sure the reader notices. FIGURE 5.8
Examples of different font sizes and their recommended applications.
TAF-ER562X_WINGR-09-0401-C005.indd 184
11/9/10 12:40:53 AM
Documentation
185
• Use 18-point font sparingly and only for critical items you want to be sure the reader notices. If used to any great extent, 18-point font would make the documentation very long.
5.7.1.2.5.2 Font Type Guideline 5.101: Serif versus Sans Serif Fonts For body text, the choice between a serif font (e.g., Times New Roman) and a sans serif font (e.g., Arial) is largely a matter of personal taste. Limited differences have been found between serif and sans serif fonts in terms of legibility or reader performance.
Guideline 5.102: Use Sans Serif Fonts for Headings Sans serif fonts are preferred for headings due to their “cleaner” appearance.
Guideline 5.103: Do Not Mix Serif and Sans Serif Fonts When used within the same text passage, a mixture of serif and sans serif fonts can slow reading speed (Boyarski et al., 1998).
Guideline 5.104: Use All Capitals and Italics Judiciously Use of text in all capital letters or all italics reduces legibility and slows reading speed by as much as 15% to 20% (Tinker, 1963). Text printed in all capital letters also takes up more space. All-capital letter phrases can be used sparingly for headings or phrases (e.g., warning signal words) as a means of emphasis.
Guideline 5.105: Minimize the Use of Multiple Font Types Frequently switching from one font type to another will distract the reader and reduce reading speed.
Guideline 5.106: Be Consistent Pick a small number of commonly used fonts and use them consistently throughout your documentation. It is common practice to use one font type for body text and another for titles, headings, and subheadings.
5.7.1.2.5.3 Line Characteristics Guideline 5.107: Choose a Line Length for Reading Ease The best line length for documentation printed in 12-point font is 4.0 ± 1.25 inches. Longer lines may strain the eye as it scans across their entire length, making it easier to jump to the wrong next line. This is an especially crucial consideration for medical devices, where the steps of each operating procedure must be performed in their correct sequence.
Guideline 5.108: Avoid Very Short Line Lengths Do not use a line length less than 2.5 inches. It slows reading speed because of the large number of back-and-forth eye movements required while reading even a single sentence.
Guideline 5.109: Use Ragged-Right Margins A ragged-right margin makes it easier to move from one line to the next than does fully justified, center-justified or right-justified text. Readers can keep track of their place better because the ragged-right profile helps distinguish one line from another. Center-justified lines produce
TAF-ER562X_WINGR-09-0401-C005.indd 185
11/9/10 12:40:54 AM
186
Handbook of Human Factors in Medical Device Design
a “riverbank” effect that makes it harder to locate the next line. Fully justified text requires the eye to adjust to variable spacing between words. In addition, fully justified text typically includes hyphenated words, which further slows reading speed.
Guideline 5.110: Minimize Hyphenation Minimize the use of hyphenation, especially with short words. Hyphenation requires the reader to remember the last syllable on the previous line. Persons with limited vision or poor memory often find this to be difficult.
5.7.1.2.5.4 Color and Contrast Guideline 5.111: Maintain High Print-to-Background Contrast Black text on a white background is a universal standard for print contrast. It provides the best contrast and is easiest to read. However, other color combinations are acceptable as long as they maintain a high contrast between the printed material and the background. Some acceptable alternative color combinations are black on yellow, dark blue on white, and dark green on white (Backinger & Kingsley, 1993; see also Chapter 8, “Visual Displays”).
Guideline 5.112: Be Careful with Color Combinations Avoid color combinations with poor contrast, such as red text on a black background or lightcolored text on a light background, which may make it hard to read the printed material.
Guideline 5.113: Avoid Highly Saturated Colors Highly saturated colors are difficult to look at for extended periods without developing eye fatigue.
Guideline 5.114: Avoid Light Blue and Pastel Colors Blue is hard for many older people to see. Eight percent of males cannot distinguish pastel colors (Backinger & Kingsley, 1993).
5.7.1.2.6 Physical Characteristics The guidelines in this section apply mostly to paper documentation. Physical characteristics of documentation affect its ease of use and subjective appeal and contribute to how well its content can be understood, followed, and remembered. Desirable physical characteristics for documentation stem from two factors: the documentation’s environment of use and its updating requirements. Guideline 5.115: Use Heavy Dull-Finish Coated Paper Paper with a dull finish is better than glossy paper, which can direct distracting reflections toward the eye. Paper should be heavy enough to prevent show-through and to resist tearing with repeated use. It should also be coated to make it resistant to spillage and soiling if the document is likely to be used under conditions in which this is probable.
Guideline 5.116: Choose between Landscape and Portrait Orientation Choose the orientation that best allows interrelated text, illustrations, and warnings to be displayed next to each other.
TAF-ER562X_WINGR-09-0401-C005.indd 186
11/9/10 12:40:54 AM
Documentation
187
Guideline 5.117: Choose Appropriate Document Size User manuals and quick reference guides must be designed for ease of access and use during device operation, which necessitates a smaller format. Other documents, such as technical manuals, are more likely to be referred to when a device is not being operated. These documents are often book length and should be sized accordingly.
Guideline 5.118: Make the Document Easy to Update Updating a paper document involves adding or deleting pages. Ring binding is ideal for meeting this requirement. Glued or spiral bindings are preferable for documents that will not be modified (Simpson & Casey, 1988).
Guideline 5.119: Ensure the Document Will Lay Flat The document should lay flat without assistance when open so that users can have both hands free to operate the device.
Guideline 5.120: Make the Document Durable The document should be sufficiently durable to stand up to repeated use over the expected lifetime of the device.
Guideline 5.121: Make the Document Easy to Clean The document should be easy to clean if liquids (e.g., blood, saliva, saline) are accidentally spilled on it. Lamination can preserve the document if it is used frequently or in a wet environment (Backinger & Kingsley, 1993).
5.7.2 GUIDELINES FOR ELECTRONIC DOCUMENTATION In addition to the general design guidelines presented in the preceding section, to create the most effective devices, document designers should consider features unique to electronic documentation. The design guidelines in this section address those features. 5.7.2.1 Navigation Compared with paper documentation, electronic documentation gives users more ways to navigate to information of interest. In paper documentation, organization aids, such as a table of contents, tabbed pages, or an index, are used to locate information. In addition to these features, electronic documentation uses other navigation aids, including navigation bars, search functions, and hyperlinks. With ineffective organization and navigation, users can become lost or disoriented as they move from one part of the documentation to another. The following guidelines will help designers create effective navigation schemes for electronic documentation. Guideline 5.122: Make Documentation Content Explicit The key to effective navigation is the logical and explicit organization of documentation content. Users should always know where they are, where they came from, how to return to previous screens or pages, and where they can go next (Koyani & Nall, 1999).
Guideline 5.123: Use the Same Navigation Aids on All Pages A consistent navigational “look and feel” will make it easier for users to move around the documentation (Koyani & Nall, 1999).
TAF-ER562X_WINGR-09-0401-C005.indd 187
11/9/10 12:40:54 AM
188
Handbook of Human Factors in Medical Device Design
Guideline 5.124: Place Meaningful Titles on Every Screen Titles help users remain oriented and be less likely to get lost (Wiklund, 1998).
Guideline 5.125: Multiple Ways to Access Information Navigation bars, a key word search function, hyperlinks, and breadcrumb trails are ways to navigate electronic documentation. A site map or menu index that makes the documentation’s content explicit also provides ways to access information (Spool, Scanlon, Schroeder, Snyder, & DeAngelo, 1997).
Guideline 5.126: Use Consistent Placement of Navigation Aids and Controls Commands such as “Exit,” “OK,” “Previous,” “Next,” and “Help” should be placed in the same location on each screen or page.
Guideline 5.127: Allow Direct Access to the Home Page or Screen Give users an easy way to return to the home page or screen. A “Go to Main Menu” or similarly labeled button or icon that takes users to the home page or screen from anywhere in the documentation accomplishes this goal.
Guideline 5.128: Provide a Key Word Search Function Key word search allows users to quickly locate information about a specific topic or term.
5.7.2.2 Hyperlinks Hyperlinks provide direct access between a given topic and related information. Hyperlinks can appear in many places in electronic documentation, including tables of contents, indexes, site maps, and header and body text. Guideline 5.129: Clearly Indicate Hyperlinks Always use underlines or some other visual indicator (e.g., a stacked list of items) to indicate text links. Using mouse-overs to designate links can confuse users and slow them down (Bailey, Koyani, & Nall, 2000).
Guideline 5.130: Reserve Blue-Colored Text for Hyperlinks Blue-colored text should not be used in hyperlinked or Web documentation unless it is a clickable hyperlink. On the Web, blue text indicates clickable text. For the same reason, try to avoid making unclickable text red or purple because these colors often denote visited links. Use blue underlined text for unvisited links and purple or red underlined text for visited links (these are standard browser-default link colors) (Nielsen, 1996a, 1996b).
Guideline 5.131: Differentiate Link Labels Make link labels different from each other to reduce navigation errors (Spool et al., 1997).
Guideline 5.132: Position Links Prominently on the Screen Place important links (and other important information) high enough on the page that they are visible without scrolling (Bailey et al., 2000).
TAF-ER562X_WINGR-09-0401-C005.indd 188
11/9/10 12:40:54 AM
Documentation
189
Guideline 5.133: Text Links Preferable to Graphic Links A text link’s descriptive title increases its predictability and facilitates locating information better than do graphic links, which are generally more difficult to interpret (Koyani & Nall, 1999). Text links change color after being visited, and graphic links do not, making it harder for users to keep track of what portions of the documentation they have visited when graphic links are used (Spool et al., 1997).
Guideline 5.134: Selectively Use Graphic Links Some international symbols, such as the crossbar through a circle meaning “no,” can be more immediately recognizable to a global audience than plain text. In such cases, the use of graphic links may be justified, especially when users are not fluent in the language used by the documentation.
Guideline 5.135: Avoid Long Link Labels Use short labels for hyperlinks. A long link label may become wrapped (appear on more than one line) and appear to be two separate links, resulting in user confusion (Spool et al., 1997).
Guideline 5.136: Use Breadcrumb Trails in Web-Based Documentation It can be surprisingly easy for users, especially novices, to get lost when navigating electronic documentation. One way to guard against this tendency is always to show users their location via a specific type of hyperlink called a breadcrumb trail: A) Breadcrumb trails show the path the user has followed to get to a given location in the documentation and allow users to move several levels in a hierarchy at a time. Put breadcrumb trails near the top of a Web page (Nielsen, 1999). B) Users often leave and then return to a Web page, viewing other Web pages in the interim. Breadcrumb trails help users reorient to a Web page when they return to it (Nielsen, 2007).
5.7.2.3 Language and Readability People typically find it harder to read electronic documentation than paper documentation. Their eyes tend to fatigue more easily, and reading speed slows by as much as 25%. Therefore, it is not surprising that most people scan electronic documentation rather than read it thoroughly (Morkes & Nielsen, 1997). Guideline 5.137: Facilitate Scanning Use phrases and short sentences to convey information. Avoid complex sentence construction.
Guideline 5.138: Reduce Word Count Limit the amount of text to about half the word count (or less) compared to paper documentation.
5.7.2.4 Organization and Layout Guideline 5.139: Use a Shallow Information Hierarchy Design electronic documentation so that its information is organized into three or fewer layers. Put as much important information in or near the top layer as possible. This will give users faster and easier access to it (Czerwinski, Larson, & Robbins, 1998).
TAF-ER562X_WINGR-09-0401-C005.indd 189
11/9/10 12:40:54 AM
190
Handbook of Human Factors in Medical Device Design
Guideline 5.140: Limit the Number of Topics per Information Layer To avoid overloading a user’s memory, each layer should have no more than seven topics. If there are more than seven topics, users will tend to lose focus. Fewer than seven topics likely will be adequate for all but the most complex devices. For example, documentation that is three layers deep with seven options per layer contains 343 topics (7 × 7 × 7 = 343), which should be enough for the large majority of medical devices. Documentation that is three layers deep with four options per layer contains 64 topics (4 × 4 × 4 = 64), which should be adequate for many medical devices.
Guideline 5.141: Keep Screen Layout Simple A simple screen layout will help users find information and perform their tasks efficiently. Use easily understood buttons, hyperlink labels, and other easy-to-use navigation aids (Lynch & Horton, 1999).
Guideline 5.142: Avoid Horizontal Scrolling Format page contents so that horizontal scrolling is not necessary. Horizontal scrolling makes it harder to understand a page because not all of its content can be viewed at once. Horizontal scrolling also annoys users.
5.7.2.5 Illustrations Guideline 5.143: Take Display Size Limitations into Account The size of the display used to present documentation may restrict the amount of detail that can be included in illustrations. Simplified line drawings that emphasize the important aspects of device use will be clearer than photographs in these cases.
5.7.2.6 Web-Specific Graphics The following guidelines apply to graphics presented via the Web. Guideline 5.144: A Few Simple Graphics Will Reduce Download Times Use only graphics that enhance content or lead to a better understanding of the information being presented. Users will only be willing to wait for graphics to download when the graphics add value for them (Nielsen, 1999; Spool et al., 1997).
Guideline 5.145: Use a Base Resolution of 800 × 600 If graphics look good at 800 × 600 pixel resolution, they will also look good at higher resolutions (Venkatacharya, 2000).
Guideline 5.146: Design Graphics Specifically for the Web Use the 216 Web-safe colors. Many Web users still use equipment that displays only 256 colors. All Web-safe colors will be visible on all 256-color systems. Using Web-safe colors eliminates many cross-browser inconsistencies in how colors appear on one browser or platform versus another.
Guideline 5.147: Respect the Four-GIF Limit Graphics Interchange Format (GIF) is an image compression technique commonly used for Web graphics. Most Web servers call up four GIFs at a time, so it is a good idea to limit GIFs to fewer than four per page. By avoiding multiple server calls for GIFs, pages will load much faster.
TAF-ER562X_WINGR-09-0401-C005.indd 190
11/9/10 12:40:54 AM
Documentation
191
Guideline 5.148: For Smaller Images, Do Not Create Blended, Indistinct Edges or Fades Keep everything in crisp colors. Provide a strong contrast between the image color and the background. Do not try aliasing (images with jagged edges) small images or fonts. Using crisp colors reduces file size and helps images look clean and bold.
Guideline 5.149: Use Aliased Images for Graphics with Text Aliased images containing text will download faster than antialiased images (images with smoothed-out edges). Aliased images have a smaller file size. Also, many sans-serif fonts are easier to read when they are aliased.
5.7.2.7 Highlighting The potential number of highlighting methods is greater for electronic documentation than for paper documentation. Most highlighting methods for paper documentation can also be used for electronic documentation. Highlighting methods unique to electronic documentation include flash coding, animation, and sound. As with paper documentation, only a small number of highlighting methods should be used in any given document. 5.7.2.7.1 Flash Coding Guideline 5.150: Use Flash (Blinking) Coding Sparingly Use flash coding only to display urgent, high-priority information.
Guideline 5.151: Do Not Flash Text Flash an icon or border associated with the text or display a “focus” area.
Guideline 5.152: Give Users Control over Flash Coding Users should be able to acknowledge the event causing the flashing and suppress it if desired (Fernandes, 1993).
Guideline 5.153: Flash Frequencies and Epilepsy Try to avoid flash frequencies in the 10- to 50-Hz range (especially 20 Hz) to minimize the possibility of inducing seizures in persons with photosensitive epilepsy. Sensitivity to this effect increases with the intensity of the light and the portion of the person’s visual field that is affected. A flashing screen or larger screen object is worse than a smaller screen object, such as a cursor. Focusing attention on a flashing object, which is the goal of flash coding, also increases the risk (http://trace.wisc.edu/docs/consumer_product_guidelines/consumer. htm; Jeavons & Harding, 1975).
5.7.2.7.2 Animation Guideline 5.154: Use Animation for Event Sequences Animation is especially effective in depicting a sequence of actions, such as the steps involved in device operation (see Chapter 11, “Software User Interfaces”).
Guideline 5.155: Animation Emphasizing Key TaskRelated Information Animation should support task performance, such as drawing attention to critical aspects of device use or to newly received information (Spool et al., 1997).
TAF-ER562X_WINGR-09-0401-C005.indd 191
11/9/10 12:40:54 AM
192
Handbook of Human Factors in Medical Device Design
Guideline 5.156: Use Animation to Enhance Explanations Animation is particularly good for explaining ideas involving changes in time, position, and/ or process. It can convey dynamic aspects of device operation.
Guideline 5.157: Positioning Animation Do not place animation too close to nonmoving text or it may draw users’ attention away from the text.
Guideline 5.158: Animating Text Control the timing of animated text so that it remains readable.
5.7.2.7.3 Sound Guideline 5.159: Effective Use of Sound Sound can be effective if it is carefully applied; otherwise, it will be only a distraction. Sound should be used to do the following: A) Provide users with information, such as the actual voice messages or tones generated by a medical device B) Reinforce the presentation of written or graphical information See also Chapter 10, “Alarm Design,” for more information.
Guideline 5.160: Integrate Sound with Text and Illustrations Use sound to help users understand how to operate a device, not just to get their attention. For example, providing the same feedback tones that a device makes during its operation is an effective use of sound, whereas playing background music is not.
Guideline 5.161: Provide Redundant Coding for Sound Always supplement sound with a visual representation of the information it conveys. Some users may have hearing impairments, use their devices in a noisy environment, or use them in a setting that requires sound to be set at a low volume. See Chapter 3, “Environment of Use.”
5.7.2.8 Typography Guideline 5.162: Text on Computer Monitors For text that appears on computer monitors, keep the following points in mind (Rubenstein, 1988): • Text appearance will vary with resolution and monitor size. For example, 10-point font will look smaller on a 1,024 × 768 resolution monitor than it will on an 800 × 600 resolution monitor. Users can adjust text appearance by self-selecting resolutions and monitors. • Point size is not comparable across font styles (see Section 5.7.1.2.5.1). Different styles may have very different size letters at the same point size. Test view the actual fonts in their intended point sizes to determine if they are legible.
Guideline 5.163: Use Web-Specific Fonts for Online Documentation If documentation is presented via the Web, use fonts that have been developed specifically for use on Web pages, such as Georgia (serif) and Verdana (sans serif).
TAF-ER562X_WINGR-09-0401-C005.indd 192
11/9/10 12:40:54 AM
Documentation
193
Guideline 5.164: Avoid Background Images, Textures, and Patterns Nonneutral backgrounds can make it harder to read the overlying text. If a background image must be used, blur it and make it a light shade or largely transparent so that the overlying text stands out (Spool et al., 1997).
5.7.2.9 Physical Characteristics Most of the physical characteristics of electronic documentation relate to the visual and auditory features that have already been addressed. However, users may, on occasion, desire to print out all or a portion of an electronic document. Guideline 5.165: Create Two Versions of Online Documentation Optimize one version for online viewing using a screen-oriented style sheet. Chunk it into several separate files joined by hyperlinks. The second version should contain the entire document in a single file in a format suitable for printing, such as PDF. This printer-friendly version should preserve the pagination, formatting, and typography found in the electronic version to facilitate easy cross-referencing from one version to the other. It should be printable on standard letter-sized paper (8.5 × 11.0 inches) and not use specialized fonts or symbols that may not be available on all printers. If the document is intended for international distribution, make it compatible with A4 paper size (8.25 × 11.7 inches) (Nielsen, 2000).
5.8 CASE STUDIES The following case studies demonstrate how to apply the human factors principles and guidelines in this chapter to create effective device documentation.
5.8.1 CONTACT LENS CARE A human factors evaluation of manufacturers’ documentation for cleaning soft contact lenses was performed because of an unexpectedly high incidence of eye health problems among soft contact lens users. This evaluation identified numerous deficiencies in both informational content and presentation format that contributed to poor user compliance. This noncompliance was responsible, at least in part, for microbial infection, allergic reactions, and corneal trauma. Furthermore, a human factors usability study found that experienced lens wearers made substantial numbers of errors when using manufacturers’ documentation to guide their lens care. A task analysis was conducted to obtain a detailed description of the actions necessary for compliant lens care. The tasks evaluated included procedures for hand washing, lens removal, lens cleaning, lens disinfection, and reinsertion of the lens or placement in a lens case. The task analysis also identified the most likely errors associated with each of these actions. This information was then combined with human factors documentation design guidelines to improve lens care documentation with the following features: • Lengthy procedures were decomposed into a series of simple individual steps. Each step features simple, direct procedural text, an accompanying illustration, and rationales explaining the reason for performing various steps. Figure 5.9 shows a representative page from the prototype documentation. • Text was written at the grade 5 level and printed in a clear, legible font. • Page layout and highlighting techniques presented the information in a clear and simple way.
TAF-ER562X_WINGR-09-0401-C005.indd 193
11/9/10 12:40:54 AM
194
Handbook of Human Factors in Medical Device Design
Putting on lenses
Step 1 – Wash, rinse, and dry hands. A. Wash your hands with soap that does not have any oils or perfumes. This will keep residues from getting on your lenses. B. Rinse your hands thoroughly after washing. C. Dry your hands with a lint-free towel. This will help keep lint and dust from getting on your lenses and irritating your eyes.
Always wash your hands before handling your lenses. This will remove dirt and oils that could get on the lenses. Hand washing also helps prevent eye infections.
FIGURE 5.9 Sample lens care booklet page showing the step-by-step instructions for a contact lens care procedure.
• Warnings and precautions were linked to the applicable procedural steps. • Detailed line drawings illustrated the most critical steps. • A quick reference guide was also provided that summarized lens care procedures in an easy-to-follow format (see Figure 5.10). Using this new documentation, both novice and experienced contact lens wearers made significantly fewer lens care errors and reported a greater understanding of why they should perform various actions as part of their lens care regimen.
Daily lens care 1. Clean lenses
Weekly lens care
2. Disinfect lenses As recommended for your lenses, use either Cold or Heat disinfection.
Be sure to perform both of the following procedures once a week.
a. Wash hands. b. Place lens in palm.
Cold (chemical) disinfection
c. Rub lens – each side for 20 seconds.
a. Fill lens case with disinfecting solution.
a. Fill lens case with saline solution.
d. Rinse lens with saline solution.
b. Soak lenses for 4 hours.
b. Place case in disinfection unit.
e. Place lens in lens case.
c. Rinse lens before wearing.
c. Start disinfection.
f. Go to Step (2).
FIGURE 5.10 lens care.
Heat (thermal) disinfection
d. Rinse lens before wearing.
Clean/disinfect lens case
Enzyme clean lenses
a. Boil a pan of tap water.
a. Follow enzyme cleaner instructions.
b. Boil empty case & caps for at least 10 minutes. c. Air dry case & caps.
b. Clean and disinfect lenses as for daily lens care.
Quick reference guide summarizing the major procedures and steps in soft contact
TAF-ER562X_WINGR-09-0401-C005.indd 194
11/9/10 12:40:54 AM
Documentation
195
5.8.2 BLOOD GLUCOSE METER OPERATION Portable blood glucose meters have become a major tool for self-care by persons with diabetes. Used properly, these devices provide accurate blood glucose measurements. The results are used as a basis for various interventions, such as injecting insulin when a high glucose reading is obtained. Accurate readings are therefore crucial to the well-being of meter users. Providing users with clear and easy-to-follow documentation for meter use and maintenance is critical to obtaining accurate readings. The order in which topics are presented in documentation is a major determinant of its usability. The table of contents for a prototype blood glucose user manual, shown in Figure 5.11, shows that its topics follow a logical flow from general considerations about device use to the specific procedures for operating and maintaining the device. A section on warnings described hazards involved in device use, along with advice on how to minimize them. A troubleshooting section was provided to help users identify and remedy problems encountered during device use. Closing sections contained supporting information and an index for locating specific topics. The text that described the operating and maintenance procedures for the device consists of a numbered sequence of steps, each of which was labeled with a short title, an excerpt of which is shown in Figure 5.12. Each procedural step is described by one or more short sentences or phrases written in a positive, active voice. A line drawing that depicts the actions described by the text accompanies each step. The illustrations enable poor readers or nonnative English speakers an alternative way to understand device operation. Warnings, when applicable, are placed next to the relevant step but set off from the text and illustrations to increase their salience.
TABLE OF CONTENTS 1. Introduction.............................................................................................. 1 Advantages of Your Blood Glucose Meter........................................... 1 Preparing to Use Your New Meter........................................................ 2 The Meter Kit............................................................................................. 3 Features of Your Blood Glucose Meter................................................ 6 Display Messages....................................................................................... 7 2. Using Your Meter.................................................................................. 10 Installing the Batteries........................................................................... 10 Calibrating the Meter............................................................................. 12 Doing a Blood Glucose Test................................................................. 15 3. Maintaining Your Meter....................................................................... 19 Replacing the Buffer Solution Reservoir............................................. 19 Replacing the Test Cap.......................................................................... 22 4. Warnings................................................................................................ 25 5. Troubleshooting Guide......................................................................... 26 Troubleshooting Display Messages..................................................... 27 Troubleshooting High or Low Test Results....................................... 30 6. Specific Performance Characteristics.................................................. 33 7. Technical Information.......................................................................... 34 8. Device Warranty.................................................................................... 35 9. Index....................................................................................................... 36
FIGURE 5.11 Table of contents showing the logical organization of information in a user manual for a portable blood glucose meter.
TAF-ER562X_WINGR-09-0401-C005.indd 195
11/9/10 12:40:55 AM
196
Handbook of Human Factors in Medical Device Design
2. Slide calibration bar to the right Slide the calibration bar to the right until it locks into place. The meter is now ready to be calibrated.
3. Apply calibration solution to test cap. Place one drop of calibration solution inside the white ring on the test cap. “CALIBRT” is displayed for 15 seconds.
4. Wait 15 seconds, then read display. After 15 seconds, the unadjusted reading from the 200 mg/dL calibration solution is displayed for 3 seconds. “CAL OK” is displayed when the automatic calibration is completed.
FIGURE 5.12 Sample page from a prototype blood glucose meter user manual describing the steps involved in the device’s calibration procedure.
RESOURCES The guidelines and principles presented in this chapter are consistent with two FDA medical device guidance documents—Backinger & Kingsley (August 1993) and Food and Drug Administration (April 2001)—both cited below. Other useful sources include: American National Standards Institute/Association for the Advancement of Medical Instrumentation (ANSI/AAMI). (2010). Human Factors Engineering Guidelines and Preferred Practices for the Design of Medical Devices. ANSI/AAMI HE-75-2010. Arlington, VA: Association for the Advancement of Medical Instrumentation. American National Standards Institute/Association for the Advancement of Medical Instrumentation (ANSI/AAMI). (2001). Human Factors Design Process for Medical Devices. ANSI/AAMI HE-74-2001. Arlington, VA: Association for the Advancement of Medical Instrumentation. Callan, J. R., Gwynne, J. W., Sawyer, D., and Tolbert, M. T. (1993, December). Principles of Medical Device Labeling (NTIS PB 94-126851). San Diego, CA: Pacific Science & Engineering Group. Retrieved June 29, 2009, from http://www.fda.gov/downloads/MedicalDevices/ DeviceRegulationandGuidance/Guidance Documents/UCM095300.pdf. Inaba, K., Parsons, S. O., and Smillie, R. (2004). Guidelines for Developing Instructions. Boca Raton, FL: CRC Press.
For additional information on the role of human factors in medical device design and development, see the following: Sawyer, D. (1996). Do It by Design: An Introduction to Human Factors in Medical Devices. Washington, DC: U.S. Department of Health and Human Services, Food and Drug Administration.
TAF-ER562X_WINGR-09-0401-C005.indd 196
11/9/10 12:40:55 AM
Documentation
197
REFERENCES American National Standards Institute. (2002a). American National Standard for Safety Color Code. ANSI Z-535.1-2002. New York: American National Standards Institute. American National Standards Institute. (2002b). American National Standard for Environmental and Facility Safety Signs. ANSI Z-535.2-2002. New York: American National Standards Institute. American National Standards Institute. (2002c). American National Standard for Safety Symbols. ANSI Z-535.3-2002. New York: American National Standards Institute. American National Standards Institute. (2002d). American National Standard for Product Safety Signs and Labels. ANSI Z-535.4-2002. New York: American National Standards Institute. Backinger, C. L. and Kingsley, P. A. (1993, August). Write It Right: Recommendations for Developing User Instruction Manuals for Medical Devices Used in Home Health Care (HHS Publication FDA 93-4258). Rockville, MD: U.S. Department of Health and Human Services, Food and Drug Administration. Retrieved June 29, 2009, from http://www.fda.gov/downloads/Medical Devices/DeviceRegulationandGuidance/GuidanceDocuments/ucm070771.pdf. Bailey, R. W. (1989). Human Performance Engineering (2nd ed.). Englewood Cliffs, NJ: Prentice Hall. Bailey, R. W., Koyani, S., and Nall, J. (2000, September 7–8). Usability testing of several health information Web sites. In National Cancer Institute Technical Report. Bethesda, MD: National Cancer Institute. Bogner, M. S. (Ed.). (1994). Human Error in Medicine. Hillsdale, NJ: Lawrence Erlbaum Associates. Boyarski, D., Neuwirth, C., Forlizzi, J., and Regli, S. H. (1998). A study of fonts designed for screen display. In Proceedings of the SIGCHI Conference on Human Factors in Computing Systems (pp. 87–94). New York: ACM Press/Addison-Wesley. Callan, J. R. and Gwynne, J. W. (1993, September). Human Factors Evaluation of Contact Lens Labeling for Wearers of Soft Contact Lenses: Task VI final report. San Diego, CA: Pacific Science & Engineering Group. Callan, J. R., Gwynne, J. W., Sawyer, D., and Tolbert, M. T. (1993, December). Principles of Medical Device Labeling (NTIS PB 94-126851). San Diego, CA: Pacific Science & Engineering Group. Retrieved June 29, 2009, from http://www.fda.gov/downloads/MedicalDevices/ DeviceRegulationandGuidance/GuidanceDocuments/UCM095300.pdf. Czaja, S. J. and Nair, S. N. (2006). Human factors engineering and systems design. In G. Salvendy (Ed.), Handbook of Human Factors and Ergonomics (3rd ed., pp. 32–49). Hoboken, NJ: Wiley. Czerwinski, M., Larson, K., and Robbins, D. (1998). Designing for navigating personal Web information: Retrieval cues. In Proceedings of the Human Factors and Ergonomics Society 42nd Annual Meeting (pp. 458–462). Santa Monica, CA: Human Factors and Ergonomics Society. Donchin, Y., Gopher, D., Olin, M., Badihi, Y., Biesky, M., Sprung, C., Pizov, R., and Cotev, S. (1995). A look into the nature and causes of human errors in the intensive care unit. Critical Care Medicine, 23(2), 294–300. Dumas, J. S. and Redish, J. C. (1993). A Practical Guide to Usability Testing. Norwood, NJ: Ablex. Fernandes, K. (1993). User Interface Specifications for the Joint Maritime Command Information System (JMCIS), Version 1.3. San Diego, CA: Naval Command, Control, and Ocean Surveillance Center, RDT&E Division, Code 4221. Food and Drug Administration. (2001, April). Guidance on Medical Device Patient Labeling: Final Guidance for Industry and FDA Reviewers. Rockville, MD: Food and Drug Administration, Center for Devices and Radiological Health. Retrieved June 29, 2009, from http://www.fda. gov/MedicalDevices/DeviceRegulationandGuidance/GuidanceDocuments/ucm070782.htm. Hollnagel, E. (2006). Task analysis: Why, what, and how. In G. Salvendy (Ed.), Handbook of Human Factors and Ergonomics (3rd ed., pp. 373–383). Hoboken, NJ: Wiley. Horton, W. (1994). Designing and Writing Online Documentation: Hypermedia for Self-Supporting Products (2nd ed.). New York: Wiley.
TAF-ER562X_WINGR-09-0401-C005.indd 197
11/9/10 12:40:55 AM
198
Handbook of Human Factors in Medical Device Design
Inaba, K., Parsons, S. O., and Smillie, R. (2004). Guidelines for Developing Instructions. Boca Raton, FL: CRC Press. International Organization for Standardization. (2001). ISO 9186:2001. Graphical Symbols— Test Methods for Judged Comprehensibility and for Comprehension. Geneva: International Organization for Standardization. Jeavons, P. M. and Harding, G. F. A. (1975). Photosensitive Epilepsy. London: Heinemann. Kaye, R. and Crowley, J. (2000, July). Medical Device Use–Safety: Incorporating Human Factors Engineering into Risk Management. Rockville, MD: Food and Drug Administration, Center for Devices and Radiological Health. Kingsley, P. A. (1995, November). Draft Initial Report on Medical Device Labeling: Health Care Practitioners’ Medical Device Information and Labeling Needs, Results of Qualitative Research. Rockville, MD: U.S. Department of Health and Human Services, Food and Drug Administration, Center for Devices and Radiological Health. Kingsley, P. A. (1999, August). Draft Report on Medical Device Labeling: Patients’ and Lay Caregivers’ Medical Device Information and Labeling Needs, Results of Qualitative Research. Rockville, MD: U.S. Department of Health and Human Services, Food and Drug Administration, Center for Devices and Radiological Health. Kirwan, B. and Ainsworth, L. K. (Eds.). (1992). A Guide to Task Analysis. London: Taylor & Francis. Kobus, D. A., Moses, J. D., and Bloom, F. A. (1994). Effect of multimodal stimulus presentation of recall. Perceptual and Motor Skills, 78, 320–322. Koyani, S. J. and Nall, J. (1999, November). Web site design and usability guidelines. In National Cancer Institute, Communication Technologies Branch Technical Report. Bethesda, MD: National Cancer Institute. Leape, L. L., Brennan, T. A., Laird, N., Lawthers, A. G., Localio, A. R., Barnes, B. A., Hebert, L., Newhouse, J. P., Weiler, P. C., and Hiatt, H. (1991). The nature of adverse events in hospitalized patients: Results of the Harvard Medical Practice Study II. New England Journal of Medicine, 324(6), 377–384. Lynch, P. J. and Horton, S. (1999). Web Style Guide: Basic Design Principles for Creating Web Sites. New Haven, CT: Yale University Press. Retrieved June 29, 2009, from http://www.webstyleguide. com/wsg3.index.html. Marcus, A. (1992). Graphic Design for Electronic Documents and User Interfaces. New York: ACM Press. McCarthy, E. J., Backinger, C., Kaczmarek, R., Withiam-Wilson, M., Alexander, G., Coppola, A., Greishaber, M. F., Kingsley, P., Porter, B., Riley, K., et al. (1992, June). Medical Devices Used in Home Health Care (HHS Publication FDA 92-4252). Rockville, MD: U.S. Department of Health and Human Services, Food and Drug Administration. Meister, D. and Enderwick, T. P. (2001). Human Factors in System Design, Development, and Testing. Mahwah, NJ: Lawrence Erlbaum Associates. Morkes, J. and Nielsen, J. (1997). Concise, SCANNABLE, and Objective: How to Write for the Web. Retrieved April 13, 2010, from http://www.useit.com/papers/webwriting/writing.html. Nemeth, C. P. (2004). Human Factors Methods for Design: Making Systems Human-Centered. Boca Raton, FL: CRC Press. Nielsen, J. (1994). Usability Engineering. San Francisco: Morgan Kaufmann. Nielsen, J. (1996a). Marginalia of Web Design. Retrieved April 13, 2010, from http://www.useit. com/alertbox/9611.html. Nielsen, J. (1996b). Top Ten Mistakes in Web Design. Retrieved April 13, 2010, from http://www. useit.com/alertbox/9605.html. Nielsen, J. (1999). When Bad Design Elements Become the Standard. Retrieved April 13, 2010, from http://www.useit.com/alertbox/991114.html. Nielsen, J. (2000). Designing Web Usability. Indianapolis: New Riders. Nielsen, J. (2007). Breadcrumb Navigation Increasingly Useful. Retrieved April 13, 2010, from http://www.useit.com/alertbox/breadcrumbs.html.
TAF-ER562X_WINGR-09-0401-C005.indd 198
11/9/10 12:40:55 AM
Documentation
199
Pilot, L. R. (1999, May). Medical device labeling in the European Union. Medical Device and Diagnostic Industry. Retrieved June 29, 2009, from http://www.devicelink.com/mddi/ archive/99/05/012.html. Rogers, W. A. (1997). Designing for an Aging Population. Santa Monica, CA: Human Factors and Ergonomics Society. Rogers, W. A. and Fisk, A. D. (Eds.). (2001). Human Factors Interventions for the Health Care of Older Adults. Mahwah, NJ: Lawrence Erlbaum Associates. Rogers, W. A., Fisk, A. D., and Walker, N. (1996). Aging and Skilled Performance. Mahwah, NJ: Lawrence Erlbaum Associates. Rubenstein, R. (1988). Digital Typography: An Introduction to Type and Composition for Computer System Design. Reading, MA: Addison-Wesley. Rubin, J. (1994). Handbook of Usability Uesting: How to Plan, Design, and Conduct Effective Tests. New York: Wiley. Ryan, J. P. (1991). Design of Warning Labels and Instructions. New York: Van Nostrand Reinhold. Simpson, H. and Casey, S. M. (1988). Developing Effective User Documentation: A Human Factors Approach. New York: McGraw-Hill. Spool, J. M., Scanlon, T., Schroeder, W., Snyder, C., and DeAngelo, T. (1997). Web Site Usability: A Designer’s Guide. North Andover, MA: User Interface Engineering. Tinker, M. A. (1963). Legibility of Print. Ames: Iowa State University Press. U.S. Code of Federal Regulations. (2003a, April 1). Title 21 Food and Drug Administration, DHHS, Part 801—Labeling. U.S. Code of Federal Regulations. (2003b, April 1). Title 21 Food and Drug Administration, DHHS, Part 809—In vitro diagnostic products for human use, Subpart B—Labeling. Vanderheiden, G. C. (2006). Design for people with functional limitations. In G. Salvendy (Ed.), Handbook of Human Factors and Ergonomics (3rd ed., pp. 1387–1417). Hoboken, NJ: Wiley. Venkatacharya, P. S. (2000, October). Techniques for Creating User-Friendly Enterprise Portals: Developing for the Oracle Internet Platform. Retrieved April 13, 2010, from http://otn.oracle. com/products/iportal/pdf/tech_create_user_friend_eportals.pdf. Wieringa, D., Moore, C., and Barnes, V. (1993). Procedure Writing: Principles and Practices. Columbus, OH: Battelle Press. Wiklund, M. E. (1998, May). Making medical device interfaces more user-friendly. Medical Device and Diagnostic Industry. Retrieved June 29, 2009, from http://www.devicelink.com/mddi/ archive/98/05/032.html. Wiklund, M. E. (2002, January). Medical device user manuals: Shifting toward computerization. Medical Device and Diagnostic Industry. Retrieved June 29, 2009, from http://www. devicelink.com/mddi/archive/02/01/003.html. Wiklund, M. E. (2004, February). Intuitive design: Removing obstacles also increases appeal. Medical Device and Diagnostic Industry. Retrieved June 29, 2009, from http://www.devicelink. com/mddi/archive/04/02/005.html. Wogalter, M. S. (Ed.). (2006). Handbook of Warnings. Mahwah, NJ: Lawrence Erlbaum Associates. Wogalter, M. S., Godfrey, S. A., and Fontenelle, G. A. (1987). Effectiveness of warnings. Human Factors, 29, 599–612. Wright, P. (1999). Writing and information design of healthcare materials. In C. Candlin & K. Hyland (Eds.), Writing: Texts, Processes and Practices. London: Addison-Wesley Longman. Young, S. L. and Wogalter, M. S. (1990). Comprehension and memory of instruction manual warnings: Conspicuous print and pictorial icons. Human Factors, 32, 637–649.
TAF-ER562X_WINGR-09-0401-C005.indd 199
11/9/10 12:40:55 AM
TAF-ER562X_WINGR-09-0401-C005.indd 200
11/9/10 12:40:55 AM
6 Testing and Evaluation Edmond W. Israelski, PhD CONTENTS 6.1 Purpose ....................................................................................................................202 6.1.1 Constraints and Limitations of This Guidance...........................................203 6.2 Types of Usability Tests ...........................................................................................203 6.2.1 Formative Usability Testing........................................................................203 6.2.2 Summative Usability Testing ......................................................................205 6.3 Principles of Good Usability Test Design ............................................................... 206 6.3.1 Critical Usability Test Design Considerations ........................................... 206 6.4 Usability Testing: Overview ....................................................................................207 6.4.1 Usability Test Plan Content ........................................................................207 6.4.1.1 Purpose ....................................................................................... 209 6.4.1.2 Setting ........................................................................................ 209 6.4.1.3 Participants ..................................................................................212 6.4.1.4 Prototypes or Simulations ...........................................................212 6.4.1.5 Methodology ...............................................................................213 6.4.1.6 Tasks ............................................................................................ 214 6.4.1.7 Usability Objectives .................................................................... 216 6.4.1.8 Data Collection ............................................................................ 217 6.4.1.9 Data Analysis .............................................................................. 217 6.4.1.10 Reporting ..................................................................................... 218 6.5 Logistics................................................................................................................... 219 6.5.1 Testing Locations ........................................................................................ 219 6.5.2 Number of Participants ............................................................................... 219 6.5.3 Recruiting Activities ...................................................................................220 6.5.4 Testing Staff Size ........................................................................................220 6.5.5 Session Length ............................................................................................220 6.5.6 Video Recording .........................................................................................221 6.5.7 Note Taker ..................................................................................................221 6.5.8 Language Translators..................................................................................221 6.5.9 Data Logging Software...............................................................................222 6.5.10 Screen Capture ...........................................................................................222 6.5.11 Eye Scan Capture........................................................................................222 6.6 Protocol-Related Activities ......................................................................................224 6.6.1 Participant Orientation................................................................................224 6.6.2 Consent Forms ............................................................................................224 201
TAF-ER562X_WINGR-09-0401-C006.indd 201
11/9/10 12:49:41 AM
202
Handbook of Human Factors in Medical Device Design
6.6.3 6.6.4 6.6.5 6.6.6
Nondisclosure Forms ..................................................................................224 Pretest Questionnaire ..................................................................................225 Participant Training prior to Testing ..........................................................225 Directed Tasks ............................................................................................225 6.6.6.1 Think-Aloud Protocol..................................................................225 6.6.6.2 Codiscovery .................................................................................226 6.6.6.3 Self-Exploration ...........................................................................226 6.6.7 Interviews ...................................................................................................227 6.6.8 Posttest Questionnaire ................................................................................227 6.6.8.1 Examples of Questionnaire Items................................................227 6.6.9 Debriefing ...................................................................................................228 6.6.10 Record Keeping ..........................................................................................228 6.6.11 Testing Team Debriefing.............................................................................229 6.7 Sources of Test Bias .................................................................................................229 6.7.1 Common Testing Mistakes .........................................................................229 6.8 Supplemental Usability Evaluation Methods ...........................................................229 6.8.1 Cognitive Walk-Through ...........................................................................230 6.8.2 Expert Review ............................................................................................230 6.8.3 Heuristic Review.........................................................................................230 Appendix 6.A: Examples of Usability Test Plans ............................................................231 Appendix 6.B: Usability Test Report Checklist ...............................................................237 Appendix 6.C: Statistical Justification for Sample Sizes in Usability Tests ................... 240 Appendix 6.D: Frequently Asked Questions ...................................................................245 Resources .........................................................................................................................248 References ........................................................................................................................248
Testing and evaluation covers usability testing and all the considerations needed to plan, execute, analyze, and report on these studies. This chapter gives guidance on principles for developing usability tests as well as advice on how variables are chosen, controlled, measured, and analyzed. Usability testing is a cornerstone of the best practices for the design of medical devices.
6.1 PURPOSE This chapter provides key methodological details on how to plan and conduct usability evaluations to obtain valid and reliable data to judge the safety and effectiveness of the medical device user interfaces. As described in Human Factors Design Process for Medical Devices (ANSI/AAMI HE-74-2001), usability testing is only one part of good human factors design methodology, and it must be used in conjunction with other important methods, such as task analysis, user profiling, use environment analysis, use error risk analysis, and iterative user-centered design. Usability testing alone will not ensure the development of safe, effective, and easy-to-use medical devices. This chapter covers a range of best practices for conducting usability tests. A great deal can be learned with simple usability tests that do not require a major financial investment or elaborate equipment. The usability testing program should be executed systematically and controlled for sources of bias and unreliability.
TAF-ER562X_WINGR-09-0401-C006.indd 202
11/9/10 12:49:41 AM
Testing and Evaluation
203
6.1.1 CONSTRAINTS AND LIMITATIONS OF THIS GUIDANCE Usability testing should never be considered a substitute for product validation. Usability testing can verify that design outputs match design inputs; that is, the actual design has been executed to match the design specifications as described in the product development plan. Further, usability testing can validate that device usability meets usability objectives. If the usability objectives have been validated with users through rigorous user research, then the usability test can contribute to the validation of the user interface. Medical device validation should include laboratory testing and possibly clinical testing to evaluate device efficacy, reliability, safety, and performance. Another important limitation of usability testing is the false belief that user-interface quality can be tested into a device. Iterative or multiple rounds of design and redesign lead to usability improvements, but if up-front user needs analysis (also called contextual inquiry) is not conducted to understand the users, their tasks and goals, and the use environment, then usability testing alone may never allow valid usability goals or objectives to be met. This may be true no matter how many rounds of iterative design, redesign and testing are conducted. Usability testing can be employed during both user-interface design verification and validation. User-interface verification can ensure that design outputs meet design inputs for measures of usability. User-interface validation can ensure that device use is consistent with user requirements. This may be accomplished by having acceptance criteria for the final summative usability test based on the following: • Valid usability objectives that truly reflect user requirements • Rigorous use of risk analysis and risk management in the selection and prioritization of usability objectives as acceptance criteria • User-interface verification and validation being only one part of overall device verification and validation, which should include all product effectiveness and safety measures Table 6.1 provides definitions for the terms used in this chapter.
6.2 TYPES OF USABILITY TESTS Usability testing includes two major categories.
6.2.1 FORMATIVE USABILITY TESTING Formative usability testing is performed early with simulations and early working prototypes and explores whether usability objectives are attainable. Formative testing typically does not have strict acceptance criteria. • Exploratory testing—Tests users performing high-level tasks or walking through the tasks using low-fidelity simulations (e.g., paper or foam models). Concepts are being tested at this stage of development (e.g., a usability test of a computer simulation of a touch-screen user interface to a heart monitor or paper sketches of the navigational buttons and menus for such a device).
TAF-ER562X_WINGR-09-0401-C006.indd 203
11/9/10 12:49:41 AM
204
Handbook of Human Factors in Medical Device Design
TABLE 6.1 Definitions Term
Definition
Contextual inquiry*
The process of observing and working with users in their normal environment to better understand the tasks they do and their workflow. Effectiveness** The accuracy and completeness with which users achieve specified goals. Efficiency** The resources expended in relation to the accuracy and completeness with which users achieve goals. Efficiency in the context of usability is related to “productivity” rather than to its meaning in the context of software efficiency. Formative usability Usability testing that is performed early in the design process with simulations testing or the earliest working prototypes and explores whether usability objectives are attainable but does not have strict acceptance criteria. Prototyping and iterative The design process that involves the rapid turnaround of user-interface design prototypes or simulations that are usability tested and improved in an iterative cycle until usability objectives are attainable. Summative usability Usability testing that is performed in the late stages of design. These tests testing contribute to design verification and validation. It is a recommended best practice to have formal acceptance criteria (e.g., usability objectives for human performance and satisfaction ratings). Task analysis* Task analysis is a family of systematic methods that produce detailed descriptions of the sequential and simultaneous manual and intellectual activities of personnel who are operating, maintaining, or controlling devices or systems. Usability inspection Inspection methods involve analytical reviews and systematic walk-throughs of methods user interactions with simulated or working user-interface designs to identify usability problems. Usability objectives* Usability objectives (or goals) are a desired quality of a user-device interaction that may be expressed in written form, stipulating a particular usability attribute (e.g., task speed) and performance criteria (e.g., number of seconds). Usability testing* Procedure for determining whether the usability goals have been achieved. Usability tests can be performed in a laboratory setting, in a simulated environment, or in the actual environment of intended use. User Any person who interacts with the device. User interface The hardware and software aspects of a device that can be seen (felt, heard, or otherwise perceived) by the user and the commands and mechanisms the user employs to control device operation and input data. The interface includes labeling, instructions for use, and training materials. User group** Subset of users who are differentiated from other users by factors such as age, culture, or expertise that are likely to influence usability (also described as a distinct user group). Use environment* The actual conditions and settings in which users interact with the device or system. Use error Use error is characterized by a repetitive pattern of failure that indicates that a failure mode is likely to occur with use and thus has a reasonable possibility of predictability of occurrence. Use error can be proactively identified through the use of techniques such as usability testing and hazard analysis. Use error can be addressed and minimized by the device designer.
TAF-ER562X_WINGR-09-0401-C006.indd 204
11/9/10 12:49:41 AM
Testing and Evaluation
205
TABLE 6.1 (CONTINUED) Definitions Term Use error risk analysis User error
User profiles*
Definition Risk analysis method focused on the use error component of fault and hazard analysis for medical devices. User error is characterized by an isolated pattern of failure that indicates a failure mode that is due to fundamental errors by humans and has no reasonable possibility of being predicted. User error is not readily preventable and cannot feasibly be addressed by the device designer. Summary of the mental, physical, and demographic traits of the end-user population as well as any special characteristics, such as occupational skills and job requirements, that may have a bearing on design decisions.
*From Human Factors Design Process for Medical Devices (ANSI/AAMI HE-74-2001). **From Common Industry Format for Usability Test Reports (ANSI/NCITS 354-2001).
• Assessment testing—Tests that give the users realistic tasks to perform on working prototypes or more fully developed simulations, usually without patients attached to the device (e.g., usability testing of the feel and control of a working prototype of a handheld glucose meter). • Comparison (contrast) testing—Tests comparing two or more design alternatives (e.g., a test to measure the effectiveness and alerting properties of two sets of competing auditory alarms for an infusion pump). • Comparison (competitive) testing—Tests that gather usability performance of competitors’ products. These tests could be part of the design exploration to understand the best features of existing devices or could be used to support marketing claims for a device (e.g., usability test comparing task success rate and time to run a blood chemistry test using a variety of on-market handheld point-of-care blood analyzers).
6.2.2 SUMMATIVE USABILITY TESTING Summative usability testing is performed late in design as part of formal device verification and validation. It is a recommended best practice to have formal acceptance criteria (e.g., usability objectives for human performance and user satisfaction ratings). Guideline 6.1: Summative Usability Testing Sample Size Summative usability testing should use a sufficient sample size to allow performance of statistical tests on the reliability of the results.
If iterative rounds of formative usability testing are performed as recommended in this chapter and in the usability literature, then typically there will be few usability surprises uncovered during late-stage summative testing. Recommendations for statistically justifiable sample sizes are given in Appendix 6.A. The remaining sections of this chapter cover best practices for both formative and summative usability testing.
TAF-ER562X_WINGR-09-0401-C006.indd 205
11/9/10 12:49:41 AM
206
Handbook of Human Factors in Medical Device Design
6.3 PRINCIPLES OF GOOD USABILITY TEST DESIGN In usability tests, representatives of the intended user population interact with one or more device models, prototypes, or production units to assess ease of learning, efficiency of use, ease of use, ease of remembering, safety, and/or user appeal (Nielsen, 1994), among the many possible attributes of interest. Usability tests can be performed in a laboratory setting, in a simulated environment, or in the actual use environment. Usability testing, especially when conducted in the actual use environment, is a particularly effective way to detect use errors. However, because the subject populations typically are small, use errors having an inherently low probability of occurrence across the entire user population may not be detected. Statistically, it can be shown that small sample size usability tests are not powerful enough to detect low-frequency use errors. For this reason, the use of additional complementary techniques, such as risk analysis and usability inspection methods, is essential. Testing some types of devices to the limit of their functional capabilities may place patients at risk, which is unacceptable.
6.3.1 CRITICAL USABILITY TEST DESIGN CONSIDERATIONS Usability testing should be done carefully and systematically, or else the resulting data may not be valid or reliable. Poor data may lead to poor design decisions and ultimately an errorprone and unsafe medical device. Guideline 6.2: Usability Testing Protocol Development Usability testing protocols should be developed in collaboration with professional-level human factors specialists. Nonspecialists under the direction and training of professionals can handle the actual execution and reporting of the test.
There are many factors and characteristics that must be considered in designing a usability testing protocol. Table 6.2 summarizes these characteristics. A usability test has some very high-level attributes that are outlined below and explained in more detail in subsequent sections of this chapter. TABLE 6.2 Characteristics of a Typical Usability Test Test Plan Content • • • • • • • •
Purpose Setting Participants Prototypes Simulations Methods Tasks Usability objectives • Data collection • Data analysis • Reporting methods
TAF-ER562X_WINGR-09-0401-C006.indd 206
Logistics • Testing locations • Number of participants • Recruiting activities • Testing staff size • Session length • Video recording • Note taker • Language translator • Data logging software • Screen capture • Eye scan capture
Protocol Activities • • • • • • • • • •
Orientation Consent Nondisclosure Prequestionnaire Type of training Directed tasks Self-exploration Interview Postquestionnaire Debriefing
Data • • • • • • • • • •
Task times Task completion Efficiency Significant errors Ratings Rankings Verbal comments Questionnaires Videos Photo images
11/9/10 12:49:42 AM
Testing and Evaluation
207
Guideline 6.3: Usability Testing Plan A usability testing plan should specify the test subjects, sample size (and rationale), tasks to be tested, test environment, device to be tested, performance measures, and specific testing procedure.
The sample size should be small but sufficient (explained in more detail in Appendix 6.C). For many usability tests, the size may be as small as five to eight for early formative testing (one round) and 15 to 20 for later summative testing against usability objectives. The test should focus on realistic tasks based on user scenarios derived from previous task analysis and risk analysis. The later stage tests should occur in a realistic simulation of the intended use environment (e.g., lighting, noise, other equipment, and so on) and use the actual device, a prototype, or a sufficiently interactive simulation. Testing should focus on actual user performance (task completion, time, errors, and so on), although participant opinions can provide valuable information. The test is typically recorded (audio and/or video).
6.4 USABILITY TESTING: OVERVIEW This section gives specific guidance on the details of designing and executing a usability test. Table 6.3 presents a high-level description of the most common types of usability tests. Guideline 6.4: Frequency of Usability Testing Usability testing should be applied iteratively at various stages in the user-centered design process.
Figure 6.1 suggests where in the device development process usability testing and its various forms should be applied. In particular, formative usability testing is an important part of the product design cycle as part of the process of specifying user-interface design output. In an iterative process, the final user-interface specification can be achieved through successive usability tests, each attempting to measure whether the usability requirements (objectives) can be attained. Summative usability testing is an important part of verification (design outputs meet design inputs) and validation (user requirements in the form of usability objectives are met).
6.4.1 USABILITY TEST PLAN CONTENT As noted earlier, usability testing is a scalable process. For simple, low-risk products, simple and inexpensive testing may be sufficient. For newer technology or more complex user interfaces, a more comprehensive usability testing plan may be required. The higher the risk indexes estimated from risk analysis, the greater the need for more comprehensive usability testing. For example, a simple drug injection device might require only a single formative test and one final summative usability test involving a few simple tasks. In contrast, a complex new infusion pump might require five or more iterative formative tests using computer simulations of the interface screens as well as a series of investigational tests of display readability and alarm effectiveness. The number of tasks included in the usability tests would be sufficient to address the most critical and high-risk use scenarios.
TAF-ER562X_WINGR-09-0401-C006.indd 207
11/9/10 12:49:42 AM
208
Handbook of Human Factors in Medical Device Design
TABLE 6.3 Types of Usability Testing and Considerations
Type
When in Design Cycle
Description
Minimum Sample Size (per group)
Formative Usability Testing Conceptual 5–8 design or higher
Considerations
Exploratory
High-level test of users performing tasks.
Assessment
Gives representative users real tasks to perform (1:1 or two working together).
Iterative throughout the design cycle
5–8
Early designs or computer simulations. Used to gauge whether usability objectives are attainable.
Comparative and contrast
Compares two or more design alternatives.
During design
5–8
Used to decide if one user interface concept or prototype is better than others.
Comparative and competitive
Tests against competitors’ user interfaces.
During design
5–8
Used to learn about competitors’ devices, best features. May be done early during conceptual design or near the end of design cycle.
Validation
Based on simulations of early concepts; could be low-fidelity paper prototypes or foam core models. Useful to use think-aloud protocols.
Summative Usability Testing Tests real users and real End of design At least 15–20 Used to verify and validate tasks with market cycle (often larger) interface design; usability ready device. objectives have acceptance criteria. Should include training, documentation, and labeling.
The summative usability test(s) would need to be comprehensive in terms of tasks and the number of different representative user groups included. Guideline 6.5: Usability Test Plan Design Whether the usability test is simple or comprehensive, the usability test plan should describe the goals for the usability test and how acceptance criteria will be determined. It is very important that the usability test plan describes the following: Purpose Setting Participants
Tasks Usability objectives Data collection
Methodology Prototypes/simulations Data analysis Reporting
Appendix 6.B contains a Usability Test Report Checklist that describes the important elements of a usability test plan. In the following, additional information is provided on
TAF-ER562X_WINGR-09-0401-C006.indd 208
11/9/10 12:49:42 AM
Testing and Evaluation
Design control activities
Concept phase
Design input
Design output
Verification
Validation
Perform studies & analysis
Design requirements
Design specifications
Test output against input
Test against user needs
Contextual inquiry
Human factors activities
Literature reviews Complaints analysis Market research
FIGURE 6.1
209
Task analysis User profiles Use environment Heuristic review Risk analysis Usability objectives
Prototyping/ simulations Iterative design Formative usability testing
Expert reviews Cognitive walkthroughs
Risk analysis
Summative usability testing
Cognitive walkthroughs
Risk analysis
Production units (or equivalent) Summative usability testing Field studies
Usability evaluation in the device design control process.
the contents of each section of the usability test plan, which is sometimes described as a usability testing protocol. 6.4.1.1 Purpose It is important to explain the high-level objectives or purpose for conducting the usability test. The objectives should at least address the following questions: • Is it exploratory or a design alternatives comparison test? • Is it an early-stage formative test or a late-stage summative test? 6.4.1.2 Setting The test setting can place constraints on how a test can be conducted. Testing in the actual clinical environment might mean that some control will be lost over the variables manipulated in the test. If a professional usability testing facility or laboratory is used, then more experimental control can be achieved, for example, by systematically varying levels of ambient light in studying a nurse’s ability to read a patient monitor. However, a laboratory can never re-create the full contextual richness of the actual use environment. In a clinical setting, there might be only one or two levels of ambient light. A trade-off needs to be made among test setting choices. High-fidelity simulations, such as mock operating rooms and intensive care units, can provide advantages for usability testing. High fidelity in this context means that the testing environment closely resembles the actual environment of use. These settings often include a computer-controlled patient simulator; a full suite of medical devices, accessories, and supplies; and even confederate clinicians who act like they are delivering actual patient care. Such high-fidelity test environments may be particularly valuable for summative usability testing performed late in the design cycle as described in Table 6.3. Figure 6.2 shows a realistic patient simulation complete with computer-controlled patient mannequins. Figure 6.3 shows a typical usability lab as seen from the control room, which is the room for controlling the videotaping equipment and systems being tested. Figure 6.4 shows a typical
TAF-ER562X_WINGR-09-0401-C006.indd 209
11/9/10 12:49:42 AM
210
Handbook of Human Factors in Medical Device Design
FIGURE 6.2
Patient simulator lab configured for usability testing.
FIGURE 6.3
Typical usability lab as seen from an observation and control room.
TAF-ER562X_WINGR-09-0401-C006.indd 210
11/9/10 12:49:42 AM
Testing and Evaluation
211
VCR & camera controls
Additional Observation
Participant Side
Observer Side Table
Sliding glass door
FIGURE 6.4
Floor plan of typical usability testing lab.
floor plan of a usability testing lab. Observers and technicians can view the test participants through one-way glass. The participants only see a mirror from their side of the one-way glass. Finally, Figure 6.5 shows a person operating the control console of a portable usability testing and videotaping system that is easily carried into actual clinical settings, which in this case is a diagnostic laboratory. Recruiting of clinical test participants would be easier in a clinical setting, but then clinicians could be more easily called away to attend to a critical, clinical situation unless they are recruited to participate in the usability test while off shift. In an off-site professional location, unplanned interruptions are less of a problem. One consideration in choosing the proper setting is whether the test moderator should be in the same room as the test participant. When together, the chances of the moderator influencing the results are greater because there are more opportunities for the observer’s body language and gestures to lead the participant, for example, inadvertently giving them biasing hints on how to complete a task. However, some tasks require a fair amount of physical interaction between moderator and participant, which necessitates physical colocation. See Wiklund (1994) for a detailed discussion regarding usability testing facility layouts and floor plans.
FIGURE 6.5
Moderator operating a portable usability testing control console.
TAF-ER562X_WINGR-09-0401-C006.indd 211
11/9/10 12:49:43 AM
212
Handbook of Human Factors in Medical Device Design
6.4.1.3 Participants Guideline 6.6: Usability Test Participants Usability test participants should be representative of the most likely users of the device. Participants should be selected on the basis of the personal characteristics described in the user profiles that are typically gathered during the initial contextual inquiry phase of design.
User profiles include information about training, skills, experience levels, demographics, job type, and motivation, among other defining characteristics. Sometimes, predesign analysis will also describe unique user groups for which the device is targeted. For example, an infusion pump might have users who are nurses, doctors, technicians, and patients, each of which is a unique user group. The number of test participants selected from each unique user group depends on whether the test is formative or summative, as described more completely in the “Number of Participants” section. Guideline 6.7: Device Development Team Members Product development team members and others with special product knowledge should not be usability test participants because they will have biases and are distinctly different from real device users. Section 6.5.3 describes considerations in recruiting test participants.
6.4.1.4 Prototypes or Simulations Early prototypes or simulations are best derived from the user-interface requirements obtained in the user analysis conducted in the early development stages. Contextual inquiry is typically used in these early stages. Employ usability testing and usability inspection methods to assess how closely the early prototype designs match user needs and usability objectives. Gather usability feedback to improve the design for further testing in an iterative fashion. Some recommended types of prototypes and simulations include: • Paper-based, low-fidelity simulations (e.g., rough hand-sketched paper screen shots of a handheld glucose meter that are placed in front of a user and sequenced as a working model might behave). See Synder (2003) for more detail on the methods of paper prototyping. • Wizard of Oz simulations whereby a human simulates a not-yet-functional system and mimics its operation during a usability test. For example, to simulate a voice-activated surgical robot, a human pretends to be a speech recognition system by listening to the words spoken by a test subject and typing them into a computer that in turn displays these words to the test subject. • Computer simulations (typically screen based, e.g., Flash™, Macromedia Director™, Visual Basic™, Java, or HTML). Such computer prototypes can realistically simulate intended user interactions while sacrificing full fidelity (e.g., touch screen to mimic actual hardware buttons on an infusion pump) yet at a lower cost. • Horizontal and vertical simulations. Horizontal simulations are broad and show the range of navigation options and high-level features (e.g., the high-level navigation bar for a computer workstation that controls a blood gas analyzer). Vertical simulations are narrow and deep and show how one particular feature works in depth (e.g., a full set of functions for all the steps in calibrating a computer-controlled immunoassay diagnostic system). Hybrid simulations combine horizontal simulations with
TAF-ER562X_WINGR-09-0401-C006.indd 212
11/9/10 12:49:43 AM
Testing and Evaluation
213
a limited set of vertical simulations of critical device features (e.g., a touch-screen infusion pump simulating the global navigation and two features, such as delayed therapy and bolus delivery). • Working early prototypes of the actual device or parts of it. The caveat is that early software can be unstable and recovery steps need to be created if the software crashes during the usability tests. 6.4.1.5 Methodology The method description section of a usability protocol and subsequent report is much like the “methodology” section of any scientific report. Guideline 6.8: Description of Test Methods The usability test plan should describe the usability study methodology so that another researcher or developer has sufficient information to be able to replicate the test. It should cover all of the items discussed in the “Test Plan Contents” (6.4.1) and “Logistics” (6.5) sections of this chapter.
A usability test is similar in many ways to a psychological experiment. Most of the rules and considerations for good experimental design are applicable. For example, if several independent variables are being manipulated in the usability test (e.g., some subjects get training and others do not, and the subjects are from two levels of relevant medical device experience), then a balanced experimental design is required to avoid “confounding” of the independent variables. Confounding occurs when two or more independent variables are mixed in such a way that their individual impact on the outcome measures cannot be determined. For example, if a study were designed to compare reading accuracy of a cathode ray tube (CRT) versus an LCD display but used larger character sizes on the CRT screen, then the variables of display type and character size would be confounded. In this case, it would not be possible to determine whether better reading scores for the CRT were due to better image quality or to having larger display characters compared to the LCD screen. The usability study design should take into account the goals of the study (i.e., the key questions it is intended to answer) and the methods to be employed. An early issue is how independent variables in the design will be controlled. The independent variables can be varied systematically by choosing multiple levels (those of most importance for design), held constant by using only one level, or randomized and possibly presented in counterbalanced orders to control for learning order effects (see Section 6.4.1.6.2, “Task Presentation Order”). 6.4.1.5.1 Experimental Design A key study design attribute is how the participants will be allocated across the independent variables. In “within-subjects” design, each person experiences multiple levels of one or more independent variables. This situation is sometimes called “repeated measures.” An example would be one group of 15 nurses where each is tested on loading syringes into three different infusion pump designs. In “between-subjects” design, separate groups are defined in which participants experience only one level of each independent variable. An example would be three different groups of five nurses, where each is tested loading syringes into only one of the three different pump designs. Mixed designs may also be appropriate, particularly when two or more independent variables are to be tested. For
TAF-ER562X_WINGR-09-0401-C006.indd 213
11/9/10 12:49:43 AM
214
Handbook of Human Factors in Medical Device Design
example, two groups of nurses are studied on three different interfaces (i.e., type of interface is the within-subjects variable) where one group gets prior training and the other group does not (training is the between-subjects variable). The choice of experimental design will determine the type of statistical analysis required and will affect the power of the study to reliably answer quantitative questions of interest. The test design must be robust and avoid all sources of test data bias (see Section 6.7 for a discussion regarding potential sources of bias during data collection). 6.4.1.6 Tasks Guideline 6.9: Describe Tasks to be Performed The usability test plan should describe the tasks that participants will be asked to perform and needs to specify the task presentation order.
6.4.1.6.1 Task Selection The selection of tasks is an important part of any usability test plan. For the usability test to have the most value, the tasks chosen should be representative of the frequent, urgent, critical (safety and performance), and most challenging tasks performed with the medical device. Task analysis studies should be used to identify the usability test tasks. Task analysis is done through contextual inquiry techniques of systematic observation and collection of data to create task flows, use cases, and typical usage scenarios, and other products (Hackos and Redish, 1998). The output of the task analysis answers the essential question of what users expect to do with the medical device. Some task analysis outputs include: • Task flows • Block diagrams, task tables, and task flowcharts • Use cases (a concept well known by software engineers used to describe software module requirements) • Task statements or usage scenarios • Narrative descriptions of the steps needed to accomplish an activity • Task statement, including actor, action verb, object of action, and purpose of task Guideline 6.10: Test Safety Critical Tasks Tasks selected for testing in a usability study should include those tasks that are safety critical, essential to successful device function, and common device-related tasks. The most safetycritical tasks should be derived from use-error–focused risk analysis, those tasks receiving relatively high risk estimations (frequency × severity), or tasks known from customer complaints and other databases to cause adverse events.
The determination of most essential tasks required to use a device can be based on frequency, importance, urgency, or difficulty. Examples of essential medical device tasks include the following: • Priming an IV line • Turning the device on and calibrating it • Executing the main therapy
TAF-ER562X_WINGR-09-0401-C006.indd 214
11/9/10 12:49:43 AM
Testing and Evaluation
215
• Reacting to a fault condition • Stopping a device during a critical incident Frequency of tasks to be performed in actual clinical settings should also be considered in task selection but is secondary to task importance, criticality, and impact on patient or user safety. Not all tasks need to be performed by all subjects in a usability test. Analytical usability inspection methods, such as cognitive walk-throughs, are discussed in Section 6.8 as recommended methods to comprehensively evaluate all user tasks. This is necessary in complex devices with many features and tasks since it would not be practical to include all tasks in a single usability test. The number of tasks in the test will also be influenced by test session length. Test session length must often be kept within reasonable time limits because of participant availability and facility constraints. Typical session lengths are 1 to 2 hours. Times may be longer for more complicated systems, such as laboratory diagnostic systems, and may range from 2 to 4 hours. It is possible to split usability task scenarios among the test participants as long as the minimum number of test subjects for any critical task meets the sample size recommendations cited in Appendix 6.C. One participant might perform 5 to 15 tasks, assuming that each task takes 3 to 10 minutes to complete. The total tasks in the test might be up to 25, depending on the complexity of the user interface and the number of safety-critical tasks. See Section 6.5.5 for additional discussion. If learning rate and learning time are of interest, then some tasks may be presented multiple times to users. The users’ performance on repeated exposures to the same tasks would allow a learning curve to be derived. The order of task presentation can affect learning. Tests must be designed to minimize training bias. 6.4.1.6.2 Task Presentation Order Another consideration when designing a usability test is the order of task presentation. The test design must avoid learning effects on task performance due to the ordering of task presentation. Guideline 6.11: Task Presentation Order Task presentation order should follow standard experimental design considerations, such as counterbalancing the order of presentation to control for learning order effects.
Counterbalancing typically means controlling for order effects by having equal subsamples of test participants perform the tasks in different orders that balance out which tasks follow each other in the different sequences. If there were two tasks, then the protocol would be to have half the participants do task A followed by B and the other half do task B followed by A. Another technique to minimize task order effects is to have each subject perform the tasks in a unique random order. Sometimes there is a logical progression of tasks that makes counterbalancing difficult. But the main point is that test designers need to be sensitive to order effects. Counterbalanced designs can get much more complicated when there are more than two conditions. Counterbalancing strategies for multiple conditions include Latin squares, where, for three variables, six orders may be presented that are completely balanced for order and sequential positioning of the independent variables. The six orders that result in each condition being counterbalanced for order and sequencing would be ABC, BCA, CAB, CBA, ACB, and BAC. See Winer (1971) for additional information on related design of experiment techniques.
TAF-ER562X_WINGR-09-0401-C006.indd 215
11/9/10 12:49:43 AM
216
Handbook of Human Factors in Medical Device Design
6.4.1.6.3 Task Scenario Examples Examples of task scenarios that might be given to a test participant to perform include the following: 1. IV infusion pump drug therapy initiation: Order details are as follows—dopamine; 3 mcg/kg/min, 400 mg in 250 ml DsW. Step 1. Load drug in disposable tubing and prime infusion set. Step 2. Program the infusion pump for above order. Step 3. Start therapy. 2. In vitro diagnostic device system example: Step 1. Based on the reagent status and load list, determine which of the reagents need to be replaced or added to modules 1 and 2. Step 2. Add or replace the reagents that are needed in modules 1 and 2. 3. Cardiac defibrillator example: Take the automatic cardiac defibrillator out of the closet, unpack it, and demonstrate how it would be used on a simulated patient who is in ventricular fibrillation with low blood pressure. 4. Cardiac monitor example: Make the following adjustments: • Change trending preference from cardiac index to cardiac output. • Change display time trend to 3 hours. • Change the cardiac output averaging mode to “urgent.” • Change the cardiac output scale to 10 L/min. • Change the lower alarm limit for cardiac output to 4.0 L/min. It is important to avoid terms in the task scenario statements that give the test participant step-by-step instructions or even a mild hint about how to perform a task. Give general instructions that are sufficient to complete the task. Do not simply lead the participant directly to the answer. For early formative usability tests, it can be useful to give very general task statements that encourage the participants to self-explore the user interface and to have a think-aloud dialogue about the interface and its features, such as navigation and organizing structure. 6.4.1.7 Usability Objectives The main goal of specifying usability objectives (also known as usability requirements, usability goals, performance goals, or human factors requirements) is to create a metric that can be applied during usability testing as a way of assessing whether the interface is acceptable for the test. Usability objectives are the best way to create a quantitative qualityrelated criteria. Moreover, usability goals should emphasize objective measures of user performance with the device (functional goals) rather than user opinions about the device. Typically, quantified usability objectives include the following: • • • • • • •
User performance goals (objective goals) Task completion time Success rate or error rate and type Learning time Accuracy Efficiency (number of total steps and missteps) Number of references to instructions or online help
TAF-ER562X_WINGR-09-0401-C006.indd 216
11/9/10 12:49:43 AM
Testing and Evaluation
217
• User satisfaction (subjective goals) • Rating scales (Likert scales, comparative rating; see Section 6.6.8.1) • Semantic differential (pick rating between two opposite adjectives) National and international standards recommend the setting of usability objectives as a best practice (ANSI/AAMI HE-74-2001, IEC/ISO 62366). Objective goals of human performance are the most important, especially from a safety perspective, and user satisfaction goals should be considered secondary or supplemental. Guideline 6.12: Usability Objective Usability objectives should be established for use in device planning early in the design cycle as well as for setting acceptance criteria for later summative usability testing. Usability objectives should be set as part of customer requirements development and compared to competitive benchmarks (usually obtained from published studies or from comparative usability testing of best-in-class competitor’s devices). Only a few critical task-related usability objectives are necessary. For medical devices, safety is of prime importance, and usability objectives should be set on the basis of the most critical safety-related tasks in using a device.
Examples of quantitative usability objectives or goals are the following: • Ninety percent of experienced nurses will be able to insert the infusion pump tubing set on the first try with no instructions, and 100% will be able to correct any insertion errors. • Ninety percent of experienced anesthesiologists will be able to calibrate the cardiac monitor within 2 minutes with no errors. • After reading the quick reference card, 90% of experienced clinicians will be able to properly configure the display on the first try to show the two ECG lead traces. • Ninety-five percent of technicians with no prior experience with this type of IVD diagnostic system will achieve the target mastery level in 6 or fewer hours of use. • Eighty percent of experienced ICU nurses will prefer the readability of the display for the latest-generation ventilator monitor compared with the existing monitors. Typical usability objectives are stated as first-time task completion rates for minimally trained users, with a success rate in the range of 80% to 95%. It is expected that after more experience or with additional training, users will achieve completion rates approaching 100%. Highly safety-critical tasks may require 100% completion rates. 6.4.1.8 Data Collection Data collection can take many forms, including simple paper-based logging forms with stopwatches, computer logging software, video recording, or even systems that track users’ eye movements. The form of data collection will depend on the stage of development, the criticality of the device, and the project’s budget and schedule constraints. Table 6.4 provides a list of some of the dependent variables possible. 6.4.1.9 Data Analysis Data analysis can range from qualitative methods for early formative usability testing (based on detailed analysis of subject behaviors and comments) to mostly quantitative methods for summative usability testing against quantitative usability objectives and
TAF-ER562X_WINGR-09-0401-C006.indd 217
11/9/10 12:49:44 AM
218
Handbook of Human Factors in Medical Device Design
TABLE 6.4 Dependent Variables in a Usability Test Task times Task completion rates Efficiency measures (e.g., missteps/total steps) Significant errors Ratings or rankings Verbal comments (categorized) Questionnaire data (many forms from yes–no to essay) Number of references to documentation or use of help systems Video and audio records Photo images Physical measures (e.g., fatigue, force, heart rate, pupil size, galvanic skin response, respiratory rate)
acceptance criteria. When video recording is done, sometimes the tapes or equivalent media are not reviewed except for archival or backup purposes. At other times, many hours will be expended reviewing recordings and classifying behaviors and timing data for in-depth analysis. The statistical analysis may be limited to simple descriptive statistics concentrating on measures of central tendency (means, medians) and dispersion (variance, range). This may be sufficient for making general design choices. However, with more advanced designs manipulating several variables, more sophisticated statistical techniques should be used. In some situations, such as summative validation testing, it will be necessary to formally ascertain whether acceptance criteria have been met. This requires testing the statistical reliability of the results using small-sample statistical techniques as described in Section 6.5.2 and Appendix 6.C. As noted earlier, task completion rate acceptance criteria are usually set for first-time users with minimal training with the expectation of a 100% completion rate with steadystate usage. Nonetheless, in analyzing usability data, especially summative test data, it is important that all task failures be analyzed from a risk analysis perspective, including descriptions of task-failure risk mitigation strategies. This practice is highly recommended even when the task completion rate exceeds the acceptance criteria target but is less than the 100% task completion rate. Guideline 6.13: Analyze All Use Errors All use errors that occur during usability tests should be analyzed for their potential to represent any use-related risk of injury to patients, users, or bystanders.
6.4.1.10 Reporting Guideline 6.14: Usability Test Reporting The reporting of usability testing results should follow accepted practices for the writing of scientific reports, especially in the case of summative usability tests.
It is recommended that report writing checklists be used. The checklist shown in Appendix 6.B was taken from the standard for usability testing report writing, ANSI/NCITS 354-2001,
TAF-ER562X_WINGR-09-0401-C006.indd 218
11/9/10 12:49:44 AM
Testing and Evaluation
219
Common Industry Format for Usability Test Reports. For early formative tests, it may not be necessary to generate a full-format CIF report. Results and recommendations for iterative design might instead be conveyed as bulleted lists in topline reports or presentation slides.
6.5 LOGISTICS 6.5.1 TESTING LOCATIONS* Several factors will influence the decision on the number of different testing locations that will be used, including user access, cost, and feasibility. Multiple regions, cities, or countries may be used if clinical practices are known to differ significantly by geographic region. Language differences might also dictate conducting the tests in multiple countries. If representative users (e.g., burn unit nurses or pediatric orthopedic surgeons) are not available in a single location, then the tests may need to be conducted in multiple locations to avoid extensive travel by test participants. In contrast, a single central location may be sufficient when subjects are available locally and there are no regional differences in critical user attributes, customs, or practices. Often individual differences in task performance will be much larger than any geographically related differences. The final decision on test location will be dictated by a variety of considerations, including: • Availability and recruiting convenience of test participants • Existence of regional differences • The need for specialized nonportable equipment or facilities (e.g., a high-fidelity realistic use environment) • Budget and time constraints
6.5.2 NUMBER OF PARTICIPANTS Guideline 6.15: Sample Size Calculations Sample size for usability testing needs to be statistically justified, especially for summative usability tests in which the goal is to assess whether prespecified usability objectives have been met.
Usability professionals generally agree that formative usability tests require only five to eight subject participants per distinct user group. (Each distinct user group has a different user profile; e.g., doctors, nurses, and patients are distinct user groups.) This should include participants with diverse capabilities, as would be expected in the actual user population. Even with a few participants per user group, major usability issues will be uncovered. The rationale is that after five subjects are tested, major usability defects will be observed repeatedly for successive subjects with little additional usability information gained. It is becoming more common for usability testing to be conducted as series of tests, each of which can have five to eight subjects. This approach yields accumulated sample sizes of 30 or more during the formative stages of design. Appendix 6.C gives additional detail and justification for these sample size recommendations. For later-stage summative usability testing, larger sample sizes are recommended to allow meaningful statistical evaluation of the attainment of acceptance criteria. As described in Appendix 6.C, summative testing may be done with as few as 15 to 20 participants per *
See also Section 6.4.1.2.
TAF-ER562X_WINGR-09-0401-C006.indd 219
11/9/10 12:49:44 AM
220
Handbook of Human Factors in Medical Device Design
distinct user group. Appendix 6.C also provides a discussion about type I and II errors, statistical power considerations and other sample size alternatives.
6.5.3 RECRUITING ACTIVITIES Recruiting test participants can be challenging. A properly designed screening instrument, which may take the form of a checklist or structured questionnaire, is recommended. Telephone interviews using the screening questions can be useful. A quota can be set for different types of representative users from different user groups as specified in the user profiles. Investing the effort early on to create effective user profiles will facilitate recruiting because targeted screening questions can be created to match the profiles of interest. Market research companies as well as usability consultants typically have subject pool databases from which participants can be drawn. Budgets should include recruiting expenses as well as participant compensation. Other recruiting methods include cold calling, Web posting, and advertising. Participants often cancel their appointments, so it is prudent to recruit extras. It is typical to expect 10% no-shows, but this rate may be higher in some subject pools and in some regions of the world. Some clinicians in high-demand specialties may be especially hard to recruit. See Section 6.6.2 for advice on how to properly handle subject consent issues.
6.5.4 TESTING STAFF SIZE Some usability tests are simple enough that one test moderator is sufficient. These tests would include usability tests with simple tasks that are performed quickly and with minimal participant effort. However, it is common to have two or more testers. One staff member can act as a moderator (either in the same physical space as the test participant or in a nearby room communicating over a voice link or intercom with less chance of subtly influencing the participant). The second staff member could be taking notes as well as controlling the equipment and possibly greeting test participants. They might also be taking care of logistics, such as getting consent forms signed and paying the participants. If more than one facilitator is used, they need to be calibrated with regard to testing procedures and data scoring to ensure that experimenter bias is avoided. Sometimes a videographer is used to control the video cameras as the test progresses, such as if dynamic movement is expected and cannot be captured with fixed cameras. It may also be useful to have a stenographer record all remarks in real time rather than transcribing the dialogue from the video or audio recordings at a later time. Multiple judges of participants’ behavior (e.g., to classify the behavior) are sometimes advantageous or necessary, such as when judgment of the subject’s performance is complex or happens very quickly. It is usually a good idea to invite members of the development team and key managers to observe the test sessions in an adjacent acoustically isolated room, either through one-way glass (half-silvered mirrors) or via multiple video feeds and two-way audio connectivity when needed. One-way glass allows observers to see the test participants, while the test participants see only their reflection in a mirror and cannot see the observers.
6.5.5 SESSION LENGTH Session length is a function of the complexity of the tasks completed during the test session, the tolerance of the test subjects, and the durability of the testing staff, among other
TAF-ER562X_WINGR-09-0401-C006.indd 220
11/9/10 12:49:44 AM
Testing and Evaluation
221
considerations. Total session times typically range from 45 minutes to 2 hours for reasonably simple usability tests (e.g., 5 to 15 tasks) as well as pre- and postsession questionnaires and a debriefing session. Half-day to full-day test sessions may be required for complex systems. It is a wise practice to schedule time between subject sessions to provide a cushion for late starts and sessions that go overtime. This buffer time can be used to conduct a debriefing on the previous test session results or to set up for the next session as well as giving the testing staff a break.
6.5.6 VIDEO RECORDING Video recording has many advantages. While many usability tests can be conducted without video, video recording allows postsession review of complex and quick-moving action or difficult-to-assess behaviors. Video analysis facilitates reliable classifications as to errors, time, and causation of a usability defect. The video is a traceable record of the events that transpired during the test. The videos allow offline transcription of dialogue or action. Multiple observers can judge participants’ behaviors after the test session has ended. Even test participants can review the video record to comment on test session events. The video can be streamed over the Internet from a test session to remote personnel who can view the test sessions from the comfort of their offices. It is sometimes desirable to create a short highlights video for viewing by other team members and key managers to more strongly make the case for specific user-interface improvements. Digital video is the most efficient method for recording and editing. There is nothing more powerful to a reluctant member of a development team than to be humbled by the remarks of test participants as they struggle with a user-interface design and fail to complete supposedly easy tasks.
6.5.7 NOTE TAKER As noted in Section 6.5.4, it can be advantageous to have a stenographer take real-time, verbatim notes of all the relevant dialogue between the moderator and test participants. Then notes are immediately available to team members rather than their having to wait for a transcript after the sessions are completed. These notes will allow rapid reporting of key findings from the usability test with supporting verbatim comments from the test participants. The notes can be linked to the video recording to identify key parts of the session that justify more detailed analysis. Notes also allow easy identification of video segments for inclusion in an edited highlights video summary of the usability test.
6.5.8 LANGUAGE TRANSLATORS Translators are useful for usability testing in foreign countries or to assess language translation accuracy and its effect on usability. There are several ways to use translators in the usability testing process. One method is to have two translators participate in the test. One bilingual individual with training in usability testing would moderate the session. A second bilingual translator would be with the observation team to simultaneously translate the dialogue between the moderator and test participant. The observation team (typically not speaking the native language of the participants) could ask the moderator to pursue some action of the participant in more detail through the bilingual translators via private audio links.
TAF-ER562X_WINGR-09-0401-C006.indd 221
11/9/10 12:49:44 AM
222
Handbook of Human Factors in Medical Device Design
A less desirable alternative is to translate the transcript of the study afterward. The disadvantages are missing real-time opportunities to correct bias errors on the part of the bilingual moderator (who is less likely to be a trained human factors professional), to make real-time modifications in the test protocol, or to ask probing questions at key points in the test session. Another consideration is that nonnative language–speaking test participants may want to speak the target language during the test but will desire occasional translation support in order to understand more technical or advanced expressions.
6.5.9 DATA LOGGING SOFTWARE A number of excellent data logging software packages are on the market, including Observer, Observant, OVO Logger, and Observational Coding System Tools (OCS Tools). All allow the easy capture of rapidly occurring subject behaviors directly into databases and spreadsheets as the sessions are being conducted or after the fact. Some allow the realtime use of function keys to identify key activities, errors, and precategorized missteps. These coded observations are usually time-stamped by the software and synchronized with video recordings of the sessions. Some of these software applications are also available on personal digital assistants (PDAs) or portable tablet PCs for use in field situations where a laptop computer may not be as convenient to use. For simple usability tests, computer data logging may be considered a luxury because of the added expense involved.
6.5.10 SCREEN CAPTURE For user interfaces that are screen based, it can be very useful to capture the screen action on video. This can be done for computer-based displays via scan converters. The resulting video quality (e.g., resolution and display flicker) is a direct function of the cost and sophistication of the scan converter. For video records for which high-quality recordings of screen action are not critical, a direct video recording of the screen might be sufficient. When using a scan converter, it is sometimes desirable to use picture-in-picture, which combines views of the screen activity with a view of the test subject. It is also possible to capture screen activity (e.g., cursor movements, opening and resizing of windows, forms, and dialog boxes) using special software that creates a direct file of the screen activity along with an audio track into a common multimedia format such as AVI or MPEG. Screen capture is considered a luxury for most usability tests. Other software can provide a record of all button presses, keystrokes, mouse clicks, and other direct interactions with the device’s user interface. Such keystroke audit software produces a rich source of user interaction data but can be time-consuming to analyze and interpret.
6.5.11 EYE SCAN CAPTURE Currently available eye gaze tracking systems generate a wealth of usability data and may sometimes be appropriate, given the budget, resource constraints, and usability test objectives. These systems generate large quantities of data that may not always be justifiable or useful. Eye tracking systems typically use an infrared beam aimed at the test participant’s cornea. The beam is reflected and captured via an infrared sensitive video camera and synchronized with the screen being viewed or with an independent video of the area being
TAF-ER562X_WINGR-09-0401-C006.indd 222
11/9/10 12:49:44 AM
Testing and Evaluation Tobii 1750 eye-tracking system and 17” monitor (set to 1024 × 768 px)
223
Face camera
Infrared cameras (integrated into the monitor) tracking participant’s eye movements
Participant Infrared beams integrated into bottom of monitor bezel, directed at participant’s eyes
Moderator instructs the participant and controls the eye tracking software
FIGURE 6.6 Picture of the eye scanner and setup (infrared light source is at the bottom of the screen bezel on the left, and the infrared sensitive cameras are in the upper bezel of the display screen).
watched. Some systems require head-mounted equipment, while others, though not headmounted, constrain the user’s head movements. The resulting data can show a playback of the screen action with the subject’s eye scan traces superimposed. Data can be gathered on the sequences of eye gaze movement along with dwell times for different areas of the interface being viewed. Most of these eye tracking systems (Tobii, ERICA, ASL, and EyeTracking Inc.) also allow pupil size to be measured dynamically. Pupil size may indicate the degree of mental workload or fatigue for a subject. For example, larger pupil sizes (an indirect measure of sympathetic nervous system activity) indicate higher workloads. The measured eye position data are more reliable than a subject’s self-reports of where in a scene they have gazed. For example, the effectiveness of attention-grabbing icons, colors, and animations can be measured with these systems. Differential attention to text versus graphics can be assessed. Screen layouts can be optimized for navigation and organization by evaluating eye scan patterns. The disadvantages of current eye scan systems are cost, the need for frequent calibration, and the inconvenience for test participants using headmounted equipment or head movement constraints (for additional detail, see Lankford et al., 1997). Figure 6.6 shows a typical eye tracking equipment setup. Figure 6.7 shows an
FIGURE 6.7
Eye tracking trace diagram.
TAF-ER562X_WINGR-09-0401-C006.indd 223
11/9/10 12:49:44 AM
224
Handbook of Human Factors in Medical Device Design
eye movement trace diagram outlining the sequence of eye fixations. Each fixation point is represented as a circle with larger circles indicating longer dwell times. The images in this case are three drug labels, each having a different presentation style for lot and expiration number.
6.6 PROTOCOL-RELATED ACTIVITIES The following sections offer recommendations on some of the details of the usability testing protocol. This section addresses the activities associated with a typical usability test. There is a logical progression of activities for the test participants, starting with their orientation and culminating in the debriefing session.
6.6.1 PARTICIPANT ORIENTATION The first step is making the test participant feel at ease with the testing session. It is important to reassure the test participants that they are not being tested for their intelligence, knowledge, or skills; rather, the usability test is an evaluation of the quality of the user interface of the medical device. The moderator should point out that the design is being tested for flaws and that participants’ candid input is important. Also, the orientation should include discussion of the logistics of the test, signing of consent and nondisclosure forms, and clarification of participant compensation.
6.6.2 CONSENT FORMS Guideline 6.16: Consent Documentation Test participants must be informed of the type of data being collected, the procedures, and the purpose of the test. A written consent document should be completed by every participant.
The consent form should give all the details necessary for an individual to make an informed decision about whether to participate in the study. Issues that must be addressed include the use of video recording, the presence of observers, and participants’ right to terminate the evaluation session at any time. The consent form should disclose the objectives of the study without influencing the participant’s behavior during the study. Some institutions may require that an Institutional Review Board (or IRB) review and approve usability testing protocols and related consent forms, especially if the test participants might be placed at some level of risk to their health or well-being (e.g., being exposed to biohazardous materials). Test participants need to be assured that any personal information collected will be kept confidential and private as required by the Health Insurance Portability and Accountability Act in the United States and other international privacy protection laws.
6.6.3 NONDISCLOSURE FORMS Test participants may be asked to sign a form stating that any company proprietary information on designs and intellectual property should not be disclosed to anyone outside the testing environment. Participants are often asked to not discuss the study with colleagues who might be future test participants in order to avoid influencing their behavior.
TAF-ER562X_WINGR-09-0401-C006.indd 224
11/9/10 12:49:45 AM
Testing and Evaluation
225
6.6.4 PRETEST QUESTIONNAIRE A pretest questionnaire is often used to obtain background information about participants to aid in understanding their subsequent performance and opinions. The questionnaire can also be used to validate that the screening instrument was used correctly by the recruiters and that participants match the desired user profile. These pretest questionnaires typically ask participants about their previous device experience, training, education, licenses, or relevant medical conditions.
6.6.5 PARTICIPANT TRAINING PRIOR TO TESTING The type of pretest training given to a test participant is a key decision during protocol design. One philosophy is to mimic worst-case clinical conditions in which a new user is asked to use a medical device without the benefit of “in-service” training. Such training, while desirable, is not always practical. For example, on a short-staffed hospital floor, registry (per diem) nurses might be assigned during an overnight shift without the availability of previously trained staff to answer any questions they might have about the operation of newly installed medical devices. In this type of “out-of-the-box” testing situation, test participants receive no training, and the self-evidency or intuitiveness of the design is evaluated. A second test session might be conducted after appropriate training is administered to the same or a different group of subjects. Sometimes the training program itself may be subjected to a usability test. In other situations, a device cannot be operated without training. Formal certification may even be required before unsupervised use is allowed. For example, the placement of carotid stents can be done only by trained and certified physicians. In this situation, the usability test needs to be preceded by device use training that is representative of that to be used in the real clinical environment. In either case (training or no training), the participant’s pretest treatment must be done consistently to remove any confounding due to differential knowledge prior to performing any test tasks. Having the trainer use a script or providing a video recording of the training material can achieve consistency. In most cases, it is advisable to give participants some practice exercises so that they are comfortable with the testing protocol.
6.6.6 DIRECTED TASKS 6.6.6.1 Think-Aloud Protocol The most common form of usability testing asks participants to perform critical task scenarios. Often participants are given the tasks in written form and are asked to read and then perform the tasks while their performance is being observed, measured, and recorded. Sessions are typically one-on-one with the moderator. The think-aloud, or directed dialogue, technique is recommended for exploratory tasks in which the participants describe their thoughts and logic aloud to the moderator. Some participants will find it difficult to perform complex or demanding tasks and describe what they are thinking at the same time. In these situations, participants could complete the task and then immediately describe their logical thought process, including what aspects of the user interface confused them. At first, some participants will need
TAF-ER562X_WINGR-09-0401-C006.indd 225
11/9/10 12:49:45 AM
226
Handbook of Human Factors in Medical Device Design
continual reminders to “think aloud.” While the think-aloud protocol provides a rich source of information, the research team must be careful in interpreting individual comments since subjects can be unintentionally biased by prior experiences, task progress, and task outcome. An alternative is to have participants watch a video recording of their task performance and describe their thoughts. In this method, participants will be subject to memory loss, memory reshaping, and hindsight bias and potentially overlook some important details regarding usability issues they experienced. For example, a cardiologist using an exchange catheter may forget the difficulties experienced while redirecting the path of the wire in some unusual patient arterial anatomy. If time to perform a task is being measured and is part of a usability objective, then sufficient sample size will be required for a split study in which a subset of participants do not think aloud while performing the task because the talking process will add time to their tasks. Split-study designs can be used with different participant groups. Alternatively, similar tasks can be given to the same participants with nonredundant task parameters (e.g., medical orders with different values). One set of tasks can be timed without talking aloud, while a second set of similar but nonredundant tasks can be performed using the think-aloud protocol. Learning effects must be taken into account, and counterbalanced presentation of task order may be required, as described in Section 6.4.1.6.2. 6.6.6.2 Codiscovery Two-person teams, or codiscovery, is a less frequently used option. Here, two participants complete tasks as a team and talk to each other, explaining their problems and observations on the user interface. This technique may be more natural for some test participants but makes the logistical planning of the test more difficult because of the added effort of scheduling two participants together. In some instances, designers are interested in whether individuals can perform a task unassisted using the medical device, as a use environment analysis would show. In these cases, two-person teams would not be a good choice. On the other hand, for systems that are frequently used concurrently by several people, codiscovery may be more informative, such as in the evaluation of adjustment tools used in spinal implants performed by a surgical team. When such dyads or teams are being tested, the participants should accurately reflect the roles and experience of intended user teams. 6.6.6.3 Self-Exploration A great deal can be learned from having participants explore a user interface without specific direction to demonstrate success or failure while attempting to complete important tasks. The test moderator will give them a brief overview and then observe as they explore the device and its features. The features with the greatest degree of self-evidency or intuitiveness will emerge, and test participants typically will not fully explore more difficult features. Use of the think-aloud protocol allows the participants to explain their initial usability problems. This technique is complementary to more directed task formative usability testing and should be done at the start of a session. Test participants should be given 15 minutes or so to conduct such undirected explorations.
TAF-ER562X_WINGR-09-0401-C006.indd 226
11/9/10 12:49:45 AM
Testing and Evaluation
227
6.6.7 INTERVIEWS Interviews are either structured with predefined questions or unstructured with only general discussion guidelines. They are typically conducted on conclusion of the directed tasks portion of the usability test. Interview questions should offer participants a chance to explore in more depth usability problems that occurred during the directed tasks portion of the protocol and to ask about overall ratings and opinions. Sometimes, satisfaction or acceptance ratings can be solicited as part of the interview or, alternatively, asked as part of a posttest questionnaire.
6.6.8 POSTTEST QUESTIONNAIRE Questionnaires can be either structured or unstructured. Structured questionnaires have a sequence of directed questions and typically include rating scales or ranking questions about device features and usability dimensions. Rating questions could be completed all at one time at the end of the session. Alternatively, the questions can be posed at the completion of each task as long as these posttask questionnaires are short and do not disrupt the testing session’s flow. Below are examples of various types of question styles, including Likert scales, semantic differential, and ranking with point allocation. Survey question design is a specialized area. Question wording can substantially influence the results. See Fowler (1995) and Rea and Parker (1997) for details on issues such as number of points on a scale, use of odd versus even numbers of rating choices, numeric versus categoric, question order, scale direction, and labeling, among many other considerations in designing survey questions. 6.6.8.1 Examples of Questionnaire Items 1. Simple rating scale On a scale of 1 to 10, where 1 is the worst possible rating and 10 is the best possible rating, how would you rate the medication experience provided by this kit? 2. Likert rating scales A) How would you rate the overall usefulness of this device (check one)? Not Useful at All
Somewhat Not Useful
Neutral
Somewhat Useful
Very Useful
B) Please rate the following aspects of the device on a scale from 1 (very difficult) to 5 (very easy):
Very Difficult 1
Somewhat Difficult 2
Neither Difficult or Somewhat Easy Easy Very Easy 3 4 5
Turning on Correcting an input value Calibrating
TAF-ER562X_WINGR-09-0401-C006.indd 227
11/9/10 12:49:45 AM
228
Handbook of Human Factors in Medical Device Design
3. Semantic differential scales. Please indicate your assessment of the device on the following scales by marking the appropriate box with an X. Easy to use Efficient
Hard to use Inefficient
Convenient
Inconvenient
Acceptable
Unacceptable
4. Ranking with interval measures We would like you to rate your satisfaction with the three different medication injection kits that you experienced. Please rate each of the three devices on a scale of 0 to 100, where the total (sum) of the three scores equals 100. For example: A = 60, B = 30, and C = 10. (In this example, A is rated twice as good as B and six times as good as C.) OVERALL ratings (taking into account all factors) Injection Kit A: ___ Injection Kit B: ___ Injection Kit C: ___ Total = 100
6.6.9 DEBRIEFING Guideline 6.17: Debriefing Each study participant shall be debriefed at the end of each test session.
Debriefing can also be used to learn about better ways to recruit test subjects, conduct test sessions, and clarify uncertainties about the usability test purposes. Additional feedback (informal comments) about the design is often obtained during debriefing. The following points should be addressed in debriefing: • Make sure the test participants are not unhappy or upset. • Ensure that participants do not feel like they are personal failures if they struggled using the device(s). • Reinforce that they found design failures. • Make sure that they feel appreciated and successful in helping you find possible design problems.
6.6.10 RECORD KEEPING Usability testing data in all its forms should be recorded and kept in secure files. Accurate record keeping and archiving are important for regulatory compliance as well as to facilitate future access (e.g., when the next version of the device is being designed).
TAF-ER562X_WINGR-09-0401-C006.indd 228
11/9/10 12:49:45 AM
Testing and Evaluation
229
6.6.11 TESTING TEAM DEBRIEFING After each test session, the testing team members should compare notes and observations. If this is not feasible because of tight scheduling, at least debrief at the end of the day. This allows any ambiguities to be clarified since different observers may have interpreted the test participants’ behaviors from different perspectives. Debriefing also allows the team to consider adjustments to the testing protocol as new data are collected.
6.7 SOURCES OF TEST BIAS There are many potential sources of bias in conducting a usability test. It is tempting to jump in and assist test participants with task completion, especially if they ask the moderator for hints or help. It takes considerable discipline and experience to avoid influencing the results. Some considerations to reduce bias include the following: • Do not lead or prompt participants. Make only neutral comments and do not speculate on what the user is thinking. For example, ask “What are you thinking?” instead of “Are you considering pushing the BACK button because you feel that will clear the display?” Or say “How are you feeling about that action?” rather than “Do you feel frustrated?” If the test participant asks “Can I do that?” you should respond “What do you think would happen if you did that?” • Watch body language and facial expressions as sources of bias if the moderator and participant are in same room during testing. • Talk only as needed to clarify and run the evaluation. Try not to interrupt. Let the participants complete tasks themselves. • If participants are lost, end the task. (A 5-minute maximum time per task is typical, but it could be longer, depending on the task.) • Only if a task is incomplete (and the next task depends on the current task) should you show the user what to do. Even then, deliberately hesitate before assisting to allow the participant to give more insight.
6.7.1 COMMON TESTING MISTAKES The following is a short list of common mistakes made during usability testing: • Not pretesting the protocol (Pretesting, or a “dry run,” can help remove ambiguous task statements and participant instructions and can ensure that the logistics work.) • Leading or biasing the participant • Helping the participant complete tasks • Talking too much or not watching carefully • Rushing the participant • Making the participant feel inferior or inadequate • Not making sessions friendly and interactive • Not keeping track of time and having to rush to finish all tasks in the protocol
6.8 SUPPLEMENTAL USABILITY EVALUATION METHODS Other human factors methods can supplement usability testing. These are sometimes called usability inspection methods (for details, see Nielsen and Mack, 1994) and are
TAF-ER562X_WINGR-09-0401-C006.indd 229
11/9/10 12:49:45 AM
230
Handbook of Human Factors in Medical Device Design
complementary to empirical measures of usability. Usability inspection methods provide valuable input to user error–related risk analysis. Because these methods are analytical, they may miss usability problems that can be discovered only when actual users use a medical device under realistic conditions. These inspection methods are sometimes the only practical way to comprehensively evaluate all the tasks that could be performed with a complex medical device. Formal usability tests would cover the most frequent and critical tasks, while the remaining tasks are evaluated using one or more of the inspection methods. Several types of usability inspection methods are described briefly below (more detail can be found in Nielsen, 1994).
6.8.1 COGNITIVE WALK-THROUGH Cognitive walk-throughs employ a structured review of user requirements associated with the performance of a sequence of predefined tasks. A trained human factors professional leads a multidisciplinary design team through the critical user tasks while noting where usability problems might occur and then recommending mitigations. Team members usually include development, medical, marketing, and quality assurance personnel. The walkthrough may involve examining a very-low-fidelity prototype of the system, such as a paper prototype or a storyboard. If the team includes a user representative, then this method is called a pluralistic walk-through.
6.8.2 EXPERT REVIEW With this method, device usability is evaluated by HFE specialists to identify design strengths and weaknesses and to recommend improvements. Expert reviews are based on applying knowledge of previous research in the field of human factors and ergonomics to the design. Expert reviews may involve only one or two human factors experts.
6.8.3 HEURISTIC REVIEW Clinical or human factors experts evaluate a device or system through the assessment of how it conforms to well-established user-interface design rules or heuristic guidelines. A heuristic review is a more formal process that requires multiple experts who develop a consensus opinion about design characteristics. Sources for these heuristics include this book and the ANSI/AAMI HE-75-2008 standard.
TAF-ER562X_WINGR-09-0401-C006.indd 230
11/9/10 12:49:45 AM
Testing and Evaluation
231
APPENDIX 6.A: EXAMPLES OF USABILITY TEST PLANS CASE REPORT 1: A SUMMATIVE USABILITY TEST PLAN EXAMPLE FOR A DIAGNOSTIC SAMPLE PREPARATION SYSTEM Project Overview Introduction The Diagnostic Sample Preparation System is an automated laboratory diagnostic instrument that uses general-purpose reagents to prepare dynamic samples for further analysis. Major components include the Diagnostic Sample Preparation System worktable, internal handling arm, robotic manipulator arm, liquid subsystem, bar code reader, racks, and carriers. The user sets up and runs protocols using the Dynamic Testing (DT) Wizard software application residing on a personal computer that is connected to the Diagnostic Sample Preparation System worktable. Objectives The objectives of the Diagnostic Sample Preparation System usability test are as follows: • Assess the performance of both trained and untrained users when carrying out basic system operations. • Assess the performance of trained operators when carrying out intermediate-level system operations, including daily maintenance tasks. • Identify critical and noncritical usability issues that affect user performance and satisfaction. Assign a priority for the resolution of each usability issue identified. • Obtain subjective user ratings with regard to ease of use and user satisfaction. • Identify top user likes and dislikes. • Document changes recommended by participants. Acceptance Criteria The following are the usability objectives that will serve as acceptance criteria for this summative usability test: • Eighty percent of the untrained participants will be able to successfully set up and initiate a DT protocol (basic tasks). • Ninety percent of the trained participants will be able to successfully set up and initiate a DT protocol (all basic and intermediate-level tasks) and to perform all daily maintenance tasks with no additional training. • Eighty percent of all participants will give the Diagnostic Sample Preparation System a rating of 5 or higher (on a 7-point scale) with respect to their overall satisfaction with the device. Method Participants A minimum of 35 laboratory practitioners will participate in the test. Each participant will be assigned to one of two groups hereafter referred to as Group 1 and Group 2. Participants in both groups will be medical technologists or technicians (or equivalent). At least 75% will have knowledge/experience with molecular diagnostics technologies.
TAF-ER562X_WINGR-09-0401-C006.indd 231
11/9/10 12:49:46 AM
232
Handbook of Human Factors in Medical Device Design
• Group 1 will consist of a minimum of 20 participants (25 will be scheduled). • Group 2 will consist of a minimum of 15 participants (20 will be scheduled). Makeup sessions will be scheduled, if necessary, to obtain these minimum numbers. Group 1 (Untrained Group) Although participants in this group will not be trained to use the Diagnostic Sample Preparation System, they will be shown a video describing the basic features of the device and its operation. Group 1 participants are intended to represent potential users with no prior experience with the Diagnostic Sample Preparation System. Group 2 (Trained Group) Participants in Group 2 will each receive a minimum of 4 hours of training on the operation and maintenance of the Diagnostic Sample Preparation System by a qualified company trainer prior to being studied. Equipment and Materials Equipment and Software The test site will have an installed Diagnostic Sample Preparation System Worktable and System Control Center. Installed software will include the DT Wizard, the application that operators use to set up and run test protocols and to perform routine maintenance tasks. The investigator will record the serial number for the Diagnostic Sample Preparation System Worktable and the version number for the DT Wizard software (if available). Simulated Samples, Controls, and Reagents Patient samples, controls, and reagents will be simulated using water, colored water, or a buffer solution. Hence, the normal procedures for safeguarding patient identity, handling of investigational reagents, distribution and control of investigational products, and materials accountability are not required. Use of simulated samples, controls, and reagents is acceptable because of the following: • The test is intended to evaluate the user interface for the Diagnostic Sample Preparation System. The suitability of the Diagnostic Sample Preparation System in performing its intended functions will not be evaluated. • The primary focus of the investigation will be protocol setup; protocols will not be run to completion, and no test results will be produced. Although participants in this study will not be exposed to biohazards, they will use appropriate personal protective equipment (e.g., lab coats, safety glasses or shields, gloves, and so on) and follow established laboratory practices for the handling of patient samples and other potentially hazardous materials. Instructional Materials • Participants in Group 1 (the untrained operators) will be shown a video that provides an overview of the Diagnostic Sample Preparation System and its basic operation. No other instructional materials will be provided.
TAF-ER562X_WINGR-09-0401-C006.indd 232
11/9/10 12:49:46 AM
Testing and Evaluation
233
• Participants in Group 2 (the trained group) will have access to the current version of the Diagnostic Sample Preparation System Operations Manual and will receive Diagnostic Sample Preparation System operator training equivalent to that being planned for actual customers. No other instructional materials will be provided. Video Recording All test sessions will be recorded on videotape unless such recordings are not feasible. Procedures Group 1 (Untrained Users) Participants will be tested individually in separate sessions at a company facility. The duration of each session is expected to be about 2 hours. Each session will begin with a general introduction. The participant will then be shown a videotape covering the primary components and subsystems of the Diagnostic Sample Preparation System and its basic operation. System maintenance and recovery from error conditions will not be covered. Following the video overview, the participant will be asked to perform a series of basic operations tasks in the order in which they appear on a prepared task list (see Table 6.A1, column 2). If the participant cannot complete a given task within 5 minutes or indicates an inability to proceed without assistance (after having made a reasonable attempt), the effort will be recorded as a task failure. The participant will then be given the necessary assistance to complete the task if that task must be successfully completed in order to advance the dialog to the next task. The following will also be done: • Participants will sign a consent form acknowledging that they will be observed and recorded and that they can discontinue the test at any time. • All user errors will be recorded and classified according to their nature, frequency, and severity. • Task times will be recorded. • At the end of the session, participants will be asked to complete a questionnaire to assess their perceptions of the usability of the device. • Participants will be debriefed at the conclusion of the testing session. Group 2 (Trained Users) Each participant in Group 2 will receive approximately 4 hours of training on the operation and maintenance of the Diagnostic Sample Preparation System from a qualified company training instructor prior to the test session. Participants will be tested individually in separate sessions at a company facility. The duration of each session is expected to be about 2.5 hours. Each session will begin with a general introduction. The participant will then be asked to perform a series of tasks in the order in which they appear on a prepared task list (see Table 6.A1, column 3). The tasks will cover basic and intermediate-level system operation and daily maintenance tasks. If the participant cannot complete a given task within 5 minutes or indicates an inability to proceed without assistance (after having made a reasonable attempt), the effort will be recorded as a task failure. In these instances, the participant will be given an additional 2 minutes to consult the Diagnostic Sample Preparation System
TAF-ER562X_WINGR-09-0401-C006.indd 233
11/9/10 12:49:46 AM
234
Handbook of Human Factors in Medical Device Design
TABLE 6.A1 Diagnostic Sample Preparation System Usability Test Task List Assigned to Tasks Component and Subsystem Identification 1. Identify the sample rack, reagent troughs, PosID, output rack, TIPs, solid and liquid waste containers, 1-ml subsystem rack, power on/off switch, and reservoir for system liquid. 2. Identify the LiHa, rack and reagent bar codes, wash/waste station, safety panel and door, 1-ml subsystem zones, TIP tray, TIP reuse rack, TIP deep-well plate, RoMa, valve connections, syringe screw, and plunger lock screw. 3. Discuss the numbering scheme for the placement of racks and carriers. Start-Up Tasks 4. Start the Diagnostic Sample Preparation System software application and comment on the main menu. 5. Check instrument status. 6. View the status of the last completed run. 7. View and print the latest protocol report. 8. Flush the system. 9. Perform instrument startup tasks. 10. View/print partial log report. 11. Select protocol for run and begin run setup. Preparation of Samples and Controls 12. Load bar-coded sample tubes and controls in Sample Rack 1 and place on the Diagnostic Sample Preparation System worktable. 13. Use PosID to scan the sample tubes and manually enter sample and/or control IDs where necessary. Worktable Preparation 14. Print the reagent report and discuss it. 15. Check system liquid container and all waste containers. 16. Load TIPs. 17. Put reagent troughs in reagent carrier and place the carrier on the Diagnostic Sample Preparation System worktable. 18. Load output tubes in Output Rack 1. 19. Discuss numbering and labeling of output tubes. 20. Place Output Rack 1 on Diagnostic Sample Preparation System worktable. 21. Load reaction vessels. 22. Load the TIP tray, reuse rack, and deep-well plate. 23. Navigate through the remaining screens and start the run. 24. View the Process Screen and comment on it. 25. Enter a user comment in the appropriate text box. 26. Abort the run and return to the main menu. Remove Racks and Other Components 27. Remove Output Rack 1 from the Diagnostic Sample Preparation System worktable and unload the output tubes.
TAF-ER562X_WINGR-09-0401-C006.indd 234
Group 1
X
Group 2
X
X X
X X X X X
X
X X X X X X X X
X
X
X
X
X X
X X X
X X
X X X X X X X X X X
X
X X X
X
11/9/10 12:49:46 AM
Testing and Evaluation
235
TABLE 6.A1 (CONTINUED) Diagnostic Sample Preparation System Usability Test Task List Assigned to Tasks
Group 1
Group 2
28. Remove Sample Rack 1 from the Diagnostic Sample Preparation System worktable and unload the sample tubes. 29. Remove the TIP tray, reuse rack, and deep-well plate.
X X
Daily Maintenance Tasks 30. Check/tighten valve connections, syringe screws, and plunger lock screws. 31. Clean/tighten TIP cones. 32. Describe step-by-step how to refill the system liquid container, empty the liquid waste container, clean the worktable, and clean the waste station. Usability Questionnaire 33. Complete usability questionnaire.
X X X
X
X
Operations Manual and make a second attempt to complete the task. If the second attempt is also unsuccessful, the participant will be given the necessary assistance to complete the task if that task must be successfully completed in order to advance the dialog to the next task. The following will also be done: • Participants will sign a consent form acknowledging that they will be observed and recorded and that they can discontinue the test at any time. • All user errors will be recorded and classified according to their nature, frequency, and severity. • Task times will be recorded. • At the end of the session, the participants will be asked to complete a questionnaire to assess their perceptions of the usability of the device. • Participants will be debriefed at the conclusion of the testing session. Instrument Mode The Diagnostic Sample Preparation System will normally be used in the “R” (run), mode. However, it may be used in the “T” (test), mode, if necessary. Measurements The following observations/measurements will be obtained during each session: 1. Task outcome (completion or failure) for each task attempted 2. Total time required to set up and initiate a run and for each task 3. All use errors and missteps committed by the users. 4. Tasks requiring reference to the Operations Manual (Group 2 only) 5. Tasks requiring assistance from the investigator 6. Observed and participant-reported usability issues 7. Subjective ratings for instrument and software usability (from the questionnaire)
TAF-ER562X_WINGR-09-0401-C006.indd 235
11/9/10 12:49:46 AM
236
Handbook of Human Factors in Medical Device Design
Data Analysis All numeric (continuous) data will be presented as mean, range, and 95% confidence intervals. Incidence data (e.g., errors per task) will be reported as percentage of possible opportunities. Qualitative data (e.g., descriptions of usability issues) will be analyzed for patterns and practical significance. Project Team Review A formal review of the results of this study will be required. The review team will be appointed by the project manager. After the review, action items will be assigned to appropriate parties. Configuration Management Configuration control will be maintained throughout the study. Record Retention All records will be deposited or referenced in the Project History File. Plan Revisions Any revision to this usability test plan requires the approval of the same areas approving the previous version.
TAF-ER562X_WINGR-09-0401-C006.indd 236
11/9/10 12:49:46 AM
Testing and Evaluation
237
APPENDIX 6.B: USABILITY TEST REPORT CHECKLIST Table 6.B1 is a usability test report writing checklist that is reproduced from ANSI/NCITS 354-2001, Common Industry Format for Usability Test Reports. Industry experts formulated this checklist. Use of this checklist would facilitate the creation of a complete and comprehensive report. It is not necessary for every element of the checklist to be used, but it would be good practice to review each element and to have a justification for exclusion of any elements. TABLE 6.B1 Usability Test Report Checklist Use the following checklist to ensure that required elements (•) are addressed in your usability test report. Recommended items are denoted by –. Title page – Company logo or name • Identify report as Common Industry Format for Usability Test Report v2.0 • Name the device and version that was tested • Who led the test • When the test was conducted • Date the report was prepared • Who prepared the report • Customer company name • Customer company contact person • Contact name(s) for questions and/or clarifications • Enter phone number • Enter e-mail address • Enter mailing or postal address Executive summary • Start on new page; end with page break • Provide a high-level overview of the test • Name and describe the device • Summary of method(s), including number and type of participants and tasks • Results expressed as mean scores or other suitable measure of central tendency – Reason for and purpose of the test – Tabular summary of performance results – If differences are claimed, the associated statistical probability Introduction Full product description • Formal device name and release or version • Describe what parts of the device were evaluated • The user population for which the device is intended – Any groups with special needs – Brief description of the environment in which it should be used – The type of user work that is supported by the device Test objectives • State the objectives for the test and any areas of specific interest • Functions and components with which the user directly and indirectly interacted – Reason for focusing on a device subset continued
TAF-ER562X_WINGR-09-0401-C006.indd 237
11/9/10 12:49:46 AM
238
Handbook of Human Factors in Medical Device Design
TABLE 6.B1 (CONTINUED) Usability Test Report Checklist Method Participants • The total number of participants tested • Segmentation of user groups tested, if more than one • Key characteristics and capabilities of user group • How participants were selected and whether they had the essential characteristics • Differences between the participant sample and the user population • Table of participant (row) characteristics (columns) – Description of groups with special needs Context of device use in the test • Any known differences between the evaluated context and the expected context of use Tasks • Describe the task scenarios for testing • Explain why these tasks were selected • Describe the source of these tasks • Include any task data given to the participants • Completion or performance criteria previously established for each task Test facility – Describe the setting and type of space in which the evaluation was conducted – Detail any relevant features or circumstances that could affect the results Participant’s computing environment • Computer configuration, including model, operating system version, required libraries or settings • If used, browser name and version; relevant plug-in names and versions Display devices • If screen based, describe screen size, resolution, and color setting • If print based, the media size and print resolution • If visual interface elements can vary in size, specify the size(s) used in the test Audio devices – If used, specify relevant settings or values for the audio bits, volume, and so on Manual input devices – If used, specify the make and model of devices used in the test Test administrator tools • If a questionnaire was used, describe or specify it – Describe any hardware or software used to control the test or to record data Experimental design • Describe the logical design of the test • Define independent variables and control variables • Describe the measures (dependent variables) for which data were recorded Procedure • Operational definitions of measures • Operational definitions of independent variables or control variables • Time limits on tasks • Policies and procedures for interaction between tester(s) and participants – Sequence of events from greeting the participants to dismissing them – Non-disclosure agreements, consent form completion, warm-ups, pretask training, and debriefing – Verify that the participants knew and understood their rights as human subjects – Specify steps followed to execute the test sessions and record data – Number and roles of people who interacted with the participants during the test session
TAF-ER562X_WINGR-09-0401-C006.indd 238
11/9/10 12:49:46 AM
Testing and Evaluation
239
TABLE 6.B1 (CONTINUED) Usability Test Report Checklist – Specify if other individuals were present in the test environment – State whether participants were paid (and how much) Participant general instructions • Instructions given to the participants (here or in appendix) • Instructions on how participants were to interact with any other persons present Participant task instructions • Task instruction summary Usability metrics • Metrics for effectiveness • Metrics for efficiency • Metrics for satisfaction Data analysis • Data scoring • Data reduction • Data management and storage procedures • Statistical analyses performed including power calculations as appropriate Results Performance results • Tabular performance results per task or task group • Summary table(s) of performance results across all tasks • Graphical presentation of performance results • Performance results: • Statistical analysis results, as appropriate Satisfaction results • Tabular satisfaction results • Summary table(s) of satisfaction results • Graphical presentation of satisfaction results Appendices • Custom questionnaires, if used • Participant general instructions • Participant task instructions – Release notes
TAF-ER562X_WINGR-09-0401-C006.indd 239
11/9/10 12:49:46 AM
240
Handbook of Human Factors in Medical Device Design
APPENDIX 6.C: STATISTICAL JUSTIFICATION FOR SAMPLE SIZES IN USABILITY TESTS NUMBER OF PARTICIPANTS This appendix provides additional justification for the usability testing sample sizes recommended in Section 6.5.2, “Number of Test Participants.” Formative Usability Test Sample Size Even with only a small number of subjects, major usability issues will be uncovered. Formative usability tests may require only five to eight participants per homogeneous user group (Virzi, 1992). After as few as five subjects are tested, the same major usability defects will be observed repeatedly for successive subjects, thus providing limited additional usability information for iterative design improvement. Figure 6.C1 illustrates how few subjects are needed in a formative usability test for exploring user-interface design concepts and early prototypes. Thus, for usability defects individually having a 25% chance of being observed in a single usability test participant, after five to eight participants studied, the cumulative percentage of usability defects found will be between 75% and 90%. Formative usability testing is mostly a qualitative endeavor, and results are not typically statistically analyzed. All usability problems uncovered during these tests need to be thoroughly investigated for their cause and their impact on device risk and safety. To further illustrate the ability of small samples to pick up usability defects, Table 6.C1 shows the cumulative probability of a usability defect being detected in a usability test given the underlying probability that a single test would show a particular problem. The table applies for all kinds of formative tests and participant populations. For example, if the underlying usability problem probability was 0.25, then with six test subjects the cumulative probability of detection is 0.82, showing that many usability defects can be discovered with small sample sizes. This table was generated with the following formula: R = 1 – (1 – P)n where n = number of subjects, P = probability of a single test showing a usability problem or defect, and R = cumulative probability of detecting a usability problem.
Percentage of defects found
100 90 80 70 60 50 40 30 20 10 0
1
2
3
4
5
6
7
8
9
10
11
Number of test subjects
FIGURE 6.C1
Test subjects needed in a formative usability test.
TAF-ER562X_WINGR-09-0401-C006.indd 240
11/9/10 12:49:46 AM
Testing and Evaluation
241
TABLE 6.C1 Cumulative Probability of Detecting a Usability Problem Usability Defect Probabilitya 0.05 0.10 0.15 0.25 0.50 0.75 0.90 a
Number of Test Subjects 1
2
3
4
5
6
7
8
9
10
25
50
75
0.05 0.10 0.15 0.25 0.50 0.75 0.90
0.10 0.19 0.28 0.44 0.75 0.94 0.99
0.14 0.27 0.39 0.58 0.88 0.98 1.00
0.19 0.34 0.48 0.68 0.94 1.00 1.00
0.23 0.41 0.56 0.76 0.97 1.00 1.00
0.26 0.47 0.62 0.82 0.98 1.00 1.00
0.30 0.52 0.68 0.87 0.99 1.00 1.00
0.34 0.57 0.73 0.90 1.00 1.00 1.00
0.37 0.61 0.77 0.92 1.00 1.00 1.00
0.40 0.65 0.80 0.94 1.00 1.00 1.00
0.72 0.93 0.98 1.00 1.00 1.00 1.00
0.92 0.99 1.00 1.00 1.00 1.00 1.00
0.98 1.00 1.00 1.00 1.00 1.00 1.00
Chance of a single test subject showing a usability problem or defect.
Other considerations when choosing sampling strategies and sample sizes include the following: • Having face validity: Will the data and resulting recommendations seem reasonable and justifiable to any skeptical development team members or others who are evaluating the results? • Statistical validity: For summative usability testing against quantitative acceptance criteria, statistical validity and reliability need to be considered. This topic is addressed in the next section. • Geographic representation: This is important if there are known regional differences in usage patterns or clinical practices. Regional differences in a nation are usually small and do not require distributed sampling. However, usage differences may be greater across national borders, necessitating multiple sample groups when testing a device intended for multinational use. • Homogeneous versus heterogeneous populations: Usability tests often require practical compromises in sample selection. Stratified, truly random sampling plans that include gender, age, experience levels, education, skill differences, and so on are rarely practical. Thus, typically a mixed or heterogeneous sample is chosen that contains some mix of user types without rigorous compliance with quota-based stratified sampling strategies. • Distinct user profiles: When the user profiles require distinct user groups (e.g., different native languages or doctors versus nurses), the usability test sampling plan must include these distinct groups. The sample size recommendations in this section would then apply to each distinct user group. Sampling strategies and their implications are complex; for more details, the reader should refer to textbooks on multivariate statistics, experimental design, or clinical trials. Ultimately, decisions about number and size of test groups and sampling should be based on a thorough understanding of the device user’s needs and relevant population attributes
TAF-ER562X_WINGR-09-0401-C006.indd 241
11/9/10 12:49:46 AM
242
Handbook of Human Factors in Medical Device Design
and on use error–related risk analysis (i.e., the risk of specific use errors across different user profile dimensions, such as experience, visual acuity, fine motor skill, and so on). Summative Usability Test Sample Sizes For later-stage summative usability testing, larger sample sizes are required so that statistical tests may be performed. There is some controversy among human factors professionals about sample size. However, through the use of exact statistical tests, such as the exact binomial test, summative testing can be done with as few as 15 to 20 participants per distinct user group. Table 6.C2 illustrates sampling plans and acceptance levels for hypothesis
TABLE 6.C2 Sampling Plans for Summative Usability Tests Probability of Acceptance if True Population Passing Rate Isb
Sampling Plana
Calculated Values
% Passing Upper Type I Objective N Accept Reject 100% 99% 97% 95% 90% 75% 50% 25% 10% 95% CLc Errord 97% 97% 97% 95% 95% 95% 90% 90% 90% 85% 85% 85% 80% 80% 80% 75% 75% 75% 70% 70% 70% a
b
c d
10 15 20 10 15 20 10 15 20 10 15 20 10 15 20 10 15 20 10 15 20
1 2 2 2 2 3 3 4 4 4 5 6 4 6 7 5 7 8 5 8 10
2 3 3 3 3 4 4 5 5 5 6 7 5 7 8 6 8 9 6 9 11
1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
0.97 0.99 0.98 1.00 0.99 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
0.91 0.96 0.92 0.99 0.96 0.98 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
0.74 0.82 0.68 0.93 0.82 0.87 0.99 0.99 0.96 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
0.24 0.24 0.09 0.53 0.24 0.23 0.78 0.69 0.41 0.92 0.85 0.79 0.92 0.94 0.90 0.98 0.98 0.96 0.98 1.00 1.00
0.01 0.00 0.00 0.05 0.00 0.00 0.17 0.06 0.01 0.38 0.15 0.06 0.38 0.30 0.13 0.62 0.50 0.25 0.62 0.70 0.59
0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.02 0.00 0.00 0.02 0.00 0.00 0.08 0.02 0.00 0.08 0.06 0.01
0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
99.5% 97.6% 98.2% 96.3% 97.6% 95.8% 91.3% 90.3% 92.9% 85.0% 85.8% 86.0% 85.0% 80.9% 82.3% 77.8% 75.6% 78.3% 77.8% 70.0% 69.8%
0.03 0.01 0.02 0.01 0.04 0.02 0.01 0.01 0.04 0.01 0.02 0.02 0.03 0.02 0.03 0.02 0.02 0.04 0.05 0.02 0.02
Accept means the null hypothesis is accepted when the number shown under Accept is the number of recorded failures. Reject is the number of failures required to reject the null hypothesis. Type II error of falsely accepting the null hypothesis for true population values less than the % Passing Objective. Upper one-sided 95% confidence level for % Passing if the number of failures equals the Accept number. Probability of falsely rejecting the null hypothesis of being better than the % Passing Objective.
TAF-ER562X_WINGR-09-0401-C006.indd 242
11/9/10 12:49:47 AM
Testing and Evaluation
243
testing of task completion rates using the exact binomial test, which is appropriate for small sample statistical testing. An example follows: • The objective is at least 90% completing a task the first time with no instructions. • Hypothesis testing. Ho (null hypothesis): pass rate ≥90%. Ha (alternative hypothesis): pass rate <90%. In Table 6.C2, using a sample size of 15 and a target task completion rate of 90%, the usability test will be considered acceptable if 4 or fewer do not complete the task successfully. The upper 95% confidence limit is 90.3% when 11 of 15 complete the task successfully (or, equivalently, 4 of 15 do not complete the task successfully). This upper confidence interval limit is above the target of 90%, so we accept the null hypothesis that the passing rate is at least 90%. If 5 of 15 failed to complete the task, then the null hypothesis is rejected, and we conclude that the pass rate is less than 90%. This would happen because if 5 of 15 failed, the upper 95% confidence limit would be less than our 90% target value. Table 6.C2 shows that larger sample sizes in general provide more statistical power in detecting usability problems. Higher statistical power means that there is a higher probability of being able to correctly reject the null hypothesis. For example, the following two sampling plans could be used for an objective of greater than 90%: Plan 1: n = 10, accept = 3, reject = 4 Plan 2: n = 20, accept = 4, reject = 5 If the true population passing rate were 50% (this population value cannot be known with certainty and must be estimated), Table 6.C2 would show the n = 10 plan yielding a (false) success rate 17% of the time, while the n = 20 plan would be falsely positive only 1% of the time. In other words, the type II error of falsely accepting the null hypothesis (i.e., concluding that the participants could perform the task successfully 90% of the time when in fact the rate in the participant population is only 50%) for a sample size of 20 would rarely occur. Table 6.C2 shows a range of values for probability of acceptance if the true population passing rate takes on the values listed across the table ranging from 100% to 10%. Table 6.C2 also shows the value of the type I error in the far right column. The type I error is the probability of falsely rejecting the null hypothesis (i.e., false negative of concluding the test failed, whereas the population’s task success rate is in fact greater than the acceptance criteria). Type I error is a single value. Conversely, type II error is a range of values represented by a curve, the receiver operating curve (ROC), that is a function of the assumed and estimated values of the true population passing rates. One-tailed tests are appropriate given the interest in whether the passing rate is above a certain target, not whether the passing rate is exactly at some target. That is, the null hypothesis is unidirectional. Two-tailed statistical tests and resulting confidence intervals are appropriate only when testing whether a pass rate is at a certain exact value. Thus, larger sample sizes will result in more usability defects being found during formative usability tests and more power for the statistical tests used in summative usability tests. Some human factors professionals advocate the use of larger sample sizes in the 30 to 100 range to pick up a higher percentage of usability defects as well as to have greater statistical confidence and higher power in summative results (i.e. lower type II error probabilities).
TAF-ER562X_WINGR-09-0401-C006.indd 243
11/9/10 12:49:47 AM
244
Handbook of Human Factors in Medical Device Design
A common formative testing strategy is to test multiple small samples iteratively during the design stages such that the cumulative sample size might range from 30 to 80 total test participants. Figure 6.C2 shows the cumulative sample size that might be achieved through an iterative development cycle, where there are six stages of formative tests and one final stage of summative testing. In this figure, for a complex device, the formative tests each have 8 participants and 25 for the final summative test. For a simple device, the formative tests have 5 participants each with 15 for the summative test. In summary, efficient usability test designs that produce confidence in the resulting design decisions are possible with modest sample sizes without sacrificing statistical reliability. It should be noted that there is a philosophy behind having an “optimistic” null hypothesis such as noted here: • Ho (null hypothesis): pass rate ≥90% • Ha (alternative hypothesis): pass rate <90% This philosophy is optimistic because we are assuming that the earlier iterative rounds of formative usability tests have produced an acceptable level of usability and that the summative usability test is a final check to validate that optimistic assumption. This philosophy is also used in manufacturing operations quality control sampling testing. It is assumed that a highly reliable manufacturing process has been iteratively designed and will produce high-quality products unless data from relatively small quality control samples indicates otherwise. The desired outcome with this philosophy is to accept the more favorable null hypothesis. An alternative philosophy that is espoused in some international standards for consumer product usability is a more conservative and pessimistic one. The pessimistic philosophy assumes that there is no prior evidence that the product has been designed for acceptable usability, and the validation test takes on a different reverse direction for the null hypothesis as follows: • Ho (null hypothesis): pass rate ≤90% • Ha (alternative hypothesis): pass rate >90% In this case, much larger sample sizes are required (e.g., 50 to 70) to get to the point of rejecting the pessimistic null hypothesis and accepting the desirable alternative hypothesis. This is more conservative and preferable when it is feasible to conduct usability tests with larger sample sizes. The pessimistic assumption is that all is not well until the data prove
Cumulative sample size
140 120 100 80 60 40 20 0 1
2
3
4
5
6
7
Phase of testing
FIGURE 6.C2
Cumulative sample size over multiple usability tests.
TAF-ER562X_WINGR-09-0401-C006.indd 244
11/9/10 12:49:47 AM
Testing and Evaluation
245
otherwise. This is in contrast to the optimistic assumption that all is well until the data prove otherwise. The larger samples sizes for the pessimistic hypothesis point of view will have the added advantages of lower type I and type II error rates but may not be practical for many usability testing situations. A variation recommended by some statisticians is to lower the value of the number of failures to reject the null hypothesis in Table 6.C2. This has the effect of shifting the ROC and thereby reducing the type II error rates (falsely accepting the null hypothesis) but with the trade-off of increasing the type I error rate (falsely rejecting the null hypothesis). Alternative statistical approaches are possible in validation or summative usability testing where acceptance criteria are important parameters and can involve statistical tests other than the exact binomial test, including using the hypergeometric statistical test, the Wilson method, the Ward method, or Bayesian statistics. The main point of the methods suggested here is that any statistical testing against validation acceptance criteria needs to be statistically justified.
APPENDIX 6.D: FREQUENTLY ASKED QUESTIONS CAN A FOCUS GROUP BE USED FOR USABILITY TESTING? No. Focus groups are not suitable for usability testing. Focus groups are best suited for measuring opinions and attitudes and for eliciting indications of possible user behaviors. They do not allow observation of actual user behavior or measurement of functional performance. The social interactions among group members and the potential for moderator bias are some of the other major drawbacks in using focus groups. A dominant group member may sway the opinion of less aggressive members. Some group members may passively agree with the predominant group opinion regarding usability, thereby failing to contribute their unique, possibly valuable, and perhaps contrary usability observations. Similarly, the following methods are not recommended as substitutes for well-planned and well-executed usability tests: • • • •
Customer preference studies using visual analog scales of subjective ratings Self-reports (which are subject to memory recall bias and errors) Use of only subjective satisfaction ratings Clinical device trials (These are more suitable for evaluating device efficacy and reliability, unless supplemented by systematic observation of users performing actual tasks on the operational system) • Simple tests of device functionality with possibility of biased instructions or use of self-report data • Anecdotal information from early device trials or from training sessions • Executive or engineering reviews
DOES THE “RULE OF 300” APPLY TO SAMPLE SIZE IN USABILITY TESTS; THAT IS, FOR STATISTICAL RELIABILITY, DON’T WE NEED LARGE SAMPLE SIZES? Usability tests rarely require large sample sizes of 300 or more participants, as might typically be used for market research. In fact, for early formative usability tests, small samples of participants from each distinct user group make up a sufficient and cost-effective use
TAF-ER562X_WINGR-09-0401-C006.indd 245
11/9/10 12:49:47 AM
246
Handbook of Human Factors in Medical Device Design
of scarce resources. For later-stage summative usability testing, larger sample sizes are required but rarely exceed 100 participants and may be as little as one-fifth that size.
IS IT APPROPRIATE FOR USER-INTERFACE DESIGNERS TO RUN USABILITY TESTS ON THEIR OWN DESIGNS? Although not optimal because of the risk of biased interpretations, it may sometimes be appropriate for trained human factors professionals to run usability tests on their own designs. This is appropriate only if the design has been developed using rigorous user-centered design methods, including contextual inquiry, use error risk analysis, prototyping, and iterative design. Moreover, the usability test plan should be designed to minimize the risk of observer and facilitator biases. The goal of usability testing is to measure if prespecified usability objectives have been achieved, thereby reducing possible bias caused by user-interface designers ignoring usability defects found by users during usability testing. It is never acceptable for developers not trained in human factors engineering to conduct usability tests on their own designs. Independent human factors experts should be used in this case.
CAN A USABILITY DESIGN CHECKLIST BE USED IN PLACE OF A USABILITY TEST? A usability checklist can supplement a usability test but should never be used as a substitute. A checklist can be used to evaluate the general conformance of a design to accepted design guidelines. Research has shown that this is never sufficient; only usability testing with realistic tasks being performed by representative users will uncover latent usability defects.
CAN AN EXPERT REVIEW OR A HEURISTIC REVIEW REPLACE A USABILITY TEST? No. Similar to design checklists, expert reviews (reviews based on previous research and experience) or heuristic reviews (reviews based on applying design rules of thumb) by userinterface design or human factors experts can uncover many but not all usability defects. The best approach may be to combine expert reviews (the analytical approach embodied in use error–focused risk analysis) with the empirical approach of observing use errors and user difficulties in a usability test. For complex user interfaces, it is usually impractical to do formal usability testing of all possible user tasks. In these cases, the expert review can look at all tasks, and the usability testing would be conducted on the most significant and highest-risk-level tasks.
WHAT IS WRONG WITH WAITING TO GATHER USABILITY DATA FROM POSTMARKET REVIEWS OF A DEVICE? Waiting until the device is released to discover usability defects substantially increases the risk of operational device–related use errors that can cause patient or user injury or death. The probability of device recalls and litigation increases significantly. Furthermore, fixing usability defects after a device is released is usually very expensive and time-consuming. Options for mitigating usability problems in marketed devices are often limited to warnings and training, which are much less effective than device design solutions at preventing future use errors.
TAF-ER562X_WINGR-09-0401-C006.indd 246
11/9/10 12:49:47 AM
Testing and Evaluation
247
USABILITY TESTING ADDS TIME AND COSTS TO MY ALREADY TIGHTLY CONSTRAINED DEVELOPMENT PROJECT, SO HOW CAN I JUSTIFY THIS ADDITIONAL INVESTMENT? The return on investment (ROI) for doing early usability testing with modest sample sizes is large. ROI for usability investments can range from 10% to over 100% (Bias and Mayhew 1994). Combining user-centered design with early usability testing actually reduces the development time primarily because expensive and time-consuming rework is avoided since the device is designed right the first time. Moreover, when considering the entire product life cycle, incorporation of usability engineering will reduce training costs, decrease support costs, increase customer satisfaction and brand loyalty, and possibly increase market life span.
WON’T USABILITY TESTING BE A WASTE OF TIME IF I DO EXTENSIVE INITIAL CONTEXTUAL INQUIRY RESEARCH AND SUBSEQUENTLY DEVELOP USER PROFILES (E.G., PERSONAS) AND TASK FLOWCHARTS AND USE ENVIRONMENT DESCRIPTIONS AS PART OF MY USER-INTERFACE SPECIFICATION? If predicting human behavior were a better-developed science, then usability testing might not be necessary. Moreover, these design methods have their limitations and biases (e.g., selection of the attributes of specific personas is subjective and may inadvertently omit or underspecify critical user attributes). Thus, even with a rigorous user-centered design process, the current state of the art is that initial designs are rarely perfect and free of usability defects. Only task-based usability testing will uncover latent usability defects.
CAN USABILITY TESTS BE USED FOR BOTH VERIFICATION AND VALIDATION TESTING? Yes. Usability testing can be used for both verification (assessing whether design outputs meet design inputs: was the device built correctly?) and validation (assessing whether the design meets valid customer requirements: was the correct device built?). Usability testing validates only the user interface and not the entire device. Full device validation requires extensive testing of its effectiveness and safety through multiple validation steps.
SHOULD USABILITY TESTING BE DONE ON THE DEVICE’S DOCUMENTATION, LABELING, AND TRAINING? Yes. The user interface of a device, by definition, includes labeling, documentation, and training materials. Usability test tasks should include requiring the users to follow the user documentation (user manuals or quick reference materials). If it is known that devicespecific training will not always be delivered and that for some users their only source of information on how to use a device is the user manual, then it is even more critical to conduct a usability test of the user manual. Usability testing of training materials (CD-ROM, instructor-led sessions, Web-based materials, and so on) will measure the effectiveness of these materials.
TAF-ER562X_WINGR-09-0401-C006.indd 247
11/9/10 12:49:47 AM
248
Handbook of Human Factors in Medical Device Design
RESOURCES American National Standards Institute. (1991). American National Standard for Safety Color Code. ANSI Z 535.1-1991. New York: American National Standards Institute. American National Standards Institute. (1991). Criteria for Safety Symbols. ANSI Z 535.3-1991. New York: American National Standards Institute. American National Standards Institute. (1998). American National Standard for Product Safety Signs and Labels. ANSI Z 535.4-1998. New York: American National Standards Institute. American National Standards Institute. (2002). American National Standard for Safety Color Code. ANSI Z-535.1-2002. New York: American National Standards Institute. American Psychological Association. (1982). Ethical Principles in the Conduct of Research with Human Participants. Washington, DC: APA Press. Bias, R. G. and Mayhew, D. J. (Eds.). (1994). Cost-Justifying Usability. Boston: Academic Press. Dumas, J. and Redish, J. C. (1999). A Practical Guide to Usability Testing (rev. ed.). Bristol, UK: Intellect Books. International Organization for Standardization. (1997). ISO 9241-1:1997. Ergonomic Requirements for Office Work with Visual Display Terminals, Part 11—Guidance on Usability. Geneva: International Organization for Standardization. International Organization for Standardization. (1999). ISO 13407:1999. Human-Centered Design Processes for Interactive Systems. Geneva: International Organization for Standardization. International Organization for Standardization. (2000). ISO/TR 18529:2000. Ergonomics— Ergonomics of Human-System Interaction—Human-Centred Lifecycle Process Descriptions. Geneva: International Organization for Standardization. International Organization for Standardization. (2001). ISO 9186:2001. Graphical Symbols— Test Methods for Judged Comprehensibility and for Comprehension. Geneva: International Organization for Standardization. International Organization for Standardization. (2002). ISO/TR 16982:2002. Ergonomics of Human-System Interaction—Usability Methods Supporting Human-Centred Design. Geneva: International Organization for Standardization. International Organization for Standardization. (2007). ISO/IEC 62366:2007. Medical Devices— Application of Usability Engineering to Medical Devices. Geneva: International Organization for Standardization. Mayhew, D. (1999). The Usability Engineering Lifecycle: A Practitioner’s Handbook for User Interface Design. San Francisco: Morgan Kaufmann. Nielsen, J. and Landauer, T. K. (1993). A mathematical model of the finding of usability problems. In CHI ’93. Conference Proceedings on Human Factors in Computing Systems. New York: ACM, pp. 206–213. Rubin, J. (1994). Handbook of Usability Testing: How to Plan, Design, and Conduct Effective Tests. New York: Wiley. Wiklund, M. E. (1995). Medical Device and Equipment Design: Usability Engineering and Ergonomics. Boca Raton, FL: CRC Press. Wiklund, M. E. and Wilcox, S. B. (2005). Designing Usability into Medical Devices. Boca Raton, FL: CRC Press.
TAF-ER562X_WINGR-09-0401-C006.indd 248
11/9/10 12:49:47 AM
Testing and Evaluation
249
REFERENCES American National Standards Institute/Association for the Advancement of Medical Instrumentation (ANSI/AAMI). (2001). Human Factors Design Process for Medical Devices. ANSI/AAMI HE-74-2001. Arlington, VA: Association for the Advancement of Medical Instrumentation. American National Standards Institute. (2002). American National Standard for Product Safety Signs and Labels. ANSI Z535.4-2002. New York: American National Standards Institute. American National Standards Institute. (2002). American National Standard for Safety Symbols. ANSI Z535.3-2002. New York: American National Standards Institute. American National Standards Institute/National Committee for Information Technology Standards (ANSI/NCITS). (2001). Common Industry Format for Usability Test Reports. ANSI/NCITS 354-2001. New York: American National Standards Institute. Fowler, F. J. (1995) Improving Survey Questions: Design and Evaluation. Thousand Oaks, CA: Sage. Hackos, J. T. and Redish, J. C. (Eds.). (1998). User and Task Analysis for Interface Design. New York: Wiley. Lankford, C. P., Shannon, P. F., Beling, P. A., McLaughlin P. J., Ellis, S. H., Israelski, E. W., and Hutchinson, T. E. (1997). Graphical User Interface Design Using Eye Gaze Tracking and Pupil Response with ERICA. Proceedings of the 41st Annual Meeting of the Human Factors Society. Nielsen, J. (1994). Usability Engineering. San Francisco: Morgan Kaufmann. Nielsen, J. and Mack, R. L. (Eds.). (1994). Usability Inspection Methods. New York: Wiley. Rea, L. M. and Parker, R. A. (1997). Designing and Conducting Survey Research: A Comprehensive Guide. New York: Jossey-Bass. Synder, C. (2003). Paper Prototyping: The Fast and Easy Way to Design and Refine User Interfaces. San Francisco: Morgan Kaufmann. Wiklund, M. E. (Ed.). (1994). Usability in Practice: How Companies Develop User-Friendly Products. Boston: Academic Press. Winer, B. J. (1971). Statistical Principles in Experimental Design. New York: McGraw-Hill Virzi, R. A. (1992). Refining the test phase of usability evaluation: How many subjects is enough?” Human Factors 34, 457–468.
TAF-ER562X_WINGR-09-0401-C006.indd 249
11/9/10 12:49:47 AM
TAF-ER562X_WINGR-09-0401-C006.indd 250
11/9/10 12:49:47 AM
7 Controls Stephen B. Wilcox, PhD CONTENTS 7.1 General Considerations ............................................................................................255 7.1.1 Control Selection .........................................................................................256 7.1.2 General Human Factors Principles for Controls .........................................257 7.1.2.1 Force ............................................................................................257 7.1.2.2 Feedback......................................................................................258 7.1.2.3 Control Layout .............................................................................258 7.1.2.4 Labeling.......................................................................................259 7.1.2.5 Icons ............................................................................................260 7.1.3 Special Considerations ................................................................................260 7.2 Design Guidelines for Specific Controls..................................................................261 7.2.1 Control Panel–Type Controls .....................................................................261 7.2.2 Push Buttons ...............................................................................................262 7.2.2.1 Push-Button Geometry ................................................................263 7.2.2.2 Push-Button Force .......................................................................263 7.2.2.3 Push-Button Mounting Surface .................................................. 264 7.2.2.4 Push-Button Feedback ................................................................ 264 7.2.2.5 Differentiating Push Buttons ...................................................... 264 7.2.3 Thumbwheels ............................................................................................. 264 7.2.3.1 Thumbwheel Geometry .............................................................. 264 7.2.3.2 Thumbwheel Force ......................................................................265 7.2.3.3 Thumbwheel Layout ....................................................................265 7.2.3.4 Thumbwheel Feedback ................................................................265 7.2.4 Rotary Knobs ..............................................................................................265 7.2.4.1 Knob Geometry ...........................................................................265 7.2.4.2 Knob Force ..................................................................................266 7.2.4.3 Knob Mounting ...........................................................................266 7.2.4.4 Knob Layout ................................................................................266 7.2.4.5 Knob Feedback ............................................................................266 7.2.4.6 Knob Increments .........................................................................266 7.2.5 Rotary Encoders..........................................................................................267 7.2.5.1 Rotary Encoder Geometry ..........................................................267 7.2.6 Toggle Switches ..........................................................................................267 7.2.6.1 Toggle Switch Geometry .............................................................268 7.2.6.2 Toggle Switch Force ....................................................................268 251
TAF-ER562X_WINGR-09-0401-C007.indd 251
11/8/10 11:43:26 PM
252
Handbook of Human Factors in Medical Device Design
7.2.7
7.2.8
7.2.9
7.2.10
7.2.11 7.2.12
7.2.13
7.2.6.3 Mounting Toggle Switches ..........................................................268 7.2.6.4 Toggle Switch Layout ..................................................................268 7.2.6.5 Toggle Switch Travel ...................................................................268 7.2.6.6 Toggle Switch Feedback ..............................................................268 7.2.6.7 Inadvertent Activation of Toggle Switches ..................................269 7.2.6.8 Toggle Switch Labeling ...............................................................269 Small Levers ...............................................................................................269 7.2.7.1 Lever Geometry...........................................................................269 7.2.7.2 Lever Force..................................................................................269 7.2.7.3 Mounting Levers .........................................................................269 7.2.7.4 Lever Layout................................................................................270 7.2.7.5 Inadvertent Lever Activation .......................................................270 7.2.7.6 Lever Labeling ............................................................................270 Rocker Switches ..........................................................................................270 7.2.8.1 Rocker Switch Geometry ............................................................270 7.2.8.2 Rocker Switch Force....................................................................270 7.2.8.3 Mounting Rocker Switches..........................................................270 7.2.8.4 Rocker Switch Layout..................................................................271 7.2.8.5 Rocker Switch Travel...................................................................271 7.2.8.6 Rocker Switch Feedback .............................................................271 7.2.8.7 Rocker Switch Labeling ..............................................................271 Sliders .........................................................................................................271 7.2.9.1 Slider Geometry ..........................................................................271 7.2.9.2 Slider Force .................................................................................272 7.2.9.3 Mounting Sliders .........................................................................272 7.2.9.4 Slider Layout ...............................................................................272 7.2.9.5 Slider Feedback ...........................................................................272 7.2.9.6 Slider Labeling ............................................................................272 Key-Operated Controls ...............................................................................272 7.2.10.1 Key/Lock Geometry ....................................................................273 7.2.10.2 Key Control Force ......................................................................273 7.2.10.3 Key Control Mounting ...............................................................273 7.2.10.4 Key Control Feedback .................................................................273 7.2.10.5 Key/Lock Labeling ......................................................................273 Fingerprint Readers.....................................................................................273 Membrane Controls and Keypads ...............................................................273 7.2.12.1 Membrane Button Geometry.......................................................273 7.2.12.2 Treatment of the Surface of Membrane Controls ........................274 7.2.12.3 Membrane Button Force ..............................................................274 7.2.12.4 Mounting of Membrane Controls ................................................274 7.2.12.5 Layout of Membrane Controls.....................................................274 7.2.12.6 Numeric Keypads .......................................................................274 7.2.12.7 Feedback from Membrane Controls ............................................275 7.2.12.8 Membrane Control Labeling ......................................................275 7.2.12.9 Differentiating Membrane Controls ............................................275 Touch Screens .............................................................................................275 7.2.13.1 Advantages of Touch Screens ......................................................277
TAF-ER562X_WINGR-09-0401-C007.indd 252
11/8/10 11:43:26 PM
Controls
7.2.14
7.2.15
7.2.16
7.2.17
7.2.18
7.2.19
7.2.20
253
7.2.13.2 Disadvantages of Touch Screens .................................................277 7.2.13.3 Touch-Screen Geometry ..............................................................277 7.2.13.4 Touch Zone Force ........................................................................278 7.2.13.5 Touch Zone Activation ................................................................278 7.2.13.6 Touch-Screen Feedback...............................................................278 7.2.13.7 Touch-Screen Labeling................................................................279 Standard Keyboards ....................................................................................279 7.2.14.1 Advantages of Keyboards ............................................................279 7.2.14.2 Disadvantages of Keyboards .......................................................279 7.2.14.3 Keyboard Geometry ....................................................................280 7.2.14.4 Keyboard Layout .........................................................................280 7.2.14.5 Keyboard Force ...........................................................................280 7.2.14.6 Keyboard Feedback .....................................................................280 7.2.14.7 Key Labeling ...............................................................................281 Mice ............................................................................................................281 7.2.15.1 Advantages of Mice .....................................................................281 7.2.15.2 Disadvantages of Mice ................................................................281 7.2.15.3 Mouse Geometry and Design Attributes .....................................281 7.2.15.4 Feedback from Mouse Activation ...............................................282 7.2.15.5 Other Design Requirements for Effective Mouse Use ................282 Styli and Light Pens ....................................................................................282 7.2.16.1 Advantages of Styli/Light Pens ...................................................282 7.2.16.2 Disadvantages of Styli/Light Pens...............................................283 7.2.16.3 Styli/Light Pen Geometry and Design Considerations................283 7.2.16.4 Forces for Styli/Light Pens ..........................................................284 7.2.16.5 Feedback from Stylus/Light Pen Activation ................................284 Trackballs ...................................................................................................284 7.2.17.1 Advantages of Trackballs ............................................................284 7.2.17.2 Disadvantages of Trackballs ........................................................284 7.2.17.3 Trackball Geometry.....................................................................285 7.2.17.4 Forces for Trackballs ...................................................................285 7.2.17.5 Trackball Feedback .....................................................................285 Displacement Joysticks ...............................................................................286 7.2.18.1 Advantages of Joysticks...............................................................286 7.2.18.2 Disadvantages of Joysticks ..........................................................286 7.2.18.3 Joystick Geometry .......................................................................286 7.2.18.4 Joystick Force ..............................................................................286 7.2.18.5 Joystick Feedback ........................................................................286 Isometric Joysticks ......................................................................................287 7.2.19.1 Advantages of Isometric Joysticks...............................................287 7.2.19.2 Disadvantages of Joysticks ..........................................................287 7.2.19.3 Isometric Joystick Geometry .......................................................287 7.2.19.4 Isometric Joystick Force ..............................................................287 7.2.19.5 Isometric Joystick Feedback ........................................................288 Other Input Devices ....................................................................................288 7.2.20.1 Pointer Sticks ...............................................................................288 7.2.20.2 Touch Pads...................................................................................288
TAF-ER562X_WINGR-09-0401-C007.indd 253
11/8/10 11:43:26 PM
254
Handbook of Human Factors in Medical Device Design
7.2.20.3 Data Gloves .................................................................................288 7.2.21 Hands-Free Controls ...................................................................................288 7.2.21.1 Eye Tracking................................................................................288 7.2.21.2 Directed Breathing ......................................................................288 7.2.21.3 Mouth Sticks................................................................................288 7.2.21.4 Head Movement ..........................................................................289 7.2.21.5 Voice-Activated Control ..............................................................289 7.3 Large Mechanical Controls .....................................................................................289 7.3.1 Cranks .........................................................................................................289 7.3.1.1 Crank Geometry ..........................................................................289 7.3.1.2 Crank Forces ...............................................................................290 7.3.2 Handwheels .................................................................................................290 7.3.2.1 Handwheel Geometry..................................................................290 7.3.2.2 Handwheel Force .........................................................................291 7.3.3 Large Levers ...............................................................................................291 7.3.3.1 Lever Geometry...........................................................................291 7.3.3.2 Lever Force..................................................................................291 7.3.3.3 Lever Feedback ...........................................................................292 7.3.4 Whole-Hand-Operated Controls ................................................................292 7.3.4.1 Hand-Operated Control Geometry ..............................................292 7.3.4.2 Hand Control Force .....................................................................292 7.3.5 Foot Controls...............................................................................................292 7.3.5.1 Foot Control Geometry and Location .........................................293 7.3.5.2 Foot Control Force ......................................................................293 7.3.5.3 Foot Control Feedback ................................................................293 Resources .........................................................................................................................293
This chapter focuses on the specification and design of controls for medical devices. A control is defined as a device component used to alter how a device functions. The guidelines presented here address traditional “control panel” controls, such as rotary knobs and toggle switches; “input devices” for electronic equipment, such as mice and touch screens; and larger control mechanisms, often of a strictly mechanical nature, such as cranks and handwheels. Controls are generally associated with displays, which provide feedback to the user concerning the results of control operations. Guidelines for displays, however, are provided separately in Chapter 8, “Visual Displays.” Also, control design should generally include usability testing to ensure that users will be able to use the controls in the intended way with minimal error. Usability testing is covered in Chapter 6, “Testing and Evaluation.” This chapter includes guidelines for input devices but leaves further discussion of control functions for computer systems to Chapter 12, “Workstations.” Furthermore, this chapter does not address the specific controls for handheld devices, which are treated in Chapter 16, “Hand Tool Design.” Because controls provide the means for clinicians, patients, or maintainers to operate devices, good control design is central to the elimination of use error. Thus, the usability of a device is only as good as the usability of its controls. We can distinguish between the control/display interface on the one hand and what we might call the “natural” interface on
TAF-ER562X_WINGR-09-0401-C007.indd 254
11/8/10 11:43:26 PM
Controls
255
the other. As devices become more complex and sophisticated, the control/display interface inevitably becomes more important at the expense of the natural interface. So, for example, medical professionals and patients historically interacted simply and directly with beds. Now they interact with elaborate control mechanisms, often provided via touch-screen graphic user interfaces as well as with the physical beds themselves. What this means is that, increasingly, from the user’s perspective, the control/display interface is the medical device. Examples of medical device controls include the following: • The buttons on blood glucose meters that allow users to check their past readings, add contextual data (e.g., insulin doses) to a given reading, change the date, and so on • The array of buttons, knobs, and trackballs that might be found on an ultrasound system • Touch-screen interfaces that are used to set parameters and operate a wide variety of devices, from cardiac imaging systems to intraaortic balloon pumps to contrastmedium injection systems • Handheld controls that incorporate rocker switches • Foot controls that activate electrocautery and harmonic surgical instruments • The plunger on a syringe • Cranks used to raise or lower beds Thus, with the exception of some relatively simple devices, such as some disposables, most medical devices incorporate some sort of controls.
7.1 GENERAL CONSIDERATIONS A control’s purpose is to provide a means for a user to alter how a device functions. Thus, the first step for control specification is to understand the function to be controlled. Some key questions in this regard include the following: • • • • • •
Is the function continuous, or does it contain discrete alternatives? If it contains discrete alternatives, how many are there, and how are they ordered? What are the boundaries of the functions to be controlled? How much control precision is necessary? How will information about the state of the control be communicated to the user? How catastrophic would it be to inadvertently alter a control setting?
Once the function to be controlled is fully understood, the next step is to understand the intended users’ needs and preferences as well as the likely use environments. Only then can the optimal type of control be chosen. The goal is to match the control to what it is going to control on the one hand and to the user and the use environment on the other. With regard to users, physical issues, such as strength and hand size, are potentially important (see Chapter 4, “Anthropometry and Biomechanics”), as are cognitive issues, such as the type of mistakes that particular users are likely to make or what their expectations are as a function of training (see Chapter 2, “Basic Human Abilities”). Environmental considerations include lighting levels and whether the environment contains vibration or other conditions that could affect the user’s ability to use controls (see
TAF-ER562X_WINGR-09-0401-C007.indd 255
11/8/10 11:43:26 PM
256
Handbook of Human Factors in Medical Device Design
Chapter 3, “Environment of Use”). Also, most controls are associated with specific information displays, so the controls and the displays must be designed together (see Chapter 8, “Visual Displays”). Once the type of control is chosen, the next step is to determine its specific parameters (e.g., shape, size, travel, and force). Then the controls have to be arranged. Of course, decisions about control arrangement should take place in parallel with the selection of specific control parameters because the form of the controls can affect the layout and vice versa, as can the design and arrangement of associated displays. One final step is to determine appropriate control labeling (see Chapter 13, “Signs, Symbols, and Markings”). Moreover, most devices require more than one control, and the integration of multiple controls and displays raises additional design considerations (see Chapter 12, “Workstations”). This section provides recommendations for selecting the type of control and for determining its various parameters. These guidelines represent a “first pass” for making design decisions. They are no substitute for empirically testing ease of use and effectiveness of control designs (see Chapter 6, “Testing and Evaluation”).
7.1.1 CONTROL SELECTION Some general principles that apply to control selection include the following: Guideline 7.1: Comply with Conventions Control conventions used by other common devices with which users are familiar should be considered. There are certainly times that such conventions can be violated, for example, when existing conventions are shown to be inferior to a newer alternative. However, it is crucial to carefully consider and address user expectations when choosing a control. Usability testing is one way to determine users’ needs and expectations.
Guideline 7.2: Logical Control/Display Relationships Control movement should have a logical relationship to associated display movement. Control feedback should be compatible with control movement such that the feedback should be consistent with user expectations. Many controls generate expectations regarding the effects of directional control movement; for example, rotating a knob counterclockwise causes something to open. However, such expectations can be culture-specific, so they should be tested with potential users.
Guideline 7.3: Prevent Inadvertent Activation Inadvertent activation of controls should be prevented, particularly when such activation can have safety consequences. Methods for minimizing inadvertent activation include the following: • • • • •
Locating the control so that the user is unlikely to contact it accidentally Shielding it with a physical barrier or movable guard Recessing it below the surrounding surface Placing it within a “well” formed by raised partitions Interlocking the control so that it cannot be operated unless other criteria are met and/ or actions are taken
TAF-ER562X_WINGR-09-0401-C007.indd 256
11/8/10 11:43:26 PM
Controls
257
• Building in resistance to movement (e.g., detents) • Requiring a control to be activated for a period of time (e.g., pressing a button for at least three seconds to adjust a function) • Requiring multiple operations to activate controls in synchrony • Requiring confirmation before initiating the controlled function—particularly relevant whenever changing the function has critical implications
Guideline 7.4: Test for Intuitiveness Intuitiveness of controls should be tested with members of the intended user population. In general, it is important to test controls, along with other aspects of a given device design, via usability testing.
Guideline 7.5: Accommodate Use of Protective Gear When users are likely to wear gloves or other protective gear, the design of controls must take this into account. Gloves decrease tactile sensitivity, increase the effective size of the hand, and can decrease the friction between the gloved hand and the control. Other types of protective gear can compromise vision or hearing as well as tactile sensitivity and can decrease mobility and dexterity. If a device will be operated by people wearing protective gear, it should be tested that way.
Guideline 7.6: Design for Safety Where hazards can be reduced, safety devices should be built into controls. Examples of safety devices include interlocks on autoclaves that do not allow doors to open while steam is generated and momentary contact controls on electrocautery devices that produce energy only when the control is affirmatively activated.
Guideline 7.7: Fail-Safe If incapacitation of the user can cause a critical condition, then a “dead-man” control should be used. A dead-man control is a type of momentary contact control that deactivates the critical state (e.g., firing a laser) when the user stops applying affirmative control (e.g., pressing a button).
7.1.2 GENERAL HUMAN FACTORS PRINCIPLES FOR CONTROLS The following are human factors principles that apply to all controls. 7.1.2.1 Force See Chapter 4, “Anthropometry and Biomechanics,” for more information about force. Guideline 7.8: Minimize Required Forces The force required for any control should be as low as possible—sufficient to provide tactile feedback and to prevent inadvertent activation. A particular advantage of low-force controls is that they facilitate use by people with physical disabilities. However, it is important to require enough resistance force—around 0.7 pounds (3.1 N)—for typical finger-operated controls to prevent “overshoot” (i.e., moving the control farther than intended), particularly with rotary selectors, track knobs, and sliders. Overshoot is a larger problem for people with impaired hand control.
TAF-ER562X_WINGR-09-0401-C007.indd 257
11/8/10 11:43:26 PM
258
Handbook of Human Factors in Medical Device Design
Guideline 7.9: Motion Pushing or pulling motions are generally preferable to rotation for impaired and elderly users. Rotation requires greater fine motor control.
7.1.2.2 Feedback There should be immediate feedback to the user that a control motion has been accepted by the device. User performance is degraded when there is any delay between the time that a user manipulates a control and when he or she perceives the relevant feedback. If there is an inherent delay in the controlled parameter, the user should still receive immediate feedback that the control command has been initiated. Guideline 7.10: Immediate Feedback Control feedback should be immediate.
Guideline 7.11: Clear Feedback Controls should provide clear information about a user’s actions. Feedback can be provided through properties of the control itself (intrinsic feedback) or from a separate display (extrinsic feedback).
Guideline 7.12: Redundant Feedback Redundant feedback is generally advantageous. Providing multimodal visual, auditory, and tactile feedback concurrently will make it more likely that all users will be able to use a control accurately and reliably. Redundant feedback is particularly important for critical controls, such as those that control drug infusion or parameter setting for imaging systems.
Guideline 7.13: Elastic Resistance For most controls, elastic resistance is optimal for tactile feedback. The recommended pattern of tactile feedback is initially a low resistance that builds up rapidly, followed by a rapid decrease when the control reaches the activation point.
Guideline 7.14: Auditory Feedback Auditory feedback is most effective in environments with low ambient noise levels. For simple tones, the recommended frequency is between 400 and 1,500 Hz. Complex tones are generally preferable to pure tones (see Chapter 10, “Alarm Design,” for more information on this topic).
Guideline 7.15: Confirmation of Critical Actions Controls for highly critical functions should incorporate at least one confirmation step to ensure verification of the command. For example, after receiving initial feedback of control activation, the user then has to perform another control operation to actuate the desired function.
7.1.2.3 Control Layout Guideline 7.16: Frequency of Use The most frequently used controls should generally be placed in the most convenient locations, but other factors need to be considered. In practice, layout is a compromise between a number
TAF-ER562X_WINGR-09-0401-C007.indd 258
11/8/10 11:43:27 PM
Controls
259
of arrangement principles—grouping according to importance (e.g., most important at the top), grouping according to function (i.e., controls with similar functions placed together), arrangement according to sequence of use, and so on (see Chapter 12, “Workstations,” for more information about the optimal arrangement of controls and displays).
Guideline 7.17: Adequate Spacing There should be sufficient space surrounding a control to rest the fingers or the hand. This greatly enhances hand comfort, particularly for elderly and impaired users, and reduces the risk of inadvertent control activation.
Guideline 7.18: Optimal Orientation Controls should be arranged so that their operation does not obscure the related displays. Displays placed to the side of related controls are often easier to use than displays placed above or below, although this is less true when there is a large matrix of closely spaced controls. Optimal placement also depends on the line of sight between the user and the control/ display pair. Generally, users are faster and make fewer errors with a horizontal control/ display layout. Placing displays to the left of controls accommodates right-handed users. An alternative is to place displays above controls so that left-handed users are not penalized. Users’ positioning should be considered when selecting control/display relationships. Both line of sight and arm/hand motion should be considered.
Guideline 7.19: Facilitate Multiple Grips Controls should allow various grips rather than requiring unique, predetermined grips that might not accommodate all users and uses.
Guideline 7.20: Comfortable Motion Controls should not require awkward postures or hand/arm positions for their operation.
7.1.2.4 Labeling Guideline 7.21: Consistent Labeling Control labeling should be consistent. Error rates decrease when labels are applied in a consistent manner. This principle also applies to coding, that is, using color, shape, or size, for example, to indicate the type of control.
Guideline 7.22: Label Centering Text labels are often most effective when situated in the center of a control with as large a size as can be accommodated. The placement of labels in the center of touchable areas attracts visual attention and eliminates confusion, although it does obscure the label as the control is used. Therefore, when it is important for the user to continue to see the label while activating a control, labels should be placed in an alternate position that remains in view as the user actuates the control.
Guideline 7.23: Readability As a general rule, the readability of labels viewed within an arm’s reach increases with point size, up to 24 points. Of course, specialized controls that have to be seen from a distance will require even larger text labels.
TAF-ER562X_WINGR-09-0401-C007.indd 259
11/8/10 11:43:27 PM
260
Handbook of Human Factors in Medical Device Design
Guideline 7.24: Use Simple Fonts Simple, block-style (i.e., sans serif) fonts are optimal for control and instrument labeling.
Guideline 7.25: Text Orientation Text labels should read from left to right for products to be used in Western culture. In the United States and Europe, reading speed significantly decreases with other letter orientations. However, this is individual- and culture-specific.
Guideline 7.26: Clear Text Descriptions Text labels should be designed to rapidly and reliably communicate their intended meaning to users. Use of common words in labels decreases communication errors. Abbreviations or acronyms (e.g., NIBP for noninvasive blood pressure) should be restricted to those that the user population clearly understands and where spelling out the abbreviation or acronym would reduce visibility/readability.
Guideline 7.27: Testing Necessary As with other aspects of device design, comprehension of labeling should be tested with users (see Chapter 6, “Testing and Evaluation”). Furthermore, comprehension of label meaning in isolation may not predict comprehension in the context of the entire device during actual use.
Guideline 7.28: Durability Effort should be made to ensure that the label is adequately durable, particularly where the user’s hand comes in contact with the label. One technique is to place labels below transparent material. It can also help to emboss a label so that the text is below the contact surface.
7.1.2.5 Icons Icons are more universal than text in that the same icons can be used across different language groups. However, it is a challenge to make icons as specific and understandable as text. In addition to the issue of language variation, illiteracy remains a global problem (see Chapter 13, “Signs, Symbols, and Markings,” for more information on this subject). Guideline 7.29: Maximize Size for Readability Larger icons and text foster accuracy and ease of use and accommodate visually impaired users.
Guideline 7.30: Provide Cues When symbols are used, it often improves comprehension to provide a cue, such as a “ground line,” to help users understand the proper orientation.
Guideline 7.31: Combine Text and Icons Text plus icon labels are typically more effective than either alone.
7.1.3 SPECIAL CONSIDERATIONS The principles of control design for medical devices are comparable to those of other products. However, there are some special considerations.
TAF-ER562X_WINGR-09-0401-C007.indd 260
11/8/10 11:43:27 PM
Controls
261
• Use error tends to be more consequential with medical devices. The same error (e.g., activating the wrong control, failing to activate a control when it should be activated, or setting a parameter incorrectly) that might cause only an inconvenience in a consumer device can cause patient injury or death with a medical device. • Unlike many consumer products, medical devices tend to be targeted toward very specific user groups, making the design of controls a matter of accommodating a much smaller group. Many medical devices are used exclusively by highly trained professionals. This fact increases the need for controls to accommodate users with very refined and skill-intensive goals. That said, the intended user population is likely to exhibit anthropometric variability similar to the general population (see Chapter 4, “Anthropometry and Biomechanics”). • The controls of home health care devices, by virtue of their targeted users, often have to accommodate people with a variety of disabilities, including visual impairments, mobility and dexterity problems, and cognitive deficits. • Rather than just one user (typically the device owner), medical devices often have a variety of users, including physicians, nurses, technicians, patients, and caregivers who might pose different control requirements. • Medical devices are often used in environments with long-standing idiosyncratic conventions that have to be understood and considered when designing controls. • Many medical devices are portable so that the controls might have to be operated while the device, such as a noninvasive blood pressure measurement device, is moving along with the patient. • It is common for medical devices to be used in emergency situations, so users might be in a high arousal state, leading them to act instinctively rather than “cognitively.” • Certain medical devices (e.g., external defibrillators) are used infrequently, so their operation has to be intuitive to a user who might not have used the device for weeks or months, if at all. • Many medical devices are used in hostile physical environments, such as extreme cold or hot environments. • The same users might switch frequently between the controls of small handheld devices (e.g., digital thermometer) and a large workstation (scanner).
7.2 DESIGN GUIDELINES FOR SPECIFIC CONTROLS 7.2.1 CONTROL PANEL–TYPE CONTROLS The guidelines in this chapter are for “full-size” devices with control panels. Different control types are better for different functions (Table 7.1). The recommendations must be adjusted for handheld and miniaturized devices. In general, dimensions can be reduced when other methods are used to address ease of use (e.g., software algorithms to prevent inadvertent operation of adjacent keys, auditory feedback to replace physical travel). The drawings in this chapter provide recommended minimum dimensions in inches. They should be taken as general recommendations only because any given device will require multiple design considerations and trade-offs. Usability testing should be the final arbiter of design quality.
TAF-ER562X_WINGR-09-0401-C007.indd 261
11/8/10 11:43:27 PM
262
Handbook of Human Factors in Medical Device Design
TABLE 7.1 Selection of Control Elements for Different Functions Applications
3
Med
OF
N
Emergency Start/Stop
L ow
5
High
4
Two States O
Multiple States
F
Discrete Steps 1 2
Continuous Scaling
Controls Control Panel–Type Controls ✓ ✓
Push buttons Toggle switches Continuous thumbwheels Thumbwheels with discrete stops Rotary knobs Levers Rocker switches Sliders Key-operated controls Membrane controls
✓
Touch screens Keyboards Mice Light pens Trackballs Joysticks Other input devices
✓ ✓ ✓ ✓ ✓ ✓ ✓
Pedals Cranks Large levers Wheels Palm buttons
✓ ✓ ✓ ✓
✓
✓
✓
✓
✓
✓
✓
✓
✓ ✓ ✓
✓
✓
✓ ✓
✓
✓
✓
✓ ✓ ✓ ✓ ✓ ✓ ✓
✓ ✓ ✓ ✓ ✓ ✓
✓
Input Devices ✓ ✓ ✓ ✓ ✓ ✓ ✓
Large Mechanical Controls ✓
✓
✓
✓
✓
✓
✓ ✓ ✓
Generally, controls should be grouped into functional categories with additional space between the groupings. However, as mentioned above, other arrangement principles, such as sequencing and importance, should also be considered.
7.2.2 PUSH BUTTONS Push buttons are particularly appropriate for initiating a cycle, for selecting an option/ channel, or for on/off actions (see Table 7.1). Push buttons are heavily dependent on tactile
TAF-ER562X_WINGR-09-0401-C007.indd 262
11/8/10 11:43:27 PM
Controls
263
and visual feedback for control-state discrimination. Push buttons are normally either momentary contact (in which the controlled function changes only while the button is pressed), state change (in which, like an on-only button, pressing the button alters a state in one direction), or two state (in which, like on/off controls, each button press alters the state back and forth between two states). 7.2.2.1 Push-Button Geometry 1. Travel: The goal is to minimize the duration of the application of force by incorporating travel distances just large enough to provide adequate tactile feedback—0.12 inches (3 mm) or so. Too little travel causes increased errors because it does not provide adequate feedback. Of course, if travel distance is difficult to provide, tactile feedback can be replaced or supplement with visual or auditory feedback. 2. Separation (see Figure 7.1 for proposed dimensions): If the space is available, between-button separation should be a minimum of 1 inch (25 mm) to allow the largest fingers to fit easily between controls, decreasing the likelihood of accidental activation of adjacent buttons. This separation also allows a user to rest his or her finger between buttons, which can reduce fatigue, particularly for users with dexterity disabilities. 3. Width and height: Keeping controls as large as necessary to accommodate people with large fingers reduces the need for fine motor control (See Chapter 4, “Anthropometry and Biomechanics,” for information about finger and hand dimensions). The critical factor in the geometry of push buttons is not shape but adequate contact area. Figure 7.1 illustrates the smallest recommended width for a push button. Push buttons can take a variety of shapes, including square, oval, round, lozenge shaped, or rectangular. 4. Surface contour: A concave surface with a small convex radius around the edges helps to keep the finger from sliding off the control while eliminating uncomfortable edges. Adding surface friction (e.g., via texture) to the button also helps to keep the finger from sliding around the control’s surface. 7.2.2.2 Push-Button Force Button resistance should be kept below 2 pounds (9 N) to accommodate the vast majority of users; resistance should be kept below 0.7 pounds (3.1 N) to accommodate many disabled users while still providing adequate tactile feedback (see Chapter 4, “Anthropometry and Biomechanics,” for more information about user-compatible force levels).
0.12” 0.8
”
0.8
”
1”
FIGURE 7.1
Proposed dimensions for control-panel push buttons.
TAF-ER562X_WINGR-09-0401-C007.indd 263
11/8/10 11:43:27 PM
264
Handbook of Human Factors in Medical Device Design
7.2.2.3 Push-Button Mounting Surface If possible, push buttons should be mounted on a horizontal rather than a vertical surface. Buttons mounted on the horizontal plane permit finer motor control than push buttons on the vertical plane. With horizontally mounted controls, the user does not have to exert as much force to keep the hand suspended, thereby providing greater control and precision. When the main body of a given device presents a vertical plane to the user, it often improves usability to place controls on a diagonal plane that might be inset into the device. 7.2.2.4 Push-Button Feedback 1. Buttons should provide immediate feedback. Immediate visual, auditory, and/ or tactile feedback about button position enhances ease of use, particularly for impaired users. 2. Back illumination (i.e., lighting the button itself when pressed) provides good feedback for push buttons because it has a direct connection to the control action. 3. A tone, click, or an audible snap is an effective form of auditory feedback for push buttons. 4. For push buttons, tactile feedback is generally produced through elastic (springlike) resistance as the button travels. 7.2.2.5 Differentiating Push Buttons 1. Differentiating buttons on the basis of shape is valuable for sight-impaired users. People can easily distinguish between five or so different shapes by feel alone. Shape is a good coding method to use redundantly with color and labeling. 2. Size is generally less effective as a coding method. Other possible coding methods include position (including in depth) and texture. 3. For arrays of buttons, it can be helpful to provide a small dimple or protrusion on certain important buttons so that they can be easily identified by feel.
7.2.3 THUMBWHEELS Thumbwheels have the advantages of saving space and providing for an infinite range of rotation, making them well-suited for continuous adjustments (see Table 7.1). When space is an issue, they may also be used to control discrete states. However, thumbwheels are inferior to rotary knobs (see below) for controlling discrete states in that the former but not the latter allow the user to see the separate states simultaneously. Thumbwheels should not be used to indicate more than nine states. 7.2.3.1 Thumbwheel Geometry 1. A thumbwheel should accommodate large thumbs to allow sufficient contact with the wheel while not being unwieldy to users with small thumbs. 2. The “opening length,” or exposed area protruding above the surface, should permit the user to attain significant wheel displacement within a single motion: 1 inch (25 mm) and 90 degrees of arc are appropriate (see Figure 7.2). A small thumbwheel contact surface obliges users to perform tedious, repeated movements to attain significant rotational displacement.
TAF-ER562X_WINGR-09-0401-C007.indd 264
11/8/10 11:43:28 PM
Controls
265
1”
90° 1”
2”
FIGURE 7.2
Proposed dimensions for thumbwheels.
3. A serrated or textured surface reduces slippage for continuous thumbwheels and provides more efficient rotational displacement. 4. For thumbwheels that control choice of discrete states, a protrusion might coincide with each of the discrete states to make the position clearer. If used, protrusions should be at least 0.13 inches (3 mm) high and separated by 0.45 inches (11 mm). 7.2.3.2 Thumbwheel Force Keeping the resistance below 20 ounces (5.6 N) will accommodate the vast majority of users; setting the force at 6 ounces (1.7 N) will accommodate many disabled users while still providing adequate tactile feedback. 7.2.3.3 Thumbwheel Layout 1. In vertically oriented thumbwheels, upward or forward rotation should represent an increase in the controlled variable. 2. In horizontally oriented thumbwheels, left-to-right motion should represent an increase in the controlled variable, at least in North America and Western Europe. The opposite may be optimal in other regions. 7.2.3.4 Thumbwheel Feedback Since continuous thumbwheels do not provide inherent feedback, changes in the controlled variable must be indicated through a separate visual display or otherwise be obvious. For discrete-state thumbwheels, position can be indicated by a detent or by auditory feedback in addition to a separate display.
7.2.4 ROTARY KNOBS There are three basic types of rotary controls: rotary selectors, continuous rotary controls, and spring-return (“momentary contact”) rotary controls. A spring-return rotary control can activate a function in either of two “directions” only while it is turned. Rotary selector controls reduce selection time and can save panel space in that they can have up to 12 specific settings (1 every 30 degrees). Rotary controls can include a push-button action to operate two-position functions. 7.2.4.1 Knob Geometry 1. Height and width: The grip surface should be large enough to accommodate the thickness of the index finger of the largest users and should not have any sharp
TAF-ER562X_WINGR-09-0401-C007.indd 265
11/8/10 11:43:28 PM
266
Handbook of Human Factors in Medical Device Design Top view Cross section 0.7”
1.5”
FIGURE 7.3
Proposed dimensions for rotary knobs.
edges (Figure 7.3). Edges can have radii as small as 3 mm, but a minimum of 9 mm is preferred for comfort. 2. Diameter: The skirt (disk below the gripped area that rotates relative to the underlying surface) of a control needs to accommodate the thumb and index-finger width of the largest users. If given an adequate width and texture, the skirt can also provide an alternative means for turning the control for a user with arthritis. 7.2.4.2 Knob Force Appropriate rotary force (i.e., torque) for overcoming detents is 1 in.-lb. (0.12 N·m). This is the minimum torque necessary to prevent “overshoot.” 7.2.4.3 Knob Mounting Depending on user arm position during use, knobs can be mounted on vertical, horizontal, or even diagonal surfaces. Testing is usually required. 7.2.4.4 Knob Layout In the United States and Western Europe, rotating a control clockwise should increase a variable. The natural mapping of numbers placed around a circular faceplate is associated with our conventions that govern clocks. However, there is an exception to this convention—for fluid control, clockwise rotation is associated with “closed” (or reduced flow) and counterclockwise with “open” (or increased flow). 7.2.4.5 Knob Feedback 1. For rotary selectors, each position should provide tactile and audible feedback. Audible clicks combined with elastic resistance and detents provide optimal feedback. Elastic resistance prevents users from placing the control between settings by an initially increasing force followed by decreasing resistance on the approach of the next setting. In general, such “off-detent” settings should be explicitly precluded since they produce ambiguous settings. 2. To communicate proper functioning and activation status to the user, it can be effective for a light to signal initial activation of rotary controls. 7.2.4.6 Knob Increments Keeping the increments at least 30 degrees apart makes it easier for manually or visually impaired users.
TAF-ER562X_WINGR-09-0401-C007.indd 266
11/8/10 11:43:30 PM
Controls
267
350
Encased numbers
FIGURE 7.4
Rotary control with encased settings.
1. Encased numerical settings in the pointer enhance placement accuracy by providing restrictive visual and physical parameters (see Figure 7.4). However, the scale should be fully exposed when it is important for the user to see the full adjustment range. Of course, the pointer should not overlap or obscure the setting label. 2. Fixed scales with a moving pointer increase precision and reduce errors more effectively than moving scales and fixed pointers. 3. Minimizing the physical separation between the pointer and the scale settings reduces reading errors by reducing parallax (i.e., visual misalignment of the pointer with the scale when viewed from an angle). 4. During operation, the user’s hand should not mask the scale settings.
7.2.5 ROTARY ENCODERS Rotary encoders (also called jog wheels and trim wheels) provide continuous (analog) movements versus the discrete (digital) movement of a rotary knob. They can provide precise adjustments of continuous variables and should be used when low force is required. 7.2.5.1 Rotary Encoder Geometry 1. Height: The height should be between 0.5 and 1 inches (13 to 25 mm). 2. Diameter: For fingertip control, the diameter should be between 0.4 and 4 inches (10 to 100 mm). For finger and thumb grasps, the diameter should be between 1 and 3 inches (25 to 76 mm). 3. Torque: The maximum torque should not exceed 0.03 Newton meters for diameters equal to or less than 1 inch (25 mm) and 0.04 Newton meters for diameters greater than 1 inch (25 mm).
7.2.6 TOGGLE SWITCHES The most common type of toggle switch is a two-state control used for on/off and start/ stop functions. Toggle switches with more than two discrete states greatly increase use errors, although they still might be optimal for some applications. In comparison to rocker switches, toggle switches are more suitable for rapid activation/deactivation of multiple minor functions. A potential problem with toggle switches, however, is confusion about
TAF-ER562X_WINGR-09-0401-C007.indd 267
11/8/10 11:43:31 PM
Handbook of Human Factors in Medical Device Design 0.8 ”
268
1.7”
30°
FIGURE 7.5
Proposed dimensions for toggle switches.
switch position. This can be addressed by keeping position coding consistent (e.g., up for on), providing clear labeling, and providing clear and immediate feedback. 7.2.6.1 Toggle Switch Geometry For length and width, the recommendations provided in Figure 7.5 should accommodate users with the largest index fingers. By selecting the larger extreme of the population, visually and physically impaired users are also accommodated by providing a larger contact surface area. 7.2.6.2 Toggle Switch Force As with push buttons, recommended force is 3.1 Newtons because toggle switches are activated via finger pushes, like push buttons. 7.2.6.3 Mounting Toggle Switches As with push buttons, placing toggle switches on horizontal or diagonal planes permits finer motor control than placing them on vertical surfaces. 7.2.6.4 Toggle Switch Layout The most common convention for vertically oriented toggle switches is for the top position to represent on and the bottom position to represent off. This translates into away for on and toward the user for off when the control is mounted on a horizontal surface. However, this convention might not hold worldwide. In the United States and much of the rest of the world, the right-hand position in horizontally oriented toggle switches represents on and the left position off. 7.2.6.5 Toggle Switch Travel 30 degrees of travel provides sufficient protection against inadvertent actuation while limiting the period of exertion, thereby accommodating physically impaired users. 7.2.6.6 Toggle Switch Feedback 1. To indicate proper lever positioning, all toggle switches should provide auditory and tactile feedback. Travel and elastic resistance are the most appropriate types of tactile feedback.
TAF-ER562X_WINGR-09-0401-C007.indd 268
11/8/10 11:43:33 PM
Controls
269
2. An audible snap should be used for auditory feedback. 3. To communicate proper functioning and activation status, a light can be used effectively to indicate switch state. 7.2.6.7 Inadvertent Activation of Toggle Switches Toggle switches are particularly prone to inadvertent activation. This can be mitigated by use of a raised collar (guard) or flip cover. 7.2.6.8 Toggle Switch Labeling Labeling should be provided on each side at the base of the toggle switch to indicate its position/state.
7.2.7 SMALL LEVERS Small levers are designed to be operated with the fingers as opposed to the whole hand. They are often associated with main power breakers and locking mechanisms. Levers are most effectively used to control alteration between two discrete states. Small levers are less easy to use when they have intermediate selection positions. 7.2.7.1 Lever Geometry 1. For height and width, as with other controls, the goal is to maximize the surface area in contact with the control and to accommodate the finger of the largest user (see Figure 7.6). 2. The control handle of a lever should have rounded edges to maximize user comfort. 7.2.7.2 Lever Force The recommended force is 3.1 Newtons. It is important to require sufficient force to provide adequate feedback and protect from overshoot. However, excessive force requirements will cause fatigue and can make lever actuation particularly difficult for users with disabilities. If higher force is needed, then the size of the lever should be increased to accommodate a full-hand grip or at least use by multiple fingers. 7.2.7.3 Mounting Levers Generally, lever motion should be restricted to the vertical or horizontal axis. Especially for elderly or arthritis-impaired users, diagonal movements can require difficult and uncomfortable angular flexion or extension of the elbow and shoulder. Top view 0.25”
9”
0.
0.9”
FIGURE 7.6
0.5”
Proposed dimensions for small levers.
TAF-ER562X_WINGR-09-0401-C007.indd 269
11/8/10 11:43:34 PM
270
Handbook of Human Factors in Medical Device Design
7.2.7.4 Lever Layout 1. For vertical orientations, up or away from the user should correspond with the on or activation setting (at least for North America and Western Europe). 2. In a horizontal orientation where movement is left to right, the left side should correspond with the “off” setting, again for users in North America and Western Europe. 7.2.7.5 Inadvertent Lever Activation Levers are particularly prone to inadvertent operation, so there should be some type of protection, such as requiring a double action (e.g., short movement 90 degrees relative to the activation motion in order to enable movement of the lever) or mechanical guarding. 7.2.7.6 Lever Labeling 1. Setting labels should be placed where they will not be covered by the user’s hand during lever operation. 2. Lever position should clearly demarcate the selected setting.
7.2.8 ROCKER SWITCHES The two-state design of rocker switches generally confines these controls to on/off or start/ stop functions (see Figure 7.7). Rocker switches generally retain the state to which they are placed until actuated again. However, they can also be spring-return momentary-contact controls that activate a function in one or two directions only while affirmatively pressed. The design of rocker switches prevents accidental activation by removing protruding handles that are present in toggle switches and levers. However, the current state of a rocker may not be as obvious as that of a toggle or lever, so it is important to provide clear feedback about the rocker switch’s position. Rocker switches can be used for more than two positions with adequate detents, but they are easier to use with only two discrete states. 7.2.8.1 Rocker Switch Geometry Length and width recommendations are designed to accommodate the index finger of the largest men, which also accommodates visually and physically impaired users by providing a large contact surface area. 7.2.8.2 Rocker Switch Force As with push buttons, the recommended force is 3.1 Newtons. Linear force should be determined at the middle position of the rocker. 7.2.8.3 Mounting Rocker Switches As with push buttons, rocker switches mounted on the horizontal plane permit finer motor control. ”
0.8
0.9
”
Cross section 30° 1”
FIGURE 7.7
Proposed dimensions for rocker switches.
TAF-ER562X_WINGR-09-0401-C007.indd 270
11/8/10 11:43:34 PM
Controls
271
7.2.8.4 Rocker Switch Layout 1. For vertically oriented rocker switches, the top position represents on. 2. For left-to-right positioning of rocker switches, the right-hand position represents on. 7.2.8.5 Rocker Switch Travel A 30-degree “tilt” provides sufficient travel and resistance without requiring prolonged exertion by physically impaired users. 7.2.8.6 Rocker Switch Feedback Rocker switches should provide tactile as well as auditory or visual feedback. Impaired users particularly benefit from clear, unambiguous feedback. An audible click conveys the proper placement or “locking in” of rocker switches. Visual feedback (e.g., onset of a light within or adjacent to the switch) can also be used to indicate the control state. 7.2.8.7 Rocker Switch Labeling Labeling should generally be provided at either end of the long axis to indicate the position/ state of the rocker switch.
7.2.9 SLIDERS Sliders are suitable for coarse operating functions such as environmental control. Their inadequacy in error- and time-sensitive operations arises from sliders’ dependency on total arm movement, which is relatively inaccurate. Sliders can be used for discrete or continuous functions. If multiple sliders are used, adequate space must be provided to protect from inadvertent operation. They can be mounted together when multiple sliders are intended to be operated simultaneously. In such situations, sliders should generally be mounted so that each slider can be operated with a separate finger (at least 1 inch [25 mm] center to center to accommodate large fingers). 7.2.9.1 Slider Geometry 1. Length/depth: A 0.9- by 0.9-inch (20 × 20 mm) surface area accommodates the largest fingers of the largest users (see Figure 7.8). It is important that the control not require an affirmative grip (i.e., that the user grasp it) but allow pushing by various parts of the hand, including finger or thumb. 2. Width: The wedge shape accommodates alternative preferred finger grip spans. 3. The gripped handle of a slider should have rounded edges to minimize user discomfort. Top view 0.25”
0.9” 0.9” 0.5”
FIGURE 7.8
Proposed dimensions for sliders.
TAF-ER562X_WINGR-09-0401-C007.indd 271
11/8/10 11:43:34 PM
272
Handbook of Human Factors in Medical Device Design
7.2.9.2 Slider Force A force of 3.1 Newtons will accommodate impaired users and protect against overshoot. 7.2.9.3 Mounting Sliders Sliders on the horizontal plane permit greater accuracy by enabling users to rest the wrist and use finger movements for control. The vertical plane requires movement of the arm for control. A diagonal plane (vertical slant) can be a good compromise. 7.2.9.4 Slider Layout 1. The on position or the high value of a scale should be at the top or away from the user (for North American and Western European users) for sliders that move in the vertical or horizontal planes, respectively. 2. The on position or the high value of a scale should be on the far right for sliders that move in the x-axis, at least in North America and Western Europe. 3. If elderly or impaired people are potential users, ample space should be provided around the control (3 inches [75 mm] below or to the right for right-handed users) so that users can rest their hands. Users achieve greater accuracy by means of finger rather than elbow and forearm movement. 7.2.9.5 Slider Feedback 1. Slider positions should be unambiguous. 2. Generally, a slider should contain a pointer that is in direct contact with the label for the chosen position. 3. Tactile or auditory feedback facilitates use, especially by visually impaired users. 7.2.9.6 Slider Labeling Labeling should be placed to the side and/or at either end of the slider movement path.
7.2.10 KEY-OPERATED CONTROLS Key-operated controls are used in situations where restricted access to the controls is important. They are appropriate for use with a small number of discrete states, preferably two or three (see Figure 7.9).
FIGURE 7.9
Key-operated control.
TAF-ER562X_WINGR-09-0401-C007.indd 272
11/8/10 11:43:34 PM
Controls
273
7.2.10.1 Key/Lock Geometry The key of key-operated controls provides the surface that is pinched between the thumb and index finger for rotation. The surface should be large enough to maintain significant contact with the thumb and index finger—at least 1 inch (25 mm) for both the vertical and the horizontal grip surface of the key. 7.2.10.2 Key Control Force The torque can be somewhat larger than that of rotary controls because of the relative effectiveness of the “key-pinch” grip. Required torque of less than 2 in.-lb. (0.23 N·m) will be comfortable to use for the majority of users. However, as with rotary controls, limiting the required torque to 0.12 Newton meters or less will accommodate many disabled users while still providing adequate tactile feedback. 7.2.10.3 Key Control Mounting Key-operated controls are generally easier to operate if the key extends horizontally from an insertion point mounted on a vertical surface. Inserting keys into horizontal surfaces requires awkward wrist movement unless the insertion point is mounted low relative to the user (e.g., below the waist when the operator is standing). 7.2.10.4 Key Control Feedback Key control positions should be unambiguous. Both tactile and visual indications of key state are important. Audible feedback may also be useful. 7.2.10.5 Key/Lock Labeling Labeling should clearly indicate the functions associated with different key positions.
7.2.11 FINGERPRINT READERS For security applications, both swipe and touch fingerprint readers are available. These have the advantage of eliminating the need for user entry of complex password text, for example. The device should be designed to allow use of the thumb or fingers of either hand.
7.2.12 MEMBRANE CONTROLS AND KEYPADS Membrane controls are suitable for noncritical operations (because of the absence of tactile feedback) in environments that require frequent cleaning and/or are exposed to fluids. Membrane controls are typically in the shape of buttons and can be assembled into a full keypad. Membrane controls are appropriate for numerical entries or commands requiring a sequence of input data. 7.2.12.1 Membrane Button Geometry 1. For width and length, the dimensions are designed to accommodate the index finger of the largest users (see Figure 7.10). This size also accommodates visually and physically impaired users by providing a large contact surface. 2. The minimum dimension of membrane controls should be 0.8 inches (20 mm). Round, oval, lozenge shaped, rectangular, or square are all acceptable.
TAF-ER562X_WINGR-09-0401-C007.indd 273
11/8/10 11:43:35 PM
274
Handbook of Human Factors in Medical Device Design
0.8”
0.8” 0.25”
FIGURE 7.10
Proposed dimensions for membrane controls.
3. A separation of 0.25 inches (6 mm) provides room for error, creating a target width of at least 1.3 inches (32 mm). Membrane switches can generally have less separation than conventional buttons because the former are more likely to be used in complex, sequential fashion. 7.2.12.2 Treatment of the Surface of Membrane Controls Providing tactile cues on a membrane surface makes it easier to use, particularly for visually impaired users. Providing ridges or raised surfaces or texture changes that indicate the shape of the buttons helps users operate them by touch only. 7.2.12.3 Membrane Button Force The appropriate actuation force for a membrane control is 0.2 Newtons. The recommended force for keypads is lower than for push buttons because hand posture is typically more like that of typing, so the muscles of the upper arm are not involved as they are for a normal push button. However, the lower force might increase vulnerability to inadvertent activation. 7.2.12.4 Mounting of Membrane Controls Membrane controls are generally easier to use if mounted on a horizontal rather than vertical surface. 7.2.12.5 Layout of Membrane Controls 1. Logically clustering membrane keys in an area below or beside visual displays creates clear groupings. 2. Spatial groupings, graphic indication, and tactile cues, such as raised borders between groups of key categories, reduce user confusion when multiple membrane keys are clustered. 3. Providing additional space between membrane-key groupings reduces accidental actuation and provides space for users to rest their fingers. 7.2.12.6 Numeric Keypads For numeric keypads, the telephone-pad layout (see Figure 7.11) should be used unless user needs and testing indicate otherwise. The telephone keypad layout, in comparison to the calculator layout or the linear keyboard layout, achieves slightly greater user speed and lower error rates because of a more natural and recognized mental mapping. The human tendency (at least in North America and Western Europe) is to read from top to bottom and begin counting at one. The telephone keypad layout is required by some national and international standards (including ISO 9995, which specifies keyboard layouts for office systems, and IEC 62366).
TAF-ER562X_WINGR-09-0401-C007.indd 274
11/8/10 11:43:35 PM
Controls
275
0.8”
0.8” 0.25”
FIGURE 7.11
Preferred telephone-style keypad layout.
However, in situations in which the keypad is used in a manner akin to a calculator, the calculator layout may be considered. User testing would be indicated in such cases. 7.2.12.7 Feedback from Membrane Controls 1. Visual and auditory feedback is important for membrane controls. Although some tactile feedback (e.g., low-travel “snap domes”) should be provided, if possible, the relative lack of tactile feedback makes visual and auditory feedback more important. Tones are effective forms of auditory feedback and are best presented using frequencies between 400 and 1,500 Hz. 2. Feedback should be instantaneous. Even a slight delay can cause repeated pressing. 3. “Press-and-hold” repeat time should be 0.09 seconds. A delay greater than 0.10 seconds causes repeated pressing. Fast typists can type two characters in 0.06 seconds. 4. As with push buttons, back illumination can be effective in creating a direct association between user action and activation of the control. Back illumination is generally not appropriate for numeric input pads, however, because of the rapid succession of entries. Numeric input pads should be accompanied by a visual display that provides immediate feedback. 7.2.12.8 Membrane Control Labeling A clear, uncluttered graphic label in the center of the key provides the most effective type of labeling. 7.2.12.9 Differentiating Membrane Controls Shape coding is effective and can be used redundantly with color, position, and/or labeling.
7.2.13 TOUCH SCREENS Touch screens can be implemented using various technologies, all of which have in common the integration of electronic visual displays, labels, and touch zones programmed to be
TAF-ER562X_WINGR-09-0401-C007.indd 275
11/8/10 11:43:35 PM
276
Handbook of Human Factors in Medical Device Design
TABLE 7.2 Touch-Screen Control Selection Matrix Touch-Screen Control Elements Functional Characteristics
Radio Buttons Check Boxes List Boxes
Easy access to control selections Appropriate for mutually exclusive options Provide intuitive mapping Provide unlimited number of choices Allow options to remain visible May require additional scrolling Provide great flexibility Can hide options Can support complex functions Less space efficient Limited selection options Can compromise intuitiveness of operation Limited precision
✓ ✓ ✓ ✓
Sliders or Scales
✓ ✓ ✓ ✓
✓ ✓ ✓ ✓ ✓
✓ ✓
Spin Boxes
✓ ✓
✓ ✓ ✓ ✓ ✓ ✓
✓ ✓
✓ ✓
control actuators. Touch-screen applications are diverse, although touch screens should not require prolonged control activation or precise control movements. Table 7.2 summarizes the characteristics of the various types of touch-screen controls, all of which are depicted in Figure 7.12. Touch screens are particularly appropriate for applications in which the following are true: • Menu selections are required. • The users’ focus is on the display.
FIGURE 7.12
Alternative touch-screen controls.
TAF-ER562X_WINGR-09-0401-C007.indd 276
11/8/10 11:43:35 PM
Controls
• • • •
277
It is time consuming or dangerous to divert attention from the display. The workload can be reduced with a limited number of inputs. Potential users are relatively inexperienced. The device will be used in a high-stress environment.
7.2.13.1 Advantages of Touch Screens 1. The only movements required are natural pointing and finger-sliding gestures. 2. The touch screen provides a direct relationship between hand and pointer movement in terms of direction, distance, and speed. 3. The possible inputs are limited by what is displayed on the screen. Thus, no memorization of commands is required. 4. The limited number of possible inputs also minimizes input errors. 5. Users require only minimal training. 6. Additional surface space is not required, unlike most other pointing devices. 7. Touch screens are durable in high-use environments. 7.2.13.2 Disadvantages of Touch Screens 1. There can be considerable fatigue with touch-screen use over extended periods of time. 2. If a finger is the touch mechanism, the finger may obscure part of the screen (a stylus is usually more accurate than a finger). 3. A touch screen may make selecting small items difficult or even impossible. 4. A finger may be too large to be accurate with small objects, particularly individual characters. 5. Complex data input may be slower, especially if only one finger is used. 6. Users can leave smudge marks and biologically contaminate the screen. 7. Touch screens may be more difficult for some disabled users (e.g., those with poor eye–hand coordination or tremor) to use. 8. Touch screens can be damaged if physical objects (e.g., pen tips) rather than fingers are used. 9. Directional, continuous-control functions can be more difficult to implement on a touch screen (vs. sliders or rotary knobs). 7.2.13.3 Touch-Screen Geometry 1. The height and width of the actuation areas should be within the range of 0.6 to 1.5 inches (15 to 38 mm) with a spacing of 0.1 to 0.25 inches (3 to 6 mm) between adjacent actuation areas (see Figure 7.13). Touch zones smaller than approximately 0.9 inches (23 mm) will be associated with greater use errors. Similarly, decreased “dead space” between keys will increase the incidence of unintentional activation of adjacent touch zones. Error reduction software algorithms (e.g., temporary disabling of adjacent objects once a given object is selected) can allow decreased between-key spacing. 2. Touch areas larger than 1.2 by 1.2 inches (30 × 30 mm) provide greater accuracy and will minimize data entry errors. It can be helpful to make the active area larger than the visible target provided on the screen. 3. Regardless of key geometry, center-to-center key spacing should not be less than 0.8 inches (20 mm).
TAF-ER562X_WINGR-09-0401-C007.indd 277
11/8/10 11:43:35 PM
278
Handbook of Human Factors in Medical Device Design 0.25”
1”
0.25”
1”
FIGURE 7.13
Ideal dimensions for “active areas” on touch screens.
7.2.13.4 Touch Zone Force For direct-touch controls, activation resistance should be within the range of 0.9 to 5.3 ounces (0.25 to 1.5 N). Force should be adjusted, where necessary, to prevent inadvertent activation. 7.2.13.5 Touch Zone Activation 1. “Up-triggers” (activate-on-release) are generally preferable to “down-triggers” (activate on initial touch). Activating items when the finger is removed decreases errors because it gives users the opportunity to slide their finger off the button, effectively canceling their selection. A good approach is to highlight an item when touched and then execute the choice when the finger is removed. 2. To facilitate activation and decrease user confusion, the entire area of a button should be touchable. Buttons with larger touchable areas (even an area larger than the visually demarcated button) produce less user confusion about the location of a valid touch. 3. In places where accurate target selection is required, the target is very small, or the user may be off angle from the touch screen, visible crosshairs indicating the location of touch can be helpful. 7.2.13.6 Touch-Screen Feedback 1. By clearly indicating which options are selected, highlighting decreases use error rates and compensates for the lack of tactile feedback. 2. Shape or color coding permits differentiation of touch zones (active areas) from inactive zones (informational text and background graphics). 3. Visual effects (e.g., concave and convex surfaces) can be used to indicate button status, such as pressed and unpressed. 4. Touch screens should provide auditory feedback (e.g., a soft click or beep) to indicate activation or selection input. Auditory feedback, including speech, is particularly helpful for users who have a visual impairment. 5. Frequently operated touch screens should provide the user with the option of muting tonal signals to prevent sound distraction or redundancy.
TAF-ER562X_WINGR-09-0401-C007.indd 278
11/8/10 11:43:35 PM
Controls
279
6. As with membrane controls, buttons on touch screens should provide immediate feedback with a repeat delay (with continuous pressing) of 0.09 seconds. 7. “Touch mice” are cursors (often crosshairs or arrows) that are controlled by the finger. They decrease mistakes in finger placement and provide clear feedback on positioning. They are typically used with up-triggers so that the user can clearly see the to-be-controlled function and touch location prior to lifting the finger to activate a choice. 7.2.13.7 Touch-Screen Labeling Placement of labels in the center of touchable areas increases proper activation. Users are drawn to labels and have a tendency to touch and select them. Labels offset from touchable areas can cause confusion and frustration, leading to use errors. There are a number of design issues with touch-screen test labels (see Chapter 8, “Visual Displays,” and Chapter 11, “Software User Interfaces,” for more information on recommended text parameters). 1. A pixel-based display with maximum contrast usually provides excellent legibility. 2. Character height of 10 points or greater with a stroke width-to-character height ratio of 1:6 is optimal. Character height of 18 points or more is preferable for persons with reduced visual acuity. 3. Letter spacing within words should be between 0.5 and 1.5 times stroke width. 4. Simple sans serif fonts enhance legibility. 5. Short text length reduces confusion and recognition time. The recommended maximum length of text labels is 40 to 60 characters. 6. Use of upper- and lowercase letters for body text and all caps for headings may be an effective method to differentiate text messages from titles and headings. 7. Grouping alphanumeric strings according to accepted and readily used formats increases clarity of text. 8. Terminology should be familiar to the intended users. 9. Structured narrative increases ease of scanning and identification.
7.2.14 STANDARD KEYBOARDS The keyboard is the most efficient way to enter alphanumeric information. Extended keyboards can provide additional characters for special applications. Keyboards are particularly appropriate where there is a need for large amounts of alphanumeric, particularly alphabetic, input and when one-handed typing is not necessary. 7.2.14.1 Advantages of Keyboards 1. No other commonly available input device allows experienced typists to enter alphanumeric data as quickly. 2. Keys can be used for various tasks (e.g., arrow keys, function keys). 3. Function keys or “key equivalents” of menu choices can speed use for skilled users. 7.2.14.2 Disadvantages of Keyboards 1. The speed advantage for text input is less significant when users are unskilled typists. 2. For optimal use, both hands are required, reducing the positional flexibility of the user.
TAF-ER562X_WINGR-09-0401-C007.indd 279
11/8/10 11:43:36 PM
280
Handbook of Human Factors in Medical Device Design
3. Keyboards are vulnerable to fluid ingress. 4. Keyboards require large horizontal surfaces. 5. The keyboard can be slower than other devices for pointing or selecting. 7.2.14.3 Keyboard Geometry 1. The standard keyboard is preferable to miniaturized or membrane keyboards because it offers greater comfort and accuracy. While alternative designs can be used, their usability should be tested. 2. Travel: Each key should travel no less than 0.05 inches (1 mm) and no greater than 0.25 inches (6 mm). 3. Shape: Keys may be circular, square, or rectangular with a dished (i.e., concave) top. Keys should not have any sharp edges. 4. Diameter: For barehand operation, the diameter of each key should exceed 0.4 inches (10 mm) but should not exceed 0.75 inches (19 mm). The preferred target diameter is 0.5 inches (13 mm). 5. Separation: As with membrane controls, a separation of 0.25 inches (6 mm) provides room for error while minimizing the required finger reach between keys. 6. Location: The keyboard should be located horizontally or at approximately the level of the elbow of the seated user and directly in front of the user. Where possible, the keyboard should be detachable to permit the user to place it in a position that facilitates comfortable typing. 7.2.14.4 Keyboard Layout 1. The key layout should match the intended user population convention. For all English-speaking countries and most other European countries, the standard QWERTY layout should be used, although there are small differences in conventions from one country to the next. 2. A separate number keypad should be provided if device interactions call for extensive numerical input. 3. Keyboards should not include inactive keys. 4. Fixed-function keys (e.g., enter) should be used for time-critical, error-critical, or frequently used control input. They should be labeled clearly and conspicuously. 5. Fixed-function keys should have a consistent meaning throughout a given system. 7.2.14.5 Keyboard Force The force required for key activation should not exceed 5 ounces (1.4 N) but should be at least 2 ounces (0.6 N) to reduce inadvertent activation. The recommended force for keyboard keys is lower than for push buttons because the optimal position of the user’s hands eliminates the use of upper-arm muscles, requiring the user to rely on finger strength. 7.2.14.6 Keyboard Feedback 1. Activation of each key should result in a user-observable action. In many applications, audible feedback is also desirable. 2. When fixed-function key activation does not result in an immediately observable natural response (e.g., causing an alarm tone to silence), some form of system acknowledgment or feedback should be provided to the user.
TAF-ER562X_WINGR-09-0401-C007.indd 280
11/8/10 11:43:36 PM
Controls
281
3. Except for toggling between two alternate states, a fixed-function key should require only a single activation to execute a function. 7.2.14.7 Key Labeling As with button and membrane controls, a clear, uncluttered label in the center of the key provides the most effective type of labeling.
7.2.15 MICE A mouse establishes a direct relationship between the hand and pointer movement in terms of distance, direction, and speed. Mice are particularly appropriate for simple pointing rather than the generation of free-drawn graphics, for example, where simultaneous selection of multiple objects is necessary. 7.2.15.1 Advantages of Mice 1. The selection methods (“buttons”) are included as part of the mouse. A scroll wheel can also be incorporated. 2. One’s view of on-screen elements is not obscured when using a mouse. 3. Wireless optical mice do not require a cable connected to the device. 4. The possible inputs are limited by what is displayed on the screen. Therefore, command memorization is not required. 7.2.15.2 Disadvantages of Mice 1. A hand must be removed from the keyboard in order to use the mouse. 2. A mouse requires a horizontal surface of at least 100 cm2. 3. At times, without substantial use experience, long movement distances may be required. 4. The use of the mouse requires some degree of eye–hand coordination. 5. A mouse is not ideal where the “rolling” surface might be contaminated by fluid. 7.2.15.3 Mouse Geometry and Design Attributes 1. A mouse should accommodate both right- and left-handed users. 2. A mouse should be shaped approximately as a rectangular solid and should have no sharp edges. 3. The width should be no less than 1.6 inches (40 mm) and no more than 2.8 inches (70 mm) (see Figure 7.14).
2.8” to 4.7” 1.0” to 1.6”
1.6” to 2.8”
FIGURE 7.14
Proposed dimensions for mice.
TAF-ER562X_WINGR-09-0401-C007.indd 281
11/8/10 11:43:36 PM
282
Handbook of Human Factors in Medical Device Design
4. The length should be no less than 2.8 inches (70 mm) and no more than 4.7 inches (120 mm). 5. The height should be no less than 1.0 inches (25 mm) and no more than 1.6 inches (40 mm). 6. Regarding location, a clear, flat, horizontal surface should be available on which the mouse may be used. The working height should be approximately at elbow level of the user. 7. A mouse should have one or more buttons that are operable without diminishing control of the mouse. Button contact surfaces should be perpendicular to displacement direction and finger motion during actuation. 8. The mouse should move easily in any direction without requiring the user to readjust his or her hand position. 7.2.15.4 Feedback from Mouse Activation 1. “Gain,” or the relationship between mouse movement and cursor movement, should be adjustable. 2. For a small screen object that requires fine mouse positioning, a “hot zone” or clicking area should be provided around it. 3. Any application that allows the user to drive the cursor off the edge of the display should provide indicators to assist the user in bringing it back. 4. If the user grasps the mouse in what seems to be the correct orientation and moves it rectilinearly along what is assumed to be straight up the y-axis, then the direction of movement of the cursor on the display should be vertical, within 10 degrees either way. 7.2.15.5 Other Design Requirements for Effective Mouse Use 1. Double clicks should not be required as the only way to accomplish essential operations. Rapid double pressing is difficult for some people and may exacerbate musculoskeletal disorders. 2. Mouse-plus-keystroke combinations should be used only for selection of multiple items from a list. 3. The user should not be required to track a moving target.
7.2.16 STYLI AND LIGHT PENS Styli and light pens establish a direct relationship between hand and pointer movement in terms of direction, distance, and speed and are also classified as ITL pointing devices because the control is on the same plane as the pointer. Light pens are typically used directly with visual displays or with a separate touch pad. Styli and light pens are particularly appropriate where complex objects, graphical input, or handwriting have to be generated or where there is insufficient room for a mouse. 7.2.16.1 Advantages of Styli/Light Pens 1. Styli and light pens are smaller than the finger for touch-screen applications, enabling users to select smaller targets or make fine-grained movements. 2. The only movement required is a natural pointing gesture. 3. The possible inputs are limited by what is displayed on the screen. Therefore, command memorization is not required.
TAF-ER562X_WINGR-09-0401-C007.indd 282
11/8/10 11:43:36 PM
Controls
283
4. Styli and light pens take advantage of users’ experience with ordinary writing implements. 7.2.16.2 Disadvantages of Styli/Light Pens 1. Light pens can be fatiguing to use over an extended period of time, particularly direct-touch light pens. 2. A place for the stylus or light pen to rest and to be stored must be provided. 3. The hand must be moved away from the keyboard in order to use a light pen. 4. Direct-touch pens can damage the screen if too much force is inadvertently applied by the user. 5. Styli and light pens are suboptimal for users suffering from tremor or with other impairments of hand control. 7.2.16.3 Styli/Light Pen Geometry and Design Considerations 1. The stylus or light pen should take the approximate form of a conventional pen. 2. The stylus/light pen should be between 4.7 and 7.1 inches (120 to 180 mm) in length and should be between 0.3 and 0.8 inches (8 to 20 mm) in diameter (Figure 7.15). 3. A convenient mounting or storage device (e.g., clip, holder, etc.) that prevents the stylus or light pen from becoming lost should be provided. 4. The contact surface of a selector button that is mounted on a stylus or light pen should have a diameter of at least 0.2 inches (5 mm). 5. If a separate tablet is provided, it should be located to maintain spatial correspondence with the display. That is, the left and right sides of the tablet should correspond to the left and right sides of the screen and the furthermost side of the tablet should correspond to the top of the screen. 6. The screen surface of an accompanying tablet should be within the comfortable reach envelope of the intended user. The working height should be at the elbow level of the seated user. 0.3” to 0.8”
4.7” to 7.1”
FIGURE 7.15
Proposed dimensions for styli and light pens.
TAF-ER562X_WINGR-09-0401-C007.indd 283
11/8/10 11:43:36 PM
284
Handbook of Human Factors in Medical Device Design
7.2.16.4 Forces for Styli/Light Pens 1. The stylus/light pen should weigh more than 0.35 ounces (10 g) but less than 0.875 ounces (25 g). 2. The maximum force required for continuous input on a tablet should not exceed 2.9 ounces (0.8 N). 3. The force required to actuate a button should be greater than 1 ounce (0.3 N) but less than 2.9 ounces (0.8 N). 4. For direct-touch light applications, a push-tip switch requiring 2 to 5 ounces (0.6 to 1.4 N) of force for activation is usually preferred. 7.2.16.5 Feedback from Stylus/Light Pen Activation 1. An on-screen cursor should be displayed to indicate the location and movement of the control. The movement of the stylus across the surface grid should result in the cursor moving in the same direction, at the same rate, and with a smooth motion. When used as a two-axis controller, contact with the tablet surface should cause the cursor to appear at the designated coordinates on the display and to remain there until the pen is removed. 2. Visible feedback should be designed to be easily seen under the point of the stylus/ pen. 3. Audible feedback should be provided as indicated by functional requirements or user feedback.
7.2.17 TRACKBALLS Trackballs are functionally similar to mice. Their ability to continuously scroll without spatial constraints makes them well suited for rapid cursor positioning and for moving onscreen objects. Trackballs are particularly appropriate when the following are true: • • • •
Only cursor positioning is required. Long cursor travel in one direction is required. Precise tracking along a curving path is not required. Surrounding surface area is at a premium.
7.2.17.1 Advantages of Trackballs 1. A trackball requires less space than a mouse, particularly because it can be integrated into a keyboard or device. 2. Trackballs offer a direct relationship between hand and pointer movement in terms of direction and speed. The ball’s speed and rotation provide direct tactile feedback. 3. Vision of on-screen elements is not obscured when using a trackball. 4. Trackballs are likely to cause fewer musculoskeletal problems than mice. 7.2.17.2 Disadvantages of Trackballs 1. There is no direct relationship between the hand and pointer movement in terms of distance.
TAF-ER562X_WINGR-09-0401-C007.indd 284
11/8/10 11:43:36 PM
Controls
285
2. The hand must be removed from the keyboard keys, particularly when the trackball is not integrated into the keyboard. 3. Some people find a trackball difficult to learn to use skillfully. 4. Trackballs can be vulnerable to fluid spills. 5. Track balls can be less durable than mice. 7.2.17.3 Trackball Geometry 1. Trackballs should be mounted in a horizontal rather than vertical orientation. 2. A trackball should be capable of rotation in any direction in order to generate any combination of x and y cursor movements. 3. Wrist support and/or arm support should be provided when trackballs are used to make precise or continuous adjustments and to prevent musculoskeletal problems. 4. The ball should have a minimum diameter of 2 inches (50 mm) and a maximum diameter of 6 inches (150 mm), with a preferred target diameter in the middle at 4 inches (100 mm) (see Figure 7.16). 5. The ball should have a minimum clearance of 2 inches (50 mm) on all sides and should be at least 4.75 inches (120 mm) from the edge of the supporting surface. 7.2.17.4 Forces for Trackballs Forces can range from a minimum of 0.9 ounces (0.25 N) to a maximum of 5.4 ounces (4.5 N), but the preferred force is 1.1 ounces (0.3 N). Initial resistance to rotation should range from 0.9 to 1.4 ounces (0.25 to 4 N). 7.2.17.5 Trackball Feedback 1. A cursor should be displayed on the screen to indicate the location and movement of the control. 2. A movement of the ball should result in a smooth corresponding movement of the display cursor. A movement in one direction should correspond to the same movement of the cursor. 3. When the control is moved in the x or y direction alone, no cross-coupling (i.e., cursor movement in the orthogonal direction) should be apparent. 4. Backlash (tracking in a direction opposite to a rapid movement) should be prevented. 5. Control/cursor movement ratios should facilitate both rapid gross positioning and smooth precise positioning. 6. Any application that allows the user to drive the cursor off the edge of the screen should provide indicators to assist the user in bringing it back onto the screen. 100° - 140°
2”
FIGURE 7.16
2” - 6”
4¾” - 9¾”
Proposed dimensions for trackballs.
TAF-ER562X_WINGR-09-0401-C007.indd 285
11/8/10 11:43:36 PM
286
Handbook of Human Factors in Medical Device Design
7.2.18 DISPLACEMENT JOYSTICKS Displacement joysticks involve hand or finger movement of a lever to control a cursor. Displacement joysticks can control three dimensions by moving in the x and y directions as well as via shaft rotation. Displacement joysticks are particularly appropriate for the following: • Tasks that require continuous control in two or three dimensions • Tasks that do not require a high degree of precision 7.2.18.1 Advantages of Joysticks 1. Joysticks occupy a small amount of space, and they can be mounted on a keyboard. 2. Over extended periods of time, joysticks may be less fatiguing than mice or trackballs. 7.2.18.2 Disadvantages of Joysticks 1. Joysticks tend to have relatively low accuracy and resolution. 2. Joysticks are difficult to use for drawing or tracing tasks. 7.2.18.3 Joystick Geometry 1. Hand-operated displacement joysticks should be between 4.3 and 7.1 inches (110 to 180 mm) long and no more than 2 inches (50 mm) wide. 2. Finger-operated displacement joysticks should be between 3 and 6 inches (75 to 150 mm) long and between 0.25 and 0.63 inches (6 to 16 mm) wide (see Figure 7.17). 3. Both hand- and finger-operated displacement joysticks should have a clearance of at least 4 inches (100 mm) to the side and 2 inches (50 mm) to the rear for full hand movement. 7.2.18.4 Joystick Force The force should be greater than 0.75 pounds (3.3 N) but should not exceed 2 pounds (8.9 N). 7.2.18.5 Joystick Feedback To provide adequate tactile feedback while minimizing required hand motion, the control should have a maximum displacement of 45 degrees from the center position in any direction. 90° 0.25” - 0.625”
3” - 6”
FIGURE 7.17
Proposed dimensions for finger-operated displacement joysticks.
TAF-ER562X_WINGR-09-0401-C007.indd 286
11/8/10 11:43:36 PM
Controls
287 0.25” - 0.625”
3” - 6”
FIGURE 7.18
Proposed dimensions for isometric joysticks.
7.2.19 ISOMETRIC JOYSTICKS The hand-operated isometric joystick has no perceptible movement, but its output is a function of the force applied (see Figure 7.18). Isometric joysticks are particularly appropriate for tasks that require the following: • Precise return to a neutral-centered position after each use. • Fine control of a system’s reaction (i.e., an arm of an industrial robot). 7.2.19.1 Advantages of Isometric Joysticks 1. As with displacement joysticks, isometric joysticks occupy a small amount of space. 2. The reliance on force rather than finger and/or forearm motor control enables isometric joysticks to attain a higher degree of precision and control than displacement joysticks. 7.2.19.2 Disadvantages of Joysticks 1. Isometric joysticks do not provide direct tactile feedback regarding the level of force exerted by users. This can be partially mitigated, however, by audible and/or visual feedback. 2. They require training and experience to achieve full and precise control. 7.2.19.3 Isometric Joystick Geometry Hand-operated isometric joysticks should be between 4.3 and 7.1 inches (110 to 180 mm) long and no more than 2 inches (50 mm) wide. 7.2.19.4 Isometric Joystick Force The force should be greater than 0.75 pounds but should not exceed 2 pounds.
TAF-ER562X_WINGR-09-0401-C007.indd 287
11/8/10 11:43:36 PM
288
Handbook of Human Factors in Medical Device Design
7.2.19.5 Isometric Joystick Feedback Feedback requirements for isometric joysticks are similar to those for displacement joysticks. However, audible and/or visual feedback is important because isometric joysticks do not provide tactile feedback.
7.2.20 OTHER INPUT DEVICES 7.2.20.1 Pointer Sticks Pointer sticks generally underperform mice but can be integrated into portable computers. In addition, pointing sticks are sometimes more efficient when used for mixed typing and pointing activities because of the time reduction for switching between the pointing stick and the keyboard. Pointer sticks can be difficult to control and are more difficult to learn to use than mice or styli. 7.2.20.2 Touch Pads Touch pads generally underperform mice but, like pointer sticks, can be integrated into portable computers. Generally, performance with a touch pad is better than with a pointer stick. 7.2.20.3 Data Gloves Data gloves are equipped with sensors that sense the movements of the hand. This movement is interpreted by a computer that responds in predefined ways. For example, data gloves can be used in virtual reality environments where the user sees an image of the data glove and can manipulate the movements of the virtual environment using the glove. Forces and torque applied by the parts of the hand can also be sensed.
7.2.21 HANDS-FREE CONTROLS Hands-free controls are ideal when the user’s hands are occupied by concurrent tasks. For example, they are particularly beneficial in multitask situations such as driving a vehicle. They are also useful to individuals who do not have controlled use of their hands or other extremities as a result of illness or disability. Besides using their hands, users can activate controls through various other means discussed here. 7.2.21.1 Eye Tracking Eye tracking relies on a camera that focuses on one or both eyes and interprets the direction of the user’s gaze in relationship to the environment. It can be used to select, manipulate, and move objects in the visual field. 7.2.21.2 Directed Breathing “Sip and puff” controls are activated by blowing and sucking air through a tube. They typically only allow for basic on and off control. 7.2.21.3 Mouth Sticks Mouth sticks are pointer sticks that users hold between their lips and teeth. They can be used to push buttons or select items on a touch screen.
TAF-ER562X_WINGR-09-0401-C007.indd 288
11/8/10 11:43:36 PM
Controls
289
7.2.21.4 Head Movement Head movements are generally interpreted by a gyroscope. They are usually limited to simple on and off control. 7.2.21.5 Voice-Activated Control Voice-activated control presents the potential to replace all or some physical controls of a device. Voice-activated control is particularly suitable for devices requiring hands-free use or for accommodating impaired users. Current voice recognition systems can process continuous speech, eliminating the need for unnatural pauses. After a period of training and error correction, systems can achieve relatively high accuracy (upwards of 95%). Length of training is program specific and depends on the user’s speech characteristics. However, voice recognition accuracy will be compromised when there is background noise, poor clarity of pronunciation, fast speech, or unusual diction. Also, 95% accuracy might be insufficient for many applications.
7.3 LARGE MECHANICAL CONTROLS The controls described in this section are generally used when significant force is necessary.
7.3.1 CRANKS Cranks are well adapted to perform continuous adjustments. By converting weight resistance into a rotational force, cranks enable users to work and manipulate heavy loads. 7.3.1.1 Crank Geometry 1. The crank grip handle should turn freely around the shaft. The friction that results from fixed grip handles with heavy loads may injure the user or create unnecessary discomfort. 2. When cranks are used for fine-tuning or other functions involving numerical selection, each rotation should correspond to a multiple of 1, 10, 100, and so on. 3. Cranks for standing users should be positioned between 2.95 and 3.94 feet (0.9 to 1.2 m) above the floor, assuming horizontal mounting. 4. The surface area in contact with the hand should be optimized (minimum diameter of 0.5 inches [13 mm]; see Figure 7.19). The handle should turn freely about its shaft if rapid rotation is necessary. However, a fixed handle can enable more precise adjustment. 3”
”
3.8
0.5
”
FIGURE 7.19
Proposed dimensions for cranks for light loads.
TAF-ER562X_WINGR-09-0401-C007.indd 289
11/8/10 11:43:36 PM
290
Handbook of Human Factors in Medical Device Design
5. The handle should be 1 to 3 inches (25 to 75 mm) in length for light loads (under 5 pounds [2.3 kg]) and greater than 3 inches (75 mm) for heavy loads (over 5 pounds [2.3 kg]). The preferred length is 1.5 inches (38 mm) for light loads and 3.8 inches (100 mm) for heavy loads. 6. The handle should be between 0.4 and 0.625 inches (10 to 16 mm) in diameter for light loads and between 1 and 1.5 inches (25 to 38 mm) for heavy loads. The preferred diameter is 0.5 inches (13 mm) for light loads and 1.0 inch (25 mm) for heavy loads. 7. For tasks requiring rotations less than 100 revolutions per minute (rpm), the turning radius should be between 1.5 and 5 inches (38 to 125 mm) for light loads and between 7.5 and 20 inches (190 to 500 mm) for heavy loads. 8. For tasks above 100 rpm, the turning radius should be between 0.5 and 4.5 inches (38 to 115 mm) for light loads and between 5.0 and 9.0 inches (130 to 230 mm) for heavy loads. 9. A folding handle should be used, when necessary, to prevent the handle from posing a hazard to passersby. Such a handle should be spring loaded to keep it extended in the cranking position when in use and folded when not in use. 7.3.1.2 Crank Forces In general, increases in the degree of resistance will reduce the maximum turning rate. Inertia will assist in the maintenance of a constant rate of rotation. Maximum force will be a function of who the users are and what posture they can adopt. A conservative maximum force to accommodate the smallest users with poor posture is approximately 5.6 pounds (25 N).
7.3.2 HANDWHEELS Handwheels are suitable where two-handed operation is required because of high operating force (see Figure 7.20). The speed of operation is typically slower than achievable with cranks operated single-handedly. 7.3.2.1 Handwheel Geometry 1. Handwheels mounted vertically should have a diameter between 7.9 and 20 inches (200 to 500 mm).
FIGURE 7.20
Handwheel.
TAF-ER562X_WINGR-09-0401-C007.indd 290
11/8/10 11:43:37 PM
Controls
291
2. Handwheels mounted horizontally on the floor should have a diameter between 11.8 and 59.8 inches (300 to 1500 mm). 3. Handwheels mounted horizontally overhead should have a diameter between 7.9 and 15.7 inches (200 to 400 mm). 4. Effective use of a handwheel is limited to 60 degrees of arc, which is the maximum that can be achieved without removing the hands from the handwheel. 5. Whether mounted horizontally or vertically, the diameter of the wheel’s rim should be between 0.75 and 1.26 inches (19 to 32 mm). 7.3.2.2 Handwheel Force Operating force at the periphery of the handwheel should be below 245 N for two-handed operation or 127 N for one-handed operation. For small arcs of movement, inertial resistance must be minimized.
7.3.3 LARGE LEVERS For situations where heavy manual force is applied, large levers are most appropriate for selection among several discrete steps. Levers are also appropriate for controls that require multidimensional movement. 7.3.3.1 Lever Geometry 1. A shaft length of 7.2 inches (180 mm) will support two-handed use for the largest users (see Figure 7.21). 2. Handle shape can be cylindrical, spherical, or shaped to fit the hand grip as required but should have no sharp edges. 3. The minimum distance between controls should be 2 inches (50 mm) for onehanded operation and 3 inches (75 mm) for two-handed operation. The preferred separation is 4 and 5 inches (100 and 125 mm), respectively. 7.3.3.2 Lever Force 1. For one-handed forward operation, the force should not exceed 30 pounds (135 N). The maximum should be 20 pounds (90 N) if the user must exert a lateral force. 2. For two-handed forward operation, the force should not exceed 50 or 30 pounds (220 or 135 N) if the user must exert a lateral force.
7.2”
FIGURE 7.21
Large lever.
TAF-ER562X_WINGR-09-0401-C007.indd 291
11/8/10 11:43:37 PM
292
FIGURE 7.22
Handbook of Human Factors in Medical Device Design
Whole-hand-operated controls.
3. For both one- and two-handed operation in forward and lateral directions, the minimum force should be 2 pounds (9 N). 7.3.3.3 Lever Feedback 1. A lever should travel far enough to provide tactile feedback but should not move further than 14 inches (350 mm) forward or more than 38 inches (965 mm) laterally. 2. To indicate proper positioning, a lever should provide auditory and/or tactile feedback.
7.3.4 WHOLE-HAND-OPERATED CONTROLS Whole-hand-operated controls allow for higher force to be exerted and are often used for emergency situations in which the user has to operate the control reflexively but precision is not required (see Figure 7.22). 7.3.4.1 Hand-Operated Control Geometry 1. A control 3 inches (75 mm) in diameter will provide conspicuity and a large surface area of contact with the hand. 2. The control should not have any sharp edges. Edges should have a minimum radius of 4 mm, but a radius greater than 9 mm is preferred. 3. Round grips provide flexibility for the user regarding postural orientation with respect to the handle. 7.3.4.2 Hand Control Force Whole-hand push controls should not require more than 13 pounds (58 N) of linear force.
7.3.5 FOOT CONTROLS Foot-operated controls are most appropriate when the user’s hands are committed to performing other tasks (see Figure 7.23). In addition, foot controls allow for the application of greater force than the upper body can produce comfortably. Foot controls can be used by a
1 - 2.5”
0.5” (minimum)
FIGURE 7.23
Foot control.
TAF-ER562X_WINGR-09-0401-C007.indd 292
11/8/10 11:43:37 PM
Controls
293
sitting or standing user. However, the disadvantages are that they often cannot be seen during operation, and they generally do not allow the same level of precision as hand-operated controls. 7.3.5.1 Foot Control Geometry and Location 1. Foot controls can be a variety of shapes, but in some applications, they should have flanges to support the heel or toe caps in order to provide a means of keeping the foot in contact with the control even when the control is inactivated. 2. Generally, nonslip surfaces should be used. 3. Horizontal separation should be a minimum of 3 inches (75 mm). Vertical separation should be a minimum of 8 inches (200 mm) when users must use multiple foot controls with the same foot. However, multiple foot controls should be avoided if possible. 4. Placement should allow the user to center the control under the toe and the ball of the foot. Operation by the heel of the foot is undesirable. 7.3.5.2 Foot Control Force 1. If the foot is not allowed to rest on the control, keeping the minimum force above 4 pounds (18 N) will accommodate the vast majority of users while providing adequate tactile feedback. 2. If the foot is allowed to rest on the control, keeping the minimum force above 10 pounds (45 N) will help to prevent accidental activation. 3. Forces required should not exceed 20 pounds (90 N). 7.3.5.3 Foot Control Feedback 1. Foot controls should provide tactile and either auditory or visual feedback to indicate that that the user has successfully activated the control. 2. To provide adequate tactile feedback, the travel distance should exceed 1 inch (25 mm) but should not exceed 2.5 inches (65 mm) for normal operation or 4 inches (100 mm) if the entire leg is engaged to activate the control.
RESOURCES Baumann, K. and Thomas, B. (2001). User Interface Design for Electronic Appliances. New York: Taylor & Francis. Bergman, E. and Johnson, E. (1997). Towards accessible human computer interaction (pp. 87–113). Advances in Human-Computer Interaction (vol. 5). Norwood, NJ: Ablex. Bobjer, O., Johansson, S., and Pigue, S. (1993). Friction between hand and handle: Effects of oil and lard on textured and non-textured surfaces; perception of discomfort. Applied Ergonomics, 24(3), 190. Bradley, J. (1967). Tactual coding of cylindrical knobs. Human Factors, 9(5), 483. Crawford, A. (1963). The perception of light signals: The effect of mixing flashing and steady irrelevant lights. Ergonomics, 6(3), 267. Crosby, A., Wehbe M. A., and Mawr, B. (1994). Hand strength: Normative values. Journal of Hand Surgery, 665. Edgar M., MacKenzie, S., and Buxton, W. (1996). A wearable computer for use in microgravity space and other non-desktop environments. Companion of the CHI ’96 Conference on Human Factors in Computing Systems, ACM, 69–70. Engineering Resource for Advancing Mobility, Ergonomics Aspects of Electronic Instrumentation: A Guide for Designers (SP-576). (1984). Warrendale, PA: Society of Automotive Engineers.
TAF-ER562X_WINGR-09-0401-C007.indd 293
11/8/10 11:43:38 PM
294
Handbook of Human Factors in Medical Device Design
Fisk, A. and Rogers, W. (1997). Handbook of Human Factors and the Older Adult. New York: Academic Press. Garrett, J. (1971). The adult human hand: Some anthropometric and biomechanic considerations. Human Factors, 13(2), 117. Greiner, T. (1991). Hand Anthropometry of U.S. Army Personnel. Washington, DC: U.S. Department of Commerce. Hoffman, E., Mannering, S., and Schoner, S. (1992). Response time as a measure of compatibility for linear displays with rotary controls. Proceedings of the Human Factors Society, Santa Monica, CA, 1483. Hsu, S. and Peng, Y. (1993). Control/display relationship of the four burner stove: A reexamination. Human Factors, 35(4), 745. Human Factors and Ergonomics Society. (2002). Human Factors Engineering of Computer Workstations (BSR.HFES 100 Draft Standard). Santa Monica, CA: Human Factors and Ergonomics Society. Imrhan, S. and Loo, C. (1989). Trends in finger pinch strength in children, adults, and the elderly. Human Factors, 31, 689–702. Kanis, H. (1993). Operation of controls on consumer products by physically impaired users. Human Factors, 35(2), 305. Kanis, H. and Van Hees, L. (1995). Manipulation of push buttons and round rotary controls. Proceedings of the Human Factors and Ergonomics Society, Santa Monica, CA, 374–378. Konz, S. (1995). Work Design: Industrial Ergonomics (4th ed.). Scottsdale, AZ: Publishing Horizons. Mathiowetz, V., Kashman, N., Volland, G., Weber, K., Dowe, M., and Rogers, S. (1985). Grip and pinch strength: Normative data for adults. Physical Medicine and Rehabilitation, 66, 69–74. McNeil, J. (1995). Americans with Disabilities: Data from the Survey of Income and Program Participation. Washington, DC: U.S. Department of Commerce, Economics, and Statistics Administration. Metz, S., Isle, B., Denno, S, and Li, W. (1990). Small rotary controls: Limitations for people with arthritis. Proceedings of the Human Factors and Society, 34, 137–140. Muckler, F. (1984). Standards for the design of controls: A case history. Applied Ergonomics, 15(3), 175. O’Hara, J. (2002). Soft Controls: Technical Basis and Human Factors Review Guidance. NUREG/ CR-6635. New York: Brookhaven National Laboratory. Odom, J. (1984). Applying Manual Controls and Displays: A Practical Guide to Panel Design. Freeport, IL: Honeywell Corporate Industrial Design. Peebles, L. and Norris, B. (1998). ADULTDATA: The Handbook of Adult Anthropometric and Strength Measurements. London: Department of Trade and Industry. Pheasant, S. (1996). Bodyspace: Anthropometry, Ergonomics, and the Design of Work. London: Taylor & Francis. Pirkl, J. (1994). Transgenerational Design: Products for an Aging Population. New York: Van Nostrand Reinhold. Rodgers, S. (1986). Ergonomic Design for People at Work (vols. 1 and 2). New York: Van Nostrand Reinhold. Salvendy, G. (Ed.). (2008). Handbook of Human Factors and Ergonomics. New York: John Wiley & Sons. Sanders, M. and McCormick, E. (1993). Human Factors in Engineering and Design. New York: McGraw-Hill. Schoorlemmer, W. and Kanis, H. (1992). Operation of controls on everyday products. Proceedings of the Human Factors Society, 36, 509–513. Scott, B. and Conzola, V. (1991). Designing touch screen numeric keypads: Effects of finger size, key size, and key spacing. Proceedings of the Human Factors and Ergonomics Society, 41, 360–364.
TAF-ER562X_WINGR-09-0401-C007.indd 294
11/8/10 11:43:38 PM
Controls
295
Sommerich, C. (2000). Inputting to a notebook computer. Proceedings of IEA/HFES Congress, Human Factors and Ergonomics Society, San Diego, CA, 671–674. U.K. Ministry of Defence. Defence Standard 00-25: Human Factors for Designers of Systems - Part 19: Human Engineering Domain. DEF STAN 00-25/1. U.K.: 2004. U.S. Department of Defense. (1999). Design Criteria Standard: Human Engineering, MIL-STD1472F. Washington, DC: U.S. Department of Defense. U.S. Federal Aviation Administration. (2003). Human Factors Design Standards (HFDS). HF-STD001. Washington, DC: U.S. Federal Aviation Administration. Valk, M. (1985). An experiment to study touch screen “button” design. Proceedings of the Human Factors Society 29th Annual Meeting, Santa Monica, CA, 29, 127–131. Van Cott, H. and Kinkade, R. (1992). Human Engineering Guide to Equipment Design. Washington, DC: American Institutes for Research. Woodson, W., Tillman, B., and Tillman, P. (1992). Human Factors Design Handbook. New York: McGraw-Hill.
TAF-ER562X_WINGR-09-0401-C007.indd 295
11/8/10 11:43:38 PM
TAF-ER562X_WINGR-09-0401-C007.indd 296
11/8/10 11:43:38 PM
8 Visual Displays William H. Muto, PhD, CPE; Michael E. Maddox, PhD, CHFP CONTENTS Scope ................................................................................................................................299 General Display Types .....................................................................................................299 Information Display ........................................................................................................ 300 8.1 General Principles .................................................................................................. 300 8.1.1 Understand and Accommodate User Population Characteristics ...............301 8.1.2 Accommodate the Range of Use Postures ..................................................301 8.1.3 Determine Typical and Variations in Mounting Positions ..........................302 8.1.4 Examine Environmental Conditions...........................................................302 8.1.5 Identify the Displayed Information Requirements .....................................303 8.1.6 Perform Objective Display Measurements .................................................305 8.1.7 Conduct Usability Tests ..............................................................................306 8.2 General Display Performance Requirements...........................................................307 8.2.1 Performance Requirements ........................................................................307 8.2.2 Display Viewing Conditions .......................................................................307 8.2.2.1 Viewing Distance .......................................................................307 8.2.2.2 Range of Viewing Angles ...........................................................308 8.2.2.3 Display Location/Orientation .....................................................308 8.2.3 Spatial Characteristics ................................................................................309 8.2.3.1 Image Quality .............................................................................309 8.2.3.2 Pixel Grid Modulation ................................................................ 310 8.2.3.3 Fill Factor ................................................................................... 310 8.2.3.4 Geometric Distortion .................................................................. 310 8.2.3.5 Moiré Patterns ............................................................................ 310 8.2.4 Temporal Characteristics ............................................................................ 311 8.2.4.1 Flicker ......................................................................................... 311 8.2.4.2 Jitter ............................................................................................ 312 8.2.4.3 Image Formation Time ............................................................... 313 8.2.5 Luminance and Color Characteristics ........................................................ 313 8.2.5.1 Luminance .................................................................................. 313 8.2.5.2 Luminance Contrast ................................................................... 314 8.2.5.3 Contrast Polarity ......................................................................... 314 8.2.5.4 Luminance Uniformity ............................................................... 315 8.2.5.5 Specular Glare ............................................................................ 315 8.2.5.6 Color Measurement..................................................................... 316 8.2.5.7 Color Uniformity ........................................................................ 316 297
TAF-ER562X_WINGR-09-0401-C008.indd 297
11/8/10 11:44:05 PM
298
Handbook of Human Factors in Medical Device Design
8.3 Display Formatting .................................................................................................. 316 8.3.1 Size and Spacing of Displayed Characters/Symbols .................................. 316 8.3.1.1 Screen Object or Character Height ............................................. 316 8.3.1.2 Character Height ......................................................................... 317 8.3.1.3 Character Width-to-Height Ratio ............................................... 318 8.3.1.4 Character Stroke-to-Width Ratio ................................................ 319 8.3.1.5 Font Style .................................................................................... 319 8.3.1.6 Antialiasing ................................................................................320 8.3.1.7 Character, Line, and Word Spacing ............................................320 8.3.1.8 Size of Color Objects and Alphanumeric Strings .......................320 8.3.2 Considerations for Displaying Data ............................................................321 8.3.2.1 Precision .....................................................................................321 8.3.2.2 Adequate Signal Duration ..........................................................321 8.3.2.3 Infrequent Signals .......................................................................321 8.3.2.4 Data Update Rates ......................................................................321 8.4 Types of Electronic Display .....................................................................................322 8.4.1 Comparison of Display Technologies .........................................................322 8.4.2 Cathode Ray Tubes (CRT) ..........................................................................322 8.4.3 Liquid Crystal Displays (LCDs) .................................................................322 8.4.4 Types of Liquid Crystal Displays................................................................324 8.4.4.1 Type of LCD by Method of Lighting..........................................324 8.4.4.2 Active-Matrix versus Passive-Matrix Displays...........................325 8.4.5 Light Emitting Diode (LED) Displays ........................................................325 8.4.5.1 Diffusing Elements and Light Pipes for LEDs ...........................326 8.4.6 Organic LEDs .............................................................................................327 8.4.7 Electroluminescent Displays.......................................................................327 8.4.8 Transilluminated Displays ..........................................................................328 8.4.9 Large-Screen/Projection Displays ..............................................................328 8.4.9.1 Application .................................................................................328 8.4.9.2 Control and Content of Displayed Information ..........................329 8.4.9.3 Viewing Distance .......................................................................329 8.4.10 Scale Indicators ..........................................................................................329 8.4.10.1 General .......................................................................................329 8.4.10.2 Applications ................................................................................330 8.4.11 Pointers .......................................................................................................330 8.4.11.1 Design Characteristics for Gauges with Pointers........................330 8.4.11.2 Numerical Progression ...............................................................332 8.4.11.3 Qualitative Indications on Scales................................................332 8.4.11.4 Break in Circular Scale...............................................................333 8.4.11.5 Scale Face Opening ....................................................................333 8.5 Special Applications ................................................................................................334 8.5.1 Introduction ................................................................................................334 8.5.2 Touch Screens .............................................................................................334 8.5.2.1 Touch-Screen Characteristics .....................................................334 8.5.2.2 Parallax .......................................................................................336 8.5.2.3 User/Environment Considerations ..............................................337
TAF-ER562X_WINGR-09-0401-C008.indd 298
11/8/10 11:44:05 PM
Visual Displays
299
8.5.3
Head-Mounted displays ..............................................................................337 8.5.3.1 Image Orientation and Position ..................................................338 8.5.3.2 Computing Bandwidth ................................................................339 8.5.3.3 Display Resolution ......................................................................339 8.6 Case Study ...............................................................................................................339 8.6.1 Specifying and Selecting a Display for a Cardiac Output Monitor ............339 8.6.1.1 Introduction ................................................................................339 8.6.1.2 Analysis ......................................................................................340 8.6.2 Display Usability Test: User Study to Assist in Display Selection for a Patient-Controlled Analgesia Infusion Pump .............................................345 8.6.2.1 Introduction ................................................................................345 8.6.2.2 Method ........................................................................................345 8.6.2.3 Procedure ....................................................................................347 8.6.2.4 Results ........................................................................................347 8.6.2.5 Conclusions/Recommendations ..................................................348 References ........................................................................................................................348 For many medical devices, the visual display is an essential and sometimes the only element of the user-device interface. The main purpose of a visual display is to present dynamic information to the user in alphanumeric, data graphic, or pictorial form. The display in some devices also allows user input, as when using a touch screen, light pen, or mouse input. The success of the visual display is determined by several factors, including the following: • How well the display subsystem supports the user’s task requirements • Whether the display matches the user’s characteristics and capabilities • The extent to which the display is compatible with the demands of the use environment
SCOPE The purpose of this chapter is to provide a brief overview of different display types (electronic as well as electromechanical/mechanical displays) and provide general guidance on their selection and use. The chapter also provides basic information regarding the formatting of display information.
GENERAL DISPLAY TYPES These guidelines apply to displays ranging from small-format numeric displays to largeformat graphical displays. These displays can be installed in a wide variety of devices, ranging from small handheld devices to helmet-mounted displays to large console-based systems or even projection systems. The types of display technologies used in these systems include the following: • Cathode ray tubes (CRTs), color and monochrome • Flat-panel (liquid crystal or plasma) displays, color and monochrome • Projection systems
TAF-ER562X_WINGR-09-0401-C008.indd 299
11/8/10 11:44:05 PM
300
Handbook of Human Factors in Medical Device Design
• Electroluminescent displays • Transilluminated displays • Light-emitting diode (LED) displays
INFORMATION DISPLAY This chapter also addresses formatting characteristics of the displayed elements. For certain kinds of displays, such as character and segmented displays, the format of displayed elements is dictated by design of the hardware. In others, such as graphical (“bitmap”) displays, the formatting of displayed elements is defined by software as well as the constraints of the hardware. The use of electromechanical and mechanical “dials” and “counters” in medical devices has diminished significantly in the past several years, with many such displays having been replaced by microprocessor-driven, electronic displays. These changes are dramatically illustrated by the evolution of the anesthesia workstation. Earlier generations of anesthesia workstations (see Figure 8.1) relied on mechanical or electromechanical gauges and flow meters. In contrast, newer systems use microprocessor-based integrated systems with electronic displays that receive information from various transducers and controls elements (Figure 8.2). Despite their diminished use, this chapter includes design guidelines for dials and counter-type displays for the following reasons: • A number of medical devices (such as pressurized tanks) continue to use such displays. • Many graphical user interfaces present analog or digital data using formats that emulate traditional gauges and dials. In such cases, many of the guidelines used are useful when graphically emulating such devices.
8.1 GENERAL PRINCIPLES Visual displays are typically one of many components in a medical device. However, from the perspective of users, displays assume a very important (if not the most important) role in their relationship with the device. Nearly all the information that is transmitted from
FIGURE 8.1 gauges.
Example of earlier-generation anesthesia workstation mechanical flow meters and
TAF-ER562X_WINGR-09-0401-C008.indd 300
11/8/10 11:44:05 PM
Visual Displays
301
5
l/min 2.5
1.0
O2 FIGURE 8.2
N2O
Virtual flow meter showing the status of two gases (O2 and N2O).
the device to users is transferred via a display. In some instances, such as touch-sensitive displays, users also employ the display to transmit information and control requirements to the device. Recognizing the key role of displays in using medical devices, it is not surprising that the general principles associated with proper display design parallel the user-centered design process for any device or system. The following principles are given in the approximate order in which the related user and task information should be considered. In general, designing displays for medical devices should start with user and task characteristics and then proceed to information and technology specifications.
8.1.1 UNDERSTAND AND ACCOMMODATE USER POPULATION CHARACTERISTICS Whether a display is appropriate for a given application depends largely on its compatibility with users’ characteristics and capabilities. Important population considerations include users’ visual characteristics (e.g., visual acuity, age-related visual impairment, and color vision deficiencies) as well as their anthropometric (relevant body characteristics, such as stature and eye height) characteristics.
8.1.2 ACCOMMODATE THE RANGE OF USE POSTURES It is essential to determine the likely positions of users’ eyes relative to the display during use. Variations in eye position could be due to a user changing posture (e.g., sitting vs. standing), moving to an off-angle viewing position, or changing viewing distance. For example, the designers of a patient monitor may assume that a user will view the device at arm’s length (or less) while standing directly in front of it. While monitoring a patient, however, a caregiver is likely to view patient parameters from the patient’s bedside or from several feet away. In addition to ideal or typical positions, designers should consider possible
TAF-ER562X_WINGR-09-0401-C008.indd 301
11/8/10 11:44:06 PM
302
Handbook of Human Factors in Medical Device Design
FIGURE 8.3 Stacking infusion pumps on an IV pole results in large difference in viewing angles. The pump in the upper right is set approximately at mean female (U.S. female, 50th percentile) eye height; the lowermost pump is set approximately at waist level.
extremes of user positions. Observing users interacting with similar devices in a real work environment can provide valuable insight into display requirements.
8.1.3 DETERMINE TYPICAL AND VARIATIONS IN MOUNTING POSITIONS Whether a device is mounted on a shelf, in a workstation, on a pole or is handheld, it is important to consider possible variations in the physical location and orientation of the device and how certain factors, such as viewing angles and glare susceptibility, could be affected. Infusion pumps, for example, typically are designed to be pole mounted with an assumed ideal mounting position at or near the user’s eye level. However, at a patient’s bedside, where space is at a premium, infusion pumps are often “stacked” with other pumps and bags of fluid on the same pole, resulting in pumps being placed significantly higher or lower than the assumed ideal viewing position (see Figure 8.3). Pumps or other devices designed without such considerations often exhibit deficiencies such as poor display visibility due to the use of displays with narrow viewing angles, bezels that obscure part of the display, buttons that do not align with the screen labels because of “parallax error,” and so forth. To avoid such problems, it is important to observe devices in actual use to determine typical as well as emergency use scenarios. Figure 8.4 depicts a common position of a portable patient monitor during the emergency transport of a patient.
8.1.4 EXAMINE ENVIRONMENTAL CONDITIONS The device’s display should be legible in the range of anticipated lighting conditions. This can be especially challenging when a single device must accommodate a wide variety of lighting conditions, ranging from patient rooms with low illumination to full sunlight. A study of use environments should include lighting measurements to characterize typical as well as worst-case viewing conditions (Muto, 2001).
TAF-ER562X_WINGR-09-0401-C008.indd 302
11/8/10 11:44:06 PM
Visual Displays
303
FIGURE 8.4 Portable patient monitor in use while transporting a patient shows that displays are not always placed in ideal viewing positions. (Courtesy of Spacelabs Medical, Inc.)
Vibration is another environmental variable that can affect the legibility of displays, such as when a device is used in a vehicle or aircraft (e.g., during patient transport). If the device must be used in such environments, especially when vibrations are in the range of 10 to 25 Hz (Sanders and McCormick, 1993), display legibility can be improved by providing vibration isolation and by increasing the size of displayed characters.
8.1.5 IDENTIFY THE DISPLAYED INFORMATION REQUIREMENTS Information required during the performance of the user’s task should be an essential consideration in the choice of displays and in the design of the overall user interface. Examples of how information requirements can drive display requirements include the following: • The specification for various display attributes, such as resolution, color range, and update rate, should be driven primarily by the specific information needed for the user’s tasks. For example, a high-resolution display may be required for users to discern extreme details in a radiological image (Figure 8.5), whereas a numeric
FIGURE 8.5 permission.)
A radiology workstation requires high image quality. (Courtesy of IBM. With
TAF-ER562X_WINGR-09-0401-C008.indd 303
11/8/10 11:44:06 PM
304
Handbook of Human Factors in Medical Device Design
FIGURE 8.6 Glucose meters incorporating low-cost, segmented LCDs. The glucose meter on the right has a custom LCD panel that contains fixed display elements (shown as units of measure (mg/dL) and AM or PM [currently displayed] labels for time of day).
display with a low-cost, segmented LCD likely will be sufficient for reading a numeric value on a handheld glucose meter (Figure 8.6). • Criticality and frequency of use should be important determinants of certain attributes of displayed elements. Critical and/or frequently used information should be made prominent with appropriate screen placement, high luminance and contrast, larger fonts, and attention-getting color. Less critical or less frequently used information can use smaller fonts, lower luminance, and more subdued colors. • Qualitative versus quantitative display information: Although medical tasks often require the user to assess precise quantitative information in order to perform their tasks, it is also often the case that the user is interested in the approximate value and trend of a continuously changing condition or parameter. Examples include trending graphs (heart rate, blood pressure, temperature, and so on) and categorical status indicators (e.g., green, yellow, and red for patient monitor conditions). • Character, symbolic, and pattern information: Certain information, such as blood pressure, is best expressed in numbers. Other information, such as whether a certain value, such as heart rate, is instantaneously increasing, decreasing, or remaining steady, is best presented as a symbol, like an arrow. In other cases, the user’s task is to assess of patterns of very rapid changes, as when evaluating ECG (see Figure 8.7) or EEG data. Some data may require both qualitative and quantitative readouts. For example, heart rate is most often displayed as a two-digit quantitative
FIGURE 8.7 Remote pager displays ECG waveform that conveys the electrical activity of the heart. (Courtesy of Spacelabs Medical, Inc.)
TAF-ER562X_WINGR-09-0401-C008.indd 304
11/8/10 11:44:08 PM
Visual Displays
305
value, whereas changes in heart rate over extended durations are typically shown graphically.
8.1.6 PERFORM OBJECTIVE DISPLAY MEASUREMENTS Although vendors almost always publish performance claims for their displays (e.g., maximum luminance, LCD contrast ratios, display “pitch”), such measures are often difficult to extrapolate to a given application. Furthermore, variations in manufacturers’ measurement methods often make it risky to base display selections only on technical claims. Consequently, it is advisable to gather quantitative display performance data to assess compatibility with the intended users and applications. Relatively straightforward measures of luminance and contrast can be useful in assessing subtle differences among displays and are more reliable than subjective judgments. Figure 8.8 shows a handheld meter that measures the luminance of small object areas (typically 1/3 to 1 degree of circular area). Similar handheld devices that measure chromaticity (color) can be used to objectively assess a display’s color uniformity, accuracy, and range of colors (color gamut). These measurements can provide valuable information for applications where accurate color discrimination is important. More advanced measures of display performance can be useful for assessing display image quality and performance, especially for demanding applications. The modulation transfer function (MTF), a measure of a display’s ability to present the contrast of an object as a function of object detail, has been used for years to measure image quality of displays objectively. However, the MTF is not a unitary measure, that is, single number. Rather, it is usually depicted as a graph showing a display’s ability to provide contrast for increasingly detailed patterns, usually sine waves. It is a useful technique to determine how display contrast falls off for smaller and smaller features. However, it does not automatically relate that capability to either user task requirements or users’ visual capabilities. A more useful
FIGURE 8.8 Luminance meter is used to measure the light emitted from a small circular area (typically 1/3 to 1 degree) on a display. The area to be measured is sighted through an eyepiece (not shown), and the trigger is activated to initiate the measurement. Smaller targets are measured with the use of close-up lenses. (Courtesy of Konica Minolta. With permission)
TAF-ER562X_WINGR-09-0401-C008.indd 305
11/8/10 11:44:09 PM
306
Handbook of Human Factors in Medical Device Design
measure in this regard is the modulation transfer function area (MTFA), which takes users’ ability to see various levels of detail into account (see the discussion of MTF and MTFA in Section 8.2.3.1). In digital X-ray applications, where the user is required to view highly detailed imagery, detective quantum efficiency (DQE), a measure of the combined effect of the noise and contrast performance of an imaging system, has been suggested as an alternative to MTF as a way of better predicting display performance in clinical settings (see http://www. gehealthcare.com/usen/xr/edu/products/digimgtutorial.html). These types of photometric measurements can be useful in selecting and qualifying candidate displays by reducing reliance on purely subjective judgments. Because some of these measures (e.g., MTF and DQE) require specialized instrumentation, it may be appropriate to engage third-party evaluators to perform these sophisticated measurements.
8.1.7 CONDUCT USABILITY TESTS User input is critical to the selection and design of displays. Display evaluations conducted early in a device development, involving people who represent the target user population, can provide both objective and subjective information regarding the suitability of candidate displays (with either the display mounted in the device or the display by itself). Later usability tests with displays integrated into operational devices (or representative prototypes) should be conducted to assess display performance in realistic use scenarios (Figure 8.9). During such assessments, care should be taken to represent the relevant tasks, environment, the range of mounting positions, the range of viewing positions, and so on. See Chapter 6, “Testing and Evaluation,” for a more information on the planning and execution of usability tests.
FIGURE 8.9 Example of a usability test of displays for an infusion pump. Participants performed simple programming tasks and subsequently rated each of several displays. Pumps were placed at various heights to assess visibility and user subjective ratings at “normal” and extreme viewing angles.
TAF-ER562X_WINGR-09-0401-C008.indd 306
11/8/10 11:44:09 PM
Visual Displays
307
8.2 GENERAL DISPLAY PERFORMANCE REQUIREMENTS 8.2.1 PERFORMANCE REQUIREMENTS This section specifies preferred performance characteristics for visual displays. Visual displays that conform to these guidelines will be legible, readable, and comfortable to use.
8.2.2 DISPLAY VIEWING CONDITIONS With few exceptions (e.g., head-mounted displays), the following guidelines apply to all displays. 8.2.2.1 Viewing Distance Viewing distance is the linear distance between the user’s eye and the center of the visual display. Specifying the viewing distance range establishes conditions under which a display will meet applicable functional and performance requirements. Guideline 8.1: Minimum Viewing Distance The minimum viewing distance the minimum distance anticipated for a given display (and will conform to the specified display parameters). For many tasks (e.g., reading and interaction with the associated device), the minimum design viewing distance (minimum distance) should be ≥40 cm.
Guideline 8.2: Maximum Viewing Distance The maximum viewing distance the maximum anticipated viewing distance for a given application. The maximum viewing distance should be defined for each display application to establish critical display parameters, such as overall display size and the required character/ symbol sizes for legibility or effective display element recognition. For example, a patient monitor at the patient’s bedside with a typical viewing distance of 6 feet may require a maximum viewing distance of 15 feet; a requirement driven by the need of users to be able to view patient parameters quickly from outside the patient’s room (through an observation window) or from the doorway. Figure 8.10 shows laparoscopic surgery being conducted using a display located several few feet from the surgeon.
FIGURE 8.10 surgeon.
Laparoscopic surgery using a CRT display placed at eye level several feet from the
TAF-ER562X_WINGR-09-0401-C008.indd 307
11/8/10 11:44:10 PM
308
Handbook of Human Factors in Medical Device Design
8.2.2.2 Range of Viewing Angles Unlike users of computer monitors in a typical office, who usually sit in an adjustable chair directly in front of their monitors, users of medical devices often assume less-thanoptimal viewing positions. This is mainly because many medical devices are not used in a fixed position with users seated or standing in a single predictable position. Figure 8.11 illustrates various factors that can affect the viewing angle of a medical device, including the display monitor’s mounting height, monitor tilt angle, user’s viewing distance, and the user’s eye height. Assumptions about users’ eye height should consider the distribution of the user population, a determination of what population extreme should be accommodated (usually the 5th or 95th percentile; see Chapter 4, “Anthropometry and Biomechanics”), and whether users will be sitting, standing, or lying down during use. Because the legibility of certain types of displays (such as passive-matrix LCDs) tends to degrade as the viewing angle increases from normal (i.e., perpendicular to the display face), users often report that such displays appear dim, are hard to read, and display unnatural colors. Photometric measurements of these displays at the actual viewing angles provide objective evidence of decreases in luminance, contrast, and color distortion (hue and saturation). Guideline 8.3: Viewing Angle The maximum anticipated horizontal and vertical viewing angles of a display should be defined. The performance of the candidate displays should be assessed at maximum viewing angles prior to making final display selections. Candidate displays should meet all viewing requirements throughout the range of viewing angles.
8.2.2.3 Display Location/Orientation The location and orientation of the display are often critical to effective task performance. Improper placement of a display can increase the user’s visual scan time, decision-making time, and error rates. For example, a display placed so that movement on the screen does not match the user’s expectation for direction of movement could result in errors or increased
Monitor
Viewing distance
Display height
0°
Monitor tilt angle
Off-normal viewing angle
Eye height (standing or seated)
Floor
FIGURE 8.11 Elements that contribute to display vertical viewing angle. (Adapted from Muto, W. H., paper presented at meeting of the Association for the Advancement of Medical Instrumentation, Baltimore, MD, 2001. With permission.)
TAF-ER562X_WINGR-09-0401-C008.indd 308
11/8/10 11:44:11 PM
Visual Displays
309
reaction time. In a study to determine the effects of varying display positions (three different horizontal positions [left, frontal, and right] at two heights [eye level and hand level]) on endoscopic task performance times and quality scores, it was found that when displays were presented in front at the hand level, performance exceeded all other display position conditions (Hanna, Shimi, and Cuschieri, 1998).
8.2.3 SPATIAL CHARACTERISTICS 8.2.3.1 Image Quality Most observers are able to identify a display that has high clarity, good color rendition, and other factors making for a “great picture.” For several decades, researchers have been striving to develop measurements of image quality that correlate with and that can ultimately predict the perception of good image quality and optimal visual task performance. One of those measurements is the MTF, which was briefly described in Section 8.1.6. The MTF is a two-dimensional curve established by making photometric measurements and then performing a series of mathematical transformations that characterize the ability of a display to reproduce details. Various types of displays are capable of producing more or less detail than can readily be perceived by human observers. The MTFA is a unitary measure that expresses the ability of a display to depict objects of such detail and contrast that exceed the threshold requirements of the human visual system. Unfortunately, there are no commonly accepted values of MTFA that can be used to judge the appropriateness of displays for specific medical tasks. In fact, at least one early study found very little relationship between MTFA and certain basic display-related tasks (Maddox, 1979). The MTFA measure is much better suited to rank ordering various displays in terms of their ability to reproduce details within the range of human visual capabilities. Image quality metrics based on spatial characteristics, such as MTFA, are dependent on the ability of a display to reproduce details, which is directly related to the density of picture elements (pixels) that form the image. Continuing development of CRTs and flatpanel displays in recent years has resulted in displays with high pixel densities. As pixel density (expressed as the number of pixels per unit distance) increases, the eye’s ability to discern individual pixels decreases, resulting in the perception of images as continuous and smooth. Displays capable of producing images that appear more continuous (where pixels are not visible at normal viewing distance) are rated as having better image quality than those that do not. For example, pixel densities of 72 pixels per inch (ppi) are generally considered adequate for text presented on desktop displays. Pixel densities of 150 to 300 ppi are considered sufficient to produce photographic images of “good” to “excellent” image quality. Unfortunately, pixel density is often confounded with displayed object size and sometimes not directly related to the underlying display technology. For example, most flat-panel computer displays are capable of being set in software to display various numbers of “pixels.” The term is in quotation marks because the vertical and horizontal pixel count defined in software is not directly related to the number of true picture elements contained on the display surface. For example, changing the display setting from 1,024 × 768 pixels to 800 × 600 pixels does not alter the number of physical picture elements on the surface of the display. Rather, the lower pixel count setting merely causes many display pixels to be used to form a single software pixel.
TAF-ER562X_WINGR-09-0401-C008.indd 309
11/8/10 11:44:11 PM
310
Handbook of Human Factors in Medical Device Design
Because of these interdependencies, in order to assess the compatibility of a display with particular task requirements, usability tests should be conducted presenting the size range of objects likely to be seen in practice. This range should definitely include the smallest or most finely detailed objects to assess compatibility of a display with the application. 8.2.3.2 Pixel Grid Modulation Displays present images with picture elements (pixels) in a pattern known as a grid or raster. However, when picture elements within the grid or raster are visible, they tend to be distracting and interfere with the legibility of displayed images. Pixel grid modulation is a method for specifying/assessing the visibility of the grid relative to its background. Because it is desirable for the pixel grid to be less visible, the corresponding pixel grid modulation should also be lower. Guideline 8.4: Pixel Grid Contrast For CRTs with a pixel density of less than 30 pixels per degree of visual angle (at the minimum design viewing distance measured perpendicular to the raster), the luminance contrast should not exceed 0.4 for monochrome displays and 0.7 for color displays with all pixels illuminated (HFES, 2002).
8.2.3.3 Fill Factor Although images tend to look uniform across a typical display, when viewed under magnification the pixels of a display are separated by nonilluminated (dark) areas. The “fill factor” of a display refers to the ratio of the total illuminated area relative to the nonilluminated area. Displays with a higher fill factor are typically judged to have higher image quality than those with a lower fill factors. Fill factor is calculated by multiplying the height of a pixel times its width divided by the area allocated to the pixel. Pixel size is determined by defining the edges of the pixel as the point at which there is 50% luminance contrast between the pixel and its background (ISO 9241, part 3, 1992). Guideline 8.5: Fill Factor For flat-panel displays having a pixel density of 30 pixels per degree of visual angle or less (at the design viewing distance), the fill factor should exceed 0.3. The preferred fill factor is ≥0.5.
8.2.3.4 Geometric Distortion Geometric distortion refers to the deviation of rows or columns of picture elements from a straight line. Geometric distortion can interfere with the legibility of characters and symbols. Guideline 8.6: Geometric Distortion The addressable area of a display screen should not have geometric distortion that exceeds 1% of the screen width or the screen height.
8.2.3.5 Moiré Patterns Moiré patterns are repeating (periodic) visual patterns caused by interaction of the pixel grid with the grid or repeating pattern of the image being displayed. Such patterns are often seen when a striped or patterned field is shown on standard broadcast television, and the
TAF-ER562X_WINGR-09-0401-C008.indd 310
11/8/10 11:44:11 PM
Visual Displays
311 Contrast Cm
10
6 JND 0.1
0.01 0.5
1.0
10 20 Spatial frequency (cpd)
FIGURE 8.12 Thresholds for visibility of moiré patterns as a function of spatial frequency (in cycles per degree) and contrast modulation.
resultant image seems to contain undulating dark and light patterns that are considered distracting or otherwise undesirable. Guideline 8.7: Moiré Patterns For color displays, the contrast of Moiré patterns should not exceed 6 JNDs (just noticeable differences) at their fundamental spatial frequency. This criterion is specified by the curve shown in Figure 8.12. (BSR/HFES 100, 2002).
8.2.4 TEMPORAL CHARACTERISTICS 8.2.4.1 Flicker Flicker refers to the perception of rapid fluctuations in brightness levels (e.g., alternating dark and light) of a screen or screen image. Users who perceive flicker complain that a display tends to “blink” rapidly—often reported as a fluctuation in the viewer’s peripheral vision. Flicker can be distracting and lead to complaints of eyestrain. Guideline 8.8: Avoiding Flicker The image and background on the display screen should be free of apparent flicker.
A number of studies have investigated the display factors that can affect the perception of flicker, which includes refresh rate, display luminance, surrounding illuminance, and field of view. Peripheral vision is more sensitive to flicker than is central retinal vision. As the refresh rate increases, the point at which flicker is no longer apparent is referred to as the “flicker fusion frequency” or “critical flicker frequency” (CFF).
TAF-ER562X_WINGR-09-0401-C008.indd 311
11/8/10 11:44:11 PM
312
Handbook of Human Factors in Medical Device Design 100
CFF (Hz)
Flicker limits
95 90
70°
85
50°
80
30°
75 70
10°
65 60 55 50
Luminance (cd/m2) 0
20
40
60
80
100
120
140
160
180
200
FIGURE 8.13 CRT flicker limits. Perception of flicker as a function of CRT luminance and field of view. (From BSR/HFES 100, Human Factors Engineering of Computer Workstations, Human Factors and Ergonomics Society, Santa Monica, CA, 2002.)
The human factors standard on computer workstations (HFES, 2002) proposes a method to determine the CFF, the refresh frequency limits at which flicker is detected, for CRTs with short to medium-short persistence phosphors. The graph shown in Figure 8.13 shows that CFF (in Hz) varies as a function of luminance and the angular field of view. To determine the CFF, do the following: 1. Measure the average luminance of the display at the intended brightness setting. 2. Calculate the field of view using the formula Viewing angle = 2tan–1(D/2V) where D is the diagonal measurement of the display (mm) and V is the minimum viewing distance. 3. Locate the intersection of the luminance value with the curve of worst-case viewing angle. 4. Read the CFF as the ordinate at the intersection in step 3. A display that has a refresh rate of that frequency or higher is not likely to be a source of flicker in the working environment. A good rule of thumb for most common displays is to provide a refresh rate of 70 to 75 Hz. Based on the content of Figure 8.13, larger displays and higher luminance might require higher refresh rates. 8.2.4.2 Jitter Unwanted movement of screen images, called jitter, can be annoying and should be kept to a minimum. Jitter is most visible at low frequencies (1 to 3 Hz), whereas at higher frequencies, jitter is perceived as image blurring. Guideline 8.9: Avoiding Jitter The peak-to-peak variation in the location of a picture element should not exceed 2 mm/cm of design viewing distance in the frequency range of 0.5 to 30 Hz.
TAF-ER562X_WINGR-09-0401-C008.indd 312
11/8/10 11:44:11 PM
Visual Displays
313
8.2.4.3 Image Formation Time Information on medical device displays often changes rapidly and dynamically. Image formation time (or latency) is a measure of how quickly a display can update the image to a usable level of contrast. Some LCDs, for example, require more than 200 ms to form the desired image. Although such displays may have the advantage of being free from flicker, they can suffer from low contrast, which is “caused by persistence of previous images during image motion or immediately following image changes” (HFES, 2002, p. 56). Long image formation times may also cause complaints of “ghosting.” Guideline 8.10: Image Formation Time Displays should have an image formation time of less than 55 ms (HFES, 2002).
8.2.5 LUMINANCE AND COLOR CHARACTERISTICS 8.2.5.1 Luminance Luminance is a measure of the light intensity emitted from a display or a light source. It is the objective measure of what most people think of as “brightness.”* The official international measurement for luminance is candelas per square meter (cd/m2), which is also equivalent to the “nit,” that is, 1 nit = 1 cd/m2 (Sherr, 1979). Virtually all modern desktop monitors and those with similar formats easily meet 35 cd/m2. Smaller-format displays that are commonly lit by low-power light sources (e.g., LEDs) sometimes do not meet these minimum levels and are typically perceived as being “dim.” The following are recommended luminance guidelines: Guideline 8.11: Minimum Display Luminance The luminance of a display should be no less than 35 cd/m2.
Guideline 8.12: Luminance under High Ambient Light Conditions Under high ambient lighting (illuminance) conditions, users often prefer higher luminance levels, such as 100 cd/m2 (ISO 9241-3, 1992).
Guideline 8.13: Highest Luminance for Demanding Applications Demanding applications, such as anatomical imaging systems, may require even higher levels of luminance. For example, for remote viewing of patient radiographic images, gray-scale monitors should have a luminance of 50 foot-lamberts, or 172 cd/m2 (American College of Radiology, 1999).
Guideline 8.14: Provide “Brightness” Control A display that is capable of a higher luminance than is necessary for its intended use (e.g., a physiological display at the patient bedside) should have a “brightness” control. * Although “brightness” is commonly used interchangeably with “luminance,” “brightness” more accurately refers to the human response to luminous intensity (analogous to the human perception of “loudness” as a function of auditory intensity in decibels).
TAF-ER562X_WINGR-09-0401-C008.indd 313
11/8/10 11:44:11 PM
314
Handbook of Human Factors in Medical Device Design
Guideline 8.15: Constraints on Displaying Dimming Brightness (or “dimming”) controls, when employed, should prevent users from setting important displays (e.g., alarms) to imperceptible levels.
8.2.5.2 Luminance Contrast One of the most important factors in display legibility is luminance contrast, defined as the difference in luminance between foreground and background of displayed elements. Luminance contrast is typically expressed as either of the following: • Contrast ratio (CR = Lmax/Lmin) • Contrast modulation (Cm= [Lmax – Lmin]/[Lmax + Lmin]), where Lmax is the higher luminance of the background or displayed symbol and Lmin is the lower luminance of the background or symbol Guideline 8.16: Minimum Luminance Contrast Ratio The luminance contrast ratio of any display, whether monochrome or color, should be no less than 3:1 (equivalent contrast modulation is 0.5; HFES, 2002). This contrast ratio should be maintained across all anticipated viewing angles.
Guideline 8.17: Preferred Luminance Contrast Ratio The preferred luminance contrast ratio, across all anticipated viewing angles, should be no less than 7:1 (equivalent contrast modulation is 0.75; HFES, 1988).
8.2.5.3 Contrast Polarity Either dark characters on a light background (positive image polarity) or light characters on a dark background (negative image polarity; see Figure 8.14) are acceptable as long as the other display requirements are met. Each of the display modes has advantages and disadvantages. With positive image polarity, screen reflections are less troublesome, edges appear sharper, and luminance balance is easier to obtain. With negative image polarity, flicker and moiré patterns are less perceptible. Where color discrimination is important, a wider range of colors can be used with negative image polarity than with positive image polarity (HFES, 2002) because the lighter colors (e.g., yellow, white) that are suitable for use on a dark background do not provide sufficient contrast to be suitable on a light background.
This text is displayed with positive image polarity.
This text is displayed with positive image polarity.
This text (bold) is displayed with positive image polarity.
This text (bold) is displayed with positive image polarity.
FIGURE 8.14 Illustration of positive image polarity (left column) and negative image polarity (right column).
TAF-ER562X_WINGR-09-0401-C008.indd 314
11/8/10 11:44:11 PM
Visual Displays
315
8.2.5.4 Luminance Uniformity A screen should appear to have uniform “brightness” across the screen face. For a display to appear uniform, the luminance, averaged across a 1-degree area at the minimum design viewing distance. Guideline 8.18: Luminance Uniformity Luminance should not vary from one area of the display (averaged across a 1-degree area at the minimum design viewing distance) to another by more than 1.7:1 (ISO 9241-3, 1992).
8.2.5.5 Specular Glare Reflections of room lights from the display face (also referred to as specular glare) can significantly reduce screen contrast and consequent legibility. Users find excessive reflections annoying (see Figure 8.15 for an illustration of specular glare). Guideline 8.19: contrast of Specular Reflections The luminance contrast of specular reflections should be less than or equal to 1.25.
Reflection contrast is defined as RF = (Lmin + Lr)/Lmin, where Lmin denotes the luminance produced by the minimum image input signal and Lr denotes the maximum luminance of the specular reflection (HFES, 2002). To control reflections, display surfaces or overlays applied to the display surface are often treated in various ways to combat glare, including chemical or mechanical etchings and the use of front surface coatings. Chemical or mechanical etching causes the surface of the glass to be roughened so that reflected light is randomly scattered. Although effective in reducing glare, etching also tends to diffuse the light emitted from the display itself, causing significant degradation in display image quality. Chemical coatings applied to the display surface can be effective against reflected glare sources, typically with less degradation to image quality than with most etchings. The disadvantages of coatings are that they tend to reduce screen luminance, show fingerprints, and produce slight color distortion. When dealing with glare-reducing overlays or touch-screen overlays (discussed later in this chapter), the trade-off between glare reduction and resolution degradation should be considered and tested (under actual use conditions) before inclusion in the application.
No anti-glare treatment
Without anti-glare coating
With anti-glare coating
FIGURE 8.15 Screen on the left shows reflective glare that can interfere with task performance and cause annoyance. The screen on the right illustrates how glare (shown on the left half of the screen) is reduced with the use of an antiglare treatment (on the right side of the screen).
TAF-ER562X_WINGR-09-0401-C008.indd 315
11/8/10 11:44:12 PM
316
Handbook of Human Factors in Medical Device Design
8.2.5.6 Color Measurement There are a number of systems used to specify display color. All electronic color displays are additive; that is, individual colors are generated by combining portions of three primary colors: red, green, and blue. The most commonly applied international standard for specifying colors on additive displays has been developed by the Commission Internationale de l’Eclairage (CIE). This standard has been modified from its original form to incorporate human perceptual data. The result is known as the “CIELUV” color space. A detailed explanation of color specification is beyond the scope of this chapter; however, interested readers can consult any of a number of sources (e.g., Thorell and Smith, 1990). 8.2.5.7 Color Uniformity Color should appear uniform across the display screen. When using chromaticity measurement instruments, the following guidelines apply. Guideline 8.20: Color Uniformity Across the Screen The chromaticity measurements Du′v′ (measured at different screen locations) at a given chrominance should not exceed 0.03.
Guideline 8.21: Color Uniformity in Small Screen Areas The chromaticity measurements Du′v′, should not exceed 0.02 within any area subtending less than 35 degrees (BSR/HFES 100, 2002).
8.3 DISPLAY FORMATTING Small-format displays often employ fixed segments or dots that can be turned on and off (“addressed”) to form alphanumeric characters or symbols. The symbols and characters in these displays, typically single- or two-line displays, are defined by the hardware (showing character segments or a fixed dot matrix) and remain in a fixed format (character size, shape, and so on). Another type of display, available in a variety of sizes, is a bitmapped display, which allows addressing of individual picture elements (pixels) anywhere on the display. With bitmapped displays, alphanumeric characters, symbols, graphics, or even pictures can be displayed in a range of sizes and styles and are all controlled by software. The following are formatting guidelines for characters and symbols, whether they are determined by hardware constraints or software.
8.3.1 SIZE AND SPACING OF DISPLAYED CHARACTERS/SYMBOLS 8.3.1.1 Screen Object or Character Height One of the common screen formatting questions for medical applications is, “How big should it [the data readout, message, or symbol] be?” The recommended size of displayed objects (characters or symbols) can be expressed as a physical dimension (in millimeters, inches, or points) or as the number of pixels from the bottom of the character to the top. Because the perceived height of the character will vary depending how far the viewer is from the display, a more convenient way of specifying character height is the visual angle
TAF-ER562X_WINGR-09-0401-C008.indd 316
11/8/10 11:44:12 PM
Visual Displays
317
Visual angle (arc), A
Height of object, H
Distance of object, D
Visual arc (in minutes) = Tan–1 (H/D) × 60
FIGURE 8.16 Illustration of visual angle and formula for calculation. (Adapted from Sanders, M.S. and McCormick, E.J., Human Factors in Engineering and Design. McGraw-Hill, New York, 1993. With permission.)
that the object subtends at the eye. Figure 8.16 illustrates the concept of visual angle (or visual arc) and how it is determined. Visual angle is usually expressed in minutes of arc, where 1 degree = 60 minutes of arc (Sanders and McCormick, 1993). 8.3.1.2 Character Height One of the most important determinants of alphanumeric character legibility is character height.* To specify a constant character height at various distances, the physical height of a character must vary to present a constant visual angle. Table 8.1 shows the relationship between selected character heights in arc minutes (24, 22, 20, 18 and 16) and physical character height in inches and points (72.27 points per inch) at five different viewing distances (16, 24, 36, 120, and 180 inches). The following guidelines present recommended character heights for various viewing scenarios (HFES, 2002). Guideline 8.22: Minimum Character Height The minimum character height should be 16 minutes of visual arc when legibility is important.
Guideline 8.23: Preferrred Character Height The preferred height of characters is 20 to 22 minutes of visual arc for displayed characters that are viewed frequently or when rapid comprehension is essential.
Guideline 8.24: Maximum Character Height for Text The maximum character height for contiguous text should be less than 24 minutes of visual arc. Characters that are too large decrease reading speed by reducing the number of characters viewed during an eye fixation.
Guideline 8.25: Essential Readout Character Height For essential readouts made up of a small number of digits and where quick recognition is required (e.g., heart rate, infusion amount), character heights of 24 to 30 minutes of arc are acceptable. * Although overall character size (width as well as height) is the essential characteristic, it is assumed that the width of readable characters is proportional to height (see Section 8.3.1.3). Consequently, character height is a common convention for specifying character size.
TAF-ER562X_WINGR-09-0401-C008.indd 317
11/8/10 11:44:12 PM
318
Handbook of Human Factors in Medical Device Design
TABLE 8.1 The Relationship between Selected Character Heights in Arc Minutes and Physical Character Height in Inches at Five Different Viewing Distances Character Height Visual Angle
Viewing Distance (inches)
Character Height (inches)a
Approximate Font Size (points)b
24 minutes
16 24 36 120 180
0.1117 0.1675 0.2513 0.8377 1.2565
16 24 36 120 180
22 minutes
16 24 36 120 180
0.1024 0.1536 0.2304 0.7679 1.1518
7 11 17 55 83
20 minutes
16 24 36 120 180
0.0931 0.1396 0.2094 0.6981 1.0471
7 10 15 50 76
18 minutes
16 24 36 120 180
0.0838 0.1257 0.1885 0.6283 0.9424
6 9 14 45 68
16 minutes
16 24 36 120 180
0.0745 0.1117 0.1675 0.5585 0.8377
5 8 12 40 61
Source: Israelski, E. (2004). Personal communication. a Character height (inches) = distance × (minutes of arc)/(57.3 × 60). b Font size (points) = character height (inches) divided by 0.013837 where 1 point = 1/72.27 inches.
In addition to viewing distance, the choice of character height can also depend on other factors, including: • User population (e.g., see Chapter 18, “Home Health Care,” for a discussion of older users) • Need for quick recognition • Degraded environmental conditions (high illuminance, vibration, and so on) 8.3.1.3 Character Width-to-Height Ratio Characters (including other symbols and icons) that are too wide or too narrow relative to their height will appear distorted and are harder to read.
TAF-ER562X_WINGR-09-0401-C008.indd 318
11/8/10 11:44:12 PM
Visual Displays
319
Guideline 8.26: Character Width-to-Height Ratio For optimal legibility and readability, the ratio of character width to height should range from 0.6:1 to 0.9:1. The ratio is determined by measuring an unaccented (i.e., sanserif) uppercase character “H” (HFES, 2002).
8.3.1.4 Character Stroke-to-Width Ratio Displayed characters are composed of line segments called “strokes.” If the strokes are too thin, they will be less likely to be detectable by the eye, rendering characters difficult to read. However, using strokes that are too wide will cause the user’s eyes and brain to “fill in” the spaces of characters, also resulting in poor readability. Guideline 8.27: Character Stroke-to-Width Ratio The character stroke width should be 8% to 20% of the character height. The stroke may be more than one pixel wide (HFES, 2002).
8.3.1.5 Font Style There are thousands of typographical styles in use today. One of the most recognizable attributes of type styles is whether a font is a serif (serifs are the small lines at the top and bottom of characters) or a sans serif font (font without serifs) (see Figure 8.17). Research in this area is somewhat contradictory. Hay (1999) attributes inconsistent results to the notion that under ideal viewing conditions (adequate font size and luminance) serif fonts perform better (increased reading speed, legibility) than san serif fonts and that under lessthan-ideal conditions (smaller fonts, low luminance) sans serif fonts perform better. Wilson (2001) argues that most current displays are not equivalent to printed materials because of their relatively low resolution. Displays have a typical resolution of 72 dots per inch, whereas print resolution is typically 180 to 300 dots per inch or higher: • When the task requires reading of continuous text, serif fonts are more readable than san serif fonts for printed materials or displays with high resolution (e.g., more than 150 dots per inch). • Sans serif fonts are often preferred by users over serif fonts when presented on displays with lower resolution (e.g., fewer than 80 dots per inch) or over luminance. Regardless of style care should be taken to choose the font that minimizes confusion for character combinations used in the specific application. For example, when using Arial, a sans serif font, uppercase “I” (as in “India”) and lowercase “l” (as in “lima”) appear identical so that the first three characters in “Illinois” are indistinguishable from each other. Similarly, with Times New Roman, a serif font, the lowercase letter “l” (“lima”) and the
Serifs
FIGURE 8.17 Comparison of serif font (left, New Times Roman) and sans serif fonts (right, Arial). Arrows point to serifs at the bottom of “A” and the top of “T.”
TAF-ER562X_WINGR-09-0401-C008.indd 319
11/8/10 11:44:12 PM
320
Handbook of Human Factors in Medical Device Design
numeral “1” (one) appear virtually identical. Although the context will often minimize the ambiguity of the text (such as in “Illinois”), care should be taken to avoid combinations where such confusions could cause users to misinterpret the readout. Changes in the font and/or the text may be necessary to eliminate confusion. 8.3.1.6 Antialiasing Computer displays are made up of distinct colored dots called picture elements, or “pixels.” Usually, pixels are placed in tightly spaced rows and columns so that adjacent pixels are to the left, right, top, and bottom of their neighbor. Because the pixels are constrained to this square array, a line on a computer screen that is not precisely horizontal or vertical can appear jagged. Likewise, characters or symbols with curved edges can also look uneven or jagged. This phenomenon is called “aliasing.” The visual effects of aliasing can be minimized with a technique called “antialiasing.” Antialiasing visually smoothes jagged character edges and lines by filling in adjacent pixels with less saturated colors or shades of gray. Antialiasing can be very effective at improving the appearance of a certain range of fonts. However, when applied to fonts that are small, antialiasing tends to fill in parts of characters, reducing their sharpness. 8.3.1.7 Character, Line, and Word Spacing Improper spacing of characters, words, and lines can interfere with readability. Guideline 8.28: Between-Character Spacing The minimum spacing between characters should be greater than or equal to 10% of the character height (ANSI/HFS 100, 1988).
Guideline 8.29: Between-Line Spacing The minimum spacing between lines of text should be greater than or equal to 15% of the character height (ANSI/ HFS 100-1988).
Guideline 8.30: Between-Word Spacing Because multiple words are often read in clusters, excessive or variable spacing between words can interfere with reading speed and comprehension. A minimum of one-half the width of character “H” should be used between words. The spacing between words should exceed the spacing between characters.
8.3.1.8 Size of Color Objects and Alphanumeric Strings As an image gets smaller, the ability of the eye to accurately perceive its color decreases. In fact, unless the colored object is greater than a specified size, the human visual system will detect only luminance (light vs. dark) and cannot perceive chrominance (color) information. In addition, because the eye has difficulty focusing on images made up of short-wavelength light, the minimum object size required for viewing blue objects is significantly larger than for most other colors (2 degrees for blue, 20 minutes for others). If color information is to be conveyed in the alphanumeric, symbolic, or pictorial information, the displayed image should conform to the following guidelines (HFES, 2002):
TAF-ER562X_WINGR-09-0401-C008.indd 320
11/8/10 11:44:13 PM
Visual Displays
321
Guideline 8.31: Color Discrimination of Alphanumeric Strings Where accurate color discrimination of alphanumeric character strings is required, character height should subtend 20 minutes of arc (1/3 degree) or greater.
Guideline 8.32: Color Discrimination of Individual Characters If accurate color discrimination of an individual character or symbol is required, the height of the symbol should subtend 30 minutes of arc (1/2 degree) or greater.
Guideline 8.33: Use of Blue Colored Text or Objects Blue should not be used to color objects subtending less than 2 degrees of visual arc.
8.3.2 CONSIDERATIONS FOR DISPLAYING DATA 8.3.2.1 Precision Displaying precision beyond the required level can be distracting and/or misleading and can interfere with timely decision making. For example, it is technically possible for heart rate in patient monitors to be calculated and displayed with multiple decimal places. However, because heart rate tends to be highly variable and clinical needs do not require high precision, a digital display of heart rate with multiple decimal places would be excessively variable (noisy) at the extreme decimal places and would provide no clinical benefit. Guideline 8.34: Precision of Display Information The precision of displayed information should be no more than that necessary for the intended user task or decision-making activity.
8.3.2.2 Adequate Signal Duration Guideline 8.35: Adequate Signal Duration Information of short duration should be displayed for a sufficient duration to be detected reliably under expected user workload and environmental conditions.
For important information that is transient, it may be necessary to artificially extend the displayed event duration or enhance the display by employing instrument memory or “history” functions that would allow the user to recall such events. In some cases it may be appropriate for the displayed data to remain until dismissed or manually updated by the user. Examples include display of transient alarms/warnings or clinical advisories. 8.3.2.3 Infrequent Signals Humans are not very proficient at monitoring data continuously for events that occur infrequently. In systems requiring user “vigilance” for critical events, designers should provide means to automatically detect the event and notify the user or make the signal more obvious. Alarms and alerts are examples of such enhancements. 8.3.2.4 Data Update Rates Guideline 8.36: Speed and Frequency of Data Updates Information should be updated at a sufficient speed and frequency to support user tasks in all operating or service modes.
TAF-ER562X_WINGR-09-0401-C008.indd 321
11/8/10 11:44:13 PM
322
Handbook of Human Factors in Medical Device Design
Guideline 8.37: Need for Updates Unnecessary updates can be distracting and should be avoided.
Guideline 8.38: Automatic Updates Data or other visual information should not be automatically updated while they are being viewed if that update will impair reliable interpretation.
Automatic updating of the display can disrupt users’ visual scan patterns and cause a misread error, for example when new items are added to a list and cause the list items to “scroll.” In such situations, lists should be updated in response to a user-initiated action such as a “refresh” or paging command.
8.4 TYPES OF ELECTRONIC DISPLAY Once the user requirements for a given application are determined, the display choice will involve consideration of a number of display performance requirements such as size, resolution, luminance, contrast, refresh rates, color capabilities, and viewing angle requirements. At the same time, application and business requirements will inevitably present constraints and design trade-offs, such as space availability, portability, power requirements, heat production, durability, reliability, and cost. The following section presents a summary comparison of major display types (see also Table 8.2) followed by a brief discussion of each display and their most salient characteristics.
8.4.1 COMPARISON OF DISPLAY TECHNOLOGIES 8.4.2 CATHODE RAY TUBES (CRT) (SEE TABLE 8.2) Despite the more recent introduction of display technologies such as plasma, liquid crystal, and digital light processors, the CRT remains a ubiquitous display technology. Continuous improvements in image quality (resolution, good color rendition, high luminance, absence of flicker, and so on) and relatively low cost allowed the CRT to be the dominant display technology for over three decades. However, because CRTs tend to be large and heavy, they are not well suited for use in portable or compact devices requiring limited space. New display technologies, however, are rapidly surpassing the CRT in medical applications.
8.4.3 LIQUID CRYSTAL DISPLAYS (LCDS) LCDs are available in a wide variety of types, sizes, shapes, and performance and cost levels. Their advantages in packaging size, addressability, low power consumption and in recent displays, high image quality, have made them the display of choice for medical devices. With recent advances in LCD technologies, particularly in color displays using thin film transistors (TFTs) combined with high-performance lighting sources (fluorescent, electroluminescent, or light-emitting diodes), many of today’s LCDs are capable of presenting
TAF-ER562X_WINGR-09-0401-C008.indd 322
11/8/10 11:44:13 PM
TAF-ER562X_WINGR-09-0401-C008.indd 323
Large
Medium to large Large
Small
Small
Small
Miniature to large Small to large Small to medium Small
Size
Yes Limited
Low Low
High
High
Medium to high High
Yes
Yes
Yes
Limited
Limited
Limited
Yes
Low
High current Low
Yes
Color Capability
High
Power/ Voltage
Medium to high High
Medium
Low to high
High to very high Low to high
Medium to high Medium to high Low
Low to high
Luminance Capability
Medium to high Medium to high Medium to high
High
High
Low
Medium to high Medium to high Low
High
High
High
High
High
High
High
Low to medium Low to medium
High
High
Resolution Contrast
Good
Fair to good
Medium Good
Wide
Wide
Viewing Angle Uniformity
Low to medium
Low to medium
Low to medium
Low to medium
Low to medium
Good
Fair
Good
Medium Good to wide Wide Good
Wide
Wide
Wide
Low with Narrow Fair backlighting High in reflective mode Low to high Wide Good
Low to medium
Low to medium
Low to medium
Compatible Lighting Levels
Yes
Yes
Yes
Yes
Yes
Yes
“Character”—no “Graphic”—yes
Yes
Yes
Yes
Matrix Addressing
Source: Adapted from Snyder, H. L., Human Visual Performance and Flat panel Display Image Quality, Virginia Polytechnic Institute and State University, Blacksburg, 1980.
DLP (digital light processor)–based projection
LCD projection
Plasma
Cathode ray tube (CRT) Liquid crystal (TFT) Liquid crystal (STN) Liquid crystal (small, e.g., one- to two-line alphanumeric) Light-emitting diode (LED) Organic lightemitting diode Electroluminescent
Technology
TABLE 8.2 Qualitative Comparison of Display Technologies
Visual Displays 323
11/8/10 11:44:13 PM
324
Handbook of Human Factors in Medical Device Design
images that rival those of high-performance CRTs. The suitability of a given LCD depends on the task requirements and the demands of the environment.
8.4.4 TYPES OF LIQUID CRYSTAL DISPLAYS 8.4.4.1 Type of LCD by Method of Lighting Unlike CRTs and other displays, LCDs do not emit light. Instead, LCDs act as “light valves” that control either transmitted or reflected light. LCD panels can be reflective, transmissive, or transflective (partially transmissive and partially reflective), depending on their design and construction. Reflective LCDs use ambient light that enters through the front of the display and reflects it back to the user. Transmissive LCDs are designed to transmit light from a light source through a liquid crystal layer and out through the front to the user. The most common light sources for transmissive LCDs are electroluminescent panels or fluorescent tubes. A new generation of transmissive LCDs uses “white” LEDs as the light source. Transflective LCDs combine the features of transmissive and reflective LCDs so that in bright lights, ambient light is reflected from a semireflective surface behind the liquid crystal layer. In low light conditions, a backlight is used behind the semireflective surface to light the front of the display. Table 8.3 summarizes the advantages and disadvantages of each technology. Guideline 8.39: Reflective LCDs and Moderate to High Ambient Illuminence Reflective LCDs should be considered ambient illuminance is moderate to high, for example, in direct sunlight where other displays would be overwhelmed by high illuminance.
Guideline 8.40: Reflective LCDs Not for Low Light Conditions Reflective displays should not be used for low light conditions unless they employ edge or front lighting to produce minimum recommended luminance levels.
TABLE 8.3 Comparison of Transmissive, Reflective, and Transflective LCDs Transmissive LCD
Reflective LCD
Power consumption
Illuminated by “backlight” (typically fluorescent or electroluminescent) High
Color rendition
Wide color gamut
Contrast in darkness, low to moderate illuminance
High
Contrast in high illuminance
Low
Illuminated by ambient illumination or front lights Low (without front lighting) Narrow color gamut (often monochrome) Low Not suitable in low illuminance conditions without front light High
Display lighting
TAF-ER562X_WINGR-09-0401-C008.indd 324
Transflective LCD Illuminated by backlight or ambient illumination Low (with backlight turned off) to high Narrow color gamut Moderate
Moderate
11/8/10 11:44:13 PM
Visual Displays
325
Guideline 8.41: Transmissive LCDs for Low to Moderate Illuminance Transmissive LCDs are best suited for low to moderate illuminance conditions. High ambient lighting conditions cause a reduction in display contrast to the point where they can become difficult or impossible to read.
Guideline 8.42: Transflective LCD Application If a transflective LCD is to be used in both high and low lighting conditions, the display should be assessed in both modes across a full range of illuminance conditions and viewing angles.
8.4.4.2 Active-Matrix versus Passive-Matrix Displays Another distinction between LCDs is the way individual pixels in the display are activated. “Active-matrix” displays incorporate TFTs that allow each pixel to be turned on or off individually and randomly. Pixels within a “passive-matrix” display are activated when corresponding rows or columns of the display are activated sequentially. Active-matrix displays typically exhibit higher contrast and luminance across wider viewing angles than do passive-matrix displays. Passive-matrix displays are lower cost and consume less power than active-matrix displays. Guideline 8.43: Passive-Matrix Displays Passive-matrix displays should not be used for applications that require dynamic updating of graphical elements because they can exhibit “ghosting” of rapidly moving graphical elements.
Small-format LCDs, such as those used in handheld devices, have typically employed passive-matrix technology (see Figure 8.18). Active-matrix LCDs are increasingly common in smaller devices, such as infusion pumps and handheld personal digital assistants (PDAs) used for medical applications (See Figure 8.19). LCD applications that require high image quality and larger viewing angles are more likely to use active-matrix technology (see Figure 8.20).
8.4.5 LIGHT EMITTING DIODE (LED) DISPLAYS LED displays, which have been available for decades, are used in a wide variety of display applications. Common display uses include segmented digit displays, bar graphs, and light bars. Small dot matrix units have also been produced that display alphanumeric information
FIGURE 8.18 Small-format LCDs used in digital thermometers (left). Pulse oximeter (right) displays pulse waveform and prominent numeric readouts. (Courtesy of Nellcor Puritan Bennett (Covidien). With permission.)
TAF-ER562X_WINGR-09-0401-C008.indd 325
11/8/10 11:44:13 PM
326
Handbook of Human Factors in Medical Device Design
FIGURE 8.19 The main display (center) of this infusion pump employs a 4.7-inch diagonal passive-matrix LCD with white LED backlights. (Courtesy of Alaris Medical Systems (Care Fusion). With permission.)
for data and message displays. LCDs have largely replaced LED displays in many applications because they are lower cost and lower power consumption while providing greater flexibility in displaying characters and symbols. Individual LED lights (also referred to as discrete LEDs) have been widely used as status indicators. Newer technologies, such as so-called superbright LEDs, are suitable for high illumination conditions. In fact, these LEDs are sufficiently intense to be appropriate for a wide array of applications, including display backlighting, flashlights, vehicular brake lights, and outdoor traffic stoplights and signage. These LEDs have become widely used as light sources for displays. 8.4.5.1 Diffusing Elements and Light Pipes for LEDs In applications where it is not possible to place LEDs in direct line of sight as a status indicator, optical fibers or “light pipes” are used to transmit light from the LED to the indicator position. Where LEDs are used as a light source for displays, fiber-optic “panels”
FIGURE 8.20 LCD application with large-format displays used as companion monitors typically employ active-matrix technology.
TAF-ER562X_WINGR-09-0401-C008.indd 326
11/8/10 11:44:14 PM
Visual Displays
327
FIGURE 8.21 Fiber-optic panels consisting of bonded stands of optical fiber can be used to transmit light from discrete LED(s) to produce relatively uniform backlight for a display or control panel. (Courtesy of Lumitex, Incorporated. With permission.)
can be used to provide uniform lighting produced by discrete LEDs. These panels, shown in Figure 8.21, consist of very thin fiber-optic strands bonded together to form a flat membrane. The fibers are terminated into a bundle with one or more LEDs so that the light is distributed across the multiple fibers that form the flat membrane. These panels are also suitable for use as backlighting for control panels and keypads.
8.4.6 ORGANIC LEDS With organic LEDs (OLEDs), polymeric dyes are deposited on a layer of glass or thin film plastic to form red, green, and blue pixels that emit light when activated. The layer on which the OLEDs is deposited can be made flexible (for use in nonrigid or curved surfaces) or transparent for use in special applications such as visor-mounted displays. Because OLEDs are self-luminous, they do not require a backlight and consequently need very shallow mounting depth and consume little power. Among their other advantages are high luminance, high contrast at wide viewing angles, wide color range, and rapid response time. Although OLEDs are available in sizes suitable for handheld devices (e.g., cell phones, digital camera view finders), the technology allows them to be produced in virtually any size. They are being considered as an alternative to LCDs (Bedell D., 2004). One of the shortcomings of current OLEDs is that they have a limited operating life (e.g., less than 10,000 hours), making them unsuitable for use in devices where the display must remain on continuously. As research and development succeeds in increasing their longevity, they may find wide application in medical devices.
8.4.7 ELECTROLUMINESCENT DISPLAYS Electroluminescent displays (ELDs) are flat-panel technologies that emit their own light and therefore do not need ambient light or backlighting. They consist of a luminescent phosphor layer between insulating layers and a matrix of electrodes (in rows and columns). When voltage is applied to a row and column, the phosphor in the area of the intersection emits light as a picture element (pixel). ELDs are useful where ruggedness, wide viewing angle, speed, high luminance, high contrast, wide temperature tolerance, and low power requirements are needed. Although color ELDs have been developed for special applications, virtually all commercially available ELDs are monochromatic amber, blue, or yellow. A few have the ability to display limited colors (e.g., eight colors and black). ELDs have been available in a variety of formats ranging from one-line fixed-format alphanumeric displays to bitmapped displays
TAF-ER562X_WINGR-09-0401-C008.indd 327
11/8/10 11:44:14 PM
328
Handbook of Human Factors in Medical Device Design
FIGURE 8.22 Example of “dead-front” transilluminated panel with backlight turned off (left). When the backlight is activated, the word or symbol in the translucent portion of the panel becomes visible (right).
ranging in size from approximately 3 to 10.4 inches in diagonal. Alternate technologies, such as active matrix LCDs, have resulted in a decline in the use of ELDs in recent years.
8.4.8 TRANSILLUMINATED DISPLAYS Transilluminated displays typically consist of a panel with single or multiple translucent areas that form meaningful words, abbreviations, numbers, symbols, on an opaque background. Light sources behind the translucent areas are activated on command to illuminate the desired words, symbols, and so on. A common type of transilluminated display is the “dead-front” panel designed so that the word or symbol being displayed remains invisible until the light source illuminates (see Figure 8.22). Other examples of transilluminated displays are simple indicator lights and legend lights. A legend light is typically a discrete light with a translucent cover on which words or symbols are inscribed. Transilluminated displays offer simple, relatively low-cost presentation of status information but have limited flexibility in terms of dynamic information.
8.4.9 LARGE-SCREEN/PROJECTION DISPLAYS Although projection displays have existed for a number of years, many of the early systems were expensive and had relatively poor image quality. Recent developments in projection technologies and computer video and graphic systems have greatly increased the availability of systems that present large, high-quality images at relatively low cost. 8.4.9.1 Application Large-screen displays are useful in situations such as the following: • When a group of users must interact as a team using the same displayed information • When one or more members of a team of users must move about while referring to the displayed information and where portable or fixed conventional displays are not usable or visible at their assigned position(s) • When constraints such as space, cost, or technical limitations preclude the use of individual displays for each team member to view commonly used information • When it is desirable to have general information available to persons who might interrupt ongoing group operations by looking over the users’ shoulders to see individual displays Guideline 8.44: Use of Large-Screen Displays Large-screen displays should be used only when the spatial and environmental conditions allow satisfactory viewing conditions for critical users. Factors such as viewing distance,
TAF-ER562X_WINGR-09-0401-C008.indd 328
11/8/10 11:44:15 PM
Visual Displays
329
range of viewing angles, lack of interference from intervening objects/personnel, and ambient illumination should be considered.
8.4.9.2 Control and Content of Displayed Information Guideline 8.45: Display Controls Where team users are likely to be in different locations, controls for such displays should either be embedded in the display’s physical packaging or located in a separate workstation or control device.
Guideline 8.46: Preventing Unauthorized Display Viewing The impact of unauthorized viewing of information should be considered. Large displays with wide viewing angles may not be appropriate if they will display confidential patient information in a semipublic area.
Guideline 8.47: Control of Group Displays Control of group displays should ensure that critical information cannot be modified or deleted accidentally or arbitrarily.
8.4.9.3 Viewing Distance Guideline 8.48: Maximum Viewing Distance for Large Displays The display should not be farther from an observer than will enable readability of critical detail presented on the display. For textual information, for example, the maximum viewing distance should be such that the smallest character height will subtend no less than 20 minutes (1/3 degree) of visual arc.
The minimum distance between the viewer, and the display should also be considered. Viewing too closely may cause the display raster or pixel structure (which would ideally not be obvious) to be visible, leading to annoyance (e.g., “screen-door” effect) or interference with readability. Guideline 8.49: Minimum Viewing Distance for Large Displays For most applications, the projection display should not be viewed closer than half the display width or height—whichever is greater. The allowable minimum viewing distance may be reduced with the use of higher-resolution projections (resulting in a greater number of pixels per inch).
8.4.10 SCALE INDICATORS 8.4.10.1 General Scale indicators, such as gauges or dials, fall into two general categories: 1. Moving pointer with fixed scale. Scales can be: • circular • curved (arc) • horizontal straight • vertical straight
TAF-ER562X_WINGR-09-0401-C008.indd 329
11/8/10 11:44:15 PM
330
Handbook of Human Factors in Medical Device Design
2. Fixed pointer with a moving scale. These can also be: • circular • curved (arc) • horizontal straight • vertical straight Although mechanical and electromechanical gauges continue to be used in medical applications, many of them are being replaced by the visual representation of such gauges via electronic displays. Whether dealing with physical scale indicators or their electronic representations, the following guidelines generally apply (ANSI/AAMI HE-48, 1993). 8.4.10.2 Applications The selection of scale indicators should be based on the criteria presented in Table 8.4. In general, scale indicators should be used to do the following: 1. Display quantitative information in combination with qualitative information (such as trend and direction of motion). 2. Display only qualitative information where the use of alphanumeric readouts or counters is inappropriate. For example, a gauge may be more appropriate than a counter when an approximate quantitative value is needed rather than an exact value. In such applications, the speed of recognition of an approximate number has more utility than presenting an exact value.
8.4.11 POINTERS 8.4.11.1 Design Characteristics for Gauges with Pointers Whether physical dials or their electronic representations, dials with pointers should have the following characteristics: Guideline 8.50: Dial Pointer Length The pointer should extend to, but not obscure, the shortest graduation marks.
Guideline 8.51: Dial Pointer Width The width of the pointer where it intercepts the graduation marks should not exceed the width of the intermediate marks.
Guideline 8.52: Coaxial Pointers Whenever precise readings are required, no more than two coaxial pointers (pointers with the same pivot point) should be mounted on one indicator face unless more than two parameters are closely related.
Guideline 8.53: Minimizing Parallax of Pointers In mechanical or electromechanical gauges, the pointer should be mounted as closely as possible to the face of the dial to minimize parallax.
TAF-ER562X_WINGR-09-0401-C008.indd 330
11/8/10 11:44:15 PM
TAF-ER562X_WINGR-09-0401-C008.indd 331
Most economical of space and illumination. Scale length limited only by number of counter drums.
Source: Reprinted from U.S. Department of Defense, MIL-STD 1472F: Design Criteria Standard: Human Engineering (1999).
General
Good Fair Pointer position readly controlled and monitored. Simplest relation to manual control motion Requires largest exposed and illuminated area on panel. Scales length limited unless multiple pointers used.
Good Most accurate monitoring of numerical setting. Relation to motion of setting knob less direct than for moving pointer. Not readable during rapid setting. N/A No gross position changes to aid monitoring.
Fair Relation to motion of setting knob may be ambiguous. No pointer position change to aid monitoring. Not readable during rapid setting. Poor No position changes to aid monitoring. Relation to control motion somewhat ambiguous. Saves panel space. Only small section of scale need be exposed and illuminated.
Tracking
Poor Numbers should be read. Position changes not easily detected.
Poor Numbers should be read. Position changes not easily detected.
Good Minimum time and error for exact numerical value.
Counters
Qualitative Good Information Location of pointer easy. Numbers and scale need not be read. Position change easily detected. Difficult to judge direction and magnitude of deviation without reading numbers and scale. Setting Good Simple and direct relation of motion of pointer to motion of setting knob. Position changes aids monitoring.
Fixed pointer Fair May be difficult to read while scale is in motion.
Moving pointer
Quantitative Fair Information May be difficult to read while point is in motion.
Use
Scales
TABLE 8.4 Application of Various Types of Mechanical/Electromechanical Displays Flags
Limited application.
N/A
N/A
Limited application.
N/A
Good N/A Minimum time and error for exact numerical value. Provides reference records. Poor Good Easily detected Economy of space.
Printers
Visual Displays 331
11/8/10 11:44:15 PM
332
Handbook of Human Factors in Medical Device Design
Guideline 8.54: Numeral Orientation on Stationary Scales Numerals on stationary scales should be oriented vertically.
Guideline 8.55: Numeral Position on Scales Numerals should be positioned on scales so that the pointer does not obscure them.
8.4.11.2 Numerical Progression Guideline 8.56: Clockwise Numerical Progression on Scales The numerical progression on fixed circular scales should increase in the clockwise direction, from left to right for linear horizontal scales, and from the bottom up for linear vertical scales.
Guideline 8.57: Numerical Progression Increments Numerals should progress by 1s, 5s, or 10s. Other progression schemes, such as by 2s or 3s, can be disruptive to rapid and accurate reading and should be considered only after careful evaluation of the option with users.
Guideline 8.58: Whole Number Prefixes for Large Values Where large numerical values are used in a scale, scale markings should display whole number prefixes to enhance readability. To read the display, the user will multiply the number by a scale factor (such as 10s, 100s, or 1000s). An example is a pressure gauge that shows gradations in whole numbers from 0 through 5, denoting pressure in thousands of pounds (the “× 1000” would be clearly displayed on the display face).
Guideline 8.59: Use Linear Scales Linear scales should be used except when user information requirements clearly dictate the use of nonlinear scales.
Guideline 8.60: Avoid Decimals Decimals should be avoided whenever possible. When decimals are used, leading zeros in front of the decimal point should be omitted, unless driven by convention or other standards (Sanders and McCormick, 1993).
8.4.11.3 Qualitative Indications on Scales In addition to numeric values, other graphical treatments on the face of scale indicators may be used to convey information, such as desirable operating range, dangerous operating level, caution, undesirable, or inefficient condition, and so on. Guideline 8.61: Operating Range Indications When operating conditions fall within an indicated range on the scale, these areas should be denoted with a pattern, shape, color, or other indication on the face of the instrument (see Figure 8.23).
TAF-ER562X_WINGR-09-0401-C008.indd 332
11/8/10 11:44:15 PM
Visual Displays
333 Normal (green)
Caution (yellow)
Caution (yellow)
Cold (yellow)
Normal (green) Hot (red)
Danger (red)
FIGURE 8.23 Illustration of color coding used in qualitative displays. (From Sanders, M.S. and McCormick, E.J., Human Factors in Engineering and Design. McGraw-Hill, New York, 1993. With permission.)
8.4.11.4 Break in Circular Scale Guideline 8.62: Circular Scale Spacing In a circular scale, a space in the scale of at least 10 degrees of arc should be provided between the two ends of the scale except on multirevolution instruments such as clocks (see Figure 8.24).
8.4.11.5 Scale Face Opening Guideline 8.63: Scale Face Opening If the display will be used for adjusting or setting a desired value, the unused portion of the dial face should be covered, and the open window should be large enough to permit at least one numbered graduation to appear on each side of any setting.
>10°
FIGURE 8.24 The ends of circular scale, such as the one on this sphygmomanometer, should be separated by a gap of 10 degrees or greater. (Courtesy of Heine. With permission.)
TAF-ER562X_WINGR-09-0401-C008.indd 333
11/8/10 11:44:15 PM
334
Handbook of Human Factors in Medical Device Design
8.5 SPECIAL APPLICATIONS 8.5.1 INTRODUCTION This section describes two special display systems. The first, touch screens, are displays that also serve as input devices. The second system, the head-mounted display, is a nonconventional type of display technology for use in special medical applications.
8.5.2 TOUCH SCREENS A touch screen combines a display with an input device that responds to contact with the user’s finger or object (e.g., stylus). A finger touch produces a displayed response, such as moving a cursor or changing the appearance of a selected screen element. The popularity of touch screens in medical devices is largely attributable to the fact that touching objects on a screen is a natural, intuitive action that can be accomplished without having to reach for another object such as a mouse, joystick, or light pen (see Figure 8.25). Most touch screens can be cleaned easily with standard cleaning materials. Touch screens fall into two general categories. The first involves the use of a transparent overlay that is affixed to the display face. The overlay contains conductive or capacitive sensing elements that translate the touch position into horizontal and vertical coordinates that are transmitted as numerical coordinates to the device’s computer. The second category of touch screen is activated when a finger touch interrupts an infrared or acoustic signal that is normally propagated across the display face. When the user touches the screen, the user’s finger interrupts the horizontal and vertical signal paths, the intersection of which determines the location of the touch (Arnaut and Greenstein, 1988). 8.5.2.1 Touch-Screen Characteristics When choosing a touch screen, it is important to match the characteristics of the touch screen to the demands of the application and the user. Candidate touch screens should be tested with users in the target environments under typical use conditions (e.g., using latex gloves).
FIGURE 8.25 The use of touch screens in medical devices is very popular because they afford natural, intuitive interaction, alleviating the need for another object, such as a mouse, stylus, or light pen. The touch screen shown here is for an in vitro diagnostic system. (Courtesy of Abbott Laboratories. With permission.)
TAF-ER562X_WINGR-09-0401-C008.indd 334
11/8/10 11:44:16 PM
Visual Displays
335
8.5.2.1.1 Touch-screen resolution Touch-screen resolution refers to the number of discrete touch positions that are resolvable in a horizontal or vertical axis (usually expressed in terms of number of touch points per linear distance or square area). Higher-resolution touch screens are theoretically better suited for applications requiring continuous motion, such as when moving a cursor or “dragging-anddropping” actions. Lower-resolution touch screens (e.g., those using infrared sensors) can be used when interactions are limited to coarse, discrete actions, such as when using a touch screen to select relatively large screen “buttons” or objects. With the possible exception of infrared touch screens using widely spaced sensors, users are not likely to notice resolution differences among various touch-screen technologies. 8.5.2.1.2 Target size and spacing for touch screens Exercise care when selecting the size and spacing of selectable screen objects or “buttons.” According to Cushman and Rosenberg (1991), touch screen targets should have the following characteristics. Guideline 8.64: Object Size for Finger-actuated touch screens For finger-actuated touch screens, the minimum dimension of selectable screen objects (such as a screen button) should be 13 × 13 mm (0.5 × 0.5 inches) or greater.
Guideline 8.65: Object Spacing for Finger-actuated touch areas For finger-actuated touch areas, spacing between them should be 6 mm (0.25 inches) or greater.
Smaller touch targets or reduced spacing should be tested with users in simulated tasks, with special attention given to input error rates. Figure 8.26 shows an example of a user interface designed for use with a touch screen. Control buttons are >13 mm in width and height, and center-to-center distances are >19 mm.
FIGURE 8.26 Example screen from a laboratory in vitro diagnostic instrument that employs a 17-inch touch screen. (Courtesy of Abbott Laboratories. With permission.)
TAF-ER562X_WINGR-09-0401-C008.indd 335
11/8/10 11:44:16 PM
336
Handbook of Human Factors in Medical Device Design
8.5.2.2 Parallax A common problem with touch screens is parallax, or the misalignment between the perceived position of an object on the touch screen and the position of the associated touch area for that object. Such misalignment can cause the user to miss the intended target or activate an unintended one. Parallax occurs when (1) there is a gap between the touch-sensing surface and the display surface and (2) the user’s eye position differs from the intended viewing position. Parallax is often more pronounced when CRTs with curved display surfaces are integrated with a flat touch-screen surface, where the curvature of the display creates a gap between the sensing surface and the display surface, especially at the edges. In Figure 8.27, viewer A and viewer B view the same image from two different vantage points. If the touch screen is calibrated from the vantage point of viewer A, a person viewing the display at position B will most likely miss the touch target. Some points to consider for combating parallax follow. Guideline 8.66: Minimize Distance between Touch and Display Surfaces To minimize parallax, the distance between the touch surface and the display surface should be minimized.
When touch screens are integrated with flat-panel displays, they are typically bonded very closely to the display surface and, consequently, result in minimal parallax. Because Apparent button position, viewer A
Line-of-sight, viewer A
Button image
Line-of-sight, viewer B Display surface (CRT)
Apparent button position, viewer B
Touch screen surface
FIGURE 8.27
Exaggerated illustration of parallax error at two different viewing positions.
TAF-ER562X_WINGR-09-0401-C008.indd 336
11/8/10 11:44:16 PM
Visual Displays
337
acoustic wave touch screens allow the touch-screen surface and the display surface to be coincident, parallax is effectively eliminated. Guideline 8.67: Increase Object Size and Distance Where parallax is unavoidable, the size of selectable screen objects and distance between them should be increased to minimize parallax effects.
8.5.2.3 User/Environment Considerations 8.5.2.3.1 Surface Contamination Designers should anticipate display surface contamination and the need for frequent cleaning with decontaminating solutions. Guideline 8.68: Safe Cleaning Materials for Touch Screens Requirements and tests should verify that cleaning (and the possible use of excessive or harsh cleaning solutions) does not harm the touch surface or the touch-screen electronics.
Guideline 8.69: Temporary “Lockout” for Cleaning of Touch Screens With devices that are likely to be cleaned during operation, a temporary “lockout” function that disables the touch screen during cleaning should be considered.
Guideline 8.70: Evaluate Touch Screen Function When Dirty Touch screens should be evaluated in situations where they are likely to become contaminated and cannot be cleaned immediately. For example, significant surface contamination on an acoustic wave touch screen can cause those spots to be inoperative until they are cleaned.
8.5.2.3.2 Use of Gloves Guideline 8.71: Use of Gloves with Touch Screens When choosing a touch screen, consideration should be given to whether gloves are likely to be used. Certain touch screens, such as those that depend on capacitive sensing, may not work reliably when used by a person wearing latex or similar gloves.
8.5.2.3.3 Need for Calibration When first installing a touch screen, it is usually necessary to calibrate the touch screen so that surface touch positions are properly aligned with the corresponding symbols or areas on the display. During use, it is not uncommon for the touch screen to require recalibration caused by, for example, differing eye positions during use or shift of the screen image (as when someone uses the monitor controls to change the screen’s horizontal or vertical position). Although a relatively simple task, performing a calibration may not be feasible at critical moments or when a device is used concurrently by several operators. If limitations of a touch screen cannot be effectively mitigated (e.g., using backup cursor control devices or locking monitor controls), a touch screen may not be a good design choice.
8.5.3 HEAD-MOUNTED DISPLAYS A disadvantage exhibited by most conventional displays is that the user may have to alternate his or her gaze between the primary point of interest (such as the patient or surgical
TAF-ER562X_WINGR-09-0401-C008.indd 337
11/8/10 11:44:17 PM
338
Handbook of Human Factors in Medical Device Design
site) and the display. This action could result in delays in acquiring necessary information, momentary disorientation, and increase of errors. Head-mounted or helmet-mounted displays (HMDs) can eliminate such problems by combining displayed information with the user’s natural field of view. In such cases, the HMD can provide spatial information not possible with conventional fixed displays. More advanced applications of HMDs are being proposed, including the use of integrating ultrasound or other imaging technologies to allow a surgeon to visualize the inside of a patient’s body during surgical or diagnostic procedures. Current HMDs can be found in the following configurations: • Virtual reality HMDs. Worn as a helmet or goggles, an HMD typically presents the user with a stereoscopic, computer-generated view of the world. Stereoscopic presentation allows the user to perceive depth cues not available in conventional twodimensional displays. The typical HMD uses miniature CRTs or LCDs viewed through lenses aligned to the line of sight for each eye and mounted into a headworn housing (see Figure 8.28). • See-through HMDs. See-through HMDs are similar to virtual reality displays except that with the see-through displays, the real world is viewed directly through semitransparent mirrors while the computer images, generated by miniaturize displays, are reflected into the user’s eyes via the same mirrors. • Retinal scanning. In a third type of HMD, the need for intermediary displays is eliminated by projecting a modulated beam of light directly into the user’s eye. The beam, controlled by horizontal and vertical scanners, produces a rasterized image directly on the user’s retina. 8.5.3.1 Image Orientation and Position In applications where computer-generated imagery is combined with real-world imagery, accurate registration of the images is essential. This is particularly challenging given that the user’s body and head are free to move and rotate. To properly combine the computergenerated images with the real-world view, any change in the position and orientation of the user’s head must be sensed in three-dimensional space using optical, magnetic, or inertial
FIGURE 8.28 outside scene.
Example of a head-mounted virtual reality display having no reference to the actual
TAF-ER562X_WINGR-09-0401-C008.indd 338
11/8/10 11:44:17 PM
Visual Displays
339
sensors. These data are fed to the imaging computer that calculates the proper image orientation and movement and updates the image with proper orientation and registration in real time. 8.5.3.2 Computing Bandwidth Early virtual reality systems provided images that were rather primitive and responded slowly when the user moved his or her head. In recent years, digital image processing bandwidth has increased significantly. However, rapid real-time update of large highresolution images combined with fast head movement without image degradation or delay may continue to be a problem. 8.5.3.3 Display Resolution Current HMDs employ small LCDs or CRTs. Despite their small size, the basic technologies that produce the displays maintain pixel densities that are equivalent to those of conventional larger displays. For example, current technologies categorize high-resolution LCDs as ranging from around 85 to 133 pixels per inch (Dell Computer Inc., 2003). However, when displays with such pixel densities are viewed under magnification to fill the entire visual field, as they would in an HMD, the images appear coarse and discontinuous. Consequently, work is under way to develop higher-resolution “microdisplays” with pixel densities of 2,000 dots per inch on a 1-inch-square format. The availability of such displays most certainly will assist in advancement of the state of the art for future medical applications.
8.6 CASE STUDY 8.6.1 SPECIFYING AND SELECTING A DISPLAY FOR A CARDIAC OUTPUT MONITOR 8.6.1.1 Introduction During the development of the Hospira Q2TM Plus CCO/SO2 Monitoring System (see Figure 8.29), an instrument that measures continuous cardiac output and oxygen saturation, a new color display was to replace the monochrome CRT of the previous-generation instrument. Because the design goals of the system included small size, low weight, and low power consumption, several LCDs were considered. However, one of the concerns of the design team was the LCD’s ability to perform at wide viewing angles. Thus, the team took a systematic
FIGURE 8.29
The Hospira Q2 TM Plus CCO/SO2 Monitoring System.
TAF-ER562X_WINGR-09-0401-C008.indd 339
11/8/10 11:44:17 PM
340
Handbook of Human Factors in Medical Device Design
approach to (1) determine the user viewing requirements for the display and (2) objectively measure the performance of candidate LCDs to meet these viewing requirements. 8.6.1.2 Analysis Members of the design team visited various departments within 10 different hospitals to gather information on current (or similar) instrument installations. They conducted user interviews, observations, and simple measurements to gather information on: • • • • • • • •
Device mounting height Viewing distance Seated/standing posture Device tilt angle Maximum horizontal viewing angle Room illumination Device brightness setting Likes and dislikes
Field visits revealed that the devices were placed in one of three basic positions: 1. High mount, above the user’s eye level (maximum height) (see Figure 8.30) 2. Medium mount; typically on a shelf at approximate eye level (see Figure 8.31) 3. Low mount, on bedside table (minimum height) (see Figure 8.32) 8.6.1.2.1 Viewing Angle Calculations Several factors directly affect the vertical viewing angle: the device’s height above the floor, the user’s viewing distance, the user’s eye height, and the device’s tilt angle (see Figure 8.11).
FIGURE 8.30
Example of a monitor placed on a high shelf.
TAF-ER562X_WINGR-09-0401-C008.indd 340
11/8/10 11:44:18 PM
Visual Displays
341
FIGURE 8.31
Example of medium mounting height.
FIGURE 8.32
Example of low mounting height.
TAF-ER562X_WINGR-09-0401-C008.indd 341
11/8/10 11:44:18 PM
342
Handbook of Human Factors in Medical Device Design
TABLE 8.5 Calculations of Worst-Case Viewing Angles for High Mounting Display Position with 5th-Percentile Male and Female (U.S. Population) Eye Heights in Seated and Standing Positions User 5th percentile female—standing 5th percentile female—stool (30″) 5th percentile female—stool (24″) 5th percentile male—standing 5th percentile male—stool (30″) 5th percentile male—stool (24″)
User Eye Ht (mm)
Difference (mm)
Viewing angle
1420 1452 1300 1595 1502 1350
587 555 707 412 505 657
52.1° 50.5° 57.1° 42.0° 47.8° 55.2°
Source: Derived from Pheasant, S., Bodyspace: Anthropometry, Ergonomics and the Design of Work, Taylor & Francis Ltd., London, 1996. Note: Bold indicates worst case scenario.
Worst-case viewing angles were calculated using the maximum anticipated monitor height at an assumed minimum viewing distance of 18 inches and lowest anticipated eye height. Eye heights were determined for the shortest anticipated users, the 5th percentile, from anthropometric tables (Pheasant, 1996) for both males and females in standing and two seated positions (see Table 8.5). From these data, the worst-case viewing angle was found to be 57.1 degrees for the 5th percentile female user seated on a 24-inch stool. This angle was adopted as the worst-case angle for the high mounting position. Similar viewing angle calculations were made for low and medium height positions. Horizontal viewing angles limits were derived from direct observations during field visits. Table 8.6 summarizes horizontal and vertical viewing angle requirements for the display. 8.6.1.2.2 Display Requirements/Acceptance Criteria A display standard (ANSI/HFES 100, 1988) was used to establish minimum requirements for the chosen display. The team consensus was that a candidate display would be acceptable if it met or exceeded the display requirements shown in Table 8.7 at the full range of viewing angles (shown in Table 8.6). 8.6.1.2.3 Display Testing Measurements Each of the candidate displays was placed on a fixture that allowed adjustments in pitch (tilt) and azimuth (left and right rotation) to simulate horizontal and vertical viewing angles, TABLE 8.6 Viewing Angle Requirements Determined for the Cardiac Monitor Viewing Angle Requirement Horizontal viewing angle High mount viewing angle (viewed from below the display) Low mount viewing angle (viewed from above the display)
TAF-ER562X_WINGR-09-0401-C008.indd 342
Angles ±45 degrees –57 degrees +64 degrees
11/8/10 11:44:19 PM
Visual Displays
343
TABLE 8.7 Display Requirements Display Characteristic
Requirement
Luminance Preferred contrast Minimum contrast Minimum character height
35 cd/m2 7:1 3:1 20 to 22 minutes of visual arc (used to specify size of displayed characters)
Source: Derived from ANSI/HFES 100, 1988)
adjustable in increments of 5 degrees. A 1/3-degree spot photometer (Minolta LS 110), mounted on a tripod, was placed level with the display, directly in front (perpendicular to the line of sight) at a distance of 1 m (see Figure 8.33). The photometer was used to measure the luminance of black or white 1-inch square targets (to measure minimum and maximum luminance, respectively) on the display face. The resultant luminance measures were used to assess compliance with minimum luminance and contrast requirements. Display measurements were initiated “normal” to the display (i.e., 0 degrees horizontal and vertical) and were incremented to a maximum angle of 65 degrees or until either luminance or contrast criteria fell below specified levels. In some cases, the boundaries were assessed first and were then decremented toward 0 degrees. 8.6.1.2.4 Candidate LCD Assessment A computer spreadsheet was developed to record the data gathered during the photometric measurements. Each of the “major” cells of the spreadsheet contained data for a specified horizontal and/or vertical tilt angle (in 5-degree increments). The center of the spreadsheet was the initiating point at 0 degrees horizontal and 0 degrees vertical (signifying that the display was pointed directly at the meter with no tilt).
FIGURE 8.33 Photometric measurement of a candidate display. The display stand allows the display unit to be tilted and swiveled in increments of 5 degrees.
TAF-ER562X_WINGR-09-0401-C008.indd 343
11/8/10 11:44:19 PM
344
Handbook of Human Factors in Medical Device Design Horizontal viewing angle Left 5° increments
Up 5° increments
0°
Right 5° increments
Acceptable viewing angles (horizontal and vertical)
Horizontal viewing angle
Vertical viewing angle
Vertical viewing angle
0°
Maximum luminance (Lmax) Minimum luminance (Lmax) Contrast (Lmax/Lmin)
Down 5° increments
FIGURE 8.34 Each of the cells of the spreadsheet (shown greatly reduced) that met or exceeded the minimum acceptance criteria for luminance and contrast is shaded in gray. The center cell (shown magnified) illustrates the contents of a cell. The display data are shown for the display incorporated in the final device.
Figure 8.34 shows the resultant spreadsheet (in reduced form) with entries that spanned –65 degrees to +65 degrees in horizontal and vertical angles. Each cell contains the position designation (tilt and horizontal angle), the maximum luminance (luminance of white square), the minimum luminance (luminance of black square), and the contrast ratio, which was automatically calculated. The data in each cell were assessed to determine whether they satisfied the acceptance criteria previously specified. The shaded portions of the spreadsheet indicate those cells that met or exceeded the minimum luminance (≥35 cd/m 2) and preferred contrast values (≥7:1). The unshaded cells in the figure are those that failed to meet either criterion. The vertically oriented rectangle shown on the matrix depicts the horizontal and vertical viewing angle requirements specified in Table 8.6. A display whose shaded area extended to the rectangle or beyond was deemed to be acceptable. Conversely, a display that failed to fill the rectangle would not fulfill the viewing angle requirements and would be deemed unacceptable. The display data depicted in Figure 8.34 show that the criteria were satisfied for all viewing angles except for the cells at the top of the spreadsheet, representing the uppermost 15 degrees (as when the display is viewed from above). This display was deemed acceptable because no other display in the test performed as well, and the 15-degree shortfall could be compensated for by the presence of a fold-down stand on the front end of the unit that allowed the display to be tilted upward by 15 degrees.
TAF-ER562X_WINGR-09-0401-C008.indd 344
11/8/10 11:44:19 PM
Visual Displays
345
8.6.2 DISPLAY USABILITY TEST: USER STUDY TO ASSIST IN DISPLAY SELECTION FOR A PATIENT-CONTROLLED ANALGESIA INFUSION PUMP 8.6.2.1 Introduction The Hospira LifeCare PCA® 3 Infusion System (shown in Figure 8.35) is an infusion pump designed to administer pain medication. The “patient-controlled analgesia” (PCA) function enables the patient to “request” a preprogrammed dose of medication by pressing a pendant-mounted button. During the development, a study was conducted to assess user preferences and usability of four candidate displays in different viewing and lighting conditions. The data were to be used to guide the design team in selecting an acceptable display. This case study illustrates how ratings of target users, collected in a well-controlled study, can be used to guide the selection of displays (or other user-interface elements). When such methods are absent, designers often depend on their own subjective judgments or manufacturer’s claims to select a device or verify a design, increasing the risk that the device does not meet the needs of users in their target environments. 8.6.2.2 Method 8.6.2.2.1 Study Participants A total of 51 participants, 45 registered nurses and six licensed vocational nurses, were recruited from local area hospitals (the Silicon Valley of California). The nurses were
FIGURE 8.35 Hospira LifeCare PCA® 3 Infusion System. The LCD (in the upper right portion of the control panel) is an LCD with a yellow-green backlight.
TAF-ER562X_WINGR-09-0401-C008.indd 345
11/8/10 11:44:20 PM
346
Handbook of Human Factors in Medical Device Design
representative of the intended user population in terms of clinical experience, age, gender, and height: • • • •
Hospital experience ranged from 1 to 30 years (average 14 years). Gender: 44 females, seven males. Height ranged from 52.5 to 73.4 inches (average 65.2 inches). Infusion pump usage: 40 participants indicated that they were frequent users of infusion pumps (three to five times per week).
8.6.2.2.2 Displays and Infusion Pumps Two sets of five infusion pumps were employed in this study. Four of the pumps were new prototypes equipped with four candidate displays. A fifth pump, a previous-generation pump with its standard display, was included in the study to serve as a “benchmark.” A description of the candidate displays and their identifying labels in the test are shown in Table 8.8. 8.6.2.2.3 Viewing Conditions • Viewing height. Duplicate infusion pumps and displays were placed in two separate test rooms. The pumps in one room were mounted on IV poles in a “low” position with the display at approximately 40 inches from the floor. The infusion pumps in the second room were mounted in a “high” position with displays at approximately 60 inches, equivalent to the average eye height for the U.S. female population (Pheasant, 1996). The heights were chosen to simulate a reasonable range of mounting heights given that multiple pumps are often mounted on the same pole. • Room illumination. Displays in both rooms were viewed under two different lighting conditions. A “high” lighting level was approximated with room overhead fluorescent lights, augmented with free-standing halogen floodlights aimed at the ceiling (to provide a relatively high level of diffuse ambient illumination). The average illuminance at the display face was approximately 500 lux. The “low” lighting level was used to simulate patient rooms in near darkness. For this condition, overhead and auxiliary lights were turned off with ambient light being furnished by hallway lights through an open door. The average illuminance at the display face was approximately 2 lux. Both high and low lighting levels were chosen on the basis of light levels gathered during field observations of typical hospital rooms. All participants were asked to perform basic tasks and rate each of the displays in each of the viewing conditions shown in Table 8.9. TABLE 8.8 Description of Candidate and Baseline Displays Display Description LCD, transflective LCD, transflective with bright LED backlighting LCD, transmissive LCD, transmissive, with bright LED backlighting Previous device, LCD
TAF-ER562X_WINGR-09-0401-C008.indd 346
Color
Test Label
Yellow-green Yellow-green Yellow-green Yellow-green Blue
X Z Y W V
11/8/10 11:44:21 PM
Visual Displays
347
TABLE 8.9 PCA® 3 Test Viewing Conditions Viewing Height
Room Illumination
Dim Bright
Low viewing position
High viewing position
“LowDim” “LowBright”
“HighBright” “HighDim”
8.6.2.3 Procedure In each room, the participants performed simple programming tasks on each of the five infusion pumps in random order. After each session, participants rated each pump display for readability and overall usability (via questionnaire). Participants were then asked to move to the next pump and repeat the process. After rating all pumps, participants were asked to rank order all five pump displays from “best” (rank = 1) to “worst” (rank = 5). The procedure was performed in both high and low illumination conditions before participants moved to the next room (with different viewing heights), where the procedure was repeated. The presentation order of lighting levels was counterbalanced, and the presentation order of pumps was randomized among viewing height sessions to minimize possible biases due to order effects. 8.6.2.4 Results 8.6.2.4.1 Display Rankings across Viewing Conditions The average ranks, shown in Figure 8.36, were analyzed statistically, showing the following results (note that ratings with lower average ranks are preferred over those with higher numeric ranks): • In the LowDim (low mounting height, dim illumination) condition, all displays received statistically equivalent ratings (p >.05).
Least favored
Most favored
Display average rank by viewing condition
4 V W X Y Z
3
2
1 Lowdim
Lowbright
Highdim
Highbright
Viewing condition
FIGURE 8.36 Average rank data for five displays (four candidate displays and benchmark). Participants ranked displays on a scale of 1 to 5 (1 was the favorite display).
TAF-ER562X_WINGR-09-0401-C008.indd 347
11/8/10 11:44:21 PM
348
Handbook of Human Factors in Medical Device Design
• In the LowBright condition, displays W, Y, and Z were ranked more favorably than X and V (p <.05). • In the HighDim condition, displays W, X, and Y were ranked more favorably than Z and V (p <.05). • In the HighBright condition, W, X, and Y were ranked more favorably than V and Z (p <.05). 8.6.2.5 Conclusions/Recommendations This study was designed to assess the different display alternatives under a wide range of lighting conditions and viewing positions. User ratings showed that not all displays performed consistently well across all viewing conditions. Two of the displays, Y and W, both of which were transmissive displays, received preferred rankings across all viewing conditions. Transflective displays X and Z also received among the highest scores in some viewing conditions but were rated relatively low in others. For example, display X was highly rated in the high viewing positions but received lower ratings in the low viewing positions. Although both of the transmissive displays were deemed acceptable, the transmissive display with superbright LED backlight (W) was the recommended choice. The fact that the chosen display received higher ratings than the benchmark (previous-generation infusion pump) display (which was presumed to be acceptable) led to the conclusion that the chosen display would be acceptable to future users.
REFERENCES American College of Radiology. (1999). ACR Standard for Teleradiology. Reston, VA: American College of Radiology. American National Standards Institute/Association for the Advancement of Medical Instrumentation (ANSI/AAMI). (1993). Human Factors Engineering Guidelines and Preferred Practices for the Design of Medical Devices. ANSI/AAMI HE-48-1993. Arlington, VA: Association for the Advancement of Medical Instrumentation. Arnaut, L. Y., and Greenstein, J. S. (1988). Human factors considerations in the design and selection of computer input devices. In S. Sherr (Ed.), Input Devices (pp. 71–121). San Diego, CA: Academic Press. Bedell, D. (2004). Bend me, shape me. Dallas Morning News, April 29. Cushman, W. H. and Rosenburg, D. J. (1991). Human Factors in Product Design. New York: Elsevier. Dell Computer Inc. (2003). High Resolution, Wide-Aspect, and Wide-Viewing Displays on Dell Portable Computers. Available: http://www.dell.com/r and d. Department of Defense (1999). MIL-STD 1472F: Design Criteria Standard: Human Engineering. Washington, DC: United States Department of Defense. Hanna, G. B., Shimi, S. M., and Cuschieri, A. (1998). Task performance in endoscopic surgery is influenced by location of the image display. Annals of Surgery, 227(4), 481–484. Hay, N. J. (1999). Reading and Typography. Available: http://hubel.sfasu.edu/courseinfo/SL99/ typography.html. Human Factors and Ergonomics Society [HFES] (1998). Human Factors Engineering of Computer Workstations. Santa Monica: Human Factors and Ergonomics Society. Human Factors and Ergonomics Society [HFES] (2002). BSR/HFES100: Human Factors Engineering of Computer Workstations. Santa Monica, CA: Human Factors and Ergonomics Society.
TAF-ER562X_WINGR-09-0401-C008.indd 348
11/8/10 11:44:21 PM
Visual Displays
349
International Organization for Standardization. (1992). Ergonomic Requirements for Office Work with Visual Display Terminals—Part 3: Visual Display Requirements. ISO 9241-3:1992(E). Geneva, Switzerland: International Organization for Standardization. Israelski, E. (2004). Personal communication. Maddox, M. E. (1979). Two-dimensional spatial frequency content and confusions among computer-generated dot-matrix characters. Proceedings of the 23rd Annual Meeting of the Human Factors and Ergonomics Society, 384–388. Santa Monica, CA: Human Factors and Ergonomics Society. Muto, W. H. (2001). Defining Human Factors Requirements When No One Seems to Know. Paper presented at meeting of the Association for the Advancement of Medical Instrumentation, 9–13 June, Baltimore, MD. Sanders, M. S. and McCormick, E. J. (1993). Human Factors in Engineering and Design. New York: McGraw-Hill. Sherr, S. (1979). Electronic Displays. New York: John Wiley & Sons. Snyder, H. L. (1980). Human Visual Performance and Flat Panel Display Image Quality. Blacksburg: Virginia Polytechnic Institute and State University. Thorell, L. G. and Smith, W. J. (1990). Using Computer Color Effectively. Englewood Cliffs, NJ: Prentice Hall. Wheeler, T. R. H. and Clark, M. G. (1992). CRT technology. In H. Widdel and D. L. Post (Eds.), Color in Electronic Displays (pp. 221–256). New York: Plenum. Wilson, R. F. (2001). HTML E-mail: Text font readability study. Web Marketing Today. Available: http://www.wilsonweb.com/wmt6/html-email-fonts.htm.
TAF-ER562X_WINGR-09-0401-C008.indd 349
11/8/10 11:44:21 PM
TAF-ER562X_WINGR-09-0401-C008.indd 350
11/8/10 11:44:21 PM
9 Connections and Connectors Joseph F. Dyro, PhD, CCE, FAIMBE, FACCE CONTENTS 9.1 Purposes of Connections and Connectors ...............................................................354 9.1.1 Electricity and Light .....................................................................................354 9.1.2 Gases and Vacuum ........................................................................................354 9.1.3 Fluids ............................................................................................................354 9.2 Environments of Use ...............................................................................................355 9.2.1 The Complex Hospital Environment ............................................................355 9.2.2 Medical Devices with Multiple Connectors..................................................356 9.2.3 Other Medical Device Environments ...........................................................359 9.2.4 Connectors in Environments outside the Hospital ........................................359 9.2.5 Internal Connectors Not Visible to the User .................................................360 9.3 Lessons Learned from Bad Connections .................................................................360 9.3.1 Infant Radiant Warmer Temperature Sensor Misconnection .......................361 9.3.2 Reversed Gas Flow through Anesthesia Vaporizer Causes Increased Anesthetic Concentration..............................................................................361 9.3.3 ECG Electrode Leads Plugged into AC Line Cord.......................................362 9.3.4 Mismatched Electrosurgical Unit Active Electrode and Hand Piece ...........363 9.3.5 CPAP Valve Connected Backward Prevented Exhalation ............................364 9.3.6 Connecting Enteral Feeding Sets to Vascular Access Tubing ......................365 9.3.7 Luer Misconnections ....................................................................................365 9.3.8 Risk Analysis ................................................................................................366 9.4 General Principles ...................................................................................................367 9.4.1 Protect against Use Error ..............................................................................367 9.4.1.1 Misconnections ..............................................................................367 9.4.1.2 Inadequate Connections .................................................................368 9.4.2 Protect Patient and User from Hazard ..........................................................368 9.4.3 Accommodate User’s Physical Characteristics .............................................368 9.4.4 Design for Reduced Complexity, Ease of Use, User Satisfaction, and Comfort .........................................................................................................369 9.4.4.1 Reduced Complexity ......................................................................369 9.4.4.2 Ease of Use ....................................................................................369 9.4.4.3 User Satisfaction and Handling Comfort .......................................369 9.4.5 Design for Physical Accessibility..................................................................370 9.5 Special Considerations.............................................................................................370
351
TAF-ER562X_WINGR-09-0401-C009.indd 351
11/8/10 11:47:44 PM
352
Handbook of Human Factors in Medical Device Design
9.5.1 Connector Uses .............................................................................................370 9.5.1.1 Critical versus Noncritical Functions .............................................371 9.5.1.2 Methods of Improving Connection Reliability ..............................371 9.5.2 Users of Connectors ......................................................................................371 9.5.2.1 Ease of Use, Intuitiveness, and User Satisfaction ..........................372 9.5.2.2 Smooth Operation ..........................................................................373 9.5.2.3 Dexterity Requirements .................................................................373 9.5.2.4 Users’ Physical Attributes ..............................................................373 9.5.2.5 Users’ Physical Characteristics ......................................................373 9.5.2.6 User Skills, Abilities, Training, and Experience............................374 9.5.2.7 Urgent Use .....................................................................................374 9.5.2.8 The User as Assembler of Connectors ...........................................374 9.5.3 Connections to Patients .................................................................................375 9.5.4 Connections to Facilities ...............................................................................375 9.5.5 Connector Use Environments .......................................................................376 9.5.6 Maintainers of Connectors ...........................................................................377 9.6 Design Guidelines....................................................................................................378 9.6.1 Operation ......................................................................................................378 9.6.1.1 Keying ............................................................................................379 9.6.1.2 Interlocks .......................................................................................380 9.6.1.3 Unique Configurations ...................................................................380 9.6.1.4 Color Coding..................................................................................381 9.6.1.5 Shapes ............................................................................................381 9.6.1.6 Labels .............................................................................................381 9.6.1.7 Connection Status Indicators: Tactile, Auditory, and Visual Cues ...............................................................................................383 9.6.1.8 Alignment Marks ...........................................................................384 9.6.1.9 Prevention of Disconnection ..........................................................385 9.6.1.10 Standards .......................................................................................386 9.6.2 Physical Interaction .......................................................................................387 9.6.2.1 Physical Interaction with the Patient ..............................................387 9.6.2.2 Protection from Connector Damage and Contamination...............388 9.6.3 Conveniences ................................................................................................389 9.7 Case Studies .............................................................................................................390 9.7.1 Electrical Connectors: ECG Cable ...............................................................390 9.7.2 Liquid Connector: Anesthesia System Vaporizer Filler................................391 References ........................................................................................................................393
Connections and connectors are found on many medical devices used in a host of settings that range from intensive care units and operating rooms to the home and office. Most of the time, connectors serve their purpose well, but too frequently poorly designed and manufactured connectors, in combination with human error, have caused death and serious injury, including fatal air embolism (Institute for Safe Medication Practices [ISMP], 2004), electrocution (Katcher, Shapiro, and Guist, 1986), and asphyxia (Dubinsky, 1992). This chapter addresses connections and connectors, their role in several adverse events,
TAF-ER562X_WINGR-09-0401-C009.indd 352
11/8/10 11:47:44 PM
Connections and Connectors
353
their applications in patient care, and their human factors design principles. The chapter concludes with case studies illustrating the application of human factors principles in connector design. A connection is defined as a means for connecting a device to another device, to a patient, or to external resources (e.g., electrical power). For example, energizing an electrically powered medical device is done by connecting the device to the facility’s power distribution system. A connector (i.e., the device’s line cord plug) attached to a mating connector (e.g., the facility’s wall receptacle) establishes the connection. Thus, a connector is a device or component that enables the connection of other devices or components to one another. Figure 9.1 shows some of the multitude of connectors found in the patient care environment. This chapter focuses mainly on temporary, as opposed to permanent, connections, for example, a patient breathing circuit to an endotracheal tube or an infusion line connected to an intravenous (IV) catheter. However, many of the design guidelines applicable to temporary connections also apply to permanent connections, those connected throughout the life of the medical device (e.g., a fastener in a patient lift that connects a patient support to the lift frame). This chapter is limited to those connectors that require mechanical contact, by far the majority of connectors and connections in patient care. While wireless connections utilizing such protocols as IEEE 802.11 and Bluetooth are becoming increasingly important for the transfer of signals, such nonmechanical connections for wireless communication among medical devices are beyond the scope of this chapter. Because the highly technological specialty of anesthesiology has been on the forefront of designing medical devices to enhance safety, many of the examples in this chapter come from the operating room environment. The principles, however, apply broadly to the design of connectors for all medical devices. 14
1
2 3
13
4
12
5
11 6
10 9
8
7
FIGURE 9.1 Typical connectors on medical devices in the patient environment: (1) threaded collar on a sigmoidoscope fiber-optic cable, (2) female Luer connector on a radial artery catheter, (3) male Luer lock on oxygen tubing, (4) threaded connectors on a noninvasive blood pressure monitor, (5) parallel blade plug on a hospital grade power cord, (6) Chemetron-style male oxygen connector on a gas hose, (7) phone plug on an airway temperature sensor, (8) barbed connector on tubing for liquid transfer, (9) molded female connector on tubing, (10) multipin electrical connector on a pulse oximeter sensor, (11) multipin electrical connector on an ECG cable, (12) Y-piece on a neonatal endotracheal tube, (13) ECG electrode connected to cable clip on ECG lead, and (14) piercing pin to intravenous solution bag.
TAF-ER562X_WINGR-09-0401-C009.indd 353
11/8/10 11:47:44 PM
354
Handbook of Human Factors in Medical Device Design
9.1 PURPOSES OF CONNECTIONS AND CONNECTORS In general, connections require that the user exert forces to bring two connectors securely together (e.g., pushing, pulling, or rotating). Security of connections can be attained through friction, for example, with tapered or barbed connectors or with mechanisms such as latches, detents, and screw collars. A mechanical connection, in addition to constraining elements in relation to each other, can also enable the transfer of energy and matter (e.g., light, electricity, gas, and liquids). For example, a pin on an ECG lead, when inserted in the socket at the distal end of an ECG cable (another connector), forms a connection that enables the transfer of the millivolt-level electrical signal on a patient’s chest to the ECG monitor, to which the ECG cable is in turn connected, which interprets that signal. Similarly, a Y-piece in a patient breathing circuit enables the connection of an endotracheal tube to the rest of the circuit and hence to a ventilator or anesthesia machine, permitting gases to flow to and from the patient. A Luer connection on an IV administration set allows IV fluid to flow through tubing and catheters into the patient. Connectors can fasten one component or device to another merely mechanically, constraining the position of one part in relation to another part (e.g., a clamp on an infusion pump securing the pump to an IV pole or a threaded screw connecting a bed rail to a bed). Purely mechanical connectors (e.g., pins, bolts, screws, plates, beams, and springs), while not the focus of this chapter, do benefit from many of the human factors design principles discussed. Connections and connectors fall into four main classifications, which are discussed below: 1. Electricity and light 2. Gases and vacuum 3. Fluids 4. Mechanical constraint only
9.1.1 ELECTRICITY AND LIGHT Connectors enable transmission of electrical power (e.g., from batteries, AC supply, or high-frequency generators) and signals (e.g., from physiological monitors, alarms, and control systems). Light for illumination or data transfer uses fiber-optic connections. Both electrical and fiber-optic connectors mechanically join components such as wires, printed circuit boards, and fiber-optic bundles.
9.1.2 GASES AND VACUUM Connectors are required for the transfer of air, vacuum, and medical gases commonly found in hospitals (e.g., oxygen and nitrous oxide). Medical devices are typically connected to a central source (e.g., wall or overhead mounted connectors) or individual gas cylinders.
9.1.3 FLUIDS Connectors enable the transfer of liquids (e.g., blood, intravenous solutions, and water) in association with a wide range of medical devices, for example, infusion pumps, urinary drainage systems, blood filters, and hemodialysis units.
TAF-ER562X_WINGR-09-0401-C009.indd 354
11/8/10 11:47:46 PM
Connections and Connectors
355
9.2 ENVIRONMENTS OF USE Connections and connectors are found in medical devices used in the hospital, ambulatory care and day surgery centers, diagnostic clinics, and extended-care facilities. Increasingly, they are found in the home, ambulance, and public places. Several examples are presented next of connector use in varied settings demonstrating how variations in the environment and the user’s knowledge, skill, and experience present unique human factors design challenges. See also Chapter 3, “Environment of Use,” and Chapter 17, “Mobile Medical Devices.”
9.2.1 THE COMPLEX HOSPITAL ENVIRONMENT Connectors and connections are used throughout the hospital environment typically as components of systems comprised of the patient, medical devices, device accessories, users, and facilities (Shepherd, 2004). The interrelation of these system components must be understood when applying human factors principles to connector design. For example, typical operating rooms contain a high density of these system components, especially when procedures such as coronary artery bypass are done. Figure 9.2 illustrates the complexity of an operating room. The patient is connected to electrodes, breathing circuit, and blood lines. Devices include physiological monitors, anesthesia workstation, heart-lung machine, and electrosurgical unit (ESU). All devices have numerous external and internal mechanical and electromechanical connections. Users include the surgeon, anesthesiologist, perfusionist, and nurse. Numerous device power cords, hoses, and tubing are connected to the Facility
Devices - Patient - Operators
A F E
Environment
D
C
B 2004
FIGURE 9.2 A complex operating room environment during a coronary bypass operation showing the interconnection of devices, patient, operators, and facilities. Devices pictured are (A) anesthesia system, (B) heart-lung machine, (C) sequential compression device, (D) electrosurgical unit, (E) electrosurgical hand piece, and (F) infusion pump.
TAF-ER562X_WINGR-09-0401-C009.indd 355
11/8/10 11:47:46 PM
356
Handbook of Human Factors in Medical Device Design
wall sources of electricity, gases, and water (facility). The complex environment of patient, personnel, multiple devices, and their associated leads, cables, fluid lines, and gas hoses presents a challenge for users, as it affects such aspects as visibility of and access to connections and connectors. Other hospital areas with similar degrees of system complexity include intensive care units, interventional radiology, cardiac catheterization, and special procedures rooms. Thus, multiple connectors connect various devices operated by many persons, often in a stressful environment. In the constrained space of a hospital room, for example, the close proximity of devices with their associated power cords, pneumatic lines, monitoring cables, and fluid lines adds to the complexity. Doctors, nurses, and technicians, each with their own specialized training, work with and among these devices. Emergent needs requiring haste, other devices obscuring vision, confusing labeling, and noise interfering with auditory cues are among many human factors issues that designers must consider. The chances for misconnections, inadequate connections, and broken connections are increased in these complex use environments. Figure 9.2 illustrates the opportunity for a misconnection. The perfusionist in the foreground has noticed a line coming from the patient that is not connected to anything else. If he did not identify the purpose of the line, he could connect it to any of a number of sources of pressurized gas. Sources (e.g., sequential compression devices, noninvasive blood pressure [NIBP] monitors, or oxygen lines) can have connectors that are compatible with those found on lines inserted into the patient (e.g., a central venous catheter). Connecting the catheter to any of these devices would inject air into the patient’s blood vessels, which could cause a fatal air embolism (BS and S, 2003; ECRI, 2004). Such misconnections have occurred because of several factors, including inadequate consideration of human factors in connector design (ISMP, 2004). While the environment depicted in Figure 9.2 appears complex, the actual assemblage of devices, accessories, and connections is even more so (Figure 9.3). During cardiac surgery (e.g., coronary artery bypass graft procedures), the patient’s blood circulation is routed through a heart-lung machine (Figure 9.3), the patient is rendered unconscious by the use of an anesthesia workstation (Figure 9.4), while cutting and coagulation of body tissue is accomplished with an electrosurgical unit. Connectors are associated with each of these devices as well as with other surgical instruments, physiological monitors, infusion pumps, sequential compression units, body temperature controllers, the surgical table, and surgical lamps.
9.2.2 MEDICAL DEVICES WITH MULTIPLE CONNECTORS Many medical devices are composed of individual components, parts, and other devices connected together into an integrated system. For example, the anesthesia workstation (Figure 9.4) is connected to a wall-mounted gas console for nitrous oxide and oxygen and to an electrical outlet to power the various alarms, monitors, and control systems. An anesthesiologist operates the system, which is comprised of pneumatics, vaporizers, a breathing circuit, and monitoring systems. The pneumatic system has many internal connectors within the oxygen, air, and nitrous oxide circuitry. Inlet connections are incorporated for accepting cylinder and facility gas distribution system sources. The vaporizers have internal connectors and connectors to allow the interchange of different vaporizers, each with a specific liquid anesthetic agent. A mechanical connector system prevents the delivery of more than one liquid anesthetic agent at the same time.
TAF-ER562X_WINGR-09-0401-C009.indd 356
11/8/10 11:47:46 PM
Connections and Connectors
357
FIGURE 9.3 Perfusionist (foreground) operating a heart-lung machine during cardiopulmonary bypass surgery in the operating room.
1
2
FIGURE 9.4 circuit.
Narkomed 6400 Anesthesia System showing (1) two vaporizers and (2) breathing
TAF-ER562X_WINGR-09-0401-C009.indd 357
11/8/10 11:47:46 PM
358
Handbook of Human Factors in Medical Device Design
4
3
2
1
FIGURE 9.5 Electrosurgical hand piece (1), active electrode (2), cable (3), and connector (4) to connect to an electrosurgical unit.
Connectors also prevent the reverse flow of fresh gas through a vaporizer and enable each vaporizer to be filled with only the liquid anesthetics for which it was designed. The breathing circuit generally is comprised of several interconnected components—the carbon dioxide absorber, ventilator, and patient breathing circuit—each of which has multiple connectors. The patient breathing circuit typically connects to an endotracheal tube connected to the patient’s airway. The breathing circuit is an interconnection of lengths of flexible hose with connecting ports to enable the monitoring of pressures, flows, and gas concentrations. Breathing circuit filters and heat and moisture exchangers are also incorporated in the circuit. The surgeon uses an ESU to cut and coagulate tissue. The surgeon holds a hand piece and activates the ESU by either a foot switch or a switch on the hand piece. Figure 9.5 shows one of many types of hand pieces used to deliver ESU energy to the patient. In minimally invasive surgery, endoscopic devices such as shears are introduced into the patient through cannulas. The hand piece is connected to a cable that in turn is connected to the ESU. Energy flows bidirectionally through cables connected to the console front panel. The return cable is connected to an ESU patient return electrode or grounding pad (Figure 9.6) that is connected to a patient’s skin.
1 2 3
(a)
(b)
FIGURE 9.6 Disposable grounding pad (i.e., patient return electrode) is connected to the patient’s leg (a). The electrode (1) is connected to the connector (2) at the end of the cable (3), which returns current to the ESU. Electrosurgical grounding pad (b) is partially pulled away from its backing showing two electrodes (arrows) for sensing adequacy of contact.
TAF-ER562X_WINGR-09-0401-C009.indd 358
11/8/10 11:47:47 PM
Connections and Connectors
359
9.2.3 OTHER MEDICAL DEVICE ENVIRONMENTS Intensive care units, special procedure rooms, and emergency rooms, like the operating room, are areas of the hospital where connections and connectors are an integral part of a complex technological environment. Devices such as patient ventilators have many internal and external connectors and are, in turn, connected to other ancillary devices such as breathing circuits. Figure 9.7 shows a typical neonatal breathing circuit connected to a humidifier. The humidifier adds moisture to the dry air from the ventilator before it travels to the patient. There are 20 connections and connectors shown in the figure: connection to the patient’s endotracheal tube (10), breathing circuit heating elements (5, 14, 19), endotracheal tube connector (11, 12), pressure monitoring tubing (13), air temperature sensors (2, 15, 16), ventilator (6, 20), patient circuit tubing (1, 17, 18), humidifier (3, 4), water supply (9), water reservoir (8), and electrical power supply (7). Other hospital locations in which connections and connectors abound include medicalsurgical units, telemetry units, clinical laboratories, labor and delivery, and radiology. Connectors are used with the thousands of types of medical devices utilized in these locations, such as electric and hydraulic beds, telemetry units, infusion pumps, and x-ray machines.
9.2.4 CONNECTORS IN ENVIRONMENTS OUTSIDE THE HOSPITAL Other settings, particularly nonhospital locations such as public places, emergency vehicles, and the home, while arguably less complex, nevertheless present design challenges (see Chapter 18, “Home Health Care”). For example, home care ventilators, while simpler in design, must work reliably in the home where the level of education, skill, and training of the caregiver is typically below that of hospital personnel (Qualls, Harris, and
1 2
3
4 5
7
10
6
9
8
12 13
14 15 16 17
18
11 19 20
FIGURE 9.7
Humidifier and neonatal patient breathing circuit showing 20 connection points.
TAF-ER562X_WINGR-09-0401-C009.indd 359
11/8/10 11:47:49 PM
360
Handbook of Human Factors in Medical Device Design
Rogers, 2001). The use environments are often less controlled and predictable (Dyro, 1998). Consumer products such as extension cords and power strips may contain poorly designed or maintained connectors that result in erratic device operation because of intermittent power. Electrical appliances such as fans, phones, and fluorescent lights emit electromagnetic interference that could adversely affect device operation. Children or other untrained individuals may make dangerous connections. Pets could chew or pull on device cables or connectors.
9.2.5 INTERNAL CONNECTORS NOT VISIBLE TO THE USER Internal connectors in medical devices, facilities, and patients are generally not visible to most device users, including clinicians, patients, visitors, and ancillary personnel. However, they are critical device components for the manufacturer, the installer, and the maintainer, and their design must receive the same consideration as external cables. Inadequate attention paid to human factors design principles of connections and connectors, wherever they are located, can adversely affect patient safety and quality of care. For example, the author found, in the course of a forensic investigation, that a medical device technician upgrading an anesthesia workstation incorrectly installed piping. The incorrect installation caused reverse gas flow through a vaporizer, delivering excessive anesthetic agent to the patient and leading to permanent brain damage. The piping design did not adequately support the technician’s tasks. Two pipes were similarly shaped and had no labels or identification. Those pipes’ connections had no unique markings or keying and could be made to fit onto either the outlet or the inlet of a vaporizer manifold. The technician mistook one pipe for the other, resulting in terrible consequences. Medical devices are connected to and depend on properly functioning utilities for safe and effective operation. These connections are typically out of sight, within the walls and ceilings of the hospital. Again, inadequate attention to human factors design has resulted in patient injury and death. For example, plumbers have inadvertently crossed nitrous oxide supply lines with oxygen lines because of inadequately labeled pipe connections, causing hypoxic brain injuries (Sato, 1991). Implanted medical devices, such as a dorsal column stimulator for pain control, have internal connections within the patient’s body. In this case, during the implantation procedure, a surgeon connects electrode leads to the patient and to the stimulator. These connections are more difficult to accomplish because of their small size, a wet and slippery environment (i.e., living tissue), and the wearing of gloves that can reduce the surgeon’s dexterity.
9.3 LESSONS LEARNED FROM BAD CONNECTIONS Failure to adhere to good human factors principles when designing connectors and connections can result in patient death or serious injury. Several examples are presented next from the literature and from the author’s experience investigating medical device–related events. The examples illustrate that connectors used in conjunction with a wide variety of medical devices can cause patient harm, such as burns from stray electrosurgical energy and death from hyperthermia, anesthetic overdose, electrocution, barotrauma, ventilation failure, and air embolism. In all the examples, the application of human factors considerations to connector design could have prevented the deaths and serious injuries.
TAF-ER562X_WINGR-09-0401-C009.indd 360
11/8/10 11:47:50 PM
Connections and Connectors
361
9.3.1 INFANT RADIANT WARMER TEMPERATURE SENSOR MISCONNECTION A hospital had two different infant radiant warmers, A and B, manufactured by two different companies. Both warmers used infant temperature sensors as part of a closed-loop system to regulate warmer radiant heat output. A nurse connected a temperature sensor to the infant with adhesive tape and connected the lead from the sensor to the warmer’s control panel (Dyro and Shepherd, 2005). The thermistors in sensors A and B had different resistance versus temperature characteristics. Nevertheless, both sensors’ connectors (phone plugs) were interchangeable; either connector could be connected to either warmer. While attempting to maintain a neonate’s body temperature, sensor A was used with warmer B. This caused warmer B to sense temperatures below the infant’s actual temperatures. Consequently, warmer B delivered more radiant heat than was necessary, causing hyperthermia and brain damage. The two warmers relied only on a secure mechanical connection between the sensor and the control panel connectors. The connectors lacked labels, coding, keying, locks or alarms, or warning labels to instruct users on proper interconnection. The lesson learned from this event leads to the first specific design guidance in this chapter. Guideline 9.1: Avoid Connector Similarity Designers should protect against the possibility of other similar connectors being incorrectly connected to the device. Simply relying on the size of connectors or tightness of fit will not reliably prevent misconnection. Medical device users may force a fit or may secure a loose fit with a shim or tape. Figure 9.8 shows how an incorrect and loose-fitting infant temperature sensor phone plug was connected to an infant radiant warmer control panel phone jack and secured by white adhesive tape (Dyro, 2004).
9.3.2 REVERSED GAS FLOW THROUGH ANESTHESIA VAPORIZER CAUSES INCREASED ANESTHETIC CONCENTRATION An anesthetist connected a stand-alone vaporizer to an anesthesia workstation by way of a flexible hose. The vaporizer was designed to add a specific percentage of anesthetic agent to the gas coming from the anesthesia machine as the gas passed through the vaporizer
FIGURE 9.8 The wrong temperature sensor plugged (arrow) into the phone jack on an infant radiant warmer control unit. Inset shows a phone plug connector. (Photograph courtesy of Marvin Shepherd.)
TAF-ER562X_WINGR-09-0401-C009.indd 361
11/8/10 11:47:50 PM
362
Handbook of Human Factors in Medical Device Design
1
2
4
3
5
FIGURE 9.9 Stand-alone anesthesia vaporizer (1) with female inlet tubing adapter (2), male outlet tubing adapter (3), rubber hose from anesthesia machine (4), and rubber hose to patient (5).
before traveling on to the patient. Proper operation of vaporizers requires that gas enter through an inlet port and exit through an outlet port. The anesthetist connected the hose from the gas outlet of the anesthesia workstation to the outlet port of the vaporizer instead of its inlet port. The resultant reverse gas flow through the vaporizer resulted in a child receiving a lethal anesthetic overdose. Inattention and poor connector design caused the adverse event. The vaporizer inlet and outlet ports were designed to connect to female and male adapters, respectively. These adapters, in turn, were connected, by way of barbed connectors, to the hoses from the anesthesia workstation and to the patient. The hose from the workstation could be and, in fact, was attached to the wrong adapter. Figure 9.9 shows a vaporizer with hoses attached to barbed connectors on adapters connected by friction fit to the vaporizer’s inlet and outlet ports. This event highlights the danger of using adapters that can be mistakenly attached to the wrong lines.
9.3.3 ECG ELECTRODE LEADS PLUGGED INTO AC LINE CORD At a neonate’s bedside, the female end of a power cord was detached from a physiological monitor while the male end of the cord remained plugged into the wall receptacle. A nurse applied electrodes to the neonate for ECG monitoring. The electrodes had leads that terminated in exposed pin contacts for connection to sockets in an ECG cable connected to the monitor. The nurse noticed the female end of the power cord, grasped it, and inserted the electrode lead pins into the sockets in the female end of the energized cord. This misconnection resulted in electrocution of the neonate from full AC line voltage applied across the chest. The female end of the power cord shown in Figure 9.10a was constructed of clear plastic enabling the internal conductors, black (hot), white (neutral), and green (ground), to be seen. The terminals of the ECG electrode leads were similarly color-coded black, white, and green. The color coding of line cord conductors matching that of the electrode lead connectors and the ability to insert an electrode lead pin contact into a power cord were factors that contributed to the death. Multiple deaths associated with inadvertent connection of ECG leads to power cords or directly to wall receptacles (Katcher et al., 1986) led to an industry-wide change in the
TAF-ER562X_WINGR-09-0401-C009.indd 362
11/8/10 11:47:50 PM
Connections and Connectors
363
(a)
(b)
FIGURE 9.10 (a) Fatal misconnection: Infant’s ECG leads with exposed contacts plugged into AC-powered line cord. (b) Electrode cable and lead redesigned with recessed lead contacts to eliminate electrocution hazard.
design of ECG leads and the cable to which the leads are typically connected (U.S. Food and Drug Administration [FDA], 1997). Figure 9.10b shows the changes made in both cable and leads to eliminate the electrocution hazard. The exposed conductive pin on the lead has been replaced by a female socket. The lead’s female socket connects to male pins recessed in the electrode cable. Many health care environments are crowded with nonstandardized devices and accompanying intermingling of cables, leads, and connectors. All these devices, connectors, and cables can confuse the user and increase the risk of misconnections. The similarity in appearance of the energized power cord and the ECG monitor’s electrode cable in Figure 9.10a confused the nurse.
9.3.4 MISMATCHED ELECTROSURGICAL UNIT ACTIVE ELECTRODE AND HAND PIECE A surgeon used an ESU to perform a tonsillectomy, a relatively minor surgery involving the removal of lymphoid tissue that lies on each side of the throat at the back of the mouth. The scrub nurse connected the conductive pin of an active electrode into the socket in an ESU hand piece (Figure 9.11a). When the surgeon activated the ESU, the electrosurgical energy traveled to the active electrode and then to the tonsils, where it cut and coagulated the desired tissue. The surgeon successfully removed the tonsils but noticed burns on the inner lining of the patient’s mouth. Forensic investigation by this author revealed that
2 mm (a)
8 mm (b)
FIGURE 9.11 (a) An active electrode and an active electrode inserted in a hand piece. The electrode and hand piece are made by different manufacturers. (b) An active electrode and an active electrode inserted in a hand piece. The electrode and hand piece are made by one manufacturer.
TAF-ER562X_WINGR-09-0401-C009.indd 363
11/8/10 11:47:50 PM
364
Handbook of Human Factors in Medical Device Design
electrosurgical energy, in addition to flowing to the surgical site from the tip of the active electrode, also flowed to the patient’s mouth by an alternate current path, resulting in the burns. This alternate path was between the patient’s mouth and the connection between the proximal end of the active electrode and the socket in the hand piece. One company had manufactured the active electrode and another had manufactured the hand piece. The shape and size of the plastic insulation on the active electrode did not allow a deep enough insertion of the metallic conductor pin into the socket of the hand piece made by the other manufacturer. The resulting distance between the pin and the patient’s mucosa was only 2 mm (see Figure 9.11a), sufficiently small to permit ESU energy to bridge the gap. Figure 9.11b shows the electrode manufactured by the manufacturer of the hand piece. Note the greater depth of insertion (i.e., 8 mm) of the conducting pin. Guideline 9.2: Competitive Product Awareness Designers should be aware of all competitive products that could be connected to their devices and the likelihood of such connections and design safeguards against the use of those products that could pose hazards.
9.3.5 CPAP VALVE CONNECTED BACKWARD PREVENTED EXHALATION An adult patient was being mechanically ventilated. The respiratory therapist connected the outlet port of a continuous positive airway pressure (CPAP) valve to the exhalation limb of the patient breathing circuit. The patient’s exhalation was thus blocked by the incorrectly connected one-way CPAP valve. The patient was unable to exhale, pressure in the airway increased beyond physiological limits, and the patient suffered barotrauma and subsequently died. The design of the valve enabled the outlet port to be connected to the exhalation limb of the breathing circuit (see Figure 9.12). Despite a label on the valve with an arrow meant
1 3 1 2
2 4
FIGURE 9.12 Continuous positive pressure airway (CPAP) device (1) and adapter (2). The adapter (2) can be connected to 22-mm corrugated breathing tubing of the patient’s exhalation circuit and then connected properly to the inlet of the CPAP device (3). The adapter (2) can also be connected to 30-mm corrugated breathing tubing and then incorrectly connected to the outlet of the CPAP device (4). The inset shows the CPAP device with outlet limb redesigned to incorporate (1) a flange on the exterior and interior of the cylindrical wall preventing tapered friction-fit adapter connection and (2) a slot in the wall to allow air to escape in the event of an occlusion.
TAF-ER562X_WINGR-09-0401-C009.indd 364
11/8/10 11:47:51 PM
Connections and Connectors
365
to show direction of flow and the word “outlet,” the user construed the arrow as pointing to where the outlet, or exhalation port, of the patient breathing circuit should be attached. Designers changed the valve (see Figure 9.12, inset) by altering the shape of the outlet port and incorporating a flow bypass in the event of reverse connections. As in the case of the reversed anesthesia vaporizer connections discussed previously, this event also illustrates the dangers of adapters, as explained in Figure 9.12.
9.3.6 CONNECTING ENTERAL FEEDING SETS TO VASCULAR ACCESS TUBING Enteral feeding delivers nutrient liquids directly into the alimentary tract of patients who cannot eat solid food. An enteral feeding set is connected to a specialized tube inserted at some location in the patient’s alimentary tract (e.g., stomach, duodenum, or jejunum). Feeding is administered by an enteral pump or by gravity. The connection of enteral feeding sets to devices other than the specialized feeding tubes has seriously injured patients. Erroneous connections have been made to parenteral (i.e., introduced into the body other than by way of the gastrointestinal tract) administration sets, indwelling intravenous catheters or ports, and epidural catheters (Cyna et al., 2002). Such misconnections were possible because of the similarity in design of enteral access devices and vascular access devices since both employed rigid Luer connections (e.g., the Luer connection on the radial artery catheter shown in Figure 9.1). Even enteral feeding set connectors incompatible with rigid Luer connections have nonetheless been connected to vascular access devices by the use of adapters. Luer connectors are standard on many intravenous tubing systems (American National Standards Institute [ANSI]/HIMA, 1983). The multiple occurrences and serious consequences of these misconnections between the incompatible enteral and parenteral systems provided the rationale for the development of a standard enteral connector design to eliminate these occurrences (ANSI/Association for the Advancement of Medical Instrumentation [AAMI], 1996).
9.3.7 LUER MISCONNECTIONS Luer connectors are widespread throughout the hospital on many different devices that should never be interconnected, for example, intrathecal or epidural devices (for administration of fluids near the brain and spinal cord) connected to IV lines, IV administration sets to intrathecal or epidural lines, and IV connection to tracheostomy cuff inflation ports (ISMP, 2001). Mismatch of pressurized gas lines terminating in Luer locks has caused fatal air embolism when connected to catheters with female Luers (ISMP, 2004). Pressurized gas lines are associated with devices such as noninvasive blood pressure devices, sequential compression devices, and insufflators. Further discussion of Luer connectors and noninvasive blood pressure monitors is found below in Section 9.6.1.6. Figure 9.13 illustrates how deadly misconnection of pressurized gases to vascular access lines can occur. All three connectors (1, 2, and 3) on the left of Figure 9.13 can connect with the one on the right (4), a needleless connector, such as is typically connected to a catheter dwelling within the patient’s vascular system. Needleless connectors are female Luer connectors with a spring-loaded seal that is forced open when a male Luer is connected to it. Connector 1 is a Luer lock connected to tubing carrying oxygen under pressure. Connector
TAF-ER562X_WINGR-09-0401-C009.indd 365
11/8/10 11:47:51 PM
366
Handbook of Human Factors in Medical Device Design A
B
1
4 2
3
FIGURE 9.13 All three connectors (1, 2, and 3) can be connected to the needleless connector (4). The male Luer part (A) depresses the spring-loaded central seal (B), allowing fluid or air to enter. Connection of 1 or 3 could cause serious injury.
2, also a Luer lock, is connected to a line carrying a medicated fluid solution. Connector 3 is a molded elastomer female connector attached to an oxygen source. While all three can be connected, only connector 2 would be a safe connection. Connecting (1) or (3) would allow air to enter the needleless connector. The pressurized oxygen in connector (3) could be sufficient to open the seal (B) of the needleless connector, likely resulting in the lethal insufflation of oxygen into the patient’s vascular system. Guideline 9.3: Pressurized Air Sources and Vascular Lines Connectors on sources of pressurized gases such as noninvasive blood pressure monitors, sequential compression units, and gas flow lines should not be compatible with or able to be forced onto connections to intravascular administration lines including Luer connections and needleless connectors.
Guideline 9.4: Luer Connections for Intravascular Connections Luer connections should be reserved for use with intravascular administration devices (e.g., arterial catheters and IV administration sets) whenever feasible.
Guideline 9.5: Intrathecal and Epidural Connections Intrathecal administration sets and intrathecal access lines and their connectors should be color-coded and well labeled. IV administration sets have been mistakenly connected to intrathecal delivery lines, leading to patient injury.
9.3.8 RISK ANALYSIS The design of a life support device, such as a ventilator, whose failure could cause death or serious injury to the patient, must include risk analysis methodologies (e.g., failure mode and effect analysis or fault-tree analysis) as specified by FDA regulations, ISO 14972, and IEC/ISO 62366. The design of even noncritical devices such as wheelchairs or beds, generally not considered to pose as great a risk to patients, should also incorporate formal risk analysis during design. Many patient injuries have occurred during use of these “lowrisk” devices. A device standing alone generally presents less risk than that same device in a complex environment containing other devices to which it might be inadvertently
TAF-ER562X_WINGR-09-0401-C009.indd 366
11/8/10 11:47:51 PM
Connections and Connectors
367
connected. Thus, risk analysis must include consideration of both internal and external connections.
9.4 GENERAL PRINCIPLES The key overarching principles of connector design are as follows: protect against use error; protect the patient and the user from hazard; accommodate the user’s physical characteristics; design for reduced complexity, ease of use, and user satisfaction and comfort; and design for physical accessibility. These general principles are discussed in more detail here along with accompanying specific design guidance.
9.4.1 PROTECT AGAINST USE ERROR Misconnections and inadequate connections are use errors. Attention to human factors issues in connector design can reduce use error. Connector-related use errors should be analyzed for opportunities for design improvements to reduce future events. 9.4.1.1 Misconnections Connections should not be susceptible to hazardous misconnections, such as connecting an enteric feeding line to a central venous catheter or connecting a patient lead to an AC power source. Guideline 9.6: Unique Connectors and Connections Connectors should be able to be connected only to their intended connectors. If oxygen supply tubes are not color-coded green for oxygen and terminate in an elastic friction-fit connector, they will more likely be connected to, for example, a needleless port on an intravenous line as shown in Figure 9.13.
Guideline 9.7: Prevent Incompatible Connections A connection should not be possible between two incompatible devices. The designer of connectors must anticipate and be aware of the connector’s compatibility with other connectors on incompatible devices.
Guideline 9.8: Users Will Force Connections Designers should consider that the user, in an attempt to make connections, will attempt to force fit mismatched connectors (such as by damaging keys) and should design connectors to withstand such forces.
Guideline 9.9: Beware of Adapters The designer should be aware of the availability and use of adapters in the intended use environments. If adapters exist that could lead to a misconnection, the device designer should reconsider connector design. A less effective strategy would be to create warnings on labels, manuals, and instructions for use to advise against adaptor use.
Guideline 9.10: Prevent Application Mismatch Connectors of one application (i.e., substance being transmitted via the connection) should not be able to connect to connectors of another application (e.g., pneumatic drill requiring high-pressure nitrogen gas and suction machine requiring a vacuum source).
TAF-ER562X_WINGR-09-0401-C009.indd 367
11/8/10 11:47:51 PM
368
Handbook of Human Factors in Medical Device Design
9.4.1.2 Inadequate Connections Inadequate connections of otherwise compatible connectors can result in shock hazard from exposed electrical conductors, loss of power, or faulty information from incomplete electrical connection. Guideline 9.11: Assurance of Complete Connection Connectors should be designed so that the user is assured by sensory clues of a complete connection.
9.4.2 PROTECT PATIENT AND USER FROM HAZARD Guideline 9.12: Connector Application and Removal Connectors for nonpermanent connections should be able to be securely mated and disconnected with ease and without harm to patient or user.
Guideline 9.13: Comfort and Safety of Connections A connector should not be uncomfortable or hazardous to the person making the connection.
Guideline 9.14: Patient Comfort and Safety A connector should not be uncomfortable or injurious to the patient in intimate contact with the connector.
Guideline 9.15: Protect against Unintentional Disconnection Connectors should be protected against unintentional disconnection. Unintentional disconnection, such as patient movement causing separation of an external connector from an indwelling venous access dialysis catheter, poses the hazard of blood loss to the patient or of the transmission of infectious agents to others.
9.4.3 ACCOMMODATE USER’S PHYSICAL CHARACTERISTICS Guideline 9.16: Anthropometric Factors in Connector Design Connector design must consider the anthropometric characteristics of the user such as hand size, dexterity, range of motion, and strength (see Chapter 4, “Anthropometry and Biomechanics”). Designers should also consider the condition of the user’s hands (e.g., wet or gloved hands).
Guideline 9.17: Sensory Factors in Connector Design Connectors incorporating visual, tactile, or auditory cues should accommodate the intended user’s range of sensory abilities, including visual acuity, proprioception, touch sensitivity, and hearing.
Guideline 9.18: Minimum Connector Size Connectors should not be so small as to be difficult or impossible to handle securely. This will be particularly important for users with physical impairments such as elderly patients with arthritis who must connect a home care device. Exceptions would apply when the user’s senses are augmented by, for example, magnifiers and tools used during surgery in which connections must be made (e.g., electrodes connected to an implanted stimulator or a catheter connected to the output port of an implanted morphine pump).
TAF-ER562X_WINGR-09-0401-C009.indd 368
11/8/10 11:47:51 PM
Connections and Connectors
FIGURE 9.14
369
Miniature neonatal electrode clips.
Examples of miniature connectors (i.e., neonatal electrodes) are shown in Figure 9.14.
9.4.4 DESIGN FOR REDUCED COMPLEXITY, EASE OF USE, USER SATISFACTION, AND COMFORT 9.4.4.1 Reduced Complexity Users of medical devices operate in a complex technological environment of diverse devices and interconnections. Numerous connectors are encountered on a routine basis in the patient care environment. Reducing complexity by standardization of colors, shapes, and mechanical operation can reduce overall system complexity, thus making users’ jobs easier. For example, replacing a connector that requires screwing down to secure a connection with one that simply snaps into place and can be released with the touch of a finger reduces complexity while improving efficiency. Likewise, a clear indication of which connector goes where is especially welcome when the need to make a connection is urgent. Mating connectors blindly with, for example, polarized plugs reduces complexity, as it makes connections possible in low-light conditions or in places where visibility is limited. Moreover, such forcing functions (i.e., connector can go in only one place in one way) significantly reduce use errors. Guideline 9.19: Minimize Connector Manipulations The manipulations required to apply a connector should be minimized. While physical constraints are most effective, clear markings to indicate how the connector should be aligned with its mate will also improve performance by shortening the time necessary to locate and make the connection. Color coding and visual style can also be valuable design features.
9.4.4.2 Ease of Use Guideline 9.20: Ease of Connection Connections should not require excessive, unusual, or sustained forces applied by the user. One-handed connection with minimal hand and finger manipulation is a desired goal.
9.4.4.3 User Satisfaction and Handling Comfort Connector design should enhance user interactions by improving task performance and boosting user satisfaction. A sensory cue (such as the sound of a click or the feel of a retainer snapping into place) that indicates a connection has been securely made provides the user with feedback and a positive sense of confidence.
TAF-ER562X_WINGR-09-0401-C009.indd 369
11/8/10 11:47:51 PM
370
Handbook of Human Factors in Medical Device Design
The appeal of a connector can be improved by refinements in the places, termed touchpoints, where users come into physical contact with the connector. Touchpoints encompass virtually the entire connector, including locking and tightening mechanisms, connector body, and connector coupling to its associated cable, lead, or tubing. Each touchpoint presents an opportunity to communicate a feeling of quality to the user. Communicating a quality feel will depend both on the material finishes and on fundamental engineering. Users will draw conclusions about a connector’s overall quality from the quality of the touchpoints, making them a worthy focus of design. But, more important, high-quality touchpoints can enhance user performance by increasing vigilance, reducing physical fatigue, eliminating distractions, or enabling users to detect subtle tactile cues.
9.4.5 DESIGN FOR PHYSICAL ACCESSIBILITY Guideline 9.21: Physical Access to Connectors Location and design of connectors should reflect the frequency with which they are accessed by the user. For example, if a connector is never disconnected once it has been connected, physical access to the connector is unimportant, and the connector and its mating connector or receptacle can be located in a place that cannot be as conveniently accessed. On the other hand, a connector that is frequently handled should be located conveniently for users, with due attention paid to the proximity of nonusers, such as visitors and ancillary personnel who might disturb the connection either intentionally or unintentionally. A connector located in a convenient location but requiring infrequent connection and disconnection may be designed, for example, with a locking connector versus a quick connect. A similar design would be appropriate if inadvertent disconnection could pose a safety risk.
9.5 SPECIAL CONSIDERATIONS The diversity of medical devices with regard to complexity and sophistication, from the MRI to the stethoscope, greatly affects the design of their connectors and connections. These devices see a wide variety of uses from life support to routine diagnosis. Devices are used to aid patients who are typically in a physiologically compromised state, often in a crowded environment where myriad other devices are employed on other patients by numerous health care providers working in close proximity. Home and other nonhospital environments, where medical devices are increasingly found, are often uncontrolled and present special device design challenges. Thus, designers of device connectors must understand the environment in which connectors are embedded, including its range of medical devices and their applications, its personnel, its patients, and its utilities required to support device operation (see Chapter 3, “Environment of Use” and Chapter 18, “Home Health Care”).
9.5.1 CONNECTOR USES The designer must first answer several critical questions with regard to connector use: • What is the purpose of the connection and connector? • How critical is the connection? • What substance is contained within the lines being connected (e.g., light, water, oxygen, blood, nutritional fluids, or electricity)?
TAF-ER562X_WINGR-09-0401-C009.indd 370
11/8/10 11:47:52 PM
Connections and Connectors
371
• What are the characteristics of the substance contained in or transferred through the connectors (e.g., pressure, temperature, and toxicity)? • What are the possibilities of connector mismatch? • If the connector is part of a circuit that connects to the patient’s vascular system, what other connectors incompatible with blood can be connected to it (e.g., those carrying gases under pressure as discussed previously)? • Is the connector part of a reusable device or a disposable device? 9.5.1.1 Critical versus Noncritical Functions There will be greater risk of harm if connections fail in life support devices, such as ventilators and anesthesia machines. For example, should the connection between the patient and the ventilator become detached, the patient would be deprived of oxygen, and brain damage or death would occur rapidly. The misconnection of gases (e.g., swap of nitrous oxide for oxygen) in an anesthesia gas delivery system can result in death. Understanding the consequences of failed connections enables the designer to incorporate fail-safe concepts. Many devices, such as beds and imaging systems, do not support life-critical functions. However, complying with human factors design principles is still important in noncritical device connections. The failure of a connection involving a noncritical device such as a hip prosthesis, bed, or diagnostic imaging system could also result in death or serious injury. For example, a user may tighten the mechanical connector that secures the side rail to the bed with the expectation that an adequate mechanical connection was achieved in the absence of visual and auditory cues to the contrary. Such loose connections have resulted in bed rails disengaging, with the consequence of patients falling out of bed (Dyro, 2004). 9.5.1.2 Methods of Improving Connection Reliability A connection should be secure regardless of its intended purpose. In designing for security, the frequency of the user’s need to engage and disengage the connection should be considered. For example, if the user never has to disengage a connection, bonding of the connectors with permanent adhesives, solder, or welds is acceptable. At the other extreme, if the user must engage and disengage a connection frequently and is always present to monitor the connection, a friction fit would be appropriate. For example, an anesthesiologist must be able to make connections rapidly between a patient breathing circuit Y-connector and an endotracheal tube. This is enabled by the use of tapered friction-fit conical connectors (International Organization for Standardization [ISO], 1986, 1991, 1996). For connections that do not require rapid disconnection, a secure fit can be achieved using threaded collars or barbs (see Figure 9.1).
9.5.2 USERS OF CONNECTORS Device users can vary considerably in their physical characteristics, level of skill, training, experience, and cognitive abilities. Users can range from a highly trained cardiovascular surgeon with 20 years of experience to a poorly educated, untrained, and unskilled parent trying to operate for the first time an infant ventilator at home. As described in this section, several key use attributes should be considered when designing connectors. See also Chapter 1, “General Principles”; Chapter 2, “Basic Human Abilities”; Chapter 3, “Environment of Use”; and Chapter 18, “Home Health Care.”
TAF-ER562X_WINGR-09-0401-C009.indd 371
11/8/10 11:47:52 PM
372
Handbook of Human Factors in Medical Device Design
9.5.2.1 Ease of Use, Intuitiveness, and User Satisfaction Ease of use must be balanced against other connector design requirements. For example, there may come a point where making a connector smaller for the sake of the safety and comfort of the patient will make it too difficult for the user to attach the connector. All aspects of connector use must be considered during design: connecting, locking, maintaining, confirming status, unlocking, disconnecting, and disposing. To ensure optimal ease of use, anthropometric characteristics should be considered. An example of a poorly designed connector is shown in Figure 9.15. The spring clip on the pneumatic tube attached to a blood pressure cuff is difficult to grasp and insert in the cuff connector jack, mounted on the back of the device. The spring clip is difficult to access and grasp because it is small and can swivel about the long axis of the tube. Fingertips can be pinched between the clip and the sharp edges of the aluminum connector block during attempted connection. Connector design should facilitate the users’ tasks. The user is often under time pressure and/or distracted when using medical devices. The concerns for critically ill patients, emergent need for intervention, low staff-to-patient ratios, and increased financial pressures to enhance care process efficiency combine to increase stress levels of medical device users. Guideline 9.22: Optimize Speed of Connecting Wherever possible, the designer should minimize the time and effort required to make device connections and disconnections. Users will not be satisfied with the connector shown in Figure 9.15 because it is difficult to apply. Users will be satisfied with a connector that is easy to connect and disconnect, easy to lock and unlock, keyed to prevent misconnections, coded and labeled to facilitate its identification and that of mating connectors, and that provides ample indication of proper mating and locking. Patients may be ill when using home health devices. Making device connections is rarely a high-priority task. Connector use must therefore be intuitive to ensure efficiency, safety, and user satisfaction.
Guideline 9.23: Intuitive Operation of Connectors Connectors should not require training or written instructions for their successful use.
FIGURE 9.15 Blood pressure cuff pneumatic tube connector is difficult to attach to the cuff connector receptacles rear mounted on the noninvasive blood pressure monitor chassis.
TAF-ER562X_WINGR-09-0401-C009.indd 372
11/8/10 11:47:52 PM
Connections and Connectors
373
9.5.2.2 Smooth Operation Guideline 9.24: Smooth Connecting Operation Operational modes should be smooth without false detents, catches, or variable friction that could give the false impression that a connector has reached the end of its travel and is adequately secured.
9.5.2.3 Dexterity Requirements Guideline 9.25: Accomodate Users’ Dexterity Connectors should be operable by users who have varying degrees of dexterity. Designers should consider the many factors influencing dexterity, including physical impairments, gloves, liquids, diminished hand-eye coordination (visual impairment), high workload, or distraction.
9.5.2.4 Users’ Physical Attributes Guideline 9.26: Connection Independent of Handedness Wherever possible, connection should be as readily accomplished whether the user is righthanded, left-handed, or ambidextrous.
Guideline 9.27: Connection Independent of User Position Wherever possible, connection should be readily accomplished from a range of user arm positions due to, for example, user height, sitting/standing, or physical constraints of the use environment.
9.5.2.5 Users’ Physical Characteristics Medical device users vary widely in physical size and strength as well as in dexterity, visual acuity, coordination, and tactile sensitivity (see Chapter 4, “Anthropometry and Biomechanics,” and Chapter 2, “Basic Human Abilities”). Since most connectors will be used by people across the spectrum of these physical characteristics, they should be considered in connector design. Guideline 9.28: Accomodate Users’ Physical Limitations Whenever appropriate (e.g., for home health devices), a connector should be able to be used by those who are the most physically challenged. Hand strength varies considerably. If a twisting motion is required to secure a connector, it should be able to be made even by those with a weak hand or arthritis.
Guideline 9.29: Accommodate Visual Impairment Labels and indicators, such as arrows indicating direction for turning of a connector, should be large enough to be seen even by those whose vision is impaired. Alternatively, tactile cues should be provided that facilitate correct and efficient connection without visual cues.
Guideline 9.30: Connector Use Testing Tests of connector design should be performed by users representing the expected diversity of physical characteristics. For example, tactile cues such as a brief mechanical pulse when a
TAF-ER562X_WINGR-09-0401-C009.indd 373
11/8/10 11:47:52 PM
374
Handbook of Human Factors in Medical Device Design
connector is fully engaged may prove during testing to be too subtle for most actual users and thus need to be increased in magnitude (see Chapter 6, “Testing and Evaluation”).
9.5.2.6 User Skills, Abilities, Training, and Experience Connector design should take into account varying levels of abilities, skill, training, and experience, which can differ widely among users. Even an otherwise skilled user may be called on to use a new connector after little or no training or experience. Home users may not appreciate the technology on which a particular device is based or the risks and failure modes of that device. For example, in the home setting, a layperson may place complete trust in device alarms and technology, thus failing to understand that alarms may fail to activate under some conditions (e.g., in a partial breathing circuit disconnection in a ventilator-dependent patient). Guideline 9.31: Home Use Connectors For connectors used in the home or by laypersons, the design should accommodate the wider range of intended users’ abilities, skills, training, experiences, attitudes, and cultures.
Guideline 9.32: Features to Support Connecting Designs should incorporate such features as prompts, bold labels, clear instructions, error detection, and visual confirmation that facilitate users’ abilities to safely and effectively make device connections.
9.5.2.7 Urgent Use A connection may need to be made under urgent conditions. As the time to accomplish a task decreases, the rate of error tends to increase. For example, a user may select the wrong connector and, in haste, try to make it fit and break or bend electrical contact pins and keys (Dyro and Morris, 2004). User attention diminishes when it is distracted by sudden changes in patient condition, by words and actions of coworkers, and by the activities of ancillary personnel and visitors. Nonintuitive, poorly identified, ill-fitting, or difficult-to-operate connectors are more likely to be used improperly under these conditions. Guideline 9.33: Connector Shape Integrity Designers should avoid materials that can be forcibly distorted to mate with connectors for which they are not intended.
Guideline 9.34: Strength of Materials Connector materials should have sufficient physical properties to withstand the forces typically exerted during hasty or forced connecting. For example, keys and connector housings made of some rigid plastics break more easily than those made of metal.
9.5.2.8 The User as Assembler of Connectors In some procedures, the user assembles the component parts of a connector. For example, in the placement of a Tesio catheter, used for convenient vascular access in hemodialysis, the surgeon must place a catheter in the patient’s superior vena cava and then insert a metal cannula, rigidly attached to a threaded part, into the lumen of the exposed end of a catheter.
TAF-ER562X_WINGR-09-0401-C009.indd 374
11/8/10 11:47:52 PM
Connections and Connectors
375
1
2
3
4
FIGURE 9.16 Tesio catheter connector assembly: (1) screw thread on end of adapter, (2) metal cannula, (3) silicone catheter, and (4) female plastic collar. The female Luer connector (far left) at the end of the catheter provides the access port for dialysis.
Figure 9.16 shows the parts being assembled. Once the cannula is inserted to its full length, the surgeon screws a female plastic collar into the threaded male connection at the end of the adapter completing the connection. Guideline 9.35: User-Assembled Connectors To facilitate the user’s task of assembling a connector, the designer should consider the connector materials’ mechanical and physical properties, such as tensile strength, compressibility, hardness, and flexibility. These properties may influence the ease with which a connection can be made and the security of the connection. User testing is essential for establishing an effective design of user-assembled connectors.
9.5.3 CONNECTIONS TO PATIENTS Patients vary widely in physiological characteristics, medical condition, level of consciousness, activity, and understanding. For example, a neonate’s skin is fragile and tears easily. Lightweight, easily applied miniature clip connectors on ECG leads reduce the forces on electrodes and the stress applied to the neonate’s skin (Figure 9.14). A moving patient, for example, may exert enough force to dislodge a connection. Movement is more likely with the patient at home, at work, or during recreation. Guideline 9.36: Patient Connectors Designers should consider the physical characteristics and mobility of the connected patient. Slim, low-profile rather than thick, bulky connectors enhance patient comfort and reduce the risk of disconnection from forces applied by, for example, putting on clothing or bumping an arm against a bed side rail.
9.5.4 CONNECTIONS TO FACILITIES Utilities to which devices are connected (e.g., medical gases and electrical power distribution systems) typically are connected to devices by way of receptacles located on walls or overhead columns. Networks for hospital information systems typically are hardwired throughout the building with connectors at distributed locations such as the patient’s bedside, nurse central stations, and operating rooms. Design of keyed and color-coded connectors have reduced the incidence of misconnection at the facility interface. Connections do, however, continue to be made inappropriately. For example, wall oxygen outlets have been connected to clear (non-color-coded) tubing
TAF-ER562X_WINGR-09-0401-C009.indd 375
11/8/10 11:47:52 PM
376
Handbook of Human Factors in Medical Device Design
with a friction-fit connector (also not color-coded) at the other end, allowing connection to a device not intended for oxygen use. Guideline 9.37: Facilities Connector Identification Whenever possible, both receptacles and their intended connectors should be clearly labeled or otherwise coded to minimize the risk of misconnection of facilities connectors.
9.5.5 CONNECTOR USE ENVIRONMENTS Designers must assess the environment (e.g., the hospital, emergency transport vehicles, or the home) in which a medical device and its associated connectors will be used (see Chapter 3, “Environment of Use”). A crowded environment, for example, increases the risk of unintentional disconnections and connections. Also, a high density of people working in a confined space increases the risk of unintentional forces applied to connectors from tripping on cords or bumping into connectors. This risk increases when nonmedical staff, such as housekeeping, salesmen, and visitors, are also present. Well-intentioned but otherwise uneducated visitors, in helping to reposition a patient, for example, may dislodge a connection such as the tracheostomy tube shown in Figure 9.17. The nonhospital environment, arguably less controlled than the hospital environment, poses a greater risk of uneducated people, young siblings, and even pets disturbing or damaging connections. Connections and connectors used in environments such as hemodialysis units and surgical suites may become wet with blood and other bodily fluids and thus become more difficult to manipulate. Connectors sometimes require connecting in low light conditions, such as a darkened patient room. Guideline 9.38: Connections in Wet Environments The designer should anticipate wet environments and design to enable connections to be made if connectors become wet.
Guideline 9.39: Blind Connections For connections typically made in low light situations or in locations that are hard to reach, designers should consider means of facilitating blind connection to mating connectors such as by shapes, keys, or unique pin configurations.
FIGURE 9.17 Tracheostomy tube and Velcro-backed retention strap that had become dislodged from the patient.
TAF-ER562X_WINGR-09-0401-C009.indd 376
11/8/10 11:47:52 PM
Connections and Connectors
377
An audible sound is one of many cues that can be used to indicate that a successful connection has been made. However, ambient sound levels in a quiet patient room at night contrast sharply with those found in an operating room during a hip prosthesis implantation when saws and drills are activated against the background of anesthesia machines, suction pumps, device alarms, audible monitoring signals, general conversation, and the radio. Guideline 9.40: Audible Connection Cues When incorporating audible cues to indicate integrity of a connection, designers should take into consideration the range of sound levels in the use environment.
Standardization on one manufacturer and model of a medical device often is not possible. Therefore, different manufacturers and models of a medical device with similar connectors may be used in the same location. For example, neonatal intensive care units often have several brands of infant incubators, each with its unique temperature sensor. The risk of misconnections is high in such a situation. Guideline 9.41: Multiple Similar Devices Increase Risks Different models and manufacturers of a particular medical device may be used in the same environment and connectors should be designed such that connection with different and incompatible models is impossible.
Guideline 9.42: Misconnections to Older Devices Several generations of the same type of medical device may be in use in the hospital. Designers should consider the possibility of misconnections with previous generations of devices.
Guideline 9.43: Medical and Nonmedical Devices The designer should anticipate the presence of consumer products whose associated connectors may be confused with those intended for patient care. For example, medical devices sent home from the hospital with the patient will share the same environment with nonmedical devices, such as coffeemakers and electric fans.
9.5.6 MAINTAINERS OF CONNECTORS Reusable connectors and those intended for long-term use may become dirty, corroded, or worn. This may adversely affect the performance and safety of the medical device or system with which they are associated. For example, an intermittent, grainy, or noisy image in a video endoscopy system may be caused by an old light cable with a degraded connector. Connectors require servicing more commonly than other medical device components. Moreover, service (e.g., repair, maintenance, and modification) of medical devices often entails the removal and replacement of component parts either internal or external to the device by the disconnection and connection of these components. For example, service personnel have made incorrect connections of gas tubing internal to anesthesia machines and have also mixed gases when repairing or installing central gas distribution systems.
TAF-ER562X_WINGR-09-0401-C009.indd 377
11/8/10 11:47:53 PM
378
Handbook of Human Factors in Medical Device Design
Guideline 9.44: Connector Repair and Replacement Connectors should be designed to allow safe and easy repair and replacement. Designers should anticipate that service may be performed by individuals of varying levels of training, experience, and skill.
Guideline 9.45: Selection of Connector Component Materials The designer should apply knowledge of the connector’s expected lifetime and conditions of use to recommend appropriate materials for connector components. Longtime use mandates robust design, whereas one-time use suggests disposable connectors and less durable materials. Disposable connectors, however, should not be so fragile or elastic as to be easily deformed to force fit into wrong connectors.
9.6 DESIGN GUIDELINES The following guidelines should help designers to produce ergonomically correct connections and connectors that perform safely and effectively. The guideline categories are the following: • Operation. Design features that ensure (e.g., keys, interlocks, and latches) or aid (e.g., color coding, shape, labels, cues, and alarms) proper connection, decrease the risk of misconnection, and guard against unintentional disconnection (e.g., latches, locks, and alarms) • Physical interaction. Features (e.g., application forces, size, and surface characteristics) affecting connector use, safety of users and patients, and connector protection • Ease of use and satisfaction. Features (e.g., visual cues and ergonomic design) that make a significant difference in the user’s ease of and satisfaction with connector operation and that, in turn, increase the probability of proper connections and lower the risk of misconnections
9.6.1 OPERATION Efficacy and safety of medical devices require that users make proper connections and avoid misconnections. In addition, the user must be able to make a connection that will not be unintentionally disconnected. Connectors should be designed to preclude the possibility of incorrect connection (e.g., a connector designed with a key that mates with only a matching keyway). Proper connections can be ensured using one of the following mechanical design strategies: 1. Keying 2. Interlocks (mechanical and electronic) 3. Unique configurations Other design strategies can assist the user in making proper connections but are not foolproof, for example, connectors designed with visual cues (e.g., a connector designed with a color matching the mating receptacle). Use error can still occur if the feature is overlooked or ignored. Design features to help the user make correct connections include the following: 1. Color coding 2. Shapes
TAF-ER562X_WINGR-09-0401-C009.indd 378
11/8/10 11:47:53 PM
Connections and Connectors
379 1
FIGURE 9.18 (arrow 2).
2
LEMO physiological signal connectors showing keyway (arrow 1) and mating key
3. Labels 4. Alignment marks 5. Connection status indicators Design features relating to intentional and unintentional disconnection include the following: 1. Strength of connection 2. Latches and locks 3. Alarms and indicators 9.6.1.1 Keying Keying is a mechanical arrangement (e.g., guide pins and guide sockets) that allows connectors of the same size and type to be mated without making a wrong connection. For example, Figure 9.18 shows a key (arrow 2), a parallel-sided piece on a plug that fits into a keyway (arrow 1) in a corresponding socket, allowing unique alignment of plug and socket. The Pin Index Safety System is an example of keying that prevents, for example, the misconnection of a nitrous oxide cylinder to an oxygen inlet. Each anesthetic gas inlet is identified by a unique pin configuration in which two index pins mate with a corresponding configuration of two holes in a gas cylinder head. Figure 9.19 is a diagram showing a Yoke assembly
Index pins
Gas inlet
Cylinder heads
Cylinder head Pin holes
Gas cylinder Oxygen
Nitrous oxide
FIGURE 9.19 Pin Index Safety System: gas cylinder, cylinder head, yoke assembly, index pins, and gas inlet (adapted from Cicman et al., 2000). Cylinder heads for oxygen and nitrous oxide show their different pinhole positions.
TAF-ER562X_WINGR-09-0401-C009.indd 379
11/8/10 11:47:53 PM
380
Handbook of Human Factors in Medical Device Design
typical gas cylinder and yoke assembly such as found on an anesthesia machine. Uniquely positioned index pins on the yoke assembly mate with corresponding pin holes on the cylinder head. Cylinder heads for oxygen and nitrous oxide show different pinhole positions. Guideline 9.46: Keying of Connectors Keying should be incorporated in connectors wherever practical.
9.6.1.2 Interlocks Interlocks are design features that physically constrain user actions. Safety interlocks preclude unsafe actions. Because mixing of anesthetic agents is capable of harming the patient, interlock systems are incorporated in anesthesia machines to prevent the connection of more than one vaporizer to the fresh gas circuit (Cicman et al., 2000). Interlocks can also decrease the risk of accidental disconnection. Guideline 9.47: Connector Interlocks Interlocks should be incorporated whenever there is a risk that the user may inadvertently connect or disconnect a connector, causing patient harm.
9.6.1.3 Unique Configurations Unique configurations (e.g., pins, geometric shapes, and size) allow connectors to be mated without making a wrong connection. For example, the electrical connectors (numbers 10 and 11 in Figure 9.1) have different shapes (round and rectangular) that would prevent them from being connected to the other’s mating connector. Other connectors with different configurations are shown mounted on a physiological monitor in Figure 9.20. The Diameter Index Safety System (DISS) utilizes specific diameters of connectors that are
FIGURE 9.20 Three connectors on a physiological monitor. Each is different from the others. Note also that each is labeled for its intended function: adult NIBP, neonatal NIBP, and SpO2. Also note, however, that the adult NIBP and neonatal NIBP connectors are female Luer and male Luer, respectively. Air is emitted through these connectors to inflate a blood pressure cuff. As discussed in Section 9.3.7, the use of Luer connectors on medical devices that emit pressurized air is not recommended because of the risk of mating Luer connectors to vascular access devices (e.g., IV administration sets), leading to potentially fatal air embolism.
TAF-ER562X_WINGR-09-0401-C009.indd 380
11/8/10 11:47:53 PM
Connections and Connectors
381
unique to specific gases to prevent hazardous misconnections (e.g., nitrous oxide supply system to an oxygen hose) (Compressed Gas Association, 1989). Guideline 9.48: Unique Connector Configurations Unique configurations should be employed whenever there is the possibility of connecting incompatible devices, which could pose a hazard to the patient. Labels (e.g., instructions for use and warnings), indicators (e.g., lights and sounds), and cues (e.g., tactile, visual, and auditory) inform users of connector identity, operation, and status (see Chapter 8, “Visual Displays”; Chapter 11, “Software User Interface”; and Chapter 13, “Signs, Symbols, and Markings”).
9.6.1.4 Color Coding Visual cues, such as using the same color for both connector and mating receptacle or two mating connectors, will assist the user in making proper connections and can indicate connector or device function. Color is used to differentiate certain gas connectors. For example, in the United States, green indicates oxygen. Color also identifies liquid anesthetic agents (e.g., blue for sevoflurane and purple for isoflurane), facilitating correct vaporizer filling. Connectors in arterial and venous blood lines (e.g., Tesio catheters for vascular access in hemodialysis) can be distinguished by their colors: red (arterial) and blue (venous). However, the designer should be aware that color coding is not foolproof and should not be relied on as the only defense against misconnections. Color coding is not universal and has been utilized in medical device connector design only fairly recently. Therefore, designers should consider that if connectors can fit together, the user will consider the connection to be properly made regardless of the color coding. Guideline 9.49: Color Coding of Connectors Connections should be color-coded wherever possible. Where applicable, colors should reflect any standards or conventions applicable to the agent transmitted through the connectors.
9.6.1.5 Shapes Shape and size cues assist the user in expeditiously making the proper connections. Guideline 9.50: Shape Coding of Connectors Mating connectors should have similar shapes (e.g., cylindrical or oblong).
9.6.1.6 Labels Labels, by various textual and graphical means, convey information to the user (e.g., in Figure 9.20, labels identify the purpose of each connector: adult NIBP, neonatal NIBP, and SpO2) (see also Chapter 13, “Signs, Symbols, and Markings”). Incorporating succinct instructions into the connector itself can assist the user in making the connection properly. For example, the two defibrillator electrodes (patient connection) for use with an automatic external defibrillator (Figure 9.21) are labeled with instructions for use (i.e., a schematic of a person with arrows directing the viewer’s eyes to electrode application points). Guideline 9.51: Connector Labeling Connectors should be labeled to indicate their intended application and use.
TAF-ER562X_WINGR-09-0401-C009.indd 381
11/8/10 11:47:53 PM
382
Handbook of Human Factors in Medical Device Design
FIGURE 9.21 Automated external defibrillator and patient electrodes. Both electrodes have labels showing the recommended point of application to patient. Arrows direct the user’s attention to electrode placement.
Guideline 9.52: Instructions for Connector Use The designer should provide instructions and guidance on proper connector use on the connector itself, if space permits, as well as on associated labels and in user manuals, service manuals, or package inserts. Instructions for use should contain graphic representations of connectors and connector application (e.g., insertion, locking, unlocking, and removal).
Guideline 9.53: Warnings about Connectors Any hazard warnings should be prominent and understandable. All danger, warning, and safety instructions should be in accordance with applicable safety standards and conventions.
Guideline 9.54: Connector Label Content A connector should bear clear identification markings, typically including manufacturer, model number, and essential specifications whenever possible. For example, an electrical connector should contain labeling indicating its electrical rating (e.g., 10 to 13 amps at 125 VAC).
Guideline 9.55: Legibility of Connector Labels Letters and numbers should be as large as possible to be read at a comfortable distance without the need for extra illumination (Shurtleff, 1980).
Care should be taken when using colors. The best relative legibility of color combinations under white light is achieved with black letters on a white background. Poor legibility results from such combinations as green on red, red on green, orange on black, and orange on white. Connectors should not use embossed lettering and symbols with little or no contrast with the connector surface color (e.g., the embossed lettering “INSERT TO LINE” barely visible on the power cord plug in Figure 9.10). Guideline 9.56: Connector Label Visibility and Orientation Labeling should be visible to the user particularly in a location and orientation that can be seen during connector use.
TAF-ER562X_WINGR-09-0401-C009.indd 382
11/8/10 11:47:53 PM
Connections and Connectors
(a)
383
(b)
FIGURE 9.22 (a) Surgiport seal adapter and trocar assembly. Black marks aid alignment of a trocar assembly. The arrow on the seal adapter knob indicates the direction the knob must be turned in order to lock the adapter to the trocar assembly. (b) Surgiport seal adapter with ample indentations on knob to enable a sure grip, thus facilitating its turning for a secure connection to the trocar assembly.
Labels in the form of symbols can aid a user in identifying the purpose of the connector and the method of connector application. For example, Figure 9.1 shows an electrical plug with a green dot indicating that it is “hospital grade,” and Figure 9.22 shows an arrow indicating the direction to turn the connector for locking an adapter to a trocar assembly. Guideline 9.57: Symbols on Connectors Labels should incorporate commonly used symbols rather than text, especially where space is limited. Symbols should also be considered when the user population is multilingual. All symbols, however, should be tested for understandability by the intended users.
Guideline 9.58: Connector Label Language The language used on connector labels should be consistent with that of the intended users.
Guideline 9.59: Familiar Words on Connector Labels Labels should employ familiar words, terms, and symbols for ease of understanding.
Connectors may be handled frequently and exposed to harsh environments that can wear away labeling if it is not durable. Decals, paper labels, and pressure-sensitive labels normally do not provide the degree of permanence needed for most applications. Guideline 9.60: Durability of Connector Labels Labels and markings should be permanent and should remain legible throughout the intended life of the connector under anticipated use and maintenance conditions.
9.6.1.7 Connection Status Indicators: Tactile, Auditory, and Visual Cues Connectors may appear to be physically connected even though electrical contacts, for example, may not be established (e.g., connector from a battery charger when it is partially inserted into a jack on a battery-operated monitor). Designers should incorporate connection status indicators with such features as visual, tactile, and auditory cues to provide acknowledgment to the user that a proper connection was made.
TAF-ER562X_WINGR-09-0401-C009.indd 383
11/8/10 11:47:54 PM
384
Handbook of Human Factors in Medical Device Design
Guideline 9.61: Connection Status Connectors should not appear to be securely connected when they are not.
Guideline 9.62: Connection Status Indicators The connector should assure the user of the status of the connection (e.g., disconnected, partially connected, or fully connected). For example, connection status cues can include a particular force, familiar to a trained user, required to make a successful connection, where too little force or excessive force could alert the user to an improper connection.
Guideline 9.63: Tactile Status Confirmation Cues Connectors should incorporate, where practical, tactile cues to assure the user that a complete and proper connection has been made. Other cues include mechanical pulses or vibrations that the user’s fingers sense when such connectors are successfully mated.
Guideline 9.64: Audible Status Confirmation Cues Connectors should incorporate, where practical, audible cues to assure the user that a complete and proper connection has been made. The ambient sound level in the environment of use should be considered when designing the loudness and frequency of the audible cue. The sound of a connection being made can give the user a sense of confidence that a proper connection has been made (e.g., making and breaking a USB connection to a personal computer may produce an ascending and descending tone sequence, respectively).
Guideline 9.65: Visual Status Confirmation Connectors should incorporate, where practical, visual cues to assure the user that a complete and proper connection has been made. For example, illumination of a light or lights may be used as a means of confirming the operational status of a connector.
9.6.1.8 Alignment Marks Marks assist the user in aligning mating connectors. They also can indicate when a connection has been made properly. For example, Figure 9.23 shows an Autosuture GIA Universal Stapler hand piece and a stapler cartridge. When connecting the cartridge to the hand piece, the user aligns the mark on the cartridge (arrow 2) with the mark on the hand piece (arrow 3). The user then inserts the cartridge and rotates it to bring arrow 1 into alignment with arrow 3.
1
2
3
FIGURE 9.23 Alignment marks on the Autosuture GIA Universal Stapler hand piece (arrow 3) and a stapler cartridge (arrows 1 and 2) aid in proper insertion (arrows 2 and 3) and serve as an indication of proper connection (arrows 1 and 3).
TAF-ER562X_WINGR-09-0401-C009.indd 384
11/8/10 11:47:54 PM
Connections and Connectors
385
Guideline 9.66: Alignment Marks Alignment marks should be used whenever possible to indicate complete and proper engagement of mating connectors, especially for connections that require rotational motion. Connectors should be designed to prevent connection of incompatible devices (e.g., patient leads to power cords or enteral feeding tubing to intravenous lines).
9.6.1.9 Prevention of Disconnection Unintentional or partial disconnection can adversely affect both patient and operator. Considering an anesthesia system as an example, a breathing circuit disconnection interrupts oxygen, endangering the patient, and a scavenging system disconnect contaminates the ambient air with harmful anesthetic vapors, endangering the operating room personnel. Guideline 9.67: Disconnection Guards Mechanisms to protect from unintentional disconnection (e.g., locks, screw collars, and clips) should be incorporated into connectors. This feature is especially important for life support and life-critical devices, such as ventilators and defibrillators.
Guideline 9.68: Force to Engage Locks Locks should be easily engaged yet, once engaged, should be resistant to accidental or intentional unlocking by, for example, a patient or user contacting the connector.
Guideline 9.69: Force and Direction to Lock and Unlock In locking connectors, the locking force and security of the connection should ensure that unintentional disconnection will not occur when, for example, the cable is pulled unintentionally.
Figure 9.24 shows a latching mechanism that secures a trocar and cannula assembly used in minimally invasive endoscopic surgery. Guideline 9.70: Latches and Locks When accidental disconnection is possible, connectors should have latches or locks that can easily and intentionally be released as with a finger actuation, a twist of a threaded collar, or a push–pull movement.
FIGURE 9.24 Thumb and forefinger depress tabs to engage and latch trocar assembly (left) to sleeve (cannula) assembly (right).
TAF-ER562X_WINGR-09-0401-C009.indd 385
11/8/10 11:47:54 PM
386
Handbook of Human Factors in Medical Device Design
The possibility of disconnection is particularly high in a heavily trafficked area where passersby can trip over cables or bump against connectors. Figure 9.18 shows connectors incorporating a latch in which an extra manipulation is required to disconnect the connector. These push–pull latching connectors require a push, or a movement in the long axis direction, to unlatch the connector and a pull away from the mating receptacle to disengage the connector. Guideline 9.71: Friction Fit Connections Friction-fit connections (e.g., between disposable corrugated polyethylene tubing and other breathing circuit connectors) should be sufficiently secure so as to withstand normal forces encountered in a patient’s care.
The patient connection is often the weakest link in the application of a medical device where partial or complete disconnection of, for example, an electrosurgical return electrode, endotracheal tube Y-piece, or vascular access device can result in serious injury or death. Some patient connections incorporate alarm systems, such as electrosurgical patient return electrodes. Guideline 9.72: Device Disconnection Alarms Connections of a critical nature, where a complete or partial disconnection presents a hazard to the patient, should activate an audible and visual alarm when disconnected.
Guideline 9.73: Patient Connector Weight The connector should not be so heavy as to apply forces on the connector–patient connection that could deform and otherwise degrade the connection.
Guideline 9.74: Patient Connector Safety and Comfort Patient connectors should not be bulky, sharp, or heavy so as to compromise patient safety and comfort. A connector’s exterior should be smooth and free of pinch points, sharp edges, and other irregularities that could injure the user.
9.6.1.10 Standards Many standards and regulations exist concerning the design and use of connectors (e.g., American Society for Testing and Materials [ASTM], 1989, 1998a, 1998b; ANSI/AAMI, 1986, 1995; ANSI/HIMA, 1983; FDA, 1997; International Electrotechnical Commission, (1998; ISO, 1986, 1991, 1996; Underwriters Laboratories, 2001). Patient injury has provided the rationale for many requirements in these documents. Standards development is an ongoing process, and designers should be aware of standards under development. For example, the European Committee for Standardization (CEN) is drafting a standard titled Small Bore Connectors for Liquids and Gases in Healthcare Applications (CEN, 2003), and the AAMI recently formed a committee for “Medical Device Tubings Connectors Standard to Prevent Misconnections.” Guideline 9.75: Compliance with Connector Standards The designer should, at a minimum, comply with any applicable standards, regulations, and recommendations relating to connection and connector design.
TAF-ER562X_WINGR-09-0401-C009.indd 386
11/8/10 11:47:55 PM
Connections and Connectors
387
9.6.2 PHYSICAL INTERACTION Design features such as size, shape, surface characteristics, and application forces affect the physical interaction between the connector, user, and patient. 9.6.2.1 Physical Interaction with the Patient The following factors affect the user’s ability to make connections: 1. Magnitude and direction of force to connect and disconnect 2. Connector size and shape 3. Magnitude and direction of force to lock and unlock 4. Force to activate and modes of activation of interlocks Connectors can be connected utilizing a range of forces and motions, including pushing, pulling, twisting, pivoting, screwing, levering, or a combination of these modes. Guideline 9.76: Force to Connect and Disconnect The force required to make a connection and a disconnection should not be excessive for the user’s comfort and ability.
Guideline 9.77: Connector Gripping Surface The connector surface should have an adequate gripping surface for the user to apply the necessary forces to connect or disconnect the connector. The designer should consider the anthropometry of both user and patient (see also Chapter 4, “Anthropometry and Biomechanics”). Connector design should address the variation in expected users’ physical characteristics (e.g., small and large hands) and intended applications (e.g., flow-directed microcatheter connections to radiographic unit high-voltage cable connectors).
The size and location of connectors affect their accessibility and use. For example, Figure 9.25 shows eight appropriately sized patient-monitoring cable connectors connected to the front panel of a monitor. These connectors do not crowd one another because there is ample space between adjacent connectors, permitting hand and finger access.
FIGURE 9.25 Patient monitoring cable connectors connected to the front panel of a monitor. Connectors are appropriately sized and spaced to permit easy access.
TAF-ER562X_WINGR-09-0401-C009.indd 387
11/8/10 11:47:55 PM
388
Handbook of Human Factors in Medical Device Design
Guideline 9.78: Connector Dimensions A connector should fit comfortably in the hand or fingers. It should not be so large that it crowds other connectors or components when it is connected. Connectors should not be so small as to be difficult to handle and operate.
Guideline 9.79: Connector Accessibility Connectors located adjacent to obstructions (e.g., walls, panels, or other device parts) should be appropriately sized to allow easy connection and disconnection.
Guideline 9.80: Connector Spacing In devices incorporating multiple connectors, connector design should allow easy connecting or disconnecting without having to disturb adjacent connectors.
Guideline 9.81: Multiple Connector Orientation In devices incorporating multiple connectors of similar design, alignment members (e.g., keys or slots) should be in the same relative orientation (e.g., all keyed connectors oriented on a panel with keys in the 12 o’clock position).
Guideline 9.82: Connections in Hostile Environments Designers should take into account hostile environmental factors (e.g., mechanical vibrations during ambulance or helicopter transport) and design to optimize ease of use. Such strategies as larger connectors, wider spacing between adjacent connectors, more pronounced keys, and guides can be employed.
9.6.2.2 Protection from Connector Damage and Contamination The designer should incorporate features that protect the connector from damage from improper use (forcing a connector into the wrong receptacle) and harmful environmental factors (fluids and dirt). Guideline 9.83: Internal Conductor Protection Conductive pins should not be exposed to bending or rotational motion that could damage them as a connector is connected or disconnected from its mating connector. Alignment of the connector housing before pin-to-socket contact is a means to protect pins from damage.
Guideline 9.84: Protection of Disconnected Connectors When connectors are disconnected, internal components (such as pins and keys) should be protected from damage.
Guideline 9.85: Strength of Connectors The force required to make a connection and a disconnection should not be greater than what the connector and its component parts can withstand without damage.
Guideline 9.86: Connector–Device Interface Designers should consider the forces likely to be exerted by the user on the connector–device interface (e.g., connector to cable, hose, or panel) and design appropriately with, for example, adequate strain relief to protect this interface from damage.
TAF-ER562X_WINGR-09-0401-C009.indd 388
11/8/10 11:47:55 PM
Connections and Connectors
389
Guideline 9.87: Connector Contamination Protection Connector surfaces should not be tacky or sticky so as to retain foreign matter. High-friction, rough surfaces, if required, should not pose a cross-contamination hazard.
Guideline 9.88: Dirty Connectors Colors and patterns on connector surfaces should not camouflage dirt or other contaminants. The user should be able to readily see if a connector is dirty.
Guideline 9.89: Cleaning Connectors Connector design should facilitate cleaning. Alternatively, designs should minimize or eliminate the need for cleaning. Surface configuration of reusable connectors should permit easy cleaning, decontamination, or sterilization.
Guideline 9.90: Protecting Connectors from the Environment Connectors should not be damaged by the environment they are likely to encounter (e.g., fluids, filth, vermin, and mechanical shock and vibration).
Guideline 9.91: Resistance to Surface Contaminants To the extent possible, connector surfaces should be devoid of any cracks, crevices, nooks, or crannies that could harbor pathogens and other contaminants. The surface should be wear resistant, as surface scratches and pits could retain bodily fluids and filth. Removable connections can be a source of intermittent contact in electrical connectors (Nielsen, 2004). Cleaning of contact surfaces is required to remove dirt and corrosion that cause poor electrical contact.
Guideline 9.92: Connector Maintenance Designers should recommend connector inspection and maintenance procedures as well as time intervals between inspections.
9.6.3 CONVENIENCES User conveniences are those design features that enhance a connector’s ease of operation and user satisfaction and, as a consequence, may lessen the risk of misconnection. For example, the Surgiport trocar assembly seal adapter, shown in Figure 9.22, has a knob with ample indentations to facilitate a sure grip by the user. Guideline 9.93: Connector Grip Enhancers A connector should have surfaces and shape that enhance the user’s ability to engage and disengage the connection. The grip should comfortably conform to the hand and fingers, facilitating the action of connecting and disconnecting.
Guideline 9.94: Connector Guides Connectors should have guides (e.g., keys and indicators) that enable the user to make a connection unambiguously without having to search for the correct orientation for connector-toconnector alignment.
TAF-ER562X_WINGR-09-0401-C009.indd 389
11/8/10 11:47:55 PM
390
Handbook of Human Factors in Medical Device Design
Guideline 9.95: Connector Appearance A connector should appear well made, be recognizable, and be of a quality similar to or exceeding that of associated medical devices, systems, and facilities. The appearance, finish, and color of a connector, as perceived by users, patients, and visitors, should reflect the connector’s high quality.
Guideline 9.96: Connector Construction Quality Connectors made of transparent and translucent materials should not contain obvious inclusions, cracks, or sections of varying refractive index that would give the impression of a broken or poorly machined, molded, or extruded piece.
9.7 CASE STUDIES 9.7.1 ELECTRICAL CONNECTORS: ECG CABLE Diagnostic-quality ECGs depend on the proper placement of electrodes. A mistake in electrode application can result, at best, in a repeat of the examination with the errors corrected and, at worst, in an inaccurate diagnosis. Stress and urgency may contribute to an ECG technician’s errors. Figure 9.26a shows five ECG leads and an ECG cable. While individual ECG leads are generally color-coded (brown, white, red, black, and green), the leads are separate and can be misplaced. If a lead of one color is lost, a lead of another color may be used as a substitute. Confusion over lead placement could then occur. The leads have exposed metal conductor pins, which, as discussed above in Section 9.3.3 (see also Figure 9.10), are a safety hazard. Human factors design techniques have yielded connectors that mitigate, if not eliminate, such errors. The use of color coding, keying, labeling, and instructions resulted in the connectors shown in Figure 9.26b. This figure shows an ECG cable connector and five mating ECG leads. Color coding is effectively used to match mating connectors. The ECG lead connectors (brown, white, red, black, and green) are held in the cable yoke. Connectors at the patient electrode end of the leads (not shown in the figure) are color-coded to match the colors of the connectors attached to the yoke. Each lead bears a color that relates to the particular point of attachment to the patient (e.g., white for right leg [RL] and black for left
(a)
FIGURE 9.26 and leads.
(b)
(a) The old: individual ECG leads and ECG cable. (b) The new: Curbell ECG cable
TAF-ER562X_WINGR-09-0401-C009.indd 390
11/8/10 11:47:55 PM
Connections and Connectors
391
arm [LA]). The color of each lead matches the color of a corresponding dot on the cable connector. The dots are labeled with a lead position (e.g., the white dot is labeled RL and the black dot is labeled LA). This color-matching scheme assists the user in making the correct connections. The lead connectors in the cable yoke are keyed so that only one connection orientation is possible. The body of the ECG cable connector also displays instructions (i.e., a schematic diagram of the patient showing points of electrode lead attachment). As with all applications of color coding, color should provide redundant information. Other pertinent information is derived from the schematic diagram of the patient showing lead placements and the labeling of each colored dot with the location for the attachment of the corresponding electrodes. The time to connect all five leads to the cable is less for the new design than it was for the old, where five separate leads had to be connected.
9.7.2 LIQUID CONNECTOR: ANESTHESIA SYSTEM VAPORIZER FILLER The vaporizer on an anesthesia system controls the rate of delivery of inhaled anesthetic. It introduces a precise volume percentage of anesthetic vapor into the fresh gas stream that travels to the patient. A vaporizer is designed to deliver a precise volume percentage but only for one specific agent. A vaporizer designed to work with halothane, for example, will give erroneous volume percentages if the vaporizer is filled with isoflurane. Filling a vaporizer with the wrong agent or adding the wrong agent to a partially filled vaporizer may give a concentration that is significantly higher or lower than the concentration set on the control dial. One to three vaporizers, each capable of delivering a specific anesthetic agent, typically are connected to an anesthesia workstation at any one time. Figure 9.4, for example, shows two vaporizers connected to an anesthesia workstation (arrow).
FIGURE 9.27 Dräger Vapor® anesthetic vaporizers for sevoflurane, halothane, enflurane, and isoflurane.
TAF-ER562X_WINGR-09-0401-C009.indd 391
11/8/10 11:47:55 PM
392
Handbook of Human Factors in Medical Device Design
2 5
3
4
1
FIGURE 9.28 Anesthetic agent bottle with anesthetic agent-specific collar on the neck (1). A keyed filler adapter (2) with check valve (3) is connected to the bottle, and the keyed filler adapter is screwed firmly (4) into the anesthetic agent bottle. The square section (5) of the keyed filler adapter is rotated so that holes are on the underside.
Interlock systems prevent the activation of more than one vaporizer at a time as this could injure the patient. Open filling techniques, where the liquid is simply poured into a port in the vaporizer from a supply bottle, risk dangerous spills and introduction of the wrong agent. To reduce these risks, the company designed and developed a filling system with keyed and color-coded connectors to preclude misconnection of bottled anesthetic liquid to the wrong vaporizer. Figure 9.27 shows four agent-specific anesthetic vaporizers. Figures 9.28, 9.29, and 9.30 describe the Dräger anesthesia vaporizer filling system.
6
8
7
FIGURE 9.29 Lever (6) is swung out to relieve the pressure in the vaporizer. The keyed sealing block (7) is pulled out of its corresponding keyed filling port (8) and folded down.
TAF-ER562X_WINGR-09-0401-C009.indd 392
11/8/10 11:47:56 PM
Connections and Connectors
393
6 5,8
9
FIGURE 9.30 The square section (5) of the keyed filler adapter is pushed completely into its corresponding keyed filling port (8) in the vaporizer until it engages. The lever (6) is swung back and tightened to ensure an adequate seal, preventing liquid from leaking into the environment, which would be an environmental health hazard. The user swings the anesthetic agent bottle upside down, holding it in this position while observing the filling level through the viewing glass (9).
REFERENCES American National Standards Institute/Association for the Advancement of Medical Instrumentation (ANSI/AAMI). (1986). Electrosurgical Devices (3rd ed.). ANSI/AAMI HF-18-RC01-1986 (revised 2001). New York: American National Standards Institute. American National Standards Institute/Association for the Advancement of Medical Instrumentation (ANSI/AAMI). (1995). ECG Cables and Lead Wires. ANSI/AAMI EC-53-1995 New York: American National Standards Institute. American National Standards Institute/Association for the Advancement of Medical Instrumentation (ANSI/AAMI). (1996). Enteral Feeding Set Adapters and Connectors. ANSI/AAMI ID-541996 (reaffirmed 2005). New York: American National Standards Institute. American National Standards Institute/Health Industry Manufacturers’ Association. (1983). Medical Material—Luer Taper Fittings—Performance. ANSI/HIMA MD-70.1. New York: American National Standards Institute. American Society for Testing and Materials. (1989). Specification for Pneumatic Tourniquet Equipment. BS 7088:1989. West Conshohocken, PA: American Society for Testing and Materials. American Society for Testing and Materials. (1998a). Medical Gas Pipeline Systems. Anaesthetic Gas Scavenging Disposal Systems. BS EN 737-2:1998. West Conshohocken, PA: American Society for Testing and Materials. American Society for Testing and Materials. (1998b). Specification for Probes (Quick Connectors) for Use with Medical Gas Pipeline Systems. BS 5682:1998. West Conshohocken, PA: American Society for Testing and Materials. American Society for Testing and Materials. (2004). Anaesthetic and Respiratory Equipment. Conical Connectors. Cones and Sockets. BS EN ISO 5356-1:2004. West Conshohocken, PA: American Society for Testing and Materials. BS and S. (2003). Another near miss with IV tubing. Biomedical Safety and Standards, 33(19), 145–146. European Committee on Standardization. (2003). Small Bore Connectors for Liquids and Gases in Healthcare Applications [draft]. CEN/TC BT/TF 123:2003. Brussels: European Committee for Standardization.
TAF-ER562X_WINGR-09-0401-C009.indd 393
11/8/10 11:47:56 PM
394
Handbook of Human Factors in Medical Device Design
Cicman, J. H., Gotzon, J., Himmelwright, C., Laubach, S., Skibo, V. F., and Yoder, J. M. (2000). Operating Principles of Narkomed Anesthesia Systems (2nd ed.). Telford, PA: North American Dräger. Compressed Gas Association. (1989). Diameter Index Safety System (Non-Interchangeable Low Pressure Connections for Medical Gas Applications) CGA V5. Arlington, VA: Compressed Gas Association. Cyna A., Simmons S., Osborn K., and Andrew, M. (2002). Fatal epidural infusion—Call for a system-wide change [letter]. Anaesthesia Intensive Care 30, 99–100. Dubinsky, I. (1992). Fictitious hypoxia—A result of misconnection of oxygen tubing to a wall source of air [letter]. American Journal of Emergency Medicine, 10(6), 613–614. Dyro, J. F. (1998). Methods for analyzing home care medical device accidents. Journal of Clinical Engineering, 23, 359–368. Dyro, J. F. (2004). General hospital devices: Beds, stretchers, and wheelchairs. In J. F. Dyro (Ed.), The Clinical Engineering Handbook (pp. 421–436). Burlington, MA: Elsevier. Dyro, J. F., and Morris, R. L. (2004). Medical device troubleshooting. In J. F. Dyro (Ed.), The Clinical Engineering Handbook (pp. 436–447). Burlington, MA: Elsevier. Dyro J. F., and Shepherd, M. (2005). Enhancing patient safety by white tape analysis. Journal of Clinical Engineering, 30, 134–144. Emergency Care Research Institute (ECRI). (2004). Fatal air embolism caused by misconnection of medical device hoses to needleless Luer ports on IV administration sets [hazard report]. Health Devices, 33(6), 223–235. Institute for Safe Medication Practices. (2001). IV connection in tracheostomy cuff inflation port reflects larger problem. ISMP Medication Safety Alert, November 28, 2001. Available: http:// www.ismp.org. Institute for Safe Medication Practices. (2004). Problems persist with life-threatening tubing misconnections. ISMP Medication Safety Alert, June 17, 2004. Available: http://www.ismp.org. International Electrotechnical Commission. (1998). Medical Electrical Equipment—Part 2-2: Particular Requirements for the Safety of High Frequency Surgical Equipment (3rd ed.). IEC 60601-2-2 (1998–09). Geneva: International Electrotechnical Commission. International Organization for Standardization. (1986). Conical Fittings with a 6 Percent (Luer) Taper for Syringes, Needles and Certain Other Medical Equipment—Part One: General Requirements, ISO 594/1. Geneva: International Organization for Standardization. International Organization for Standardization. (1991). Conical Fittings with a 6 Percent (Luer) Taper for Syringes, Needles and Certain Other Medical Equipment—Part Two: Lock Fittings, ISO 594/2. Geneva: International Organization for Standardization. International Organization for Standardization. (1996). Anesthetic and Respiratory Equipment, Conical Connectors—Part 1. Connectors and Sockets 5356-1:1996. Geneva: International Organization for Standardization. Katcher, M. L., Shapiro, M. M., and Guist, C. (1986). Severe injury and death associated with home infant cardiorespiratory monitors. Pediatrics, 78(5), 775–779. Nielsen, R. (2004). Identifying and fixing intermittent problems. Biomedical Instrumentation and Technology, 38(3), 207–208. Qualls, C. D., Harris, J. L., and Rogers, W. A. (2001). Cognitive-linguistic aging: Considerations for home health care environments. In W. A. Rogers and A. D. Fisk (Eds.), Human Factors Interventions for the Health Care of Older Adults (pp. 47–68). Mahwah, NJ: Lawrence Erlbaum Associates. Sato, T. (1991). Fatal pipeline accidents spur Japanese standards: Classic O2-N2O pipe switch causes 2 deaths before problem caught. Anesthesia Patient Safety Foundation Newsletter 6(2),13–24. Shepherd, M. (2004). Systems approach to medical device safety. In J. F. Dyro (Ed.), The Clinical Engineering Handbook (pp. 246–249). Burlington, MA: Elsevier. Shurtleff, D. A. (1980). How to Make Displays Legible. La Mirada, CA: Human Interface Design.
TAF-ER562X_WINGR-09-0401-C009.indd 394
11/8/10 11:47:56 PM
Connections and Connectors
395
Underwriters Laboratories. (2001). Attachment Plugs and Receptacles, Standard 498. Melville, NY: Underwriters Laboratories. U.S. Food and Drug Administration. (1997). Medical devices: Establishment of a performance standard for electrode lead wires and patient cables [final rule]. Federal Register 62(90), 25477–25498.
TAF-ER562X_WINGR-09-0401-C009.indd 395
11/8/10 11:47:56 PM
TAF-ER562X_WINGR-09-0401-C009.indd 396
11/8/10 11:47:56 PM
10 Alarm Design Stephen B. Wilcox, PhD, FIDSA CONTENTS 10.1 General Principles ..................................................................................................399 10.1.1 Central Goals When Developing Auditory Alarm Signals .................... 400 10.1.2 Inherently Meaningful versus Abstract Auditory Alarm Signals............401 10.1.3 Speech-Based (or Verbal) Auditory Alarm Signals................................ 402 10.2 Special Considerations .......................................................................................... 402 10.3 Design Guidelines ..................................................................................................403 10.3.1 Step 1: Gathering Information .................................................................403 10.3.2 Step 2: Determining the Conditions That Require Alarm Signals ......... 404 10.3.3 Step 3: Creating Signal-Processing Algorithms ......................................405 10.3.3.1 Adjustable Alarm Limits..........................................................405 10.3.3.2 Disabling Alarm Signals or Portions of Alarm Systems ........ 406 10.3.3.3 Other Alarm Condition Considerations .................................. 407 10.3.4 Step 4: Determining the Information That Should Be Communicated about Each Alarm Condition .................................................................. 408 10.3.4.1 Likelihood ............................................................................... 408 10.3.5 Step 5: Allocating Signaling Modalities to Alarm Conditions............... 409 10.3.6 Step 6: Creating a Simulated Use Environment ...................................... 411 10.3.6.1 Visual Environment ................................................................. 411 10.3.6.2 Auditory Environment ............................................................. 411 10.3.7 Step 7: Creating Auditory Alarm Signals ................................................ 414 10.3.7.1 Constructing Abstract Auditory Alarm Signals ....................... 414 10.3.7.2 Design of Alarm Signal Pulses ................................................ 414 10.3.7.3 Design of Bursts ....................................................................... 417 10.3.7.4 Communicating Urgency ......................................................... 417 10.3.8 Step 8: Creating Visual Alarm Signals.................................................... 418 10.3.8.1 Information-Providing Visual Alarm Signals (Information Displays) .................................................................................. 418 10.3.8.2 Attention-Getting Visual Alarm Signal Signals ...................... 418 10.3.8.3 Other Considerations regarding Visual Alarm Signals ............ 419 10.3.9 Step 9: Creating Other Alarm Signals ..................................................... 419 10.3.10 Step 10: Testing Prototype Alarm Systems with Potential Users ............ 419 10.3.11 Step 11: Refining Alarm Systems Based on Test Results ........................420
397
TAF-ER562X_WINGR-09-0401-C010.indd 397
11/8/10 11:55:24 PM
398
Handbook of Human Factors in Medical Device Design
10.4 Case Study .............................................................................................................420 10.4.1 Auditory Alarms on an Intravenous Infusion System .............................420 Acknowledgments............................................................................................................421 References ........................................................................................................................421 This chapter focuses on the design of alarm systems, which we define as systems for indicating the presence of (generally) temporary, potentially dangerous conditions in real time. This distinguishes alarms from warnings, which are not real time and which tend to refer to permanent rather than temporary conditions, and from other displays and indicators, which do not necessarily refer to dangerous conditions. We use the term alarm systems rather than simply alarms to make it clear that we are referring not just to the signal that an alarm system generates but rather to the “parts of medical … equipment … that detect alarm conditions and … generate alarm signals” (International Electrotechnical Commission [IEC], 2004). An ideal therapeutic medical device would have no alarm systems at all because no dangerous conditions would arise in the first place. Figure 10.1 shows, in outline form, a simple device that alters therapy as a function of patient conditions while precluding dangerous conditions. Figure 10.2 adds the potential for dangerous conditions along with a means to address those conditions automatically. An alarm system is not required for this ideal therapeutic medical device (colloquially called a “smart” device) because it automatically recognizes and addresses dangerous conditions without the need for human intervention. Of course, this model is relevant only for therapeutic medical devices. Nontherapeutic (i.e., diagnostic) devices, such as patient monitors, are specifically designed to provide information about patient states, including conditions that require alarm signals. Because, by definition, nontherapeutic devices do not provide therapy to address relevant patient conditions, only therapeutic devices, at least in principle, could dispense with alarms. Devices such as the one illustrated in Figure 10.2 are rarely, if ever, achievable, which leads to devices, as shown in Figure 10.3, where human intervention is necessary to ensure safety. The purpose of alarm signals is to solicit human intervention. Although we assume that alarm signals are necessary, the need for human intervention to maintain safety introduces a failure mode due to the inevitable imperfection of human users. Operator dependence is a safety flaw because it puts humans, who are not always reliable, in the causal chain for achieving safety. Humans are prone to fatigue, bad judgment, “habituation” (the loss of awareness of hazards with repeated exposure), attentional deficits, inaccurate movements, and so on. It is seldom possible to eliminate all operator dependence, nor is Patient
Sensors
Therapy
Device
FIGURE 10.1 A model of a hypothetical therapeutic medical device that provides therapy based on patient parameters and does not contain any hazards.
TAF-ER562X_WINGR-09-0401-C010.indd 398
11/8/10 11:55:24 PM
Alarm Design
399
Patient Sensors
Patient Sensors
Device
Therapy Device
Sensors
Therapy
Intervention
Sensors Incident prevention system
Intervention Incident prevention system
FIGURE 10.2 An ideal therapeutic device does not include an alarm system because potentially dangerous conditions are dealt with automatically by the device itself.
Alarm signals
Human intervention
FIGURE 10.3 A model of an actual therapeutic device that requires human intervention to maintain safety.
it always desirable to do so, because automated systems have their own problems. Thus, designers should think in terms of the “gaps” in safety that can be addressed, in effect, by a person/machine “partnership.” However, everything else being equal, operator dependence should be eliminated where feasible. A system that remains safe without human intervention is safer than an equivalent system requiring human intervention to remain safe.
10.1 GENERAL PRINCIPLES The purpose of an alarm system is to elicit appropriate human intervention. This means that alarm systems have three basic requirements: 1. To accurately and reliably detect dangerous conditions 2. To immediately obtain the operator’s attention when a dangerous condition occurs 3. To clearly and accurately tell the operator what the problem is and, if feasible, what to do to correct the problem The first requirement is crucial because alarm signals, no matter how well designed and presented, are effective only to the extent that they are triggered by conditions that require human intervention. Thus, prior to designing the alarm signals, the alarm designer must answer two crucial questions: “What should the alarm signals indicate?” and “What alarm conditions should trigger the alarm signals?” In sum, an effective alarm system requires both the appropriate choice of alarm conditions and good design of alarm signals to address each of the conditions. This chapter provides a series of steps to make it as likely as possible that alarm systems will, in fact, be properly designed. These steps are as follows: 1. Gathering information, including relevant standards regarding requirements for the particular alarm system under development 2. Generating a list of alarm conditions that require alarm signals (i.e., an “alarm condition inventory”)
TAF-ER562X_WINGR-09-0401-C010.indd 399
11/8/10 11:55:24 PM
400
Handbook of Human Factors in Medical Device Design
3. Creating signal-processing algorithms to detect the proper alarm conditions 4. Determining what information needs to be communicated for each alarm condition in the inventory 5. Allocating one or more alarm-signaling modalities—visual, auditory, and/or other—to each alarm condition 6. Creating a simulated environment for testing the alarm signals 7. Creating alarm signals for each signaling modality 8. Testing the alarm system with potential users, ultimately under realistic use conditions 9. Refining the alarm system to reflect the testing results (and, typically, returning to step 7 and again to step 8 in a cyclical fashion until the alarm system is properly designed) A key issue arising from this proposed design process is that alarm signals, if they are to be effective, must be designed to be compatible with human perceptual/cognitive capabilities. Thus, this chapter provides some information about alarm signal design (specifically auditory alarm signal design because auditory alarm signals typically require the largest effort on the part of the alarm designer) in relation to human capabilities. Section 10.2, “Special Considerations,” provides information that is specific to medical alarm systems (see Chapter 2, “Basic Human Abilities,” for additional information regarding human capabilities and limitations).
10.1.1 CENTRAL GOALS WHEN DEVELOPING AUDITORY ALARM SIGNALS There are a number of goals that any auditory alarm signal should meet, including the following (see Association for the Advancement of Medical Instrumentation [AAMI], 1993; Block, 1994; Edworthy, 1998b; Edworthy and Stanton, 1995; IEC, 2004; Sanders and McCormick, 1987; Stanton and Edworthy, 1999; Weinger and Smith, 1993): • Ensure that auditory alarm signals can be heard above the background noise and/ or other auditory alarm signals • Minimize interference with communication (at the very time that communication can be crucial) • Avoid startling users or pushing them into a “high-arousal” state • Avoid disturbing patients or other people for whom the alarms are not intended • Minimize confusion due to multiple simultaneous auditory alarm signals • Ensure perception by users with partial hearing loss • Avoid irritating users or causing them hearing damage or pain • Clearly communicate the appropriate degree of urgency • Ensure that a given auditory alarm signal is distinct from others • Clearly communicate the source of the alarm signal (i.e., the alarm condition) • Provide consistent auditory alarm signals for similar conditions/devices Unfortunately, as discussed by Edworthy and Stanton (1995) and Welch (1999), these goals are rarely met. Furthermore, according to nurses (see Edworthy and Stanton, 1995; Welch, 1999), auditory alarm signals poorly communicate the degree of urgency and are too loud, too high-pitched, too persistent, too irritating, too “harsh toned,” too sudden,
TAF-ER562X_WINGR-09-0401-C010.indd 400
11/8/10 11:55:25 PM
Alarm Design
401
too confusing, and are often inappropriate. When surveyed, anesthesiologists (Griffith and Raciot, 1992), said that they wanted the following: • • • • •
All devices in a category to sound alike Not all devices to sound To look at monitors before acting For auditory alarm signals to be standardized For auditory alarm signals to be “graded” (to get more extreme as a problem persists) • Four or five distinct sounds to distinguish monitors These desires were not included in the goals listed above because they may not be relevant for a particular alarm system. However, they should be considered when designing monitors for the operating room. Therefore, the burden is on the alarm system designer to carefully consider the above goals and avoid the problems identified by Edworthy and Stanton (1995) and Welch (1999). The specific guidance provided below is designed to meet many of these goals.
10.1.2 INHERENTLY MEANINGFUL VERSUS ABSTRACT AUDITORY ALARM SIGNALS Most auditory alarm signals are abstract rather than inherently meaningful, representative sounds. However, the use of abstract auditory alarm signals usually limits the information that they can convey to notify the user that there is a problem, the source of the problem, and perhaps the degree of urgency associated with the relevant condition. A good case can be made for the use of meaningful, or representational, auditory alarm signals, or what Blattner, Sumikawa, and Greenberg (1989) call “earcons” (see Stanton and Edworthy, 1999). Gaver (1986) distinguishes between three types of meaningful sounds: 1. Symbolic, such as applause 2. Nomic, such as a door closing 3. Metaphoric, such as a falling pitch More meaningful auditory alarm signals have the potential to reduce confusion. Even two simultaneous auditory alarm signals can cause appreciable confusion. People can differentiate, learn, and/or remember only a few different abstract auditory alarm signals (Patterson, 1982; Stanton and Edworthy, 1999). In fact, medical professionals appear to be poor at correctly identifying auditory alarm signals in their environments (Momtahan, Hetu, and Tansley, 1993). On the other hand, human beings can simultaneously perceive an enormous number of meaningful sounds. For example, imagine walking in the country at night and hearing crickets, various birds, and other animal sounds while the family next door has an argument, cars drive by from various directions, a baseball game is going on down the street, and a helicopter flies over. This scenario contains much more auditory information than the typical array of simultaneous auditory alarm signals from a set of medical devices but does not confuse the average person. It follows that medical alarm–related confusion may be largely an artifact of the artificiality of auditory alarm signals—of the fact that the human perceptual system did not evolve to perceive such abstract sounds.
TAF-ER562X_WINGR-09-0401-C010.indd 401
11/8/10 11:55:25 PM
402
Handbook of Human Factors in Medical Device Design
Edworthy and Stanton (1995) found that representational auditory alarm signals (e.g., a heartbeat for ECG-related alarm signals, bubbles for a syringe pump, or a nursery chime for an infant warmer) were recognized the best by novices rather than experienced nurses who had already learned the meaning of conventional auditory alarm signals. Also, Belz, Robinson, and Casali (1999) reported that “earcons” (including screeching tires or honking horns for time to collision with drivers) can be effective. Ultimately, meaningful auditory alarm signals have great potential to reduce learning time, improve memorability, reduce confusion, increase differentiation between alarms, and increase the information content of displays (see Stanton and Edworthy, 1999). An example of a naturally meaningful auditory alarm signal is the sound of an approaching automobile. The person hearing it knows intuitively not only that there is a problem, but also what to do about it without the need for any additional information. However as discussed by Edworthy (1998a), although abstract sounds are harder to learn and remember, the artificiality of conventional auditory alarm signals makes them stand out from their backgrounds. Abstract auditory alarm signals may also be easier to understand if they are given names that, in effect, make them more meaningful (Edworthy and Meredith, 1997).
10.1.3 SPEECH-BASED (OR VERBAL) AUDITORY ALARM SIGNALS One type of meaningful auditory alarm signal is a verbal annunciator that states the problem and/or gives instructions on how to correct it. Although verbal alarm signals can be effective if used selectively (particularly for critical conditions; see Hakkinen and Williges, 1984), they have a number of potential problems, as summarized in Table 10.1 (see Block, 1994; Stanton and Baber, 1997).
10.2 SPECIAL CONSIDERATIONS One of the most important facts that any alarm designer should know is that many medical environments contain a high frequency of false alarms. Manufacturers appear to have a strong tendency to bias their alarm systems in favor of “false positives” rather than “false negatives.” This forces medical professionals to deal with many false-alarm conditions. For example, Kestin, Miller, and Moore (1986) found that only 3% of the alarm conditions TABLE 10.1 Potential Problems with Verbal Spoken Auditory Alarm Signals • • • • • • • • •
Patients and other care providers may overhear them to ill effect. The user may speak a different language. They can compete with conversation. Multiple verbal alarm signals can be very distracting and confusing. They can elicit an emotional response that interferes with optimal performance. They are easily masked by ambient sounds. The wording can be too specific or concrete. Speech is generally perceived more slowly than text. Their compelling nature may lead to overreaction or inappropriate response to false alarm.
TAF-ER562X_WINGR-09-0401-C010.indd 402
11/8/10 11:55:25 PM
Alarm Design
403
annunciated during various types of pediatric surgery related to conditions that posed a risk to patients. O’Carroll (1986) found that only eight (<1%) of 1,455 alarm conditions annunciated in an intensive therapy unit in the United Kingdom were for “potentially serious problems.” As Stanton and Edworthy (1999) point out, false-alarm conditions can engender confusion and mistrust. Weinger and Smith (1993) note that false alarms can undermine performance by: • • • •
Drawing attention to unimportant events or causing inappropriate actions Increasing distraction and stress Inducing people to ignore alarm signals for important alarm conditions Inducing people to disable part of the alarm system’s ability to annunciate on detection of potentially dangerous conditions
Indeed, Block, Nuutinen, and Ballast (1999) reported that 70% of anesthesiologists sometimes disabled auditory alarm signals, most commonly because of false-alarm conditions. As Sorkin, Kantowitz, and Kantowitz (1988) discuss, the signals used to annunciate an alarm condition should be evaluated not just in terms of their effectiveness but also as to whether they undermine performance on other tasks. In fact, Gilson, Mouloua, Graft, and McDonald (2001) provide evidence that annunciation of auditory alarm signals for falsealarm conditions not only undermines performance on a given device but also undermines performance on adjacent devices. It follows that the ideal alarm system not only activates whenever the relevant condition occurs but also never activates when the relevant condition does not occur. To paraphrase Block, et al. (1999), the ideal alarm system not only has perfect sensitivity but also has perfect specificity. Manufacturers appear to have paid more attention to the former issue than the latter. This bias toward false positives is understandable because a manufacturer is much more likely to be held liable for an alarm system that fails to annunciate a dangerous condition than for any problems caused by false alarms. However, this causes medical professionals to be annoyed by sound pollution, distracted from important tasks, and to sometimes ignore important alarms. Therefore, alarm system designers should make appreciable effort to minimize false alarms.
10.3 DESIGN GUIDELINES The following guidance is sufficiently general to apply to alarm systems in devices ranging from simple home health care devices to the complex array of alarm systems in a cardiac intensive care unit. Thus, any particular device’s alarm system will have specific requirements that are beyond the scope of this chapter. Alarm system designers must have a thorough knowledge of the relevant clinical issues associated with the alarm system that they are designing and with the users whose behaviors the alarm system is designed to affect, as discussed in the next section. What follows are step-by-step guidelines for developing medical alarm systems.
10.3.1 STEP 1: GATHERING INFORMATION There is no substitute for direct, real-world observation and interviews (ideally documented via video) to understand what actually happens in a medical device’s potential use
TAF-ER562X_WINGR-09-0401-C010.indd 403
11/8/10 11:55:25 PM
404
Handbook of Human Factors in Medical Device Design
environment (e.g., hospital, office, or home) as opposed to what is “believed” to happen. It is also crucial to understand device users, what they know and do not know, their tendencies and biases, and so on. Although a discussion of field research is beyond the scope of this chapter (since such research is relevant to the development of medical devices in general, not just for device alarm systems), we urge developers of medical alarm systems to become familiar with field research methods (see Wiklund and Wilcox, 2005, and other texts). The other important category of information pertains to standards and related information about alarms. Particularly relevant are the general alarm standards, the alarm-related collateral standard to the IEC standard for medical electrical equipment (IEC 606011-8:2004), and the Japanese standard for medical alarm systems (Japanese Industrial Standard [JIS], 1991). Another good reference is Neville Stanton’s book Human Factors in Alarm Design (Stanton, 1994). Guideline 10.1: Consult Applicable Design Standards Alarm system designers should consult national and international standards that apply to their alarm system.
10.3.2 STEP 2: DETERMINING THE CONDITIONS THAT REQUIRE ALARM SIGNALS As discussed above, an alarm signal is designed to indicate the presence of a dangerous condition that requires human intervention. IEC 60601-1-8:2004 points out that there are two basic categories of alarm conditions: 1. Conditions that stem from patient characteristics (“physiological alarm conditions”) 2. Conditions that arise from equipment characteristics (“technical alarm conditions”) A third, hybrid category of alarm conditions exists if the coupling between the patient and the medical device fails, even though both the medical device and the patient are satisfactory when considered separately. Guideline 10.2: Determine Device Parameters and Attributes Designers should use multiple methods to rigorously determine what device parameters should be included in the alarm system and the attributes of the alarm signal (i.e., criticality, onset, threshold, and so on) associated with each alarmed parameter.
The candidate conditions to be indicated by alarm signals are those that pose a danger, particularly to the patient but also to others (or to the medical device itself, which, in turn, can eventually pose a danger to people). Developing an “alarm condition inventory” is an extremely important and potentially complicated step. Two key considerations in identifying alarm conditions are the severity of the risk to the patient and the speed with which a response is required (see IEC 60601-1-8:2004; Weinger and Smith, 1993). Methods for identifying alarm conditions include gathering existing incident reports and reviewing the relevant medical literature, conducting research in the environment of use, and generating scenarios of use. Traditional hazard/risk analysis methods, such as fault tree analysis and failure modes and effects analysis (see Kirwan and Ainsworth, 1992), can also provide input to the creation of the alarm condition inventory.
TAF-ER562X_WINGR-09-0401-C010.indd 404
11/8/10 11:55:25 PM
Alarm Design
405
One other consideration that becomes more relevant as technology makes it possible to distance the medical professional physically from the patient is whether the person who is notified of an alarm condition is in a position to reasonably respond to the alarm notification. Guideline 10.3: Appropriate Responder The alarm signal should be directed at a person who is able to respond appropriately.
10.3.3 STEP 3: CREATING SIGNAL-PROCESSING ALGORITHMS After the alarm condition inventory is created, a set of particular parameters should be identified that will initiate alarm signals. These parameters can be chosen in several ways, including relying on a consensus of expert users and/or consulting clinical data (see Beneken and Van der Aa, 1989; Weinger and Smith, 1993). One issue that must be addressed is the setting of limits for what conditions will initiate an alarm signal. Alarm limits can be linked to parameters in many ways, including rate of rise of a parameter, duration above a defined threshold, and “area under the curve” (Weinger and Smith, 1993). The most important consideration is to select a parameter (or set of parameters) that, when alarm limits are applied, results in true positive detection of an actual incipient hazard. However, it is important to detect the alarm condition (and to initiate alarm signals) early enough to head off a crisis (Weinger and Smith, in press). Thus, the alarm system designer has the challenge of providing enough lead time without burdening the user with premature or unnecessary alarm signals. The ample evidence of false alarms suggests that many alarm systems are tied to parameters in ways that do not adequately reflect clinically significant events. This puts the burden on the user to weigh the notification of an alarm condition that he or she receives against other available data to decide what to do. However, the very nature of alarm signals is to increase the user’s level of (sympathetic nervous system) arousal, which, in turn, undermines his or her ability to weigh alternatives and make optimal decisions. In high-arousal states, people tend to revert to instinctive or habitual behavior and lose the ability to attend to peripheral items and events (see Bacon, 1974; Cratty, 1973). It follows, then, that a greater burden of “intelligence” should be placed on alarm systems. Guideline 10.4: Alarm Limits Where possible, alarm systems should be designed to base alarm limits on inputs from multiple parameters or to integrate data from multiple devices.
Such “smart” alarm systems have been advocated/proposed for many years (see Block, 1994; Giles, Edworthy, Brown, and Davies, 1998; Nasir, 1998). Two fundamentally different approaches to smart alarm systems are rule-based algorithms (see Watt, Navabi, Mylrea, and Hameroff, 1989) and neural networks (see Orr and Westenskow, 1993). 10.3.3.1 Adjustable Alarm Limits User adjustability of alarm limits is a controversial topic. The ideal alarm system would not require limit adjustments because the “factory” or “default” settings would be perfect in every situation. Because truly intelligent alarm systems are not yet available, most devices allow users to adjust alarm limits on the basis of their clinical experience and judgment.
TAF-ER562X_WINGR-09-0401-C010.indd 405
11/8/10 11:55:25 PM
406
Handbook of Human Factors in Medical Device Design
However, there is some evidence that clinical professionals do not often set alarm limits properly (see Deller, Konrad, Kilian, and Schuhle, 1992), and it is doubtful that lay users of home health care devices would do so. Thus, manufacturers should carefully consider whether alarm limit adjustability should be provided at all, particularly where the alarm systems are related to very hazardous conditions, such as cardiac standstill (asystole). If alarm limits are adjustable, it is important to impose constraints on the adjustments so that alarm limits make sense and are not dangerous. The following requirements should be applied to decrease the likelihood that users will accidentally set the wrong limits or be mistaken about what the limits are. Guideline 10.5: Adjustable Alarm Limits Adjustable alarm limits should be storable, but the device should clearly indicate when the limits have been altered from factory settings (Block, 1994; IEC, 2004).
Guideline 10.6: Multiple Adjusted Limits If more than one set of user-adjusted limits are stored in a device, then the user should be required to make an affirmative choice of alternatives.
Guideline 10.7: Return to Default Settings When limits are altered from factory settings, it should be easy to return them to the default settings.
10.3.3.2 Disabling Alarm Signals or Portions of Alarm Systems An extreme version of limit adjustment is the disabling of an alarm signal or part of an alarm system altogether, either temporarily or permanently. Although allowing alarm disabling poses an obvious hazard, it is often an essential feature of medical devices (particularly in-hospital devices and for noncritical alarm conditions) for the reasons listed below and will remain so as long as medical devices are prone to annunciate false-alarm conditions. Circumstances in which alarm signals may need to be disabled include the following: • Prior to putting a device or device component in service, when an alarm signal would simply be an unnecessary distraction • When a physiological measurement associated with a given parameter is no longer needed • When the user is always in close proximity to the device and has already been alerted (see Stanton and Edworthy, 1999) IEC 60601-1-8:2004 mandates use of specific symbols and recommends specific Englishlanguage terms for four different ways that alarms signals can be disabled: • AUDIO PAUSE—temporarily silencing an auditory alarm signal without affecting other alarm signals • AUDIO OFF—permanently silencing an auditory alarm signal without affecting other alarm signals
TAF-ER562X_WINGR-09-0401-C010.indd 406
11/8/10 11:55:25 PM
Alarm Design
407
• ALARM PAUSE—temporarily disabling all or part of an alarm system’s ability to generate any alarm signals • ALARM OFF—permanently disabling all or part of an alarm system’s ability to generate any alarm signals The following should be considered regarding disabling of alarm signals: Guideline 10.8: Temporary or Permanent Disabling Temporary or permanent disabling should be clearly indicated (American Society for Testing and Materials [ASTM], 1999; Block, 1994; Weinger and Smith, 1993).
Guideline 10.9: Temporary Disabling “Time Out” Temporary disabling should “time out” after a designated period of time (Weinger and Smith, in press). Given the wide variation in conditions relevant to different alarm systems, it is not possible to specify a disabling time. However, JIS (1991) does specify that the disabling period should be less than 10 minutes, which should be considered when designing products for the Japanese market. A common practice in critical care alarms is a maximum time period of 2 to 3 minutes. The choice of maximum time should be based on the variability and criticality of the parameter or condition being in alarm.
Guideline 10.10: Visual Indicators and Periodic Reminders An alarm system should provide a continuous visual indication and may provide a periodic reminder signal of the disabled state.
Guideline 10.11: Confirmation of Permanent Disabling Permanent disabling should require confirmatory actions (Block, 1994).
Guideline 10.12: Disabling for Auditory Signals Only Disabling should apply only to auditory alarm signals, not visual ones (Block, 1994; Weinger and Smith, 1993).
Guideline 10.13: Condition-Specific Disabling Disabling should apply only to specific conditions that, if changed, will reenable the alarm signal.
Guideline 10.14: Reset on Reboot Alarm limits and active status should reset on the next rebooting of the device (Block, 1994).
10.3.3.3 Other Alarm Condition Considerations There is a conceptual difference between alarm systems for devices that are closely monitored and those that are not (Seagull, Xiao, Mackenzie, and Wickens, 2000). Seagull et al. (2000) argue that, particularly for closely monitored devices such as those used by anesthesiologists in the operating room, the device designer should think more in terms of informing rather than alerting the user in order to avoid sympathetic nervous system arousal and unnecessary distraction.
TAF-ER562X_WINGR-09-0401-C010.indd 407
11/8/10 11:55:25 PM
408
Handbook of Human Factors in Medical Device Design
Another issue is how to handle the situation when an alarm condition initiates alarm signals but quickly returns to normal. The question is whether the alarm signals should be latching, that is, continue to sound until the user addresses them, even if the condition has already returned to normal. Perhaps the right approach is to ensure that the auditory alarm signal sounds for a minimum period of time and that a clear indication of the alarm condition is provided, either as a latching visual alarm signal or in an alarm log (see Block, 1994). With respect to sounding of an auditory alarm signal, IEC 60601-1-8:2004 requires this minimum period to be one full burst of the medium-priority auditory alarm signal or one half burst of the high-priority auditory alarm signal. (See Section 10.3.7, Creating Auditory Alarm Signals for more information). A third consideration is that, in emergency situations, attention-getting alarm signals for less important conditions may interfere with alarm signals associated with more critical conditions. Guideline 10.15: Less critical Alarm Conditions Auditory alarm signals for less critical alarm conditions should be disabled or otherwise “understated” when they compete for the user’s attention with more critical alarm signals.
Guideline 10.16: Alarm System Test Especially for devices designed for professional use, there should be a way to test that the alarm system is working properly. A common approach is to sound an audible tone and flash a visual alarm signal as part of a device’s power-on sequence, although this perhaps incorrectly assumes that a knowledgeable user will be present and notice if the signals fail to enunciate.
Guideline 10.17: Alarm System Failure If the alarm system itself fails, this should be clearly indicated (IEC, 2004).
10.3.4 STEP 4: DETERMINING THE INFORMATION THAT SHOULD BE COMMUNICATED ABOUT EACH ALARM CONDITION Once the “alarm condition inventory” is created, it is necessary to determine what information needs to be communicated about each of the conditions. The key information that the alarm signals have to convey is the source of the problem, the nature of the problem, what needs to be done (if possible), and the urgency. As Kerr (1985) points out, urgency is affected by the rate of change of the relevant parameters, the condition of the patient, the response time of the relevant devices, and the likely response time of the user. There is a consensus among the various standards (e.g., ASTM, 1999; IEC, 2004; JIS, 1991) for a three-level hierarchy of urgency: 1. High—requiring immediate action (i.e., emergency) 2. Medium—requiring prompt action (i.e., warning) 3. Low—requiring awareness (i.e., caution) 10.3.4.1 Likelihood Although not specifically addressed in the various standards, likelihood is another type of information that could be conveyed by alarm signals. Likelihood is certainly relevant if false-alarm signals are common. The user will inevitably use judgment of the likelihood
TAF-ER562X_WINGR-09-0401-C010.indd 408
11/8/10 11:55:25 PM
Alarm Design
409
of a dangerous condition to determine how best to respond when an alarm condition is annunciated. In fact, there is evidence that the reliability of alarm signals influences the probability of a user response (Bliss, Gilson, and Deaton, 1985). By coding alarm signals into four levels of likelihood—none, possible, probable, or likely—by color or by the addition of a verbal signal, Sorkin et al. (1988) showed improved performance in terms of appropriate response to alarm signals. The presentation of such probability coding improved alarm response when the relevant tasks were difficult (but not when the tasks were easy), and the additional complexity (i.e., inclusion of this information) did not add to overall task complexity or interfere with performance on other non–alarmrelated tasks. They argue that conveying likelihood informs good decision making and improves allocation of “attentional resources.” Guideline 10.18: Inclusion of Likelihood or Reliability Designers should consider including likelihood information in alarm signals if it is possible to do so.
10.3.5 STEP 5: ALLOCATING SIGNALING MODALITIES TO ALARM CONDITIONS As discussed in the preceding section, alarm system users need to know the location of the problem, the cause of the problem, and what action is necessary to address it. Annunciation of an alarm condition must get the user’s attention before the system can provide information about the problem and how to fix it. Other requirements are shown in Table 10.2 (see AAMI, 2010; Griffith and Raciot, 1992; IEC, 2004; International Organization for Standardization [ISO, 1996]; Weinger and Smith, 1993; and Woodson, Tillman, and Tillman, 1992). This variety of requirements has led to several calls over the years for integrated systems that take input from multiple devices and organize them into one coherent, prioritized, and logical alarm system (e.g., Block, 1994; Weinger and Smith, 1993). There is no doubt that such integrated alarm systems would be infinitely preferable to the complexity, even chaos, that users face when using a variety of alarm systems from multiple devices designed by different manufacturers at different times. For the foreseeable future, however, designers must incorporate alarm systems into their devices. TABLE 10.2 Universal Alarm Signal Requirements • • • • • •
Do not startle or unnecessarily arouse, distract, or annoy the intended recipient. Capture the user’s attention long enough to fix the problem. Avoid interfering with speech or with other alarm signals. Minimize interference due to alarm signals that relate to less important alarm conditions. The same alarm conditions should always trigger the same alarm signal. Ensure that the alarm signals are clear, understandable, and intuitive under all foreseeable environmental conditions. • Provide alarm signals that are likely to produce a fast and accurate response. • Provide feedback for correct responses. • Provide alarms signals that, to the extent possible, can be detected by users with perceptual disabilities (particularly for home health care devices).
TAF-ER562X_WINGR-09-0401-C010.indd 409
11/8/10 11:55:25 PM
410
Handbook of Human Factors in Medical Device Design
Guideline 10.19: Allocating Signaling Modalities Designers should consider the potential interactions with alarm signals from other devices routinely found in the intended use environment.
Which signaling modalities should be allocated to a specific alarm condition? Available signaling modalities include auditory, visual, olfactory, vibratory, and other types of tactile/ haptic stimulation. However, the available literature provides guidance primarily about the use of nonspeech auditory as well as visual signaling modalities. Guidance related to use of speech usually recommends limited use under very specific conditions. One excellent example is the use of speech prompts by automatic external defibrillators. Even with this, however, there is a risk of speech prompts, when inappropriate, biasing users to follow device prompts against their better judgment. The literature provides limited guidance for the use of olfactory, vibratory, or other types of tactile/haptic stimulation as signaling modalities. However, there are medical device examples of vibratory alarm signaling. The most common example involves patient monitoring devices that send alarm information to commercial pagers, cellular phones, or personal digital assistants as secondary (not primary) alarm notification signals. Many insulin pumps also employ vibratory alerts. The typical and generally logical strategy for selecting alarm signaling modalities is to rely primarily on auditory alarm signals for getting the user’s attention and on visual alarm signals for conveying information regarding a specific alarm condition. Thus, the primary intent of the auditory alarm signal is alarm notification (i.e., a condition exists) and possibly to indicate its urgency and source. The visual alarm typically provides information about the cause of the alarm and possibly corrective action. This means that most alarm conditions have both visual and auditory alarm signals. AAMI (2007), IEC (2004), ISO (1996), and Weinger and Smith (1993) all advocate this for all but “low-priority” alarm conditions. Low-priority alarm conditions are required to have a visual alarm signal but do not require an auditory alarm signal. Visual alarm signals may capture the user’s attention and also convey specific information regarding an alarm condition. The classic example is wiring a doorbell or a teletypewriter to the lights in the home of someone who is deaf. In contrast, auditory alarm signals may be limited in their ability to convey specific information related to an alarm condition unless speech-based signals are used. In general, other signaling modalities (i.e., neither auditory nor visual) have yet to find common health care applications. However, with suitable user training, non–speech-based auditory, olfactory, vibratory, and tactile/haptic signaling modalities may all be viable methods of providing specific detailed information regarding an alarm condition. For example, many vision-impaired people get specific information either via auditory signals or tactilely by reading Braille. When selecting alarm-signaling modalities, care must be taken to consider whether the user of the medical device may have a sensory limitation, such as impaired hearing or vision. In the case of a hearing-impaired user, an auditory alarm signal, if provided, cannot capture his or her attention, so designers must provide other signaling modalities to fulfill that role. An attention-getting auditory alarm signal may still be valuable in this situation because it may attract the attention of someone else (e.g., a family member) nearby who has normal hearing. Table 10.3 summarizes the advantages and disadvantages of visual and auditory alarm signals, respectively. It draws from a number of sources, including Griffith and Raciot, (1992), Stanton (1994), Stanton and Edworthy (1999), and Weinger and Smith (1993).
TAF-ER562X_WINGR-09-0401-C010.indd 410
11/8/10 11:55:25 PM
Alarm Design
411
TABLE 10.3 Comparison of Auditory and Visual Alarm Signals Conditions Conducive to Auditory Alarm Signals
Conditions Conducive to Visual Alarm Signals
When there is a visually cluttered environment When the operator moves around When the display is out of the user’s visual field When a rapid user response is desired When information does not involve a temporal order When information will not be referred to later When information is continuously changing When the message is simple When information comes from a single source
When there is a noisy environment When the user stays in one place When the user constantly monitors a visual display When there can be a delay prior to a user response When understanding of temporal order is important When information will be referred to later When information is constant When the message is complex When information comes from multiple sources
10.3.6 STEP 6: CREATING A SIMULATED USE ENVIRONMENT When designing alarm signals for an alarm system, it is crucial to understand the environment in which those alarm signals will be used. Guideline 10.20: Realistic Testing Environment Testing of alarms should be done in an environment that realistically re-creates the essential attributes of the intended use environment.
10.3.6.1 Visual Environment It is useful to measure light levels and to record (via video or still photography) the usual landscape in typical use environments so that they can be re-created for testing purposes. Guideline 10.21: Visual Testing Environment During the evaluation of visual alarm signals, the visual aspects of the intended use environment should be reproduced as closely as possible.
10.3.6.2 Auditory Environment High-quality audio recordings are useful for determining the characteristics of the auditory environment where a device will be used. Such environments are very different, in terms of their acoustic characteristics, from the typical office or lab where an alarm system might be developed (see Figure 10.4). It is important to obtain a good representative sample of use environments. Guideline 10.22: Auditory Testing Environment The testing environments should include the sounds that will typically accompany the device, such as music, people talking on the phone, other alarm signals, and so on.
Audio recording requires a good nondirectional microphone and a good recorder. The audio equipment should be of sufficient quality to ensure that the sound is as realistic as possible. This means having a microphone with a frequency range of at least
TAF-ER562X_WINGR-09-0401-C010.indd 411
11/8/10 11:55:25 PM
412
Handbook of Human Factors in Medical Device Design
FIGURE 10.4 The operating room is an example of a high-noise environment that can be quite different than other medical care environments.
20 to 20,000 Hz and a recorder with a frequency range of at least 20 to 40,000 Hz (to ensure good frequency response up to 20,000 Hz). Both the microphone and the recorder should have flat frequency-response curves. These recordings, once obtained, should be analyzed with a spectrum analyzer (see Figure 10.5) to determine what frequencies are represented. Everything else being equal, a given sound can be heard better if it is not “masked” by other sounds with similar frequencies. Guideline 10.23: Avoid Frequencies Common in Use Environment Auditory alarm signals should not share frequencies with other sounds commonly found in the intended use environment.
To facilitate realistic testing of potential alarm tones, an accurate sound environment should be created. This too requires good equipment (see Figure 10.6), including one computer to play digitally recorded ambient sound (through large speakers), a second computer to play new alarm signals (through small speakers in the middle), and a sound meter. A third computer might be used to mimic the device user interface (e.g., response times or alarm signal identification). Again, it is important to have an amplification and speaker system that supports a 20- to 20,000-Hz range with a flat
FIGURE 10.5
Output of a spectrum analyzer.
TAF-ER562X_WINGR-09-0401-C010.indd 412
11/8/10 11:55:25 PM
Alarm Design
FIGURE 10.6
413
Setup for reproducing sound and testing new auditory alarm signals.
frequency-response curve. Because sound degradation is cumulative, high-quality recording and playback equipment are needed. For example, if each of the components (microphone, recorder, amplifier, and speakers) degrades the sound 10%, the end result will be a 34% degradation. Testing should occur in a room as similar as possible to the room where the sounds were recorded, particularly with regard to its size. Also, a sound meter should be used to make sure that the sound is played back at the correct volume. If there is wide variation in the acoustic properties of typical use environments, then this range of variation should be represented in the simulation tests. Guideline 10.24: Simulated Use Environment Auditory alarm signals should be tested in the simulated use environment, using speakers that are as similar as possible to the speakers to be used in the actual device (see Figure 10.7).
For optimum results, the mounting, orientation and other physical characteristics of the speakers in the simulated use environment should be as similar as possible to the speakers used in the actual device. Figure 10.7 shows two alternative prototype speakers for testing. The clear boxes simulate the space behind the speakers that the final device will provide.
FIGURE 10.7
Two prototype alarm signal–producing speakers undergoing usability testing.
TAF-ER562X_WINGR-09-0401-C010.indd 413
11/8/10 11:55:25 PM
414
Handbook of Human Factors in Medical Device Design
10.3.7 STEP 7: CREATING AUDITORY ALARM SIGNALS 10.3.7.1 Constructing Abstract Auditory Alarm Signals Patterson is generally credited for proposing a systematic approach to the design of auditory alarm signals—one that is in wide use today (see Patterson, 1982, 1989, 1990). Patterson, Edworthy, Shailer, Lower, and Wheeler (1986) also proposed a standardization scheme for the operating room in which there is a specific auditory alarm signal for each type of device (ventilators, infusion devices, and so on). Although such a scheme seems logical (although Weinger and Smith (1993) do mention some potential drawbacks), for standardization to be effective, it must be widely adopted. To date, the Patterson et al. scheme remains a concept that has been only rarely, if ever, adopted. The guidance provided by the standards of ASTM (1999), IEC (2004), and JIS (1991) tend to be consistent with what one might call the “Patterson” approach. This involves (see also Stanton, 1994) designing pulses, or specific sounds, and organizing those pulses into bursts, or groups of pulses. 10.3.7.2 Design of Alarm Signal Pulses Designing an alarm signal pulse requires specifying its frequency components, its timing, and its loudness. For determining the frequency of alarm signal pulses, the designer, using sound-waveform-generation software, should work directly with sine waves rather than the complex waveforms (e.g., triangle waves) provided, as an alternative, by some software systems. These alternative waveforms are made up of combinations of sine waves, so they make it more difficult for the alarm signal designer to understand the nature of a given auditory signal. The approach proposed by AAMI (1973), ASTM (1999), and Stanton and Edworthy (1999b), among other sources, is to create an alarm signal that consists of multiple frequencies, typically four or more. There are two ways that different frequencies combine to make a given sound: 1. A tone with a specific frequency, a fundamental, can be combined with additional frequencies, or harmonics, that have lower amplitudes relative to the harmonic. The “note” that is heard is the fundamental. The pattern of harmonics gives the sound its timbre. The timbre is what differentiates the same note played on different musical instruments, say, the middle C of a piano, vs. a clarinet, vs. a guitar. The various standards generally envision auditory-alarm signals that consist of a single fundamental with multiple harmonics. 2. Tones, with or without harmonics, can be combined at roughly the same amplitude to create chords. A chord is typically a combination of 3 or more such fundamentals. Although not typically envisioned by alarm standards, using multiple fundamentals to create chords can provide the alarm designer with additional options. Some reasons for creating pulses that have multiple frequencies include the following: • Inclusion of harmonics tends to make a sound less harsh-sounding. • The addition of harmonics can give a sound a distinctive timbre that can be used to differentiate it from other sounds.
TAF-ER562X_WINGR-09-0401-C010.indd 414
11/8/10 11:55:26 PM
Alarm Design
415
• A user with frequency-specific hearing loss will be less likely to have a detection problem when the sound is a chord made up of frequencies that are reasonably far apart. • A chord is less likely to be masked by ambient sound since such masking is relatively frequency-specific. Guideline 10.25: Constructing Alarm Frequencies Designers should be familiar with and apply basic musical principles in constructing the frequencies of alarm signals since the human ear is attuned to music.
It follows then, that an auditory-alarm signal should generally contain a fundamental with multiple harmonics, but there are advantages to using multiple fundamentals to make a chord. Sounds that are “semimusical” seem to work best, that is, chords that are somewhat musical but that are not commonly used in music. A sound that is too musical may be confused with music and may not attract attention, particularly in an environment, such as an operating room, where music is often played. On the other hand, if the sounds are not somewhat musical or are too dissonant, they can be annoying. One source for useful chords is early 20th-century music by composers such as Stravinsky or Bartók. The next step is to adjust the key and the octave to make the frequencies as different as possible from the ambient sound. At this stage, it is good to generate various candidate sounds. By listening to sounds in the re-created ambient sound environment, a “short list” of candidates can be identified that are appropriately conspicuous, distinctive, and reasonably pleasant. Ideally, the sounds should intuitively relate to what they indicate; however, this can be difficult to achieve. As with the ambient sounds, the candidate sounds should ideally be generated by the actual system (speakers and amplifier) that will be incorporated into the final device. Table 10.4 summarizes frequency guidance from various sources. We recommend fundamental frequencies in the range of 150 to 1,000 Hz with harmonics in the range of 300 to 4,000 Hz. The Japanese standard (JIS, 1991) is clearly the outlier since it does not allow for harmonics, and it specifies a much narrower range of frequencies. Thus, it may be difficult to create an alarm signal that conforms to the specific acoustic guidelines of the U.S. and European standards on the one hand and the Japanese standard on the other. TABLE 10.4 Pulse-Frequency Recommendations from Various References Reference IEC 60601-1-8:2004 ASTM F 1463-93:1999) Patterson (1982) Sanders and McCormick (1987) JIS T 1031-1991
TAF-ER562X_WINGR-09-0401-C010.indd 415
Frequency Range for Fundamental (Hz)
Frequency Range for Harmonics (Hz)
150–1,000 150–1,000 150–1,000 200–5,000 1,000–2,000 (emergency) 500–600 (warning) 300–400 (caution)
300–4,000 300–4,000 500–5,000 200–5,000 — — —
11/8/10 11:55:26 PM
416
Handbook of Human Factors in Medical Device Design
In addition, lower frequencies are better for covering distance (hence fog horns), but higher frequencies allow better localization. Thus, the use of multiple frequencies in a pulse may facilitate both good localization (via higher frequencies) and good spread (via the lower frequencies). Guideline 10.26: Use Frequencies below 1000 Hz Frequencies below 1,000 Hz should be included in the selection of fundamentals when the signal has to travel a significant distance (Senders and McCormick, 1987), and fundamentals below 500 Hz should be included when the sound has to travel around an obstacle.
As discussed above, the pattern of harmonics creates that timbre, or the distinctive sound akin to that of a particular musical instrument. Guideline 10.27: Timbre Held Constant Timbre should typically be held constant for equivalent signals, and timbre can be used to differentiate between different signals (see Gerth, 1993).
People with hearing loss tend to lose higher frequencies first, so it makes sense to avoid relying on particularly high frequencies. Timing relates to the length of the pulse, the length of the “off” period, and the rise time, (i.e., the time it takes the pulse to attain its full loudness). An unduly rapid rise time will startle the user. To prevent this, Patterson (1989) suggests that an auditory alarm signal begin relatively quietly and gradually increase in loudness. Table 10.5 summarizes some general recommendations regarding pulse timing. Since the literature and existing standards vary, the designer has appreciable freedom in determining pulse timing. In fact, IEC (2004) contains an “escape clause” stating that its requirements can be violated if the candidate alarm system is validated by adequate usability testing. This is an important waiver that also applies to other characteristics of alarm signals. For loudness (the psychological variable that correlates with the intensity, or amplitude, of the signal—see Chapter 2, “Basic Human Abilities”), the ideal approach is to use algorithms that consider the intensity and frequency of ambient sounds (see Laroche, Tran Quoc, Hetu, and McDuff, 1991; Lazarus and Hodge, 1986; Momtahan, et al., 1993).
TABLE 10.5 Pulse-Timing Recommendations from Various Sources Source
Pulse On (Seconds)
IEC 60601-1-8:2004
0.075–0.20 (high) 0.125–0.25 (medium and low) ASTM F 1463-93:1999 0.10–1 Patterson (1982) 0.10–0.20 JIS T 1031-1991
TAF-ER562X_WINGR-09-0401-C010.indd 416
0.25 (emergency) 0.75 (warning) 1 (caution)
Pulse Off (Seconds)
Rise Time (Seconds)
0.05–0.125 (high) 0.125–0.25 (medium and low) — ≤0.15 (urgent) >0.30 (nonurgent) 0.25 (emergency) 0.50 (warning) ≥4
10%–20% of pulse duration (on) ≥0.015 0.02–0.03 — — —
11/8/10 11:55:26 PM
Alarm Design
417
Edworthy (1998b) proposes that an auditory alarm signal be 15 to 25 dB above the “masked threshold,” the determination of which requires one of the algorithms mentioned above. Both Patterson (1982) and JIS (1991) advocate an alarm loudness of 15 to 25 dB above the ambient noise. Guideline 10.28: Loudness to Assure Alarm Perception The auditory alarm signal should be perceivable under foreseeable use conditions.
Because masking is strongly affected by frequency, a specific intensity or a given number of decibels above ambient sound cannot be uniformly specified. Thus, we present the above recommendations for the reader’s information but do not endorse them. 10.3.7.3 Design of Bursts The designed pulses should be arranged into bursts (i.e., repetitive groups of pulses). Patterns of bursts can be created by varying the order, timing, intensity, fundamental frequency, or timbre of the pulses that make up a burst. Whereas the pulse is akin to a chord in music, the burst is akin to a melody or rhythm. Composers such as Bartók and Stravinsky are a good source of melodies that are somewhat but not overly musical. IEC (2004) provides specific melodies that might be used as well as some general timing guidance for bursts. It also specifies that individual pulses within a burst should not differ in intensity by more than 10 dB. As with pulse design, the designer should test various burst candidates in a simulated use environment. An important fact relevant to bursts is that temporal patterns have a disproportionate influence on which sounds are remembered as similar (Edworthy, 1998a; Patterson, 1982). 10.3.7.4 Communicating Urgency As described by Edworthy (1994), auditory alarm signals can be designed to communicate urgency in an intuitive and consistent way. Table 10.6 summarizes findings from a number of studies. Algorithms are available that specify a consistent increase in perceived urgency as a function of given acoustic changes in auditory alarm signals. In general, a signal’s temporal characteristics are more effective than its sound-quality characteristics at increasing the perception of urgency. For example, a 50% increase in perceived urgency can be achieved by a 1.3-fold increase in speed of pulses, a 2.2-fold increase in the number of repeating TABLE 10.6 Auditory Parameters That Will Increase Perceived Urgency • • • • • • • •
Higher fundamental frequency Greater dissonance Shorter interpulse intervals Decreasing interpulse intervals within a burst More regular temporal patterning of bursts More repeating elements Larger differences between the fundamental frequencies of pulses within a burst Random pitch variation between bursts (as opposed to up or down)
TAF-ER562X_WINGR-09-0401-C010.indd 417
11/8/10 11:55:26 PM
418
Handbook of Human Factors in Medical Device Design
units, a 2.8-fold increase in fundamental frequency, or a 28.5-fold increase in dissonance (see Edworthy, 1994).
10.3.8 STEP 8: CREATING VISUAL ALARM SIGNALS As mentioned above, there is a general tendency for auditory alarm signals to be used to attract attention and for visual alarm signals to communicate information. However, visual alarm signals can complement the attention-getting function of auditory alarm signals. Doing so can address the needs of those with hearing impairments or people working in noisy environments. Visual alarm signals also tend to be more position-specific than auditory alarm signals. 10.3.8.1 Information-Providing Visual Alarm Signals (Information Displays) JIS (1991) divides visual alarm signals into “annunciators” and “display phenomena” that provide instructional information. Such visual display signals should clearly transmit meaning through the manipulation of text, figures, colors, and positions. These requirements are no different than those for any visual display (see Table 10.7 and Chapter 8, “Visual Displays”). IEC (2004) provides a textual convention for indicating urgency with three exclamation marks for high urgency, two for medium urgency, and one for low urgency. Adams and Edworthy (1995) suggest that alarm text can be perceived as more urgent by the following: • • • •
Increasing the font size (e.g., “warning”) Providing more space around the text (“w a r n i n g”) Increasing the width or borders Using red rather than black text
10.3.8.2 Attention-Getting Visual Alarm Signal Signals Attention-getting visual alarm signals are often point sources such as warning lights. However, other implementations are possible. Table 10.8 provides guidance for the color and timing of visual alarm signals. Because the alarm standards are similar, it is possible, if desired, to comply with all of them. Some standards also provide guidance for the TABLE 10.7 General Requirements for Information Displays • Legible at a distance of 1 m (AAMI, 1993; ASTM, 1999; IEC, 2004; JIS, 1991) • Accompanied by a specific auditory annunciator—if the same display is used for non–alarmrelated information (JIS, 1991) • Legible in bright (1,500-lux [AAMI, 1993; ASTM, 1999; JIS, 1991]) and dim (100-lux [AAMI, 1993; ASTM, 1999]) ambient environments • Synchronized with other alarm signals if they flash (JIS, 1991) • Alarm information differentiated from other displayed information by color, reverse video, or change in luminance or by inclusion in a box or the use of symbols (Block, 1994; JIS, 1991) • Display of only information essential to perform the intended task (AAMI, 1993) • Use of reverse video rather than flashing for readability (Block, 1994)
TAF-ER562X_WINGR-09-0401-C010.indd 418
11/8/10 11:55:26 PM
Alarm Design
419
TABLE 10.8 Guidance for Attention-Getting Visual Alarm Signals Color Source
High/Emergency
Medium/Warning
Low/Caution
IEC 60601-1-8:2004 ISO (1996) JIS T 1031-1991 ASTM F 1463-93:1999
Red Red Red Red
Yellow Yellow or yellow/red Yellow Yellow
Yellow or cyan — Yellow Yellow
Timing IEC 60601-1-8:2004 ISO (1996) JIS T 1031-1991 ASTM F-1463-93:1999
1.4–2.8 Hz, 20%–60% on 2–3 Hz, equal on/off Synchronized with auditory alarm 1.4–2.8 Hz
0.4–0.8 Hz, 20%–60% on — Synchronized with auditory alarm 0.4–0.8 Hz
Continuous — Continuous Continuous
brightness of visual alarm signals and require them to be visible at a viewing distance of at least 4 m (e.g., ASTM, 1999; IEC, 2004). Another common requirement is that they be visible from 30 degrees off the perpendicular viewing axis. 10.3.8.3 Other Considerations regarding Visual Alarm Signals The presentation of excessive visual information slows processing time. Athenes, Chatty, and Bustico (2000) describe research with air traffic controllers indicating that reaction time is faster with more global changes to visual displays (vs. localized), to more opaque images (vs. transparent), and to more steplike changes (vs. gradual). There is evidence that different forms of visual alarm signals are better at communicating different types of alarm conditions (Stanton and Stammers, 1998). Text may be better for time-based reasoning. “Mimic” formats (e.g., a picture of the facility showing the problem) may be best for communication of spatial location, while annunciator-type visual alarm signals (e.g., flashing displays) may be best for communicating spatial patterns. ISO (1996) makes a provision for “area-source” visual alarm signals—increasing the background luminance 5-fold for “warning” or 10-fold for “emergency.” It also suggests varying signals with different meanings by at least two of the four dimensions—color, location, relative position, or temporal pattern.
10.3.9 STEP 9: CREATING OTHER ALARM SIGNALS As discussed above, there is relatively little information available about tactile or other atypical alarm signals for medical applications. However, these may be appropriate and even preferred in some situations. Vibratory alarm signals, in particular, can be effectively used to get the user’s attention, as indicated by their use with cell phones.
10.3.10 STEP 10: TESTING PROTOTYPE ALARM SYSTEMS WITH POTENTIAL USERS For a detailed discussion of usability testing, see Chapter 6, “Testing and Evaluation.” Alarm systems should be tested initially with re-created ambient sounds and later in the real environment of use with the same rigor as any other critical device function.
TAF-ER562X_WINGR-09-0401-C010.indd 419
11/8/10 11:55:26 PM
420
Handbook of Human Factors in Medical Device Design
For usability testing of alarm systems, the key criteria for success include the following: • • • •
Attention-getting ability Distinctiveness Clear communication of the desired information (source, urgency, and meaning) Freedom from annoyance and aversion
In general, the most important thing to test is the desired user behavior—successful correction of the alarm condition. As with other aspects of testing of a device, two or more rounds of usability testing are typically required for reasonable assurance that an alarm system is adequately usable/successful. For devices designed for use by people with disabilities (often the case with home health care devices), these users must be included in the test population. Edworthy and Stanton (1995) have proposed several testing strategies that can improve alarm system design, including “appropriateness” rankings of alternative alarm signals, “confusion” tests that provide evidence of whether two alarm signals are adequately distinctive, “urgency” ratings, and recognition/matching tests.
10.3.11 STEP 11: REFINING ALARM SYSTEMS BASED ON TEST RESULTS The purpose of usability testing is to refine the design. Thus, after each round of testing, the alarm system should be refined (or redesigned if needed) in response to the findings. A detailed discussion of such refinements is beyond the scope of this chapter. However, the goal is to continue to refine the design until the alarm system meets the criteria for success discussed throughout this chapter.
10.4 CASE STUDY 10.4.1 AUDITORY ALARMS ON AN INTRAVENOUS INFUSION SYSTEM Medical device reports and customer complaint data reinforced the need for more effective visual and auditory alarms to alert intravenous (IV) pump users to device fault conditions such as air in line, occlusion in IV tubing, pending battery failure, IV bag near empty, or unsafe dosage rates for a particular drug in a specific clinical application. The development team also decided to adopt the recommendations of the international standard (IEC 606011-8:2004) for medical device auditory alarms to use unique melody patterns to distinguish IV pump alarms from those of other critical care devices, such as ventilators and vital sign patient monitors. These auditory alarms were then subjected to extensive lab and field studies for effectiveness and acceptability. An early test in actual hospital settings with extended use showed user dissatisfaction with the harshness of some of the alarm melodies. The international standard recommended a discordant set of tone melodies for the highest alarm level, which clinicians, patients, and their families complained was too harsh and irritating. Ironically, the alarm tone standards chose discordant tones to purposefully create this type of sound. Some clinician complaints were of the nature that they would not use these devices at all unless the alarms were modified. Or, worse, they would permanently disable the alarms, creating
TAF-ER562X_WINGR-09-0401-C010.indd 420
11/8/10 11:55:26 PM
Alarm Design
421
a potentially dangerous use scenario. This outcome highlights a well-known dilemma for human factors professionals that laboratory studies are imperfect predictors of user behavior and attitudes in a real-world extended use setting. The previous lab-based usability studies had determined that these alarm sounds were effective and alerting, but the studies did not capture long-term subjective preference ratings. A tone design specialist was engaged to redesign the tones to be more acceptable while still alerting the users and remaining in compliance with the IEC alarm standard. Subsequent comparative usability evaluations (group demonstrations and interviews) demonstrated the acceptability of the redesigned melodies.
ACKNOWLEDGMENTS The author would like to acknowledge Carl Pantiskas for the extensive input and editing that he provided for this chapter.
REFERENCES Adams, A. S. and Edworthy, J. (1995). Quantifying and predicting the effects of basic text display variables on the perceived urgency of warning labels: Tradeoffs involving font size, border weight and colour. Ergonomics, 38, 2221–2237. American Society for Testing and Materials. (1999). F1463-93. Standard Specification for Alarm Signals in Medical Equipment Used in Anesthesia and Respiratory Care. West Conshohocken, PA: American Society for Testing and Materials. Association for the Advancement of Medical Instrumentation, Arlington, VA. (2010). HE 75:2010. Human Factors Engineering Guidelines and Preferred Practices for the Design of Medical Devices. Athenes, S., Chatty, S., and Bustico, A. (2000). Human factors in ATC alarms and notifications design: An experimental evaluation. Third USA/Europe Air Traffic Management R&D Seminar, Naples, Italy. Bacon, S. J. (1974). Arousal and the range of cue utilization. Journal of Experimental Psychology, 101, 81–87. Belz, S. M., Robinson, G. S., and Casali, J. G. (1999). A new class of auditory signals for complex systems: Auditory icons. Human Factors, 41, 608–618. Blattner, M. M., Sumikawa, D. A., and Greenberg, R. M. (1989). Earcons and icons: Their structure and common design principles. Human-Computer Interaction, 4, 11–44. Beneken, J. E. W. and Van der Aa, J. J. (1989). Alarms and their limits in monitoring. Journal of Clinical Monitoring, 5, 205–210. Bliss, J. P., Gilson, R. D., and Deaton, J. E. (1995). Human probability matching behavior in response to alarms of varying reliability. Ergonomics, 38, 2300–2312. Block, F. E. (1994). Human factors and alarms. In C. Lake (Ed.), Clinical Monitoring for Anesthesia and Critical Care (2nd ed.). Philadelphia: W. B. Saunders. Block, F. E., Nuutinen, L., and Ballast, B. (1999). Optimization of alarms: A study on alarm limits, alarm sounds, and false alarms, intended to reduce annoyance. Journal of Clinical Monitoring and Computing, 15, 75–83. Cratty, B. J. (1973). Movement Behavior and Motor Learning. Philadelphia: Lea & Febiger. Deller, A., Konrad, J., Kilian, J., and Schuhle, B. (1992). Alarms in an operative intensive care unit— Response of the nursing staff. In J. Hedley-Whyte (Ed.), Operating Room and Intensive Care Alarms and Information Transfer. Philadelphia: American Society for Testing Materials. Edworthy, J. (1994). Urgency mapping in auditory warning signals. In N. Stanton (Ed.), Human Factors in Alarm Design. London: Taylor & Francis.
TAF-ER562X_WINGR-09-0401-C010.indd 421
11/8/10 11:55:26 PM
422
Handbook of Human Factors in Medical Device Design
Edworthy, J. (1998a). Does sound help us to work better with machines? A commentary on Rautenberg’s paper “About the importance of auditory alarms during the operation of a plant simulator.” Interacting with Computers, 10, 401–409. Edworthy, J. (1998b). What makes a good alarm. In Medical Equipment Alarms: The Need, the Standards, the Evidence. London: Institute for Electrical Engineers. Edworthy, J. and Stanton, N. (1995). A user-centered approach to the design and evaluation of auditory warning signals: 1. Methodology. Ergonomics, 38, 2262–2280. Edworthy, J. and Meredith, C. (1997). Influence of verbal labeling and acoustic quality on the learning and retention of medical alarms. International Journal of Cognitive Ergonomics, 1, 229–243. Gaver, W. W. (1986). Auditory icons: Using sound in computer interfaces. Human-Computer Interaction, 2(2), 167–177. Gerth, J. (1993). Identification of sounds with multiple timbres. Proceedings of the Human Factors and Ergonomics Society, 37, 539–543. Giles, C. S., Edworthy, J., Brown, R. D. H., and Davies, S. C. (1998). ITU alarm confusion—A smart solution. In Institute for Electrical Engineers, Medical Equipment Alarms: The Need, the Standards, the Evidence. London: Institute for Electrical Engineers. Gilson, R., Moulaua, M., Graft, A., McDonald, D. (2001). Behavioral influences of proximal alarms. Human Factors, 43, 595–610. Griffith, R. L. and Raciot, B. M. (1992). A survey of practicing anesthesiologists on auditory alarms in the operating room. In J. Hedley-Whyte (Ed.), Operating Room and Intensive Care Alarms and Information Transfer. Philadelphia: American Society for Testing and Materials. Hakkinen, M. T. and Williges, B. H. (1984). Synthesized warning messages: Effects of an alerting cue in single- and multiple-function voice synthesis systems. Human Factors, 26, 185–195. International Electrotechnical Commission. (2004). IEC 60601-1-8—Medical electrical equipment— Part 1–8: General Requirements for Safety—Collateral Standard: General Requirements, Tests and Guidance for Alarm Systems in Medical Electrical Equipment and Medical Electrical systems. Geneva, Switzerland: International Electrotechnical Commission. International Organization for Standardization. (1996). 11428—Ergonomics—Visual Danger Signals—General Requirements, Design and Testing. Geneva, Switzerland: International Organization for Standardization. Japanese Industrial Standard. (1991). JIS T 1031-1991. General Rules for Alarms of Medical Equipment. Tokyo, Japan: Japanese Industrial Standards Organization. Kerr, J. H. (1985). Warning devices. British Journal of Anaesthesiology, 57, 696–708. Kestin, I. G., Miller, B. R., and Moore, C. H. (1986). Auditory alarms during anesthesia monitoring. Anesthesiology, 69, 106–109. Kirwan, B. and Ainsworth, L. K. (Eds.). (1992). A Guide to Task Analysis. London: Taylor & Francis. Laroche, C., Tran Quoc, H., Hetu, R., and McDuff, S. (1991). Detectsound: A computerized model for predicting detectability of warning signals in noisy environments. Applied Acoustics, 33, 193–214. Lazarus, H. and Hoge, H. (1986). Industrial safety: Acoustic signals for danger situations in factories. Applied Ergonomics, 17, 41–46. Momtahan, K., Hetu, R., and Tansley, B. (1993). Audibility and identification of auditory alarms in the operating room and intensive care unit. Ergonomics, 36, 1159–1176. Nasir, B. (1998). A medical device alarm based on the pulse oximeter. In Institute for Electrical Engineers, Medical Equipment Alarms: The Need, the Standards, the Evidence. London: Institute for Electrical Engineers. O’Carroll, T. M. (1986). Survey of alarms in an intensive therapy unit. Anaesthesia, 41, 742–744. Orr, J. A. and Westenskow, D. R. (1993). A breathing circuit alarm based on neural networks. Journal of Clinical Monitoring, 9, 31–37. Patterson, R. D. (1982). Guidelines for auditory warning systems on civil aircraft. Civil Aviation Authority Paper 82017. London: Civil Aviation Authority.
TAF-ER562X_WINGR-09-0401-C010.indd 422
11/8/10 11:55:26 PM
Alarm Design
423
Patterson, R. D. (1989). Guidelines for the design of auditory warning sounds. Proceedings of the Institute of Acoustics, 11, 59–71. Edinburgh, Scotland. Patterson, R. D. (1990). Auditory warning sounds in the work environment. Philosophical Transactions of the Royal Society of London, 327, 485–492. Patterson, R. D., Edworthy, J., Shailer, M., Lower, M. C., and Wheeler, P. D. (1986). Alarm sounds for medical equipment in intensive care areas and operating theatres. Institute of Sound and Vibration Research Report No. AC598. Sanders, M. S. and McCormick, E. J. (1987). Human Factors in Engineering and Design (6th ed.). New York: McGraw-Hill. Seagull, F. J., Xiao, Y., Mackenzie, C. F., and Wickens, C. D. (2000). Auditory alarms: From alerting to informing. Proceedings of the IEA/HFES Congress, 1, 223–226. Sorkin, R. D., Kantowitz, B. H., and Kantowitz, S. C. (1988). Likelihood alarm displays. Human Factors, 30, 445–459. Stanton, N. (1994). Alarm initiated activities. In N. Stanton (Ed.), Human Factors in Alarm Design. London: Taylor & Francis. Stanton, N. A. and Baber, C. (1997). Comparing verbal alarm displays: Speech versus textual systems. Ergonomics, 40, 1240–1254. Stanton, N. and Edworthy, J. (1999). Auditory warnings and displays: An overview. In N. A. Stanton and J. Edworthy (Eds.), Human Factors in Auditory Warnings. Brookfield, VT: Ashgate. Stanton, N. A., and Stammers, R. B. (1998). Alarm-initiated activities: Matching visual formats to alarm handling “tasks.” International Journal of Cognitive Ergonomics, 2, 331–348. Watt, R., Navabi, M., Mylrea, K., and Hameroff, S. (1989). Integrated monitoring smart alarms can detect critical events and reduce false alarms (abstract). Anesthesiology, 71, A338. Weinger, M. B. and Smith, N. T. (1993 and in press). Alarms, and Integrated Monitoring Systems. In J. Ehrenwerth and J. B. Eisenkraft (Eds.), Anesthesia Equipment: Principles and Apllications (1st and 2nd eds.). Malvern, PA: Mosby. Welch, J. (1999). Auditory alarms in intensive care. In N. A. Stanton and J. Edworthy (Eds.), Human Factors in Auditory Warnings. Brookfield, VT: Ashgate. Wiklund, M. and Wilcox, S. (2005). Designing Visability into Medical Products. New York: CRC Press. Woodson, W. E., Tillman, B., and Tillman, P. (1992). Human Factors Design Handbook (2nd ed.). New York: McGraw-Hill.
TAF-ER562X_WINGR-09-0401-C010.indd 423
11/8/10 11:55:26 PM
TAF-ER562X_WINGR-09-0401-C010.indd 424
11/8/10 11:55:26 PM
11 Software User Interfaces Michael E. Wiklund, MS, CHFP CONTENTS Sample Software User Interfaces .....................................................................................427 Influences on Software User-Interface Design ................................................................428 11.1 Design Considerations ...........................................................................................429 11.1.1 General Considerations..............................................................................430 11.1.1.1 Initial Ease of Use ......................................................................430 11.1.1.2 Focus on User Tasks ...................................................................430 11.1.1.3 Provide User Guidance............................................................... 431 11.1.1.4 Safeguard against Use Error.......................................................432 11.1.1.5 Optimize Interaction Requirements ...........................................432 11.1.1.6 Support Product Evolution .........................................................432 11.1.1.7 Improve Software and Hardware Integration .............................433 11.1.1.8 Interaction Style .........................................................................433 11.1.2 Special Considerations ..............................................................................434 11.1.2.1 Screen Size .................................................................................434 11.1.2.2 Compatibility..............................................................................435 11.1.2.3 Information Priority ...................................................................436 11.1.2.4 Information Legibility ................................................................436 11.1.2.5 User Population ..........................................................................436 11.1.2.6 Standardization ..........................................................................437 11.1.2.7 System Integration ......................................................................437 11.2 Design Guidelines ..................................................................................................437 11.2.1 Conceptual Model .....................................................................................438 11.2.2 User-Interface Structure ............................................................................439 11.2.2.1 Linear Structure ........................................................................ 440 11.2.2.2 Branching Structure .................................................................. 440 11.2.2.3 Web (or Network) Structure ...................................................... 440 11.2.3 Interaction Style ........................................................................................ 440 11.2.4 Screen Layout ............................................................................................441 11.2.5 Legibility ...................................................................................................442 11.2.6 Aesthetics ................................................................................................. 446 11.2.7 Data Entry .................................................................................................447 11.2.8 Color ..........................................................................................................450
425
TAF-ER562X_WINGR-09-0401-C011.indd 425
11/9/10 2:21:22 AM
426
Handbook of Human Factors in Medical Device Design
11.2.9
Dynamic Displays (See Chapter 8, “Visual Displays”)............................452 11.2.9.1 Trend Displays .........................................................................452 11.2.9.2 Waveform Displays ..................................................................453 11.2.9.3 Numeric Displays ....................................................................456 11.2.10 Special Interactive Mechanisms...............................................................457 11.2.10.1 Soft Key User Interfaces .........................................................457 11.2.10.2 Control Wheel User Interfaces ................................................458 11.2.10.3 Touch-Screen User Interfaces..................................................459 11.2.10.4 On-Screen Keyboards and Keypads ........................................463 11.2.10.5 Speech-Emitting User Interfaces............................................ 464 11.2.11 User Support ............................................................................................465 11.3 Case Studies .......................................................................................................... 466 11.3.1 Diabetes Management System ...................................................................467 11.3.2 Neuromodulation System ..........................................................................468 Resources .........................................................................................................................470 References ........................................................................................................................470
A growing proportion of medical devices incorporate some kind of a software user interface—the medium through which people and computers interact. This trend is driven in part by the cost advantages of microprocessor-based control systems over electromechanical ones. It is also being driven by the marketplace demand for “smarter” devices that perform more functions without added hardware cost. As a result, users now operate many devices by selecting an option from an on-screen menu rather than pressing a mechanical switch, throwing a lever, or turning a knob. A medical device’s software interface might be small and simple, such as the stampsized display on a digital thermometer, or it might be large and complicated, such as the multiple displays and controls associated with a CT scanner (Figure 11.1). Regardless, the software user interfaces of all types of medical devices should facilitate user tasks, prevent use error, and satisfy the device users’ needs.
FIGURE 11.1 The software user interfaces of a digital thermometer and a CT scanner differ greatly in terms of the size and complexity of their displays and controls, affecting the amount of information transmitted bidirectionally between users and the device. (PENDING With permission.)
TAF-ER562X_WINGR-09-0401-C011.indd 426
11/9/10 2:21:22 AM
Software User Interfaces
427
Unfortunately, many device-related use errors and user complaints stem from shortcomings in those devices’ software user interfaces. For example, users can become lost in menu systems that locate frequently used features three or more levels deep. They can become overly dependent on a device’s automated functions, reducing their ability to cope with operational upsets, such as a partial device failure that requires manual actions that are relatively unfamiliar. They also might struggle to form a complete and accurate mental model of how a device works when it subtly shifts between operational modes or is riddled with inconsistencies, such as the following: • Differences in the placement of the same information on a series of screens • Variations in how data are entered • Alternative means of moving between screens and selecting information (e.g., when to click on a button versus select options from a pull-down list) The usability problems described above and the shortcomings that lead to them are avoidable. The application of the design guidelines recommended in this chapter will produce designs that are more likely to satisfy users’ needs and reduce use errors.
SAMPLE SOFTWARE USER INTERFACES Embedded software user interfaces—those found in special purpose medical devices—are plentiful. Examples include patient monitors, infusion pumps, and defibrillators. These devices tend to incorporate a set of dedicated controls, such as a number pad, four-way cursor control, and special purpose keys that allow the users to interact effectively with the associated software user interface. As medical devices get larger, it is common to find their software user interface presented in a form more like a conventional PC application, running within common operating systems (e.g., Windows™) on a standard computer display (e.g., 15-inch diagonal color LCD). Nurses’ central station monitors, such as the one shown in Figure 11.2, are a common example. These devices tend to incorporate conventional input devices, such as a keyboard and mouse (or trackball).
FIGURE 11.2
Some medical devices have the physical characteristics of computer workstations.
TAF-ER562X_WINGR-09-0401-C011.indd 427
11/9/10 2:21:24 AM
428
Handbook of Human Factors in Medical Device Design
INFLUENCES ON SOFTWARE USER-INTERFACE DESIGN Many factors drive the structure, interaction style, and appearance of the software user interfaces of medical devices. A medical device’s size can be a dominant factor; for example, smaller medical devices are frequently preferred for use in congested workspaces. The device’s smaller size limits the size of both its display and its controls that, in turn, affect its information presentations and the mechanisms used to interact with that information. For example, complex user interactions might call for the device to have a larger display (e.g., 800 × 600 pixels), but its size constraints might require a smaller one (e.g., 320 × 240 pixels). Naturally, the amount of available screen area will have a substantial effect on the entire user-interface design. A small display can require users to interact with information from a series of screens instead of one (Figure 11.3). A large display can support presentation of all task-related information on a single screen. However, care must be taken to ensure that the screen does not look congested or intimidating to users. Access to reliable power is another important factor influencing the type of display technology and whether the display is constantly turned on or put to “sleep” after seconds or minutes of inactivity. Medical devices that stay in one place and have continuous access to AC power (e.g., wall-mounted patient monitors, MRI scanners) can use bright, color displays that consume a lot of energy. However, portable devices that draw power mostly from rechargeable batteries (e.g., digital thermometers, infusion pumps, and transport monitors) generally incorporate less power-hungry displays, such as a monochrome LCD that is not backlit. Additional factors influencing display selection and therefore the user-interface design include functional complexity, the degree to which a device will be controlled automatically as opposed to manually, and cost. However, perhaps the most significant influence on the design of many medical devices is the consumer software industry and its standard operating systems and requisite interaction styles, browsers, and associated development tools. The latter factor explains why many medical devices, such as those used in laser-based eye surgery, have software user interfaces that have the same “look and feel” as PC software applications and Web sites. This commonality with consumer software is beneficial in that it can reduce software development costs and allow users to apply their prior experience using PC applications and browsing the Internet to their use of the medical device. A potential downside to such common platform medical applications is that they limit design freedom, thereby reducing options to optimize user-interface elements specific to
FIGURE 11.3 Smaller displays must divide the information that could be presented on a single, large display into four chunks (i.e., four separate screens).
TAF-ER562X_WINGR-09-0401-C011.indd 428
11/9/10 2:21:24 AM
Software User Interfaces
429
unique clinical tasks or functions. Many medical devices serve a single purpose and do not benefit from a user-interface setup to perform tasks such as “open file” or “close window.” Also, some designs are better suited to use with specific pointing devices, such as a mouse versus a touch screen, than others. Commercial PC platforms have significant user-interface design shortcomings that can propagate to the medical application, potentially with disastrous consequences. Therefore, designers face many trade-offs when deciding whether to use a common platform or to develop a unique solution. Regardless, usability testing under realistic use conditions is the most reliable method of assessing user-interface design decisions. The balance of this chapter provides focused guidance on the design of medical device software user interfaces. This chapter excludes basic design topics that are adequately addressed by other standards and guides for software user-interface design. In place of basic guidance, it discusses specific aspects of medical device software design that are essential to ensuring safety while bolstering effectiveness, usability, and appeal. Accordingly, readers should view the content as supplemental to the larger body of guidance available through the listed resources. Readers seeking basic software user-interface design guidance are referred to the resource listing at the end of this chapter as well as the web site of the ACM’s Special Interest Group on Computer-Human Interaction (SIGCHI), which provides updated references to design standards and guides.
11.1 DESIGN CONSIDERATIONS Microprocessors have “turbocharged” medical devices, substantially increasing their ability to present diagnostic information, deliver therapy, and offer safeguards against use error. Setting the example of multi-million-dollar imaging equipment aside, consider the humble digital thermometer. In place of a mercury-filled tube that posed readability problems, there are now electronic devices that work faster, provide a large numerical readout of the patient’s temperature, and allow one to cycle through recent readings to determine temperature trends. However, early generation digital thermometers posed significant usability challenges, such as determining how to review the last few measurements by pressing cryptically labeled buttons. Accordingly, even the most rudimentary software user interface can pose design challenges. The recent history of medical device incidents, involving products as diverse as patientcontrolled analgesia pumps, radiation therapy machines, and apnea monitors, teaches us the importance of optimizing the software user interface to meet users’ needs and expectations and to reduce the likelihood of use error and foreseeable misuses. Clearly, patient safety depends on users receiving the right information and control options at the right time so that they understand what is happening and what needs to be done. Problems with the usability of medical software range from minor inconveniences to major, error-inducing flaws. An error committed while using a consumer software application might cause the loss of a document. In contrast errors committed while using a software-based medical device could cause injury or death. Most companies already recognize that software user-interface development is not solely a programming task any more than hardware user-interface development is solely a metal fabrication task. People with knowledge of human capabilities and the demands of the medical environment, as well knowledge of good user-interface design practices, should play leading roles in the development process (see ANSI/AAMI HE 74-2001).
TAF-ER562X_WINGR-09-0401-C011.indd 429
11/9/10 2:21:24 AM
430
Handbook of Human Factors in Medical Device Design
Assuming that people who possess the right knowledge and skills are participating in the development process, the next challenge is to drive the design process in the direction of the simplest solution that addresses the users’ needs. Designers must accept the fact that most users do not want technology per se. Rather, they want help attaining a specific goal or outcome. The proposed technology should be as simple to operate as possible. One might recall the adage that woodworkers ultimately need holes rather than the drills that make the holes; the drills represent a way to meet a need rather than the fulfillment of the need itself. Similarly, nurses ultimately need useful information about patients rather than the monitors that display the information. Accordingly, designers are well served to adopt a “keep-it-simple stupid” or KISS philosophy and to resist the temptation to add extra features or to make the software user interface more visually dramatic than necessary. Software user interfaces embedded in medical devices have foremost a utilitarian function. The user interface can be “softened” to give it a friendly look and feel as compared to an analytical application intended for use by computer scientists. Many medical devices need to work in harmony with others and should be “good citizens.” This means that they should be compatible with other products and not draw undue attention to themselves for reasons other than conveying useful information or facilitating useful work.
11.1.1 GENERAL CONSIDERATIONS Some basic principles associated with software user-interface design are presented below (see also Chapter 1, “General Principles”). 11.1.1.1 Initial Ease of Use While most medical device manufacturers welcome the opportunity to provide customer training sessions, called “in-services,” that might last 30 minutes or so, invariably some medical staff will not attend them. Instead, caregivers will ask a peer (e.g., an experienced nurse who will serve as the “power user”) to show them how to use a given device or try to figure it out for themselves—possibly even during a crisis. The latter situation is just one of several scenarios in which a medical device user, including people using medical devices in their homes, might not receive formal training. Consequently, it is important for medical devices to be easy to use. At a minimum, untrained users should be able to accomplish basic and critical tasks, such as turning the device on and off. Fortunately, there are many ways to make a software user interface usable. Just a few of the options include limiting the number of functional options, providing clear screen titles and information labels, prompting the user in a step-by-step fashion, and providing realtime learning resources, such as online help. 11.1.1.2 Focus on User Tasks With the exception of certain workstations, such as an ultrasound scanner, few medical devices demand a user’s continuous attention (Figure 11.4). Rather, users divide their attention among multiple elements in their workspace, including various devices, other people, and the patient. When the user pays attention to a given device, it is often to perform a discrete task before shifting attention to something else. This creates a strong need for medical device user interfaces to be task oriented, allowing users to quickly access options, take action, and confirm the results. Therefore, a detailed task analysis that identifies the
TAF-ER562X_WINGR-09-0401-C011.indd 430
11/9/10 2:21:24 AM
Software User Interfaces
431
FIGURE 11.4 Some medical devices call for frequent or continuous user interaction while others do not. Left: An ultrasound workstation requires the user to pay almost continuous attention to the software user interface and its associated information displays (i.e., scanned images). Right: In contrast, caregivers might pay close attention to a portable patient monitor only when checking a patient’s status, making occasional adjustments to settings, and when there is an alarm. (Courtesy of Philips Healthcare. With permission.)
frequent, urgent, and critical tasks provides data essential to designing the software user interface. 11.1.1.3 Provide User Guidance Many caregivers favor user interfaces that provide step-by-step procedural instructions rather than leaving the user to infer the proper operational sequence from an array of options (Figure 11.5). Therefore, while providing instructions in the form of pop-ups or prompts might be perceived as an impediment to rapid task performance, it can be an appropriate and desirable means to ensure that people accomplish complex tasks correctly. Directions are also helpful to caregivers when they are learning to use a device. However, such procedural support can make users feel locked into one particular way of interacting with a device, and more experienced users might consider the interface inefficient and less satisfying. Therefore, designers should carefully consider the advantages and disadvantages of directed procedural support.
FIGURE 11.5 Inexperienced users generally consider medical devices that provide explicit, step-by-step instructions to be easier to use. Many experienced users might also value step-by-step instructions when implemented wisely.
TAF-ER562X_WINGR-09-0401-C011.indd 431
11/9/10 2:21:24 AM
432
Handbook of Human Factors in Medical Device Design
11.1.1.4 Safeguard against Use Error As medical devices get “smarter,” they present the opportunity to safeguard the user against actions or inactions that could have serious negative consequences. For example, a device might be designed to verify that the user has entered a safe medication dosage or ventilation pressure and to alert the user—or prevent operation altogether—if safe limits have been breached (Figure 11.6). Designers should look for opportunities to implement safeguards, taking care to ensure that they do not interfere with or prevent appropriate use, especially during emergencies. 11.1.1.5 Optimize Interaction Requirements Specific types of caregivers, such as interventional cardiologists and ultrasound technicians, spend hours every day operating specific equipment and, naturally, develop a high level of mastery. In contrast, many medical device users, including laypeople, are busy and usually lack the time or desire to master a software user interface. Most caregivers want to spend less time interacting with devices and more time interacting with patients. Work pressures mean that these caregivers cannot devote much time to learning to use a device. Designers must also appreciate that their device is not the only one that the caregiver must learn to use. Furthermore, caregivers might use dozens of complex devices, each of which has a unique user interface. Therefore, software interface designers should generally assume that they have the user’s attention for only brief periods of time at best. This suggests bringing essential information and controls to the interface’s top level. It also suggests designing software so that users receive essential information about important events, even if they are not attending to the device when such events occur. 11.1.1.6 Support Product Evolution Software user-interface designers face the difficult but important challenge of anticipating future changes. A long (multiyear) service life suggests that software will be updated intermittently. New capabilities will be added, and an application’s look and feel might be changed to reflect a new branding scheme. Accordingly, designers should make their user interface as flexible as possible without compromising its performance. In some cases,
FIGURE 11.6 Medication dosing safety software asks caregivers to confirm dose settings that exceed hospital-specified limits.
TAF-ER562X_WINGR-09-0401-C011.indd 432
11/9/10 2:21:25 AM
Software User Interfaces
433
this might mean leaving some unused space for an expanded list of parameters or allowing the application’s outward appearance to be changed (reskinned) without significantly affecting the user interface’s underlying functionality and overall organization. It is prudent for designers to ask users and perhaps marketing personnel to describe their vision of the future, including new needs and functions. 11.1.1.7 Improve Software and Hardware Integration More often than not, product development teams face pressure to “freeze” the hardware user interface well ahead of the software user interface. In fact, it is common for hardware designers to settle on a screen size and control panel design before many details of the software user-interface design have been adequately explored and established. This can lead to awkward design solutions that pose usability problems. The need for better timing and integration is reflected in some medical devices that look as if the teams designing the hardware and software user interfaces rarely communicated. Incompatibilities might include clashing color schemes, misalignment between hardware buttons and on-screen information, insufficient direct access to critical information, and inappropriate control mechanisms or location. It is far better to synchronize the hardware and software development efforts so that they inform each other. For example, the selection of a compact, 320 × 240 quarter VGA display for size and weight requirements might be balanced against the need to present a large amount of trend data on a single screen. Alternatively, the need to rapidly silence an alarm by pressing a dedicated, physical button might be weighed against the option of selecting an on-screen target using a mouse, trackball, or touch screen. Integrated solutions have a complementary appearance that effectively relates hardware controls and displays to those presented by the software (Figure 11.7). 11.1.1.8 Interaction Style Software user interfaces employ varying interaction styles. Common styles include the following: • Menus ask a user to select a desired option from a set of options that might be presented in a list (fixed or scrolling), tool palette, ATM-like interface employing “soft keys,” or other format. For example, a nurse might select the option “Pelvis” from a list of body parts by moving a highlight bar through a long list and then pressing the Enter key. Treatment − Step 1 Option 1 Option 2 Option 3 Option 4
FIGURE 11.7 This schematic design illustrates a high level of integration between the hardware (soft keys) and the software (on-screen menu options).
TAF-ER562X_WINGR-09-0401-C011.indd 433
11/9/10 2:21:26 AM
434
Handbook of Human Factors in Medical Device Design
• Direct manipulation asks a user to interact directly with interface elements, such as “clicking” or “dragging” on displayed objects, to achieve a task goal. For example, a physician might open a patient chart by clicking on an object that looks like a folder or file it away by “dragging” it to an object that looks like a file cabinet. This example exhibits how a direct manipulation style of interaction might employ a metaphor (e.g., folders and file cabinets) to enhance understanding. Perhaps the most common example of an interface metaphor is the “desktop,” where one expects to find documents and tools related to document manipulation. • Question-and-answer dialogue leads users through a series of questions requiring a response over the course of a task. For example, a diagnostic device might ask, “What is the patient’s medical record number,” to which the nurse would respond by typing in a series of numbers. Some “wizards” that guide users through complex tasks might be considered a special form of questionand-answer dialogue. • A command line calls for users to know a special vocabulary and syntax to perform tasks. For example, a scanner might present some prompt, such as a blinking colon, indicating that the user must type in a key word or concatenated expression, such as “patient/openrecord,” to proceed with the task. This dialogue style, one of the oldest, is how people interacted with DOS-based personal computers in the 1980s. “Keyboard shortcuts” are a modern use of the command-line dialogue style. Lacking an absolute definition of what constitutes a “style,” some designers might add spreadsheets, natural (spoken) language, data entry forms, and dialog boxes to the list. In practice, a majority of software user interfaces employ a combination of dialogue styles. In fact, the lines of distinction among dialogue styles quickly blur when examining actual design solutions. For example, one might regard a tool palette to be a menu that enables direct manipulation and employs a metaphor. When designing a medical product’s software user interface, it is important to select the dialogue style(s) that are best suited to the task at hand. Readers should refer to the resources provided at the end of this chapter, particularly ANSI/HFES 200 and the ISO 9241-100 series of software ergonomics standards, to develop a full understanding of a given style’s strengths and weaknesses. In summary, interaction styles that limit the need for special knowledge are best suited to naive users and those who have limited motivation or capacity to master a special command language and/or syntax. Conversely, special commands and/or syntax may help to accelerate tasks performed repeatedly by persons who can be expected to receive special training.
11.1.2 SPECIAL CONSIDERATIONS The software user interface of a medical device warrants the following special considerations. 11.1.2.1 Screen Size Many medical devices must be compact to work well in their intended use environments. Critical care units, operating rooms, and ambulances, for example, are typically congested with equipment and constrained in terms of space. Similarly, medical devices might also be
TAF-ER562X_WINGR-09-0401-C011.indd 434
11/9/10 2:21:26 AM
Software User Interfaces
435
used in tight spaces within people’s homes, such as bathrooms, which have limited counter space. Therefore, small displays are common not because a larger display might not be preferable from the users’ perspective but rather because a larger display will not fit within the given environmental constraints or meet portability goals for the device (Figure 11.8). The need to compress information into a small display space forces user-interface designers to prioritize and layer information within a screen hierarchy. In the process, designers need to give users the information they need at the right time with a minimal amount of user effort. For example, users should not have to take several steps to accomplish frequent or urgent tasks because the required information and/or controls are “buried” in a menu system. 11.1.2.2 Compatibility Generally, device manufacturers strive to make their products compatible in terms of interaction style as well as appearance; characteristics often referred to as a product’s “look and feel.” Accordingly, a given manufacturer might develop a common “look and feel” to be implemented across entire product lines. Such a business strategy leads to products that share common hardware and software elements, such as a touch screen or control wheel, that then dictate specific kinds of user interactions. It also leads to devices that look similar in terms of their labeling, menus, use of windows, and many other features. While user-interface consistency or product compatibility might be good for the manufacturer, it may or may not be good for the users. Some devices, even within a single manufacturer’s product line, might be used in such different scenarios that they call for different types of user interfaces. For example, a menu-based software user interface might be an acceptable means to acquire information presented on a bedside patient monitor but unacceptable as a means to interact with a defibrillator. Thus, designers must strike a proper balance between user-interface compatibility and addressing an individual device’s unique user-interface requirements.
FIGURE 11.8 A portable ultrasound scanner’s small screen enhances its portability while placing space constraints on the software user interface. (Courtesy of SonoSite, Inc. With permission.)
TAF-ER562X_WINGR-09-0401-C011.indd 435
11/9/10 2:21:26 AM
436
Handbook of Human Factors in Medical Device Design
TABLE 11.1 Urgent Visual Acquisition Tasks Determine the source of a high-priority alarm Determine the change in blood pressure over the last 5 minutes Inspect an ECG waveform for arrhythmias
11.1.2.3 Information Priority Caregivers are often under intense time pressure to perform tasks quickly as well as to perform several tasks in parallel (i.e., multitask). Consequently, users might have only seconds to acquire information from a given device (Table 11.1) before they need to move on to another task. So it is important for many medical devices to communicate critical information quickly, accurately, and reliably. In many cases, this means limiting the amount of information presented on the resting screen (i.e., home screen) and making the most important data more conspicuous than the rest. For example, the patient’s heart rate might be the most dominant piece of information displayed on a portable patient monitor’s resting screen—a screen that also shows waveforms and several other numeric values in a less dominant manner (Figure 11.9). In this way, the heart rate will draw the user’s eye at a glance, and there will be a reduced chance that the user will misread the numeric value. Notably, the designers of multiparameter medical devices often violate this important principle. 11.1.2.4 Information Legibility The legibility of on-screen information is critical to preventing use errors. Mistaking one number for another can have lethal consequences. Therefore, user-interface designers should not compromise legibility for the sake of fitting more information on a screen or to achieve an aesthetic goal (Figure 11.10). 11.1.2.5 User Population Some medical devices intended for use by trained clinicians in a hospital might also be used by laypeople in their homes. Accordingly, designers need to build an appropriate amount of flexibility into the software user interface so that it serves multiple user groups. This might suggest creating both advanced and simple operating modes, for example, or
FIGURE 11.9 The heart rate (80 beats per minute) and oxygen saturation level (97%) are prominent on this portable patient monitor. (Courtesy of Welch Allyn, Inc. Unauthorized use not permitted.)
TAF-ER562X_WINGR-09-0401-C011.indd 436
11/9/10 2:21:26 AM
Software User Interfaces
437
FIGURE 11.10 The limited software user interface on this electrocautery device incorporates a large display to ensure the legibility of numerical values.
perhaps different software versions or devices. Of course, in such a situation, care must be taken to prevent accidental or undetected shifts between the modes. 11.1.2.6 Standardization In some cases, user-interface designs must comply with standards and regulations or with established conventions that apply to medical devices. For example, the International Electrotechnical Commission (IEC) has established standards for alarm systems, including those presented via a software user interface (IEC 60601-1-8). There are also established conventions for the use of certain colors to represent medications and hemodynamic parameters. Designers should follow such conventions whenever possible. Designers should resort to novel solutions only if there are no existing or emerging standards or user conventions or if the proposed innovation offers disproportionate benefits in safety, effectiveness, and usability. Only a relatively small number of design practices have become common and standardized, so medical device designers retain considerable creative freedom. 11.1.2.7 System Integration It is increasingly important for medical devices to be able to share information with other devices. Sometimes, information sharing takes places in the background and does not require user intervention. However, at other times, when information sharing requires user intervention, it is advantageous for the connected systems to “speak” the same language. In such cases, it is optimal from a user standpoint for the devices to employ the same or similar user-interface mechanisms (i.e., “widgets,” such as scrolling lists and drop-down menus). That way, the user does not need to learn and remember two or more ways of handling the same data management task.
11.2 DESIGN GUIDELINES The following guidelines should be helpful to software user-interface designers who wish to optimize an application in terms of its functional capabilities and usability. The guideline categories are as follows: • Conceptual model—Providing an overarching construct for how users are likely to think about a user interface—the mental model or picture that users form about how the software user interface works
TAF-ER562X_WINGR-09-0401-C011.indd 437
11/9/10 2:21:27 AM
438
Handbook of Human Factors in Medical Device Design
• User-interface structure—Placing individual screens in a hierarchy that complements how people prefer to approach frequent, urgent, and critical tasks • Interaction style—Establishing a pattern of interaction between the user and the software application that facilitates tasks, accommodates users’ capabilities and eases users’ navigation among interface elements • Screen layout—Organizing information on a computer display so that users can locate specific items quickly and make the appropriate associations • Legibility—Presenting textual and graphical information in a clear manner so that users can read it and discriminate important details • Aesthetics— Presenting information in a manner that is visually pleasing, which can have a positive influence on task performance while not intimidating new users • Data entry—Establishing rules for how users input data or make selections via the software user interface • Color—Making color contribute in a meaningful way to the clarity of information as well as drawing attention to the most important information • Dynamic displays—Using active graphical and textual elements to convey information in a more compelling manner than possible using static displays • Special Interactive Mechanisms— Some specialized methods of software user interface interaction (e.g., touch screens, control wheels) provide unique advantages as well as disadvantages for medical device designs. • User support— Giving users information at the right time and in the right format to help them perform tasks safely, quickly and effectively
11.2.1 CONCEPTUAL MODEL Software user interfaces vary widely in terms of their purpose, structure, method of navigation, and visual appearance, to name just a few design variables. However, the need to establish a relatively simple, overarching organizing principle (also called a conceptual model) is a common attribute of user-friendly solutions. This is because people have a much easier time understanding how to interact with software user interfaces that can be “boiled down” to a simple representation of how they interact with the device. Guideline 11.1: Number of Elements As a general rule, it is easier to interact with software based on a conceptual model with a relatively small number of basic elements—perhaps 10 or fewer. Limiting the number of basic elements helps users form a simple mental model (i.e., big picture) of how the user interface is organized.
Guideline 11.2: User Task Orientation Conceptual models should be based on how a designer expects the users to think of the software user interface and/or how the user approaches tasks. The conceptual model should be grounded in a thorough understanding of how the software will be used to accomplish real-world tasks. Accordingly, the conceptual model should reflect a logical organization of user tasks rather than the device’s electromechanical functions or software modules.
TAF-ER562X_WINGR-09-0401-C011.indd 438
11/9/10 2:21:27 AM
Software User Interfaces
439
11.2.2 USER-INTERFACE STRUCTURE The following design guidelines pertain to the overall organization of information and control functions provided by the software user interface. Guideline 11.3: Compatibility with the Conceptual Model The software user-interface architecture—the structure or hierarchy of individual screens (also called nodes) and their relationships (also called links)—should reflect the underlying conceptual model. For example, a conceptual model prescribing a software user interface with five major elements (e.g., Setup, Calibration, Treatment, Reports, and System Administration) should be complemented by a user-interface architecture with five major pathways (e.g., menu options) that users may follow to perform the associated tasks.
Guideline 11.4: Menu Depth Generally, people prefer menu systems that are relatively shallow, requiring the user to navigate no more than two or three levels deep in a menu hierarchy to reach the desired content/ options. This approach reduces the chance that users will consider certain features to be “buried” in the software user interface. It also reduces the time required to select a menu option. Scanning and choosing among three menu layers containing 10 options per layer is likely to proceed faster that five menu layers containing six options per layer, in part because of the number of required “clicks” and the need to shift visual focus from one listing to another.
Guideline 11.5: Menu Breadth Generally, people prefer menu systems that do not have an overwhelming number of toplevel options. A medical device with too many top-level options might intimidate new users and can make it more difficult for users to form an accurate mental model of how the device works. Recognizing that user preferences and device requirements vary widely, the optimal number of options seems to fall in the broad range of 3 to 12 options. Some human factors practitioners would advocate an even narrower range of five to nine options, citing limits on the ability of people to mentally manipulate a greater number of options at times when the options might not be displayed. If more than five or six choices are desirable, designers should provide users with additional methods of facilitating easy and rapid selection (e.g., functional grouping).
User-interface structures can be linear, branching, or networked (weblike), as shown in Figure 11.11. Each kind of structure has its advantages and disadvantages in terms of ease of use, task speed, and overall compactness (number of screens). Many user interfaces are a blend or hybrid of the three types.
FIGURE 11.11 Linear (left), branching (middle), and networked or weblike (right) user-interface structures offer different advantages and disadvantages (see text for details).
TAF-ER562X_WINGR-09-0401-C011.indd 439
11/9/10 2:21:27 AM
440
Handbook of Human Factors in Medical Device Design
11.2.2.1 Linear Structure A linear user-interface structure presents screens to the user in a predetermined order. The structure essentially forces the user to view information and perform tasks in the order intended by the designers, although it is usually possible to go backward to repeat a step. The structure tends to prevent users (e.g., nurse programming an infusion pump) from skipping procedural steps or from becoming lost in the screen hierarchy. It is well suited to applications used by individuals with little or no training because it leads users through tasks, thereby reducing the demand on users to choose their own approach to tasks. This is why so-called wizards, which are often used to configure software, usually follow a linear sequence of steps. 11.2.2.2 Branching Structure A branching user-interface structure presents multiple options to users, enabling them to focus on the screen content and tasks of particular interest. The structure relieves the user from having to view all content, regardless of its relevance. As such, a software user interface with a branching structure (e.g., interfaces found on many physiological monitors) might be perceived as simpler than other types because the user is not forced to view all its contents but rather sees only the pertinent elements. The structure tends to enable users to perform tasks quickly. It is also well suited to applications used by several types of trained individuals who might have substantially different task goals. 11.2.2.3 Web (or Network) Structure Web user-interface structures enable users to follow alternative paths to access screen content and options of interest. Rather than starting at the top of the user interface each time to initiate a task (e.g., calibrating an instrument), a user might initiate the task by selecting options presented on a lower-level screen. As such, it is harder to draw a chart that effectively illustrates task flows because of the high number of possible pathways, hence the use of the term “web.” This structure is well suited to applications (e.g., patient information management systems) used by experts who value the ability to perform multiple tasks concurrently and as quickly as possible. In short, a web structure enables users to “jump” from one task to another. Such structures introduce the risk that users will not complete tasks and might ultimately become confused or lost in a stream of screens that appear to have no apparent hierarchy.
11.2.3 INTERACTION STYLE As discussed earlier, computer-based medical devices, depending somewhat on the associated hardware, present designers with a choice of interaction styles. The chosen interaction style not only influences the appearance and number of specific screens but also has a significant effect on task intuitiveness and efficiency. Recapping the preceding discussion, some of the contemporary and legacy interaction styles include question-and-answer dialogue, menu, and direct manipulation, among others. Each style can be implemented in different ways. For example, a menu-driven user interface might call for users to interact with a pointing device such as a mouse, to touch targets on a screen, to press soft keys adjacent to the display, or to speak commands. In fact, the lines usually blur between interaction styles, a good example being using a pointing device
TAF-ER562X_WINGR-09-0401-C011.indd 440
11/9/10 2:21:28 AM
Software User Interfaces
441
to select options from a menu. Such a solution could be classified as a menu-driven or a direct manipulation style of interface because both terms apply. Ultimately, the key to successful user interactions is to choose one or more styles that suit the task at hand. For example, a question-and-answer dialogue might be the best solution to entering data carefully, such as for radiation treatment machine settings. In this case, performing tasks sequentially could prevent use errors that might be more frequent if the user were to directly manipulate multiple settings on a single screen. However, the first approach would probably consume more time, which is why designers need to analyze the trade-offs among interaction styles. A variety of human factors standards and textbooks provide guidance on how to select an appropriate interaction style and implement it properly (see the resources at the end of this chapter). Key guidelines are presented below. Guideline 11.6: Task Pace A device’s interaction style should enable users to perform tasks at the pace necessary to accomplish the intended tasks safely and efficiently. For example, users should not have to endure delays when requesting laboratory results to be displayed on a bedside patient monitor. Neither should they have to repeatedly navigate to a patient data entry page because the associated device “timed out” and automatically returned to a “resting screen” or top-level display.
Guideline 11.7: Pointing Device Compatibility Interaction style should be carefully matched to the type of pointing device so that users can accomplish tasks without hindrance. For example, direct manipulation calls for a pointing device suited to selecting and “dragging” on-screen objects. By comparison, ATM-like soft keys facilitate a menu-based approach.
Guideline 11.8: Interaction Style Consistency While a large proportion of user interfaces employ more than one interaction style, designers should apply them consistently. For example, the same user interface should not require users to select from a comparable set of options by clicking on a “radio button” in one case and typing an option into a blank field in another.
11.2.4 SCREEN LAYOUT The term screen layout refers to the arrangement of visual elements on a given display screen, which can have a significant influence on the quality of user interaction with the elements as well as the device’s visual appeal. There is no magic formula for developing a good screen layout. Generally, screens that appear symmetrical, balance content effectively, and use blank space to set major elements apart are effective designs. But sometimes an asymmetrical arrangement of elements or a high-density layout is just as effective, depending on the users and use scenarios. Accordingly, it makes sense to follow the guidelines for screen layout as a starting point, develop several variants on the basis of professional judgment, and then conduct user tests to determine the best possible solution. Guideline 11.9: Alignment Grid Generally, screen content should ascribe to an alignment grid that gives the screen a consistent, orderly appearance, thereby facilitating rapid scanning and the clear demarcation of
TAF-ER562X_WINGR-09-0401-C011.indd 441
11/9/10 2:21:28 AM
442
Handbook of Human Factors in Medical Device Design
functional groups and information hierarchies. Content placement that does not conform to the grid should do so purposefully. For example, a pop-up dialog box might be deliberately offset from the grid to draw more attention.
Guideline 11.10: Content Hierarchy Screen content should ascribe to a hierarchy appropriate to the selected design schema, the tasks being supported, and the user population. Accordingly, the most important information or the first step in a multistep procedure should be presented in the highest priority (i.e., most conspicuous) location, which is typically the top and/or left side of the screen for most cultures. On very small screens (e.g., one-line LCD displays), the middle position is usually the most conspicuous, although display element size might most affect how well people notice it.
Guideline 11.11: Content Distribution Screens should have a balanced appearance created by a relatively even distribution of content (figure) and blank space (ground). That way, portions of the screen will not appear overly congested or sparse but rather will present content in a manner that is visually pleasing as well as readable.
The following guidelines pertain to the provision of blank space (also called white space or background) surrounding a screen’s major visual elements. They ensure proper spatial separation so that specific visual elements do not look crowded. A crowded or congested-looking screen might intimidate some users and make it more difficult to acquire information. Guideline 11.12: Gutters Readability is usually enhanced by adding a gutter or margin between the screen’s edge and content. To ensure visual separation from the screen’s edge, gutters should be at least a few pixels in width or height. A larger gutter might be warranted if the screen has a bezel (a faceplate that covers the edges of a display and extends to shade the display) and the bezel position is somewhat variable. Note that background fills should go all the way to the edge of the screen.
Guideline 11.13: Padding Padding (blank space) should separate blocks of content (including graphics and text) that share the same background. Padding between on-screen content, such as the space between two columns of text in a newspaper article, helps to reinforce functional groupings and to avoid a congested screen appearance (Figure 11.12). Typically, padding alone is simpler and more visually pleasing than using a line in addition to padding to divide the elements.
11.2.5 LEGIBILITY In many cases, clinicians have misread information displayed on a screen because of legibility problems. Such use errors can cause patient injury or death from incorrect treatment based on the false readings. For example, a patient received a morphine overdose when the number “7” was mistakenly read as a “1” (Figure 11.13). Therefore, extra care must be taken to select a font style, size, and resolution that users will be able to read correctly during the full range of use scenarios, including high-workload and time-pressured periods.
TAF-ER562X_WINGR-09-0401-C011.indd 442
11/9/10 2:21:28 AM
Software User Interfaces
443
Padding
Gutters
FIGURE 11.12 Screens include gutters and padding to produce a clean-looking design that separates related information into functional groups. Demarcation lines separating list of patient names adds unnecessary visually complexity.
FIGURE 11.13 Font design influences the user’s ability to differentiate numbers, such as “1” and “7.” The middle of these rows is least likely to be misread because the numbers are less stylized and important details (e.g., the top of the 7) are more salient.
Guideline 11.14: Text Style On-screen text should have a simple style that is optimized for legibility. Normally, this means using sans serif fonts—letter forms that do not have extra details or “flourishes” to make them look more attractive or to convey a mood (Table 11.2). Common sans serif fonts include Arial and Helvetica, although there are many more. Fonts should have a smooth rather than a
TABLE 11.2 Comparison of Serif (with Flourishes) and Sans Serif (without Flourishes) Fonts Less Preferred Fonts (serif)
More Preferred Fonts (sans serif)
White Background with Black Text Times
Courier
Arial
Helvetica
Black Background with White Text Times
TAF-ER562X_WINGR-09-0401-C011.indd 443
Courier
Arial
Helvetica
11/9/10 2:21:28 AM
444
Handbook of Human Factors in Medical Device Design
FIGURE 11.14 Comparison of normal text and antialiased text (reflecting font smoothing). (a) Text that has a jagged look that might reduce legibility and aesthetics. (b) Text that is antialiased looks smooth and can improve legibility and aesthetics. “jaggy” appearance. Normally, this is accomplished by using “scalable” fonts and a moderate amount of antialiasing (a method of adding shading to otherwise jagged edges to make them look smoother) (Figure 11.14). Avoid using fonts that simulate the look of a readout produced by a segmented display (Figure 11.15).
Guideline 11.15: Text Size On-screen text should be sized to ensure reliable communication at the maximum expected viewing distance. For some applications, such as a digital thermometer, the expected viewing distance will be less than or equal to an arm’s reach away from the face. For other applications, such as a heart rate monitor, the maximum expected viewing distance might be 25 feet or more, as one finds in some critical care environments. The recommended character size of critical information is 1/150th the viewing distance, suggesting that a key parameter value on a patient monitor viewed from 3 feet away should be greater than or equal to a quarter of an inch. The recommended character size of important, but noncritical information is 1/300th the viewing distance, suggesting that a key parameter value on a patient monitor viewed from 3 feet away should be greater than or equal to an eighth of an inch (a 9-point font, given that 1 point equals 1/72 inches). Viewed from a distance of 25 feet, critical and important information would be displayed using characters at least 2 inches tall.
Guideline 11.16: Figure-to-Ground Contrast Text and its associated background should have sufficient contrast to ensure readability but contrast might be increased or decreased according to their relative importance. Critical
FIGURE 11.15 High-resolution displays should present information with maximum legibility using optimized fonts (bottom) rather than mimicking lower-resolution displays, such as segmented LCD readouts (top).
TAF-ER562X_WINGR-09-0401-C011.indd 444
11/9/10 2:21:29 AM
Software User Interfaces
445
information should contrast sharply against its background, suggesting the use of white text on a black background or black text on a white background. Less critical information might be visually subdued by using somewhat lower-contrast pairings, such as medium blue text on a white background. An effective test of the legibility of colored pairings, particularly for people with impaired color vision, is to check the pairing when displayed in gray scale (with the color removed using an editing program or by setting the monitor to display in monochrome). Certain color pairings, such as saturated red characters on an equally saturated blue background, should be avoided because they have a garish appearance and can cause eyestrain because different wavelengths of light focus at different depths in the eye (a visual perception phenomenon called chromostereopsis).
Guideline 11.17: Text Capitalization Using ALL CAPITAL LETTERS can draw attention to important textual information. However, it generally takes people longer to read long strings of capitalized text than lowercase text (Table 11.3) and might increase reading errors. Full capitalization also takes up more space that could be used to accommodate lowercase letters presented in a larger font (as shown in the sample). Therefore, the use of capitalization as a highlighting technique is best limited to individual letters (e.g., capitalizing the key letters of a drug name, such as “DOPamine,” to avoid confusion with “DOBUTamine,” which is called using TALLman letters), words (e.g., OFF), and short phrases (e.g., DO NOT UNPLUG). Capitalization can also be an effective way to indicate the top level of a set of hierarchical labels.
Guideline 11.18: Line Spacing Lines of text should be spaced far enough apart to ensure that a gap of at least 1 pixel or greater exists between the ascending letterforms (e.g., bdfhklt) and descending letterforms (e.g., gjpqy). Additional space (i.e., leading) between lines will make text more attractive and readable by avoiding a crowded appearance (Table 11.4).
Guideline 11.19: Text Justification For languages, text should be left justified with a ragged-right margin. This method of alignment enables rapid scanning. It also makes text appear as a more unified block of information, standing apart from other blocks of information and supporting the goal of creating functional groupings. Full justification (lines of text starting and ending at a uniform point) becomes less readable when such alignment leads to insufficient or excessive space between words.
TABLE 11.3 Long Strings of Capitalized Words Take Longer to Read Than Those Using MixedCase Letters PLACE A TINY DROP OF BLOOD ON THE TEST STRIP, THEN PRESS THE GREEN KEY. YOUR TEST RESULT WILL APPEAR ON THE SCREEN IN ABOUT A MINUTE.
Place a tiny drop of blood on the test strip, then press the green key. Your test result will appear on the screen in about a minute. The lowercase example uses a larger font and still takes up less space.
TAF-ER562X_WINGR-09-0401-C011.indd 445
11/9/10 2:21:30 AM
446
Handbook of Human Factors in Medical Device Design
TABLE 11.4 Comparison of Text with Varying Degrees of Spacing between Lines (Leading) Poor
Acceptable
Better
Insufficient space between lines causes ascenders and descenders to overlap.
Minimum space between lines keeps ascenders and descenders from touching but looks congested.
Sufficient space between lines keeps ascenders and descenders from touching and enhances readability.
Oxygen Saturation Cardiac Output ST Analysis
Oxygen Saturation Cardiac Output ST Analysis
Oxygen Saturation Cardiac Output ST Analysis
Guideline 11.20: String Length Adequate space should be allocated to accommodate text that will be translated into other languages, some of which employ graphical characters requiring more vertical space to ensure clarity (Table 11.5).
Guideline 11.21: Touch Screens The legibility of text appearing on a touch screen can be degraded by fingerprints and other screen surface contamination as well as by the opacity of any overlaying filmlike materials that give a touch screen its sensing capability. Therefore, touch-screen text should be oversized and/or have higher-than-normal contrast with its background. See Chapter 7, “Controls,” and Chapter 8, “Visual Displays,” for more design guidance on touch screens.
Guideline 11.22: Icons (Symbols) Icons should be of sufficient size and detail to ensure proper perception of their form and details from the expected viewing distance. The proper perception of an icon’s form and details is a different objective than ensuring proper icon interpretation, which depends on what the icon’s features communicate (i.e., semantics) rather than its legibility.
11.2.6 AESTHETICS Aesthetics might not be on a par with other software user-interface design objectives (e.g., safety), but the contribution of good aesthetics should not be discounted. A visually TABLE 11.5 Comparison of Text Strings Translated into Four Different Languages Language
Sample Prompt
English German Japanese
To fill the breathing circuit with pure oxygen, press the Green button labeled “O2 Flush.” Um den Atemkreis mit reinem Sauerstoff zu füllen, drücken Sie die grüne mit “O2 Flush” beschriftete Taste. 純粋な酸素で呼吸回路を満たすためには、 “O2 Flush” と表示された緑のボタンを押して下さい
Spanish
Para llenar el circuito de respiración con oxígeno puro, golpe el botón verde etiquetado O2 inunda.
TAF-ER562X_WINGR-09-0401-C011.indd 446
11/9/10 2:21:30 AM
Software User Interfaces
447
appealing screen can be less intimidating to new users and enhance the interactive experience. It can also help users focus on key information rather than being distracted by unattractive elements. Guideline 11.23: Use of Color Besides communicating special meaning and establishing an information hierarchy, designers can use color to give medical software a pleasing and appropriate appearance. A few moderately saturated colors that fit a harmonious color palette tend to give software user interfaces a softened, professional appearance. In contrast, many highly saturated colors tend to give interfaces a harsh, garish, and unappealing appearance. See also Section 11.2.8.
Guideline 11.24: Use of Graphics Users tend to be critical of screens that are packed with text, commenting that the screens look congested and hard to read. By adding graphics, designers can often convey more information in less space than required by text and can give screens visual distinctiveness, thereby facilitating navigation and information retrieval. Generally, graphics should have a useful function rather than being simply decorative.
Guideline 11.25: Screen Density After considering several other factors that should dictate what information belongs on a given screen, decisions about information density should also consider aesthetic effects. Users like screens that have a modest amount of white space separating major screen elements. Accordingly, screens tend to look best when 20% to 30% of the total screen area is blank.
Guideline 11.26: Branding While acknowledging a manufacturer’s need to brand their medical devices, logos and similar branding elements should not interfere with device function. More specifically, they should not distract users or impede them from accomplishing tasks. For example, a logo should consume only a small portion of a top-level software screen and should not compete for attention with important information. Relocating a logo to the hardware and reserving the screen for clinically significant information is usually the best solution.
11.2.7 DATA ENTRY Certain medical devices, such as ultrasound scanners, point-of-care blood gas analyzers, and infusion pumps, can require a substantial amount of data entry. Required data might include patient information, physician name, diagnostic codes, fluid volumes, medication names, or clinical notes. It is important, sometimes even critical, that users enter the information in a complete, accurate, and efficient manner. Data entry errors have the potential to cause patient injury and even death. Moreover, an inefficient data entry process can impair patient safety and reduce clinician productivity. Guideline 11.27: Use of Labels All data entry fields and data sets should be labeled. The use of acronyms and abbreviations should be limited to those that are universally recognized in the given clinical setting (e.g., HR = heart rate and NIBP = noninvasive blood pressure).
TAF-ER562X_WINGR-09-0401-C011.indd 447
11/9/10 2:21:31 AM
448
Handbook of Human Factors in Medical Device Design
Guideline 11.28: Label Placement Labels should be placed adjacent to the associated data, preferably on top or to the left side.
Guideline 11.29: Label Appearance Labels should appear somewhat subordinate to the primary information, presuming that they remain legible at the intended viewing distance (Table 11.6). This focuses the user’s attention on the key, variable information. For example, blue or medium gray text can be used on a white background to label primary information presented as black characters, presuming that the contrast ratio between text and background remains high.
Guideline 11.30: Units of Measure Where applicable, data entry fields should include units of measure (e.g., “psi,” “mmHg,” and “bpm”) of the associated parameter. Such units should be presented in close proximity to the associated parameter value and be sufficiently large to ensure readability but not so large as to compete for attention with the value. Care should be taken to avoid mixing English and metric units, except in cases where medical conventions dictate such mixing.
Guideline 11.31: Data Format Data entry fields should indicate the input format. For example, a medical record number field might be divided into three separate blocks separated by dashes. As another example, the data entry field for a patient’s birth date (e.g., September 10, 1960) can be divided into three separate blocks for the month, day, and year. Moreover, the separate fields can be labeled “mm,” “dd,” and “yy,” thereby guiding the user to input the appropriate two digits in each field (e.g., 09, 10, and 60).
Guideline 11.32: Data Entry Fields Data entry fields should be visually distinct from other information, such as a static presentation of other data. For instance, data fields can appear highlighted (e.g., a recessed window with a white background) compared to other data presentations, such as black text on a light gray background.
Guideline 11.33: Data Entry Field Labeling Data entry label fields should be labeled. Where appropriate, they should include an example of the data entry format (e.g., mm/dd/yyyy for entering the date).
Guideline 11.34: Data Entry Field Size Data entry fields should be sized to accept the largest expected data strings (i.e., words, phrases, or sentences). Extra room should be provided to accommodate translation of the
TABLE 11.6 Typography Can Make the Most Important Information Stand Out
TAF-ER562X_WINGR-09-0401-C011.indd 448
11/9/10 2:21:31 AM
Software User Interfaces
449
software user interface into other languages. Multiple data entry fields appearing together should have a common size to produce a more symmetrical, easy-to-scan appearance, except where the data have a fixed string length, such as a patient’s birth date.
Guideline 11.35: Functional Grouping As appropriate, data should be placed in functional, related groups that lend themselves to hierarchical labeling.
Guideline 11.36: Data Justification Text entered into a data field should appear left justified (Table 11.7). Values entered into a data field might more appropriately be right justified or decimal point justified. Decimal point justification (i.e., alignment) is preferred in cases where users will access precise magnitudes and draw comparisons among vertically stacked values.
Guideline 11.37: Data Arrays Generally, people have an easier time reading narrow columns of data (e.g., 20 rows × 2 columns) rather than wide rows of data (e.g., 2 rows × 20 columns). However, the nature of the data and user population conventions should be considered when choosing the final format. For example, wide rows of data might be more appropriate if the data set is presented in conjunction with a time-based graph of blood pressures that includes a horizontal time axis. Visual aids, such as striping (i.e., alternating line shading) on every other line (Figure 11.16), can be added to arrays to help users trace data along a given row or column.
Guideline 11.38: Automatic Fill-In Reliable data available from other sources (including associated screens in the same application) should fill in automatically rather than requiring retyping, thereby reducing the user’s workload and eliminating the potential for data entry (i.e., transcription) errors. However, users should either enter and/or verify critical values as appropriate.
TABLE 11.7 Comparison of Text Justification and Formatting Schemes Comments Poor Right justifying the parameter values is not worthwhile considering that the text strings do not vary dramatically in length. Decimals points are not aligned. Better Left justifying parameter labels of similar-length text strings makes them easier to read. Rounding off the parameter values avoids excess precision and allows for right justification. Additional enhancements include visual subordination of the labels (use of gray letters) and units (smaller, lowercase letters) and the elimination of abbreviations.
TAF-ER562X_WINGR-09-0401-C011.indd 449
Parameter Display
Press. Vol. Cycle Progr.
12.6 PSI 300.0 PSI 15 minutes 2
Pressure 13 psi Volume 300 mL Cycle Time 15 min Program No. 2
11/9/10 2:21:31 AM
450
FIGURE 11.16
Handbook of Human Factors in Medical Device Design Lastname1 Firstname MI
000-000-0000
100
Lastname1 Firstname MI
000-000-0000
100
Lastname1 Firstname MI
000-000-0000
100
Lastname1 Firstname MI
000-000-0000
100
Lastname1 Firstname MI
000-000-0000
100
Lastname1 Firstname MI
000-000-0000
100
Lastname1 Firstname MI
000-000-0000
100
Lastname1 Firstname MI
000-000-0000
100
Lastname1 Firstname MI
000-000-0000
100
Lastname1 Firstname MI Lastname1 Firstname MI
000-000-0000 000-000-0000
100 100
Striping helps users read across horizontal lines.
Guideline 11.39: Data Checking When appropriate, automatic checks should be made on entered data to confirm their format and that they are within an appropriate, safe range. For example, checks should be made to confirm that the dosage set for delivering a medication does not exceed the levels established by the health care organization. In certain cases, the software should provide the user with the option to override the set limits if necessary to deliver effective medical care.
11.2.8 COLOR The use of color can either enhance or degrade the information acquisition process. A positive result depends on designers using color in a meaningful way rather than as decoration. Curiously, a large proportion of medical professionals downplay the need for aesthetic design while at the same time being drawn toward attractive designs. Therefore, color can play a functional role in user-interface design while also boosting a product’s general appeal. For example, color might be used to differentiate realtime versus historical data. The following guidelines address both functionality and aesthetics. Guideline 11.40: Number of Colors To achieve a pleasing aesthetic, experienced designers try to limit their basic color palette to a small number of colors—perhaps three to five—that have a generally complementary appearance while also ensuring good information legibility. One or more of the colors used for backgrounds and frames may conform to a corporate branding standard color scheme.
Guideline 11.41: Color Conventions Table 11.8 presents some of the conventional uses of color in the USA to code information on medical device displays. Color conventions differ among countries.
Guideline 11.42: Nonreliance on Color Considering that about 10% of some adult populations have impaired color vision, color should not be used as the sole means of coding information. Other possible means of coding include shape, size, and labeling.
TAF-ER562X_WINGR-09-0401-C011.indd 450
11/9/10 2:21:32 AM
Software User Interfaces
451
TABLE 11.8 Color Codes for Medical Applications Color
Meaning
White Red
Conventional practice: Primary information on a black background Alarm: High priority Hazard: Danger (an associated hazard will be deadly; will cause property damage) Conventional practice: Arterial (oxygenated) blood pressure Conventional practice: OFF, power OFF Conventional practice: Stop, emergency stop Conventional practice: Fault condition Conventional practice: Energy being delivered (e.g., laser firing) Conventional practice: Stay clear Association: Warm, hot Orange (amber) Alarm: Medium priority Hazard: Warning (an associated hazard may be deadly or injurious; may cause property damage) Yellow Alarm: Low priority Hazards: Caution (an associated hazard may be injurious; may cause property damage) Gas: Air Conventional practice: Slow Association: Warm, sunny Green Conventional practice: ON, power ON Conventional practice: Go/continue Conventional practice: All OK (normal) Conventional practice: Ready (available for use) Gas: Oxygen Association: Good Association: Environmentally friendly Blue
Pink
Gray Brown Black
Conventional practice: Secondary information on a white background Conventional practice: Deoxygenated lungs or blood Gas: Nitrous oxide Association: Frozen, cold Conventional practice: Oxygenated lungs or blood Association: Healthy Association: Female Conventional practice: Unavailable or nonapplicable option/information Gas: Carbon dioxide Gas: Helium Conventional practice: Primary information on a white background Gas: Nitrogen
Some color conventions defined by: Compressed Gas Association, Inc., Standard Color Marking of Compressed Gas Containers Intended for Medical Use, Arlington, VA, 1988, reaffirmed September 1993.
TAF-ER562X_WINGR-09-0401-C011.indd 451
11/9/10 2:21:32 AM
452
Handbook of Human Factors in Medical Device Design
Guideline 11.43: Color Combinations Information legibility depends on the use of appropriate colors for background and foreground information, taking into consideration hue (wavelength), value (brightness), and saturation (amount of gray mixed into the color) so that the color pair has a sufficiently high contrast ratio. For example, yellow numbers will stand out effectively against a dark blue background, while red letters do not stand out as well against a green background, even when viewed in gray scale (Figure 11.17).
Guideline 11.44: Color Associations Designers should pay close attention to color associations common to the user population and adjust color schemes accordingly. For example, pink coloring might suggest a healthy condition, while dusky gray coloring might suggest an unhealthy one.
Guideline 11.45: Color Customization Because the effectiveness of color codes depends on their widespread use, password protected administrative controls should be included in software user interfaces to prevent users from inappropriately customizing the colors on a given display. This will help prevent critical misinterpretations that could occur if users encountered different versions of the same device within the same instituation.
Guideline 11.46: Demarcating Using Color Color can be used effectively in place of lines and other boundary markers to differentiate functional groups.
Guideline 11.47: Indicating Status Using Color Color can be used to indicate a change in status. For example, a battery symbol might change from green to yellow or amber and then to red to indicate the shift from fully charged to nearly discharged.
11.2.9 DYNAMIC DISPLAYS (SEE CHAPTER 8, “VISUAL DISPLAYS”) 11.2.9.1 Trend Displays The following design guidelines pertain to the design of trend displays, which are commonly found on vital signs monitors, time-based therapeutic devices, and diagnostic devices. Generally, trend displays show how a particular parameter value has changed over
98.6
98.6
98.6
98.6
FIGURE 11.17 (See color insert following page 564.) Red letters on a green background (left) have a lower contrast ratio than red letters on a dark blue background (right), as illustrated when the pairs are converted to gray scale, roughly simulating how an individual with color-impaired vision is likely to perceive it.
TAF-ER562X_WINGR-09-0401-C011.indd 452
11/9/10 2:21:32 AM
Software User Interfaces
453
time. The trend might or might not include real-time data. Some trend displays plot points on an x-y graph. Others show a continuous line (i.e., waveform). With either type of trend display, designers need to ensure that the user will be able to properly interpret the data, distinguish older values from newer ones, determine the units of measure, and extract the necessary level of detail. Guideline 11.48: Displaying Real-Time Data Trend displays should clearly differentiate current values from older values that provide a historical perspective.
Guideline 11.49: Data Resolution Trend data should be presented at a resolution that facilitates interpretation of the data’s meaning. The time frame may be the last few minutes, hours, or even days, depending on various clinical factors. For example, a clinician might need more resolution in an ECG waveform than a CO2 waveform to extract the necessary diagnostic information.
Guideline 11.50: Time Frame Trend data should be presented for a period of time that facilitates determining the clinical requirements. Where the time frame can be adjusted, the means of adjustment should be readily obvious, and the time frame being used should be clearly indicated.
11.2.9.2 Waveform Displays The following design guidelines pertain to the design of waveform displays—time-based presentations of measured values, such as blood pressure. Such tracings are common among medical devices used to monitor critically ill patients continuously during and after medical procedures. Some devices, such as monitors used in open heart surgery, can display six or more traces at a time. Central monitoring workstation displays can show waveforms associated with a dozen or more patients. Distinct from trend displays, waveform displays usually present parameter values over a short period of time, such as the past 20 seconds as opposed to the past 20 minutes. Guideline 11.51: Waveform Color As appropriate, waveforms should be color coded to conform with local, national, and international medical conventions. For example, an arterial blood pressure waveform is typically colored red to associate it with the color of well-oxygenated blood.
Guideline 11.52: Waveform Cycles Waveform displays should present a sufficient number of cycles for users to interpret the data effectively. For example, cardiologists and anesthesiologists want to view at least three or four complete ECG waveforms at a time to derive sufficient diagnostic information about rate and rhythm.
Guideline 11.53: Stopped Motion Users should be able to stop the motion of (i.e., “freeze”) a waveform display in cases where a more detailed and potentially more time-consuming assessment of a specific waveform
TAF-ER562X_WINGR-09-0401-C011.indd 453
11/9/10 2:21:32 AM
454
Handbook of Human Factors in Medical Device Design
FIGURE 11.18 Digital patient monitoring systems enable clinicians to study changes in a patient’s ECG waveform (e.g., perform an ST segment analysis to determine if the patient is becoming ischemic) over time. (Courtesy of Philips Healthcare. With permission.) component is warranted (Figure 11.18). An alternative to “freezing” the display is to provide users with the capability to take a digital “snapshot” at a moment of interest for later review.
Guideline 11.54: Refresh Mechanism Designers should obtain users’ feedback to determine the most effective means of refreshing a waveform display. Either a moving “erase bar” that refreshes a stationary waveform or a moving waveform might be more appropriate for specific applications or might be preferred by certain user populations. When used, an erase bar should be visually distinct so that users can easily detect the location of the most current data. The advantage of a refresh bar is that the user can study details associated with a stationary waveform rather than having to visually track a waveform that is moving across the display. However, most physiological monitors use a moving waveform to more closely emulate ongoing physiological activity.
Guideline 11.55: Waveform Resolution The resolution of waveforms—a function of display resolution and waveform trace (or sweep) speed—should be sufficient for users to extract the necessary clinical information. Notably, a slow-moving trace effectively has fewer display elements (e.g., pixels) with which to provide waveform detail. Conversely, a fast-moving waveform effectively has more display elements, thus providing more detail, although increased trace speed means that it might not be possible to fit the whole waveform on the display. The same resolution trade-offs apply when determining the vertical scale of a waveform.
Guideline 11.56: Line Thickness Waveform line thickness should strike a balance between visibility and resolution. The minimum-thickness (i.e., 1-pixel-thick) line might not stand out from other information,
TAF-ER562X_WINGR-09-0401-C011.indd 454
11/9/10 2:21:32 AM
Software User Interfaces
455
particularly if the line and the background colors do not contrast sharply. However, a multiplepixel line might obscure waveform details. Accordingly, waveform lines should be optimized through user testing to provide maximum diagnostic value.
Guideline 11.57: Background Color Anecdotally, clinicians generally prefer to view white and light-colored waveforms drawn on a black background. Such waveform displays emit less light that might interfere with certain medical procedures, such as eye surgery, that are performed in dimly lit rooms. That said, black backgrounds can be more susceptible to glare. Therefore, both types of backgrounds have their advantages and disadvantages, explaining why design practice varies among manufacturers and why one approach cannot be deemed best. To maintain a good figure-to-ground contrast ratio, the waveforms displayed on a white background need to be darker—even though they might appear to be the same color (Figure 11.19).
Guideline 11.58: Synchronization Generally, multiple waveforms associated with the same physiological process (e.g., the same patient’s cardiovascular system) should be visually synchronized so that they scroll together or refresh at the same time. Thus, a pair of blood pressure waveforms should advance in unison across a screen so that values at a single point of time are aligned (presumably vertically) on the time axis. An exception would be cases where waveform cycles and the required resolution are so different that their synchronization would degrade the readability or clinical value of one or more of the waveforms.
Guideline 11.59: Overlapping Waveforms Generally, waveforms reflecting nominal conditions and the majority of abnormal conditions should not overlap each other because this would make it more difficult to visually trace and interpret waveforms, even if they are differentiated by color. One exception might be situations where a waveform value has “spiked” dramatically beyond its normal levels, crossing into another waveform’s allotted screen space rather than truncating. Some overlap might also be necessary to fit several waveforms on a single display or allow the user to compare the shape or magnitude of related waveforms. Waveforms displayed on a monochrome monitor should not overlap unless (a) the waveform patterns are sufficiently distinctive that there is virtually no chance of misidentification and/or (b) the waveforms are complemented by
FIGURE 11.19 Comparison of waveforms displayed on dark and light backgrounds. The waveform colors from top to bottom are white, red, yellow, green, and blue.
TAF-ER562X_WINGR-09-0401-C011.indd 455
11/9/10 2:21:33 AM
456
Handbook of Human Factors in Medical Device Design
numeric readings that eliminate any chance of waveform misinterpretation. User testing to determine waveform readability is advisable.
Guideline 11.60: Associating Waveforms with Parameters Waveforms should be functionally grouped with associated numeric readings. For example, an ECG waveform typically will be horizontally aligned with the heart rate reading (i.e., numeric value).
Guideline 11.61: Scale Resolution Waveform scales and markings should allow the user to accurately estimate the associated parameter values with sufficient resolution to meet user needs.
Guideline 11.62: Waveform Arrangement For certain applications, users should be provided with the capability to arrange waveforms (and associated numeric values) according to their preference, presumably based on waveform importance or clinical interrelationship. The option to make such rearrangements might best be password protected to prevent inappropriate and unauthorized ones.
11.2.9.3 Numeric Displays The following design guidelines pertain to the design of numeric displays. Guideline 11.63: Numeric Size At a minimum, numeric displays should be legible from the intended viewing distance. However, the size of a particular number should reflect its relative importance as compared with other numeric information presented concurrently on one or more displays within a medical workstation. For example, a patient monitor might display a patient’s heart rate using much larger characters than used to display body temperature.
Guideline 11.64: Numeric Color When shown on a color display, numerics should be colored to match medical industry conventions. The value (lightness versus darkness) of specific colors may be varied to ensure legibility against the given background, noting that medium-value backgrounds can pose legibility problems when using anything other than very dark or very light colored numeric backgrounds. When paired with a waveform or trend plot of the same variable, the same color should be used.
Guideline 11.65: Numeric Placement Numerics should be placed in proximity to associated information, such as waveforms (e.g., a trace of the patient’s expired CO2 level) and dynamic symbols (e.g., an IV fluid bag).
Guideline 11.66: Numeric Alignment Numerics should ascribe to an alignment grid and reflect whatever application and medical industry conventions exist for text justification. A scattered or random-looking arrangement of numerics should be avoided.
TAF-ER562X_WINGR-09-0401-C011.indd 456
11/9/10 2:21:33 AM
Software User Interfaces
457
Guideline 11.67: Numeric Flashing Numeric flash (i.e., blink) rates should be in the range of 1 to 3 Hz, a duty cycle that should catch the attention of a person glancing at the display. In some cases, it might enhance numeric legibility to bias the flash rate (i.e., duty cycle) so that the numeric is on two-thirds of the time and off one-third of the time. Another way to enhance legibility is to alternate the display element from full intensity to a dimmed state rather than completely off. However, flashing might not be the most effective way to draw attention to important information, particularly because flashing necessarily degrades readability compared with continuous data presentation. Alternative methods of highlighting numeric values include adding an outline around the numeric or displaying it in inverse video.
11.2.10 SPECIAL INTERACTIVE MECHANISMS 11.2.10.1 Soft Key User Interfaces The following design guidelines pertain to the design of soft key–driven user interfaces (like those commonly used in automatic teller machines and cellular telephones) that employ unlabeled buttons that map to on-screen labels defining the purpose of the buttons. This style of user interface offers considerable flexibility as well as good tactile feedback with the use of physical keys. Guideline 11.68: Aligning Soft Keys and Content Soft keys should be aligned precisely with their on-screen labels to avoid erroneous associations, particularly in cases where the hardware design is prone to parallax problems (the misalignment of keys and labels that can occur when a user does not view the display exactly head-on). Parallax problems can be partially mitigated by adding leader lines or other onscreen cues that strengthen the visual linkage between soft keys and associated labels.
Guideline 11.69: Coding On-screen information associated with a soft key should share the same coding. For example, on-screen information may be colored green to match a green soft key. Care should be taken to ensure that the colors look the same, which might require hue (wavelength), value (brightness), and saturation (amount of gray mixed with the color) adjustments to produce the perception of equivalence. One could also code the soft key–label pairs in a graphical manner.
Guideline 11.70: Differentiating Soft Key Labels On-screen information associated with a soft key should be visually distinct from other types of on-screen information. Methods of differentiation include placing the information within a box or circle, varying the font or character size, and using inverse video and coloration, as shown in Figure 11.20.
Guideline 11.71: Consistent Association It is generally inadvisable to change the association between on-screen labels and soft keys between data screens. That is, for example, if the leftmost soft key is labeled “ENTER” on one screen, it should be “ENTER” on all screens where that label is used.
TAF-ER562X_WINGR-09-0401-C011.indd 457
11/9/10 2:21:33 AM
458
Handbook of Human Factors in Medical Device Design
FIGURE 11.20 The ventilator’s soft key labels—the numeric values placed at the bottom of the screen—are presented in inverse video to differentiate them as selectable.
11.2.10.2 Control Wheel User Interfaces The following design guidelines pertain to the design of control wheel–driven user interfaces. (Note: Control wheels go by many other names, including encoders and jog wheels.) Guideline 11.72: Coding of Control Wheel Rotation Local user population conventions might call for turning a control wheel clockwise to move a highlight bar downward versus upward on a menu, for example (Figure 11.21). The most appropriate response to rotating a control wheel should be determined by testing its use with the intended user population.
Guideline 11.73: Control Wheel Responsiveness The focus (e.g., highlight or cursor) should respond nearly instantaneously in response to control wheel rotation, regardless of the wheel rotation rate, so that there is no perceivable lag that could lead to under- and overshooting.
Guideline 11.74: Use of Detents Control wheels should have detents (evenly spaced breakthrough points that feel a bit like miniature speed bumps that provide tactile feedback, perhaps in the form of a clicking sensation) to give users a better sense of control over on-screen adjustments. Control wheels lacking detents can be more difficult to use when making discrete adjustments to on-screen elements.
Guideline 11.75: Number of Detents Typically, control wheels have a “power-of-two” number of detents. Accordingly, most control wheels have 8, 16, or 32 detents spread evenly across the wheel’s 360 degrees of rotation. The number of detents should be chosen to give the control wheel a good “feel” for its size (specifically its diameter) and so that users can move efficiently among the on-screen elements. Accordingly, smaller-diameter wheels should have fewer detents than larger-diameter wheels, making it possible for users to rotate the wheels the same incremental distance along its outside edge. There should be a good match between number of detents per 360 degrees of wheel rotation and the speed with which the user will want to move among on-screen targets and adjust numeric values, for example.
TAF-ER562X_WINGR-09-0401-C011.indd 458
11/9/10 2:21:33 AM
Software User Interfaces
FIGURE 11.21
459
Alternative on-screen responses to rotating a control wheel clockwise.
Guideline 11.76: Numerical Adjustment When adjusting numeric values, rotating the wheel a fixed amount—from one detent to another, for example—should produce a consistent, proportional change in the on-screen element. For example, when a user rotates a control wheel and feels three “clicks,” the clicks should produce a proportionally consistent change in an associated parameter value (i.e., increase a heart rate alarm from 120 to 121 to 122 to 123 beats per minute. The exception to this guideline is when a widely ranging value requires precise adjustment (see next guideline).
Guideline 11.77: Acceleration Algorithms Because control wheels have detents that suggest consistent proportional change, acceleration algorithms (e.g., changing a value faster if one turns the wheel more abruptly) should be avoided if possible. However, when, as is often the case, a control wheel must be used to control a variable over a wide dynamic range (several orders of magnitude), an appropriate acceleration algorithm should be selected on the basis of user testing of alternatives.
Guideline 11.78: Highlight or Cursor Size The moving highlight (i.e., cursor) should be sufficiently large that it is easy to detect its change/movement while rotating the control wheel. In many cases, the highlight might be naturally large, taking the form of an entire block containing a numerical value. However, a typical size cursor (e.g. a thin vertical line matching the height of associated text) is too small for most medical applications. A blinking curser will facilitate visual acquisition.
Guideline 11.79: Two-Dimensional Movement While control wheels alone are not effective solutions for software applications that require precise two-dimensional movement, a hybrid control wheel/joystick provides the added capability (Figure 11.22). However, such a multivariate control might be difficult to use, so its suitability should be determined through user testing.
11.2.10.3 Touch-Screen User Interfaces The following design guidelines pertain to the design of touch screen–based user interfaces (see also Chapter 7, “Controls”). A touch screen–based user interface presents certain design freedoms and constraints. A touch-screen interface offers a more direct mapping of controls to on-screen information than do rotary knobs, trackballs, or mice. A touch-screen user interface can present users with only
TAF-ER562X_WINGR-09-0401-C011.indd 459
11/9/10 2:21:33 AM
460
Handbook of Human Factors in Medical Device Design
(a)
(b)
FIGURE 11.22 (a and b) A patient monitor is equipped with a control wheel with a built-in joystick that permits precise, two-dimensional cursor movement. (Courtesy of Philips Healthcare.)
those controls required to perform a selected task as opposed to a physical control panel that contains all the controls a user might need to perform myriad tasks. However, a touch screen might limit the speed of data input tasks requiring a keyboard, and touch target sizes might limit the amount of information that fits on a single screen when compared to less direct selection mechanism, such as a mouse/trackball. Also, use of a touch screen can be problematic when hands are contaminated with blood or other substances. Accordingly, designers should carefully consider the nature of the user tasks, the use environment, and general maintenance requirements before committing to a touch screen–driven user interface. Guideline 11.80: Activation States Touch-screen targets should have a range of appearances to differentiate when they are unselected, selected but not actuated, and actuated (like a “latching” button), as shown in Figure 11.23. Methods to distinguish the activation states include inverting the button and highlighting it (making the “table” and surrounding bevel a lighter shade than normal). Threedimensional targets are preferable to two-dimensional targets because they are visually distinct from nonselectable objects on the display (Figure 11.24).
Guideline 11.81: Target Size Touch-screen targets should be sufficiently large to facilitate rapid, error-free inputs by individuals with large fingers. Similar to physical keys on a keypad, a target size of no less than
FIGURE 11.23 Variation in the appearance of touch targets depending on their activation state. From left to right, these illustrate buttons that are unselected, selected (finger in contact), and actuated (latched).
TAF-ER562X_WINGR-09-0401-C011.indd 460
11/9/10 2:21:34 AM
Software User Interfaces
FIGURE 11.24
461
Comparison of flat-looking versus three-dimensional-looking buttons.
0.5 inches (1.3 cm) is preferred. Sometimes, it is advantageous to oversize the touch area so that the associated target is actuated if the user touches anywhere close to it, as shown in Figure 11.25.
Guideline 11.82: Target Spacing Generally, the centers of touch-screen targets should be spaced 0.75 inches (1.9 cm) apart to help users avoid pressing the wrong target. However, it might be necessary to reduce the spacing to accommodate more targets on a small display. In such cases, other error-prevention methods (e.g., liftoff actuation or strong highlighting) should be employed to counteract the potential for users to touch the wrong target.
Guideline 11.83: Response Time There should be no perceptible delay between touching an on-screen target and receiving visual and, possibly, audible feedback.
Guideline 11.84: Consistent Target Placement Whenever possible, place touch targets in the same location on every screen. This allows users to develop “muscle memory,” which can increase reliability and speed. It will also reduce use errors such as the user touching the wrong target because the desired selection was in that same spot on another screen.
Guideline 11.85: Target Placement When a given medical device is likely to be placed at or above the user’s normal line of sight, it is advantageous to locate touch targets on the lower portion of the screen where they are easier to reach. This has the added benefit of enabling users to touch the targets without blocking other portions of the screen with their hand.
Guideline 11.86: Dragging To the extent possible, avoid the use of scrolling lists and equivalent slider bars that require dragging one’s finger across the screen. People tend to have difficulty sliding their fingers across flat surfaces and stopping at a precise spot. Also, dragging tends to smear the screen.
FIGURE 11.25 Comparison of minimum versus oversized touch targets (the area within the dashed rectangle).
TAF-ER562X_WINGR-09-0401-C011.indd 461
11/9/10 2:21:35 AM
462
Handbook of Human Factors in Medical Device Design
FIGURE 11.26 For touch screens, discrete input keys are preferred over a slider for scrolling through a list. A slider bar is hard to control precisely and smears a touch screen. However, a slide bar can be paired with arrow keys, for example, to provide effective control over a scrolling list.
Guideline 11.87: Scrolling Where scrolling is required, buttons are generally easier to use than a slider to scroll up and down (Figure 11.26). However, even though the dragging action associated with a slider is more difficult to control precisely and causes smear marks, a slider can be used to facilitate moving rapidly through a set of options. A slider also offers the advantage of indicating one’s place within a scrolling list.
Guideline 11.88: Number Presets Where appropriate, touch screens should present users with an array of preset options rather than requiring them to scroll or toggle through multiple options. For example, instead of adjusting a rate from a value of 10 to 50 by pressing and holding down a virtual key as it increments, one might present 10 presets ranging from 0 to 100 in increments of 10 and then enable the user to adjust the value more precisely using arrow keys, for example.
Guideline 11.89: Graphical Buttons Rather than placing an icon on a button, it can be visually simpler and more appealing to make a graphic (icon) the actual button, as shown in Figure 11.27. This approach enables one to enlarge the graphic without using up more screen space. However, depending on the target’s shape, the total target area could be diminished by abandoning the basic key shape.
Guideline 11.90: Audible Feedback It is often desirable to provide audible feedback, such as a soft “click” or “beep” when a user touches an on-screen target. Different sounds can be associated with valid versus invalid selections. However, because added noise can be a source of distraction and annoyance in medical environments where there might be the need for quiet or where there is already a lot of noise, users should be given the option to turn off the audible feedback.
FIGURE 11.27 looking button.
Sometimes a graphical button is preferable to placing a graphic on a standard-
TAF-ER562X_WINGR-09-0401-C011.indd 462
11/9/10 2:21:35 AM
Software User Interfaces
463
Guideline 11.91: Actuation Use errors can be reduced by employing a “liftoff” rather than “touchdown” method of actuation. Employing a liftoff approach, selections take effect only when users remove their fingers from the screen. This approach enables users to deselect an option by sliding their fingers off a given target. The selected target would change its appearance (e.g., a key appears pressed in) on touchdown to acknowledge contact and then provide explicit feedback on liftoff, such as adding a letter to a data entry field. The dynamic experience can be enhanced by adding distinctive sounds (different tone “clicks,” for example) upon touchdown and lift-off. Another strategy to reduce errors is to create a touch-reactive “hot zone” that is larger than the visual target, but only when enlarging the actual target will not induce other kinds of errors, such as unexpected actuations due to touching blank spaces.
11.2.10.4 On-Screen Keyboards and Keypads While on-screen keyboards and keypads can be displayed when needed to enter alphanumeric data, if an application requires frequent alphanumeric data entry, the on-screen keyboard and keypad should be present at all times to avoid the time-consuming and potentially annoying need to request them. Guideline 11.92: Inclusion of Number Keys Keyboards should include number keys except in cases where there is insufficient vertical space on the screen.
Guideline 11.93: On-Screen Keyboard Layout A QWERTY key arrangement is best suited to most medical workers who are familiar with computer applications requiring the use of a keyboard. While a classic QWERTY arrangement that presents rows of keys in an offset pattern might be optimal because of its familiarity, moving the keys into a grid (Figure 11.28) might not negatively affect keyboard usability. However, an alphabetical key arrangement might be required on touch screens that have insufficient width, such as those found on “portrait” displays that are taller than they are wide, to fit a reasonably sized keyboard. Alternatively, keypads may assume the normal layout found on telephones and may be presented as a configuration option.
Q
W
A
E
S
Z
D
X (a)
C
Q
W
E
A
S
D
Z
X
C
(b)
FIGURE 11.28 QWERTY keyboard arrangements that include versus exclude offsets reflect a trade-off between input speed and space requirements. (a) A classic QWERTY keyboard arrangement, which includes offset rows of keys, may afford the best typing performance but requires more screen area. (b) A QWERTY keyboard arrangement that ascribes to a grid affords good typing performance and requires somewhat less screen area.
TAF-ER562X_WINGR-09-0401-C011.indd 463
11/9/10 2:21:35 AM
464
Handbook of Human Factors in Medical Device Design
11.2.10.5 Speech-Emitting User Interfaces The following design guidelines pertain to the design of speech-prompting user interfaces, such as defibrillators intended for use by the general public and glucose meters designed for use by individuals with visual impairments. A key advantage of speech-prompting user interfaces is that users can focus on other things while listening to instructions and other kinds of audible feedback rather than having to read instructions on the screen. For example, someone using a defibrillator can follow a set of prescribed steps, including attaching electrical leads, without repeatedly looking back to a display. This can save precious seconds in the performance of an emergency procedure while also helping the user concentrate on the hands-on tasks and bolstering their confidence. Speech prompts also allow users to sequentially follow spoken directions, relieving them from having to memorize procedures. However, in some situations, speech prompts could increase the burden on users’ memories because the signal does not persist like a displayed message. Speech prompts are also viable only if the user understands the spoken language, the sounds will be heard clearly in the use environment, and there are no other auditory stimuli (e.g., lots of talking) in the environment. Thus, user testing of speech prompts is always advisable. Guideline 11.94: Source of Audible Prompts Where possible, spoken prompts should be analog or digital voice recordings rather than synthesized speech because voice recordings are generally easier to understand.
Guideline 11.95: Tone of Voice Prompts The tone of voice prompts should be tailored to suit the medical device user and use scenarios. Some users and uses might be best served by voice prompts that sound directive (more severe and insistent) as opposed to advisory (more relaxed). Also, some users might want to hear a female versus male voice because of subjective preference or their ability to discriminate a certain tone of voice against the background noise. This need suggests giving users a choice of voice gender if possible. Notably, a synthesized male voice may be more understandable than a synthesized female voice, and in an emergency situation, a female voice may be more attention getting. Regardless, the tone of voice should be tested to ensure that users consider it to be appropriate and intelligible.
Guideline 11.96: Delivery Rate of Voice Prompts Voice prompts should be delivered at a rate of about 170 words per minute to ensure comprehensibility.
Guideline 11.97: Voice Prompt Volume The volume of voice prompts should be sufficient to be intelligible given the ambient noise levels. Some applications benefit if volume adjustment and/or a mute controls are provided. However, users should not be allowed to suppress or reduce the attention-getting ability of critical prompts (e.g., a spoken alarm message).
Guideline 11.98: Conciseness of Voice Prompts Voice prompts should be as brief as possible while communicating information reliably without ambiguity.
TAF-ER562X_WINGR-09-0401-C011.indd 464
11/9/10 2:21:35 AM
Software User Interfaces
465
Guideline 11.99: Repetition For many applications, voice-enabled devices should provide users with the option of repeating a message, particularly if the given device will be used in a noisy environment and/or users are subject to distraction. Although potentially annoying, critical messages—or some variant designed to be even more attention getting—should automatically repeat at appropriate intervals until the user responds appropriately.
11.2.11 USER SUPPORT Software user-interface designers must balance giving users enough supporting information, such as prompts, to perform tasks properly without causing information overload. One solution is to both segregate and subordinate the supplemental information, such as presenting prompts or instructions at the bottom of the screen in a relatively lower-contrast, smaller but still legible font. Another solution is to allow users to request additional information that goes beyond basic instructions as needed. A third option is to provide help only upon request by means of a “Help” key or the equivalent. The following guidelines describe ways to provide helpful information to medical device users without impeding their work or causing distraction or annoyance. Guideline 11.100: Pop-Up Messages Larger displays provide enough space to present pop-up messages as needed to provide user guidance, for example, to correct a problem, such as proceeding with a task before completing a data entry form. The number of such pop-ups should be kept to the minimum necessary to offer essential guidance and protect against actions that would waste a considerable amount of time or compromise safety. While pop-ups offer the advantage of prompting (or requiring) user action at a given point in an interactive sequence, they can cover underlying information and be annoying or distracting, especially to expert users. Accordingly, pop-ups should be used at times when a specific user action is required, and they should be placed where they do not hide important information. A perceptibly excessive number of pop-ups is one indication that user tasks and associated user-interface elements are not optimally designed and may need to be restructured.
Guideline 11.101: Conciseness Information intended to guide the user should be meaningful and concise. It helps to keep paragraphs and sentences short and to use terminology familiar to the user (Table 11.9).
TABLE 11.9 Meaningful and Concise Messages Facilitate User Tasks, Particularly Recovery from Errors and Malfunctions Poor
Better
The pump has unexpectedly stopped because of pump error condition 21. Under no conditions should the operator attempt to restart the pump unless the error condition has been resolved, in which case pumping may be resumed.
Pump has stopped because the fluid reservoir is empty. Refill the reservoir before restarting.
TAF-ER562X_WINGR-09-0401-C011.indd 465
11/9/10 2:21:36 AM
466
Handbook of Human Factors in Medical Device Design
Guideline 11.102: Differentiation Supporting information should be visually distinct from primary information so that the user’s eye goes to the primary information first.
Guideline 11.103: Style Prompts, instructions, and similar supporting information should employ a consistent style. While no single style is best for all applications, the styles presented in Table 11.10 are usually effective. The infinitive construction, which starts with the expression “To [accomplish the stated goal],” offers the advantage of stating the result of an action before prompting action. This writing style can avoid errors associated with acting in response to a prompt before reading the consequence of the action.
Guideline 11.104: Graphics In many cases, a graphic can be more effective than words at guiding users, particularly when members of the user population might have low literacy skills or not speak the native language very well. As with written prompts, such graphics should be as simple as possible, focusing the user’s attention on the most important details. For this reason, a line drawing will often be more effective than a photograph because photographs usually include many extraneous details (Figure 11.29). Graphics should be validated through user testing to ensure that they communicate effectively and are not subject to critical misinterpretations.
Guideline 11.105: Animations Animations sometimes provide a superior means to guide users. For example, an animation might be preferable as the means to show users how to calibrate a sensor or replenish fluids. A simple animation might be superior to video because it eliminates unnecessary or unwanted detail, focusing the user’s attention on the most important details.
11.3 CASE STUDIES The following case studies exemplify the effective application of human factors principles to the design of software user interfaces for a diabetes management system and a neuromodulation system.
TABLE 11.10 Alternative Wording for User Prompts Poor: No definitive or consistent style; poor writing
Acceptable: Describe the action, Better: Describe the goal, then the goal. then the action.
The Red Button will start up the pump. Press the Red Button to start the To start the pump, press the Zeroing the pressure sensor requires pump. Red Button. pressing the Zero Key. Press the Zero Key to zero the To zero the pressure sensor, When the battery is charged up to the pressure sensor. press the Zero Key. maximum level, you can switch over Be sure the battery is fully charged Before switching from AC to from AC power to DC power. before switching from AC to DC DC power, be sure the battery power. is fully charged.
TAF-ER562X_WINGR-09-0401-C011.indd 466
11/9/10 2:21:36 AM
Software User Interfaces
467
FIGURE 11.29 Line drawings show only the important details. A photograph of the same scene would include many unnecessary details, such as the bed frame and bed controls.
11.3.1 DIABETES MANAGEMENT SYSTEM Roche Diagnostics is a global company that produces a broad line of medical technologies, including diabetes management devices (e.g., glucose meters, insulin pumps, and infusion sets) and associated software products. The company’s diabetes management software product, called Accu-Chek 360º, helps people with diabetes track their blood glucose values as well as other important physical parameters, such as weight, insulin intake, and blood pressure. The core of the software is a multitab set of graphs and reports (Figure 11.30). One of the application’s main purposes is to replace manually filled logbooks with easy-to-use software to transfer data directly from users’ glucose meters and/or insulin pumps, thereby eliminating the chance of data recording errors. Data visualizations help users better understand the effect that their physical activity, eating habits, and medications have on their blood sugar levels. The application produces reports that users can take to their health care professionals so that they get a better picture of their patients’ disease management efforts. Users navigate the application using a computer mouse, interacting with familiar graphical user-interface widgets, such as drop-down lists and hyperlinks (similar to a Web site). Accordingly, the application’s interactive style is familiar to people who use personal computers at home and work. Software user-interface features that contribute to this software’s ease of use include: • A setup wizard that helps users install the application on their computer and customize it to their needs and preferences • Screen layouts that provide direct access to the most frequently used and critical functions • Algorithms that detect significant changes to insulin pump settings and that require users to confirm changes that could affect patient safety • Controls that alert users to incorrect settings on their insulin pump and provide instructions for resolving them
TAF-ER562X_WINGR-09-0401-C011.indd 467
11/9/10 2:21:36 AM
468
Handbook of Human Factors in Medical Device Design
FIGURE 11.30 Three display screens of a diabetes management system: standard day, summary, and insulin pump use. (Courtesy of Roche Diagnostics Corporation. With permission.)
• An autodetection capability that enables users to connect one of a half dozen different glucose meters and insulin pumps to the software’s host computer via a cable and special communication device, enabling data to then transfer directly into the software application • Controls enabling users to automatically generate their favorite reports each time they use the application • The use of harmonious and subtle colors to differentiate information groups • Constrained information density so that screens do not look congested and, consequently, intimidating, especially to new users • Limited number of controls options on screens, again to avoid intimidating users • Online help, including animated tutorials, to guide users through tasks In the course of developing the application, the company conducted extensive user research on how people with diabetes and caregivers, such as diabetes nurse educators, collect and use patient data. This research helped shape the initial user-interface design. Importantly, testing also helped to identify and resolve several design shortcomings that led to use error. Then the research team conducted a series of usability tests in multiple countries to identify opportunities for user-interface design refinement and to validate the final design.
11.3.2 NEUROMODULATION SYSTEM Uroplasty, Inc., is a global medical device company that develops, manufactures, and markets devices that address voiding disorders, such as urinary incontinence, and
TAF-ER562X_WINGR-09-0401-C011.indd 468
11/9/10 2:21:36 AM
Software User Interfaces
FIGURE 11.31
469
A neuromodulator stimulator. (Courtesy of Uroplasty, Inc. With permission.)
pelvic disorders. The company’s Urgent® PC is a telephone handset–sized device (Figure 11.31) that delivers electrical stimulation to a patient’s tibial nerve, which runs through the ankle, from which current travels to the sacral plexus within the pelvis. The device, which is typically used in a doctor’s office, offers patients an alternative to drug therapies that can have undesirable side effects or, in some cases, nontreatment. As such, the device can have a very positive impact on a patient’s quality of life and sense of dignity. The software user-interface challenge was to prompt users to operate the device properly and to provide appropriate feedback using a small, segmented LCD screen (Figure 11.32). Unlike dot-matrix LCD displays, which provide a large array of pixels, segmented displays allow only a limited set of dedicated symbols—or subelements therein—to be displayed or hidden. The technology allows for smooth rather than jagged-looking graphics, such as a battery or electrical spark symbol, but it does not allow for multiple words to be displayed using the same screen space, for example.
FIGURE 11.32
Sample LCD screens that appear during different stages of stimulator use.
TAF-ER562X_WINGR-09-0401-C011.indd 469
11/9/10 2:21:36 AM
470
Handbook of Human Factors in Medical Device Design
Designers explored multiple screen layouts for the following information: • • • • •
An indication that a stimulation level (i.e., current) test was in progress Time remaining in a 30-minute treatment Stimulation level Battery level An indication that electrical stimulation was either on or off
The final layout establishes a strong link between the stimulation level adjustment controls (“+” and “–” hard keys) and the displayed stimulation level. It also groups together the indications that stimulation is active (a moving arrowhead) with the remaining stimulation time (a countdown timer) on the display’s right side. The battery symbol is displayed constantly on the display’s left side.
RESOURCES Ameritech Web Page User Interface Standards and Design Guidelines. http://www.ameritech. com:1080/corporate/testtown/library/standard/web_guidelines/index.html. ANSI/HFES 200: Human Factors Engineering of Software User Interfaces, Santa Monica, CA: Human Factors and Ergonomics Society, 2008. Cooper, A. About face: The Essentials of User Interface Design. Foster City, CA: IDG Books, 1995. Helander, M., Landauer, T., and Prabhu, P. (Eds.). Handbook of Human-Computer Interaction. Amsterdam: North-Holland, 1997. International Standards Organization (ISO). ISO 9241-100 series—Software ergonomics. Jacko, J. A., and Sears, A. (2003). The Human-computer Interaction Handbook: Fundamentals, Evolving Technologies, and Emerging Applications. Mahwah, NJ: Erlbaum. Macintosh Human Interface Guidelines. http://www.usability.gov/pdfs/guidelines.html. Accessed on 6-2-09. National Cancer Institute. Research-Based Web Design and Usability Guidelines. http://www. usability.gov/pdfs/guidelines.html. Accessed on 6-2-09. Sears, A. and Jacko, J. A. (2009). Human-computer Interaction: Designing for diverse users and domains. Boca Raton, FL: CRC Press. Shneiderman, B. Designing the User Interface: Strategies for Effective Human-Computer Interaction. Reading, MA: Addison-Wesley, 1997. Sun Guide to Web Style. http://www.usec.sun.com/styleguide. U.S. Department of Defense. (1996). Human Engineering Design Criteria for Military Systems, Equipment, and Facilities. MIL-STD-1472F. Washington, DC: U.S. Department of Defense. Web Content Accessibility Guidelines 2.0.2008. http://www.w3.org/TR/WCAG. Accessed on 6-2-09. Yale C/AIM Web Style Guide. http://info.med.yale.edu/caim/manual.
REFERENCES Compressed Gas Association, Inc. “Standard Color Marking of Compressed Gas Containers Intended for Medical Use,”Arlington, VA, 1988, reaffirmed September 1993. Miller, G.A. (1956) The magical number seven, plus or minus two: some limits on our capacity for processing information. (http://www.musanim.com/miller1956/) Psychological Review, 63, 81–97.
TAF-ER562X_WINGR-09-0401-C011.indd 470
11/9/10 2:21:37 AM
12 Workstations Michael E. Wiklund, MS, CHFP CONTENTS 12.1 Considerations........................................................................................................476 12.1.1 Safety .........................................................................................................477 12.1.1.1 Protect Users from Hazards .....................................................477 12.1.1.2 Protect against Use Error..........................................................477 12.1.2 Usability ....................................................................................................478 12.1.2.1 Avoid Excess Complexity .........................................................478 12.1.2.2 Allocate Functions Appropriately to the User(s) versus the Workstation...............................................................................478 12.1.2.3 Arrange Controls and Displays to Facilitate User Tasks ..........478 12.1.2.4 Accommodate the Users’ Physical Characteristics ..................479 12.1.2.5 Provide Affordances .................................................................480 12.1.2.6 Consider the Intended Use Environment(s) ..............................480 12.1.3 User Satisfaction ........................................................................................480 12.1.3.1 Choose an Appropriate Visual Style ........................................480 12.1.3.2 Add Refinements ......................................................................481 12.2 Special Considerations ...........................................................................................482 12.2.1 Workstation Uses .......................................................................................482 12.2.1.1 Serve Life-Critical Purposes ....................................................482 12.2.1.2 Clinical Practices Evolve ..........................................................482 12.2.1.3 Users Assume Unusual Positions..............................................482 12.2.2 Workstation Users......................................................................................483 12.2.2.1 Users Can Possess Varying Degrees of Skill, Training, and Experience ................................................................................483 12.2.2.2 Workstations Might Be Used by Individuals with Disabilities ................................................................................483 12.2.2.3 Patients Might Be under Stress.................................................483 12.2.2.4 Caregivers Can Be under Stress ...............................................484 12.2.3 Workstation Use Environments .................................................................484 12.2.3.1 Migration from Hospital to Home Use .....................................484 12.2.3.2 Frequent or Infrequent Cleaning ..............................................485 12.2.3.3 Need for Compactness ..............................................................485 12.3 Design Principles ...................................................................................................485 12.3.1 Operational Factors ...................................................................................486 12.3.1.1 Modes of Operation ..................................................................486 471
TAF-ER562X_WINGR-09-0401-C012.indd 471
11/9/10 12:21:13 AM
472
Handbook of Human Factors in Medical Device Design
12.3.2
12.3.3
12.3.4
12.3.5
12.3.6
12.3.1.2 Use Error Prevention ................................................................487 12.3.1.3 Automatic versus Manual Control ............................................489 12.3.1.4 Patient and User Safety and Security .......................................489 12.3.1.5 Power Supply ............................................................................491 12.3.1.6 Readiness ..................................................................................492 12.3.1.7 Security ....................................................................................493 12.3.1.8 Privacy ......................................................................................493 Communication .........................................................................................494 12.3.2.1 Alarms ......................................................................................494 12.3.2.2 Warnings ..................................................................................494 12.3.2.3 Labeling ....................................................................................495 12.3.2.4 Instructions for Use ..................................................................497 Component Configurations ........................................................................498 12.3.3.1 Demarcation .............................................................................502 12.3.3.2 Display Integration (See Chapter 8, “Visual Displays”) ...........503 12.3.3.3 Storage Space ...........................................................................505 12.3.3.4 Job Aids ....................................................................................506 12.3.3.5 Cable (Wire and Tube) Management (See Chapter 9, “Connections and Connectors”) ......................507 12.3.3.6 Housings ...................................................................................508 Physical Interaction ...................................................................................508 12.3.4.1 Anthropometric Characteristics (See Chapter 4, “Anthropometry and Biomechanics”) ......................................509 12.3.4.2 Physical Accessibility ............................................................... 510 12.3.4.3 Multiple Users .......................................................................... 511 12.3.4.4 Clinician and Patient Position................................................... 511 12.3.4.5 Line of Sight .............................................................................512 12.3.4.6 Handedness............................................................................... 513 12.3.4.7 Repetitive Motion and Cumulative Trauma (See Chapter 4, “Anthropometrics and Biomechanics” and Chapter 16, “Hand Tools”) ........................................................................... 513 12.3.4.8 Compactness ............................................................................. 514 12.3.4.9 Mobility (See Chapter 17, “Mobile Medical Devices”) ............ 514 12.3.4.10 Stability .................................................................................... 517 12.3.4.11 Adjustability ............................................................................. 517 User Accommodations .............................................................................. 518 12.3.5.1 Seating ...................................................................................... 518 12.3.5.2 Hospital Beds and Examination Tables ....................................520 12.3.5.3 Work Surfaces ..........................................................................523 12.3.5.4 Keyboards (See Chapter 7, “Controls”) ....................................524 12.3.5.5 Foot Controls (See Chapter 7, “Controls”) ...............................526 12.3.5.6 Remote Controls (See Chapter 7, “Controls”) ..........................527 12.3.5.7 Grips and Handles ....................................................................528 12.3.5.8 Supports and Restraints ............................................................530 Surface Characteristics ..............................................................................530 12.3.6.1 Appearance...............................................................................530 12.3.6.2 Color ......................................................................................... 531
TAF-ER562X_WINGR-09-0401-C012.indd 472
11/9/10 12:21:13 AM
Workstations
473
12.3.6.3 Material Finish .........................................................................532 12.3.6.4 Cleanliness ...............................................................................532 12.3.6.5 Maintenance .............................................................................534 12.3.7 Environmental Factors (See Chapter 3, “Environment of Use”) ...............534 12.3.7.1 Task Lighting ............................................................................535 12.3.7.2 Noise (See Chapter 3, “Environment of Use”) .........................535 12.3.7.3 Vibration ...................................................................................535 12.3.7.4 Venting .....................................................................................536 12.4 Case Studies ...........................................................................................................536 12.4.1 Anesthesia Workstation .............................................................................536 12.4.2 Ultrasound Imaging Workstation ..............................................................537 12.4.3 Hospital Bed ..............................................................................................539 Resources .........................................................................................................................540 References ........................................................................................................................541 Medical care environments are populated with special purpose medical workstations, such as those enabling ultrasound imaging, anesthesia delivery, and blood chemistry analysis, to name just a few. Designing medical workstations according to accepted human factors principles and practices should contribute substantially to the safety, effectiveness, and efficiency of the associated medical procedures. It should also contribute to the physical and emotional well-being of the workstation users, whether they are caregivers, technicians, maintainers, or patients. Therefore, an investment in the human factors of a medical workstation is clinically beneficial and can provide a measurable business payoff to the manufacturer (e.g., increased market share, reduced need for customer support, and reduced risk of product liability claims). Looking beyond basic functional requirements, operating a medical workstation should also be a satisfying experience. Users, such as an ultrasound technician, a nurse anesthetist, and an electrophysiologist, should feel empowered by their particular workstations because they help them accomplish tasks in an effective and efficient manner. A well-designed workstation should be easy to set up and use. It should place information and controls where they are needed at the right moments during a given medical procedure. Both clinicians and patients should be physically and emotionally comfortable and protected against hazards throughout all phases of workstation operation. But most important from a safety perspective, workstations should avert use errors or at least help the user overcome errors that occur. In summary, a good workstation should function like a trusted assistant, helping to make tasks proceed smoothly and successfully. Unfortunately, not all workstations achieve this ideal, particularly those that are a haphazard assemblage of originally independent components (i.e., improvised workstation). A workstation with inferior ergonomic design can, for example, give the user a backache after just a short period of use. A workstation incorporating a confusing software user interface might frustrate the user by presenting a confusing set of options or insufficient feedback on how tasks are progressing. A poorly designed workstation can increase task performance time if, for example, the component layout does not complement the expected sequence of use. Controls can be prone to accidental actuation if they are located where they can be bumped and/or they lack guards against such contact. Warnings can be poorly worded or missing altogether, increasing the likelihood and consequences of misuse.
TAF-ER562X_WINGR-09-0401-C012.indd 473
11/9/10 12:21:14 AM
474
Handbook of Human Factors in Medical Device Design
So, the lesson from decades of design experience and workstation use is clear; developers have many reasons to invest research and development resources in the human factors engineering of medical workstations. Other factors being equal, customers are far more likely to choose a user-friendly workstation over one that is difficult to use. Error-resistant workstations will shield a manufacturer from product liability claims by reducing the chance of mishaps. Easy-to-use workstations will lessen the burden on training departments and user manual writers because there is less to explain. Meanwhile, customers will reap the benefits of greater worker health, effectiveness, productivity, increased satisfaction and morale, and improved care efficiency and safety, all of which will enhance a health care delivery organization’s bottom line. Before presenting workstation design considerations and guidelines, a more detailed definition of terminology is warranted. You might ask, What is a workstation? How do workstations differ from the things we call systems, workplaces, workspaces, machines, devices, products, and tools? Arguably, the differences are partly a matter of perspective and semantics. While one person might refer to a magnetic resonance imaging (MRI) scanner as a workstation, another person might refer to it as a complex tool or workspace. The ambiguous terminology is due to the fact that medical technologies spread across a continuum of functional and physical complexity that has no strict boundaries (Figure 12.1). For the purposes of this handbook, the term “workstation” describes medical devices that have some, if not all, of the following characteristics: • Capable of performing numerous functions that might otherwise be performed by an assemblage of “stand-alone” or discrete devices • Functionally integrated, offering performance advantages over the use of several discrete devices • Operated continuously in a hands-on manner for an extended period of time (i.e., several minutes or even hours rather than just seconds) • Physically large by comparison to other devices and tools, sometimes as large as or larger than the user • Designed for use over many years, thereby justifying their relatively higher capital cost Workstations also tend to do the following: • Function in a stand-alone manner other than requiring utilities, such as electrical power, piped-in gases, data communication lines, and so on. Tool
Syringe
FIGURE 12.1
Device
Thermometer
Workstation
Patient monitor
X-ray machine
Workplace
CT scanner
Medical devices are spread across a continuum of functional complexity.
TAF-ER562X_WINGR-09-0401-C012.indd 474
11/9/10 12:21:14 AM
Workstations
475
• Require users (including patients) to assume static postures for extended periods. Some postures might be both unconventional (e.g., lying head down or with one arm held overhead) that call for extensive ergonomic analysis to ensure comfort and safety. • Necessitate special user training to enable operation to their fullest potential. • Provide a high level of monitoring and control over many functions to enable a full understanding of the technology’s operational status (i.e., situational awareness) and the patient’s well-being. As discussed earlier, there are many types of medical workstations (Figure 12.2), each designed to facilitate specific medical procedures, such as the following: • • • •
Anesthesia delivery Cardiac catheterization Cardiopulmonary bypass Patient monitoring
FIGURE 12.2
Sample medical workstations.
TAF-ER562X_WINGR-09-0401-C012.indd 475
11/9/10 12:21:14 AM
476
• • • • • • • • •
Handbook of Human Factors in Medical Device Design
Hemodialysis Laser-assisted in situ keratomileusis (LASIK) Magnetic Resonance Imaging (MRI) systems Peritoneal dialysis Positron Emission Tomography (PET) scanning Radiation therapy Tooth whitening Ultrasound imaging X-ray imaging
Workstations enabling these procedures have many characteristics in common with a familiar, nonmedical workstation:—aircraft cockpits. Each is a specific place (i.e., a station) where skilled users perform designated tasks. They provide substantial capability in an integrated, compact, and sometimes deceptively simple-looking package. In contrast, other workstations can look rather sophisticated, even perplexing, to the layperson’s eye—just like an aircraft cockpit. This is because workstations are usually intended for use by experts who understand the technology. In fact, a sophisticated-looking workstation, as opposed to one that appears simpler because of a reduced number of surface components, might be the optimal solution for a medical specialist because it gives him or her immediate access to all essential displays and controls. However, some workstations (e.g., one used for home dialysis) are intended for use by a layperson who manages his or her own care or provides care to someone else in a residential or workplace setting. These people are usually served better by workstations that are simpler because of the elimination of nonessential features. The term “workstation” also applies to items such as hospital beds, dental chairs, and operating room tables. Although one might think of these items as special purpose medical furniture, they also fit the workstation definition. Unquestionably, they perform special functions necessary to support the delivery of effective medical care. Many of them incorporate displays and controls and pose the same kind of design challenges one associates with a more typical-looking workstation, such as an ultrasound scanner. For example, a hospital bed could include controls and displays to adjust its position as well as weigh the patient, trigger an alarm if the patient leaves the bed, vary the pressure applied to the patient’s body by an air mattress, call for a nurse, and control a television.
12.1 CONSIDERATIONS Human factors in workstation design is an especially broad topic because of the diverse and sometimes complex nature of workstations. Accordingly, there are many general principles to consider (see Chapter 1, “General Principles”). However, all the general principles are summed up by the following statement: Adapt the workstation to the user rather than making the user adapt to the workstation. In fact, adapting designs to meet user needs rather than the reverse is a basic tenet of human factors engineering. Workstations should accommodate the physical, intellectual, and psychological needs of a diverse user population wherever possible. Otherwise, the quality of user interactions with the technology may suffer, thereby reducing the quality of patient care and the associated outcomes.
TAF-ER562X_WINGR-09-0401-C012.indd 476
11/9/10 12:21:14 AM
Workstations
477
Indeed, designing a workstation that accommodates a diverse user population can be a daunting task. Users’ characteristics, such as physical size, dexterity, intellectual capability, training, and learning style, can vary widely, thereby generating a broad set of user-interface requirements. Also, workstation design inevitably involves design compromises and trade-offs that extend well beyond human factors considerations. Mechanical, electrical, software, and manufacturing engineers, for example, will face their own daunting challenges. As a result, user-interface designs almost always involve compromise— sometimes substantial compromise, such as using a smaller computer display than is ideal from a usability standpoint. Accordingly, the key to a successful workstation design is to balance human factors engineering requirements with competing design requirements, but to meet the critical user-interface requirements (e.g., ready access to an emergency stop button), particularly those presented in the remainder of this chapter.
12.1.1 SAFETY 12.1.1.1 Protect Users from Hazards Protecting users against hazards is a workstation design imperative. Obvious workstation hazards include sharp points and edges, moving components, and sources of extreme heat and harmful energy (e.g., electrical current). Somewhat less obvious hazards include undersized casters that could cause a workstation to tip over and dangling cables that could pose a tripping hazard. Ideally, designers will consider every foreseeable hazard associated with intended and unintended uses in the course of workstation design. After identifying all potential hazards, the focus switches to trying to eliminate the hazards altogether—the preferred solution. If this is not possible, then the risk should be reduced by adding protective features, such as a physical guard or interlock. Rather than eliminating the hazard altogether, which might be impossible, designers frequently take a defense-in-depth approach that includes at least two means of protection (e.g., adding a guard and a warning). However, designers need to avoid overloading a design with numerous protective features that ultimately impede operation. The goal should be to balance the need for user protection with practical use considerations, ensuring that reasonably skilled users can operate the workstation safely under all reasonably likely conditions as well as under less common but higher-risk conditions (e.g., emergencies). Naturally, the same standard of care applies to protecting the patient, maintenance personnel, and any other people who might interact with the workstation. 12.1.1.2 Protect against Use Error Use error is a documented cause of many medical mishaps, some involving medical workstations such as radiation treatment machines and anesthesia workstations. As an indication of progress, a majority of the reported mishaps involving anesthesia machines occurred years ago before the establishment of equipment standards aimed at reducing the chance of use error, so the rate of mishaps has dropped considerably. Improved anesthesia workstation design exemplifies how an investment in human factors engineering can dramatically reduce the rate of use errors leading to patient injury and death. Still, it is virtually impossible to eliminate all chance of use error, considering that human beings are not perfect. Committing mental slips and mistakes (Norman, 1988) is part of being human. Also, new technologies can introduce new opportunities for use error, as has been
TAF-ER562X_WINGR-09-0401-C012.indd 477
11/9/10 12:21:14 AM
478
Handbook of Human Factors in Medical Device Design
the case with introduction of anesthesia machines with electronic controls in place of mechanical ones. Human fallibility means that workstation designers need to anticipate use errors and provide resources enabling users to recover from them before there is an adverse outcome. For example, designers might need to build in software checks so that users perform tasks in the proper sequence. As another example, designers might need to employ more than one method of information coding (e.g., use both shape and direction of control movement coding) to ensure that users recognize a critical control and its current status on grasping it with a wet, gloved hand in a dimly lit room.
12.1.2 USABILITY 12.1.2.1 Avoid Excess Complexity Health professionals frequently complain that workstations have extraneous features. They assert that manufacturers add extra features to make a technology statement and to prevail in “feature wars” waged among competitors for the sake of looking better on paper and making sales. This leads to workstations loaded with features of limited utility and increased interactive complexity. One solution is to make the supplemental features unobtrusive by placing them behind a panel or relegating them to a second-level software menu, for example. However, designers are well advised to adopt a more minimalist philosophy and question the benefit versus cost to users of all nonessential features. Requiring a strong justification to include a particular feature, instead of requiring a strong justification to exclude it, will probably lead to a workstation that has the right number of features. 12.1.2.2 Allocate Functions Appropriately to the User(s) versus the Workstation Most workstations incorporate some automation. On the positive side, this relieves the user of tedious tasks better suited to mechanization or computerization, such as sensor calibration or fluid priming. Machines certainly do many things—especially repetitive tasks requiring precise timing—better than humans. On the negative side, automation sometimes limits the user’s understanding of the workstation’s operational status (i.e., degrades his or her situational awareness) and the user’s ability to respond to emergencies. In cases of excess automation, users might find themselves feeling “behind the curve,” uncertain what is happening and how to intervene effectively. Therefore, developers need to assign the right kinds of functions to the user(s) versus the machine (see Chapter 2, “Basic Human Abilities”). Making the right assignments requires a careful analysis of all required functions and decisions about which ones best suit the user, taking into account the need to limit mental and physical workload. Moreover, for those tasks performed by the machine, care must be taken to keep the user informed about operations (i.e., tasks) in progress. 12.1.2.3 Arrange Controls and Displays to Facilitate User Tasks To begin, designers select or develop appropriate controls and displays to meet a user’s task requirements (see Chapter 7, “Controls,” and Chapter 8, “Visual Displays”). Then the challenge is to arrange them in a task-oriented manner. The design of control–display relationships should evolve from a careful analysis of how people will interact with the controls and displays in the course of performing common, critical, and time-sensitive tasks (Figure 12.3). Unfortunately, control and display arrangements often reflect underlying engineering requirements more than user needs and associated task flows. Potential consequences
TAF-ER562X_WINGR-09-0401-C012.indd 478
11/9/10 12:21:14 AM
Workstations
FIGURE 12.3
479
Examples of good and poor control and display relationships.
include wasted body, hand, and eye motion as well as erroneous actions. However, these consequences can be avoided while still accommodating engineering requirements. The key is to conduct a detailed task analysis along with other engineering studies to identify the full set of layout options and constraints. Assuming a user-centered approach to interface design, developers should start with a task-oriented layout and make only those compromises that are absolutely necessary. 12.1.2.4 Accommodate the Users’ Physical Characteristics The physical diversity of human beings poses substantial workstation design challenges. For example, a workstation might need to accommodate young children as well as large adults and/or both thin and heavy individuals (Figure 12.4). Similarly, a workstation might need to accommodate individuals with disparate grip strength and those with specific disabilities, such as impaired vision, impaired hearing, and immobilized limbs. Fortunately, there are abundant data on human physical characteristics and many analysis methods and tools to address the challenge (see Chapter 4, “Anthropometry and Biomechanics”). Anthropometric databases include a plethora of physical dimensions, such as overall stature, standing elbow height, hand length and breadth, and chest depth. There are also abundant data
FIGURE 12.4 Bariatric beds are built wider than conventional beds to accommodate very large individuals. (Courtesy of Hill-Rom.)
TAF-ER562X_WINGR-09-0401-C012.indd 479
11/9/10 12:21:14 AM
480
Handbook of Human Factors in Medical Device Design
on human strength, reach capabilities, and vision and auditory capabilities. By drawing on these data, designers can establish a range of appropriate workstation dimensions, determine ranges of adjustability, and accommodate people with special needs. 12.1.2.5 Provide Affordances Features that enhance user interactions are called affordances (Norman, 1988). An affordance could be an oversized handle that enables the user to move the workstation comfortably and with sufficient leverage from various angles. Or the affordance could be a built-in spotlight that enables the user to read a document during the course of a surgical procedure performed in dim lighting conditions. Sometimes, a simple affordance makes a huge difference in a workstation’s usability and popularity with users. For example, caregivers appreciate patient monitors that have visual alarms that are detectable from a considerable distance (see Chapter 10, “Alarms”). Affordances can also contribute significantly to a workstation’s efficacy and safety. For example, the same handle that enables a user to move a workstation with less effort can also provide the grasping point that helps the user keep his or her balance while reaching for a piece of equipment. 12.1.2.6 Consider the Intended Use Environment(s) The nature of the intended use environment is an important workstation design consideration (see Chapter 3, “Environment of Use”). After all, design requirements will shift dramatically depending on whether caregivers will operate a particular workstation in the same environment each time or in several different environments, including some that will expose the workstation to harsh climate conditions or rough handling. For example, a mammography machine may spend its entire service life in the same temperature- and humidity-controlled room. By comparison, an emergency ventilator might spend most of its service life inside an ambulance’s tight quarters (Figure 12.5), and an infant incubator might be used in both kinds of environments.
12.1.3 USER SATISFACTION 12.1.3.1 Choose an Appropriate Visual Style Some medical professionals chafe at the suggestion that a workstation’s appearance is important, insisting that it is not. However, workstation appearance does matter in subtle but sometimes powerful ways. Clearly, developers should not produce scary-looking workstations that fuel patient fears of a medical procedure or trigger claustrophobic reactions. Similarly, developers should avoid designs that draw undue attention to unimportant workstation attributes or that visually clash with the care environment. So, appearance can matter, at the minimum on an emotional or subconscious level. Appropriate visual design can also offer functional benefits. For example, it is easier to tell when a workstation needs cleaning if it is lightly colored and shows grime (see also Guideline 12.213: Light Colors Show Contamination). Conversely, workstations that rarely (if ever) receive a thorough cleaning should employ surface finishes that hide grime so that they do not always look dingy. Also, certain workstation elements, such as connection ports (see Chapter 9, “Connections and Connectors”) can be color coded to draw attention, identify their function, and reduce the risk of misconnections.
TAF-ER562X_WINGR-09-0401-C012.indd 480
11/9/10 12:21:15 AM
Workstations
FIGURE 12.5
481
Diverse medical workstation use environments.
12.1.3.2 Add Refinements Designers often refer to a product’s “touchpoints.” Touchpoints are the places where users make physical contact with a given product. In the case of an automobile, the touchpoints include the door handle, steering wheel, and turn indicator. In the case of an anesthesia workstation, the touchpoints might include the gas flow valve controls, the pivoting arm supporting the patient monitor, and the cart’s drawer handles. In the case of a dental chair, the touchpoints would include the headrest, foot control, and the cushions. Each touchpoint presents an opportunity to communicate the impression of quality to the user. The impression of quality will depend both on the material finishes and on mechanical engineering. Users will draw conclusions about a workstation’s overall quality from the quality of the touchpoints, so they are a worthy focal point for designers. But more important, highquality touchpoints can enhance user performance in important ways, such as increasing task vigilance by reducing physical fatigue, eliminating distractions, and enabling users to detect subtle tactile cues.
TAF-ER562X_WINGR-09-0401-C012.indd 481
11/9/10 12:21:15 AM
482
Handbook of Human Factors in Medical Device Design
12.2 SPECIAL CONSIDERATIONS Medical workstations have many characteristics in common with workstations found in the transportation, manufacturing, and communications industries. However, medical workstation design warrants the following special considerations.
12.2.1 WORKSTATION USES 12.2.1.1 Serve Life-Critical Purposes Some workstations perform life-critical functions, such as supporting a patient’s respiration and circulation. Certain kinds of use errors can cause patient injury or death (Bogner, 1994). Therefore, such workstations need to be designed to be virtually fail-safe from a user interaction point of view. This implies the incorporation of affordances, constraints, and safeguards to prevent inadvertent or erroneous actions (or inactions) during use. Such human factors design features will complement the requisite degree of electromechanical redundancy to prevent single-point and cascading failures. 12.2.1.2 Clinical Practices Evolve Medical workstations often remain in continuous use for many years, even decades (see Chapter 15, “Product Life Cycle”). This places a burden on designers to consider not only a workstation’s durability but also how clinical practice might change over time, leading to new and different workstation uses. Designs should incorporate flexibility to accommodate such change, which might include sequential technological add-ons over time. It also requires workstations to be robust so that they operate properly for a long time without needing excessive maintenance. 12.2.1.3 Users Assume Unusual Positions Medical procedures often require patients and caregivers to assume unusual positions during use. For example, a patient might need to lie on his or her side or to be tilted with his or her head down (i.e., placed in the Trendelenberg position) during a procedure (Figure 12.6).
FIGURE 12.6 Lithotripter places patient in a level and head-down position. (Courtesy of Storz Medical. With permission.)
TAF-ER562X_WINGR-09-0401-C012.indd 482
11/9/10 12:21:16 AM
Workstations
483
Accordingly, workstation designers need to identify the full range of possible patient and caregiver positions—not just sitting and standing in a conventional pose—in the course of an anthropometric analysis prior to establishing design requirements.
12.2.2 WORKSTATION USERS 12.2.2.1 Users Can Possess Varying Degrees of Skill, Training, and Experience Often there is wide variation in the natural and acquired skill level of workstation users even among those who have extensive education, training, and prior experience. Some people simply demonstrate greater aptitude for learning to operate a specific workstation. Certain individuals can learn quickly but then have difficulty recalling procedures later on, particularly if they have not used the workstation in several months. Other users might never receive formal training, for a variety of reasons, and learn instead more informally from experienced colleagues or by trial and error. Accordingly, designers should not depend too heavily on user training to overcome fundamental operational complexities (see Chapter 1, “General Principles”). Rather, workstations should ideally provide extra support to less experienced and/or unskilled users. These features could include on-screen prompts, bold labels, a quick reference card, an online help system, or automated error detection (e.g., detecting that the user has programmed an infusion pump to deliver an extremely high dose of medication and asking for confirmation) (see Chapter 11, “Software User Interfaces”). Ultimately, given the nature of today’s medical workforce and administrative procedures, manufacturers are well advised to produce workstations that are safe to use with little if any training, even though a naive operator might not be able to operate them effectively. Of course, users should embrace their responsibility to develop the necessary skills before operating complex workstations. However, the real world sometimes thrusts users into situations in which they might not have developed such skills and must rely instead on their experience with other technology and intuition. 12.2.2.2 Workstations Might Be Used by Individuals with Disabilities Many patients (and some caregivers) will have significant disabilities that place limits on their ability to interact with a medical workstation. As discussed earlier, potential disabilities, whether temporary or permanent, include limited range of limb motion, limited hand dexterity, reduced muscle strength, impaired hearing and vision, cognitive slowing, and memory deficits, to name just a few. Such functional limitations should be considered in the design of workstations, thereby accommodating a diverse patient population (see Chapter 18, “Home Health Care”). Similarly, accommodations should be made for workstation users who might also have mental or physical disabilities. Generally, such accommodations will benefit all users, which is the fundamental goal of universal design (North Carolina State University, 2004). 12.2.2.3 Patients Might Be under Stress An intimidating or even scary-looking workstation can place added stress on an unstable patient. Therefore, designers should create workstations that look benign and perhaps even comforting (Figure 12.7). There are several ways to accomplish this goal, including assuring continuous visual contact between the patient and caregiver, hiding mechanical
TAF-ER562X_WINGR-09-0401-C012.indd 483
11/9/10 12:21:16 AM
484
Handbook of Human Factors in Medical Device Design
FIGURE 12.7 MRI scanner is decorated with a sand castle graphic (harmonized with other room decor) to make it less intimidating to children. (Courtesy of Children’s Hospital Boston. With permission.)
components that do not require constant access, and using surface finishes and colors that look and feel soothing. 12.2.2.4 Caregivers Can Be under Stress Workstation designers need to consider the effects of stress on the caregivers as well. Caregivers often work under extreme time pressures on tasks that have life-and-death consequences. So, while caregivers become accustomed to working under such stress, it can take its toll in the form of mental lapses and frustration. Workstation designers can help to relieve the stress on caregivers—or at least not exacerbate it—by taking all possible steps to accommodate caregivers’ needs, including providing information when it is needed rather than making the caregiver ask for it, providing alerts and alarms that are contextually appropriate rather than a nuisance, and building in safeguards against possible use errors so that caregivers can correct their mistakes before the mistakes lead to major problems.
12.2.3 WORKSTATION USE ENVIRONMENTS 12.2.3.1 Migration from Hospital to Home Use A workstation designed for hospital use by sophisticated health care workers can find its way into homes where it is used by laypersons (see Chapter 18, “Home Health Care”). This
TAF-ER562X_WINGR-09-0401-C012.indd 484
11/9/10 12:21:16 AM
Workstations
485
places a considerable burden on workstation designers to develop solutions that will be usable and safe in a spectrum of use environments, each presenting unique design requirements and challenges. For example, a child might tamper with a workstation’s controls or place body parts near moving mechanisms. 12.2.3.2 Frequent or Infrequent Cleaning Hospital maintenance personnel clean some kinds of workstations (especially those found in the operating room) several times a day, after every patient exposure, using strong disinfecting chemicals (e.g., bleach). Meanwhile, other kinds of workstations found on general units (i.e., wards) might not be cleaned for months or years at a time. Therefore, designers must identify a workstation’s potential use environments to determine the associated cleaning requirements. Workstations that will receive frequent cleaning should be designed to facilitate and withstand rigorous cleansing. This means eliminating irregular features that trap contaminants and ensuring that maintenance workers can access all exposed surfaces without having to use special tools. Regardless of cleaning frequency, workstations should resist contamination and look presentable. 12.2.3.3 Need for Compactness Some patient care settings are cramped, packed with equipment, or both. This is often the case in critical care environments in which patients can require so much supportive technology that there is limited space for the caregiver (Figure 12.8). This means that certain types of workstations need to be quite compact and/or must not occupy too much floor space (i.e., have a small “footprint”). Conversely, other kinds of workstations can occupy an entire room specifically designed to accommodate them.
12.3 DESIGN PRINCIPLES The following guidelines should help designers produce ergonomically correct workstations that are well suited to the associated diagnostic and therapeutic regimes. The guideline categories are the following: • Operational factors, such as a workstation’s readiness for use, clarity of operational status, degree of automation, and user security and privacy
FIGURE 12.8
Critical care environments can be packed full of equipment.
TAF-ER562X_WINGR-09-0401-C012.indd 485
11/9/10 12:21:16 AM
486
Handbook of Human Factors in Medical Device Design
• Communication of information to the users through various means including displays, labels, and warnings • Component configurations that facilitate user tasks by placing controls and displays where they are needed • Physical interaction between the workstation and the users, including comfortably standing, sitting, or lying down to operate the workstation or be treated • User accommodations that make a significant difference in a workstation’s ease of operation and user satisfaction • Surface characteristics, such as materials finish and color and their effects on users • Environmental factors that must be addressed to ensure that users can operate workstations effectively
12.3.1 OPERATIONAL FACTORS The following design guidelines pertain to the operating characteristics of workstations. 12.3.1.1 Modes of Operation A medical workstation might have several operational modes, including start-up, normal use, emergency use, calibration, simulation, and service. Each mode might require adjustments to the workstation’s physical component configuration and enable specific functions and disable others. Alarm limits might need to be reset. As such, a workstation’s behavior and the demands on the user(s) can vary substantially among operational modes. A workstation’s operational modes should be simple and readily apparent. Given the critical functions performed by many medical workstations, there is little room for ambiguity in the user’s mind about the operational mode. Such ambiguity could induce use errors that place caregivers, patients, or property at risk. For example, patient transport personnel and ICU nurses used a patient monitor that was inadvertently placed in a static “demonstration mode;” they thought their patient had a classically normal heart rate of 72 beats per minute and blood pressure of 120/80 when the actual values were 140 beats per minute and 80/60, respectively (Gosbee, 2002). Thus, workstations should give users the information and control capabilities required to exert an appropriate level of control over workstation functions. Following the design guidelines presented below will help achieve this goal. Guideline 12.1: Mode Indication Workstations should continually indicate their operational mode and status. Specifically, workstations should indicate whether they are in an automatic or manual mode. They should also indicate whether they are in an active (in-use), passive (standby), or turned off mode.
Guideline 12.2: Mode Selection Workstations should provide a clear means of selecting an operating mode, such as “neonatal,” “pediatric,” or “adult” mode (Figure 12.9).
Guideline 12.3: Training Mode Indication Workstations should provide a clear indication when they are in a demonstration or training mode.
TAF-ER562X_WINGR-09-0401-C012.indd 486
11/9/10 12:21:17 AM
Workstations
487
FIGURE 12.9 patient mode.
Neonatal/pediatric/adult ventilator includes an LED readout to indicate the selected
Guideline 12.4: Default Mode Automatic functions, particularly those that are life critical, should default to a safe operating mode in the event of a component failure. The workstation should immediately alert the user to the mode change by means of a high-priority message or alarm (see Chapter 10, “Alarms”).
Guideline 12.5: Mode Changes Workstations should alert users immediately to any mode changes when an awareness of the current operational mode is critical to maintaining situational awareness or being prepared to act quickly and effectively in an emergency.
12.3.1.2 Use Error Prevention Taking steps in the design process to reduce the likelihood and effect of use errors is imperative to meeting regulatory requirements, fulfilling functional requirements, and ensuring a workstation’s commercial viability. Fortunately, there are many ways to reduce the chance that users will commit an error while operating a medical workstation. Moreover, there are many ways to reduce the risk that a use error will lead to adverse consequences. As discussed earlier (see 12.1.1.1 Protect Users from Hazards), human factors engineers use a multilayered approach to preventing and mitigating the effects of use errors. The preferred technique has been to modify a given design to eliminate the potential for a use error. For example, if a push button is found to be subject to inadvertent actuation, it might be appropriate to change it to a dual-action, flush-mounted control. Another effective method is to build in safeguards, such as a transparent cover over a critical control, to prevent inadvertent actuation. Less effective but still potentially useful preventive measures include warning users about possible use errors and or training them to avoid them. Each design strategy has its place in an overall approach for preventing use errors. The following guidelines address specific measures that designers can take to make a workstation less vulnerable to use error. Guideline 12.6: Protective Features Workstations should incorporate features that prevent critical use errors (e.g., interlocks, confirmation requests, and physical guards). An analysis of past use errors involving the same or similar types of workstations can identify opportunities to incorporate protective features.
TAF-ER562X_WINGR-09-0401-C012.indd 487
11/9/10 12:21:17 AM
488
Handbook of Human Factors in Medical Device Design
Guideline 12.7: Confirmation of User Intent Workstations should require users to confirm critical and irreversible machine functions. This will give users time to detect and correct slips and mistakes that could waste time, waste resources (e.g., a tubing set), cause property damage, and possibly harm the user or patient.
Guideline 12.8: Set Safety Limits Workstations should preclude dangerous settings, such as high ventilator pressures or high radiation dose levels, or at least require their confirmation before the setting takes effect. In this way, health care organizations can establish normal limits for specific clinical actions and require clinician users to confirm their intent to exceed them (if allowed).
Guideline 12.9: Inadvertent Actuation Workstations should incorporate guards against the inadvertent activation (enabling) of functions that could be dangerous. Moreover, workstations should require users to deliberately activate potentially dangerous functions when needed and otherwise lock them out. In this way, the user is sure to be aware of when potentially hazardous portions of the workstation are active and can take appropriate precautions.
Some automatic functions, such as movement of a boom on a C-arm fluoroscopy machine, could pose a hazard if activated at the wrong moment during therapy (Figure 12.10). For example, a moving boom might strike the user, the patient, or another piece of equipment. Or an energy source might be armed and ready to deliver a dose of radiation. Guideline 12.10: Prompt User Input Workstations should conspicuously indicate when there is a need for user input or intervention. This will help expedite tasks and avoid errors of omission, such as failing to restart a pump that is delivering a critical medication to a patient.
Guideline 12.11: Automatic Safeguards If a user fails to act in a timely or appropriate manner, workstations should automatically perform the functions necessary to ensure the safety of the user and patient. Users should not
FIGURE 12.10 C-arm type of fluoroscopy machine requires safeguards to ensure that the moving “arm” does not move unintentionally and injure a patient or caregiver. (Courtesy of Siemens Medical Solutions USA, Inc.)
TAF-ER562X_WINGR-09-0401-C012.indd 488
11/9/10 12:21:17 AM
Workstations
489
be able to deactivate automated safety features—particularly critical ones—except perhaps in extraordinary situations.
Guideline 12.12: Layers of Protection Optimally, workstations should provide patients and users with multiple layers of protection against potential hazards. The protective “layers” might include two or more of the following: a mechanical interlock, a protective cover, an access code, an alarm, a printed warning, and user training.
12.3.1.3 Automatic versus Manual Control Workstation functions should be automated when the device can perform the required function more effectively than a human user and can do so without reducing the user’s “situational awareness.” In other words, the workstation’s level of automation versus manual control should ensure that the user remains fully aware of the system’s current operational state and anticipates its likely state in the near future. This situational awareness allows the user to respond quickly and effectively to emergencies. Guideline 12.13: Manual Overrides of Automatic Functions Users should be able to override automatic functions except for those associated with critical protection systems. This approach grants users ultimate control and accommodates unanticipated circumstances and needs.
12.3.1.4 Patient and User Safety and Security Designers should strive to make workstation users as physically and psychologically comfortable and secure as possible. Do not presume that workstations serving a utilitarian purpose should be optimized to achieve only their clinical purposes. In most cases, it is possible to ensure a workstation’s functional effectiveness while also protecting the users’ sense of well-being. Guideline 12.14: Patients’ Physical Comfort Workstations should ensure the physical well-being of the patient, making the patient as comfortable as possible while holding him or her securely in the position necessitated by the associated medical procedure.
Guideline 12.15: Patient Restraint Design Patient restraints should require simple, intuitive steps to apply and remove without causing patient discomfort. Rapid removal might be essential in an emergency, such as a fi re.
Guideline 12.16: Personal Control and Communication Workstations that could pose a hazard to patients should include a means, such as a tethered control (i.e., pendant), for patients to request help (Figure 12.11), indicate an emergency, or stop potentially hazardous functions (as appropriate).
Guideline 12.17: Avoid Claustrophobic Conditions Where possible, workstations should not place the patient in an enclosed space that could trigger feelings of claustrophobia. For example, some MRI machines place patients inside a
TAF-ER562X_WINGR-09-0401-C012.indd 489
11/9/10 12:21:17 AM
490
FIGURE 12.11
Handbook of Human Factors in Medical Device Design
Patient holds a tethered pendant that includes a nurse call button.
tunnel that provokes claustrophobia in some individuals. The latest generation of so-called open MRI machines reduces the feeling of being in a tunnel, giving patients the feeling of greater freedom versus being trapped (Figure 12.12).
Guideline 12.18: Guards Against Injury Where possible, workstations should incorporate physical guards to protect users from moving components (e.g., gears and pulley belts) and components with sharp points or edges.
Guideline 12.19: Indicate Disabled or Failed Components Workstations should clearly and continually indicate whether important components either are not working properly or have been disabled.
Guideline 12.20: Redundanct Components and Backup Workstations should incorporate backup or redundant systems and components as necessary to protect the patient and user from hazards and allow them to complete critical tasks.
FIGURE 12.12 Whole-body scanner keeps the “tunnel effect” to a minimum for the sake of patient comfort. (With permission.)
TAF-ER562X_WINGR-09-0401-C012.indd 490
11/9/10 12:21:17 AM
Workstations
491
12.3.1.5 Power Supply Virtually all workstations—even today’s hospital beds—require electrical power that is supplied from either an AC outlet or a battery pack. Regarding workstations that deliver critical therapies, a total power loss can place patients at risk. Therefore, maintaining a reliable source of power to workstations is usually an imperative. The following guidelines describe how to ensure that a workstation user can determine a power supply’s status, can take the precautions necessary to avoid power interruptions, and can continue to deliver safe care in the event of an unavoidable power failure. Guideline 12.21: Backup Power Workstations should have a backup power source, such as an onboard battery, when a power failure would be disruptive or dangerous to the patient or user. The need for battery backup power is especially important for mobile workstations that might not be able to be connected to a central power source (i.e., AC power) under certain circumstances.
Guideline 12.22: Indication of Power Source Workstations should indicate when they are drawing power from a backup source, such as an internal battery (Figure 12.13).
Guideline 12.23: Power Remaining Indication When technically feasible, workstations operating on battery power should provide an approximation of how long they will continue to operate before running out of power (Figure 12.13).
Guideline 12.24: Low-Battery Warning Workstations should provide users with a warning when battery power is running low. The warning should be issued when there is enough battery power remaining to enable the user to find an alternative power source or safely discontinue use of the workstation.
Guideline 12.25: Fail-Safe Before a backup power supply becomes exhausted, workstations should save critical information and place physical components in their safest possible state.
Guideline 12.26: Power Capacity The maximum amount of time a workstation can operate on battery power should be determined on the basis of the most demanding use scenarios.
45 minutes remaining
FIGURE 12.13
Graphical symbols representing AC and battery power.
TAF-ER562X_WINGR-09-0401-C012.indd 491
11/9/10 12:21:18 AM
492
Handbook of Human Factors in Medical Device Design
Guideline 12.27: Power Switches Power switches should be placed in a visually accessible location that is not subject to inadvertent actuation (Figure 12.14). This will alleviate the need for users to “hunt” for power switches, sometimes by feeling around the back of a device with their hand for something that feels right.
Guideline 12.28: Screen Savers Screen savers, which are intended to save power, avoid screen burn-in (mostly a problem with earlier display technologies), and ensure information privacy, should not mask, obscure, or displace information essential to safe operation. Even when no operational changes or user interactions have occurred for an extended period of time, users should be able to glance at a workstation display to obtain critical information, such as a hemodynamic measurement or a dosage level, as appropriate.
12.3.1.6 Readiness While some workstations might receive continuous use, others might sit idle for hours, days, or longer between uses. Idle workstations might be placed in a standby mode that consumes little power, or they might be turned off. Consequently, it might take a while for idle workstations to return to a fully operational status. Depending on the type of workstation, certain components might need to energize, pressurize, heat up, complete a calibration cycle, and so on. While engineers should strive to minimize the “warm-up period,” to the extent that users might need to wait, workstations should indicate when they are or will be ready for use. Guideline 12.29: Indication of Use Readiness Workstations should indicate when they are powered up and ready for use. Conversely, workstations should indicate when they are not ready for use and why (e.g., require calibration or resupply).
Guideline 12.30: Start-Up Duration Workstations should have a start-up (warm-up) duration that is compatible with its urgency of use and instructions for use. Accordingly, workstations that must be functional within
FIGURE 12.14 Power switch is visually accessible and covered by a clear plastic guard to prevent accidental actuation.
TAF-ER562X_WINGR-09-0401-C012.indd 492
11/9/10 12:21:18 AM
Workstations
493
30 seconds, for example, should have a start-up cycle of ≤30 seconds. Workstations that perform lifesaving functions in urgent or emergency conditions (e.g., defibrillators) should have a minimal start-up time.
Guideline 12.31: Warning of Slow Start Up Times Workstations with a start-up duration that could interfere with the pace of critical care delivery should include a prominent warning label stating the start-up duration.
Guideline 12.32: Time Remaining Indication During the start-up period, a workstation should indicate how much time remains before it will be ready for use. This can be accomplished using a countdown timer or a progress bar, for example. If possible, workstations should alert users to wait if they try to use the equipment before it is ready for use.
Guideline 12.33: Critical Functions Available Quickly During the start-up period, a workstation should make critical functions available as quickly as possible, even if other functions are not yet available.
12.3.1.7 Security Only qualified and authorized individuals should operate medical workstations, particularly those delivering life-supporting therapy. Strategies for preventing unauthorized use should consider the possibility of mistaken and accidental actions as well as malicious ones. Guideline 12.34: Prevent Unauthorized Operation Workstations should prevent operation by unauthorized individuals. For example, therapeutic devices used in critical care environments might require caregivers to enter a special code or press an enabling button before they can change the device’s settings, thereby preventing hospital visitors from making adjustments.
12.3.1.8 Privacy Medical care environments have not generally been known for preserving personal dignity. Patients often must endure diagnostic and therapeutic procedures that not only cause pain but also intrude on their privacy. Workstation designers should pursue opportunities to preserve patient privacy and dignity. Guideline 12.35: Maintain Patient Privacy Workstations should not unnecessarily compromise the patient’s or user’s physical privacy. To the extent possible, workstations should enable users to assume comfortable and dignified positions as well as incorporate features, such as “modesty panels,” that afford an added measure of physical privacy.
Guideline 12.36: Protect Patient Confidentiality Workstations should prevent unauthorized individuals from viewing sensitive, personal information that may appear on documents or computer screens, for example.
TAF-ER562X_WINGR-09-0401-C012.indd 493
11/9/10 12:21:18 AM
494
Handbook of Human Factors in Medical Device Design
12.3.2 COMMUNICATION The following design guidelines address various means of communicating important information to the user and/or patient through channels other than conventional, computer-based displays. 12.3.2.1 Alarms The advent of microprocessor-based medical technology has led to the proliferation of alarms in the medical environment. It seems as though every device in medical care environments has its special way of beeping, which results in a chaotic and noisy situation. Thus, there is a movement toward alarm harmonization (see Chapter 10, “Alarm Design”). Guideline 12.37: Alarm Integration Where possible, alarms should be presented in an integrated manner (i.e., consolidated in one physical location and logically ordered) that gives users a focal point for identifying and accessing alarm conditions.
Guideline 12.38: Compliance with Alarm Standards Visual and audible alarms should follow the guidance provided in IEC 60601-1-8. In summary, alarms should ascribe to a hierarchy that includes high (red), medium (orange), and low (yellow) levels. Their acoustic characteristics (frequency, repetition pattern, and volume) should ensure detection by appropriate individuals under all expected operating conditions.
12.3.2.2 Warnings Some experts believe that warnings are virtually useless except perhaps for reducing a manufacturer’s vulnerability to product liability claims in case of an accident. Other experts assert that well-designed warnings can effectively reduce the chance of use errors. The actual value of warnings is difficult to measure, particularly because many warnings are designed to prevent rare events, so their effectiveness is difficult to observe and document. Nevertheless, regulatory agencies, lawyers, and conventional wisdom call for warnings against significant residual hazards. So, workstation designers face the task of producing the most effective warnings possible (see Chapter 13, “Signs, Symbols, and Markings”). Guideline 12.39: Warn Against Personal Hazards Warning labels should alert users (including patients) to potential hazards, such as moving components, hot surfaces, and hazardous chemicals.
Guideline 12.40: Warn Against Property Damage Warning labels should alert users to the potential for equipment damage, such as breakage due to rough handling or electrical malfunctions due to fluid contamination.
Guideline 12.41: Conspicuity of Warnings Warnings should be placed conspicuously and in close proximity to the workstation element(s) associated with the hazard or concern. Warnings should not be placed where they are likely to be obstructed by operating personnel, patients, or other equipment and materials, for example.
TAF-ER562X_WINGR-09-0401-C012.indd 494
11/9/10 12:21:18 AM
Workstations
495
Guideline 12.42: Comply with Warning Standards Warnings should generally conform to the requirements found in applicable standards, such as ANSI Z535 (generally focused on signage), and national and international regulatory requirements.
Guideline 12.43: Warning Saturation In general, workstation warnings should be limited to those addressing the major hazards and vulnerabilities. Otherwise, users might experience information overload, question the credibility of the warnings, or disregard the warnings altogether.
Guideline 12.44: Combine Text and Graphics Warnings should employ a combination of text and graphics (i.e., a pictograph) to communicate hazards at a glance.
Guideline 12.45: Start with Signal Word Signal words, such as “DANGER,” “WARNING,” “CAUTION,” and “NOTICE,” which indicate a hazard’s severity, should be highly visible and precede other warning information in a hierarchical communication scheme.
Guideline 12.46: Warning Hierarchy When presenting several warnings, the most critical ones should be presented in the most visually dominant manner so that they are the first to draw the user’s attention. For example, in a vertical arrangement of warnings, those incorporating the signal word “DANGER” should be placed above those incorporating the signal word “WARNING,” leading users to read the most important ones first. Note that this order presumes a cultural convention that leads people to gaze at objects placed on top of a set before those placed below.
Guideline 12.47: Concise Wording Warning messages should be concise—stating the nature of the hazard, its potential consequences, and the means to avoid the hazard.
Guideline 12.48: Test Warning Effectiveness Warning developers should conduct the necessary legibility and comprehension tests to confirm the warning message’s effectiveness.
12.3.2.3 Labeling In this section, the term “labeling” refers to text and/or graphics that identify workstation components and their operational states, or provide operational cues (see also Chapter 13: Signs, Symbols and Markings). Labels can improve a user’s recognition of specific workstation features, their functions, and their means of operation. Too few or too many labels might degrade a workstation’s usability. The challenge is to optimize a workstation’s labeling scheme to communicate an appropriate amount of useful information to users without oversaturating them or impairing usability. Of course, different types of users will have different labeling needs. Inexperienced users will benefit from a greater number of labels, some of which identify the most basic
TAF-ER562X_WINGR-09-0401-C012.indd 495
11/9/10 12:21:18 AM
496
Handbook of Human Factors in Medical Device Design
workstation features. Experienced users might prefer a “cleaner” design that limits labeling to unusual (i.e., unfamiliar) and critical workstation features. Given the spectrum of users’ needs associated with most workstations, designers should conduct task and risk analyses to determine the best possible labeling scheme. Usually, such analyses tend to favor a thorough labeling scheme that addresses the needs of less experienced users. Regulatory and legal requirements will also have a strong influence on the final labeling scheme. While past product liability suits might lead a manufacturer to protect itself by “papering” a product with labels and warnings, this could prove counterproductive if it is distracting, obscures critical information, or makes the device less usable. Guideline 12.49: Labels as Affordances Workstation components should include labels that facilitate initial ease of use and mitigate the chance of use errors.
Guideline 12.50: Label Legibility Labels should be legible under all expected viewing conditions. Although some caregivers might be accustomed to using a flashlight to read labels and control settings (e.g., to avoid waking patients at night or avoid interfering with medical procedures performed in dim lighting conditions), developers should consider illuminating these design elements (i.e., provide a spotlight or a backlight) because a flashlight might not always be available.
Guideline 12.51: Label Accessibility The user’s view of control labels should not be obstructed while operating the associated control (Figure 12.15). Normally, this suggests placing control labels above controls so that users do not block them when they reach for the controls (see Chapter 7, “Controls”). When a control is located well above eye level, labels should be placed to one side to maintain visibility and avoid labels being blocked by the user’s hand. However, every effort should be made to ensure label-to-control placement consistency, which might require relocating controls.
Guideline 12.52: Symbols as Labels As appropriate, labels should incorporate symbols/icons in place of text to accommodate people who speak different languages (as occurs in European countries and sections of the United States with large immigrant populations) or have low literacy. However, the intended user’s ability to recognize symbols (icons) should be tested.
Start pump Start pump Start
FIGURE 12.15
pump
User’s hand will block a label placed below a control (middle).
TAF-ER562X_WINGR-09-0401-C012.indd 496
11/9/10 12:21:18 AM
Workstations
497
Guideline 12.53: Label Language As necessary to facilitate use in predominantly multilingual regions, text labels, like warnings, should be presented in two or more of the most prevalent languages spoken in the region where the workstation will be used (Figure 12.16). However, care should be exercised to avoid creating a congested-looking user interface where it becomes difficult to link labels with their appropriate controls and displays. Where possible, users should have the option to display labels (particularly labels presented on computer-driven displays) in their preferred language. However, complications can arise when different users within the same working environment need to interact with the same device but have different language preferences. In such cases, an institution might choose to lock in a default language.
Guideline 12.54: Label Terminology Labels should include terms and/or symbols that are generally familiar to the intended users.
Guideline 12.55: Label Durability All markings should be sufficiently durable so as to be readable for the expected life of the device. Specifically, markings should resist being marred or rubbed off over an extended period of use and cleaning. One strategy to achieve this goal is to print labels on the underside of a transparent membrane.
12.3.2.4 Instructions for Use It is often said that medical products—workstations included—should be so intuitive to operate that there is no need for instructions. However, this ideal is rarely achieved. Moreover, many regulatory bodies require manufacturers to provide operating instructions. So, hospitals and other care environments have bookcases and/or file cabinets full of user manuals—documents that most caregivers consider to be their last resort if they cannot figure out how to operate something on their own or with the help of a more experienced colleague. General disdain for consulting user manuals has created a negative reinforcement loop that leads many manufacturers to focus less attention on making manuals truly helpful because the users profess not to use them. Instead, many manufacturers write the user manuals predominantly to meet regulatory requirements. Only occasionally will they invest heavily in designing and writing a user manual that will serve as an excellent learning tool for inexperienced users, people performing uncommon tasks, and those trying to solve a problem. Of course, users are pleased to discover a well-written, user-centered manual or related learning tools, such as quick reference cards and online help. These learning tools can be very helpful to all types of users at some point in time, particularly when there are problems requiring a quick resolution. The challenge for institutions and managers is to make such documentation readily available without cluttering frontline work areas. Danger - Explosion hazard. Do not use in the presence of flammable anesthetics. Danger - Resque d’explosion ne pas employer en presence d’anesthesiques inflammables.
FIGURE 12.16
Dialysis machine warning written in both English and French.
TAF-ER562X_WINGR-09-0401-C012.indd 497
11/9/10 12:21:18 AM
498
Handbook of Human Factors in Medical Device Design
Emergencies do not provide users with the time to search through cabinet drawers for a manual. This is another reason for the growing popularity of online help, particularly in association with workstations incorporating a computer display. Users also appreciate quick reference cards that can stay connected to a workstation as well as critical instructions printed on the associated workstation component(s) by the manufacturer. Guideline 12.56: Basic Operation Instructions Workstations should provide at least the basic instructions for use, if not a complete guide, in some form. Such instructions may take the form of a quick reference guide addressing operational fundamentals and emergency procedures.
Guideline 12.57: Availability of Operating Instructions Basic and critical operating instructions should be available at all times, including when the workstation is powered off. Accordingly, such operating instructions should not be provided solely by means of computer displays that might be turned off.
12.3.3 COMPONENT CONFIGURATIONS Sometimes, designers are granted considerable freedom regarding the configuration of workstation components. This may be the case when designing the user interface of a large, computer-driven workstation, particularly when the user interface is a separable component. However, workstation designers often find themselves heavily constrained by engineering considerations unrelated to human factors. For example, the user interface of a smaller workstation might pose mechanical, electrical, and package design requirements that dictate the location of specific components, such as a power switch, an LCD display, or a keypad. Despite such constraints, designers should identify a component configuration that still enables users to perform tasks in a safe and effective manner, generally treating efficiency as an important but secondary consideration. A thorough analysis of user requirements provides important insights into the optimal way to arrange workstation components so they reflect the necessary trade-offs while avoiding safety and usability problems. For example, the frequency and urgency of tasks might suggest placing certain controls and displays closer to the user. Another possible arrangement might make it easier for users to form a clear understanding of how the workstation works as a whole. A third arrangement of controls and displays might match users’ expectations developed by using similar workstations. So, designers need to weigh the user-oriented advantages and disadvantages of the various options, obtain user feedback on the options, and account for other engineering constraints to converge on an optimal solution. Important factors to consider while exploring component configuration options are listed below. Guideline 12.58: Arrangements Should Optimize Task Performance Components should be arranged in a consistent manner that maximizes user performance of the most important (i.e., frequent, critical, and urgent) tasks.
Guideline 12.59: Arrangement Consistent with User Expectations Workstation components should be arranged in the same manner as similar and/or previousgeneration workstations unless there is a compelling reason to adopt a new arrangement, such
TAF-ER562X_WINGR-09-0401-C012.indd 498
11/9/10 12:21:19 AM
Workstations
499
as a demonstrable performance improvement or avoiding patent infringement. Doing so will reduce the need for training and will help ensure positive transfer (see Guideline 12.60).
Guideline 12.60: Arrangement Supports Prior Experience Component arrangements should match patterns previously established by the manufacturer and/or the medical device industry to ensure positive (rather than negative) transfer of operational experience with similar devices. In this context, positive transfer refers to the ability of a user to apply what they know about operating one workstation to another because the workstations function in a consistent manner. Negative transfer refers to cases when a user tries to operate a workstation but finds that it does not function the same way that he or she expects based on previous experience. Negative transfer can lead to use errors, such as rotating a valve control clockwise to decrease a flow rate when it actually increases a flow rate. Note that transfer can also occur based on experience with nonmedical devices (e.g., a cellular phone that has a “telephone” style as opposed to a “calculator” style of number pad).
Guideline 12.61: Arrangement Consistent with User Conventions Generally, controls and displays should be arranged to facilitate a left-to-right (primary) and top-to-bottom (secondary) movement of the eyes and hands. However, designers should evaluate the conventions of the intended user populations in order to determine the optimal arrangement.
Guideline 12.62: Comply with Arrangement Standards Component arrangements should comply with applicable standards, such as those published by international and national standards organizations.
Guideline 12.63: Avoid Mirror Image Arrangements Similar sets of components should not be mirror imaged. In other words, one set should not appear to be the mirror reflection of the other (i.e., a horizontally or vertically flipped arrangement) (Figure 12.17).
Guideline 12.64: Differentiate Components Sets of similar-looking components should be differentiated in some manner (e.g., color coding, labeling, and/or background panel color) that makes them easy to locate.
Guideline 12.65: Component Visibility Displays and controls should be placed at a consistent, comfortable distance from the user’s eye. A distance of 18 to 24 inches is appropriate for many video display terminals (see Chapter 8: Visual Displays). Good
Poor
Similar sets of controls and displays assume the same arrangement.
Similar sets of controls and displays are mirror imaged, increasing the chance of a use error.
FIGURE 12.17
Examples of good (left) and poor (right) control and display relationships.
TAF-ER562X_WINGR-09-0401-C012.indd 499
11/9/10 12:21:19 AM
500
Handbook of Human Factors in Medical Device Design
Guideline 12.66: Reach Distance of Controls Controls should be placed where users can reach them without disrupting any ongoing tasks. Generally, this means placing controls within the extended (i.e., slightly stretched) reach of a seated user (see Chapter 4, “Anthropometry and Biomechanics”).
Guideline 12.67: Location of Emergency Controls Emergency controls should be within the reach of users in all expected use positions (Figure 12.18). If necessary, workstations should provide emergency controls (e.g., an interrupt switch that stops a potentially hazardous function) in multiple locations.
The following guidelines on component configuration describe the need for consistency and logic in the relationship between individual controls and displays as well as component groupings. Guideline 12.68: Control-Display Arrangement Consistency The arrangement of similar groupings of controls and displays should be consistent. However, they should be kept simple, leaving out nonessential details.
Guideline 12.69: Emphasize Functional Relationships Displays and their associated controls should be arranged in a manner that emphasizes their functional relationship. For example, gas control valves can be placed directly below the associated flow rate indicators (flow tubes) to emphasize their functional relationship (Figure 12.19).
FIGURE 12.18 This volume computed tomography (CT) scanner’s emergency stop button (upper right corner) is placed in close proximity to the primary controls.
TAF-ER562X_WINGR-09-0401-C012.indd 500
11/9/10 12:21:19 AM
Workstations
501
FIGURE 12.19 (See color insert following page 564.) An anesthesia machine’s gas flow meters (tubes) reside directly above their associated controls. The controls are also differentiated by color and tactile feel.
Guideline 12.70: Mimics of Physical System Components can be arranged in a manner that represents or “mimics” the real system. This means arranging valve controls and associated displays in a pattern that is similar to the actual valves and piping. The relationship among components can be reinforced by the addition of flow lines and special symbols. Such mimics can help users form an accurate mental model of the workstation’s functions. However, they should be kept simple, leaving out nonessential details.
Guideline 12.71: Visibility Controls and displays should be placed where they will not be blocked by other items (e.g., surgical drapes, treatment devices, and/or IV fluid bags and lines) in the intended use environment.
Guideline 12.72: Component Spacing Components should be spaced a sufficient distance apart to provide easy physical access and to avoid inadvertent actuation. Factors such as the range of user hand sizes, the possible use of gloves, and the potential for workstation vibration should be considered when determining the optimal spacing (Figure 12.20).
Guideline 12.73: Control-Display Proximity A display that should be monitored while operating a particular control should be placed in close proximity to that control (Figure 12.21). The conventional solution is to place the display directly above the associated control so that the display will not be blocked by the user’s hand.
TAF-ER562X_WINGR-09-0401-C012.indd 501
11/9/10 12:21:19 AM
502
FIGURE 12.20
Handbook of Human Factors in Medical Device Design
Examples of good (left) and poor (right) component spacing.
Guideline 12.74: Protect Components Against Hazards Components that are vulnerable to damage or accidental actuation should be placed in protected locations to make them less susceptible to incidental contact. Additionally, components requiring physical manipulation should be kept a safe distance away from hazards, such as sources of heat and moving parts.
12.3.3.1 Demarcation Sometimes there is a need to visually distinguish related groups of workstation components in order to help users locate specific components and form a clearer mental model of workstation functions. Effective demarcation can make a workstation look simpler and improve usability.
FIGURE 12.21
Examples of good (top) and poor (bottom) control and display relationships.
TAF-ER562X_WINGR-09-0401-C012.indd 502
11/9/10 12:21:20 AM
Workstations
503
FIGURE 12.22 Functional groups of controls and displays demarcated by a line (top), a colored background (middle), and space (bottom).
Guideline 12.75: Demarcate Functional Groups Functional groupings of displays, controls, and other workstation elements should be demarcated by appropriate means (e.g., panel color, demarcation lines, and/or unused panel space) (Figure 12.22).
Guideline 12.76: Subdivide and Label Components Components forming large grids should be visually subdivided into smaller groupings with unique labels.
12.3.3.2 Display Integration (See Chapter 8, “Visual Displays”) Automobile manufacturers go to considerable lengths to produce dashboard layouts that work well for drivers. For example, they take care to ensure that the speedometer is fully visible through the open area of the steering wheel. Workstation designers should take similar steps to ensure that all displays provide information to the users in an effective manner. Caregivers should not have to adjust their body position or twist their neck during a medical procedure just because the display is blocked by another component or is placed outside the normal line of sight. Nor should caregivers find themselves guessing about a display reading because the numbers are too small. Guideline 12.77: Visibility of Displays Displays ranging from conventional monitors to small, two-line displays and single-parameter gauges should be placed where they can be seen clearly from all expected user positions and under all expected viewing conditions (e.g., sunlit and dimly lit rooms). When multiple users must view the same display from different positions, the display position should provide a better view to the user(s) who view it the most often while not precluding any user from viewing critical information. Alternatively, multiple displays should be provided, each placed in an optimal viewing position for one of the primary users.
Guideline 12.78: Legibility of Displayed Elements Displays should present information in a size that can be read comfortably at the maximum specified viewing distance. For many workstations, the expected viewing distance of information viewed straight on will not exceed an arm’s reach (about 20 to 24 inches). Generally, the character height of critical information should be equal to the viewing distance (measured in
TAF-ER562X_WINGR-09-0401-C012.indd 503
11/9/10 12:21:21 AM
504
Handbook of Human Factors in Medical Device Design
the same units) divided by 150. The character height of information that has the next-highest importance should be equal to the viewing distance divided by 300. Assuming a viewing distance of 24 inches, displayed critical and important text should be 0.16 inches high and 0.08 inches high, respectively.
Guideline 12.79: Viewing Distance for Multiple Displays Multiple displays should be arranged at approximately the same viewing distance to limit the user’s need to refocus his or her gaze.
Guideline 12.80: Optimal Display Height To optimize viewing comfort, the top portion of a display should be slightly (about 10 degrees) below the user’s normal, horizontal line of sight.
Guideline 12.81: Display Position to Prevent Fatigue To avoid neck strain, displays should be arranged so as to limit the user’s need to repeatedly turn or tilt his or her head. Because the most fatiguing head movement is to gaze upward, displays should optimally be placed near or slightly below the horizontal line of sight from the expected user’s eye position.
Guideline 12.82: Display Adjustability As necessary, displays should incorporate adjustment (e.g., tilt and swivel) mechanisms that enable users to position them for optimal viewing as well as to avoid glare (Figure 12.23). Displays should be angled laterally no more than 30 degrees from the user’s line of sight to ensure good legibility.
FIGURE 12.23 position.
Swivel arm enables caregivers to place a patient monitor in the optimal viewing
TAF-ER562X_WINGR-09-0401-C012.indd 504
11/9/10 12:21:21 AM
Workstations
505
Guideline 12.83: Handles on Displays Large and mobile displays should incorporate handles or secure grasping points for easy movement.
Guideline 12.84: Labeling of Displays As necessary, displays should be labeled to clarify their purpose or the source of the displayed information (e.g., Bed 1, Bed 2, or Bed 3).
Guideline 12.85: Stow Unneeded Displays It should be possible to place movable displays (e.g., displays on a pivoting arm) out of the way when they are not needed so that they do not interfere with other user tasks.
12.3.3.3 Storage Space There are many supplies and accessories that accompany most medical workstations. For example, a given workstation can require bottles of lubricating gel, a roll of replacement paper for an integrated printer, or a lead-lined, protective cloak. It facilitates associated medical procedures when workstations provide adequate and conveniently located storage for such items. Guideline 12.86: Sufficient Storage Capacity Workstations should include storage space for items such as associated accessories, consumables, and frequently used spare parts where their ready availability will enhance overall task performance (Figure 12.24).
Guideline 12.87: Storage Design for Intended Uses Storage spaces should be shaped properly to accommodate the stored items without the need for special arrangement or forceful placement.
Guideline 12.88: Labeling of Storage Spaces Storage spaces should be labeled to facilitate easy identification and retrieval of stored items. Alternatively, space should be provided for the users to apply their own labels (e.g., handwritten on tape).
Guideline 12.89: Cleanability of Storage Spaces Storage spaces should be easy to clean. For example, removable bins would be easier to clean than built-in (nonremovable) bins.
Guideline 12.90: Security of Storage Spaces Storage spaces used to store items that are vulnerable to damage or are dangerous, have significant monetary value, or are controlled substances should include a locking mechanism. However, storage spaces should not be lockable if they are likely to contain items that must be available in an emergency and the means to unlock the storage space quickly at the necessary moment cannot be ensured.
Guideline 12.91: Position Storage Spaces Optimally Workstation shelves should be positioned so that placing and removing objects does not cause physical strain (e.g., a strained back or shoulder). Generally, shelves at waist to shoulder level
TAF-ER562X_WINGR-09-0401-C012.indd 505
11/9/10 12:21:21 AM
506
Handbook of Human Factors in Medical Device Design
FIGURE 12.24 Workstations incorporate various storage solutions, ranging from sliding baskets to lockable drawers to open bins. should be the deepest (≤24 inches), shelves placed below waist level should be somewhat shallower (≤18 inches), and shelves placed above the shoulders should be the shallowest (≤12 inches) unless the workstation incorporates a convenient means to increase the user’s reach (e.g., stepladder).
Guideline 12.92: Provide Storage for Documentation An easily accessed storage location should be provided for any device user manuals or associated learning tools (e.g., quick reference guide) that users might need to access while performing tasks.
12.3.3.4 Job Aids A job aid is practically anything that helps people do their jobs. The job aid can be as simple as a hook in the right place. Guideline 12.93: Provide Useful Job Aids Workstations should incorporate helpful features or “affordances,” such as document holders, pen holders, and headsets, that enhance the patient’s and/or the user’s ability to perform routine, emergency, and maintenance tasks.
TAF-ER562X_WINGR-09-0401-C012.indd 506
11/9/10 12:21:21 AM
Workstations
507
Guideline 12.94: Workstation Customization Where appropriate, a workstation should have mechanisms that enable users to customize its configuration to suit their preferences. Ideally, the workstation should be able to memorize an individual’s custom configuration for convenient selection at a later time. For example, a workstation can incorporate presets for certain examination table positions.
12.3.3.5 Cable (Wire and Tube) Management (See Chapter 9, “Connections and Connectors”) Medical workstations are often encumbered with many wires, cables, and tubes—the essential connections to the patient, power sources, information conduits, and supplies (e.g., medical gases). Caregivers routinely ask for solutions to the disorganized mess that can develop in care environments, such as a pediatric intensive care unit, where one finds a considerable amount of technology in use. Workstation designers can contribute to the solution by limiting the number of lines (where possible), providing a means to route them in an organized manner, and generally recognizing that the workstation will be used in conjunction with other equipment. Guideline 12.95: Effective Cable Routing Cables should be routed in an intuitive and convenient manner. Any cables running to the workstation should be placed where they will not interfere with tasks or pose a tripping hazard (Figure 12.25).
Guideline 12.96: Prevent Inadvertent Disconnection Cables should be protected against accidental disconnection. Moreover, users should be able to detect visually an incomplete (i.e., loose) cable connection or be alerted to a disconnection by means of an alarm.
Guideline 12.97: Protect Against Cable Damage Cables should be protected against damage (e.g., crushing and crimping).
FIGURE 12.25
Movable arm is equipped with snaps to hold tubes in place.
TAF-ER562X_WINGR-09-0401-C012.indd 507
11/9/10 12:21:21 AM
508
Handbook of Human Factors in Medical Device Design
Guideline 12.98: Visual Inspection of Cables Cables requiring continuous visual inspection should be placed in a visually conspicuous location.
Guideline 12.99: Coding of Cables Cables and their associated ports should be coded (e.g., colored, textured, and/or shaped) to facilitate proper connections and prevent inappropriate ones. In particular, cable and port designs should preclude erroneous connections (e.g., placing a patient sensor lead into a power supply outlet) that could lead to injury or death.
Guideline 12.100: Labeling of Cables Cables should be labeled for rapid and accurate identification. Labels should be obvious and readable when users hold the cables at arm’s length.
12.3.3.6 Housings In addition to protecting internal components, housings keep workstations from looking intimidating. In fact, they can even give a workstation a pleasing appearance—an important goal given that the appearance of medical technology can impact both clinician and patient performance and satisfaction. Housings also serve important functional purposes, such as protecting users from hazards and providing work surfaces. Guideline 12.101: Opening/Closing of Housings Housings should close securely so as to avoid inadvertent opening at a time that could be inconvenient or hazardous. Opening mechanisms should be designed to prevent accidental actuation.
Guideline 12.102: Avoid Housing Hazards Housings should be designed to eliminate pinch points, sharp edges, and sharp points.
Guideline 12.103: Minimize Noise and Vibration Housings should be designed to avoid vibration and noise (rattling) that could be annoying and distracting to users.
Guideline 12.104: Avoid Blockage of Housings Housings should be designed to prevent any vents or access points from becoming blocked.
Guideline 12.105: Functional Utility of Housings When possible and appropriate, housings should serve useful, secondary purposes, such as providing a writing surface or an attachment point for accessories.
12.3.4 PHYSICAL INTERACTION The following design guidelines pertain to the physical interactions between a workstation and the people who might interact with it, including clinicians, patients, technicians, and maintainers.
TAF-ER562X_WINGR-09-0401-C012.indd 508
11/9/10 12:21:22 AM
Workstations
509
12.3.4.1 Anthropometric Characteristics (See Chapter 4, “Anthropometry and Biomechanics”) Most adults have had the experience of driving an automobile that feels like a perfect fit with their body type versus one that is an awkward fit. The fit quality depended on the position of the seat relative to the dashboard, steering wheel, gas and brake pedals, armrest, shifter, and other interior components. It also depended on the provision and effectiveness of adjustment mechanisms for the seat, steering wheel, and pedals (a relatively new innovation). In other words, fit quality depended on a fundamentally sound interior design that considered the physical characteristics of the driver population and appropriate adjustment mechanisms to fine-tune the relationship between the user and the interior components. In the case of medical workstations, the fit between the users and the technology is usually critical to medical efficacy and user satisfaction. A dental patient will be dissatisfied and uncomfortable reclining in a dental chair that presses sharply on her spine but will rest comfortably and with greater satisfaction in one that provides perfect support after a few minor adjustments. Similarly, a woman in the course of labor and delivery could experience more discomfort in a birthing bed that does not elevate in a manner that fits her body but may find relief in one that does, allowing her to find a position that makes it easier to endure contractions (Figure 12.26).
FIGURE 12.26 Birthing bed accommodates various patient positions during labor and delivery. (© 2009 Hill-Rom Services, Inc. Reprinted with permission–all rights reserved.)
TAF-ER562X_WINGR-09-0401-C012.indd 509
11/9/10 12:21:22 AM
510
Handbook of Human Factors in Medical Device Design
Guideline 12.106: Accommodation of Diverse Users Minimally, workstations should accommodate individuals with physical traits within the range of the 5th-percentile female to the 95th-percentile male for the intended user population. However, designing according to this range of human characteristics can sacrifice the needs of about 10% of the population. Where feasible, the range should be expanded to span the first to 2.5th percentile female and the 97.5th- to 99th-percentile male, thereby accommodating many more people at the extremes. Similarly, the lower limit for standing eye height should be drawn from data associated with a generally shorter user population, while the upper limit for standing eye height should be drawn from data associated with a generally taller user population. When a single workstation cannot accommodate all users within a sufficiently large size range, manufacturers should consider producing models of varying size.
Guideline 12.107: Ample Legroom Sit-down workstations should provide ample clearance for the user’s legs. To accommodate the majority of users, the minimum depth of the leg space, measured from the work surface’s front edge, should be 15 inches (38.0 cm) at knee level and 23.5 inches (59.0 cm) at floor level. The minimum width of the clearance envelope should be 20 inches (50.8 cm). The minimum height of the clearance envelope, which provides room for the user’s knees, should be 26.2 inches (66.5 cm), measured from the bottom of the work surface to the floor.
Guideline 12.108: Kick Spaces for Standing Workstations Standing workstations should provide clearance at their base for the user’s feet. Nominally, cabinets should have kick spaces (i.e., a space to accommodate the front portion of the foot when a user presses up against the cabinet) measuring 4 inches (10 cm) high and 4 inches (10 cm) deep (Figure 12.27).
12.3.4.2 Physical Accessibility Section 508 of the Rehabilitation Act (enacted by the U.S. Congress in 1973 and subsequently amended in 1998 to address the accessibility of information) promoted enhancements to the physical environment, and now the electronic world, to improve access to places and services by people with disabilities. The act influenced the design of hospitals,
FIGURE 12.27
Desk provides a kick space for feet to rest comfortably on the floor.
TAF-ER562X_WINGR-09-0401-C012.indd 510
11/9/10 12:21:23 AM
Workstations
511
leading to new kinds of entrances, workspaces, and bathrooms, for example. Just as architects needed to give more consideration to the accessibility of their living and workspaces, workstation designers should give due consideration to the accessibility of their designs. Sometimes, features added to improve accessibility benefit all users, affirming the concept of “universal design.” Guideline 12.109: Provide Assistive Features Large workstations should include features (e.g., platforms, ramps, stairs, ladders, and/or video camera views) necessary to provide users full access to all components requiring direct interaction.
Guideline 12.110: Accommodate Users with Disabilities Where appropriate, workstations should be designed to accommodate the needs of individuals with disabilities (e.g., deafness, low hearing, blindness, low vision, muscle weakness, and/or limited range of motion.) This need is particularly important regarding workstations designed for use by patients and laypersons (This topic is covered in more depth in Chapter 18: Home Health Care.).
12.3.4.3 Multiple Users Medical workstations sometimes engage several users at the same time. This requires workstation designers to consider such factors as how people will interact and how they will access specific workstation components. Guideline 12.111: Facilitate Communication Between Multiple Users Workstations designed for concurrent use by multiple users should enable visual contact and verbal communication between users, as appropriate.
Guideline 12.112: Facilitate Collaboration and Coordination Individual work areas should be sized and arranged to facilitate collaborative tasks (e.g., one technician passing a blood or tissue sample to another technician or two people viewing a displayed image) (Figure 12.28).
12.3.4.4 Clinician and Patient Position Pictures in sales brochures typically show workstation users in traditional poses: clinicians standing up straight while the patient lies supine on a treatment table. However, many medical workstations place users—clinicians and patients alike—in a wide range of body positions during the course of a medical procedure. Patients might need to elevate their legs or lie on their side. Clinicians might need to sit or to extend their reach across the patient’s torso to place a sensor correctly. Workstations need to facilitate the wide range of possible user positions. Guideline 12.113: Sitting and Standing Operation Where possible, workstations should enable operation from either a sitting or a standing position. For example, some anesthesiologists prefer to stand up while working, whereas others are more comfortable sitting down for significant periods of time. Therefore, specific controls and displays should be easily accessed from either a standing or seated position.
TAF-ER562X_WINGR-09-0401-C012.indd 511
11/9/10 12:21:23 AM
512
Handbook of Human Factors in Medical Device Design
FIGURE 12.28 Catheterization laboratory requires people to work in a coordinated fashion. (With permission.)
Guideline 12.114: Aid User Position Maintenance As appropriate, workstations should provide the means (e.g., a chin cup and forehead rest on an eye treatment device) (Figure 12.29) necessary to keep the user or patient securely in the preferred position.
Guideline 12.115: Allow Body Position Adjustments The workstation should allow the user to adjust his or her body position frequently in order to mitigate muscle fatigue. Conversely, workstations should not require the user to remain in a single position for a long period of time.
12.3.4.5 Line of Sight Like the airline pilot scanning the cockpit for flight information, clinicians scan their operating environment for key information, such as the patient’s vital signs, the patient’s
FIGURE 12.29 Chin cup and forehead rest hold the patient’s head in the proper position for an eye examination. (© Philip Berck 2005. With permission.)
TAF-ER562X_WINGR-09-0401-C012.indd 512
11/9/10 12:21:23 AM
Workstations
513
physical appearance, the amount of IV fluid infused, a therapeutic device’s state of readiness, and the location and current setting of important controls. As such, it is important to place critical workstation elements and associated information within the clinician’s field of view without obstructing his or her view of the patient. Conversely, it might be appropriate to keep certain workstation elements and associated information outside the patient’s field of view in order to control his or her anxiety, for example. Guideline 12.116: Unobstructed View of Workstation Components Workstations should offer users an unobstructed view of all workstation components (e.g., controls and displays) they need to see at the appropriate time to accomplish tasks (Figure 12.30).
12.3.4.6 Handedness Generally, workstations should enable equivalent ease of use by right- and left-handed users, who represent roughly 90% and 10% of the population, respectively. (Note: The precise proportion of right-handed and left-handed individuals is widely debated, but righthanded individuals still greatly outnumber left-handed individuals.) Guideline 12.117: Use of Dominant Hand Workstations should enable users to use their dominant hand to perform tasks requiring significant manual dexterity.
12.3.4.7 Repetitive Motion and Cumulative Trauma (See Chapter 4, “Anthropometrics and Biomechanics” and Chapter 16, “Hand Tools”) Repetitive motion disorders, a major concern among certain types of factory and office workers, are a growing concern within the medical industry. Disorders afflicting the hands, wrist, and shoulder are a particular concern to clinicians whose livelihood depends heavily on the use of the upper extremities. Accordingly, designers should (1) determine if
FIGURE 12.30 Wall-mounted patient monitor provides an unobstructed view for multiple members of a clinical team at once.
TAF-ER562X_WINGR-09-0401-C012.indd 513
11/9/10 12:21:24 AM
514
Handbook of Human Factors in Medical Device Design
workstation use necessitates repetitive motions or other actions that could induce cumulative trauma and (2) pursue designs that limit such motions or actions. Guideline 12.118: Minimize Repetitive Motions Workstation operations should not expose the user to repetitive motions that could lead to or aggravate repetitive motion or cumulative trauma disorders.
Guideline 12.119: Adjustable Wrist Supports Workstation keyboards and cursor control devices (e.g., trackballs and mice) should be equipped with adjustable wrist supports.
Guideline 12.120: Neutral Grip Workstation controls should use a neutral grip (neither flexed nor extended) that minimizes musculoskeletal strain.
Guideline 12.121: Limit Need to Reach Workstation tasks should limit or eliminate the need for users to reach across their body.
12.3.4.8 Compactness Space is at a premium in many care environments, particularly operating rooms and critical care environments. This places pressure on workstation designers to produce compact designs that are nonetheless functional and usable. Often, it is enough to produce a workstation that has a relatively small footprint (taking up little floor space) or to suspend some or all of the workstation’s components. However, cubic volume can also be a concern in some of the more cramped workspaces. Guideline 12.122: Compact Footrints Workstations should be designed for compactness if intended for use in environments where space is quite limited (e.g., in an intensive care unit or in an ambulance). However, their compactness should not hinder their operability (Figure 12.31).
Guideline 12.123: Intuitive Folding Mechanisms The means to configure a workstation into a more compact unit (e.g., the movement of folding components) should be intuitive. Accordingly, adjustment mechanisms should be self-evident and easy to operate while ensuring that all possible configurations are stable and secure.
12.3.4.9 Mobility (See Chapter 17, “Mobile Medical Devices”) Many workstations, both large and small, need to be mobile. Workstations, such as portable X-ray machines, need to go to the patient on an as-needed basis and otherwise stay out of the way. Therefore, with the possible exception of workstations designed to be shipped in a crate, rolled into place on a dolly, and permanently bolted in place, most workstations require features that enhance their mobility. Mobility requirements will vary depending on a workstation’s range of potential use environments (e.g., operating room, patient room, hallways, helicopters, and ambulances) and the type and number of people who will be moving it.
TAF-ER562X_WINGR-09-0401-C012.indd 514
11/9/10 12:21:24 AM
Workstations
515
FIGURE 12.31 Electrocardiograph and its cart have a relatively small “footprint” that improves access to tight workspaces.
Guideline 12.124: Ease of Relocation Generally, workstations should be designed to permit their relocation throughout the expected use environment(s). Some workstations need to accommodate limited movement within a specific area of a hospital, while others need to accommodate movement between buildings (Figure 12.32). Some workstations even need to be moved from an ambulance or helicopter to an accident site, such as a densely wooded mountainside.
Guideline 12.125: Obstacle Navigation Mobile workstations should be capable of passing over expected obstacles on the floor (e.g., power cords, raised thresholds) without significantly deviating from the intended path or requiring excessive force to overcome the obstacle (Figure 12.33). Also, workstation’s rolling mechanisms should be designed to prevent damage to expected obstacles.
Guideline 12.126: Locking Mechanisms and Brakes Workstations should incorporate features to hold them in a fixed location (e.g., locking casters) for as long as necessary. This enables workstations to remain motionless when parked on an inclined hospital corridor ramp, for example, or during medical procedures during which the patient might move or a clinician might press against the workstation. Such features should be easy to engage rapidly and should include guards against accidental release.
Guideline 12.127: Guards and Bumpers By virtue of special features (e.g., bumpers), workstations should prevent damage to other elements in the working environment (e.g., door frames, furniture, and other workstations) due to the routine, incidental impacts that frequently occur during transport.
TAF-ER562X_WINGR-09-0401-C012.indd 515
11/9/10 12:21:24 AM
516
Handbook of Human Factors in Medical Device Design
FIGURE 12.32 This CT scanner is designed to move about the hospital rather than remain fixed in the radiology unit. (Courtesy of Neurologica Corp. With permission.)
Guideline 12.128: No Loose Parts Mobile workstations should be free of loose parts that could strike objects or people while the workstation is in motion.
Guideline 12.129: Store/Secure Necessary Accessories Mobile workstations should include a means to store/secure necessary accessories (e.g., power cords, lead wires, and tubes). For example, an ultrasound workstation might include a drawer for paperwork, a bracket on which to wrap a power cord, clips to secure the sensor wands, and a “cup holder” for gel bottles.
Guideline 12.130: Minimize Weight to Ease Movement Mobile workstations should be sufficiently lightweight to enable their movement without causing undue fatigue.
FIGURE 12.33 Relatively large casters make it easier for this hospital bed to roll over wires, cables, tubes, and thresholds, for example. Note that the locking mechanism is foot actuated.
TAF-ER562X_WINGR-09-0401-C012.indd 516
11/9/10 12:21:25 AM
Workstations
517
Guideline 12.131: Label with Lifting Instructions Workstations designed for manual lifting should include a label indicating the object’s weight, specifying the number of people required to perform a safe lift, and illustrating the proper lifting technique (if not obvious).
Guideline 12.132: Strain Relief or Quick Disconnect Mechanisms Workstations should eliminate or minimize the risks associated with moving the workstation while various cords and tubes remaining connected to a fixed location (e.g., wall, patient, and so on). Accordingly, power cords and equivalent connections (e.g., patient tubes and gas supply lines) should include strain relief mechanisms to avoid damage due to moving a workstation that is still plugged in.
Guideline 12.133: Grips for Moving Workstations Mobile workstations should include nonslip grips that ensure continuous control over the workstation when it is in motion.
12.3.4.10 Stability Equipment disasters are common in television shows about hospital life. One common scene involves people accidentally tipping over a workstation, breaking off parts, spilling supplies across the floor, and ripping away cords and tubes. Fortunately, such tip-overs are uncommon in real life, partly because there are mechanical design standards that help to preclude tip-overs. However, designers should look for all possible ways to ensure workstation stability not only to protect against tip-overs and uncontrolled movement but also to enhance medical procedures. Guideline 12.134: Motion Damping Where user tasks require extreme stability, workstation motion due to anticipated force applications (e.g., the user leaning on the work surface) should be limited to an acceptable level.
Guideline 12.135: Eliminate Tip-Over Risks Workstations should not be vulnerable to tipping due to anticipated forces that could be encountered while at a standstill (e.g., someone backing into it) or while being moved (e.g., rolling over thick cable that crosses its path). Accordingly, a workstation’s center of gravity should be low enough to avoid tipping over when pushed or pulled in any likely manner.
12.3.4.11 Adjustability A core human factors principle is to adjust the device (i.e., workstation) to the user, not the user to the device. Of course, this does not rule out people adjusting their sitting position to get more comfortable. But it calls for workstation designers to provide the adjustment mechanisms necessary to match the needs of the user population. Guideline 12.136: Adjustability for Comfort and Effectiveness Workstations should be adjustable in the various ways necessary to enable comfortable and effective use.
TAF-ER562X_WINGR-09-0401-C012.indd 517
11/9/10 12:21:26 AM
518
Handbook of Human Factors in Medical Device Design
Guideline 12.137: Intuitive Adjustment Mechanisms Adjustment mechanism operation should be intuitively obvious, and the mechanism should be clearly labeled.
Guideline 12.138: Smooth Adjustment Operation Adjustment mechanisms should operate smoothly when users apply no more than a modest force.
Guideline 12.139: Dexterity Requirements for Adjustment Adjustment mechanism operation should not require undue user dexterity or concentration.
Guideline 12.140: Accessible Adjustment Mechanisms Adjustment mechanisms, such as table height controls, should be accessible to users from all potential use positions.
Guideline 12.141: Accidental Actuation of Adjustment Mechanisms Adjustment mechanisms should be protected against accidental actuation, whether by people or by contact with other equipment. This suggests the use of locking mechanisms and twostep actuations.
Guideline 12.142: Interlocks on Adjustment Mechanisms When an untimely adjustment of the workstation could be problematic or even dangerous, interlocks should be provided that require the user to perform at least two deliberate steps to make an adjustment.
Guideline 12.143: Adjustment Mechanism Coding Adjustment mechanisms should be coded (e.g., colored, textured, and/or uniquely shaped) to distinguish them from other types of controls.
Guideline 12.144: Adjustable while Multitasking Where necessary, workstations should enable users to make adjustments while simultaneously performing other tasks. For example, the adjustment mechanism might need to facilitate actuation by use of the hand, foot, knee, elbow, or even voice.
12.3.5 USER ACCOMMODATIONS The following design guidelines pertain to the features intended to accommodate the workstation users and specific physical needs associated with performing work or receiving medical care. 12.3.5.1 Seating With so much attention paid to the ergonomics of seats, it is a relatively simple task to find a comfortable, stand-alone workstation seat (i.e., an ergonomic office chair or stool). The task becomes more complicated if the given workstation requires built-in seats. In such cases, designers should exercise care to match the seat characteristics to
TAF-ER562X_WINGR-09-0401-C012.indd 518
11/9/10 12:21:26 AM
Workstations
519
the needs of the users (i.e., size and shape) and the task (i.e., the need for specific adjustments and the frequency and duration of use). The usual solution is to adapt an existing ergonomic seat—perhaps by procuring the necessary components—rather than to design an original one. Guideline 12.145: Ergonomics of Seat Design Seating should be “ergonomically correct,” meaning that seating should be designed to (1) accommodate the physical requirements of a wide range of users by virtue of its basic design and adjustability, (2) minimize physical fatigue, and (3) facilitate user tasks.
Guideline 12.146: Seat Stability Seating should have a sturdy, stable base to resist overturning.
Guideline 12.147: Seats Impervious to Fluids Seating materials should be impervious to fluids if intended for use in environments where contamination (e.g., blood splatter and/or IV fluid spills) is likely (Figure 12.34). Otherwise, seating materials should be “breathable.”
Guideline 12.148: Easy Cleaning of Seats Seating materials should be easy to clean or sterilize as warranted by the use environment.
Guideline 12.149: Seat Adjustment Mechanisms Seating adjustments should be accessible from the seated position and be intuitive to operate without the user having to look at the adjustment mechanism.
Guideline 12.150: Seat Comfort Seating should be designed for comfort over the expected duration of user tasks.
FIGURE 12.34 Operating room chairs reflect conventional office chair ergonomics while also being impervious to fluids and suitable for easy cleaning with harsh chemicals.
TAF-ER562X_WINGR-09-0401-C012.indd 519
11/9/10 12:21:26 AM
520
Handbook of Human Factors in Medical Device Design
Guideline 12.151: Lumbar Support Seating should provide lumbar support to protect patients and users against back strain.
Guideline 12.152: Seat Cushions Seat cushions should be made of a relatively dense material that compresses no more than about 1 inch (25 mm) when a user of average weight and size sits on it.
Guideline 12.153: Soft Cushion Edges Seat cushions should have a relatively soft, waterfall-type front edge to avoid placing pressure against the back of the user’s leg.
Guideline 12.154: Footrests Seating should incorporate footrests in cases where users might not be able to place their feet flat on the floor because of the seat height or angle.
Guideline 12.155: Seats Assure Postural Stability Seating should enable the user to find a stable position from which to perform precise work (e.g., eye surgery).
Guideline 12.156: Even Pressure Distribution in Seated User Seating should distribute pressure relatively evenly to the user’s buttocks and thighs in order to avoid pressure points that can cause discomfort and possibly exacerbate patient injuries.
Guideline 12.157: Seat Swiveling or Rolling Workstations that frequently require the user to rotate laterally should include swivel seating. Similarly, workstations that require the user to move greater distances between workstation elements should include rolling seats.
Guideline 12.158: Armrests on Seats Armrests should be provided on seating when they will add sufficiently to the user’s comfort, reduce the chance of arm fatigue, increase task efficiency, and provide added security (i.e., prevent the user from falling out of the seat).
12.3.5.2 Hospital Beds and Examination Tables Hospital beds and examination tables have historically been viewed as furniture. However, some of the latest generation of hospital beds incorporate advanced features, such as vibration and percussion systems, patient weight scales, nurse call systems, TV and radio controls, and bed position controls. Examination tables have also become more complex, particularly with regard to position adjustment. Thus, the U.S. Food and Drug Administration defines them both as medical devices, a broad term that includes workstations. Hospital beds and examination tables are likely to be the most common of workstations in hospitals where they are found in various care settings. These settings, ranging from emergency rooms to intensive care units, pose varying requirements based on the nature of the user—it could be a nurse or patients, for example—and the use environment. Hospital
TAF-ER562X_WINGR-09-0401-C012.indd 520
11/9/10 12:21:26 AM
Workstations
521
beds in particular pose additional requirements associated with the movement of patients throughout the hospital. Notably, workstation designers need to consider the physical requirements of a patient who is alert but also ensure the protection of sleeping or unconscious patients. Also, patients need to be comfortable while at rest and during medical procedures. Supporting surfaces should accommodate possible impairments, physical abnormalities, and injuries so as to maximize accessibility, physical comfort, emotional comfort (i.e., maintain the patient’s dignity), stability, mobility, and protection. Designers also need to consider the needs of caregivers who place patients in beds or on tables, adjust the specialized workstations to suit the particular patient’s needs, and perform specific procedures. Guideline 12.159: Bed Length Beds should be long enough to accommodate large adults and equipment that might be placed at the end of the bed, such as a defibrillator or portable patient monitor.
Many beds incorporate shelves to hold supplemental gear. Also, some devices, such as portable patient monitors, include accessory features permitting them to be hung on or attached to the bed. Guideline 12.160: Load Capacity of Beds and Tables Hospital beds and examination tables should be capable of supporting a “safe working load,” which includes combined weight of the patient, mattress, and accessories. The current standard (IEC 60601-2-38, Electrically Operated Hospital Beds [International Electrotechnical Commission, 1996]) calls for a safe working load of 382 pounds (1,700 N), which includes 303 pounds (1,350 N) for the patient, 45 pounds (200 N) for the mattress, and 34 pounds (150 N) for accessories. Based on an assessment of the needs of particularly heavy (i.e., bariatric) patients, a substantially higher load capacity might be warranted. The load capacity of shelves and other supporting mechanisms should be designed based on worst-case use scenarios. Load capacity should be indicated on hospital beds, examination tables, and associated weight-supporting features to prevent overloading.
Guideline 12.161: Mattress Design Hospital bed mattresses should ensure patient comfort and, to the extent feasible, prevent injuries associated with prolonged use (e.g., bedsores) as well as provide therapeutic benefits. For example, some mattresses are capable of varying the pressure applied to a stationary body, are sculpted to relieve pressure on the heels, and deliver intermittent percussion and vibration to the upper torso to lessen the chance of lung problems.
Guideline 12.162: Grips on Beds and Tables Grips should be provided to help patients get on and off an examination table that is more than 18 inches (457 mm) off the floor.
Guideline 12.163: Bed Guardrail Releases Guardrail release mechanisms should be protected against unintended actuation, particularly due to the forces produced when bumped by the patient, bumped into a wall, or being jolted during transport (e.g., when a bed rolls across a raised threshold).
TAF-ER562X_WINGR-09-0401-C012.indd 521
11/9/10 12:21:26 AM
522
Handbook of Human Factors in Medical Device Design
FIGURE 12.35 Bed entrapment scenarios that should be avoided by properly sizing a hospital bed’s components.
Guideline 12.164: Bed Guardrail Height Guardrails found on adult beds should be set at least 8.7 inches (220 mm) above the mattress to help prevent falls while also enabling patients to reposition themselves. Guardrails found on children’s beds should be at least 29 inches (737 mm) above the mattress to prevent younger children from climbing out.
Guideline 12.165: Bed Guardrail Safety Bed guardrails should be designed to protect users and patients from injury or discomfort due to incidental or continuous contact.
Guideline 12.166: Bed Entrapment Risks Hospital beds should be designed to prevent entrapment of patient body parts (e.g., head, neck, and/or chest) in the gaps between components, such as the guardrails (i.e., side rails), head panel, foot panel, and mattress (Figure 12.35). For more detailed guidance on proper bed component sizing, see IEC 60601-2-38 (1996) and “Hospital Bed System Dimensional and Assessment Guidance to Reduce Entrapment—Guidance for Industry and FDA Staff” (2006) from the References list for this chapter.
Guideline 12.167: Bed Control Design Bed controls should be intuitive to use by virtue of their shape, location, and labeling (Figure 12.36). Text should be avoided as the sole type of labeling because some patients might not be able to read, might not understand the text language, or might have impaired vision. IEC 60601-2-38 (1996) provides basic guidance for control labeling. Foot pedals, which are often used in conjunction with examination tables and beds, should be spaced at least 3 inches (75 mm) apart to reduce the chance of accidental actuation.
Guideline 12.168: Exam Table and Bed Height Adjustment An examination table’s height should be adjustable over the range necessary to accommodate different-size users and to facilitate tasks. The optimal height for a table supporting standing uses requiring the application of only light forces is 42 inches (107 cm). The optimal
TAF-ER562X_WINGR-09-0401-C012.indd 522
11/9/10 12:21:26 AM
Workstations
523
FIGURE 12.36 Bed control with snap-dome switches uses mostly symbolic labels in place of text. “CPR Drop” enables the hospital staff to place the bed in the ideal position to perform cardiopulmonary resuscitation. Several raised bars protect the controls from accidental actuation. height for a table supporting standing uses requiring the application of a substantial downward force is 91 cm. Ideally, tables will be adjustable across an even larger range. Hospital bed mattress height should be adjustable over a range of about 15.7 to 31.5 inches (400 to 800 mm) above the floor. Designers should review accessibility standards (e.g., Section 508 of the Rehabilitation Act) to determine appropriate and/or required ranges of height adjustment to accommodate users with special needs, such as patients in wheelchairs who must transfer on to an exam table or into a hospital bed.
Guideline 12.169: Table and Bed Adjustment Limits The range of component adjustments should be limited to prevent configurations that could cause injury to a patient (e.g., extending a patient’s knee in the wrong direction), injury to a clinician (e.g., crushing a foot placed underneath the table or bed), or instability, noting that a moving patient can apply considerable force at various points on a table or bed as well as in various directions.
Guideline 12.170: Forces Required for Bed Movement To protect caregivers from physical strain, the forces required to initiate and maintain hospital bed movement should not exceed about 36 and 19 pounds (160 and 85 N), respectively.
Guideline 12.171: Ease of Patient Transfer Examination tables built into workstations should be configured to enable patients to transfer themselves onto the table or to enable health providers to perform a safe and effective transfer. Where possible, examination tables and beds should incorporate features that facilitate patient transfers, thereby reducing the physical strain on the patient and on care providers who might need to assist the patient.
12.3.5.3 Work Surfaces Work surfaces should be sized and located to facilitate the intended tasks (e.g., entering information into a patient record or preparing a tissue sample) and the wide-ranging physical requirements of the intended user population.
TAF-ER562X_WINGR-09-0401-C012.indd 523
11/9/10 12:21:26 AM
524
Handbook of Human Factors in Medical Device Design
Guideline 12.172: Work Surface Size Work surfaces should provide ample room for the necessary accessories (e.g., telephone, computer keyboard, and charging units) and associated paperwork (e.g., medical charts).
Guideline 12.173: Work Surface Depth Conventional work surfaces (i.e., desktops) used by one person at a time should be a minimum of 10 inches (25 cm) and a maximum of 16 inches (40.5 cm) deep (Association for the Advancement of Medical Instrumentation, 1993, p. 29, 7.4.6) to ensure a sufficiently large writing surface and keep items within reach.
Guideline 12.174: Work Surface Height Generally, work surfaces should be placed at elbow height. Accordingly, fixed work surfaces should be placed at a height representing a compromise between the needs of smaller and larger individuals. One alternative is to provide a means for users to adjust the work surface heights to their preferred level. Another alternative is to provide adjustable height seating that enables users to raise or lower themselves to a comfortable height relative to a fixed work surface.
Guideline 12.175: Movable Work Surfaces Where appropriate, work surfaces should be movable (e.g., mounted on a pivoting arm).
Guideline 12.176: Minimal Work Surface Inclination Work surfaces should be flat or inclined no more than 5 degrees so that items placed on them do not roll or slide off. If a work surface is inclined, it should have a “lip” on the lower edge to prevent items, such as pens, syringes, and vials, from rolling off.
Guideline 12.177: Rounded Edges on Work Surfaces Work surface edges (including “lips”) should be rounded to avoiding chaffing the user’s arms and hands.
Guideline 12.178: Avoid Glare off Work Surfaces Work surfaces should not produce glare that causes the user visual discomfort or interferes with tasks, such as reading information on a computer display.
Guideline 12.179: Light Textured Work Surfaces Work surfaces should be lightly textured so that they are not too slippery and remain easy to clean.
12.3.5.4 Keyboards (See Chapter 7, “Controls”) An alphanumeric keyboard is a standard feature on many medical workstations because of the associated need to enter patient information, search for records, annotate forms and images, and so on. Medical manufacturers rarely design and produce their own mechanical keyboards. Rather, they select a keyboard that offers good durability and ergonomics. QWERTY key arrangements are the norm because they are ubiquitous and enable touch typing—an increasingly common skill among medical professionals. However, keyboard arrangements vary among different countries and incorporate special characters; a keyboard that is appropriate for use in the United States might not work as well in Scandinavian countries, for example. Therefore, manufacturers should offer modified QWERTY keyboards to
TAF-ER562X_WINGR-09-0401-C012.indd 524
11/9/10 12:21:26 AM
Workstations
525
suit local populations. Alternatively, touch-screen keyboards enable rapid reconfiguration, even between a QWERTY and alphabetical arrangement. The latter might be considered to accommodate nontypists or fit into a geometrically constrained space that requires a more vertical orientation. Other factors to consider when specifying a keyboard include placing it in an ergonomically correct position that is unlikely to cause cumulative trauma, protecting it from contamination, and, if necessary, getting it out of the way when it is not in use. Guideline 12.180: Key Arrangement Keyboards should be selected or designed to suit the expected amount of data entry. Accordingly, conventionally sized, QWERTY keyboards that provide good tactile feedback should be provided when the application requires extensive data entry or there is no reason to compromise the keyboard design. Other types of keyboards, such as touch-screen keyboards, downsized keyboards, and alphabetically arranged keyboards, could be appropriate solutions when the application requires less data entry and will be used by inexperienced typists.
Guideline 12.181: Keyboard Placement Where possible, keyboards should be placed directly in front of the user and at a comfortable height (elbow height) to enable the user to touch-type while maintaining a neutral wrist position (Figure 12.37).
Guideline 12.182: Wrist Supports Keyboards associated with extensive data entry tasks should incorporate wrist supports as a means to prevent repetitive motion disorders.
Guideline 12.183: Impervious Keyboard Keyboards that will be exposed to spilled or splattered liquids (i.e., urine, blood, IV solutions, and so on) should be impervious to contamination and facilitate rapid cleanup.
Guideline 12.184: Keyboard Stowage When a keyboard will be used infrequently, it should be possible to move it out of the way of other tasks.
FIGURE 12.37 Keyboard is placed at a comfortable height and angle to facilitate data entry (e.g., inputting nursing notes) but lacks a wrist support.
TAF-ER562X_WINGR-09-0401-C012.indd 525
11/9/10 12:21:26 AM
526
Handbook of Human Factors in Medical Device Design
12.3.5.5 Foot Controls (See Chapter 7, “Controls”) Some workstations, such those used to perform fluoroscopy, incorporate foot controls to allow users to control primary functions while using their hands for other purposes. Recognizing that people are less dexterous with their feet and that users will naturally obtain less tactile feedback through their feet (presumably covered by footwear) than their hands, designers need to limit the use of foot controls to appropriate actions. For example, it might be appropriate to use a foot control to start and stop a function or even vary its amplitude. Foot controls should not require fine motor control because most people cannot move their feet as precisely as they can move their hands. Also, foot controls require protection against accidental use and can pose a tripping hazard. Guideline 12.185: Foot Control Actuation Foot controls should not require especially precise adjustments that exceed people’s fine motor control capability. Rather, they should require only low to moderately precise user inputs by means of gross motor control.
Guideline 12.186: Footwear Compatibility Foot controls should be operable by people wearing footwear common to the particular work environment. For examples, emergency medical technicians and paramedics usually wear heavy boots, while some hospital workers wear “clogs.”
Guideline 12.187: Foot Control Feedback Variable position foot controls should provide unambiguous feedback to the user regarding the control’s current position. Specifically, foot controls should incorporate a greater degree of travel and require greater actuation forces than normally associated with hand controls. Moreover, foot controls should be operable without having to look concurrently at the controls. (Figure 12.38).
Guideline 12.188: Unintended Actuation of Foot Controls Foot controls should be protected from unintended actuation.
FIGURE 12.38
Users are likely to operate a large foot control without looking at it.
TAF-ER562X_WINGR-09-0401-C012.indd 526
11/9/10 12:21:27 AM
Workstations
527
Guideline 12.189: Multiple Foot Controls Multiple foot controls should be separated by a sufficient distance to prevent users from unintentionally actuating more than one control at a time. Also, they should be physically and/or operationally distinct to prevent the user from confusing one with another.
Guideline 12.190: Foot Control Tripping Hazards Foot controls should not present a tripping hazard.
12.3.5.6 Remote Controls (See Chapter 7, “Controls”) Remote controls might be needed when user access to a workstation is limited by the presence of other equipment and people or users need to stand away from the workstation during certain procedures (e.g., procedures involving electromagnetic radiation). A properly designed remote control should be as usable as the primary control(s) it replaces. Designed poorly, a remote control can interfere with workstation operations and potentially lead to use error. Guideline 12.191: Need for Remote Remote controls should be provided when they will increase the safety and efficiency of medical procedures. For example, remote controls should be provided when the user could be positioned a substantial distance from the workstation and require rapid access to controls to ensure patient safety and/or therapeutic effectiveness. They should be provided when they will reduce the users’ stress or physical strain.
Guideline 12.192: Not to Critical Functions A remote control should not be required to perform critical, safety-related workstation functions, except where the user must perform operations remotely for user safety (e.g., operating an X-ray machine).
Guideline 12.193: Remote versus Primary Controls A remote control should reflect the same basic control scheme and associated design characteristics as a workstation’s primary controls. Differences between remote controls and primary controls should be limited to sizing and completeness. Remote controls should be compact and reliable and include only essential functions.
Guideline 12.194: Remote Control Labeling Workstation remote controls should be labeled so that users can identify their general purpose and specific functions, particularly when they could be placed alongside other remote controls (e.g., remote controls for a bed and a patient-controllable analgesic pump).
Guideline 12.195: Remote Control Feedback When workstation functions are remotely controlled, there should be a means to provide users with the necessary feedback to ensure safe and effective operation.
Guideline 12.196: Loss Prevention of Remote Controls Remote controls that perform a critical function should incorporate a means (e.g., tethering) to prevent them from being removed from the operating area, lost, or stolen (Figure 12.39).
TAF-ER562X_WINGR-09-0401-C012.indd 527
11/9/10 12:21:27 AM
528
FIGURE 12.39
Handbook of Human Factors in Medical Device Design
Tethered remote control lets patients communicate directly with caregivers.
Guideline 12.197: Remote Control Cords Remote control cords should resist tangling, enable easy retrieval, and minimize the risk of cord damage.
12.3.5.7 Grips and Handles Many workstations require considerable handling. Users frequently move them between and within use locations, such as operating rooms and patient rooms. Also, caregivers and patients might need to stabilize themselves continually or intermittently by holding on to part of a given workstation. Therefore, designers should give careful consideration to the number, type, and placement of grips and handles—components that may be one and the same. Guideline 12.198: Provision of Multiple Grips Generally, workstations should provide multiple grips to accommodate different user positions.
Guideline 12.199: Prevent Improper Gripping Grips should prevent or discourage users from gripping the workstation in an improper manner that could pose a physical risk (e.g., cause wrist strain) or damage the workstation (e.g., break a component).
Guideline 12.200: Impact Resistance of Grips Grips should be able to withstand collisions with other equipment and walls, for example.
Guideline 12.201: Nonslip Grips Nonslip grips should be provided where they will enable users to position and stabilize themselves. Grips should be suitable for use by a bare or gloved hand that could be covered with substances common to medical environments, such as blood, saline, bodily fluids, antiseptic solution, or powder (Figure 12.40).
Guideline 12.202: Visual Distinction Grips should be visually distinct from controls and other workstation elements that should not be used as grips (Figure 12.41).
TAF-ER562X_WINGR-09-0401-C012.indd 528
11/9/10 12:21:27 AM
Workstations
FIGURE 12.40
529
Rugged grip provides a sturdy, slip-resistant means to move this workstation.
Guideline 12.203: Conspicuous Grips Grips should be visually conspicuous to permit their rapid identification and to encourage proper workstation handling (Figure 12.41).
Guideline 12.204: Grip Style Grips should enable a power grip (i.e., a full, closed palm grip) as opposed to alternative grips (e.g., a pinch grip) that could be insecure and cause strain.
Guideline 12.205: Strain Resistance Grips should allow the user to maintain a neutral wrist position that will help to avoid strain.
Guideline 12.206: Grip Separation From Other Components Handgrips should be spatially separated from controls and delicate components to avoid unintended control actuations or component damage caused by grasping the wrong component.
FIGURE 12.41
Grips stand out clearly from other components.
TAF-ER562X_WINGR-09-0401-C012.indd 529
11/9/10 12:21:27 AM
530
Handbook of Human Factors in Medical Device Design
12.3.5.8 Supports and Restraints There are passive and active restraints. A workstation may incorporate passive support or restraint features, such as a contoured seating surface, or active ones, such as a wholebody harness. Designers should perform an analysis of the potential user positions associated with medical procedures as well potential security needs to determine what types of supports and restraints are warranted. Such an analysis should consider needs associated with patients who might be physically aggressive, uncooperative, spastic, weak, or unconscious. Guideline 12.207: Patient Restraint Design Restraints should be provided where necessary to hold the patient in the proper position for the associated procedure. In most cases, such restraints should serve to protect a patient from harm. In some cases, they might also be needed to protect caregivers from a spastic or unruly patient. Such restraints should not release unintentionally. They should enable users or patients to release themselves deliberately, except where such action could be hazardous or restraint is mandated (e.g., use in a psychological unit). The means to disconnect restraints should require a quick, one-handed operation. Mobile workstations should incorporate a way to stow restraints that are not in use so that they do not become entangled with or damaged by other objects during transit.
Guideline 12.208: Workstation Headrests Workstations should incorporate headrests when the user or patient will be placed in positions, such as tilted backward, that could cause neck strain or injury, particularly to patients with neck ailments.
Guideline 12.209: Workstation Armrests Workstations should incorporate adjustable armrests if they will contribute to user comfort and help them stabilize themselves while at rest and during transfers (Figure 12.42).
Guideline 12.210: Workstation Footrests Workstations should incorporate adjustable footrests if they will contribute to the user’s comfort and help them stabilize themselves while at rest and during transfers.
Guideline 12.211: Cushioning of Patient Contact Points Where appropriate, workstations should be cushioned at body contact points to ensure patient and user comfort.
12.3.6 SURFACE CHARACTERISTICS A workstation’s surface (including cosmetic) characteristics can have a strong influence on its functional effectiveness as well as user satisfaction. 12.3.6.1 Appearance All medical workstations serve a diagnostic and/or therapeutic purpose. In theory, their appearance should be driven by functional requirements, just like space vehicles or earthmoving equipment. However, most users care about a workstation’s appearance because it has emotional impact, and can influence perceptions of health care quality, create a sense
TAF-ER562X_WINGR-09-0401-C012.indd 530
11/9/10 12:21:28 AM
Workstations
531
FIGURE 12.42 A-dec 300 Dental Chair incorporates ergonomic features to make both the patient and the dentist more comfortable. (Courtesy of A-dec.)
of professional pride, and make medical procedures seem less intimidating. As such, the appearance of medical workstations is important. Guideline 12.212: Workstation Appearance Is Important Workstations should not be overly intimidating to patients (particularly children) or users because of factors such as a harsh-looking color scheme or the unnecessary exposure of threatening or complex-looking mechanisms. Rather, workstations should possess a visual style and incorporate features that make them appear well built, constructed of quality materials, comfortable, effective, efficient, and safe. Given the subjectivity of these perceived characteristics, developers should test alternative designs with users to determine which ones are most preferred.
12.3.6.2 Color Workstation color is an important functional as well as aesthetic consideration. The right color can facilitate proper maintenance (i.e., cleaning), enhance the readability of labels, diminish glare from overhead lighting, and help to reduce patient anxiety, for example. The wrong color can have the opposite effects. Accordingly, workstation colors should be chosen carefully. Guideline 12.213: Light Colors Show Contamination Where cleanliness (or sterility) is paramount, workstations should be colored in a manner that reveals contamination. Normally, this means using a light color scheme. However, where workstations will be cleaned less frequently and this is not considered detrimental to operability or safety, exposed surfaces should be colored to mask the normal grime that accumulates during routine use. Typically, this means using a somewhat darker color scheme, although
TAF-ER562X_WINGR-09-0401-C012.indd 531
11/9/10 12:21:28 AM
532
Handbook of Human Factors in Medical Device Design
very dark schemes are prone to show scuffing and dust and can give the workstation an ominous or intimidating appearance.
Guideline 12.214: Comply with Color Conventions Critical workstation components should be color-coded in a manner consistent with established industry conventions for warnings, medications, gases, specific functions, etc.
Guideline 12.215: Harmonization with Work Environment Workstations should be colored in a manner that will harmonize with other medical equipment rather than drawing undue attention to themselves. That said, some workstation features, such as an emergency stop button, should be colored in a manner that readily draws attention.
Guideline 12.216: Durability of Workstation Colors Colors that convey important information should have a consistent appearance that is not subject to change because of aging, exposure to radiant energy, repeated cleaning, or other factors. Color durability is also important to preserving a workstation’s visual appeal.
12.3.6.3 Material Finish Similar to a workstation’s color, a workstation’s material finish(es) can influence user performance and satisfaction. A given workstation might incorporate several different surface finishes to achieve functional goals and to ensure design appeal. There is a role for smooth finishes versus textured ones, warm-feeling finishes versus cool ones, and soft, forceabsorbing surfaces versus hard ones. Guideline 12.217: Durable Finishes Material finishes should resist wear and damage caused by the most extreme use expected over the workstation’s intended life cycle.
Guideline 12.218: Texture of Workstation Surfaces Material finishes should be textured to provide traction (i.e., friction) where it is helpful to secure objects and enable patients and users to move or stabilize themselves. Conversely, they should be smooth where a slippery surface offers functional advantages, such as making it easier to slide a heavy object into position or clean a surface.
Guideline 12.219: Feel of Workstation Finishes As appropriate, workstation components that physically contact people (e.g., handgrips) should have warm-feeling material finishes, such as plastic coatings that have a suede-like texture. Such finishes should be sufficiently durable and cleanable.
Guideline 12.220: Cushioning of Workstation Finishes Certain workstation surfaces should be cushioned to ensure user comfort and absorb shock.
12.3.6.4 Cleanliness Certain kinds of workstations, like an operating room table, receive a thorough cleaning several times a day. Others might receive a thorough cleaning far less often. Therefore,
TAF-ER562X_WINGR-09-0401-C012.indd 532
11/9/10 12:21:28 AM
Workstations
533
designers should give careful consideration to the likely cleaning routines associated with their particular workstation. Such consideration might lead to designs that show or hide contamination, for example. In all cases, designers should design workstations with ease of cleaning in mind. Guideline 12.221: Workstation Ease of Cleaning Workstations should be designed for easy and effective cleaning, particularly following a therapeutic or diagnostic procedure. As a rule, seamless and smooth surfaces will stay cleaner and will be easier to thoroughly clean than textured surfaces that have multiple seams.
Guideline 12.222: Cleaning Methods Material finishes should be easy to clean using means that are conventional within the expected use environments.
Guideline 12.223: Damage Resistance with Cleaning Workstations should resist damage caused by cleansing solutions that are commonly used in the use environment regardless of whether they are the correct, specified solution for cleaning the workstation.
Guideline 12.224: Minimize Risk of Contamination Workstations should be designed to eliminate or minimize hygienic risks and facilitate cleaning in accordance with expected protocols. This requirement can place substantial restrictions on the workstation’s physical form, for example, by prohibiting the use of certain porous materials and the presence of grooves between parts that can trap bacteria (Figure 12.43).
Guideline 12.225: Easy Sterilization of Components As warranted, workstations should enable users to sterilize specific components. This need might necessitate the use of removable components. It might also require the workstation to tolerate washing with fluid soaked sponges, for example.
FIGURE 12.43
Grooves, gaps, slots, and holes present cleaning challenges.
TAF-ER562X_WINGR-09-0401-C012.indd 533
11/9/10 12:21:28 AM
534
Handbook of Human Factors in Medical Device Design
Guideline 12.226: Splash Guards and Protective Covers Workstations used in environments where there is a likelihood of contamination (e.g., blood splatter and/or IV fluid spills) should incorporate special features (e.g., covers and/or splash guards) that limit the contamination and facilitate cleaning and maintenance.
12.3.6.5 Maintenance When designing a workstation for ease of use, one should not overlook the need for easy maintenance. After all, many medical workstations require frequent, perhaps even daily, maintenance (including calibration). Maintenance workers have the same need and desire as care providers to work productively and without hindrance. Guideline 12.227: Ease of Maintenance Workstations should include features that facilitate efficient and effective maintenance, such as back panels that are easy to remove without the need for special tools.
Guideline 12.228: Indicate When Maintenance Due Workstations should indicate when maintenance is required or past due. This can be accomplished by incorporating a computer-based maintenance log and associated alert system to indicate when maintenance is due.
Guideline 12.229: Indicate Performance Degradation Risk Workstations should indicate when deferred maintenance causes a significant reduction in their operational effectiveness.
Guideline 12.230: Indicate Need for Consumables Workstations should indicate when the amount of a consumable, such as a chemical reagent, is getting low and will soon need replenishment. They should also indicate when the amount of a consumable is too low to complete another procedure, such as a blood chemistry analysis.
Guideline 12.231: Emergency Maintenance Support Workstations should be equipped with the parts, tools, and instructions that are necessary to make any emergency adjustments or repairs that are required to ensure patient safety.
Guideline 12.232: Authorized Maintenance Personnel Only A warning should be provided in cases where users might attempt to perform maintenance tasks that should be performed only by trained maintenance personnel.
12.3.7 ENVIRONMENTAL FACTORS (SEE CHAPTER 3, “ENVIRONMENT OF USE”) The following design guidelines pertain to optimizing the environmental factors associated with a workstation so that people can interact with it in a comfortable manner.
TAF-ER562X_WINGR-09-0401-C012.indd 534
11/9/10 12:21:28 AM
Workstations
535
12.3.7.1 Task Lighting Surprisingly, caregivers often struggle to see what they are doing while operating certain workstations. This is particularly true in care environments in which the lighting level is kept low, for example, to facilitate visualization during minimally invasive surgical procedures. Therefore, designers should consider the need for workstation task lighting to augment the normal lighting sources in certain use scenarios. Guideline 12.233: Workstation Task Lighting Workstations should incorporate task lighting where area or ambient lighting will not be sufficient to facilitate tasks.
Guideline 12.234: Assure Support for Critical Tasks Given the potential for a sudden loss of room or ambient lighting, critical functions should be facilitated in dim light or total darkness. Accordingly, critical controls might need to be spotlighted or backlighted or provide sufficient tactile cues and feedback to ensure proper operation in the dark.
Guideline 12.235: Adjustable Task Lighting As appropriate, task lights should have readily accessible adjustments for brightness, focus, and direction.
Guideline 12.236: Task Lighting Induced Glare Task lights should not cause glare that obscures the user’s view of critical displays and controls or interfere with activities necessitating dim lighting (e.g., certain minimally invasive surgical procedures during which room lights are turned off).
Guideline 12.237: Battery-Powered Task Lighting Where lighting is essential to the performance of a critical task, a battery-powered light source should be available in the event of a workstation or facility power failure.
12.3.7.2 Noise (See Chapter 3, “Environment of Use”) People have different levels of tolerance for noise. While one person might not object to noise in the care environment, another person might find it quite distracting. As such, workstation designers should pay close attention to the kind and level of noise emission, seeking to limit device-induced noise wherever possible. Guideline 12.238: Limit Workstation-Induced Noise Workstations should not produce noise that is disturbing to users.
Guideline 12.239: Avoid Masking Auditory Cues Workstations should not produce noise that masks critical auditory cues generated by the workstation or other equipment in the care environment.
12.3.7.3 Vibration Workstations often incorporate moving mechanisms that can cause vibration. Such vibrations might be disturbing to the user or patient.
TAF-ER562X_WINGR-09-0401-C012.indd 535
11/9/10 12:21:28 AM
536
Handbook of Human Factors in Medical Device Design
Guideline 12.240: Limit Workstation-Induced Vibration Workstations should not produce vibration that is disturbing to users.
12.3.7.4 Venting Workstations often produce undesirable by-products, such as noxious odors and heat. Such by-products might be disturbing to the user or patient. Guideline 12.241: Workstation Venting Workstations should not vent directly onto the user or patient.
Guideline 12.242: Workstation Heat Emission The heat emitted by a workstation should be limited to ensure that the use environment does not become uncomfortably warm.
12.4 CASE STUDIES The following three case three studies exemplify the effective application of human factors principles to the design of an anesthesia workstation, an ultrasound imaging workstation, and a hospital bed.
12.4.1 ANESTHESIA WORKSTATION GE Healthcare’s Datex-Ohmeda ADU Plus Carestation is a sophisticated anesthesia workstation that integrates multiple functions, including anesthetic agent vaporization, gas delivery to the patient, patient monitoring, and anesthesia record keeping (information management) (Figure 12.44). In the course of developing the workstation, the company conducted extensive user needs research in hospitals worldwide to define functional requirements and user-interface design preferences. This research led to detailed design specifications and a style guide for the workstation’s user interface. Then the company developed several prototype user-interface designs and evaluated them through usability testing in an operating room simulator. The workstation’s user-interface features the following: • • • • •
Up to three flat-screen monitors to provide complete information simultaneously Easy physical and visual access to surface-level displays and controls Several drawers to store medication, tools, and disposable items Large handgrips to facilitate moving the workstation A relatively small footprint that enables the user to place the workstation in an optimal position • A shelf to support supplemental gear • An expandable work surface for preparing medications and completing nonelectronic paperwork • An integrated alarm system The design incorporates many features common to state-of-the-art patient monitors into the workstation, eliminating the need for customers to assemble several stand-alone devices
TAF-ER562X_WINGR-09-0401-C012.indd 536
11/9/10 12:21:28 AM
Workstations
(a)
537
(b)
FIGURE 12.44 (a) The Datex-Ohmeda ADU Plus Carestation. (b) The anesthesia workstation in the operating room. (Courtesy of GE Healthcare. With permission.)
(e.g., patient monitors, anesthesia gas machine, and ventilator). The workstation’s high degree of functional integration provides the user with a comparatively simple and consistent means to acquire clinical information and take appropriate actions by means of a few, common controls. The manufacturer’s major design challenges were to: • Determine which displays and controls needed to be available at all times (i.e., on the surface) while also limiting the workstation’s visual complexity • Ensure that users would feel comfortable transitioning from a mechanical means to a computer-based means of controlling functions, such as setting the amount of gas flow to the patient • Produce a final ergonomic form that was sufficiently compact and mobile to enable users to position it near the patient while providing the user with a comfortable place to work from either a sitting or a standing position • Accommodate the needs of users in many countries who have different approaches to anesthesia delivery, including working alone versus in teams
12.4.2 ULTRASOUND IMAGING WORKSTATION GE Medical System’s LOGIQ 9 ultrasound workstation is an advanced diagnostic tool with a modern appearance that was designed to be more reassuring rather than intimidating to patients (Figure 12.45). The user-friendly appearance is due in part to the workstation’s
TAF-ER562X_WINGR-09-0401-C012.indd 537
11/9/10 12:21:29 AM
538
Handbook of Human Factors in Medical Device Design
FIGURE 12.45 The LOGIQ 9 Advanced Ultrasound Workstation. (Courtesy of GE Healthcare. With permission.)
rounded corners and soft color scheme (ivory and medium blue) as well as the user interface’s orderly appearance that belies the extensive number of surface-level controls and displays. In fact, the workstation has over 100 surface-level features that enable technicians to work efficiently. User-centered design features include the following: • Placement of the keyboard at an ergonomically correct height • Placement of the computer screens at angles that provide a roughly orthogonal view to the user • Functional grouping, shape coding, and color coding (light blue) of the image adjustment controls • Built-in holders for the imaging wands and bottles of conductive gel
TAF-ER562X_WINGR-09-0401-C012.indd 538
11/9/10 12:21:29 AM
Workstations
539
• Storage bins for workstation accessories • Multiple, soft-edged grips to facilitate moving the workstation The workstation design represents the natural evolution of the company’s ultrasound imaging product line. Technicians familiar with previous models should be able to operate the LOGIQ 9 with little additional training. Despite its increased functional capabilities, which include advanced three-dimensional imaging and hemodynamic flow analyses, the workstation’s footprint remains about the same, ensuring that it will fit into spaces already dedicated to ultrasound imaging. The manufacturer’s major design challenges were to: • Ensure that the user interface presented the correct display and control features on the front panel versus computer screen so that technicians could make a smooth transition to the workstation and maintain or increase their productivity • Produce a visually pleasing design without compromising the workstation’s operation and usability • Retain all the design features that technicians prefer, such as trackball control over on-screen images and menus and slider control over image quality, while pursuing design advances, such as on-screen recording and playback controls.
12.4.3 HOSPITAL BED Hill-Rom’s TotalCare® Connect Bed is capable of transforming itself from a hospital bed into a chair while meeting the physical needs of a diverse user population (Figure 12.46). By offering patients the choice of several supine and seated positions, it enhances patient health and comfort. The bed also enhances caregiver comfort and safety during patient transfers by improving physical access to the patient. While developing the bed, the company performed extensive anthropometric analyses of male and female patients. Human factors design studies extended beyond properly sizing the bed and determining how it should articulate to studies of possible bed rail configurations and control layouts. The designers converged on the final design by means of extensive user testing, including the use of advanced prototypes during usability tests. Notable user interface design features include the following: • Multiple gripping features that make it easier for hospital staff to move the bed and that help both caregivers and patients stabilize themselves • Large casters that enable the bed to overcome low-lying obstacles during relocation • Backlit, intuitive controls that can be operated by individuals who are illiterate, have somewhat degraded vision, or do not speak the native language • A flat control panel (embedded in the bed rails) that does not snag clothing, will not jut uncomfortably into the patient’s body, is impervious to fluid contamination, and is easy to clean • A pivoting, touch-screen display used to control the bed’s advanced features, including in-bed patient weighing, percussion and vibration therapy to help keep a patient’s lungs clear, and the movement of air throughout the mattress to help avoid pressure sores
TAF-ER562X_WINGR-09-0401-C012.indd 539
11/9/10 12:21:30 AM
540
FIGURE 12.46
Handbook of Human Factors in Medical Device Design
TotalCare® Connect Bed. (Courtesy of Hill-Rom. With permission.)
• A cushioned surface that hinges at the optimal points to accommodate various-size patients • A large foot control used to lock the bed in place (i.e., apply the brakes) Considering its numerous functions, the bed is not particularly intimidating, an important consideration for hospitals that strive for positive patient experiences. The designers used material finishes and colors that look clean, modern, and comfortable. The manufacturer’s major design challenges were to: • Match the bed’s functional capabilities not only to the patient’s need for comfort in various positions but also to the caregivers who take care of the patients’ needs, including performing various medical procedures at the bedside • Ensure the patient’s safety at all times by providing proper cushioning, avoiding pinching hazards during bed articulation, alarming if a nonambulatory patient tries to get out of bed, and ensuring that bed articulations would not interfere with ongoing therapies, such as an IV infusion • Produce a design that functions in a manner familiar to caregivers accustomed to operating the company’s other hospital beds
RESOURCES American National Standards Institute/Association for the Advancement of Medical Instrumentation (ANSI/AAMI). (2010). Human Factors Design Process for Medical Devices. ANSI/AAMI HE-75-2010. Arlington, VA: Association for the Advancement of Medical Instrumentation. Canadian Centre for Occupational Health and Safety. http://www.ccohs.ca/oshanswers/ergonomics/ office. Eastman Kodak Company Health and Environmental Laboratories, Ergonomics Group. (1989). Ergonomic Design for People at Work. (Vol. 1). New York: John Wiley & Sons. Salvendy, G. (Ed.). (2006). Handbook of Human Factors (3rd ed.). New York: Wiley InterScience. Sanders, M., and McCormick, E. (1993). Human Factors in Engineering and Design (7th ed.). New York: McGraw-Hill. U.S. Department of Defense. (1996). Human engineering design criteria for military systems, equipment, and facilities. MIL-STD-1472F. Washington, DC: U.S. Department of Defense. U.S. Department of Labor, Occupational Safety and Health Administration. Website: http://www. osha.gov/SLTC/etools/computerworkstations/index.html. Accessed on 6-1-09.
TAF-ER562X_WINGR-09-0401-C012.indd 540
11/9/10 12:21:30 AM
Workstations
541
Wiklund, M. (1995). Medical Device and Equipment Design. Boca Raton, FL: CRC Press. Wiklund, M. and Wilcox, S. (2005). Designing Usability into Medical Products. Boca Raton, FL: CRC Press. Woodson, W., Tillman, B., and Tillman, P. Human Factors Design Handbook (2nd ed.). New York: McGraw-Hill Professional.
REFERENCES Bogner, M. S. (Ed.). (1988). Human Error in Medicine. Hillsdale, NJ: Lawrence Erlbaum Associates. Gosbee, J. (2002). “Human Factors Engineering and Patient Safety.” Quality and Safety in Healthcare 11:352–354. http://qshc.bmj.com/cgi/content/full/11/4/352?maxtoshow=&HITS=10& RESULTFORMAT=&fulltext=gosbee&searchid=1&FIRSTINDEX=0&sortspec=relevance &resourcetype=HWCIT. Accessed on 6-1-09. International Electrotechnical Commission. (1996). Electrically Operated Hospital Beds. IEC 60601-2-38. Geneva: International Electrotechnical Commission. Norman, D. A. (1988). The Psychology of Everyday Things. New York: Basic Books. North Carolina State University, Center for Universal Design. (2004). Web site: http://design.ncsu. edu/cud/. Accessed on 6-1-09. United States Food and Drug Administration. Hospital Bed System Dimensional and Assessment Guidance to Reduce Entrapment—Guidance for Industry and FDA Staff. Issued March 10, 2006. Website: http://www.fda.gov/cdrh/beds/guidance/1537.html. Accessed on 6-1-09.
TAF-ER562X_WINGR-09-0401-C012.indd 541
11/9/10 12:21:30 AM
TAF-ER562X_WINGR-09-0401-C012.indd 542
11/9/10 12:21:30 AM
13 Signs, Symbols, and Markings Michael J. Kalsher, PhD; Michael S. Wogalter, PhD CONTENTS 13.1 What Should Be Labeled? .....................................................................................545 13.1.1 Who Sets the Rules? ..................................................................................545 13.2 General Labeling Requirements ............................................................................547 13.2.1 Label Content ............................................................................................547 13.2.1.1 Identity Statement.......................................................................547 13.2.1.2 Use Statement .............................................................................549 13.2.1.3 Handling Statement .................................................................... 551 13.2.1.4 Warning/Precaution Statements ................................................. 551 13.2.1.5 Specific Hazards Statements ......................................................552 13.2.1.6 Investigational Use Statement ....................................................553 13.2.2 Label Form ................................................................................................553 13.2.3 Location and Size ......................................................................................554 13.2.3.1 Prominence.................................................................................554 13.2.3.2 Principal Display Panel ..............................................................554 13.3 Labels for Device Identification, Instructions, and Hazards ..................................554 13.3.1 Labels for Electromechanical Components ...............................................556 13.3.1.1 Electrical Receptacle and Connector Labels ..............................556 13.3.1.2 Fuse and Circuit-Breaker Labels ................................................557 13.3.2 Section Summary ......................................................................................557 13.4 Human Factors Principles for Designing Medical Device Labels .........................558 13.4.1 Safety-Related Goals of Labels .................................................................558 13.4.2 Hazard Control Hierarchy .........................................................................559 13.4.3 Medical Device Liability and Warning Labels..........................................560 13.4.4 Communication-Human Information Processing (C-HIP) Model.............560 13.5 Components of the C-HIP Model ..........................................................................562 13.5.1 Source ........................................................................................................562 13.5.2 Channel......................................................................................................562 13.5.3 Delivery .....................................................................................................562 13.5.4 Receiver .....................................................................................................562 13.5.4.1 Attract Attention.........................................................................563 13.5.4.2 Hold Attention ............................................................................564 13.5.4.3 Label Comprehension.................................................................564 13.5.4.4 Fit with User Beliefs and Attitudes ............................................565 13.5.4.5 Motivation ..................................................................................566 543
TAF-ER562X_WINGR-09-0401-C013.indd 543
11/9/10 3:45:58 AM
544
Handbook of Human Factors in Medical Device Design
13.5.4.6 Label Must Produce Compliance ...............................................567 13.5.4.7 Labeling May Influence Users Indirectly ...................................567 13.5.5 Other Human Factors Issues ......................................................................567 13.5.5.1 Social Influence ..........................................................................567 13.5.5.2 Stress and Workload ...................................................................567 13.6 Using Human Factors Principles to Enhance Components of Medical Device Labeling .....................................................................................................568 13.6.1 Label Content ............................................................................................568 13.6.1.1 Abbreviations, Initials, and Symbols..........................................569 13.6.2 Location Aids and Functional Relationships .............................................570 13.6.3 Position and Placement ..............................................................................571 13.6.3.1 Visibility .....................................................................................571 13.6.3.2 Orientation..................................................................................571 13.6.3.3 Shape ..........................................................................................572 13.6.3.4 Location......................................................................................572 13.6.4 Gestalt Principles .......................................................................................573 13.6.5 Population Stereotypes and Expectations ..................................................574 13.6.6 Durable Materials ......................................................................................575 13.6.7 Legibility ...................................................................................................576 13.6.7.1 Highlighting and Contrast ..........................................................577 13.6.7.2 Typography.................................................................................579 13.6.7.3 Other Strategies ..........................................................................581 13.6.8 Coding .......................................................................................................581 13.6.8.1 Redundant Codes .......................................................................582 13.6.8.2 Color Coding ..............................................................................582 13.6.8.3 Size Coding ................................................................................586 13.6.8.4 Location Coding .........................................................................586 13.6.8.5 Shape Coding .............................................................................586 13.6.8.6 Graphics and Symbols................................................................587 13.6.9 Section Summary ......................................................................................587 13.7 Conclusions ............................................................................................................588 Acknowledgments............................................................................................................589 References ........................................................................................................................589
Recent advances in medicine, science, and technology have led to a considerable and growing number of medical devices. Technical innovations have generally been welcomed by health care providers and the general public. However, technological sophistication does not necessarily mean that such devices can be used effectively and safely by users. Medical device use has been shown to be associated with hazards to patients and clinicians. Virtually all medical devices include labels and markings. Labels can assist users in correctly operating a device and at the same time reduce the likelihood of use error. Medical device use environments have expanded beyond doctors’ offices and hospitals to outpatient, community, and home care. The intended users have broadened from trained medical professionals (e.g., physician, nurse, or other health care provider) to include lay patients and caregivers. A variety of users and use environments require different ways
TAF-ER562X_WINGR-09-0401-C013.indd 544
11/9/10 3:45:59 AM
Signs, Symbols, and Markings
545
of presenting information. Designers must consider multiple use-related factors to design appropriate labels. Medical device labeling consists of directions on how to use and care for medical devices as well as information necessary for ensuring users’ understanding and safety, including information about risks, precautions, and potential adverse reactions. This chapter provides guidance to designers for their decisions regarding positioning, formatting, and designing of labels and markings for controls, displays, panels, and associated equipment. The information presented here is derived from legal requirements, human factors research, and existing practices. The chapter is divided into three sections, organized around key design issues. The first section addresses the critical issue of what should be labeled and focuses on the specific legal requirements and voluntary standards that guide medical device labeling in the United States. The second section is an overview of relevant principles from the human factors literature that designers should consider when making labels for a medical device. This section uses a communication-human information processing (C-HIP) framework as a means of organizing the labeling literature. Designers can use this conceptual model as a developmental tool, and investigators can use it as an analytical tool. The third section provides specific recommendations for developing effective medical device labels. Examples of label designs across a range of medical devices are presented. In general, designers need to consider numerous factors in deciding how to label a device. Label characteristics that should be considered include color, anticipated viewing distances and illumination levels, time constraints of users, and understandability criteria, among many others. Labels on medical devices should appropriately attract and hold attention, be understandable and believable, and motivate users to comply with the directives they present. In addition, designers should take into account local conventions and meanings associated with specific markings as well as the reading abilities, visual acuity, and other relevant characteristics of the user population. For example, older adults and people with disabilities have different medical device labeling needs compared to health care professionals. Controls, displays, and other components of medical devices should be appropriately and clearly labeled to permit rapid and accurate human performance. The importance of gathering user input—to meet the needs of users—when designing medical device labels is emphasized. The general principles of usability testing as it relates to the evaluation of medical device labeling, including the basic processes involved in iterative design and testing, is detailed in Chapter 6, “Testing and Evaluation.”
13.1 WHAT SHOULD BE LABELED? This section reviews the regulations for medical device labeling.
13.1.1 WHO SETS THE RULES? In the United States, medical device labeling is regulated by the Center for Devices and Radiological Health (CDRH) of the Food and Drug Administration (FDA), a part of the U.S. Department of Health and Human Services. The regulations for medical device labels are provided in the Code of Federal Regulations (CFR). Title 21 of the
TAF-ER562X_WINGR-09-0401-C013.indd 545
11/9/10 3:45:59 AM
546
Handbook of Human Factors in Medical Device Design
CFR concerns food and drugs, and medical device labeling is discussed in Part 801 (Title 21—Food and Drugs). According to the FDA, a label is “a display of written, printed or graphic matter upon the immediate container of any article.” Labeling, a more inclusive term, is defined as “all labels and other written, printed, or graphic matter (1) upon any article or any of its containers or wrappers, or (2) accompanying such an article.” In Europe and some other parts of the world, medical device labeling is governed by regulations established by the European Economic Area (EEA) (Study Group 1 of the Global Harmonization Task Force, 2002 http://www.ghtf.org retrieved 7/3/09). Products that meet these requirements receive a stamp of approval termed the “CE mark” that allows them to be marketed in European Union (EU) countries (refer to Figure 13.1). EU products do not have to be evaluated by a third party to receive the CE mark (see Underwriters Laboratories, Inc., n.d.). Instead, the mark is provided contingent on the manufacturer’s word that the device meets the necessary requirements. This chapter focuses primarily on legal requirements for medical devices manufactured, distributed, and sold in the United States, but some of the similarities and differences between the FDA’s and EU’s regulations for medical device labeling are noted at various points. In addition to legal requirements, there are also voluntary standards relevant to various aspects of medical device labeling, including those established by the International Organization for Standardization (ISO), the International Electrotechnical Commission (IEC), and the American National Standards Institute (ANSI), particularly the ANSI Z-535 standards relating to the design of product safety warnings. In 1992, a voluntary group of representatives from national medical device regulatory authorities throughout the world formed the Global Harmonization Task Force (GHTF; see http://www.ghtf.org). Since its inception, the GHTF has worked to enhance the safety, effectiveness, performance, and quality of medical devices; to promote technological innovation; to achieve congruence in regulatory practices; and to facilitate international trade. The GHTF also serves as an information source that countries with medical device regulatory systems under development can use to guide their efforts. The GHTF’s principles for labeling include informing the user of the following: • A device’s identity and intended use • Instructions for use, maintenance, and storage • Risks and warnings In addition, the GHTF has a goal of promoting symbols as a way to communicate information to international audiences appropriate to users’ technical knowledge, experience, education, and training.
FIGURE 13.1 Products that meet requirements established by the EEA receive a stamp of approval, termed the “CE mark,” that allows them to be sold and marketed in EU countries.
TAF-ER562X_WINGR-09-0401-C013.indd 546
11/9/10 3:45:59 AM
Signs, Symbols, and Markings
547
13.2 GENERAL LABELING REQUIREMENTS The guidelines below will help medical device manufacturers follow various regulations and industry standards that govern medical device labeling. The categories for these requirements include the following: • Content of labeling, such as identity statements, instructions of use, and model numbers • Form of this labeling, including completeness, language, and conspicuity • Location and size of the labels on medical devices Additional labeling requirements apply to nonprescription, also called over-the-counter (OTC), medical devices. Given the rapid proliferation of both prescription and nonprescription medical devices for use at home by laypersons, designers should consider the abilities and limitations of users with respect to label design.
13.2.1 LABEL CONTENT The following guidelines pertain to the design of medical device labels. 13.2.1.1 Identity Statement Guideline 13.1: Identity of Manufacturer Both the EEA and the FDA require that medical device labels contain the name, or trade name, of the manufacturer, packer, or distributor, including the complete address (21 CFR 801.1) (see Figure 13.2). The FDA allows that if the name and address of the manufacturer are available in a local phone directory, the address may be omitted from the label.
Guideline 13.2: Identity of Other Entity When a device is not manufactured by the entity whose name appears on the label, the name must be modified by a phrase that reveals the connection the entity has with the device. For example, if one company manufactures the device for another company, the label might read, “Manufactured for X company by Y company.”
Guideline 13.3: Identity of Importer In addition to having information that identifies the manufacturer of the device, the labeling of a device marketed in the EU must include the name and address of the person or party responsible for importing the device into the EU.
Guideline 13.4: Catalog or Model Number Medical devices must contain a label indicating the device’s distinctive catalog or model number.
Guideline 13.5: Electrical Rating Labeling must include the device’s electrical rating.
For nonprescription (OTC) devices, FDA- and CE-marked devices must contain identity statements that include both what the device is and what it does (refer to Figure 13.3). The
TAF-ER562X_WINGR-09-0401-C013.indd 547
11/9/10 3:45:59 AM
548
Handbook of Human Factors in Medical Device Design
FIGURE 13.2 Both the EEA and the FDA require that medical device labels contain certain information on medical devices, including the name of their manufacturer, including their complete address, and the device’s distinctive catalog or model number and electrical rating.
FIGURE 13.3 For nonprescription (OTC) devices, FDA- and CE-marked products must contain identity statements that include what the product is and what the product does.
TAF-ER562X_WINGR-09-0401-C013.indd 548
11/9/10 3:45:59 AM
Signs, Symbols, and Markings
549
requirements of a CE-marked device are very similar to those of devices regulated by the FDA (Council Directive 93/42/EEC concerning Medical Devices, 1993). Additional details concerning specifics on nonprescription (OTC) medical device labeling can be found in U.S. 221 CFR 801. Guideline 13.6: Identity Labeling on nonprescription (OTC) medical devices and their packaging must include a statement that provides the identity of the device (i.e., what it is). The identity statement should be printed in boldface type and should be in text reasonably similar in size in relation to the most prominent text on the panel.
Guideline 13.7: OTC Label Indicates Use The next part of the identity statement should tell users what the nonprescription (OTC) device does (21 CFR 801.61).
Guideline 13.8: Instructions for Multiple Uses If a nonprescription (OTC) medical device has multiple uses, instructions for each of the uses must be included with the device.
Guideline 13.9: Device Serial Number CE-marked nonprescription (OTC) medical devices must be labeled with either the serial number or the word “lot” followed by a batch code of its manufacture. The FDA requires the presence of a control number on each unit, lot, or batch of devices if the device will be used to sustain life or is intended for surgical implantation.
Guideline 13.10: Custom-Made OTC Devices Custom-made nonprescription (OTC) medical devices must be marked with the phrase “custom-made device.”
Guideline 13.11: Quantity of Package Contents The CFR also requires a declaration of the net quantity of contents of the package by weight (pounds and ounces), numerical count, measure (size), or a combination of the three on labeling of nonprescription (OTC) devices in package form (see 21 CFR 801.62). Metric equivalents should be provided as applicable.
13.2.1.2 Use Statement While all medical devices should be accompanied by a use statement, the elements of which are described below, this information does not need to appear in on-device labels. Such information more often appears on device packaging or in the instructions for use. Guideline 13.12: Provide Instructions for Use Manufacturers are required to provide instructions that allow the device to be used safely for its intended purpose by its intended users (21 CFR 801.4-5).
TAF-ER562X_WINGR-09-0401-C013.indd 549
11/9/10 3:46:00 AM
550
Handbook of Human Factors in Medical Device Design
Guideline 13.13: Foreseeable Alternative Uses If there is a reasonable probability that a particular medical device will be used for purposes other than those originally intended, then the manufacturer must provide labeling discussing these alternative uses.
Guideline 13.14: Complete List of Uses and Use Conditions For FDA approval, the directions for use must state all of the device’s intended uses as well as all of the conditions under which the device should be used. For devices produced or marketed in the EU, the GHTF recommends that the instructions be sufficiently thorough to allow consumers to use the device safely.
Guideline 13.15: Duration of Use and Manufactured Date Duration of use should be specified. Devices with a CE mark must include label information that indicates the date after which the device should no longer be used. If an expiration date does not apply to a particular device, then the year the device was manufactured must be provided on the label. FDA-recognized symbols for date of manufacture and expiration date are shown in Figure 13.4.
Guideline 13.16: Timing of Use If appropriate to device use, the time of administration in relation to other factors (e.g., a meal, another treatment) must be indicated.
Guideline 13.17: Methods of Use The instructions should include the route and/or method of application and any other preparations that are necessary before the device can be used. Labels on devices marketed in the EEC must also state any special handling instructions.
Guideline 13.18: Frequency of Use For certain devices, instructions must include the dosing schedule for each use (e.g., “apply to affected area twice per day”), including the usual quantity and frequency for people of different ages and different physical states.
YYYY-MM
Use by YYYY-MM-DD or YYYY-MM
Date of manufacture
Expiration date
FIGURE 13.4 The symbol on the left is used by the ISO to depict “date of manufacture,” while the one on the right is used to list a device’s expiration date.
TAF-ER562X_WINGR-09-0401-C013.indd 550
11/9/10 3:46:00 AM
Signs, Symbols, and Markings
551
13.2.1.3 Handling Statement Guideline 13.19: Storage Devices that have a CE mark must have a label that indicates any special storage instructions (see Council Directive 93/42/EEC concerning Medical Devices, 1993).
Guideline 13.20: Sterilization Instruction Labeling Medical device labels in the EU must include the appropriate method of sterilization, if necessary. Nonprescription (OTC) medical devices that must remain sterile to be used for their intended purpose must include information specifying the part of the device that must remain sterile prior to use (21 CFR 801.10). The EU counterpart for medical devices that must be used under sterile conditions requires that labeling include the word “sterile” and the appropriate method of sterilization directly on the CE mark label (see Figure 13.5).
13.2.1.4 Warning/Precaution Statements Guideline 13.21: Warnings for Safe Use Warnings or precautions required to ensure safe use must be included on a device or its accompanying labeling (Council Directive 93/42/EEC concerning Medical Devices, 1993; 21 CFR 801.5).
Guideline 13.22: Foreseeable Risks The GHTF recommends that manufacturers warn intended users of any risks that are reasonably foreseeable. For example, labeling would indicate that radiation may be emitted from a device or that there is the potential for electromagnetic interference from other equipment.
It is noteworthy that the EEA and the FDA have adopted similar positions concerning the use of warnings on nonprescription (OTC) medical device labeling.
Sterile
Sterilized using steam or dry heat
Sterile
A
Sterilized using aseptic processing technique
Sterile
R
Sterilized using irradiation
Do not re-sterilize
Sterilize
FIGURE 13.5 These labels show examples of FDA-recognized symbols used to specify the medical devices that must remain sterile for their intended use as well as the appropriate methods of sterilization, such as steam or dry heat, aseptic processing, radiation, or a chemical process.
TAF-ER562X_WINGR-09-0401-C013.indd 551
11/9/10 3:46:00 AM
552
Handbook of Human Factors in Medical Device Design
FIGURE 13.6 Clear labeling with text about specific hazards can help reduce the likelihood of user error and injury.
13.2.1.5 Specific Hazards Statements Guideline 13.23: Specific Hazards According to the FDA (2000), device labeling must identify specific hazards associated with OTC medical devices and warn users of these hazards (refer to Figure 13.6). Additional considerations for dealing with specific hazards are described below.
Guideline 13.24: Additional Hazard Information There is no prohibition against providing additional hazard-related information and warnings on labels when it is warranted. The information provided will depend on the particular medical device and the nature of the hazard. Optimal label content would be revealed by analytic methods described later in this chapter.
Guideline 13.25: Warnings for Specific Risks, Including Latex The FDA requires specific warnings for some risks. For example, labeling on devices that contain natural rubber latex must contain one of four variants of the following statement (in bold print): “Caution: This Product Contains Natural Rubber Latex Which May Cause Allergic Reactions” (see 21 CFR 801.437 for all four statements). This labeling requirement derives from reports documenting instances in which people have had serious allergic reactions from natural latex proteins in a wide range of medical devices (e.g., Brehler and Kutting, 2001; Dyck, 2000; Zak, Kaste, and Schwarzenberger, 2000).
The importance of specific warnings is further illustrated in injuries from magnetic resonance imaging (MRI) systems, which operate using an intense static magnetic field to generate images. The machine’s magnetic field strength, about 100,000 times that of the earth’s, has a serious downside—it can pull metal items into its interior. In one instance, a hospital reported expensive damages when a floor buffer left nearby by the janitorial staff was drawn into the machine. Patients can be seriously injured if they happen to be in the path of metal objects, such as scissors and IV poles, under the influence of the magnetic field. Patients have been killed after being struck by oxygen tanks that were magnetically pulled into the opening of MRI machines. In addition, the presence of metallic or magnetized items can adversely affect the proper function of the scanner, resulting in poor-quality images. Warning labels need to be placed not only on the machine itself but also elsewhere in the use environment to warn users and other individuals of the risks before they get too close to the magnet.
TAF-ER562X_WINGR-09-0401-C013.indd 552
11/9/10 3:46:00 AM
Signs, Symbols, and Markings
553
13.2.1.6 Investigational Use Statement Guideline 13.26: Indication of Investigational Situations Device Labeling for medical devices that are only to be used in investigational situations must contain the statement: “CAUTION: Investigational device. Limited by Federal (or United States) law to investigational use” (see 21 CFR 812.5). In the EEA, devices made for investigational use must include the phrase “exclusively for clinical investigations” on the label to receive the CE mark (Council Directive 93/42/EEC concerning Medical Devices, 1993).
13.2.2 LABEL FORM Guideline 13.27: Accessibility of Label Information Label information should be readily accessible but should not interfere with use of the device. This information can be included in the context of labeling claims, advertising, or separate written statements.
Guideline 13.28: Avoid Misleading Statements Labeling should be worded carefully to minimize the risk of misunderstandings that could lead to misuse of a device and/or injury (refer to 21 CFR 801.6).
Guideline 13.29: Primary User Languages Labels should be in the primary language of the intended user population (see Figure 13.7). If the medical device is intended for sole use in places where the predominant language is
FIGURE 13.7 When it can be reasonably anticipated that non-English-speaking people will also be exposed to a hazard, text should be duplicated in the secondary language. Symbols and pictorials can also help ensure that people will understand the intended message.
TAF-ER562X_WINGR-09-0401-C013.indd 553
11/9/10 3:46:00 AM
554
Handbook of Human Factors in Medical Device Design
not English (e.g., Spanish), designers may use this language on labeling (see 21 CFR 801.16). Manufacturers, distributors, or other entities are allowed to substitute the predominant language for English when developing the device’s labeling.
13.2.3 LOCATION AND SIZE 13.2.3.1 Prominence The FDA has established requirements that address the issue of prominence of labels on medical devices (21 CFR 801.15). These regulations describe the size and location of labels both on the medical devices themselves and in device packaging. Guideline 13.30: Conspicuity of Labels Labels on medical devices must be prominently displayed so that users are likely to notice and read them (21 CFR 801.15). Factors that tend to increase the likelihood that users will notice and read labeling include print size, color, contrast, and sufficient label space. These and other factors are discussed in greater detail in the third section of this chapter.
Guideline 13.31: Avoid Mixing Essential and Unessential Information Designers should avoid embedding required information within other less important textual or graphic materials. For example, required information should not be placed within a label containing marketing-type information.
13.2.3.2 Principal Display Panel All mandatory label information on an OTC medical device and its packaging must be placed on the principal display panel (CFR 801.60). The term “principal display panel” refers to the part of a device or its packaging that is most likely to be seen by consumers when the device is displayed for retail sale. Frequently, this is the largest available surface area. Regardless of shape (e.g., rectangular, cylindrical, other shapes), the principal display panel must be large enough to accommodate all the required label information without reducing conspicuity or legibility of the information contained therein. Designers should consult 21 CFR 801.60 for specific guidance on how to determine the location and area of the principal display panel. Designers should also follow this guideline for nonprescription (OTC) medical device labels. Guideline 13.32: Label Text Orientation In general, label text should be located parallel to the base of the package or to the base of the device during normal use.
13.3 LABELS FOR DEVICE IDENTIFICATION, INSTRUCTIONS, AND HAZARDS Labels and markings are used on a wide variety of medical device components, including controls, keyboards, keypads, legend switches, displays, and access openings. In this section, the rationale for and general characteristics of such labels and markings are described. Labels and markings on medical devices are frequently used to identify, locate,
TAF-ER562X_WINGR-09-0401-C013.indd 554
11/9/10 3:46:01 AM
Signs, Symbols, and Markings
555
and functionally group user display and control components. Markings are also important to users when reading mechanical displays and adjusting mechanical controls. Legends list and/or explain symbols included in labeling. Labels for equipment components need to include certain information in particular formats. The sections that follow describe the design attributes of labels, including markings and legends. Guideline 13.33: Label Shouldn’t Impede Device Use Labels should facilitate device use. Label size, location, and design should not interfere with device use.
Guideline 13.34: Identification Statement Medical devices should provide identification information that is readily accessible (refer to Figure 13.2). The identification statement should include the following: • Name of manufacturer • Catalog or model number • Electrical rating (if electromechanical device)
Guideline 13.35: Label Legibility Labels should be legible to users under expected use conditions.
Guideline 13.36: Label Illumination Generally, except for devices used routinely under low-ambient-light conditions, labels should be legible without internal illumination.
Guideline 13.37: Hazard Warning Statement Users and maintenance personnel should be warned of hazards that could be encountered during the use, handling, storage, maintenance, or repair of the device. Examples of key-words for hazard statements are “fire,” “radiation explosion,” “shock,” and “infection.”
Guideline 13.38: Statement of Flammability Electrical medical devices should be labeled to indicate whether they should or should not be used in the presence of flammable substances or oxygen-rich atmospheres.
Guideline 13.39: Label Formatting Labels should communicate effectively and quickly.
Guidance on label formatting (e.g., Lehto and Miller, 1986; Wogalter, DeJoy, and Laughery, 1999; Wogalter and Vigilante, 2006) can generally be obtained by considering the following: • Existing law/regulations concerning labels, which have recently been more specific about label format. • Voluntary standards, such as the most current version of ANSI Z535.4, titled “American National Standard for Product Safety Signs and Labels,” which has reasonably good label format specifications.
TAF-ER562X_WINGR-09-0401-C013.indd 555
11/9/10 3:46:01 AM
556
Handbook of Human Factors in Medical Device Design
• Many formatting factors can enhance the effectiveness of labeling, including size, list format, “chunking” through bulleted points, and white space. More information can be found in the literature (e.g., Laughery, Wogalter, and Young, 1994; Wogalter, Young, and Laughery, 2001).
13.3.1 LABELS FOR ELECTROMECHANICAL COMPONENTS The specific labeling requirements for electromechanical components of medical devices are derived from standards developed by the IEC. The IEC is an international organization that promotes international standardization in electronics. IEC requirements are widely recognized throughout the world. IEC 60601-1 (XXXX) addresses the general requirements for electromedical devices (see IEC 60601-1, subclause 2.2.15). Examples of devices fitting the definition of electromechanical devices include battery-operated thermometers, MRI and gamma imaging systems, endoscopic cameras, and infusion pumps. Accessories to this equipment can also fall under this standard. 13.3.1.1 Electrical Receptacle and Connector Labels Guideline 13.40: Functionality Receptacles and connectors should be labeled with their intended function or connecting cable.
Warning labels that identify specific hazards are particularly important when similar design and compatible receptacles and connectors offer the possibility of misconnecting them. For example, deaths have occurred from the incorrect connection of portable blood pressure monitors to patient IV lines. Manufacturers have issued warning letters describing the hazard, but such after-market notifications are known to be a weak risk mitigation strategy. When possible, the use of specific-shaped connectors that cannot be misconnected is advisable (see Figure 13.8). Guideline 13.41: Electrical Load Information Convenience receptacles should be labeled with their maximum allowable load presented in amperes or watts, as shown in Figure 13.9 (see IEC 60601-1).
FIGURE 13.8 (See color insert following page 564.) The use of specific-shape connectors and color coding can help reduce user error.
TAF-ER562X_WINGR-09-0401-C013.indd 556
11/9/10 3:46:01 AM
Signs, Symbols, and Markings
557
FIGURE 13.9 Labels listing voltage load help decrease the likelihood of equipment damage or injury to the people who use or come into contact with medical devices.
Guideline 13.42: Color Coding of Connectors Distinguishing connectors and having connectors, colors match their respective receptacles can help to avoid misconnections (see Figure 13.10). See Chapter 9, Connections and Connectors.
13.3.1.2 Fuse and Circuit-Breaker Labels Guideline 13.43: Fuse Rating The type and current amperage rating of fuses accessible from the outside of the equipment should be permanently marked adjacent to the fuse holder. Fuse ratings should be indicated either in whole number, common fractions, or whole number plus common fractions (see IEC 60601-1).
Guideline 13.44: Spare Fuse Holders The term “SPARE,” printed in all uppercase letters, should be marked adjacent to each spare fuse holder (see IEC 60601-1).
Guideline 13.45: Fuse/Circuit Breaker Legibility Labeling of fuses and circuit breakers should be legible under ambient lighting conditions expected in the likely environments in which these devices will be used (see IEC 60601-1).
13.3.2 SECTION SUMMARY There are numerous legal requirements that manufacturers and other entities must meet with respect to the design of medical device labeling. Manufacturers, distributors, and other
FIGURE 13.10 (See color insert following page 564.) Color coding can be a useful technique for improving labeling on (a) receptacles and (b) connectors.
TAF-ER562X_WINGR-09-0401-C013.indd 557
11/9/10 3:46:02 AM
558
Handbook of Human Factors in Medical Device Design
persons involved with the design or implementation of medical device labeling should consult these requirements directly for complete, up-to-date regulations and standards. The designer may also need to seek advice from experts who know the current state of the art and science of labeling design. The FDA, EEA, and other organizations set regulations and standards that shape medical device label design and encourage manufacturers to make labels that are conspicuous, legible, understandable, and durable, among other aspects. Labels meeting the regulations and standards, however, do not ensure that the labels are, in fact, effective. Given the importance of enabling the proper safe use of the device, human factors engineering must be incorporated into the label design process. The effectiveness of label design can be assessed using human factors techniques, including particularly usability testing with representative users under realistic use situations. The next section is an overview of human factors principles that are applicable to the design of medical device labeling.
13.4 HUMAN FACTORS PRINCIPLES FOR DESIGNING MEDICAL DEVICE LABELS There is a considerable body of human factors research on warnings that has been conducted across various products and domains, yielding principles that can be generalized to labeling practices. Nevertheless, the most effective labeling for a particular device is likely to be different in form from one device to another. This chapter concerns labels across all medical devices without a focus on a particular medical device or even a class of devices. The general principles provided should be considered during label design even though all of them may not be applicable to a specific device. The right mix of design attributes is dependent on the device and other factors, and it is beyond the scope of this chapter to particularize the factors for a given medical device. Because of the substantial differences in users and settings, a device that may be useful in one situation (e.g., by physicians in a clinic) may introduce considerable risk during use in another situation (e.g., by lay users in their homes). Indeed, even experienced, highly trained individuals can make mistakes under time-constrained, heavy-mental-workload conditions (e.g., Weinger, Slagle, Kim, and Gonzales, 2001; Weinger et al., 1994). Under some emergency conditions, medical device users may have little or no time to refer to device labels, instructions, or warnings. Even under the best of conditions, users may extract little or nothing from poorly designed labels. Human factors research provides not only a database of information on label design but also methods that can be used to test label efficacy.
13.4.1 SAFETY-RELATED GOALS OF LABELS When designing device labels that have safety implications, the principles associated with warning label design apply even if the warning is not explicit. There are three goals of medical device labels with respect to safety: informing, changing behavior, and reminding. Guideline 13.46: Inform Users Labeling should inform users about the consequences (e.g., potential hazards) of the use of a medical device or its applicable component(s).
TAF-ER562X_WINGR-09-0401-C013.indd 558
11/9/10 3:46:03 AM
Signs, Symbols, and Markings
559
Guideline 13.47: Affect User Behavior Labeling should encourage or promote appropriate user behavior. For example, users should be discouraged from performing unsafe acts that they might otherwise do without the benefit of exposure to the device’s labeling.
Guideline 13.48: Remind Users On-device labels should serve to remind users about conditions or hazards even if they may already know some or all about. Thus, one of the functions of labeling is to cue recall of pertinent information.
13.4.2 HAZARD CONTROL HIERARCHY Although this chapter focuses on device labels as an aid for proper device use and injury prevention, it is important to note that engineering design solutions are usually preferred over warning labels to guide proper use and reduce hazards. Device hazards need to be discovered and, to some extent, managed by manufacturers. Manufacturers need to conduct a systematic use hazard analysis to discover the hazards of the device. Once the use hazards are discovered and analyzed, manufacturers should reduce or eliminate them when possible and practical. The basic hazard-control hierarchy (Sanders and McCormick, 1993) offers a useful framework to guide decisions concerning limiting potential injury from use and foreseeable misuse. The levels of the hierarchy are presented below in order of priority based on their likely effectiveness in preventing user injury: 1. Design to remove the hazard. The best method of hazard control is to remove the hazard. If the hazard is eliminated, then the likelihood of injury is greatly reduced. But hazards cannot always be eliminated by design and still yield a functional, usable device. For example, one cannot eliminate all the hazards associated with the use of electricity or radiation in medical devices that require energy sources. 2. Design to guard against contact with the hazard. For hazards that cannot be eliminated, the next best hazard control strategy is to guard against contact with the hazard. An example of built-in guarding is the “dead-man” switch that shuts off the power when a portable fluoroscope handle is released. 3. Ensure prior training and/or experience. This is a form of process guarding. For example, users may be required to train as or work with experts before they can use a device. Alternatively, users may need to obtain a prescription for certain medical devices. Because it depends to a greater extent on the user, this is a less effective form of risk mitigation. 4. Use warning labeling. Not all hazards can be eliminated or guarded against. In such cases, warnings in the device’s labeling are necessary. However, this is the weakest form of risk mitigation, and success requires careful label design and use testing. Thus, good device design procedures attempt initially to design out or eliminate hazards. It is far better to design a control switch so as to reduce the likelihood of inadvertent
TAF-ER562X_WINGR-09-0401-C013.indd 559
11/9/10 3:46:03 AM
560
Handbook of Human Factors in Medical Device Design
activation than to give easy access to a control with a warning label being the only method to prevent inappropriate use. Labels should be considered as a supplement to good design as opposed to a substitute for proper design (Lehto and Salvendy, 1995). Moreover, training can vary in quantity and quality, and labeling can aid or supplement training.
13.4.3 MEDICAL DEVICE LIABILITY AND WARNING LABELS Previously, manufacturers could reasonably assume that the use of most complex medical devices would be restricted to highly trained health care personnel. Until recently in the United States, if manufacturers provided adequate warning to qualified health care professionals, they were shielded from liability. This is called the “learned intermediary doctrine” (LID). LID is based on the notion that the well-trained prescriber is in the best position to communicate all the relevant warning information to end users (Sterling Drug v. Cornish, 370 F.2d 82, 85, 8th Cir. 1966). However, a growing body of case law has weakened this legal protection. Increasingly, manufacturers of poorly designed medical devices have been found responsible in liability cases if their devices were shown to have caused patient or user injury. Now that more devices are being used at home, frequently with little or no health care practitioner involvement, whatever liability protection is still afforded by the LID is being further weakened. In the legal arena, defective device design can include defective labeling. More specifically, if instructions and warnings are necessary to operate the device properly and safely and if that information is inadequate, then the device can be rendered defective.
13.4.4 COMMUNICATION-HUMAN INFORMATION PROCESSING (C-HIP) MODEL The C-HIP model (Wogalter, DeJoy, et al., 1999) combines elements of two simple models from communication theory and human information processing to describe warning and other related processing. In the basic model, people’s mental activities are described as a sequence of stages that begin with a source of that information that uses one or more channels to convey the information to a receiver. The receiver must then notice and attend to the information. The attended-to information must be understood, and for it to be believed, it must be consistent with the person’s belief system so as to motivate (energize) behavioral compliance. Usually the goal of a warning label is to produce behavioral compliance to the directive (although sometimes the goal of a warning is to convey information or remind the user of existing information). Newer conceptions of the C-HIP model are provided in Wogalter (2006) and include the aspects of other environmental stimuli, receiver characteristics, and delivery. At each stage of the model, information may be processed by “flowing through” to the next stage, or it can produce a stoppage or bottleneck to information flow before the process yields behavioral compliance. Depending on the circumstances, processing might not attain the goal of behavioral compliance, but the labels might still be somewhat effective in the role of providing understandable information and reminding the user about a previously known hazard. For example, information can positively influence comprehension about the hazard but still be discrepant with the person’s beliefs and attitudes. If so, this could block any effect on motivation and behavior; that is, the individual might disregard the warning and not comply. While a warning could produce better understanding and lead to
TAF-ER562X_WINGR-09-0401-C013.indd 560
11/9/10 3:46:03 AM
Signs, Symbols, and Markings
561
somewhat more informed decision making, it may be considered ineffective according to a strict behavioral criterion in that it does not necessarily produce the desired safe behavior. The C-HIP model not only decomposes processing into basic stages to better understand the process but can also assist in understanding why a warning might not be effective. Suppose that a warning label is not meeting the goal of high levels of behavioral compliance. One possible solution might be to increase the size of the warning label so that more people will likely see it. But noticing the warning might not be the problem. User testing might reveal that most users see, read, and understand the warning and believe the message but are still not complying with the directed behavior. According to the C-HIP model, the problem then is likely to be at the motivation stage. Users may not be complying because it is difficult to carry out the directed behavior (e.g., because of time, effort, money, or physical disability), or the warning does not adequately indicate the severity of the consequences. In these cases, a more explicit description of the consequences and a way to facilitate performance of the behavior should be considered. Thus, the C-HIP model can be used to determine the specific causes of failure, thereby redirecting limited resources toward correcting the critical aspects of the label’s design. Processing of a label may be nonlinear. The most current version of the C-HIP model contains feedback loops, as illustrated in Figure 13.11 (see Wogalter, 2006). As a result of repeated exposures, users could become habituated to a label. As a consequence, they will be less likely to attend to it on subsequent occasions. Here, memory, as part of the comprehension stage, affects the attention stage. Another example of how a later stage of processing can affect initial label perception is that some people might not believe that a medical device is hazardous. A third example is that the person may not understand the information contained in the labeling the first time they read it. As a result, they may return to an earlier stage (attention) and read it again. Source
Channel
Re ceiver
Attention switch and maintenance
Comprehension
Attitudes and Beliefs
Motivation
Behavior
FIGURE 13.11 The C-HIP model. (Adapted from Wogalter, M.S., DeJoy, D.M., and Laughery, K.R., Warnings and Risk Communication, Taylor & Francis, London, 1999.)
TAF-ER562X_WINGR-09-0401-C013.indd 561
11/9/10 3:46:03 AM
562
Handbook of Human Factors in Medical Device Design
13.5 COMPONENTS OF THE C-HIP MODEL In the following sections, a short description of the each of the main stages of the C-HIP model are described.
13.5.1 SOURCE The source is the originator or initial transmitter of the risk information and can be a person(s) or an organization (e.g., the manufacturer, the FDA). Research shows that the source of a warning can affect people’s perceptions about the material presented. Information perceived coming from a reliable, expert source (e.g., the Surgeon General, the FDA) adds credibility to the message being considered (Cohen, Cohen, Mendat and Wogalter, 2006; Cox and Wogalter, 2006; Morris and Mazis, 1999; Wogalter, Kalsher, and Rashid, 1999).
13.5.2 CHANNEL The channel is the way in which information is transmitted from the source to one or more users. There are two basic dimensions of the channel. The fi rst dimension is the media in which the information is embedded. Device use information, including warnings, can be presented in various ways, such as on-device labels, supplemental labeling information (e.g., operator manuals), live and audiovisual training, Web sites, and so on. The second dimension is the sensory modality. Product safety information is most commonly presented visually (e.g., device use instructions and warnings, pictorials and symbols) or auditorily (e.g., alert/alarm tones, voice). Potentially valuable warning information could also be presented via tactual (e.g., vibration) and olfactory (e.g., odor) modalities.
13.5.3 DELIVERY This delivery process considers the interface between channel and the receiver. Safety information that never actually reaches the user has no practical utility (Wogalter, 2006). It needs to be “delivered” to the receiver to have any chance of being perceived. A label that cannot be seen at the device user’s position will obviously be less influential compared to another that the user can see. Although location and placement are key aspects of the delivery process, label designers need to also consider users’ knowledge, abilities, and skills. For example, some users will be novices to the task or may have perceptual difficulties (e.g., sound insensitivity, color blindness). Also, some users may not receive device use training, may not seek out safety information beyond the device itself (i.e., on-device label), or may not have access to the user’s manual (Wogalter, Vigilante, and Baneth, 1998). Thus, particularly for lay use devices, on-device labels should be well designed and be effective without need for supplemental materials.
13.5.4 RECEIVER The next main aspect of the model, the receiver, encompasses a set of human information processing stages. First, the user must attend to the label content for sufficient duration to ensure that the information is perceived. In subsequent stages, the label must be easily
TAF-ER562X_WINGR-09-0401-C013.indd 562
11/9/10 3:46:03 AM
Signs, Symbols, and Markings
563
understood and fit with the user’s existing beliefs and attitudes. If it does not fit, then the information must persuade the user to change his or her beliefs and attitudes. Finally, the information must motivate the user to perform the desired actions (or avoid unsafe actions). The following sections describe stages within the receiver portion of the C-HIP model. 13.5.4.1 Attract Attention Generally, use of medical devices occurs in environments that have many stimuli competing for people’s attention. The label must initially attract attention. The more noticeable the labeling, the more likely attention will be switched to it. Several characteristics affect noticeability. “Salience” is a broad term that refers to aspects that aid in making the label more conspicuous or prominent. Guideline 13.49: Salience Labeling should stand out from background and other competing information.
Some users will not be actively seeking information about potential hazards because they are focused on the tasks they are trying to accomplish. New users may take longer to extract the relevant information from labeling than experienced users. This is also true of users who are under stressful, distracting, or high-workload conditions. Some features that increase salience include the following (see also Figure 13.12): • Large print (Wogalter et al., 1987) • High color contrast and brightness contrast (Sanders and McCormick, 1993) • Use of distinguishable colors (Braun and Silver, 1995; Sanders and McCormick, 1993) • Pictorial symbols (Kalsher, Wogalter, and Racicot, 1996; Young, Wogalter, Laughery, Magurno, and Lovvoll, 1995) • Prominent and appropriate label placement (Wogalter and Young, 1994; Wogalter et al., 1987)
FIGURE 13.12 (See color insert following page 564.) Salience refers to aspects that aid in making the label more conspicuous or prominent. In this example, the labeling is made more conspicuous through the use of color, placement, and coherent information grouping.
TAF-ER562X_WINGR-09-0401-C013.indd 563
11/9/10 3:46:03 AM
564
Handbook of Human Factors in Medical Device Design
13.5.4.2 Hold Attention Labeling that is noticed but fails to maintain attention long enough for its content to be encoded may not be useful (unless it serves as a reminder). People may notice device labels but not stop to examine them. For adequate processing of label content to occur, attention must be maintained on the message for a sufficient duration of time (Wogalter and Leonard, 1999; Wogalter and Vigilante, 2006). Some of the same design features that capture attention (Section 13.5.4.1) can also be used to maintain attention (e.g., Barlow and Wogalter, 1991; Wogalter, Forbes, and Barlow, 1993). Some additional design features to hold attention include the following (see also Figure 13.12): • Aesthetically pleasing; • Easy-to-read print (large enough to be read easily) (Wogalter, Magurno, Dietrich, and Scott, 1999) • White space (Wogalter and Vigilante, 2003) • Coherent information groupings (Hartley, 1994) • Bulleted lists as opposed to long, continuous paragraphs (Desaulniers, 1987; Wogalter and Post, 1989) • Ragged-right justification (i.e., only the left margin justified) Guideline 13.50: Optimal Quantity of Label Information Users are more likely to rapidly acquire the meaning of labeling that is brief rather than lengthy. Labels containing greater amounts of information may, however, be needed for completeness. Because such labels need to be examined for longer periods of time, they should incorporate qualities that both attract and hold attention as well as reduce the effort required to acquire the label’s information content.
13.5.4.3 Label Comprehension Labeling that is attended to and examined may have little or no value if the user does not understand (comprehend) the intended message. Guideline 13.51: Informative Labels The label should give the user an appreciation of hazards and their consequences, provide useful instructions (dos and don’ts), and enable informed judgment.
Guideline 13.52: Explicit Information The information presented should be explicit. The information should be specific rather than general (Laughery and Paige-Smith, 2006; Laughery, Vaubel, Young, Brelsford, and Rowe, 1993). Vague statements can more easily be misinterpreted. For example, the statement “Hazardous to your health” does not provide an appreciation of potential consequences in a situation where the specific hazard is poisonous vapor that, if inhaled, can cause heart failure and brain damage. When possible and practical, labels should explicitly describe the risks, what actions users should take (or avoid taking) to avoid injury, and the consequences of not complying with the recommended behaviors.
Guideline 13.53: Target Lowest-Level Abilities Whether labeling information will be understood depends on characteristics of both the label and the user. To maximize comprehension, label information should be written to take into account the lower-level abilities and skills of the target population.
TAF-ER562X_WINGR-09-0401-C013.indd 564
11/9/10 3:46:04 AM
Signs, Symbols, and Markings
565
Do not design for the average user because this will miss about half the users. User groups include the following: • Professional users. If the labeling is intended only for trained health professionals such as surgeons and operating room nurses, then designers reasonably can assume some level of user knowledge and skill when making decisions concerning labeling content and format. Even so, there is a broad range of training of health care professionals. Assumptions about background knowledge should be verified with a random sample of representative users. • Lay users. If the medical device is being designed for use by laypersons (i.e., a broad spectrum of the general population), designers cannot assume the same level of knowledge and skills as that of trained health care professionals. A much wider range of education, experience, and skills should be expected. Labels should be designed to accommodate persons with a seventh-grade reading ability if practical or possible. • Non-English speakers. Not everyone in the United States using medical devices reads English, and many devices are designed to be marketed outside the country. Potential solutions to the language problem include the use of simple terminology, increasing label size to accommodate translations, and the use of readily understood pictorial symbols (refer to Figure 13.7) • Disabled or impaired users. Particularly for medical devices designed for patients, labels must be designed to accommodate users with cognitive, perceptual, or other impairments (see Chapter 18, “Home Health Care”). Some of these considerations with respect to older adults are described in Mayhorn and Podany (2006). 13.5.4.3.1 Habituation to Label Message Repeated and long-term exposure to device labeling—even if well designed—may produce habituation, diminishing the labels’ ability to attract and hold attention (Wogalter and Laughery, 1996). One way to reduce habituation is through a periodic change of labeling. This may not be possible for a variety of reasons, including regulations that mandate labeling on medical devices be relatively permanent under ordinary conditions of use. Moreover, periodic changes to a medical device user interface could affect use and requires validation testing. 13.5.4.4 Fit with User Beliefs and Attitudes According to the C-HIP model, even if device labeling successfully captures and maintains attention and is understood, it still might fail to elicit the desired safety behavior because of discrepant beliefs or attitudes held by the receiver relative to the label’s message. According to the C-HIP model, labeling will be successfully processed at this stage if the information concurs with the user’s current beliefs and attitudes (see DeJoy, 1999; Riley, 2006). A message that is in accordance with the user’s beliefs/attitudes will tend to activate and reinforce what the user already knows and expects, thereby increasing compliance with label instructions. Conversely, if device information conflicts with the user’s existing beliefs and attitudes, the labeling message may not be processed further, and compliance will likely be decreased (Wogalter and Laughery, 2006). To overcome this, added salience and other changes to the device may be needed.
TAF-ER562X_WINGR-09-0401-C013.indd 565
11/9/10 3:46:04 AM
566
Handbook of Human Factors in Medical Device Design
Two important beliefs and attitudes related to labels are perceived familiarity and perceived hazard. In general, when people believe that they are familiar with a device, task, or environment, they are less likely to look for or read safety-related information (e.g., Godfrey, Allender, Laughery, and Smith, 1983; Wogalter, Brelsford, Desaulniers, and Laughery, 1991; Wright, 1982). Familiarity beliefs are formed from past similar experience where relevant information has been acquired and remembered. Familiarity produces the belief that nearly everything of relevance is already adequately known (Wogalter et al., 1991). A person who is familiar with a device might assume that a similar device operates the same way. If these expectations and reality do not match, use errors can occur. Familiar devices tend to be perceived as less hazardous than less familiar ones. Perceived hazard is also closely associated with beliefs about the expected injury severity level and is less closely tied with injury probability beliefs (Wogalter et al., 1991). People who do not perceive a device as hazardous are less likely to notice or read its label (Wogalter, Brems, and Martin, 1993; Wogalter et al., 1991). Guideline 13.54: Alter User’s Beliefs and Attitudes Labels should be designed to alter user’s existing beliefs and attitudes when they are not concordant with the realities of device use (e.g., actual use hazards). This difficult task is facilitated if the information is presented in a form that will be noticed, read, and understood.
Design elements that facilitate persuasion include the following: • Salience. Labeling that is salient is more likely to capture the attention of a person who is not looking for warnings or other important device information either because of familiarity effects or low hazard perception. Salience may also enhance a user’s belief that the label information is important. • Credible source. A credible source (e.g., expert authority such as the FDA) can favorably affect beliefs concerning label importance and relevance. • Severity of the consequences. Explicit information about the severity of potential consequences increases perceived hazard and intentions to comply. 13.5.4.5 Motivation Once device labeling is noticed, read, and understood and is consistent with a person’s beliefs and attitudes (or brings about a change in discrepant beliefs and attitudes), the next stage of C-HIP is motivation. The label must sufficiently energize the user to carry out the desired behavior. When the label is asking a person to do something that he or she would otherwise not do, a considerable amount of motivation may be needed. Motivation is affected by the relative trade-offs between the competing costs of compliance and noncompliance. The costs of compliance include money, time, and convenience. One way to reduce the costs of compliance is to make the behavior requested in the label easier to perform. For example, if personal protective equipment (PPE) is necessary when using a medical device, the device manufacturer might consider making the PPE more available to users, perhaps by providing the PPE with the device and including a convenient storage place for it in the device. In addition, the costs of noncompliance (i.e., consequences), such as severe injury and monetary loss, should be clearly and conspicuously presented (Wogalter, Allison, and
TAF-ER562X_WINGR-09-0401-C013.indd 566
11/9/10 3:46:04 AM
Signs, Symbols, and Markings
567
McKenna, 1989; Wogalter et al., 1987). This is another reason that giving explicit statements with specific negative outcomes rather than general ones is preferable (Laughery et al., 1993). 13.5.4.6 Label Must Produce Compliance If the user is sufficiently motivated, then he or she is likely to carry out the desired behavior. Behavioral compliance research shows that warnings and other safety-related materials are usually effective if properly designed and implemented (e.g., Cox, Wogalter, Stokes, and Murff, 1997; Kalsher and Williams, 2006; Laughery et al., 1994). 13.5.4.7 Labeling May Influence Users Indirectly Information from labels may reach users indirectly. Indirect methods of communication include user-to-user transmission and changes in use environment culture or norms. An example is an experienced user orally telling a new user pertinent safety information based on information that he or she acquired at an earlier time. That person, in turn, may change his or her behavior accordingly. To the extent that labeling information alters the behavior of a sufficient number of users, the behavior may become a norm of the use environment (i.e., part of “Culture”), thereby propagating and reinforcing the behavior even in the absence of future direct contact with labeling (see also 13.5.5.1).
13.5.5 OTHER HUMAN FACTORS ISSUES Other factors that influence motivation to comply are external to the labeling but nevertheless may affect labeling effectiveness. A medical device manufacturer cannot control the local situation in which a device is used. Thus, it is foreseeable that devices will not always be used under optimal conditions. For example, a device might be used in low lighting. Other factors that affect motivation to comply include social influence (Wogalter et al., 1989), time stress (Wogalter, Magurno, Rashid, and Klein, 1998), and mental workload (Wogalter and Usher, 1999). 13.5.5.1 Social Influence Observation of how other users behave can affect an individual users’ behavior with respect to a device. If people observe others not complying with a label’s directive to wear protective equipment while using a particular medical device and further observe them not being harmed, they may conclude that it is unnecessary to wear protective equipment themselves (Wogalter et al., 1989). By contrast, observing others complying with a label’s directive can have a positive influence. Device manufacturers should try to positively influence use behaviors through effective device design, labeling, and training. 13.5.5.2 Stress and Workload In high-stress and high-workload situations, competing activities limit the cognitive capacity or resources available for processing label information and complying with desired use behavior. Under these conditions, considerable emphases on safety and reduced cost of compliance may be required to overcome the barriers. Efforts that reduce stress and workload should also facilitate compliance.
TAF-ER562X_WINGR-09-0401-C013.indd 567
11/9/10 3:46:04 AM
568
Handbook of Human Factors in Medical Device Design
13.6 USING HUMAN FACTORS PRINCIPLES TO ENHANCE COMPONENTS OF MEDICAL DEVICE LABELING This section provides specific guidance on how to apply human factors principles to the design of medical device labels
13.6.1 LABEL CONTENT One of the most important, if not the most important, aspects of labels is the message content. Clarity, consistency, completeness, and brevity of label information are critical. As simple as these criteria may seem, achieving them requires systematic effort during the development process. Designers should consider the following guidelines when developing label content. Guideline 13.55: Understand Label Users Designers should understand the device’s intended user population. Lay users are likely to have little or no medical or technical education.
Guideline 13.56: Identify Essential Label Information Designers should determine what information is needed on the label. This may entail review of the literature, observation of users, and interviews with experts and representatives of potential users. Prioritize label content on the basis of input from subject domain experts, human factors experts, and especially potential users.
Guideline 13.57: Choose Words Carefully Words should be chosen to express exactly the idea or action intended. The wording should be clear, direct, accurate, complete, and succinct.
Guideline 13.58: Nontechnical content The use of unusual or technical terms should be avoided, particularly for device labels intended for lay users.
Guideline 13.59: Understandable to All Label Users Labels should be understood by those users with the lowest expected level of cognitive abilities. For example, information for lay users should be written at or below the seventh-grade reading comprehension level (age 13) (see 21 CFR 801.5).
Guideline 13.60: Be Consistent with User Expectations Recognized practices, expectations, and conventions of the target users (e.g., laypersons vs. health care workers) should be considered.
Guideline 13.61: Testing of Label Content Test label content using a representative sample of users to ensure that the intended message is being conveyed and to identify any incorrect or misleading information.
TAF-ER562X_WINGR-09-0401-C013.indd 568
11/9/10 3:46:04 AM
Signs, Symbols, and Markings
569
Guideline 13.62: Comprehension of Label Language If it is determined that non-English users will be unable to comprehend the on-device text, then supplemental labeling should be developed that accurately transmits the relevant hazard information to those users (see Figure 13.7). The specific implementation will depend on a number of constraints: • Simplicity of the English text or the number of languages required to meet the needs of expected device users • Availability of acceptably comprehensible symbolic representations (i.e., ones that meet established criteria; see Guideline 13.64) • The amount of space available to present label information in alternative languages and/ or to include pictorials that are language independent • Current regulations and standards
13.6.1.1 Abbreviations, Initials, and Symbols Graphical symbols or text abbreviations may be used on labels, for example, when space is limited or they are expected to be more effective than the represented text. If symbols or abbreviations are used due to space constraints, understandability should be at least equivalent to full text labeling. Guideline 13.63: Sparing Use of Abbreviations and Initials Abbreviations and initials should be designations or names that are well known to the population of intended users. If there is any question, then a formal evaluation with a representative sample of device users should be conducted.
Guideline 13.64: Understandable Symbols Symbols should have a meaning commonly understood by most users. According to safety symbol standards, 85% or more of a representative sample of 50 potential users should understand a symbol’s intended meaning, with less than 5% critical confusions (ANSI Z535.3, 2002). Critical confusions are errors of understanding in which people report the opposite of the intended meaning or answer in a way that is potentially dangerous. These performance criteria and guidelines can be reasonably extended to the assessment of the understandability of textual notations, such as abbreviations and initials. However, if use errors due to symbol confusion can have safety implications, 100% of users tested must not make use errors or other effective risk mitigations must be employed.
Designers should also familiarize themselves with other organizations that offer guidance on symbol selection for medical device labels, such as the IEC and the ISO. Some of the more relevant specific standards on the use of symbols include EN 1041, EN 1658, ISO 780, ISO 7000, ISO/TR 15223, IEC 601-1, IEC 601-2, IEC 60417-1, IEC 60417-2, IEC 878, and EN 980. More than 7,000 symbols are described in these standards, many of which are applicable to medical devices, diagnostic kits, and associated equipment and instrumentation. They also include specifications for position, size, and unit measurements. In the EU, manufacturers are allowed to devise their own symbols and use them on labels or instructions for use if they are fully explained and their safety is evaluated. The ANSI Z535.3 (2002) symbol standard contains appendices on how to develop and test safety symbols. Topics such as iterative design and testing, user feedback, and cost-saving methods in assessing comprehension from concept to symbol are also discussed (see also
TAF-ER562X_WINGR-09-0401-C013.indd 569
11/9/10 3:46:04 AM
570
Handbook of Human Factors in Medical Device Design
Deppa, 2006; Goldsworthy and Kaplan, 2006; Miller and Parent, 2006; Sojourner and Wogalter, 1998; Wogalter, Silver, Leonard, and Zaikina, 2006).
13.6.2 LOCATION AIDS AND FUNCTIONAL RELATIONSHIPS Location aids such as demarcation (e.g., boundaries or borders), color coding, shading, mimics (physical representations of medical device components and their relationship to one another), and flashing lights may be used to indicate the positions of and relationships among related controls and displays. Guideline 13.65: Redundant Location Coding Redundant location aids should be used particularly when a single method of presenting label information cannot be expected to be adequate, such as for devices that are expected to be used under degraded environmental conditions (e.g., low-light or changing light conditions).
Guideline 13.66: Demarcation and Shading Designers should use some form of demarcation or shading to group together related items such as controls and displays as illustrated in Figure 13.13.
Guideline 13.67: Mimics Mimics should be used to enhance users understanding of device function or system relationships. Mimics are displays that help users simultaneously monitor multiple components that compromise a medical device or system (Wiegmann et al., 2002). Mimic displays can help operators detect and diagnose problems as they arise. Mimics differ from other location aids in that they reflect functional and/or spatial relationships among components of the medical device. Mimics integrate representations of displays and controls into a composite graphic or pictorial. Properly designed mimics enhance a user’s ability to identify, monitor, and/or manipulate medical device displays and controls in real time. Additional factors that can further enhance the utility of mimic displays include color contrast and consistency.
Guideline 13.68: Consistent Mimic Colors Lines depicting flow of the same contents (e.g., blood, oxygen) should be colored the same.
Guideline 13.69: Minimize Parallel Mimic Lines Designers should minimize the number of similarly colored lines, termed “sensor” lines, running parallel to one another so that users can quickly identify any one of the lines if needed (see Chapanis and Yoblick, 2001).
FIGURE 13.13
Use of demarcation to group related information displayed on a medical device.
TAF-ER562X_WINGR-09-0401-C013.indd 570
11/9/10 3:46:04 AM
Signs, Symbols, and Markings
571
13.6.3 POSITION AND PLACEMENT One of the most important factors contributing to label effectiveness is the positioning of labels and other markings on medical devices and their packaging. Designers must take into account the aspects of visibility, spatial orientation, proximity, and shape. 13.6.3.1 Visibility Guideline 13.70: Viewing Angles Labels should be positioned to ensure visibility and legibility from expected vertical or horizontal viewing angles as well as from angles above or below eye level.
Guideline 13.71: Flat, Nonglossy Surface The use of a flat, nonglossy surface will prevent veiling glare from obscuring the display.
13.6.3.2 Orientation Improperly oriented labels can lead to confusion and cause delays in locating and identifying important controls and/or displays. Guideline 13.72: User Orientation vs. Label Orientation The orientation of labels should be consistent with the user’s likely physical orientation while operating the device (see Figure 13.14).
FIGURE 13.14 The orientation of labeling should make it easy for users to read and should be consistent with the user’s likely orientation while operating the device.
TAF-ER562X_WINGR-09-0401-C013.indd 571
11/9/10 3:46:05 AM
572
Handbook of Human Factors in Medical Device Design
Guideline 13.73: Horizontal Orientation Orient labels horizontally so that they may be read quickly from left to right. In most languages, including English, people read from top to bottom and left to right. However, a different orientation may be appropriate for users of other languages (e.g., Hebrew, Arabic, Chinese).
13.6.3.3 Shape Guideline 13.74: Shape of Labels and Controls The shape of controls and their labels should strengthen the association between the control and its function (Sanders and McCormick, 1993).
Guideline 13.75: Curved Labels Do not use curved text on labels, except for setting delimiters for rotary controls and displays (see Figure 13.15).
13.6.3.4 Location Guideline 13.76: Proximity of Labels and Controls Labels and other markings should be placed near the controls that they describe—either on the control itself or immediately adjacent to it—so that it is easy for the user to make the desired association between the two objects (see Figure 13.16).
Guideline 13.77: Graphics Adjacent to Text Graphics should be placed adjacent or as close as possible to its associated text so that it is easy to make the connection between the two.
Guideline 13.78: No Obstructions to Label Viewing Labels and other markings should not be blocked from view by hand positions or other equipment components. Labels should remain visible and legible once the user’s hand is placed on the control.
FIGURE 13.15 and display.
Avoid using curved text on labels, except as setting delimiters for rotary controls
TAF-ER562X_WINGR-09-0401-C013.indd 572
11/9/10 3:46:05 AM
Signs, Symbols, and Markings
573
FIGURE 13.16 Example of a device on which the labels and markings are placed on or near the controls they describe.
13.6.4 GESTALT PRINCIPLES A well-established set of perceptual guidelines, termed Gestalt principles, may facilitate design decisions concerning medical device labeling (and device components) (e.g., Wolfe et al., 2006). Gestalt principles describe people’s tendency to see separate, isolated parts as organized wholes (e.g., Baron and Kalsher, 2005). The focal point of Gestalt theory is the idea of “grouping,” or how we tend to interpret visual patterns in certain ways. The main factors that determine grouping are the “laws” of similarity, proximity, and good continuation. A description and representation of each of these principles is provided in Figure 13.17.
The law of similarity states that objects which share visual characteristics such as shape, size, color, texture, value or orientation tend to be perceived as a group or pattern
a
c
d
b
The law of proximity states The law of good that objects near each other continuation states that tend to be seen as a unit. objects arranged in either a straight line or a smooth curve tend to be seen as a unit. People tend to see two lines: one from “a” to “b” and another from “c” to “d”, even though this graphic could represent another set of lines, one from”a” to “d” and another from “c” to “b”.
FIGURE 13.17 continuation.
Examples of Gestalt principles of grouping: similarity, proximity, and good
TAF-ER562X_WINGR-09-0401-C013.indd 573
11/9/10 3:46:05 AM
574
Handbook of Human Factors in Medical Device Design
FIGURE 13.18 Labels should be placed close to the related controls, displays, ports, and so on to which they refer. These photos of medical devices show how the use of Gestalt principles can enhance the effectiveness of medical labeling.
Guideline 13.79: Grouping Labels or Label Elements Related label elements or labels should be grouped together. Labels should be placed with the controls and displays with which they are associated (see Figure 13.18).
Guideline 13.80: Separate Adjacent Device Elements Adjacent controls and displays and their labeling should be separated by sufficient space or other design elements so that they are viewed as separate.
13.6.5 POPULATION STEREOTYPES AND EXPECTATIONS Population stereotypes are social and cultural norms affecting people’s expectations on how a device works. Different parts of the world may have different expectations about how some things work or should work. Furthermore, people are likely to generalize from one device (e.g., light switch) to something similar (e.g., the power switch on a medical device). In North America, most people expect that to turn on the light, one moves the switch to an up position. Thus, the population stereotype for “on” is up and for “off” is down. In Europe, the expected movement to turn on a light switch is opposite (“on” is down). When device function is consistent with the user population stereotypes, use errors decrease. However, stereotypes are based on learning and experience, which may differ to some degree between individuals in a population. In general, label designs that make use of knowledge about people’s tendencies will facilitate performance in terms of faster time and fewer errors.
TAF-ER562X_WINGR-09-0401-C013.indd 574
11/9/10 3:46:06 AM
Signs, Symbols, and Markings
575
Guideline 13.81: Best Labels Will Not Compensate for Poor Interface Design When device design is not compatible with users’ expectations, labeling must, in a sense, work harder to control users’ behavior. However, labeling should not be used as a substitute for good user interface design.
13.6.6 DURABLE MATERIALS Guideline 13.82: Durability Labels should be resistant to wear and tear over the expected life span of the device (Glasscock and Dorris, 2006). To accomplish this goal, designers should take into account the environments in which the device will likely be used, user characteristics, materials, inks, coatings, and so on. Examples of worn labeling on medical devices are presented in Figure 13.19.
Guideline 13.83: Difficult to Detach The label should not be easily removed or be severely abraded when the device is subjected to ordinary wear and use. In most instances, labels should be difficult to detach.
Guideline 13.84: Adhesive Residue from Labels Nonsterilizable labels on devices such as surgical tools should be removable to permit sterilization. In such cases, there should be no adhesive residue.
Guideline 13.85: Replaceable Labels Removable labels must be readily replaceable. In general, the labels should be printed redundantly in other places, such as device manuals or on the Internet. Assigning part numbers to the labels is advisable to facilitate reordering.
FIGURE 13.19
Examples of worn medical device labels.
TAF-ER562X_WINGR-09-0401-C013.indd 575
11/9/10 3:46:06 AM
576
Handbook of Human Factors in Medical Device Design
Guideline 13.86: Device Manufacturer Label A durable label containing manufacturer contact information (along with device identification and serial number) should be placed where it is accessible.
Guideline 13.87: Static Electricity A buildup of static electricity can pose a safety hazard when medical devices are used in certain environments. In these instances, labels should be constructed of polyester, polyamide, or other materials known to reduce the buildup of static electricity.
13.6.7 LEGIBILITY Legibility is the ease with which the details of displayed material can be accurately discriminated. Poor legibility can reduce people’s ability to read and comprehend medical device labels. Labels should be easy to read so that users are able to extract and encode the information; otherwise, attention will not be held, and the user may attend to something else. As described earlier, legibility is part of the attention maintenance or holding stage of the C-HIP model, thus affecting whether a warning is understood, affects beliefs, and so on. Poor legibility may stem from label design, labeling material, choice of font and font size, or other factors, such as aspects of the environment (e.g., illumination levels) or user characteristics (e.g., visual acuity). In particular, older adults with agerelated perceptual or cognitive limitations may not have the ability to read small print in dimmer lighting conditions. Given the increasing percentage of older adults, designers should attempt to design labels that compensate for age-related perceptual and cognitive decrements. Guideline 13.88: Environmental Effects on Label Legibility To attain adequate label legibility, designers should consider the following environmental influences that can degrade legibility (Sawyer, 1993): low levels of ambient light, glare-producing surfaces, damage from heat, use of improper cleaning products, humidity, and moisture.
Guideline 13.89: User Attributes Affecting Label Legibility Designers should consider the following user attributes that could affect legibility of labels (Mayhorn and Podony, 2006; Smith-Jackson, 2006b): • • • • • •
Age Literacy and numeracy Perceptual limitations (e.g., color blindness, visual acuity) Expectations (Vredenburgh and Zackowitz, 2006) Cultural differences (Smith-Jackson, 2006a) Experience and training
Because medical devices are being used increasingly by laypeople in nonmedical settings (i.e., in the home), the labeling intended for such users should take into account the wide range of sensory and cognitive characteristics in the general population. When designing for the general population, designers should anticipate much higher levels of variability on most relevant user characteristics.
TAF-ER562X_WINGR-09-0401-C013.indd 576
11/9/10 3:46:07 AM
Signs, Symbols, and Markings
577
Guideline 13.90: Contextual Factors Affecting Label Legibility Designers should consider a variety of situational or contextual factors that could adversely affect label legibility, including the following: • • • • •
Heavy task load (Wogalter and Usher, 1999) Stress (e.g., Vredenburgh and Helmick-Rich, 2006) Fatigue (e.g. Vredenburgh and Helmick-Rich, 2006) Use of alcohol and other drugs Physical health and illness
Guideline 13.91: Typical Label Reading Distances Relevant text and pictorial components on medical device labels should be legible at expected viewing distances and angles. Advice on size and font characteristics is provided in Section 13.6.7.2.
Guideline 13.92: Use Conditions Affecting Label Legibility Characters and symbols should be designed to be legible by intended users under the full range of expected use conditions. A medical device user’s ability to physically read the warning label is, of course, crucial to compliance. An instrument known as the Lockhart Legibility Instrument has been developed to conduct tests in which the amount of light necessary to read a label is controlled (Bix, Lockhart, Cardoso, and Selke, 2003). The test participant rotates a filter within the instrument until legibility is achieved. A number of factors, including color, font size and other typography, and distance, both alone and in combination, can be varied under different lighting conditions to determine their impact on legibility of labeling.
13.6.7.1 Highlighting and Contrast Legibility can be enhanced through highlighting and contrast (Frascara, 2006; Wogalter and Vigilante, 2006). Guideline 13.93: Highlighting for Labels Designers should use highlighting to call attention to important aspects of medical device operation. The use of highlighting can provide visual relief, emphasize important points, and attract user attention to particularly important sections of text. Highlighting techniques include the use of the following: • • • • • • • • •
Color Bolding Underlining Italics Reverse printing (e.g., white text on black background) Varied font styles Boxing in of text Offsetting borders and backgrounds White space
Guideline 13.94: Consistent Use of Label Highlighting Highlighting techniques should be used consistently throughout all device labels.
TAF-ER562X_WINGR-09-0401-C013.indd 577
11/9/10 3:46:07 AM
578
Handbook of Human Factors in Medical Device Design
Guideline 13.95: Overuse of Label Highlighting Highlighting should not be overused, as its effectiveness will then be diminished. For example, overuse of italics or all uppercase lettering are known to reduce legibility. Another example of poor use of highlighting is the use of a gray background as a highlight for black print, as this reduces contrast (see Guideline 13.97). Designers must carefully choose which important information to highlight; this should be the most important information and information that might otherwise be missed.
Guideline 13.96: Uppercase Lettering Uppercase lettering is recommended for signal words and also may be useful for accentuating the salience of a few words of text. An example is the use of uppercase lettering to distinguish different drug names (e.g., DOBUTamine vs. DOPamine).
Guideline 13.97: Contrast of Label Print Contrast of light/dark or dark/light print is another technique designers can use to enhance legibility (Bix et al., 2003; Sanders and McCormick, 1993). Generally, black and white provides the best contrast, but labeling does not have to be achromatic: Many color combinations may be used as long as there is a large difference in light–dark and color contrast. Designers should generally use dark characters against a light background. The reverse can also be used, but if the labeling is implemented as a projected display (such as LEDs), the potential for “irradiation” should be taken into account. Irradiation is the tendency for white lettering on a black background to “spread out.” Thus, when using white print on a black background, designers should compensate by employing fonts with a thinner stroke width.
Guideline 13.98: Color and Ambient Light Color can be used to differentiate important words or text (e.g., hazard warnings). Choose color combinations that produce adequate contrast. Various color combinations of background to differentiate sections of label text are illustrated in Figure 13.20.
The perception of color can be affected by the label’s materials, reflective gloss, and ambient lighting conditions as well as user characteristics (color blindness and tinted eyewear).
FIGURE 13.20 (See color insert following page 564.) The use of different background colors can help differentiate various portions of labeling.
TAF-ER562X_WINGR-09-0401-C013.indd 578
11/9/10 3:46:07 AM
Signs, Symbols, and Markings
579
The anticipated range of these factors should be considered to determine the colors used and their presentation (e.g., materials, inks, coatings). The impact of anticipated light conditions should also be considered. A related factor with respect to lighting is the angle at which a label is presented relative to the user’s position. Although direct, perpendicular viewing is usually best, some lighting conditions may cause veiling glare, whereby the label information is obscured by reflected light, particularly on high-gloss surfaces (see Figure 13.20). In such a case, adjustment of the viewing angle or a change in labeling materials should be considered. Guideline 13.99: Color Coding of Warnings on Labels When warnings are used, it is generally advisable to use red, orange, and yellow for hazardrelated messages since those colors are associated with hazards as specified in medical device standards, such as IEC60601-1-8 (unless user testing shows other colors to be acceptable).
13.6.7.2 Typography Typography is the arrangement, style, and general appearance of the component alphanumeric material. It encompasses various characteristics of print, including type fonts, type size, and type styles. Typography affects legibility, information transmission, and search (Frascara, 2006; Simpson and Casey, 1988). Several of the most important factors are described below. Guideline 13.100: Type Size for Labels Designers should use a type size large enough for the relevant information to be extracted from the label at eye distances in which the device is being operated by the intended user audience and under the anticipated lighting conditions (Wogalter and Vigilante, 2003). Guidelines such as ANSI Z535.4’s appendix provide minimum suggested types sizes based on distances and good/poor viewing conditions.
Guideline 13.101: Type Font on Labels Many fonts in common use to display text are comparably legible. Fancy fonts like Old English and script should not be used, as they are less familiar and could slow reading speed and, in extreme cases, comprehension. Serif fonts, such as Times Roman, have embellishments on the component parts of the letters that distinguish the letters to a greater extent than sans serif letters (e.g., Helvetica, Arial). Times Roman is one of the most frequently used fonts and thus is highly familiar to most people. Whether to use serif or sans serif fonts can be determined as follows: • Serif fonts. Use serif fonts for smaller-sized print (i.e., 9 to 14 points) because they are easier to read than sans serif fonts. • Sans serif fonts. Use sans serif for larger type, such as signs or posters. Sans serif fonts are generally preferred for electronic labels.
Guideline 13.102: Multiple Fonts on a Level While the use of a different font can highlight particular text passages, generally avoid the use of multiple font types on the same label. Multiple fonts can be distracting, unattractive, and can reduce the speed at which the information is encoded. Also, if the label is not aesthetically pleasing, it could have less attention-holding power.
TAF-ER562X_WINGR-09-0401-C013.indd 579
11/9/10 3:46:08 AM
580
Handbook of Human Factors in Medical Device Design
FIGURE 13.21 Mixed-case letters should, in general, be used in medical labeling, including warnings. The label also shows the uppercase WARNING after the international symbol for alert or warning.
Guideline 13.103: Sentence Case Is Preferred Mixed-case (using both upper- and lowercase) letters should, in general, be used (see Figure 13.21). All uppercase (all capital) letters is a poor choice for users with low visual acuity, under low legibility (e.g., small print), and in glare exposure conditions. Lowercase letters are usually more legible than uppercase letters, as their shapes are more distinguishable, even though they are smaller in size. Uppercase letters have more similar components, making them less distinguishable (Backinger and Kingsley, 1993).
Guideline 13.104: Uppercase for Signal Words Uppercase should be used for signal words, such as “DANGER” and “CAUTION.” Warning standards (e.g., ANSI Z535) specify this as part of a panel that also includes an alert symbol (triangle enclosing an exclamation point) and a corresponding color. Together, these components may form a trigger that facilitates recognition and response (Wickens, Lee, Liu, and Gordon-Becker, 2003). In selective instances, uppercase may be used to highlight important text but not for a large grouping of words (refer to Figure 13.21).
Guideline 13.105: Text Emphasis Italics, underlining, or bolding can be used to highlight important text. Excessive use of any such emphasis coding will diminish its benefit.
Guideline 13.106: Vertical Text Spacing (Leading) The space between lines—known as leading—should be at least 25% to 30% of the text size (e.g., Hartley, 1994; Misanchuk, 1992; Sanders and McCormick, 1993). This distance helps reduce inadvertent switching between lines of text (see Figure 13.22).
Guideline 13.107: Kerning Kerning is adjustment of spacing between letters to make large print appear consistently spaced and to fit the text into relatively short columns (e.g., Hartley, 1994; Misanchuk, 1992; Sanders and McCormick, 1993). Kerning should be adjusted to maximize readability.
TAF-ER562X_WINGR-09-0401-C013.indd 580
11/9/10 3:46:08 AM
Signs, Symbols, and Markings
581
FIGURE 13.22 The leading and spacing of text on this label show an improper proportion, with the type size of some of the message text too small in comparison to the signal word and label size.
Guideline 13.108: Horizontal Spacing Between Letters The distance between letters (i.e., horizontal spacing) should not be so limited that one character appears to touch the sides of the next. Nor should letters be spread so far apart that additional eye movements are required to read the text (Watanabe, 1994) (see Figure 13.22).
13.6.7.3 Other Strategies Usually, space for labeling is limited. Rather than make the print too small to satisfy the completeness criterion or making the print so large that some important information is omitted, designers should consider other strategies to display the information, such as increasing the available surface area and prioritizing components of a label.
Guideline 13.109: Overall Label Size Designers could consider increasing the surface area available for the label (Wogalter and Vigilante, 2003; Wogalter and Young, 1994).
Guideline 13.110: Prioritization of Label Components Prioritization refers to ordering the components of a label with respect to importance. The most important information should be presented first and/or otherwise enhanced by highlighting (e.g., larger size, color). Decisions can be based on judgments of overall importance, severity and probability of injury, and whether the information is already known by users (Vigilante and Wogalter, 1997). When high-profile space is limited, items of the lowest priority may need to relegated to other labeling (e.g., supplemental presentation materials, such as the user manual).
13.6.8 CODING Coding refers to the use of physical attributes within a presentation to signify or designate some association, organization, or meaning. Coding methods include color, shape, graphical elements and location. The purpose of coding is to help users distinguish important characteristic features and identify functionally related and/or critical features. For example, code markings on gauges are placed to convey the desirable operating range, dangerous operating levels, status information, or alarm conditions.
TAF-ER562X_WINGR-09-0401-C013.indd 581
11/9/10 3:46:08 AM
582
Handbook of Human Factors in Medical Device Design
13.6.8.1 Redundant Codes Guideline 13.111: Redundancy of Coding Coding used to convey safety-critical information, actions, or device functions should have redundancy. That is, the same information should be presented via two or more modes to ensure that users receive it.
Several types of coding can be used, including color, size, location, shape, and symbols (see Figure 13.23). Color should not be the sole means for identifying and/or distinguishing critical information elements, nor should it be the primary means of doing so. Redundant coding, especially of critical information, should be provided for several reasons. For example, coding only with color can lead to use error for the following reasons: 1. Lighting can vary and ‘wash out’ some colors (e.g., because of glare), accentuating the need for a “backup” code. 2. Some users are likely to have some form of color blindness. In the United States, approximately 7% of men and 1% of women are red-green color blind and cannot distinguish between these two colors. A smaller percentage of people have a blueyellow color weakness. For this reason, more than one code in addition to color should be used. Both the ISO and ANSI recommend the use of redundant coding of warning information to ensure that the level of hazard is accurately conveyed. ANSI guidelines accomplish this goal by pairing each of three colors with one of three specific signal words in a warning’s header (refer to Figure 13.24). ISO standards for devices marketed in the EEA tend to incorporate shape coding of the external borders along with color in safety symbols (see Figure 13.25a). Recently, there have been efforts to harmonize ANSI and ISO guidelines for warnings (see Figure 13.25b). 13.6.8.2 Color Coding Color coding can be used to enhance the transfer of relevant information to device users by making important labeling information stand out (refer to Figures 13.8, 13.10, and 13.20). Color coding of labels facilitates visual identification and reduces the likelihood that users
FIGURE 13.23 (See color insert following page 564.) Redundant coding can help ensure that users receive the information they need to operate medical devices safely.
TAF-ER562X_WINGR-09-0401-C013.indd 582
11/9/10 3:46:08 AM
Signs, Symbols, and Markings
583
FIGURE 13.24 ANSI Z-535 (2002) recommends the use of three different colors to connote differing levels of hazard.
will inadvertently manipulate the wrong device element. For many people, color is highly effective in drawing attention to and distinguishing important device information (Wiklund and Dolan, 1996). Color can also be used to communicate varying conditions, such as levels of hazard or other quantity, as illustrated in Figure 13.26. Important considerations when using color coding are consistency, color choice, and number of colors. Despite the potential problems of using color as a method of coding, consumers have a strong preference for the use of color in applications. However, some additional considerations are discussed below. Guideline 13.112: Consistency of Color Coding Colors that have a designated meaning or that are expected to elicit a specific user response should be used consistently throughout a device.
Guideline 13.113: Color Coding Consistent with Expectations Color coding should, whenever possible, be consistent with users’ expectations, customs, and prior experience with similar devices.
Guideline 13.114: Most Readily Identified Colors Red, green, yellow, orange, and blue are the most easily identified (named and recognized) colors.
TAF-ER562X_WINGR-09-0401-C013.indd 583
11/9/10 3:46:09 AM
584
Handbook of Human Factors in Medical Device Design
(a) ISO standards for devices sold in the EEA tend to incorporate shape coding along with color in safety symbols.
Pacemaker prohibition–white and Strong magnetic field warning–yellow triangle with black border (ISO).Graphic red circle with red slash (ISO) and black graphic must be black.
Harmonized ISO and ANSI label combining three colors with a specific signal word and ISO standard safety shapes.
(b) Examples of harmonized warnings.
Mandatory action–blue circle with white graphics (ISO) (above) and harmonized ISO and ANSI label (right): indicates an action to take to avoid the hazard.
Label format combining a warning, a prohibition, and mandatory action, incorporating ISO standards shapes with ANSI formatting.
FIGURE 13.25 (a) Examples of ISO labels, advocating more shape coding, and (b) harmonized warnings that incorporate both ISO and ANSI formatting features.
The number of colors used for coding should be kept to the minimum needed to provide sufficiently distinctive information. Unfortunately, existing guidelines are inconsistent regarding the maximum number of colors that should be used. Some guidelines recommend using no more than four colors on a label, whereas others are more lenient. For instance, the National Aeronautics and Space Administration (NASA) allows the use of up to six colors on device labels. NASA also recommends no more than three shades of gray if users must either recall the meaning of each color or make identifications on the basis of color or shade. A general recommendation is to use no more than five colors (refer to
TAF-ER562X_WINGR-09-0401-C013.indd 584
11/9/10 3:46:10 AM
Signs, Symbols, and Markings
585
FIGURE 13.26 Color can be an effective tool for conveying safety critical information such as the identity of breathing gases. Note that in this case, shape and position coding are also used to reduce the risk of use error.
Figures 13.8, 13.10, 13.12, 13.15, 13.20, and 13.23). If color coding is not self-evident, a legend should be provided in the device labeling to assist the user in determining the meaning of each color code. Users are better able to distinguish colors presented concurrently (e.g., on the same fixed label) than those presented sequentially (e.g., on different screens of an electronic display). Guideline 13.115: Minimize the Number of Colors In general, using only two or three colors on a label is better than using eight or nine colors. If more categories are needed, then other methods of coding should be incorporated, such as shape, patterns, and so on.
Guideline 13.116: Meaning of Color Codings Color conventions and meanings should play influence decisions about the choice of color coding.
Worldwide, there are differences with respect to color conventions, such as color coding of medical gas cylinder contents used in medical procedures. For example, different combinations of green and gray are used to designate oxygen, carbon dioxide, and nitrogen cylinders. Patient deaths have occurred when a mix-up between carbon dioxide (nonflammable) and oxygen (flammable) has taken place and an ignition source was applied (e.g., a surgical laser). As mentioned earlier, there is a device warning label standard that recommends certain colors for designating hazards (ANSI Z535, 2002). This standard recommends the use of three different colors to indicate differing levels of hazard (in decreasing order): red, orange, and yellow (refer to Figure 13.24). ISO recommendations for the use of color for warning labels are similar. The colors blue and green may be used for safety-related and other important information. See ANSI Z535.1 (2002) for specifications of color in terms of CIE color space and associated Pantone chips. Guideline 13.117: Use of Red Color Designers should use red to indicate hazards that, if not avoided, will lead to death (ANSI Z535, 2002). Red is also used for fire safety and emergency stop control. The use of red to indicate other kinds of conditions should be minimized.
TAF-ER562X_WINGR-09-0401-C013.indd 585
11/9/10 3:46:10 AM
586
Handbook of Human Factors in Medical Device Design
Guideline 13.118: Use of Orange Color According to ANSI Z535 (2002), the color orange should generally be used to indicate hazards that, if not avoided, can lead to serious injury or death. In the signal word panel, the print is black, and the background is orange. Some rendered oranges have insufficient contrast in certain lighting conditions and should be avoided on labels.
Guideline 13.119: Use of Yellow Color The color yellow should generally be used for advisory messages, including warning of hazards that, if not avoided, could lead to minor injury or property damage (ANSI Z535, 2002). However, people generally tend to view orange and yellow as connoting similar levels of hazard. Black print on a yellow background is much more legible than black print on an orange background.
13.6.8.3 Size Coding Size coding generally applies to actual controls and connectors but might also be used in labels to provide an additional means of enhancing visual discrimination. However, size coding is generally less effective than most other kinds of coding methods (Sanders and McCormick, 1993). Guideline 13.120: Consistency of Label Element Size Similar elements (or elements used for similar functions) could be coded with the same-size labels or labeling elements.
Guideline 13.121: Number of Sizes Generally, no more than three different sizes should be employed. The ability to make reliable distinctions among different-size elements depends on the magnitude of these differences: • Large differences. Generally, larger differences between the size code components make them more distinguishable. • Twenty percent bigger. In general, the largest device element should be at least 20% bigger than the smallest.
Optimal size coding also depends on viewing distance: Greater expected viewing distances would require larger differences in size between the device elements. 13.6.8.4 Location Coding Guideline 13.122: Location Coding Designers can use location coding to relate device elements according to functional groups or sequence of use. Location coding should be applied consistently across devices and, where possible, systems.
13.6.8.5 Shape Coding Shape coding in labeling can strengthen the association between a control and its function. The ISO advocates the use of shape coding in labeling (Warburton, 2004). However, ANSI is less enthusiastic about shape coding because it may be a weaker coding modality than, for example, color. See Sanders and McCormick (1993) for an overview of various shape and control dimensions.
TAF-ER562X_WINGR-09-0401-C013.indd 586
11/9/10 3:46:10 AM
Signs, Symbols, and Markings
587
Guideline 13.123: Recognizable Shapes When devising shape-coding systems for control devices, designers should incorporate shapes that are highly recognizable, such as circles, squares, and triangles (Riley, Cochran, and Ballard, 1982).
Guideline 13.124: Tactile Coding Shape coding can also involve tactile cues, such as raised labels or buttons (e.g., Mendat and Wogalter, 2004). This is a type of texture coding that is otherwise uncommonly used in label design.
13.6.8.6 Graphics and Symbols Guideline 13.125: Use Recognizable Symbols Graphics and symbols should be selected or developed that are easily recognizable by the intended user population. In general, graphics and symbols that closely resemble their referents are more easily understood than those used to represent abstract concepts, such as radiation and biological risk (see Figure 13.27). Such abstract symbols and graphics may require training and/or accompanying text to ensure that users will interpret them correctly. When using abstract graphics or symbols for critical information transmission, other forms of coding should be considered.
Guideline 13.126: Consistency and Convention in Symbol Choice Symbols integrated into graphic depictions of flow paths (e.g., pumps, filters, valves, gain controls) should be based on consistency and convention.
Guideline 13.127: Symbol Testing The effectiveness of symbols intended to convey critical information should be established through user testing. Testing protocols are described in ANSI Z535.3 (2002) for safety symbols and in several chapters in Wogalter (2006) for warning labels (e.g., see Deppa, 2006).
13.6.9 SECTION SUMMARY This section provided specific design guidelines for medical device labels. The guidelines are based on human factors principles and will facilitate label designs that are both effective and meet prevailing legal requirements. The guidelines provided here are intended as general recommendations and should not be applied in a “cookbook” fashion. There is an almost infinite variety of medical device designs and configurations, and exceptions or better alternatives are inevitable. Indeed, medical device use environments can vary greatly (e.g., operating rooms, emergency rooms, hospital units, ambulances, homes), and many factors contribute to label effectiveness, such as user characteristics (e.g., current knowledge, stress), the expected use environment (e.g., lighting), and the task demands and exigencies (e.g., viewing distances). The effectiveness of any particular label design will depend on how it is used. Testing under the various expected use conditions can help to evaluate label effectiveness and determine the adequacy of the label’s design. Existing standards provide some guidance for label development and testing, such as the addendum/appendix of ANSI Z535.3 (2007). Wogalter, Conzola, and Vigilante (2006) outline usability principles that could be used when developing label text. Without testing, the label designer does not know whether
TAF-ER562X_WINGR-09-0401-C013.indd 587
11/9/10 3:46:10 AM
588
Handbook of Human Factors in Medical Device Design
Examples of abstract symbols
Biological risk
Laser radiation
Symbols on medical devices without accompanying text
Symbols on medical devices accompanied by text
FIGURE 13.27 Users may require training or accompanying text to understand the meaning of symbols and graphics. Iterative testing with a representative sample of likely users can help determine comprehensibility of symbols.
the label will do its job—fulfill its intended mission—to inform, to facilitate compliance, and to remind.
13.7 CONCLUSIONS This chapter consisted of three sections organized around key design issues. The first section addressed the critical issue of what should be labeled and focused on the specific legal requirements and voluntary standards that guide medical device labeling in the United States. The second section presented an overview of the relevant human factors literature that should be considered when designing labels for a medical device. This section used a C-HIP (Communication-Human Information Processing) framework as a means of organizing and understanding the labeling literature. Designers can use C-HIP as a developmental tool, while investigators can use it as an analytical tool. In the third section, specific guidelines were provided for developing effective medical device labels and their
TAF-ER562X_WINGR-09-0401-C013.indd 588
11/9/10 3:46:10 AM
Signs, Symbols, and Markings
589
associated components. Because of the breadth and scope of medical devices, specifics for any given application could not be provided. All the principles will not be applicable to any given device, and usability evaluation (testing) will be required for critical labeling design decisions. The overall goal of the chapter, of course, is to provide designers with practical guidelines for developing on-device labels for medical devices. It is intended to advance the labeling of medical devices by assisting designers in designing, formatting, and positioning of labels and markings for controls, displays, panels, and associated equipment. Labels should be considered as a supplement to, not a substitute for, good device design. (Lehto and Salvendy, 1995). An exposed switch mounted on the surface of a control panel is more likely than a recessed one to be inadvertently activated despite an effective warning label. The guidance provided for labels in this chapter may apply to other instructional and warning materials besides on-device labeling (see Chapter 5, Documentation). Legal requirements, standards, and human factors principles specify certain characteristics to ensure that medical device labeling is effective. Labels and markings on medical devices should be attention getting, understandable, believable and motivate compliance. In addition, label designs should take into account local conventions and meanings associated with specific markings as well as the abilities and limitations of the intended user population. Controls, displays, and other components of medical devices should be labeled appropriately and clearly to assure rapid and accurate human performance and to prevent user errors which could cause user or patient injury. Finally, gathering user input during the development process is vital to ensure that labels meet users’ needs.
ACKNOWLEDGMENTS The authors are grateful to Peter B. Carstensen and Charles (Dick) Sawyer from the U.S. Food and Drug Administration for their work in helping to develop a preliminary version of this chapter. The authors are grateful for the assistance in allowing photographs to be taken of various medical devices at Albany Medical Center (Albany, NY) and Duke Medical Center (Durham, NC) hospitals. Thanks are also due to Jennifer A. Cowley, a graduate student in the Department of Psychology at North Carolina State University, and Courtney Ellert, a student in the Electronic Media, Arts, and Communication program at Rensselaer Polytechnic Institute.
REFERENCES Barlow, T. and Wogalter, M. S. (1991). Increasing the surface area on small product containers to facilitate communication of label information and warnings. In Proceedings of Interface 91 (pp. 88–93). Santa Monica, CA: Human Factors Society. Baron, R. A. and Kalsher, M. J. (2005). Psychology: From Science to Practice. Needham Heights, MA: Allyn & Bacon. Bix, L., Lockhart, H., Cardoso, F., and Selke, S. (2003). The effect of color contrast on message legibility. Journal of Design Communication. Available: http://scholar.lib.vt.edu/ejournals/JDC/ Spring-2003/colorcontrast.html. Braun, C. C. and Silver, N. C. (1995). Interaction of signal word and colour on warning labels: Differences in perceived hazard and behavioural compliance. Ergonomics, 38, 2207–2220. Brehler, R. and Kutting, B. (2001). Natural rubber latex allergy: A problem of interdisciplinary concern in medicine. Archives of Internal Medicine, 161, 1057–1064.
TAF-ER562X_WINGR-09-0401-C013.indd 589
11/9/10 3:46:11 AM
590
Handbook of Human Factors in Medical Device Design
Chapanis, A. and Yoblick, D. A. (2001). Another test of sensor lines on control panels. Ergonomics, 44, 1302–1311. Council Directive 93/42/EEC concerning Medical Devices. (1993, June 14). Retrieved May 5, 2005, from http://europa.eu.int/smartapi/cgi/sga_doc?smartapi!celexapi!prod!CELEXnumdoc&lg= en&numdoc=31993L0042&model=guichett Cohen, H. H., Cohen, J., Mendat, C. C., and Wogalter, M. S. (2006). Warning channel: Modality and media. In M. S. Wogalter (Ed.) Handbook of Warnings (pp. 123–134). Mahwah, NJ: Lawrence Erlbaum Associates. Cox, E. P. III, and Wogalter, M. S. (2006). Warning source. In M. S. Wogalter (Ed.) Handbook of Warnings (pp. 111–122). Mahwah, NJ: Lawrence Erlbaum Associates. Cox, E. P., III, Wogalter, M. S., Stokes, S. L., and Murff, E. J. T. (1997). Do product warnings increase safe behavior? A meta-analysis. Journal of Public Policy and Marketing, 16, 195–204. DeJoy, D. M. (1999). Attitudes and beliefs. In M. S. Wogalter, D.M. DeJoy, & K.R. Laughery (Eds.) Warnings and Risk Communication (pp. 189–219). London: Taylor & Francis. Deppa, S. W. (2006). U.S. and international standards. In M. S. Wogalter (Ed.), Handbook of Warnings (pp. 477–486). Mahwah, NJ: Lawrence Erlbaum Associates. Desaulniers, D. R. (1987). Layout, organization, and the effectiveness of consumer product warnings. In Proceedings of the Human Factors Society 31st Annual Meeting (pp. 56–60). Santa Monica, CA: Human Factors Society. Dyck, R. (2000). Historical development of latex allergy. AORN Journal, 72(1), 27. Glasscock, N. F. and Dorris, N. T. (2006). Warning degradation and durability. In M. S. Wogalter (Ed.), Handbook of Warnings (pp. 499–511). Mahwah, NJ: Lawrence Erlbaum Associates. Godfrey, S. S., Allender, L., Laughery, K. R. and Smith, V. L. (1983) Warning messages: Will the consumer bother to look? In Proceedings of the Human Factors Society 27th Annual Meeting (pp. 950–954). Santa Monica, CA: Human Factors Society. Goldsworthy, R. and Kaplan, B. (2006). Warning symbol development: A case study on teratogen symbol design and evaluation. In M. S. Wogalter (Ed.), Handbook of Warnings (pp. 739–755). Mahwah, NJ: Lawrence Erlbaum Associates. Frascara, J. (2006). Typography and the visual design of warnings. In M. S. Wogalter (Ed.), Handbook of Warnings (pp. 385–406). Mahwah, NJ: Lawrence Erlbaum Associates. Hartley, J. (1994). Designing Instructional Text (3rd ed.). London: Kogan Page; East Brunswick, NJ: Nichols. International Electrotechnical Commission. (2007). IEC 60601-1 Standard, Medical Electrical Equipment—Part 1: General Requirements for Safety. Geneva: International Electrotechnical Commission. Kalsher, M. J., Wogalter, M. S., and Racicot, B. M. (1996). Pharmaceutical container labels and warnings: Preference and perceived readability of alternative designs and pictorials, International Journal of Industrial Ergonomics, 18, 83–90. Laughery, K. R. and Paige-Smith, D. (2006). Explicit information in warnings. In M. S. Wogalter (Ed.), Handbook of Warnings (pp. 419–428). Mahwah, NJ: Lawrence Erlbaum Associates. Laughery, K. R., Vaubel, K. P., Young, S. L., Brelsford, J. W., and Rowe, A. L. (1993). Explicitness of consequence information in warning. Safety Science, 16, 597–613. Laughery, K. R., Wogalter, M. S., and Young, S. L. (Eds.). (1994). Human Factors Perspectives on Warnings: Selections from Human Factors and Ergonomics Society Annual Meetings 1980– 1993. Santa Monica, CA: Human Factors and Ergonomics Society. Lehto, M. R., and Miller, J. M., (1986). Warnings, Volume 1: Fundamentals, Design and Evaluation Methodologies. Fuller Technical, Ann Arbor, MI. Lehto, M., and Salvendy, G. (1995). Warnings: A supplement not a substitute for other approaches to safety. Ergonomics, 38, 2155–2163. Mayhorn, C. B. and Podony, K. (2006). Warnings and aging: Describing the receiver characteristics of older adults. In M. S. Wogalter (Ed.), Handbook of Warnings (pp. 355–362). Mahwah, NJ: Lawrence Erlbaum Associates.
TAF-ER562X_WINGR-09-0401-C013.indd 590
11/9/10 3:46:11 AM
Signs, Symbols, and Markings
591
Mazis, M. B. and Morris, L. A. (1999). Channel. In M. S. Wogalter, D. M. DeJoy, & K. R. Laughery (Eds.) Warnings and risk communication (pp. 99–121). London: Taylor & Francis. Mendat, C. C. and Wogalter, M. S. (2004). The effectiveness of tactile cues in cellular phones. Proceedings of the Human Factors and Ergonomics Society, 48, 717–720. Miller, J. M. and Parent, C. (2006). Appendix: Bibliography of standards. In M. S. Wogalter (Ed.), Handbook of Warnings (pp. 795–804). Mahwah, NJ: Lawrence Erlbaum Associates. Misanchuk, E. R. (1992). Preparing Instructional Text: Document Design Using Desktop Publishing. Englewood Cliffs, NJ: Educational Technology. Riley, D. M. (2006). Beliefs, attitudes, and motivation. In M. S. Wogalter (Ed.) Handbook of Warnings (pp. 289–300). Mahwah, NJ: Lawrence Erlbaum Associates. Riley, M. W., Cochran, D. J., and Ballard, J. L., 1982. An investigation of preferred shapes for warning labels. Human Factors, 24, 737–742. Sanders, M. and McCormick, E. (1993). Human Factors in Engineering and Design. New York: McGraw-Hill. Sawyer, D. (1993, June 14). Do it by design: An Introduction to Human Factors in Medical Devices. Retrieved May 5, 2005, from http://www.fda.gov/cdrh/humfac/doit.html Simpson, H. and Casey, S. M. (1988). Designing Effective User Documentation: A Human-factors Approach. New York: McGraw-Hill. Smith-Jackson, T. L. (2006a). Culture and warnings. In M. S. Wogalter (Ed.), Handbook of Warnings (pp. 363–372). Mahwah, NJ: Lawrence Erlbaum Associates. Smith-Jackson, T. L. (2006b). Receiver characteristics. In M. S. Wogalter (Ed.), Handbook of Warnings (pp. 335–344). Mahwah, NJ: Lawrence Erlbaum Associates. Sojourner, R. J. and Wogalter, M. S. (1997). The influence of pictorials on evaluations of prescription medication instructions. Drug Information Journal, 31, 963–972. Study Group 1 of the Global Harmonization Task Force. (2002, March 1). Labelling for Medical Devices (including in vitro diagnostic devices). Retrieved May 3, 2004, from www.ghtf.org/ sg1/inventorysg1/sg1pd-n043r3.pdf Underwriters Laboratories, Inc. (n.d.). CE Mark Info. Retrieved May 5, 2004, from http://www. ul.com/regulators/CEmarkinfo.html U.S. Food and Drug Administration. (2000, July 18). Guidance for Industry and FDA Premarket and Design Control Reviewers—Medical Device Use-safety: Incorporating Human Factors Engineering into Risk Management. Retrieved May 5, 2004, from http://www.fda.gov/cdrh/ humfac/1497.html#_Toc486653483 Vigilante, W. J., Jr., and Wogalter, M. S. (1997). The preferred order of over-the-counter (OTC) pharmaceutical label components. Drug Information Journal, 31, 973–988. Vredenburgh, A. G. and Helmick-Rich, J. (2006). Extrinsic nonwarning factors. In M. S. Wogalter (Ed.), Handbook of Warnings (pp. 373–381). Mahwah, NJ: Lawrence Erlbaum Associates. Vredenburgh, A. G. and Zackowitz, I. B. (2006). Expectations. In M. S. Wogalter (Ed.), Handbook of Warnings (pp. 345–354). Mahwah, NJ: Lawrence Erlbaum Associates. Warburton, D. (2004, April). Are your product safety labels state-of-the-art? Medical Device and Diagnostic Industry. Retrieved May 4, 2004, from http://www.devicelink.com/mddi/ archive/04/04/001.html Watanabe, R. K. (1994). The ability of the geriatric population to read labels on over-the-counter medication containers. Journal of the American Optometric Association, 65, 32–37. Weinger, M. B., Herndon, O. W., Paulus, M. P., Gaba, D., Zornow, M. H., and Dallen, L. D. (1994). Objective task analysis and workload assessment of anesthesia providers. Anesthesiology, 80, 77–92. Weinger, M. B., Slagle, J. M., Kim, R. S., and Gonzales, D. C. (2001). A task analysis of the first weeks of training of novice anethesologists. Proceedings of the Human Factors and Ergonomics Society, 45, 404–408. Wickens, C. D., Lee, J., Liu, Y. D., and Gordon-Becker, S. (2003). Introduction to Human Factors Engineering (2nd ed.). Englewood Cliffs, NJ Prentice Hall.
TAF-ER562X_WINGR-09-0401-C013.indd 591
11/9/10 3:46:11 AM
592
Handbook of Human Factors in Medical Device Design
Wiklund, M. and Dolan, W. (1996, October). Why choose color displays? Medical Device and Diagnostic Industry Magazine. Retrieved May 4, 2004, from http://www.devicelink.com/ mddi/archive/96/10/015.html Wogalter, M. S. (2006). Handbook of Warnings. Mahwah, NJ: Lawrence Erlbaum Associates. Wogalter, M. S., Allison, S. T., and McKenna, N. A. (1989). The effects of cost and social influence on warning compliance. Human Factors, 31, 133–140. Wogalter, M. S., Brelsford, J. W., Desaulniers, D. R., and Laughery, K. R. (1991). Consumer product warnings: The role of hazard perception. Journal of Safety Research, 22, 71–82. Wogalter, M. S., Brems, D. J., and Martin, E. G. (1993). Risk perception of common consumer products: Judgments of accident frequency and precautionary intent. Journal of Safety Research, 24, 97–106. Wogalter, M. S., Conzola, V. C., and Vigilante, W. J., Jr. (2006). Applying usability engineering principles to the design and testing of warning text. In M. S. Wogalter (Ed.), Handbook of Warnings (pp. 487–498). Mahwah, NJ: Lawrence Erlbaum Associates. Wogalter, M. S., DeJoy, D. M., and Laughery, K. R. (1999). Warnings and Risk Communication. London: Taylor & Francis. Wogalter, M. S., Forbes, R. M., and Barlow, T. (1993). Alternative product label designs: Increasing the surface area and print size. In Proceedings of Interface 93 (pp. 181–186). Santa Monica, CA: Human Factors Society. Wogalter, M. S., Godfrey, S. S., Fontenelle, G. A., Desaulniers, D. R., Rothstein, P. R., and Laughery, K. R. (1987). Effectiveness of warnings. Human Factors, 29, 599–612. Wogalter, M. S., Kalsher, M. J., and Rashid. R. (1999). Effect of signal word and source attribution on judgments of warning credibility and compliance likelihood. International Journal of Industrial Ergonomics, 24, 185–192. Wogalter, M. S. and Laughery, K. R. (1996). WARNING: Sign and label effectiveness. Current Directions in Psychology, 5, 33–37. Wogalter, M. S. and Leonard, S. D. (1999). Attention capture and maintenance. In M. S. Wogalter, D. M. DeJoy, and K. R. Laughery (Eds.), Warnings and Risk Communication (pp. 123–148). London: Taylor & Francis. Wogalter, M. S., Magurno, A. B., Dietrich, D., and Scott, K. (1999). Enhancing information acquisition for over-the-counter medications by making better use of container surface space. Experimental Aging Research, 25, 27–48. Wogalter, M. S. and Post, M. P. (1989). Printed computer instructions: The effects of screen pictographs and text format on task performance. In Proceedings of Interface 89 (pp. 133–138). Santa Monica, CA: Human Factors Society. Wogalter, M. S., Silver, N. C., Leonard, S. D., and Zaikina, H. (2006). Warning symbols. In M. S. Wogalter (Ed.), Handbook of Warnings (pp. 159–176). Mahwah, NJ: Lawrence Erlbaum Associates. Wogalter, M. S. and Usher, M. (1999). Effects of concurrent cognitive task loading on warning compliance behavior. Proceedings of the Human Factors and Ergonomics Society, 43, 106–110. Wogalter, M. S. and Vigilante, W. J., Jr. (2003). Effects of label format on knowledge acquisition and perceived readability by younger and older adults. Ergonomics, 46, 327–344. Wogalter, M. S. and Vigilante, W. J., Jr. (2006). Attention switch and maintenance. In M. S. Wogalter (Ed.), Handbook of Warnings (pp. 245–265). Mahwah, NJ: Lawrence Erlbaum Associates. Wogalter, M. S., Vigilante, W. J., and Baneth, R. C. (1998). Availability of operator manuals for used consumer products. Applied Ergonomics, 29, 193–200. Wogalter, M. S. and Young, S. L. (1994). Enhancing warning compliance through alternative product label designs. Applied Ergonomics, 25, 53–57. Wogalter, M. S., Young, S. L., and Laughery, K. R., Eds. (2001). Human Factors Perspectives on Warnings, Volume 2: Selections from Human Factors and Ergonomics Society Annual Meetings 1994–2000. Santa Monica: Human Factors and Ergonomics Society, in press.
TAF-ER562X_WINGR-09-0401-C013.indd 592
11/9/10 3:46:11 AM
Signs, Symbols, and Markings
593
Wolfe, J. M., Kluender, K. R., Levi, D. M., Bartoshuk, L. M., Herz, R. S., Klatzky, R. L., et al. (2006). Sensation and Perception. Sunderland, MA: Sinauer Associates. Wright, P. (1982). Some factors determining when instructions will be read. Ergonomics, 25, 225–237. Young, S. L., Wogalter, M. S., Laughery, K. R., Magurno, A., and Lovvoll, D. (1995). Relative order and space allocation of message components in hazard warning signs. Proceedings of the Human Factors and Ergonomics Society, 39, 969–973. Zak, H. N., Kaste, L. M., and Schwarzenberger, K. (2000, September/October). Health-care workers and latex allergy. Archives of Environmental Health, 55, 336–346.
TAF-ER562X_WINGR-09-0401-C013.indd 593
11/9/10 3:46:11 AM
TAF-ER562X_WINGR-09-0401-C013.indd 594
11/9/10 3:46:11 AM
14 Packaging Michael E. Maddox, PhD, CHFP; Larry W. Avery, MA CONTENTS 14.1 General Principles ..................................................................................................596 14.1.1 Packaging as Part of Device Design ........................................................596 14.1.2 Use in the Home or in a Health Care Facility..........................................597 14.1.3 Use by Professionals or by Consumers (Patients) ....................................597 14.2 Special Considerations ...........................................................................................598 14.2.1 Drastically Different Use Environments..................................................598 14.2.2 Widely Disparate User Populations .........................................................599 14.3 Design Guidelines ................................................................................................. 600 14.3.1 Access ..................................................................................................... 600 14.3.2 Removal and Replacement.......................................................................603 14.3.3 Assembly or Sequential Use of Components .......................................... 604 14.3.4 Cleaning/Sterilization ............................................................................. 607 14.3.5 Identification ........................................................................................... 609 14.3.6 Coding ..................................................................................................... 611 14.3.7 Labeling ................................................................................................... 612 14.3.8 Design for Packaging ............................................................................... 614 14.3.9 Handling .................................................................................................. 615 14.3.10 Shipping/Movement................................................................................. 615 14.3.11 Inventory .................................................................................................. 616 14.3.12 Disposal ................................................................................................... 619 14.4 Case Studies ...........................................................................................................620 14.4.1 Hearing Aid Batteries ..............................................................................620 14.4.2 Mixing a Dental Adhesive .......................................................................621 Resources .........................................................................................................................621 References ........................................................................................................................621
Packaging plays a role in every stage in the life cycle of a medical device. This chapter is most concerned with the aspects of medical device packaging that (1) directly affect the identification, handling, and use of medical devices; (2) are unlikely to be addressed by standards and guidelines from other specialties or organizations; and (3) are amenable to guidance based on existing human factors data and methods. Certain aspects of medical device packaging have received a great deal of attention. For example, the discovery of HIV and other infectious diseases increased the emphasis on the 595
TAF-ER562X_WINGR-09-0401-C014.indd 595
11/9/10 12:35:53 AM
596
Handbook of Human Factors in Medical Device Design
safe packaging, use, and disposal of sharps (“sharps” is the name given to needles, scalpels, and so on). Developments in sterilization have prompted manufacturers to create packaging materials that can survive ionizing radiation and gaseous environments. More recently, device designers have paid greater attention to the ease of use of home health care devices such as glucometers and the disposable elements packaged and used with them. Human factors aspects of medical device packaging have been neglected by the human factors community and the U.S. Food and Drug Administration (FDA). The index of any standard human factors textbook is unlikely to include the term “packaging.” In product design–oriented texts, such as Cushman and Rosenberg (1991), packaging information is limited to two sentences and a table—all related to the safety of packaging. In the FDA’s human factors design guidelines Do It by Design (Sawyer, 1996), medical device packaging rates two paragraphs of general information (in a 55-page document). A classic human factors text devoted entirely to medical device design (Wiklund, 1995) devotes nine pages to a general discussion of packaging design and provides anecdotal descriptions of poor packaging design. Much of the material in this chapter derives not from publications within the human factors community, but from healthcare industry sources. For example, Hunstiger (1988) addresses a number of usability and “user friendliness” issues in an early article that appeared in the TAPPI (Technical Association of the Pulp and Paper Industry) Journal. Hultberg (1990), in a TAPPI conference proceedings paper, discusses many of the topics addressed in this chapter. The principles and guidelines in this chapter do not address topics related specifically to device development and production. Likewise, the chapter does not address industrial design or industrial hygiene issues, such as color and material selection, toxicity, and so on, except to the extent that they directly affect user tasks. These issues are adequately addressed elsewhere (e.g., Leventon, 2001; Rossi, 2003). Instead, this chapter addresses ergonomic, perceptual, and cognitive issues of packaging. These should be considered early in the design process because they ultimately determine users’ ability to interact with device packaging. It is important that the appropriate user and environmental characteristics are considered during the packaging design phase. It is also imperative that packaging supports user tasks in the real-world settings where the devices are to be used. This chapter explicitly distinguishes among devices and products that will be used primarily by professional health care providers (typically in institutional settings) and those targeting the consumer market (in noninstitutional settings).
14.1 GENERAL PRINCIPLES The general principles of user-centered design are covered in other parts of this book (see Chapter 1, “General Principles”). The following section describes certain principles that are related to packaging. However, the core concept is the same: Packaging should reflect the capabilities and limitations of the targeted user population(s) as well as the tasks that users must perform and the environments in which users must work.
14.1.1 PACKAGING AS PART OF DEVICE DESIGN Too often, packaging is left as the very last step in the product design process. This is unfortunate since critical packaging issues are much easier to address when they are considered
TAF-ER562X_WINGR-09-0401-C014.indd 596
11/9/10 12:35:53 AM
Packaging
597
as integral requirements as the design progresses. There are professional packaging designers in the medical device design community and in the more general industrial design domain. Such expertise is as important to device design as is expertise in materials, electronics, software, usability, or risk management. General principles of packaging design within the broader context of overall device design are: • Packaging requirements should be addressed during the initial device design process rather than after the prototypical design is complete. • The design characteristics of a medical device can facilitate or complicate the design of its packaging. • The ultimate “usability” of a medical device can depend as much on its packaging as on its inherent design characteristics. • The general packaging design considerations related to medical devices are similar and often identical to the considerations for other consumer and special-use products.
14.1.2 USE IN THE HOME OR IN A HEALTH CARE FACILITY Using medical devices in the home environment results in many potentially dangerous use errors (Landro, 2004). From a human factors perspective, the use of a medical device in the home environment is distinctly different from the use of a similar device in the health care environment. The most obvious differences are the education, training, and experience of the end users in these settings. However, there are many other, less obvious differences that can affect the design, storage, use, maintenance, disposal, and other aspects of a device’s packaging. Principles of packaging for the home care environment include: • The home is a fundamentally different use environment than is a health care facility. • Users in health care facilities have resources at their disposal that are not available in the home environment. • Users in the home environment are likely to be untrained in the use of medical devices, to be unfamiliar with medical terminology, and to be affected by illness or infirmity.
14.1.3 USE BY PROFESSIONALS OR BY CONSUMERS (PATIENTS) The consumer medical device market includes patients and home/lay caregivers. It is likely that significant differences between these targeted user groups will exist at all levels of usability. At the lowest level—that is, compatibility with physical, psychological, and perceptual capabilities—consumers are likely to be older and exhibit more manipulative, perceptual, and cognitive limitations than professional health care workers. The reason patients use medical devices is that they are sick, injured, or have chronic medical or other conditions that can impair their physical and/or cognitive capabilities. An example of such a chronic condition is rheumatoid arthritis, which often can affect a patient’s ability to remove an injectable rheumatoid arthritis medication from
TAF-ER562X_WINGR-09-0401-C014.indd 597
11/9/10 12:35:53 AM
598
Handbook of Human Factors in Medical Device Design
its packaging and self-administer it. Principles of product packaging for patients and lay caregivers include: • Health care professionals are familiar with various types of medical device– specific packaging, but typical consumers are not likely to have such familiarity. • Home-use consumers will not be familiar with the proper disposal of some types of medical packaging, such as sharps, biohazards, and so on.
14.2 SPECIAL CONSIDERATIONS In certain respects, medical device packaging requirements are not dissimilar from those of other commercial products. For example, the requirements for cleanliness are at least as stringent for products used in the manufacture of electronic components and computers. The ranges of age, education, and experience of potential users of consumer electronics is at least as broad as for the users of home health medical devices. Users of medical products do not appear to be particularly prone to specific types of packaging-related errors. Yet medical devices and their packaging are undoubtedly subjectively viewed and evaluated differently than products and packaging from other domains. A key difference for medical device packaging is that poorly designed, badly executed, and otherwise deficient packaging can have direct and serious effects on users’ and patients’ lives. The risks that consumers routinely assume in other aspects of their daily lives are considered unacceptable when applied to the use of medical products and procedures. Scientists (and readers of scientific studies) commonly accept a 5% risk that the results of a particular statistical test are in error (i.e., p < 0.05). How many of those same individuals would accept a 1 in 20 chance that a routine medical procedure, a medical device, a drug, or the results of a diagnostic test will seriously injure or kill them? Some medical device packaging characteristics are clearly outside the scope of human factors considerations. For example, certain products must be packaged in materials that allow (or prevent) gas diffusion to occur at certain rates. This type of requirement seems to be squarely in the “materials science” domain. Other issues are equally easy to classify as falling within the realm of human factors requirements. An example is the requirement for packaging to be easy to open by individuals wearing latex gloves. Certainly, there are “materials” issues here, but this is essentially a usability requirement. There are many gray areas that touch on issues from different domains, including human factors. For example, the decision to use recyclable packaging materials is not primarily a human factors issue. However, once the decision has been made to use recyclables, the clear depiction of understandable recycling symbols is within the human factors domain. From a device packaging perspective, there appear to be at least two considerations that, at least in combination with the potentially dire consequences of packaging-related errors, are unique to the medical domain. The packaging design issues that pertain to each of these special considerations are described in the following sections.
14.2.1 DRASTICALLY DIFFERENT USE ENVIRONMENTS Certain medical devices will be used in specific and often extremely well controlled environments. For example, cardiologists and surgeons insert arterial stents while working in the sterile environment of a catheterization lab or operating suite. Their packaging must
TAF-ER562X_WINGR-09-0401-C014.indd 598
11/9/10 12:35:53 AM
Packaging
599
take into account the sterile field in which they will be opened and used. Rheumatoid arthritis medications are designed and packaged for consumer use in their homes. Thus, package designers can count on the lack of a sterile field, protective gloves, and so on when the medication is opened. Moreover, users of rheumatoid arthritis medications are likely to have compromised manipulative capabilities, and the packaging must reflect these conditions of use. On the other hand, some medical devices are used in an extremely wide variety of environments by a broad range of users and under extremely disparate circumstances. Examples include syringes, intravenous (IV) fluids and medications, venous and arterial catheters, cardiac and pulmonary monitors, defibrillators, and dressings. Their use environments range among controlled hospital settings, chaotic emergency rooms, on-scene and in-transit emergency medical situations, nursing homes, and even personal residences (end-use consumers). The packaging for such medical devices must meet their usability requirements in environments that include widely divergent levels of lighting, noise, temperature, and movement. They might be used outside in summer or winter, day or night, in precipitation, in a moving vehicle, and so on. Some key principles of packaging for divers medical device use environments include: • Packaging that works well in one use environment might be completely inappropriate for another use environment. • Packaging design should accommodate the full range of use environment(s) of all potential users. • The human factors basis for packaging design should be well documented. • The intended use environments and their implications for device packaging should be described explicitly in the design documentation.
14.2.2 WIDELY DISPARATE USER POPULATIONS Designing medical packaging would be much easier if there was a guarantee that every user would be a trained medical professional with no perceptual, cognitive, or physical problems. This is obviously not the case. Sometimes, medical devices are actually designed differently for each targeted population. A very good example of this approach is the much-simplified packaging, instructions, and user interface for automated defibrillators designed for use by nonmedical professionals. More often, however, the same device is used by both very skilled and completely novice users. Examples include infusion pumps and IV administration supplies. The existing “packaging” literature is replete with discussions of various technical requirements related to materials, technology, and labeling. For example, many recent papers describe the pros and cons of “unit dose” packaging and bar code labeling (Frieden, 2002; Lipowski et al., 2002). These packaging considerations are usually framed in terms of handling by pharmacists and the ultimate impact on medication safety. However, very little attention has been paid to the diversity of users who ultimately interact with medical device packaging. The impact of different population capabilities on users’ ability to interact with packaging properly and efficiently may be dramatic. A recent study by the Agency for Healthcare Research and Quality (Berkman et al., 2004) identified significant connections between illiteracy, which affects approximately 90 million people in the United States, and the
TAF-ER562X_WINGR-09-0401-C014.indd 599
11/9/10 12:35:53 AM
600
Handbook of Human Factors in Medical Device Design
TABLE 14.1 Representative Differences between Medical Professionals and Lay Users of Medical Devices User Issues
Medical Professionals
Lay Users
Older vs. younger
Younger
Older
Physical, perceptual, and cognitive limitations Likely to suffer chronic health conditions that might affect ability to interact with packaging Understand medical terminology Understand “standard” medical icons, symbols, and graphics Well trained in the use of medical devices Commonly use devices that can affect the well-being of themselves or others Broad experience with medical device packaging Experience using and disposing of toxic materials, sharps, and so on Typical use setting for medical devices
No No
Yes Yes
Yes Yes
No No
Yes Yes
No No
Yes Yes
No No
Institutional and office
Residential
ability to understand and interpret medical terminology, directions, device labeling, and so on. For example, a study in two large hospitals found that 35% of English-speaking patients had insufficient literacy to function in a health care setting. The average reading level of patient materials is in the 11th- to 14th-grade range, but the average reading skill level of many patients is generally much lower. Readers should note, however, that significant usage and safety issues pertaining to medical professionals have also been identified (Wagner, 2002). The two main user populations for medical devices are “medical professionals” and “nonmedical-professional consumers” (usually known as “lay users” or “patients”). Typical lay users are family members or good Samaritans using automated defibrillators. There are vast and important differences among members of these two user populations with regard to medical product packaging. Some of the more salient differences are given in Table 14.1.
14.3 DESIGN GUIDELINES The guidelines in this section are categorized according to the device packaging use they support. The categories are presented roughly in the order in which products are designed and used. Within each category, guidelines are presented more or less in their order of priority for supporting the use category.
14.3.1 ACCESS One of the more obvious uses of packaging is to facilitate or prevent access to certain medications. We are all familiar with “childproof” containers, although it might appear that many childproof packages can also be “adult proof” (Consumer Product Safety Commission, n.d.).
TAF-ER562X_WINGR-09-0401-C014.indd 600
11/9/10 12:35:53 AM
Packaging
FIGURE 14.1
601
Corner-opening package.
In studies conducted with elderly populations, opening childproof containers was often beyond the capacity of those individuals with decreased strength or manipulative capabilities (Kendrick and Bayne, 1982; Rohles, Moldrup, and Laviana, 1983). Medical device packaging incorporates a number of methods to allow or deny access. Figures 14.1 through 14.3 depict examples of common types of access methods built into medical device packages. All these access methods require the user to manually separate two packaging components. Another very common example of this access method is the standard bandage package, which requires users to grasp two opposing paper flaps and pull the outer covering apart. Elderly users and those with dexterity impairments often have great difficulty with such “pull-apart” packaging. Another common access method is shown in Figure 14.4. This “slit-edge” method is prevalent in some types of medical devices, most notably IV infusion bags. Typical instructions for using this type of opening method are shown in Figure 14.5. This type of access method is actually quite common for food products and has simply been adapted to the medical domain. The same advantages and disadvantages pertain in both domains. For example, sometimes the edge slits are not cut deep enough, making it more difficult to tear open the package. In the medical domain, there is also the problem of the user’s gloves being covered with water or other fluids, making it difficult to grasp the areas adjacent to the opening slit with enough force to tear the package. An advantage of slit-edge packaging is that it is very easy to produce, requires no tools, and, because of its use in other domains, is fairly obvious. The issue of package accessibility is directly related to identifying the intended uses and user population for the device. Rather than provide specific guidelines for designing the proper level of access into medical packaging, it is more appropriate to supply guidance
(Top)
(Bottom)
FIGURE 14.2
Edge-opening package.
TAF-ER562X_WINGR-09-0401-C014.indd 601
11/9/10 12:35:53 AM
602
Handbook of Human Factors in Medical Device Design
Top only
FIGURE 14.3
Opening flap on the back of a see-through package.
regarding how and when the proper level of access should be identified. With this perspective, accessibility guidelines are given below. Guideline 14.1: Open and Close Package Medical packaging access requirements, specifically the ability to open (and reclose) the packaging, should be well within the strength and manipulative capabilities of all intended users.
Guideline 14.2: Open Maintaining Sterility Medical device packaging that must be opened in a sterile environment should incorporate an access method that does not expose the contents to contamination during opening.
Guideline 14.3: Support Opening Wearing Gloves Medical device packaging that will be opened by individuals wearing protective gloves should incorporate an access method that is compatible with the use of gloves (e.g., use of large, nonslippery surfaces).
Guideline 14.4: Free Access to Device Packaging should not prevent access to a medical product or device unless there are explicit, specific, and unambiguous reasons for doing so.
(Tear)
FIGURE 14.4
Slit-edge opening on an IV bag outer covering.
TAF-ER562X_WINGR-09-0401-C014.indd 602
11/9/10 12:35:53 AM
Packaging
603
(Slit edge)
FIGURE 14.5
Opening instructions for a slit-edge package.
Guideline 14.5: Reasons for Restricted Access If packaging is intended to prevent access to its contents by specific types of individuals, then this intent and the reasons for such limitations should be made explicit in the front-end design documentation.
Guideline 14.6: Restricted Access Group Identity If packaging is intended to prevent access to its contents by specific types of individuals, then the identity of these groups should be clearly stated on the packaging.
14.3.2 REMOVAL AND REPLACEMENT A number of issues relate to taking a device out of its packaging and, when necessary, replacing it. Packaging designs must support the opening of the package and removal of the device inside, even by very young and old users. If the device is designed to be replaced in its packaging, then the design of the packaging must accommodate replacement and reclosing. Thus, the packaging must be easy to open and reclose. There are issues of childproof and tamperproof packaging and the loss of strength and manipulative capabilities with increasing age and for certain medical conditions, such as arthritis. Also, there is the requirement for appropriate markings and indicators to appear in or on the packaging as to how it should be opened and reclosed. Guideline 14.7: Obvious Methods to Open and Close The tasks required to open (and reclose) packaging should be obvious to intended users.
Guideline 14.8: Open without Tools Users should be able to open (and reclose) device packaging without the use of tools.
Guideline 14.9: Unintended Users Unable to Open If the packaging is meant to be a barrier to device access, unintended persons (e.g., children) should not be able to open the package.
Guideline 14.10: Removal of Device from Packaging The tasks required to remove (and replace) devices and device components should be obvious to users.
TAF-ER562X_WINGR-09-0401-C014.indd 603
11/9/10 12:35:54 AM
604
Handbook of Human Factors in Medical Device Design
Guideline 14.11: Repackaging of Devices If the device is designed to allow replacing devices or components in the original packaging, then this should be clear to users.
Guideline 14.12: Repacking Components Available All packaging components required to repack devices should be contained in the original packaging.
14.3.3 ASSEMBLY OR SEQUENTIAL USE OF COMPONENTS Some medical devices must be assembled prior to use. Others have components intended to be used in a particular sequential order. There are a number of human factors issues related to the packaging of devices and components that are meant to be assembled or used sequentially. The three most salient human factors issues regarding the assembly of components are to ensure that the end user knows that components must be assembled, understands how components should be assembled, and is physically capable of assembling the components. The most obvious packaging issue related to device assembly is determining the level of knowledge that potential users might possess regarding assembling components or using them in a particular order. In the case of nonmedical professionals, the most conservative (and safest) assumption is that users will not be familiar with the requirement to assemble components or the need for use in a certain order. There are several methods of depicting the presence of multiple parts in a package. In Figure 14.6, the contents of a package are shown diagrammatically (and by name) on the cover. In Figure 14.7, the contents are listed on the cover. Of course, the most direct method of determining which parts are in a package and whether there are multiple parts is to simply look through transparent packaging. Many
FIGURE 14.6
Diagram of the parts in a package.
TAF-ER562X_WINGR-09-0401-C014.indd 604
11/9/10 12:35:54 AM
Packaging
FIGURE 14.7
605
List of parts in a package (on left).
medical packages are made of a clear plastic part, covered with an opaque material on which the label is printed. Figure 14.8 is an example of a clear package through which users can see the parts inside. Regarding the assembly of multiple parts, some packages contain collections of parts that can be assembled in building-block fashion to construct functional units. An example of this type of device is shown in Figure 14.9, which shows a collection of parts that can be used to build intravascular pressure measurement manifolds in whatever configuration the user might need. This type of device packaging is generally intended only for experienced, professional medical users. However, its success depends on existing standards (e.g., regarding the types of connections and user expectations, see Chapter 9, “Connections and Connectors”). Finally, the required assembly tasks must be within users’ physical, perceptual, and cognitive capabilities. From a packaging perspective, this implies that packages must be easy to open and that the various parts must be easy to grasp and remove. It should also be obvious which parts fit together and how they connect. For example, a patient with rheumatoid arthritis described how she could not open the package and assembly parts of a multiplepart device containing her weekly dosage of injectable medication when she was having an arthritis flare.
FIGURE 14.8
Transparent container showing multiple parts inside.
TAF-ER562X_WINGR-09-0401-C014.indd 605
11/9/10 12:35:54 AM
606
Handbook of Human Factors in Medical Device Design
FIGURE 14.9
Multiple parts intended to be used as needed.
Guideline 14.13: Pre-Assembly of Home Use Devices If possible, devices aimed at nonmedical professionals should not contain multiple parts that require assembly or sequential use.
Guideline 14.14: Convey Assembly Information For multiple-piece devices that do require assembly, the device packaging must convey assembly or usage-order information.
Guideline 14.15: Ease of Assembly If assembly steps are required, they should be well within the physical, perceptual, and cognitive capabilities of intended users.
Guideline 14.16: Obvious Assembly Attributes The following properties of multiple-piece devices should be immediately obvious to users: • The presence of multiple parts or pieces in the package • The fact that the parts must either be assembled prior to being used or be used in a specific order • The method(s) of assembling each part to the others • The appearance, order of assembly, and function of the assembled device
TAF-ER562X_WINGR-09-0401-C014.indd 606
11/9/10 12:35:54 AM
Packaging
607
Guideline 14.17: Components Visible inside Package All device components contained in a package should be immediately visible before or when the package is opened (i.e., nothing hidden).
Guideline 14.18: Contents Presented on Package If the contents of a package are not visible through the packaging, then the contents should be listed or shown on the outside of the packaging.
Guideline 14.19: Ease of Component Handling The packaging should allow all product components to be handled easily with whatever personal protective equipment (PPE) is typically present in the usage environment.
Guideline 14.20: Obvious Order of Assembly If a sequential assembly or usage order is required, that ordering should be made obvious in the packaging.
Guideline 14.21: Risk of Component Intermingling If device components should not be interchanged with components from other packages, the packaging should unambiguously indicate this restriction.
14.3.4 CLEANING/STERILIZATION Most of the packaging issues related to cleaning and sterilization appear to be more directed to materials selection than usage (Leventon, 2002). However, users need to be able to clean/ sterilize devices in their packaging and to ascertain the current sterilization status of devices that are still in their packages. Two key human factors issues regarding cleaning and sterilization are (1) depicting the current status of sterilization and (2) clearly indicating when packaging has lost its integrity (i.e., has been opened or otherwise lost its sterility). These issues are nicely illustrated by Hunstiger (1988). Most medical facilities, including physician practices, are able to sterilize medical instruments. A typical method of sterilizing instruments is to place the articles inside a sterilizing pouch, seal the pouch, and then place the pouch inside an autoclave, which sterilizes with heat. The same pouches can be used for gas sterilization. A sterilization pouch is shown in Figure 14.10. Note the two sterilization symbols and associated labels at the top of the pouch label (Figure 14.11). Also note the “sterile access” instructional figure at the top of Figure 14.11. Figure 14.11 is reproduced in black and white. However, the sterilization symbols shown in the figure are pastel colors in their unsterilized state. The “steam” symbol on the left is a desaturated red, and the “E.O. gas” symbol is a desaturated green/blue. When the package is exposed to the proper sterilization conditions, these symbols change color to show that the package contents are now sterile. The pouch closes with a self-adhesive strip, shown in Figure 14.12. However, in some locations, it is common practice to supplement this strip with a piece of sterilization tape, shown in Figure 14.13. When the tape has been exposed to a sterilizing environment, the diagonal stripes change color, as shown in Figure 14.14. A sterilization pouch is shown in action in Figure 14.15.
TAF-ER562X_WINGR-09-0401-C014.indd 607
11/9/10 12:35:55 AM
608
FIGURE 14.10
Handbook of Human Factors in Medical Device Design
Front of a sterilization pouch.
Guideline 14.22: Indicate Ability to Sterilize Packaging should clearly indicate whether the device within can be cleaned or sterilized while still in its packaging.
Guideline 14.23: Indicate Sterilization Method Packaging should clearly indicate the appropriate/allowable methods of cleaning or sterilizing the device within.
Guideline 14.24: Indicate Sterile Opening Packaging should clearly indicate whether it must (or can) be opened in a sterile environment.
FIGURE 14.11
Sterilization symbols.
TAF-ER562X_WINGR-09-0401-C014.indd 608
11/9/10 12:35:55 AM
Packaging
FIGURE 14.12
609
Self-adhesive sterilization pouch closure.
Guideline 14.25: Indicate State of Sterility Packaging should clearly indicate the current state of cleanliness or sterilization of the contained device.
Guideline 14.26: Indicate if Sterility Comprimised Sterile packaging should clearly indicate when its integrity has been compromised.
Guideline 14.27: Indicate if Single Use Packaging for “single-use” devices should clearly indicate this characteristic and the fact that the device should not be resterilized.
14.3.5 IDENTIFICATION The ability of users to correctly identify medical devices accurately by their appearance is a cardinal element of good packaging design. In most cases, a user’s first look at a medical device is actually a view of the device package. Certain medical devices can be similar or identical in appearance but might have very different effects on patients. For example, prefilled IV flush syringes can contain saline solution or heparin. The heparin contained in these syringes can be in different concentrations. Manufacturers of these devices routinely bar code and color code the syringes and their packaging. They also provide word labels describing the volume, content, and concentration of the liquid in the syringe. Thus, properly identifying medical devices through labeling, coding, and other packaging attributes is a fundamental requirement for their safe use. The topic of product identification is extremely broad if we consider both pharmaceuticals and medical devices. There are a number of new packaging initiatives being considered for drugs, including bar coding, radio-frequency identification tagging, and unit-dose packaging (Ambrose et al., 2002; Hankin, 2002; Lipowski, Campbell, Brushwood, and Wilson, 2002). Much of the recent work for drug packaging and identification is aimed more at making drugs easier and safer to dispense and to enhance charge capture. For
FIGURE 14.13
Sterilization tape before it is sterilized.
TAF-ER562X_WINGR-09-0401-C014.indd 609
11/9/10 12:35:55 AM
610
FIGURE 14.14
Handbook of Human Factors in Medical Device Design
Sterilization tape after it is sterilized.
medical devices, it is more important (from a usability perspective) for the devices to be quickly and easily identified by users without resorting to technology. Guideline 14.28: Unambiguous Identification Packaging should enable the device(s) they contain to be unambiguously identified from the outside the package.
Guideline 14.29: Avoid Form Factor Confusion The identity of medical devices must be discernible by all users who might need to obtain this information. The form factor (overall shape, color, size, and so on) for packaging should not appear substantially similar for different devices, especially if such user confusion might cause harm. Users who are familiar with a particular brand of device tend to identify the device by the general “look” of the packaging.
Guideline 14.30: Identification Consistent with Use Environment The identity of medical devices must be depicted in a manner consistent with the environment in which the device will be used. For example, the packaging devices that are used in
FIGURE 14.15
A sterilized device inside a sterilization pouch.
TAF-ER562X_WINGR-09-0401-C014.indd 610
11/9/10 12:35:55 AM
Packaging
611
emergency or field trauma situations should provide obvious identification information and should not require users to read the “fine print.”
14.3.6 CODING The concept of “coding,” that is, designating membership in a specific category of objects by one or more features of the object’s appearance, location, size, and so on, is important to medical device packaging (see also Chapter 13, “Signs, Symbols, and Markings”). Common coding dimensions in medical device packaging include shape and color (almost always used for pharmaceuticals in pill form), size, and labeling (including graphics and alphanumeric labels). A small but exemplary use of coding in device packaging is the use of a graphic symbol to indicate that the device has a finite shelf life and a “use-by” date (Figure 14.16). Guideline 14.31: More than One Package Coding Method The use of a single coding method, especially color, should be avoided. Packaging elements that are coded should always employ at least two coding dimensions (e.g., color and graphics).
Guideline 14.32: Less than Four Package Coding Methods Coding should be limited to a maximum of three levels—regardless of the coding dimension (except for labeling, which is described below).
Guideline 14.33: Package Coding Consistent with User’s Expectations Coding should consistent and compatible with the experience and expectations of the intended users.
Guideline 14.34: Avoid Shape Coding if Gloved Users Shape coding should not be used for packaging that will be used in an environment in which users will be wearing protective gloves since gloves reduce users’ ability to detect different shapes by touch unless demonstrated to be effective by usability testing.
Guideline 14.35: Test Package Coding Coding should undergo usability testing with intended users and in realistic usage scenarios.
FIGURE 14.16
Graphic that indicates that the device has a use-by date.
TAF-ER562X_WINGR-09-0401-C014.indd 611
11/9/10 12:35:56 AM
612
Handbook of Human Factors in Medical Device Design
14.3.7 LABELING Labeling can be an explicit form of package coding. In addition, labels for medical device packaging usually provides specific information about the use of a device, cautions and warnings, identification of the device and manufacturer, and so on (Kenagy and Stein, 2001). Labels can contain graphic elements, alphanumeric information, and color. Labels can also be functional with respect to device use (Stewart, 1997). For example, labels can contain pull-off tags that are intended to aid users by acting as reminders to take medication at certain times or on certain days. The most important aspect of labels from a usability perspective are legibility, understandability, and effectiveness (see Chapter 13, “Signs, Symbols, and Markings”). Labels must display important information so that users can easily see it. Figure 14.17 shows an interesting juxtaposition of identical packages. The bottom package is oriented so that the device name, expiration date, and package contents are visible. The top package is turned 90 degrees, and the label does not show either the use-by date or the fact that the package contains spinal needles, though the needle graphic is generally helpful. Labels on sterile packages need to show, in an obvious way, that they are sterile. Figures 14.18 and 14.19 show a label with the sterility indication depicted in very small type amid myriad other information. Figure 14.20, by contrast, shows a large “sterile” indication on a package label. Packaging labels need to be durable; they must survive typical storage, cleaning, sterilization, handling, and other usage activities. Labels that deteriorate during storage and usage could cause harm. Guideline 14.36: Readable Package Labels Alphanumeric and symbolic information on packaging labels should be large enough to be read by intended users at the range of distances appropriate for product-related tasks.
FIGURE 14.17 Two identical packages oriented differently. Not all critical information is shown on the top orientation.
TAF-ER562X_WINGR-09-0401-C014.indd 612
11/9/10 12:35:57 AM
Packaging
FIGURE 14.18
613
The label on a vascular catheter set has a very small indication of sterile status.
Alphanumeric information must, at a minimum, subtend 20 minutes of visual angle at the longest “typical” viewing distance.
Guideline 14.37: Package Label Achromatic Contrast Alphanumeric and symbolic information on labels should exhibit a minimum achromatic contrast modulation of 0.5, which corresponds to a minimum contrast ratio of 3:1. The recommended achromatic contrast modulation (between background and text) is 0.8 or higher, which corresponds to a contrast ratio of 9:1.
Guideline 14.38: Priority Package Label Information Label information should be prioritized by importance and arranged hierarchically on the label. The most important information should be the most salient on the label.
Guideline 14.39: Appropriate Package Label Terminology Label terminology should be appropriate for the intended user population. For example, medical jargon should not be used for labels aimed at lay users.
Guideline 14.40: Color Package Labels The use of colored backgrounds on labels should be avoided. This does not preclude the use of colored areas or symbols on labels, but those areas should not also contain alphanumeric information.
FIGURE 14.19
A close-up of the sterile indication on the label in Figure 14.18.
TAF-ER562X_WINGR-09-0401-C014.indd 613
11/9/10 12:35:58 AM
614
FIGURE 14.20
Handbook of Human Factors in Medical Device Design
A large sterile indication on a label.
Guideline 14.41: Package Label Color Contrast If colored backgrounds are used, then the background and foreground colors should provide effective achromatic and color contrast (see Chapter 13, “Signs, Symbols, and Markings,” and Chapter 8, “Visual Displays”).
Guideline 14.42: Package Label Usability Tests Users’ understanding of label information should be determined through formal tests with representative users.
14.3.8 DESIGN FOR PACKAGING Medical device packaging is often considered an element that is “layered” on a completed product design. In fact, packaging should not be an afterthought to device design. Just as any other device component, packaging requirements should be considered early and then throughout the entire design process (ANSI/AAMI HE-74, 2001). Medical practitioners agree that “device manufacturers should think of the package as part of the product” (Hunstiger, 1988). For example, if a device is likely to be assembled from separate pieces, then design elements that support this characteristic, including packaging and labeling, should be considered at the earliest possible design stage. Likewise, if it is clear that a device will have to undergo sterilization either prior to or between uses, then this requirement should be included in the device’s needs analysis documentation. Only certain packaging materials and designs will reliably support and survive common sterilization techniques, and this will influence other aspects of device design. Hultberg (1990) provides a very good discussion of the role that packaging should play in the overall design process. Guideline 14.43: Early Packaging Needs Analysis Early needs analysis and other user research should identify design requirements relevant to device packaging.
Guideline 14.44: Packaging Design Requirements Pertinent packaging considerations should be noted and described as part of the front-end design documentation.
TAF-ER562X_WINGR-09-0401-C014.indd 614
11/9/10 12:35:59 AM
Packaging
615
Guideline 14.45: Device-Packaging Compatability Device characteristics, such as color, text, and other visible features, should be matched to the visibility effects of likely packaging materials, considering the effects of handling, sterilization, prolonged shelf-life, and so on.
14.3.9 HANDLING Users must be able to grasp and manipulate packaged devices during the intended uses. The first concern, of course, is for users’ safety, especially regarding packaging-induced needle sticks, paper cuts, burns, exposure to toxic materials, and so on. Here again, there can be many substantive differences between lay and professional users. For example, professionals are more likely to be trained in the safe handling of sharps and biohazards. Some users may be allergic to certain materials commonly used in medical products. For example, since many people are allergic to latex, packaging that is composed of latex should be avoided, if possible, or clearly labeled if the use of latex is unavoidable. Finally, the form factor of the packaging should be consistent with the use environment and the capabilities of the intended users. For the former, a large, bulky package without appropriate grips would be unsuitable for a device used in a moving care environment (i.e., ambulance or helicopter). Similarly, a device intended for use by elderly, diabetic patients should not be packaged such that appreciable manual dexterity is required. Guideline 14.46: Safe Handling of Package Users should be able to safely handle a device, while it is still in its packaging without employing personal protective equipment (PPE).
Guideline 14.47: Indicate Risks of Handling The packaging should provide salient indications of any significant safety risks associated with handling the device both within and after it has been removed from its packaging.
Guideline 14.48: Indicate Device Handling Requirements The packaging of a device should inform users unambiguously that specific PPE will be required to handle the device once the package is opened.
Guideline 14.49: No PPE Requirements for Lay Users Lay users should not need to use PPE when handling a device either in or out of its packaging.
Guideline 14.50: Safe Handling in All Use Environments Packaging must support safe and practical handling under all conditions likely to be encountered in the device’s use environments. For example, packaging should not become slippery when exposed to fluids common in trauma or emergency care areas.
14.3.10 SHIPPING/MOVEMENT Medical devices are typically aggregated at distribution centers and shipped to their final usage location. The human factors issues associated with shipping relate to how devices
TAF-ER562X_WINGR-09-0401-C014.indd 615
11/9/10 12:35:59 AM
616
FIGURE 14.21
Handbook of Human Factors in Medical Device Design
Receiving dock near a hospital distribution area.
are packaged for shipping and how packaging supports tasks at the receiving end of the shipping chain. Large health care institutions are not significantly different than other large organizations with respect to purchasing in large quantities. Figure 14.21 shows a typical receiving dock at a large hospital, which is not substantially different than any large company’s shipping/receiving area. The guidelines related to shipping again draw a clear distinction between packaging for devices shipped to institutional users (e.g., hospitals, doctors’ offices) and those sent to (or destined for) residential lay users. Guideline 14.51: Design for Manual Handling Packaging for devices designed for handling by humans (as opposed to heavy-lift equipment) should be compatible with the anthropometry and strength capabilities of intended users.
Guideline 14.52: Packaging Supports Aggregation Devices that must be aggregated for shipment should have packaging that supports such aggregation.
Guideline 14.53: Packaging Supports Separation Devices that must be separated into smaller units after shipment should have packaging that supports such separation.
14.3.11 INVENTORY Medical devices are routinely shipped to distributors and to health care institutions in quantity so that they are often stored until they are used. The packaging for devices must accommodate and survive both shipping (see above) and storage (typically referred to as “inventory”). In large health care institutions, such as hospitals, the receiving and inventory processes for medical devices are assigned to a specific entity. Figure 14.22 shows a small part of the central distribution facility at a large hospital. This facility uses a location scheme that consists of naming each row of shelves sequentially with an alphabetic identifier and then labeling each section of shelving in that row and each individual area, or “bin,” in that section. In some locations, frequently used devices are aggregated into a smaller area of the overall storage facility. Two examples of this type of inventory scheme are shown in Figures 14.23 and 14.24. In Figure 14.23, various types of catheters and other frequently used items
TAF-ER562X_WINGR-09-0401-C014.indd 616
11/9/10 12:35:59 AM
Packaging
FIGURE 14.22
617
A view of a small part of a hospital distribution center.
are placed on a section of shelving that is apart from the standard storeroom shelves. In Figure 14.24, a movable cabinet is filled with packages of sterile towels. Figure 14.25 shows what might be termed an “open” or “bulk” storage area, in which individual items are not packaged separately. The ability of packaging to support various inventory requirements extends beyond human factors considerations. (Phillips, 1998). For example, packages must survive the expected shelf life of the product. An example of a human factors–specific consideration is the ability of users to identify devices while they are stored in inventory. Another design consideration is the ability of packaging to support inventory control procedures (i.e., knowing how many of each device is in stock at any time). Guideline 14.54: Packaging Supports Inventory Control Packaging should facilitate storing and retrieving devices to and from inventory. Examples of packaging characteristics that facilitate inventory tasks are stackable form factors, lifting indentations and handles, and device information placed on all surfaces of packaging.
Guideline 14.55: Easy Access When Stored Packaging should allow easy user access to individual devices even when multiple devices are stored in a location.
FIGURE 14.23
Part of a storage shelf for frequently used items.
TAF-ER562X_WINGR-09-0401-C014.indd 617
11/9/10 12:35:59 AM
618
FIGURE 14.24
Handbook of Human Factors in Medical Device Design
A movable cabinet filled with sterile towel packages.
Guideline 14.56: Easy Identification When Stored Users should be able to identify devices easily when they are stored in their inventory configuration.
Guideline 14.57: Critical Information Visible When Stored Use-by dates and other critical device information should be visible when multiple devices are stored in their inventory configuration.
FIGURE 14.25
An “open” or “bulk” storage area.
TAF-ER562X_WINGR-09-0401-C014.indd 618
11/9/10 12:36:00 AM
Packaging
619
Guideline 14.58: Intact Packaging During Storage Processes Placing devices into or out of storage should not damage or alter the device packaging in any way.
14.3.12 DISPOSAL What should users do with medical devices and packaging after the device has been used? In most cases, they dispose of the packaging. For single-use devices (e.g., syringes, catheters, bandages, and so on), the device itself is disposed after use. Proper disposal of medical devices may require replacing the device into its original packaging. In some cases, a disposal container might be part of or included in the original packaging. A number of disposal issues have human factors elements (Hunstiger, 1988; Phillips, 1998) including but not limited to type of disposal (e.g., recycle, toxic waste) and protection of users and others from harm (e.g. biohazard exposure or sharps injury). Appropriate disposal information, such as symbols, warnings, acronyms, prose, and so on (see below), should be placed on packaging (and devices) in such a way that it is easily seen and understood by users. Guideline 14.59: Indicate Device Disposal Methods Packaging and devices that require specific disposal methods, such as being placed in toxic waste containers, should provide a clear indication to that effect.
Guideline 14.60: Package as Disposal Method Packaging that is meant to be the disposal container for a medical device should be so identified.
Guideline 14.61: Indication of Readiness for Disposal Device packaging should provide obvious indications that it has or has not been used and is ready for disposal.
Guideline 14.62: Recyclable Packaging Indications Recyclable packaging should be marked as such using standard recycling symbols and wording (see Chapter 15, “Device Life Cycle”).
Guideline 14.63: Unambiguous Disposal Instructions Packaging and devices that should or should not be disposed in particular ways (e.g., “Do Not Recycle”) should be so labeled.
Guideline 14.64: Usability Testing of Disposal Instructions Users’ understanding of nonstandard or safety-critical disposal information should be determined through usability testing. An example of “standard” disposal information is the use of plastic recycling symbols and acronyms.
TAF-ER562X_WINGR-09-0401-C014.indd 619
11/9/10 12:36:01 AM
620
FIGURE 14.26
Handbook of Human Factors in Medical Device Design
The Easytab Hearing Aid Battery Delivery System.
14.4 CASE STUDIES 14.4.1 HEARING AID BATTERIES The Easytab Hearing Aid Battery Delivery System (Duracell, Bethel, CT) features a package design that facilitates the primary user task: retrieving a fresh battery and placing it in a hearing aid. The design developed from a decision to specifically address the needs of individuals 55 years of age and older who might have vision and manual dexterity impairments. As shown in Figure 14.26, an easy-to-open container holds eight fresh batteries. Normally, users might be challenged to pick up the tiny batteries by hand and insert them properly into their hearing aid. This task was simplified by mounting the batteries on removable, colored tags that are visually conspicuous (i.e., easy to find) and provide a large gripping surface for battery retrieval from the storage case and precise insertion. Once the battery is in place, the tab peals off. This simple but effective solution enables users who might have blurred vision and hand tremor to replace their own hearing aid batteries (see http://www.duracell. com/products/hearingaid.asp).
FIGURE 14.27
Packaging for self-etching dental adhesive.
TAF-ER562X_WINGR-09-0401-C014.indd 620
11/9/10 12:36:01 AM
Packaging
621
14.4.2 MIXING A DENTAL ADHESIVE Adper™ Prompt™ L-Pop™ Self-Etch Adhesive (3M, St. Paul, MN) facilitates the mixing and application of a self-etching dental adhesive. The user’s first task is to mix the adhesive components that are initially contained in separate chambers. To do so, the user squeezes on the distal yellow chamber, which combines the reactive components in the red chamber, incorporating an indentation that pops up to indicate that the material has properly transferred between chambers. Next, the user spins or churns the combined components with the applicator stick to ensure complete mixing. Finally, the user withdraws the stick from the package to apply adhesive directly to a patient’s tooth. The adhesive subsequently performs the necessary steps of etching, priming, and bonding. The package design stores the adhesive, facilitates mixing, and provides the means of application in one simple device rather than requiring a more complicated kit. The result is faster dental procedures and reduced workload for the dental professional.
RESOURCES American National Standards Institute/Association for the Advancement of Medical Instrumentation. (1993). Human Factors Engineering Guidelines and Preferred Practices for the Design Of Medical Devices. ANSI/AAMI HE-48-1993. Arlington, VA: Association for the Advancement of Medical Instrumentation. Needham, A. D., Natha, S., and Kaye, S. (2001). Similarities in the packaging of cyanoacrylate nail glue and ophthalmic solutions: An ongoing problem. British Journal of Ophthalmology, 86, 496–497. Pilchik, R. (2002). Validating Medical Packaging. Boca Raton, FL: CRC Press. Sherman, M. (Ed.). (1998). Medical Device Packaging Handbook. New York: Marcel Dekker. TAPPI. (1999). 1995–1999 TAPPI Polymers, Lamination, and Coatings Conference. Norcross, GA: TAPPI Press.
REFERENCES Ambrose, P. J., Saya, F. G., Lovett, L. T., Tan, S., Adams, D. W., and Shane, R. (2002). Evaluating the accuracy of technicians and pharmacists in checking unit dose medication cassettes. American Journal of Health-System Pharmacists, 59(12), 1183–1188. American National Standards Institute/Association for the Advancement of Medical Instrumentation. (2001). Human Factors Design Process for Medical Devices. ANSI/AAMI HE-74-2001. Arlington, VA: Association for the Advancement of Medical Instrumentation. Berkman, N. D., DeWalt, D. A., Pignone, M. P., Sheridan, S. L., Lohr, K. N., Lux, L., Sutton, S. F., Swinson, T., and Bonito, A. J. (2004) Literacy and Health Outcomes: Evidence Report/ Technology Assessment No. 87. AHRQ Publication No. 04-E007-2. Rockville, MD: Agency for Healthcare Research and Quality. Consumer Product Safety Commission. (n.d.). Child Resistant Packaging Saves Lives. Document #5019. Washington, DC: Consumer Product Safety Commission. Cushman, W. H., and Rosenberg, D. J. (1991). Human Factors in Product Design. New York: Elsevier. DeRespinis, P. A. (1990). Cyanoacrylate nail glue mistaken for eye drops. Journal of the American Medical Association, 263, 2301. Frieden, J. (2002). Drugmakers, hospitals give FDA earful about bar-coding pros, cons. Drug Topics, 16, 27. Hankin, R. A. (2002). Bar coding in healthcare—A critical solution. In Business Briefing: Medical Device Manufacturing and Technology. London: Touch Briefings.
TAF-ER562X_WINGR-09-0401-C014.indd 621
11/9/10 12:36:02 AM
622
Handbook of Human Factors in Medical Device Design
Hultberg, W. M. (1990). Package design of medical devices. In Proceedings of the TAPPI Polymers, Laminations, and Coatings Conference, 719–722. Norcross, GA: TAPPI (Technical Association of the Pulp and Paper Industry). Hunstiger, C. (1988, December). Life cycle of packaging materials in health care institutions. TAPPI Journal, 71, 130–132. Kendrick, R., and Bayne, J. R. (1982). Compliance with prescribed medication by elderly patients. Canadian Medical Association Journal, 127(10), 961–962. Landro, L. (2004, January 15). Medication errors can occur outside the hospital. Wall Street Journal. Leventon, W. (2001, January). Adapting packaging technology to meet device industry needs. Medical Device and Diagnostic Industry, 23(1), 114. Leventon, W. (2002, September). Medical device sterilization: What manufacturers need to know. Medical Device and Diagnostic Industry, 24(9), 52. Lipowski, E. E., Campbell, D. E., Brushwood, D. B., and Wilson, D. (2002). Time savings associated with dispensing unit-of-use packages. Journal of the American Pharmaceutical Association, 42(4), 577–581. Phillips, P. (1998). Storage of Medical Disposables and Dressings. Bridgend: Medical Disposables Resource Centre. Rohles, F. H., Jr., Moldrup, K. L., and Laviana, J. E. (1983). Opening jars: An anthropometric study of the wrist-twisting strength of the elderly. In Proceedings of the 27th Annual Meeting of the Human Factors Society, 112–116. Santa Monica, CA: Human Factors and Ergonomics Society. Rossi, P. (2003, June). Innovations in pharmaceutical packaging. Packaging and Bottling International, 72–75. Sawyer, D. (1996). Do it by Design: An Introduction to Human Factors in Medical Devices. Washington, DC: U.S. Food and Drug Administration. Stewart, I. (1997). “Single use only” labeling of medical devices: Always essential or sometimes spurious? Medical Journal of Australia, 167, 538–539. Wagner, D. (2002). How to use medical devices safely. AORN Journal, 76(6), 1059–1061. Wiklund, M. E. (1995). Medical Device Equipment Design: Usability Engineering and Ergonomics. Buffalo Grove, IL: Interpharm Press.
TAF-ER562X_WINGR-09-0401-C014.indd 622
11/9/10 12:36:02 AM
15 Device Life Cycle Michael E. Maddox, PhD, CHFP; Larry W. Avery, MA CONTENTS 15.1 15.2 15.3 15.4
The Medical Life Cycle .........................................................................................623 Life Cycle and Errors .............................................................................................625 Different Types of Users ........................................................................................625 General Principles ..................................................................................................626 15.4.1 Use in the Home or in a Health Care Facility ...........................................626 15.4.2 Use by Lay Users or by Experts................................................................626 15.5 Special Considerations ...........................................................................................627 15.5.1 Contact with Biohazards ...........................................................................627 15.5.2 Use Environments .....................................................................................627 15.5.3 Device Longevity ......................................................................................627 15.5.4 Introduction of New Technology ..............................................................627 15.5.5 Ruggedness ...............................................................................................628 15.5.6 Pull Demand for Devices ..........................................................................628 15.6 Design Guidelines ..................................................................................................628 15.6.1 Expendables ..............................................................................................629 15.6.2 Single-Use Devices ...................................................................................630 15.6.3 Storage ......................................................................................................631 15.6.4 Setup and Use ...........................................................................................631 15.6.5 Training ....................................................................................................632 15.6.6 Routine Maintenance ................................................................................634 15.6.7 Repair........................................................................................................635 15.6.8 Upgrades ...................................................................................................638 15.6.9 Disposal ....................................................................................................639 15.7 Case Study ............................................................................................................ 640 Resources .........................................................................................................................643 References ........................................................................................................................643
15.1 THE MEDICAL LIFE CYCLE The term “life cycle,” when applied to a device, typically means all facets of the device’s existence, from the time it is conceived until the time it is discarded. The term “cradle to grave” is often used when referring to a device’s life cycle. A device’s life cycle is sometimes depicted as a circle in which the demise of one product version is followed by the incarnation of a new version. Experience in the marketplace can feed into the development 623
TAF-ER562X_WINGR-09-0401-C015.indd 623
11/9/10 1:38:34 AM
624
Handbook of Human Factors in Medical Device Design
cycle for follow-on products, but that requires processes that do not affect the current device in the market. Figure 15.1 depicts the typical life cycle for a medical device. It includes three nominal phases: “Design and Development,” sometimes called the “front end” of the life cycle; “Fielding and Implementation,” sometimes called the “middle” of the life cycle; and “End of Life,” also referred to as the “back end” of the life cycle. The components of these phases vary depending on the nature of the specific device, but, in general, the terminology will be appropriate. From a human factors perspective, all life cycle requirements of a medical device must be addressed during the “Design and Development” phase. That is, regardless of when, in its life cycle, a user interacts with a device, the relevant characteristic must be identified, designed, and put into place as the device is being designed and developed. Certainly, there are rare exceptions to this assertion. For example, recycling stickers can be added to a device after it is placed in the use environment. However, such “afterthought” actions tend to be expensive, logistically more difficult, and prone to noncompliance. The design aspects of medical devices, including the critical aspects of front-end analysis, function allocation, conceptual design, prototyping, iterative usability testing, hardware and software development, production, and rollout, are addressed in HE-74-2001, HE-752010, and other chapters of the book as well as in other publications, (e.g., Cushman and Rosenberg, 1991). In this chapter, we identify and describe the specific life cycle characteristics that should be included as components of any medical device. This chapter includes a detailed discussion of the differences between device users who are medical professionals versus lay users (i.e., patients and other nonprofessional users
Strategic marketingidentify product features
User and device requirements analysis
Conceptual design
Development and testing
Design and development
Production
Device use, maintenance, and upgrade
Rollout
Fielding and implementation
Disposal
Obsolescence - No further support
End of life
FIGURE 15.1
Integration of device design cycle and market device life cycle.
TAF-ER562X_WINGR-09-0401-C015.indd 624
11/9/10 1:38:34 AM
Device Life Cycle
625
who may not be well versed in medical methods and terminology). The goal is to ensure that these end-user differences are considered during the device specification, design, and development process. Guidance related to how this issue is considered is not within the scope of the chapter but can be found in Chapter 18, “Home Health Care,” as well as other documents that describe how human factors considerations and methods should be incorporated into the device design process.
15.2 LIFE CYCLE AND ERRORS There is general agreement that medical errors exact a huge toll on patients, caregivers, and health care institutions, both financially and in terms of pain and suffering (Kohn et al., 2000). The typical range of estimates for the contribution of human error to serious medical incidents is between 50% and 80%, similar to the incidence of errors in nonmedical domains (Proctor and Van Zandt, 1994; Senders and Moray, 1991). For example, in a study of adverse outcomes from blood transfusions in Britain, 61% were found to be due to the erroneous transfusion of the wrong blood type (Love et al., 2002). We now know that a majority of medical device errors are due to design flaws. However, when device-related incidents involving injuries, death, and property damage are attributed to “human errors,” these accounts are referring to device users, not device designers. Even for incidents in which poor design plays a prominent role, such as the now-famous Therac accident described by Casey (1998) in his well-known book Set Phasers on Stun, the accident description focuses on errors by users of the device, not by designers. Thus, the typical claims of human error contributions of 50% to 80% do not include design errors. If we consider human error in terms of the medical device life cycle and we exclude errors by designers and manufacturers, then most errors occur in the middle or back end of the device life cycle. That is, most use errors that result in adverse consequences occur when a device is being used and maintained or when users improperly dispose (or fail to dispose) of the device. Obviously, the place to address design-related errors is in the design process. However, design-related errors are “latent” in the taxonomy of Reason (1990). They do not actually cause damage until an unfortunate user encounters the design error in just the right set of circumstances to precipitate an incident.
15.3 DIFFERENT TYPES OF USERS Of particular importance is the distinction among devices aimed at professional versus lay users. In addition, some guidelines in this chapter will distinguish among devices that are intended to be used in a professional use environment, such as a hospital or professional medical practice, and those intended to be used in private homes or similar noninstitutional settings. For example, Ison and Miller (2000) make a persuasive argument that institutional purchasing decisions should be driven by the life cycle costs associated with medical devices rather than their purchase price because the purchase price can be a small fraction of the total life cycle costs. In summary, the emphasis in this chapter is on the easy and safe use of medical devices over their entire expected life. A device that is effective to use inherently promotes proper use and reduces the likelihood of use errors. In addition, device use must not expose patients, health care professionals, or the general public to unacceptable risks.
TAF-ER562X_WINGR-09-0401-C015.indd 625
11/9/10 1:38:34 AM
626
Handbook of Human Factors in Medical Device Design
15.4 GENERAL PRINCIPLES 15.4.1 USE IN THE HOME OR IN A HEALTH CARE FACILITY From a human factors perspective, using a medical device in the home environment is likely to be different, at least in some respects, from using the same or a similar device in a health care environment, such as a clinic or hospital. The most obvious differences are the education, training, and experience of the end users in each of these settings (see the next section). However, there are many other, less obvious differences that can affect the design, storage, use, maintenance, disposal, and other aspects of a device’s life cycle. Table 15.1 describes the primary differences between typical home and professional device use environments. Another obvious difference between these environments is that users in institutional environments, including most professional health care facilities, have medical device resources at their disposal that are not available in the home environment. An example is the availability of trained technicians who can diagnose device problems and make repairs.
15.4.2 USE BY LAY USERS OR BY EXPERTS As noted above, the expertise and education of likely users is the most obvious difference between the home market for medical devices and the institutional market. There can also be significant differences in fundamental physical, perceptual, and cognitive abilities between these targeted user groups. For example, lay users of some medical devices are likely to be older and exhibit more manipulative, perceptual, and cognitive limitations than professional health care workers. Lay and professional users will have vastly different training and experience related to the terminology and procedures typically associated with medical devices. The device should meet the needs of the primary user population. When designing a device for both user populations, meeting the needs of lay users should generally take precedence. Lay users may not understand medical terminology related to
TABLE 15.1 Differences between Institutional and Home-Use Settings Health Care Institution
Home Setting
Closely controlled environmental conditions, including lighting, temperature, noise, and so on Established procedures to ensure safe device usage
Some environmental control but typically varies more than in institutional setting No established procedures for device usage; many procedures ad hoc No established procedures for handling and disposal of biohazards
Established procedures to ensure safe handling and disposal of biohazards and other hazardous materials (sharps, chemicals, batteries, and so on) Formal training in the use of medical devices Informal support network of individuals who are experienced in the use of specific medical devices
TAF-ER562X_WINGR-09-0401-C015.indd 626
Informal, ad hoc, or no training on the use of medical devices Possible use of manufacturer’s customer service and technical support, but other experienced users not usually available
11/9/10 1:38:34 AM
Device Life Cycle
627
the setup, use, or maintenance of medical devices. Neither will they exercise the same degree of medical device monitoring as do health care professionals.
15.5 SPECIAL CONSIDERATIONS Regardless of whether we are discussing electronic medical devices, pharmaceuticals, diagnostic and treatment tools, dressings, or other medical devices, their life cycle components are similar to those for consumer electronics, garden tools, clothing, and most any other device destined for particular markets. However, several aspects of medical devices (and their use) set them apart from other device categories.
15.5.1 CONTACT WITH BIOHAZARDS Many medical devices are, by their nature, likely to come into contact with biohazardous materials, thereby having the inherent capability to cause injury or death. Common examples are disposable dressings and syringes.
15.5.2 USE ENVIRONMENTS Professionals tend to interact with medical devices either in institutional settings, such as hospitals and nursing homes, or in ambulatory care settings, such as physician offices. Lay users typically use medical devices in their homes or in other private settings, such as hotel rooms and job sites. The implications for particular life cycle components, such as training and maintenance, will be quite different, depending on the type of user and the use environment.
15.5.3 DEVICE LONGEVITY Many durable medical devices, such as infusion pumps and surgical tables, have long life spans. For example, a portable X-ray machine might be used clinically for 20 years or more. Others, such as syringes and catheters, are meant to be used once and then discarded. The longevity of certain devices has implications for the support that manufacturers must provide. It has been estimated that manufacturers must support electronic medical devices for at least seven years after the last unit has been shipped (Newman, 2003).
15.5.4 INTRODUCTION OF NEW TECHNOLOGY The longevity of certain medical devices stems, at least in part, from the conservative nature of medical professionals. That is, once a professional learns to use a particular device, he or she is loathe to switch to a device that embodies newer technology, especially if the current device functions well. Balancing that tendency is the relatively recent trend of moving new technology quickly into what are known as “low-acuity” settings. Technology that used to take 20 years to move from teaching hospitals to physician offices is now being pushed outward much more quickly, in some cases in three to five years (Myers and Burchill, 2002). That being said, a durable medical device that starts out in a tertiary care hospital for its first five years may still have a 20-year life span as it moves to a rural hospital, then to an outpatient clinic, and finally to a physician’s office.
TAF-ER562X_WINGR-09-0401-C015.indd 627
11/9/10 1:38:34 AM
628
Handbook of Human Factors in Medical Device Design
15.5.5 RUGGEDNESS Medical devices must be rugged enough to survive (and perform reliably) in the rather harsh use environments common to health care users. Among the design requirements for infusion pumps, for example, is that they must survive being dropped, bumped into obstacles, and so on. Medical devices share this design requirement with other “heavy-duty” commercial devices, such as commercial-grade fitness equipment.
15.5.6 PULL DEMAND FOR DEVICES Another life cycle issue that has only recently pertained to medical devices is the ability of manufacturers to appeal directly to consumers. This type of direct appeal has become quite common for prescription medications, such as Viagra and various allergy and bloodpressure control drugs. A medical device example of such direct marketing is television advertising for glucometers. By using the direct marketing approach, device manufacturers are specifically aiming their devices at consumers instead of health care professionals. Such direct appeals allow device vendors to “pull” demand via patients’ direct purchases and requests to physicians, whereas past marketing efforts were directed at “pushing” physician demand through sales calls and targeted advertising in professional journals. Appealing directly to the consuming public places a greater potential usability design, training, and device support burden on vendors. Now, instead of being able to rely on the prior knowledge, training, and experience of medical professionals, device vendors must account for the much greater range of skills, knowledge, and experience of the general consumer population.
15.6 DESIGN GUIDELINES These design guidelines are presented in the approximate order in which they will be encountered in the device life cycle (see Figure 15.2). ry
e rag Sto
live De
ta Ins
io llat
n
e anc
ade
en int Ma
De
eu vic
gr Up
se
le bso
ce
sen
O re p
air al
p os
Dis
FIGURE 15.2 Steps in the medical device life cycle, from upper left to lower right. (Courtesy of Wiklund R&D.)
TAF-ER562X_WINGR-09-0401-C015.indd 628
11/9/10 1:38:34 AM
Device Life Cycle
629
15.6.1 EXPENDABLES Many medical devices use or contain elements that must be periodically refilled, replaced, or otherwise replenished. A common term for these items (such as ink cartridges, batteries, test strips, reagents, and so on) is “expendables.” The human factors design issues with expendables are that they are recognizable as such, that it is obvious when they should be replaced, that they are easy to remove and replace, and that the tasks required for safe and proper disposal of exhausted components are easy and obvious. These issues acquire added importance for expendables such as needles, tubing coverings, dressings, and so on that might be sharp, contaminated, or toxic or that might expose users to other risks. Also, the risk of improper interchangeability of expendables can have serious implications. For example, suppose that an analysis device requires various types of reagents to be replenished. If all the reagents are in containers with the same form factor and connectors, it might be possible to install the wrong reagent in the wrong location, leading to device malfunction or false analytical results. Guideline 15.1: Identity of Expendables The fact that an item is expendable should be clear and obvious to users. While the single-use characteristic of glucometer test strips may be obvious, it may be unclear whether batteries are rechargeable.
Guideline 15.2: Ready for Replacement Indication Devices should indicate to users when expendable items need to be replaced. An example is using clear containers for cleansing and mixing fluids in blood test equipment to allow users to tell at a glance when containers need to be replaced.
Guideline 15.3: Improper Installation of Expendable The design of devices should physically preclude the improper installation of expendables, especially in such a way that might endanger users or patients. For example, it should not be possible to install single use alkaline batteries when rechargeable lithium ion batteries are required. An alternative but less desirable option is to block improper installation by rendering the device inoperable if expendables are installed incorrectly.
Guideline 15.4: Automatic “Safe” Mode if Wrong Expendable If it is possible to install inappropriate expendables in a device, then the device should automatically switch into a “safe” mode that will prevent harm to users, patients, or the device. If single use instead of rechargeable batteries can be installed, then sensing circuitry should recognize this use error and not allow recharging.
Guideline 15.5: “All OK” Indication When Correct Expendable The design of devices and expendables should clearly indicate when new expendables are properly matched with the device. For example, containers of reagents and cleaning fluids can be made to conform to unique shapes and colors. When the proper containers are inserted in the proper receptacles, the colors and shapes show unambiguously a complete and proper layout.
TAF-ER562X_WINGR-09-0401-C015.indd 629
11/9/10 1:38:34 AM
630
Handbook of Human Factors in Medical Device Design
Guideline 15.6: Easy Access to Expendables Users should have easy access to expendables and their receptacles; that is, they should not have to partially or completely disassemble the device to replace expendables.
Guideline 15.7: Tools for Expendable Replacement Replacing expendables should not require the use of special tools or fixtures. If special tools are required, then they should be supplied and packaged with the replacement expendables.
Guideline 15.8: Mitigate Exposure to Hazards Device users should not be exposed to hazards such as electrical shock or toxic materials when replacing expendables.
Guideline 15.9: Avoiding Risks of Expended Items To the extent possible, users should be automatically protected from the risks posed by expended items. For example, items contaminated with blood may be automatically routed to a sealed container that can be removed and replaced without breaking the seal.
Guideline 15.10: Proper Method of Expendable Disposal The disposal method for expendables should be obvious and should protect users from risk of injury, infection, and so on. If containers, wrappers, or other items are required for disposal, they should be easy to obtain and use.
15.6.2 SINGLE-USE DEVICES Some medical devices or components of those devices are intended to be used only once and then disposed of. An example of a single-use device is a disposable hypodermic needle. Other devices, such as infusion pumps, ECG machines, and so on, are meant to be used repeatedly. The design requirements for single-use devices differ in significant ways from those of multiuse or durable devices. Even when it is obvious that a device is meant to be used only once, certain characteristics of the device (mainly its cost) may prompt some users to clean and reuse it (Health Industry Manufacturers Association, 1999; Stewart, 1997). This situation might pertain for several reasons. The device really might be capable of several uses, but the manufacturer has labeled it as single use to increase sales (Horwitz, 2002) or because it failed to qualify for or obtain regulatory approval for multiple uses. In certain cases, the device really might be a single-use device, but its design, cost, and functionality allow (or even encourage) reuse. The international “single-use” symbol is shown in Figure 15.3 (EN 980, 2003; ISO 15223, 2000). Guideline 15.11: Obvious Ready for Use A device’s readiness for use should be obvious.
Guideline 15.12: Number of Allowed Uses It should be obvious to users whether a device is meant to be used only once or many times. For example, a single-use catheter should be clearly labeled as such both on its packaging and on the device itself.
TAF-ER562X_WINGR-09-0401-C015.indd 630
11/9/10 1:38:34 AM
Device Life Cycle
FIGURE 15.3
631
International symbol indicating a device intended for single use.
Guideline 15.13: Disposition of Single-Use Device For single-use devices, it should be obvious to users what to do with the device after it has been used. In the catheter example above, this might be accomplished by placing “toxic waste” graphics on the device and on its packaging.
Guideline 15.14: Status of Single-Use Device The “used/not used” status of a single-use device should be obvious to users.
Guideline 15.15: Unfit for Reuse For single-use devices, when feasible, their first use should render them functionally and visually unfit for reuse.
15.6.3 STORAGE Devices, their components, and expendables should be designed to facilitate storage prior to and between uses. See Chapter 14, “Packaging.”
15.6.4 SETUP AND USE The issue of setting up and using a device is central to any discussion of medical device human factors. This section is not about monetary life cycle costs. However, human factors professional experience is that well-designed devices will minimize costs associated with installation, training, rework, support, service, etc. Moreover, good human factors design will minimize use errors with their associated risks and liability. The applicability of guidance related to this topic might vary, depending on the expertise of the users and the nature of the use environment. Guideline 15.16: Out-of-Box Ready for Use Ideally, devices should be ready for use when they are removed from their packaging. If setup activities are required, they should be brief and simple.
Guideline 15.17: Minimal Setup Whenever feasible, devices intended for lay users should require minimal assembly or setup prior to use.
TAF-ER562X_WINGR-09-0401-C015.indd 631
11/9/10 1:38:35 AM
632
Handbook of Human Factors in Medical Device Design
Guideline 15.18: Intuitive Setup Tasks The need for setup or initial tasks should be obvious. Moreover, the tasks required should be easily performed by intended users. For example, the steps to assemble a device, if assembly is required, should be obvious and straightforward.
Guideline 15.19: Setup Consistent with User Abilities Setting up and using a device should be within the physical, perceptual, and cognitive capabilities of the intended users. Unfortunately, this guideline has not been observed, for example, with respect to some rheumatoid arthritis (RA) drugs intended for patient self-administration. The steps required to open the package, assemble the syringe, mix the drug, draw it into the syringe, and administer the injection were often beyond the physical capabilities of many RA patients or their caregivers.
Guideline 15.20: No Special Tools Required Setting up and using a device should not require special tools (e.g., a jig for holding a hearing aid in place so a battery can be inserted easily) fixtures, or other equipment. If special tools are required, they should be included with the device and be stored on or attached to the device to minimize the risk of their being misplaced.
Guideline 15.21: “Ready for Use” Indication The device should indicate unambiguously that it is (or is not) ready for use. For example, home defibrillators indicate unambiguously when they are ready to use.
Guideline 15.22: Include All Necessary Parts All parts required for setup should be attached to or packed with the device.
15.6.5 TRAINING A comprehensive treatment of the human factors components of training is beyond the scope of this chapter. However, when viewed as a life cycle issue, there are a few aspects of training that are especially relevant to medical devices. One of these issues is the necessity and appropriateness of training. For certain types of medical devices, such as anesthesia machines, significant training is entirely appropriate. For other types of devices, significant training is neither appropriate nor allowable in certain use scenarios. For example, portable defibrillators are now a common fixture in airports and other public locations. They are expected to be able to be used by individuals with no prior training and with only minimal on-device instructions during use. For home use devices, part or all of the user population often will consist of older adults who have cognitive and perceptual impairments, or who are experiencing symptoms of chronic or acute illness. Certain nonobvious effects of these conditions, such as comprehension and memory decrements, make even plainspoken directions problematic (Park and Skurnik, 2004). In addition, training often tends to be second- or even third-hand, routinely passed from the original user to other people who interact with the device. While the original training, often provided by the manufacturer or its representatives, may be exemplary, the second-hand training (i.e., from one user to another) is likely to be inconsistent or incomplete.
TAF-ER562X_WINGR-09-0401-C015.indd 632
11/9/10 1:38:35 AM
Device Life Cycle
FIGURE 15.4
633
Homemade warning sign advising users to avoid turning off a device.
Users (and maintainers) will create their own training aids if a particular device does not provide sufficient cues regarding its proper use. Figure 15.4 illustrates how users have developed their own training aid to prevent a device’s rechargeable battery from being accidentally exhausted by powering down the device via the small main power switch located on the rear panel. To understand why users might switch the power off using this back-panel switch, note the location and size of the manufacturer’s “preferred” power switch, which is shown in Figure 15.5. This switch is “hidden in plain view”; that is, it appears to be one of the control keys on the keyboard and often escapes notice when users wish to turn off the device. Another training supplement, in this case added by the hospital’s engineering department, is shown in Figure 15.6. If the device is unplugged when it is turned off, the internal battery will not recharge. This often results in user complaints that the device is “broken” when all it needs is to be plugged into an AC power outlet. Such user behavior should have been anticipated and mitigated during design.
FIGURE 15.5
The “preferred” power switch.
FIGURE 15.6
Reminder to leave the device plugged in.
TAF-ER562X_WINGR-09-0401-C015.indd 633
11/9/10 1:38:35 AM
634
Handbook of Human Factors in Medical Device Design
Guideline 15.23: Training Fit To User Capabilities Training should be designed based on the knowledge and skills required of targeted users. Conversely, training designers should identify the knowledge and skills that are not required for a user to benefit from the training materials or program.
Guideline 15.24: Minimize Formal Training Requirements Minimize formal training requirements for devices intended for lay users. For professional users, such as anesthesiologists using inherently complex devices, formal training may be necessary. However, the training should be designed with a clear appreciation of professionals’ inherent time constraints and competing demands.
Guideline 15.25: On-Device Training Materials To the extent possible, limit training to material that can be presented within the form factor and immediately associated labeling (including quick reference cards) of the device.
Guideline 15.26: Take Advantage of Users’ Prior Experience Device designers should consider and be consistent with the user-interface attributes and functional requirements of other similar or analogous devices with which intended users are likely to be familiar. The resulting “positive transfer training” will minimize training requirements (and use errors) with the new device.
Guideline 15.27: Embed Training within Software For devices that primarily rely on a computer interface, embed additional training within the software user interface.
Guideline 15.28: Written Materials Write all user manuals, instructions, and other training and support documents at a level compatible with the minimum training, experience, and reading skills of the target user population. For population with a wide range of existing training, experience, and reading skills, more than one version of written materials might be required. Alternately, a single set of training materials could be targeted at the least-skilled user subpopulation. See Chapter 5, “Documentation”.
15.6.6 ROUTINE MAINTENANCE Routine maintenance (i.e., device maintenance tasks that are known and planned in advance), includes calibration, cleaning, inspection for worn parts, and other tasks that are not initiated as the result of device failures. Human factors guidance related to routine maintenance concerns both the design of the device (design for maintainability), which is covered elsewhere, and the mechanics of maintenance. Guidance also relates to the resources required during maintenance and the training and availability of maintainers. The main points of departure for maintenance guidelines are the location of the device and the type of end user. Guideline 15.29: Minimize Nonprofessional Maintenance Routine maintenance requirements for nonprofessional maintainers—that is, individuals whose jobs are not primarily maintenance-oriented—should be limited to the replacement of expendables. Nonprofessional maintainers include both frontline clinicians and lay users.
TAF-ER562X_WINGR-09-0401-C015.indd 634
11/9/10 1:38:35 AM
Device Life Cycle
635
Guideline 15.30: Need for Routine Maintenance Medical device user interfaces should indicate obviously and unambiguously when routine maintenance is needed.
Guideline 15.31: Design for Maintenance Device design should facilitate routine maintenance tasks. For example, some dialysis machines have tilt-out panels that allow access to internal valves and tubing. Clearances should be adequate to allow maintainers to reach, grasp, adjust, remove, and replace internal device components.
Guideline 15.32: Special Maintenance Tools Not Required Routine maintenance tasks should not require special tools, fixtures, or test equipment. For example, a battery access door should be able to be opened without tools and should include voltage test points onto which common voltmeter probes can be placed to confirm battery charge.
Guideline 15.33: Minimize Risks of Maintenance Tasks Maintenance tasks should not expose maintainers to hazards, such as electrical shock, pinch points, hazardous chemicals, and so on. To the extent that such exposure is unavoidable, maintenance procedures should explicitly account for these hazards, and maintainers should be trained and warned to take adequate precautions to avoid the hazards.
Guideline 15.34: Minimum Out of Service Time Routine maintenance tasks should remove the device from service for the minimum amount of time necessary.
Guideline 15.35: Automatic Device Calibration Whenever possible, routine device calibration should be automatic. If manually performed, calibration should be treated as a routine maintenance task.
15.6.7 REPAIR Repairs, sometimes known as “nonroutine maintenance tasks,” are required when medical devices break. In certain domains, such as military aviation, routine and nonroutine maintenance costs are major contributors to the life cycle costs of equipment. The following guidelines focus on knowing when a device needs to be repaired, being able to diagnose and repair it correctly, and knowing when a repair has been completed successfully, that is, that a device has been properly returned to service. A frequently overlooked complement to ease of repair is designing medical devices so that they are sufficiently reliable and durable to not often require repair. One of the primary deterrents to easy repair is obtaining access to the internal parts that must be observed, tested, manipulated, and replaced. Two common methods of providing maintenance access are to allow sections of a device to be tilted out and to include removable panels that expose major replaceable or repairable components. Figure 15.7 shows a typical tilt-out panel that provides fairly good access to internal wiring and parts. Figure 15.8 shows the back of a device that contains an internal, rechargeable battery. Many
TAF-ER562X_WINGR-09-0401-C015.indd 635
11/9/10 1:38:35 AM
636
Handbook of Human Factors in Medical Device Design
FIGURE 15.7
Tilt-out access panel.
FIGURE 15.8
Removable rear panel through which the battery is accessed.
medical devices, especially those used in hospitals, contain rechargeable batteries that are a major maintenance/repair component. To access the battery in the depicted device, the lower half of the back cover plate must be removed. Some medical devices, because of their inherent properties, present difficult maintenance challenges. For example, portable X-ray devices tend to be quite heavy and bulky, as seen in Figure 15.9. Note that this portable X-ray device is suspended from the shop ceiling by a chain hoist that has been fastened to the steel superstructure of the building (above the suspended ceiling). Also note the open access panel on the side of the device. Both of these design features represent reasonable approaches to conducting various types of repairs on such heavy devices. Guideline 15.36: Need for Repair The need for repair should be obvious to the user. A worst-case scenario for a medical device is for it to look as though it is functioning properly when it is not (Harris et al., 2005). For example, the continuous self-test of the overpressure sensor in an infusion pump is an internal component that is not easily observable. If the internal test subsystem detects an error in this component, it should shut down the machine and issues a clearly worded repair message.
TAF-ER562X_WINGR-09-0401-C015.indd 636
11/9/10 1:38:35 AM
Device Life Cycle
FIGURE 15.9
637
Heavy device undergoing maintenance.
Guideline 15.37: Self-Monitoring of Device Function When technically feasible, devices should monitor themselves for functional integrity, such as with the use of internal self-tests of critical components.
Guideline 15.38: Protection from Device Damage Devices should be designed to minimize the risk of damage during normal use, including when stored, shipped, handled, installed, operated, or maintained. Susceptibility of components to damage should be clearly communicated to the user. For example, devices that include delicate moving parts are often labeled with instructions to lock these internal parts before moving the device. Procedural guidance and suitable warning labels should be provided to help mitigate such damage.
Guideline 15.39: Accessibility of Components Device components with the highest failure rates should be easily accessible.
Guideline 15.40: Modularity of Device Components Heavy, large, or complex devices should be designed in modules so that individual modules can be removed, replaced, and repaired without affecting other modules. For example, modern dialysis machines are designed to facilitate cleaning, calibration, and repair (see the case study in Section 15.7).
Guideline 15.41: Minimize Time to Repair The design should minimize the overall time required for repairs.
TAF-ER562X_WINGR-09-0401-C015.indd 637
11/9/10 1:38:36 AM
638
Handbook of Human Factors in Medical Device Design
Guideline 15.42: Only Common Tools for Repair Repairs should require only common tools and test equipment.
Guideline 15.43: Minimize Special Repair Tools Requirements for the use of special tools, fixtures, and test equipment should be minimized.
Guideline 15.44: Minimize Hazardous Repair Conditions Devices should be designed to minimize exposure to hazardous conditions during repair tasks. For example, devices should be designed with power interlocks that remove all AC power when any access panel is removed.
Guideline 15.45: Repair Locations Easily Accessible Authorized repair locations and/or contact information should be easily available to maintainers.
Guideline 15.46: Lay Users Are Not Repairers Lay users should not be expected or encouraged (by the device design or otherwise) to perform any repairs on a medical device.
Guideline 15.47: Unauthorized Repair or Access Not withstanding previous guidelines, if unauthorized access to device components could pose a serious hazard, design should prevent or at least actively discourage such access.
15.6.8 UPGRADES For durable devices (especially electromechanical ones), designers must consider the effects of potential future improvements in hardware or software. An upgrade consists of the addition or replacement of, or any change to, a hardware, software, or procedural component of a device. Note that the following guidelines also generally apply to manufacturer-initiated device changes intended to correct design problems, bugs, and recurrent errors. In the best case scenario, software upgrades should be able to occur automatically without any intervention by users. Users may accomplish other upgrades or changes manually. More extensive upgrades might be more akin to maintenance and require special personnel or equipment. Guideline 15.48: Avoid Interaction Changes with Upgrades Upgrades should not significantly affect how users interact with the device. If interaction changes are unavoidable, make users aware explicitly of any modification that requires them to interact differently (than before the upgrade) with the device.
Guideline 15.49: Upgrade Tasks Upgrade tasks must be consistent with the training, experience, and skills of the individuals who will perform the upgrades. For example, it is not appropriate to expect a nurse to perform manual software upgrades. Rather, this task is more appropriate for a biomedical engineer.
TAF-ER562X_WINGR-09-0401-C015.indd 638
11/9/10 1:38:38 AM
Device Life Cycle
639
Guideline 15.50: Minimize Risk of Upgrades Upgrades should not decrease the safety or efficacy of a device, even transiently.
Guideline 15.51: Upgrade Version Obvious The upgrade version of every device, such as its software version, should be clearly depicted and obvious to users and maintainers.
Guideline 15.52: Home-Use Device Upgrades Lay users should not be expected to perform upgrade tasks on medical devices. This is not to say that home-use devices cannot be upgraded, merely that lay users should not have to perform such upgrades.
Guideline 15.53: Documentation Upgrades All user documentation, including training materials, repair manuals, and user manuals, should be concurrently updated to reflect any device, procedural, and maintenance upgrades. Ideally, these changes should be made automatically (e.g., for all online documentation).
Guideline 15.54: Version of Documentation The most current version of user documentation should be visible on any page or display of the documentation.
Guideline 15.55: Avoid Version Mismatches There should be a clear and unambiguous association between the version of documentation and the current version of the device (so that version mismatches are obvious to users). Moreover, in a facility with multiple devices to be upgraded, care should be taken to prevent concurrent use of different versions of the same device, particularly if there have been changes in its functionality or the user interface. For example, it could be hazardous if a nurse uses two different versions of software in an infusion pump on the same ward.
15.6.9 DISPOSAL All medical devices will eventually wear out, become obsolete, or be damaged beyond repair. Disposing of devices is the last step in the device’s life cycle and can present challenges to users. There are essentially two concerns related to disposal of medical devices. The first concern is user safety. Medical devices can contain toxic substances, including both biological elements (e.g., blood) and inherently toxic materials, such as heavy metals. Some devices, because of their toxicity (e.g., solvents, batteries), require special disposal procedures and cannot be disposed of with regular trash. In some instances, the device itself can be inherently hazardous, as in the case of sharps, uncharged capacitators, and so on. The second disposal concern is environmental. Certain devices (or at least their components) can be recycled. It is imperative that the device and its individual components are disposed of using appropriate methods. Figure 15.10 shows the standard recycling symbols for various types of plastic. These markings can be seen on device coverings, such as for IV bags, and on many clear plastic trays (socalled form-and-fill containers) that hold various combinations of medical devices and parts.
TAF-ER562X_WINGR-09-0401-C015.indd 639
11/9/10 1:38:38 AM
640
Handbook of Human Factors in Medical Device Design
1
2
PETE
FIGURE 15.10
HDPE
3
4
V
LDPE
5
6
7
PP
PS
OTHER
Recycling symbols for plastics.
Guideline 15.56: Instructions for Disposal Devices and components should be labeled clearly with disposal instructions. Where feasible, markings should be placed directly on the disposable component as well as on its packaging. Packaging can become lost or otherwise separated from the device. Also, users might confuse packaging disposal instructions with device disposal instructions.
Guideline 15.57: Identify Hazardous Components for Disposal Toxic or otherwise hazardous components should have markings/labels that indicate clearly and unambiguously their toxic or hazardous nature.
Guideline 15.58: Recycling of Devices or Components Devices or components that are recyclable should be marked/labeled clearly and unambiguously. They should also have the appropriate recycling stream markings. For example, most people are familiar with the common plastic recycling designation of “2-HDPE,” or what is typically called “number 2” plastic. The specific types of recyclables are often referred to as the recycling “stream,” such as clear glass, newsprint, cardboard, “number 1” plastic, and so on.
Guideline 15.59: Keep Out of Landfills Devices or components that should be kept out of landfills should have markings/labels that indicate clearly and unambiguously that they should not be placed in refuse destined for landfills.
Guideline 15.60: Alteration of Markings Disposal and recycling instructions, graphics, and other indicative features should be permanent and difficult to alter. Disposal and recycling instructions, graphics, and other indicative features should be designed to survive the planned life span of the device or component.
15.7 CASE STUDY Dialysis machines typically require routine inspection, maintenance, and calibration. A dialysis machine, by its nature, contains numerous tubing and pumps, several filters, and a number of electronic subsystems. One common dialysis machine has pullout drawers containing much of the complex tubing and other components. An example of one of these
TAF-ER562X_WINGR-09-0401-C015.indd 640
11/9/10 1:38:38 AM
Device Life Cycle
FIGURE 15.11
641
Example of original pullout drawers design.
drawers is shown in Figure 15.11. This is an example of a design that provides access to internal components for maintenance and repair, albeit one that presents a very crowded and complex work area for maintainers. The manufacturer of this dialysis machine recognized the life cycle costs associated with such a complex design and set out to develop an entirely new device layout. The result, shown in Figure 15.12, is a foldout case that exposed all the internal workings of the device. This design also considerably simplified the tubing layout and provided easy access to all mechanical and electronic subassemblies. This design is also modularized and functionally grouped; that is, similar functional components are grouped into subassemblies that are easy to reach, test, and service. Because electronic communication has proved an important aspect of device use, the new design integrated technology to record and transfer individual patient data from machine
FIGURE 15.12
New foldout device design.
TAF-ER562X_WINGR-09-0401-C015.indd 641
11/9/10 1:38:39 AM
642
Handbook of Human Factors in Medical Device Design
FIGURE 15.13
Internal CD-ROM drive and card cage.
FIGURE 15.14
Internal PCMCIA card slot.
FIGURE 15.15
External Ethernet connection.
TAF-ER562X_WINGR-09-0401-C015.indd 642
11/9/10 1:38:39 AM
Device Life Cycle
643
to machine. Embedded technology included an internal CD-ROM and PCMCIA card slots (Figures 15.13 and 15.14, respectively). In addition to these onboard data transfer devices, the new design also included networking capability (Figure 15.15).
RESOURCES Brown, S. A., Merritt, K., Woods, T. O., and Hitchins, V. M. (2001). The effects of use and simulated reuse on percutaneous transluminal coronary angioplasty balloons and catheters. Biomedical Instrumentation and Technology, 35(5), 312–22. Hall, R. (2005). Packaging issues. Medical Device Technologies, 16(8), 32–34. Nolan, P. J. (2006). The ISO 11607 revision. Medical Device Technologies, 17(8), 41–42. Pilchik, R. (2004). Medical packaging: Towards global harmony. Medical Device Technologies, 15(8), 16–18. Wood, J. M., and Heyman, G. F. (2001). Reuser friendly: A review of the regulation of and the product liability regarding the reuse of single-use medical devices. Tort Trial and Insurance Law Journal, 37(1), 41–78.
REFERENCES American National Standards Institute/Association for the Advancement of Medical Instrumentation ANSI/AAMI. (2010). Human Factors Engineering Guidelines and Preferred Practices for the Design of Medical Devices. ANSI/AAMI HE-75-2010. Arlington, VA: Association for the Advancement of Medical Instrumentation. American National Standards Institute/Association for the Advancement of Medical Instrumentation ANSI/AAMI. (2001). Human Factors Design Process for Medical Devices. ANSI/AAMI HE-74-2001. Arlington, VA: Association for the Advancement of Medical Instrumentation. Casey, S. (1998). Set Phasers on Stun—and Other True Tales of Design, Technology, and Human Error (2nd ed.). Santa Barbara, CA: Aegean Publishing Company. Cushman, W. H. and Rosenberg, D. J. (1991). Human Factors in Product Design. New York: Elsevier. EN 980. (2003). Graphical Symbols for Use in the Labeling of Medical Devices. Brussels, Belgium: European Commission. European Union Committee on Standards and Technical Regulations. Harris, B., and Weinger, M.B. (2006). An insidious failure of an oxygen analyzer. Anesthesia & Analgesia, 102, 1468–1472. Health Industry Manufacturers Association. (1999, February 18). Position Paper on the Reuse of Single-Use Medical Devices. Washington, DC: Health Industry Manufacturers Association. Horwitz, B. (2002, April 8). If you want to be King Gillette, remember to patent the blades. Mass High Tech. Accessed from http://www.masshightech.com/stories/2002/04/08/focus5-If-youwant-to-be-King-Gillette-remember-to-patent-the-blades.html International Standards Organization (ISO). 2000. ISO 15223. Medical Devices—Symbols to be Used with Medical Device Labels, Labeling, and Information to be Supplied. Geneva: International Organization for Standardization. Kohn, L. T., Corrigan, J. M., and Donaldson, M. S. (Eds.). (2000). To Err is Human: Building a Safer Health System. Washington, DC: Institute of Medicine, National Academies Press. Ison, E. and Miller, A. (2000). The use of LCA to introduce life-cycle thinking into decision-making for the purchase of medical devices in the NHS. Journal of Environmental Assessment Policy and Management, 2(4), 453–476. Love, E., Asher, D., Atterbury, C. L. J., Chapman, C., Cohen, H., Jones, H., Norfolk, D. R., Revill, J., Soldan, K., Todd, A., and Williamson, L. M. (2002). The serious hazards of transfusion. In Royal College of Pathologists Annual Report 2000–2001. Manchester, England.
TAF-ER562X_WINGR-09-0401-C015.indd 643
11/9/10 1:38:40 AM
644
Handbook of Human Factors in Medical Device Design
Myers, C. and Burchill, T. (2002). The short life of a medical device. Health Forum Journal, September/October, 35–38. Newman, C. (2003, April). Life-cycle management: A long-term challenge. Medical Device Technology, 14(3), 32–33. Park, D.C., and Skurnik, I. (2004) Aging, cognition, and patient errors in following medical instructions. In M.S. Bogner (Ed.) Misadventures in healthcare: inside stories. Chapter 11, 165–182. Mahwah, NJ: Lawrence Erlbaum Associates. Proctor, R. W. and Van Zandt, T. (1994). Human Factors in Simple and Complex Systems. Needham Heights, MA: Allyn & Bacon. Reason, J. (1990) Human Error. New York, NY: Cambridge University Press. Senders, J. W. and Moray, N. P. (1991). Human Error—Cause Prediction, and Reduction. Hillsdale, NJ: Lawrence Erlbaum Associates. Stewart, I. (1997). “Single use only” labeling of medical devices: Always essential or sometimes spurious? Medical Journal of Australia, 167, 538–539.
TAF-ER562X_WINGR-09-0401-C015.indd 644
11/9/10 1:38:40 AM
16 Hand Tool Design Richard Botney, MD; Mary Beth Privitera, M. Design; Ramon Berguer, MD; Robert G. Radwin, PhD CONTENTS Medical Hand Tools .........................................................................................................647 16.1 General Principles .................................................................................................654 16.1.1 Biomechanics ..........................................................................................654 16.1.2 Handgrips and Positions Associated with Tool Use ................................655 16.1.3 Forces Associated with Hand Tool Use ...................................................656 16.1.4 Injuries and Discomfort Related to Tool Use ..........................................657 16.1.5 Risk Factors for the Development of Injuries and Musculoskeletal Disorders .................................................................................................659 16.1.6 Key Elements of the Medical Hand Tool Design Process ...................... 660 16.1.7 Compensatory Strategies .........................................................................662 16.1.8 Trade-Offs ...............................................................................................663 16.2 Design Guidelines ................................................................................................ 664 16.2.1 Context of Use .........................................................................................665 16.2.2 Location and Environmental Factors ......................................................667 16.2.3 The End Effector (Tool Interactions with Anatomy) ...............................669 16.2.3.1 Grasping Instruments .............................................................669 16.2.3.2 Force Output from Tool ..........................................................670 16.2.4 Considerations for the Whole Tool ..........................................................670 16.2.4.1 Handle Angle ..........................................................................671 16.2.4.2 Handle Shape ..........................................................................672 16.2.4.3 Handle Length ........................................................................674 16.2.4.4 Handle Diameter (Cross-Sectional Size) ................................675 16.2.4.5 Handle Cross-Sectional Shape................................................676 16.2.4.6 Handle Material ......................................................................678 16.2.4.7 Handle Surface and Texture ...................................................678 16.2.4.8 Tool Weight and Center of Gravity .........................................679 16.2.4.9 Safety ......................................................................................681 16.2.5 User Characteristics and Related Design Considerations ........................682 16.2.5.1 Posture ....................................................................................682 16.2.5.2 Shoulder ..................................................................................683 16.2.5.3 Elbow ......................................................................................684 16.2.5.4 Wrist and Hand .......................................................................684
645
TAF-ER562X_WINGR-09-0401-C016.indd 645
11/9/10 1:55:58 AM
646
Handbook of Human Factors in Medical Device Design
16.2.6
Grip .........................................................................................................685 16.2.6.1 Gender ....................................................................................685 16.2.6.2 Wrist Position..........................................................................685 16.2.6.3 Grip Span ................................................................................685 16.2.6.4 Number of Fingers Utilized ....................................................687 16.2.6.5 Handedness .............................................................................688 16.2.6.6 Gloves and Other Personal Protective Equipment ..................688 16.2.6.7 Grip Force ...............................................................................689 16.2.6.8 Frequency (Repetition) and Duration of Effort ......................690 16.2.7 Control Type and Placement ...................................................................691 16.2.7.1 Triggers ...................................................................................692 16.2.7.2 Slide Controls .........................................................................693 16.2.7.3 Push Button.............................................................................694 16.2.7.4 Rotation ..................................................................................694 16.2.7.5 Sensory Feedback ...................................................................695 16.3 Special Considerations ..........................................................................................696 16.3.1 Laparoscopic Surgery ..............................................................................696 16.3.1.1 Handles ...................................................................................698 16.3.1.2 Internal Mechanics .................................................................699 16.3.1.3 Fixed Insertion Points .............................................................699 16.3.1.4 Contextual Factors ..................................................................700 16.3.1.5 Guidelines for Laparoscopic Instrument Design ....................700 16.3.2 Minimally Invasive Catheter-Based Procedures .....................................701 16.4 What to Do When Guidelines Are Not Available .................................................704 16.5 Case Studies ..........................................................................................................705 16.5.1 Case Example 1: Accidental Needle Punctures with Catheter Use .........705 16.5.2 Case Example 2: The Harmonic Scalpel .................................................707 References ........................................................................................................................708
Ninety seconds into the surgery, my shoulders and neck began to ache. My arms began to tremble just a little. And I couldn’t afford to shift position or take a break. After all, the scope I was trying to hold steady was inserted into a patient’s abdomen … the instruments were difficult to manipulate, limited depth cues contributed to a few stab wounds, and a very restricted field of view hampered navigation and planning … the postures I would need to assume might seem “awkward.” “Contorted” and “agonizing” are the terms I would have used. (Reprinted with permission from M. Carswell in Ergonomics in Design, Vol. 12, No. 3, 2004. © 2004 by the Human Factors and Ergonomics Society. All rights reserved.)
The above description represents the experience of surgeons who use laparoscopic tools, yet virtually every health care worker is at risk for injury or discomfort when using medical hand tools. Perhaps more importantly, patient injury may occur when poorly designed medical hand tools are used (Gawande et al, 2003b; Samore et al., 2004; Stone and McCloy, 2004). Remarkably, hand tool designs that result in pain, numbness, muscle fatigue, injury, difficulty with use, or inefficiency are accepted as normal by many health care workers.
TAF-ER562X_WINGR-09-0401-C016.indd 646
11/9/10 1:55:58 AM
Hand Tool Design
647
Design for the health care environment is unique and challenging. Well-designed medical hand tools must satisfy clinical and safety requirements. Applying human factors (HF) during design helps to ensure that tools are efficient, effective, intuitive, comfortable, and safe to use (Table 16.1) (Association for the Advancement of Medical Instrumentation, 2001). One hallmark of a well-designed tool is that during use, the user’s attention will be focused on accomplishing the task and not on how to use the tool. In other words, the hand– handle interface “disappears,” and the tool becomes an extension of the user rather than a separate implement held in the hand (Baber, 2003). Hand tools are used in every environment where health care is delivered (Table 16.2), and their use affects patient care outcomes every day. There is a tremendous diversity in the types of situations, users, conditions of use, and modalities in which a tool might be used, and medical hand tool design requires consideration of all these factors (Table 16.3). While numerous texts provide HF design guidelines for nonmedical hand tools (Cacha, 1999; Helander, 1995; Salvendy, 1997; Woodson et al., 1992), this chapter is the first to do so specifically for medical hand tool design. This chapter is a comprehensive review of HF issues and considerations related to medical hand tool design and use. General guidelines and frequent examples are provided, and specific design recommendations are made when data exist to support them. However, with so many types of medical tools, it is impossible to provide comprehensive or specific HF design guidance for each and every tool or even every class of tool.
MEDICAL HAND TOOLS Any device used to perform or facilitate manual or mechanical work can be considered a tool. Tools are often regarded as necessary to carry out one’s occupation or profession, or may be used in the performance of some procedure or operation. Hand tools amplify strength, extend reach, and concentrate forces (Cacha, 1999). Tools enhance human motor capabilities and control, and help individuals perform tasks that otherwise would be more difficult or impossible to accomplish using only bare hands (Patkin, 1967; Radwin and Haney, 1996; Dababneh and Waters, 1999). TABLE 16.1 Benefits of Having Well-Designed Medical Hand Tools Design Considerations
Potential Benefits
User safety
Reduced chance of direct (acute) injuries from tool use (e.g., needlestick injury) Reduced incidence and severity of musculoskeletal (chronic) injuries related to tool use Reduced risk of unintended patient injuries from tool use Minimized extent of tissue trauma during tool use Improved quality of care Increased efficiency and productivity Reduced need for awkward positions or motions, or excessive force Reduced muscle fatigue (muscle fatigue decreases productivity and efficiency, and increases the risk of injuries to workers and patients)
Patient safety
Comfort Ease of use Cost-effectiveness
TAF-ER562X_WINGR-09-0401-C016.indd 647
11/9/10 1:55:58 AM
648
Handbook of Human Factors in Medical Device Design
TABLE 16.2 Examples of Medical Hand Tools and Where They Are Commonly Used Use Locations
Examples of Tools Used by Users in Each Setting
Operating room (surgeon)
Forceps, clamps, scissors, scalpels, retractors, staplers, screwdrivers, drills, hammers, needle drivers, cautery probes, endoscopic instruments Laryngoscopes, stethoscopes, suction catheters, ventilation bags, fiber-optic laryngoscopes, forceps, syringes Otoscopes, sphygmomanometers, stethoscopes, staple removal instruments, minor surgical instruments, thermometers Probes, catheters, guide wires, tubes, ultrasound probes Minor surgical instruments, otoscopes, ophthalmoscopes, stethoscopes, sphygmomanometers, suction catheters, ventilation bags, forceps Forceps, surgical instruments, ultrasound probes Otoscopes, ophthalmoscopes, sphygmomanometers, suction catheters, ventilation bags, fiber-optic laryngoscopes, various catheters and catheter assemblies, ultrasound probes Pipettes, tubes and containers, syringes Otoscopes, ophthalmoscopes, sphygmomanometers, minor surgical instruments, thermometers Mechanical measuring devices, ultrasound probes Multiple handheld instruments, powered drills Syringes, laryngoscopes, forceps, stethoscopes, surgical instruments Lancets, syringes, catheters and tubes, thermometers, dose dispensers, toothbrushes, flossing tools
Operating room (anesthesiologist)
Hospital wards
Radiology Emergency room
Labor and delivery Intensive care
Clinical laboratory Clinics Rehabilitation clinics (e.g., PT, OT) Dental office Mobile sites (e.g., helicopter, ambulance) Home (usually by patients)
Notes: There is considerable variation in the types of tools and their appearance, function, and use (see also Figures 16.1–16.9). Two entries are provided for the operating room because, although surgeons and anesthesiologists are in the same room, they work in different locations, with very different sets of tools, tasks, and so on. Smith and Smith (1983) provide a comprehensive listing of surgical tools, with illustrations, but do not include laparoscopic/arthroscopic instruments or nonsurgical tools. Anderson and Romfh (1980) provide a comprehensive description of many surgical instruments, including how they are held and used, with illustrations of the many and varied handgrips used by surgeons.
For the purposes of this chapter, a medical hand tool is a device that is held and operated in the hands to perform a clinical task. Tools require motive power, which may be provided by the user or by external sources. The term instrument describes precision tools used by trained professionals, such as surgeons. Hand tools typically consist of a handle and a working end (the end effector) and sometimes a shaft or body in between. Thus, a great many medical devices are medical hand tools, ranging from simple tools such as syringes and hypodermic needles to very complex tools such as endoscopes and endovascular catheters.
TAF-ER562X_WINGR-09-0401-C016.indd 648
11/9/10 1:55:58 AM
Hand Tool Design
649
TABLE 16.3 Considerations Related to Hand Tool Use in Health Care Consideration
Comments
Many types of users and individuals handle tools, and they may or may not be adequately trained.
Physicians, nurses, physician assistants, technicians, clinical engineers, paramedics, students, patients, lay caregivers. Harm may be mechanical, chemical, electrical, or thermal from power tools (e.g., drills), lasers, electrocautery units, sharp instruments (e.g., needles or scalpels), or radiation. Sterile/nonsterile, alone or in a team setting, emergency/routine, reusable/disposable. Injury may occur during handoff. Scissors may be used for cutting or dissection, for example. Gloves, gowns, masks, face shields, or lead aprons may affect vision, grip, or comfort.
Tools may cause injury, either intentionally (e.g., surgery) or unintentionally, and in different ways. Tools are used in a variety of conditions and use environments. Tools are often handed off between users. Tools are used for a wide range of tasks, and there are a large number and variety of tools. Personal protective equipment is often used and may affect the tool’s use.
Examples of hand tools used in health care are shown in Figures 16.1 through 16.12 (see also Table 16.2); note the various handgrips used with these tools. Medical hand tools differ from conventional hand tools in that they act on and affect patients, whereas most conventional tools act on inanimate objects. Medical hand tools are used to obtain diagnostic information (e.g., an otoscope) or, in the case of invasive procedures and surgery, to physically alter the human body (see Table 16.4). Human tissues are unlike many inanimate materials; they are made of living cells that can regenerate or repair themselves and bleed, and often have highly complex and nonlinear mechanical properties. Thus, special consideration must be given to a tool’s interactions (planned or inadvertent) with human tissues. Tools must perform their intended function(s) on patients without causing unintended harm. There is also the need to preserve the integrity of tissue specimens obtained during medical procedures (e.g., for pathological diagnosis). Most commonly, there is a direct physical interaction between a health care worker, a tool, and a patient. The interaction with patients’ tissues (i.e., technical efficacy) has
FIGURE 16.1
Syringe (control type).
TAF-ER562X_WINGR-09-0401-C016.indd 649
11/9/10 1:55:58 AM
650
Handbook of Human Factors in Medical Device Design
FIGURE 16.2
Scalpel held in an internal precision grip.
FIGURE 16.3
IV catheter assembly held in a tip pinch.
FIGURE 16.4
Otoscope held in typical power grip.
TAF-ER562X_WINGR-09-0401-C016.indd 650
11/9/10 1:55:59 AM
Hand Tool Design
FIGURE 16.5
Forceps held in an external precision grip.
FIGURE 16.6
Laparoscopic instrument with pistol grip and finger rings.
FIGURE 16.7
Pneumatically powered drill.
TAF-ER562X_WINGR-09-0401-C016.indd 651
651
11/9/10 1:55:59 AM
652
Handbook of Human Factors in Medical Device Design
FIGURE 16.8
Flexible endoscope.
FIGURE 16.9
Endoscope hand piece with multiple controls.
FIGURE 16.10
Scissors held in a dynamic tripod grip.
TAF-ER562X_WINGR-09-0401-C016.indd 652
11/9/10 1:56:00 AM
Hand Tool Design
653
FIGURE 16.11
Suction catheter held with a palmar pinch grip.
FIGURE 16.12
Syringe and needle, with needle held in a lateral (key) pinch grip.
traditionally been the focus of medical tool design, with much less emphasis placed on handle design or the controlling interface. However, with the advent of more complex clinical applications and sophisticated devices, and the growth of a significantly competitive marketplace, handle design has become both an enabling and a differentiating element in the usability of hand-operated devices. TABLE 16.4 Some of the Functions Performed by Hand Tools in the Medical Setting Tool Functions Cutting, dissecting, sawing Evacuating, washing, cleansing Grasping, gripping Lifting Illuminating, magnifying, viewing Poking, probing Tying, sewing, taping Piercing, inserting, withdrawing
TAF-ER562X_WINGR-09-0401-C016.indd 653
Pulling, pushing, moving Hammering, percussing Drilling, boring Burning Destroying (tissue) Abrading, scraping Rotating, turning Writing, inscribing, etching
Spreading apart, separating Measuring, injecting, aspirating Stapling Vibrating, making a sound Dabbing, swabbing Listening, looking Gluing, adhering, attaching Pressing, compressing
11/9/10 1:56:01 AM
654
Handbook of Human Factors in Medical Device Design
Medical hand tools also differ from their nonmedical counterparts in the need for sterility or to be resterilized for reuse, both of which dictate certain tool materials and construction requirements. Single-use (disposable) devices may use plastic or have metal parts that are not intended to be cleansed after use. There are also reposable (partially reusable, partially disposable) devices (Ostlie and Holcomb, 2003). Reusable tools need to be made of materials and have surfaces and actuating mechanisms that permit mechanical and chemical cleansing. They must also withstand sterilization by one or more means, such as chemical baths, steam, and gas. General design recommendations for instruments used in sterile environments can be found in ANSI/AAMI/ISO 14397 and FDA guidance documents (American National Standards Institute, 2000; Food and Drug Administration, 1983, respectively).
16.1 GENERAL PRINCIPLES Several important considerations underlie the design guidelines provided later in this chapter. These include an understanding of biomechanics, the various handgrips and forces associated with tool use, and the nature of and risk factors for injuries and disorders that can be associated with tool use. Also discussed in this section are key elements of the medical hand tool design process, compensatory strategies employed by users to accommodate a tool’s shortcomings, and trade-offs that must often be made during tool design.
16.1.1 BIOMECHANICS Knowledge of the body’s tolerance or capacity for loading allows tasks and tools to be designed that enhance performance and minimize the risk of injuries associated with tool use (Kroemer, 1999). Although a detailed review of biomechanical principles is beyond the scope of this chapter, aspects of biomechanics related to posture and the shoulder, elbow, wrist, and hand are discussed. Refer to Chapter 4, “Anthropometry and Biomechanics,” or a biomechanics text (e.g., Nordin and Frankel, 2001) for more detailed anthropometric and biomechanical data. Substantial anthropometric data come from U.S. military sources and
FIGURE 16.13 A second blood specimen tube is held between the palm and the fourth and fifth fingers in a pulp (ulnar) pinch grip, also known as ulnar storage grip.
TAF-ER562X_WINGR-09-0401-C016.indd 654
11/9/10 1:56:01 AM
Hand Tool Design
FIGURE 16.14
655
Hammer held in a power grip.
may be a reasonable starting point. However, such data must be used with caution in user populations with gender and ethnic diversity, such as health care (Kaplan, 1981; Morse and Hinds, 1993). The data may not be applicable at the extremes of size and may not necessarily correlate with other design parameters (e.g., strength). One design procedure might be to select the most critical parameter (e.g., grip span), then identify the group with the largest number of users, and finally determine the appropriate percentile cutoff points (highest and lowest values). Typically, this is the 5th to 95th percentile. However, for critical or high risk functions, it is preferable to design for the 1st to 99th percentile. Note that this approach will still exclude two percent of users.
16.1.2 HANDGRIPS AND POSITIONS ASSOCIATED WITH TOOL USE Handgrips are usually classified as either power grips or precision grips. The hand exerts force with a power grip and uses a precision grip to perform delicate, precise tasks. There are numerous variants for each type, some of which are described below. It is common to
FIGURE 16.15
Neutral hand position.
TAF-ER562X_WINGR-09-0401-C016.indd 655
11/9/10 1:56:02 AM
656
Handbook of Human Factors in Medical Device Design
observe surgeons and other health care providers using a variety of unique grips (Figures 16.1–16.7) (Anderson and Romfh, 1980). A precision (pinch) grip, as the name implies, is used for high-precision, low-force tasks, such as those performed during microsurgery. Precision grips involve the intrinsic muscles of the hands and fingers, which provide the fine motor control needed for precision tasks. A precision grip typically employs a pinch grasp between the fingers or fingertips, such as when holding a pen or pencil. The two precision grips seen most commonly in surgery are the internal precision grip (Figure 16.2) and the external precision grip (Figure 16.5). Variants on the precision grip include the dynamic tripod (Figure 16.10), tip pinch (Figure 16.3), palmar pinch (Figure 16.11), lateral (key) pinch (Figure 16.12), and pulp (ulnar) pinch (Figure 16.13). Patkin (2001) has called the latter the ulnar storage grip. In a power grip, all the fingers wrap around the handle, with four fingers on one side of the handle and the thumb on the other side (Figures 16.7 and 16.14). Extending the thumb along the handle’s shaft can be considered a power grip with a precision component that allows the tool to be better guided and controlled during use (Figure 16.4) (Patkin, 2001). The large extrinsic arm muscles provide the strength needed for the forceful exertions associated with a power grip. The hand is less capable of performing precise movements and exertions when using a power grip because recruitment of the many motor units in the large muscles needed to generate high forces limits motor control. Power grip strength can be as much as five times greater than that of the precision grip (Mital and Karwowski, 1991). There are three basic variants in the application of a power grip, distinguished by the direction of force relative to the forearm: (1) the force is parallel to the forearm (e.g., a saw), (2) the force is at an angle to the forearm (e.g., a hammer), or (3) there is torque around the forearm (e.g., a screwdriver) (Helander, 1995; Mital and Karwowski, 1991). The body is generally strongest when positioned with neutral postures: the shoulders holding the arms adjacent to the torso, the elbow at about 90 degrees of flexion, and the hand in a neutral position. Neutral wrist and hand position can be taken as the position of the hand when preparing to shake hands: there is none or slight (<20 degrees) ulnar deviation and no radial deviation of the wrist, the fingers are open and slightly flexed, and there is a slight amount (<20 degrees) of wrist extension (Figure 16.15) (Nordin and Frankel, 2001). Larger amounts of wrist extension or ulnar deviation or any wrist flexion or radial deviation will decrease grip strength and the efficiency of transmission of force from the forearm to the hand and tool.
16.1.3 FORCES ASSOCIATED WITH HAND TOOL USE The forces associated with hand tool use can be divided into four classes. The force exerted by the user to grasp, hold, and squeeze instruments is commonly referred to as grip force or grip strength and is discussed further in the section on grip (see Section 16.2.6). A second force is that needed to activate controls. A third is the force that is transmitted by the user, via the tool, to the target. Important considerations include the exertion needed by the hand and upper extremity to actuate the tool’s intended motion or effect and the output force transmitted to the target. These motive forces are discussed in the end-effector section. Finally, reactive forces are transmitted back to the user as a result of the exertion of grip force and motive force. Several factors, such as the magnitude and direction of force, how long the force is applied, and the frequency with which the force is applied, affect how much force the user can exert safely (Iridiastati and Nussbaum, 2006).
TAF-ER562X_WINGR-09-0401-C016.indd 656
11/9/10 1:56:02 AM
Hand Tool Design
657
16.1.4 INJURIES AND DISCOMFORT RELATED TO TOOL USE Injuries associated with tool use may be broadly classified as acute or chronic. Acute injuries occur “instantaneously,” generally involve large but infrequent exertions, and are relatively common (Table 16.5). A variety of occupational injuries are reported by health care workers, but of those associated with medical tool use, puncture injuries are among the most common (Hefflin et al., 2004). The most important of these are accidental needlestick injuries, estimated to occur at least 800,000 times per year, which can result in the transmission of infectious diseases such as hepatitis or AIDS (Tan et al., 2001). Needlestick injuries are discussed from the design perspective in Case Example 1 later in this chapter. Unintended patient injuries also occur; for example, bile duct injuries during laparoscopic cholecystectomy (removal of the gallbladder) can be related to instrument design (Tang et al., 2004). Chronic injuries are often caused by cumulative trauma, or the repeated loading of an anatomic structure (e.g., tendon); the gradual, insidious buildup of trauma produces biomechanical insults to tissues (Table 16.6) (Cacha, 1999). Chronic musculoskeletal injuries and disorders characteristically cause long-term pain and disability. Musculoskeletal disorders associated with repetitive use are known by a variety of names, such as cumulative trauma disorders or repetitive motion disorders; the currently accepted term is work-related musculoskeletal disorder. Musculoskeletal injuries can involve the soft tissues (muscles, ligaments, tendons), nerves, blood vessels, or the skeleton (bones, cartilage). Detailed descriptions of the various upper-extremity musculoskeletal disorders can be found in occupational health and other texts (Herington and Morse, 1995; Levy and Wegman, 1995; McCunney, 1994; National Research Council, Institute of Medicine, 2001; Rempel et al., 1992; Violante et al., 2000). Musculoskeletal disorders are common among health care workers (Burda, 1995; Evanoff et al., 1999; Orr, 1997; Stout, 1992). Musculoskeletal disorders associated with medical hand tool use are underreported, and the magnitude of the problem is poorly documented (Wauben et al., 2006). Dentistry has long recognized that musculoskeletal injuries can occur as a result of poor instrument design, and much progress has been made in improving the design of dental hand tools. Nevertheless, the effect of prolonged or repetitive use of dental instruments, in conjunction with the awkward postures that dental workers must assume, continues to cause chronic injuries (Lalumandier and McPhee, 2001; Murphy, 1998). As already noted, problems such as numbness, pressure sores, and various other musculoskeletal disorders have been associated with laparoscopic instrument use (Berguer et al., 1999; Majeed et al., 1993; Sackier and Berci, 1992; Verma, 2004). Nonlaparoscopic surgical procedures and tools have been similarly implicated. Raynaud’s phenomenon (a condition resulting in poor circulation to the extremities) has been associated with the
TABLE 16.5 Acute Injuries Associated with Tool Use Types of injuries
Examples
Puncture wounds Cuts, lacerations Burns, blistering Electrical shock
Needlestick injuries Accidental scalpel wound Airway fire, accidental burns from electrocautery Shock to user while using electrocautery
TAF-ER562X_WINGR-09-0401-C016.indd 657
11/9/10 1:56:02 AM
658
Handbook of Human Factors in Medical Device Design
TABLE 16.6 Chronic Injuries and Other Disorders Associated with Tool Use Anatomic Involvement
Types of Injuries and Disorders
Tendon injuries
Tendonitis Tenosynovitis Trigger finger de Quervain’s tenosynovitis Ganglion cysts Epicondylitis, medial or lateral Rotator cuff tendonitis Raynaud’s phenomenon Hypothenar hammer Thoracic outlet syndrome Hand–arm vibration syndrome Carpal tunnel syndrome Cubital tunnel syndrome Back and neck pain Sciatica Degenerative joint disease
Neurovascular conditions
Neuropathy Miscellaneous
use of pneumatically powered surgical instruments (Cherniack and Mohr, 1994). Surgical retractors are uncomfortable to hold, and poor surgical exposure (visualization of the area where the surgeon is working) has been related to improper handle design (Figure 16.16) (Brearly and Watson, 1983; Patkin, 1980). Halford and Birch (2005) report that podiatrists experience hand pain that is in part due to medical tool use. Carpal tunnel syndrome has been reported in anesthetists related to the performance of direct laryngoscopy and the need to hold face masks for extended periods (Diaz, 2001). Overuse syndromes, possibly related to instrument design, have been described for gastroenterologists performing endoscopy (see Figure 16.8) (Buschbacher, 1994; Hirschowitz, 1994; Siegel et al., 1994). Work-related musculoskeletal disorders have been related to the use of ultrasonographic equipment (Horkey and King 2004;
FIGURE 16.16 A Deaver retractor, which, if held as shown for any length of time, concentrates forces in localized regions of the palm, producing pain and discomfort for the user.
TAF-ER562X_WINGR-09-0401-C016.indd 658
11/9/10 1:56:02 AM
Hand Tool Design
659
Magnavita et al., 1999; Mercer et al., 1997; Schoenfeld, 1998). Clearly, there are occupational hazards associated with medical hand tool use. However, the full extent of this problem, the role of hand tool design attributes, and the impact on patient care are largely unknown (Babar-Craig et al, 2003; Frazier et al., 1995; Hefflin et al., 2004; Öhman et al., 2002; Small, 2004; Wilkinson et al., 1992).
16.1.5 RISK FACTORS FOR THE DEVELOPMENT OF INJURIES AND MUSCULOSKELETAL DISORDERS In general, whenever a tool is used to exert large forces or requires a repetitive or sustained effort, there is an increased risk of user discomfort, fatigue, or musculoskeletal injury. Risk factors fall into three broad categories: task-related, environmental, and compounding factors (Table 16.7). The magnitude of exerted force, the duration, and the frequency of application (repetition) are undoubtedly the most important factors but posture and positioning, environmental conditions such as cold working temperatures, and vibration that is transmitted to the hand also significantly increase the risk of injury. TABLE 16.7 Risk Factors Associated with Hand Tool Use That May Contribute to the Development of Work-Related Musculoskeletal Disorders Category
Risk Factors
Task related (physical and temporal factors associated with tool use and the task to be performed)
Large force exertions (peak, average) High frequency (repetition rate) Long duration of use Fatigue with inadequate recovery time or an unbalanced work–rest cycle Extreme postures and nonneutral hand and wrist positions
Environmental
Noise Cold or hot temperatures Vibration transmitted to the hand
Compounding factors (user-related and other considerations)
Physical health Prior injury or disability Worker psychological attributes (e.g., attitude, stress, job satisfaction) Gender Ethnic background New technologies Work complexity Personality traits Economic instability and costs, financial compensation Social and political factors
Sources: Helander, M., A Guide to the Ergonomics of Manufacturing, Taylor & Francis, Bristol, PA, 1995; Kerk, C.J., Occup Med, 13, 583–598, 1998; McGorry, R.W., Appl Ergon, 32, 271–279, 2001; Morse, L.H., and Hinds, L.J. Occup Med, 8, 721–731, 1993; Strazdins, L. and Bammer, G., Soc Sci Med, 58, 997–1005, 2004.
TAF-ER562X_WINGR-09-0401-C016.indd 659
11/9/10 1:56:03 AM
660
Handbook of Human Factors in Medical Device Design
Other factors may compound the effects of the above primary risk factors. The user’s physical health, strength and endurance, and anthropometric and biomechanical considerations (e.g., body dimensions and joint flexibility) affect the risk of developing a musculoskeletal complaint. These factors may increase the risk beyond that related to tool use (e.g., body weight increases biomechanical loading of joints beyond that due to tool use alone). Alternatively, preexisting musculoskeletal disorders may be exacerbated by tool use. Social factors may influence individuals’ perceptions that complaints of musculoskeletal pain are work related. Social factors include the introduction of new technologies, economic issues and financial compensation, labor activism, environmental concerns, cultural stereotypes, media attention, marketing efforts, military conflicts, and political action. Psychosocial factors can also be associated with complaints of musculoskeletal pain and may relate to aspects of the work environment such as work complexity, job attitude and satisfaction, and other personality traits (Kerk, 1998). Control of major tool-related risk factors by incorporating HF principles into tool design should result in more usable tools, increased productivity, and fewer user or patient injuries (Bernard, 1997; Kerk, 1998; National Research Council, Institute of Medicine, 2001; Radwin and Haney, 1996).
16.1.6 KEY ELEMENTS OF THE MEDICAL HAND TOOL DESIGN PROCESS The product development process is cyclical. For hand tools, design generally starts by defining the end effector’s clinical objective and then working toward a method of controlling it. To determine the overall hand tool configuration, the clinical target, surrounding anatomical structures, user position, and purpose must be considered. This is accomplished through a detailed analysis that identifies the key elements of the clinical procedure, including when, why, and by whom it will be performed. While direct user observation and simulations are important parts of this process, there may be cases where there is no previous clinical counterpart. In this instance, the design team may use tables or storyboards to define new clinical procedures that are under development. The user’s ability to visualize the clinical target during tool use is critically important (Patkin, 1977). This may be direct visualization, as in open surgical procedures; through microscopes; or on monitors, such as for laparoscopic and endoluminal procedures. Findeffector visualization is an important consideration for overall handle configuration, control selection, and location. For example, if the clinician is looking at monitors to visualize the interaction between a tool and its target, as in laparoscopic procedures, the tool in use should not require the user to also look at the handle in order to use the device effectively. In some cases, there may be obstacles to visualizing the target or to using the tool (see Case Example 2 later in this chapter). Knowledge of these impediments helps determine the necessary angles of use and the overall geometry of both the end effector and the hand piece. The user’s postures relative to the clinical target will also contribute toward determining handle angles, orientation, leverage, pivot points and distances, and ultimately the best overall handle configuration. On the basis of this information, grip types, such as pistol, in-line, T-grip, ball, and so on, can then be determined (see Figures 16.25 and 16.27). The means of access to the target should also be considered, both those currently used and possible future methods. For example, many new surgical procedures are first performed as open procedures and then subsequently as minimally invasive procedures. The possible means of target access can affect decisions about shaft cross sections and other elements of tool design.
TAF-ER562X_WINGR-09-0401-C016.indd 660
11/9/10 1:56:03 AM
Hand Tool Design
661
Control layout should be logical with respect to the user’s expectations of purpose and functionality. In selecting control types, an initial consideration is to map the expected outcome of the end effector to an appropriate control type and then consider the control’s location. For example, if the end effector rotates about an axis, then its control element should rotate about the same axis (Figure 16.17). The control and end-effector relationship should also provide feedback or confirmation that an action has been taken. Control labels should be used only to supplement other types of feedback, such as audible clicks when fully closed (see Chapter 7, “Controls,” for specific recommendations). Controls selected for use on hand tools are generally limited to the access and forceproducing capabilities of individual fingers. Based on a conceptual design, it is possible to predict the force requirements of each control through mechanical analysis. These requirements are matched with human capabilities to optimize control design and, ultimately, initial product specifications. This initial conceptual design is then evaluated with intended end users, often with three-dimensional sketch models or proof-of-concept physical models, which may or may not be functional. Eventually, preclinical studies are conducted. Ideally, users are given a working prototype to perform tasks representative of the intended use, often on inanimate models or in animal studies. These tests can be accomplished concurrently with technical design verification studies. Tests should consider the overall tool mass and geometry necessary for the user to feel that the device is well controlled. Force requirements of tissue manipulation, use in the face of expected anatomical constraints, and whether device use is truly one-handed can also be determined. Design teams should look for unintended motions or postures that users assume during prototype tool use. Both formal and informal user input should be sought early and often. Informal evaluations may consist of interviews or observations with a limited number of users, while formal evaluations include full usability testing (refer to Chapter 6, “Testing and Evaluation,” for additional information). Regardless, both qualitative and quantitative feedback should be sought, including the following:
FIGURE 16.17 Fiber-optic laryngoscope with the tip rotating in the same direction as the thumb control.
TAF-ER562X_WINGR-09-0401-C016.indd 661
11/9/10 1:56:03 AM
662
Handbook of Human Factors in Medical Device Design
Qualitative measures: • • • • • •
How easy was it to learn to use the tool? Did clinicians use the tool as the designers had intended? If not, why not? Which handle style was preferred? What improvements, if any, would users make to the handle? What are the perceived challenges with handle design? What is the perceived risk of trauma to clinical target tissues and surrounding tissues?
Quantitative measures: • What forces were applied to clinical target tissues and surrounding tissues? • Did tissue injury or user discomfort occur? • How much force or torque did the user need to exert to use the device? To operate each control? • Did interaction with the device create pressure or pinch points? • How much time was necessary to perform the task? • How many and what type of errors did the user make?
16.1.7 COMPENSATORY STRATEGIES Even under the best of circumstances, it is likely that a tool’s ergonomics will not be optimal. For example, a tool may be used for a novel application, different users may have different use characteristics, the task might change, or the user-interface design requirements might conflict with other considerations. When a tool is not optimally designed, users often develop methods to accommodate a tool’s shortcomings. These are known as compensatory strategies. Examples of compensatory strategies include adjustments to surgical table height and the use of step stools to allow surgeons of varying heights to be accommodated, use of a wrist rest during neurologic surgery to minimize tremor (Greenberg, 1981), and the use of special mammillated gloves to increase friction (Patkin, 1967). Such strategies can reduce the risk of injury or other problems associated with the tool’s use. For example, many surgeons use a palmgrasping position (with the thumb outside and the palm resting on the thumb ring) to reduce the muscle forces used for grasping an object with a laparoscopic grasper (Figure 16.18) (Berguer, 1998). While compensatory strategies may permit more effective tool use, they can have other unintended or undesirable consequences. For example, if the surgical table height is adjusted for a short surgeon, a tall surgeon will be required to hunch over in an undesirable and uncomfortable posture. It is important to consider how a design might be influenced by or interact with compensatory strategies. Whenever possible, undesirable or potentially risky compensatory strategies should be reduced through tool or care process redesign. For example, even if the tool’s design cannot be altered to mitigate a compensatory strategy, the manufacturer could warn about use-related issues, thereby making the user aware of the concerns and how to reduce the risks. Furthermore, if the user has to adopt a compensatory strategy to use a tool, this can be valuable information to guide subsequent designs. Compensatory strategies that users may employ include the following:
TAF-ER562X_WINGR-09-0401-C016.indd 662
11/9/10 1:56:03 AM
Hand Tool Design
663
FIGURE 16.18 Laparoscopic instrument held in a palm-grasping position. Such a hand position reduces the muscle forces necessary to grasp an object with the instrument.
• Controlling hazards by various means, including engineering (e.g., tool alterations), administrative methods (e.g., policies on tool use), the use of personal protective equipment (which may affect tool use or contribute to the risk of injury), or the use of supplemental equipment that may impact on tool use (e.g., wrist rests to reduce tremor or fatigue). • Using a variety of sizes to ensure proper hand fit. Alternatively, users may alter the tool through ad hoc grip size changes or extensions. • Methods to reduce the impact of the effects of tool use, such as (1) rotation of workers, which can result in different user types; (2) exercises or therapies to mitigate the effects of tool use; (3) redesign of the workspace or task, which may affect how a tool is used; (4) training and education, which can improve tool use; or (5) adjusting one’s posture to optimize tool use (Buffington et al., 2006; Quick et al., 2003). • Preventative maintenance of tools (Halford and Birch, 2005). • Medical management of musculoskeletal disorders to minimize their impact on work and performance.
16.1.8 TRADE-OFFS No single device form or configuration will be ideal for all users. Good design results from a balance between the core elements found in business, design, and engineering, coupled with a close relationship with clinical practice. Thousands of decisions need to be made by the development team to bring a product to market. Thus, compromises are inevitable during the development process. In most cases, there will be a trade-off between the HF considerations and other design issues, such as functionality, cost, size, and so on. Moreover, there may be trade-offs among multiple HF considerations. For example, grip force must be high enough to exceed reactive forces, thereby permitting effective tool use, but it should not be so high as to result in an injury. The need for higher grip forces may necessitate a control or handle design that is not optimal for fine manipulation or hand comfort. There may also be trade-offs that affect how the tool will be used. For example, there is a trade-off in the use of nonpowered tools between force and repetition. The ability to exert greater force will reduce the number of repetitions required to accomplish a task. Conversely, less force can be exerted, but this will
TAF-ER562X_WINGR-09-0401-C016.indd 663
11/9/10 1:56:03 AM
664
Handbook of Human Factors in Medical Device Design
FIGURE 16.19 Minimally invasive dissector with a rotary control permits one-handed adjustment of the end-effector articulation. (Photo courtesy of AtriCure, Inc. With permission.)
increase the number of repetitions required to perform a task. A tool cannot be designed to minimize both the force required and the number of repetitions. The use of soft foam grips or other protective materials will reduce the transmission of vibration to the hand but will have the opposite effect if it requires the use of greater grip force (Helander, 1995; Pelmear and Leong, 2000). Innovative approaches to design trade-offs may yield significant improvements in tool functionality or usability. For example, in designing a minimally invasive dissector, the initial intent was to mirror the articulating action of the end effector with a slide control on the body of the hand piece. However, rigorous force analysis and mechanical design iterations determined that the output force of the slide control would not meet the clinical requirements of the end effector, as the slide control would require the use of two hands to move the end effector. By changing to a rotary control to control articulation, infinite adjustment of the articulating end effector and ideal torque ranges for the user became possible (Figure 16.19).
16.2 DESIGN GUIDELINES It is important to realize that these guidelines are not rigid prescriptions for medical hand tool design. They highlight issues that should be addressed during design and, when possible, provide specific quantitative HF design recommendations. However, data to support design recommendations for medical hand tools are limited. Many recommendations are based on studies involving nonmedical hand tools, which may or may not be applicable to medical hand tool design. Moreover, many recommendations found in the nonmedical literature have little data to support them and must be cautiously applied. Where deemed especially important, recommendations that may not be directly applicable to medical tools will be identified (e.g., Guideline 16.57: The recommendation on precision tool weight, not to exceed 1.8 kg, may be excessive for some precision medical tools). In addition, medical hand tool design recommendations are often based on studies from a single institution, have not been broadly validated, are based on opinion, or are derived from benchtop and not clinical testing. Consequently, during design, guidelines and assumptions should be tested rigorously within the specific use context and with the intended user population. The numerous factors that influence hand tool design and use are discussed below. These include but are not limited to the context of use, the use location and environment, the end effector and its interaction with the target (patient tissues), characteristics of the tool itself (e.g., handle configuration, grip type, control type, and location), and user characteristics (including anthropometry and biomechanics).
TAF-ER562X_WINGR-09-0401-C016.indd 664
11/9/10 1:56:04 AM
Hand Tool Design
665
16.2.1 CONTEXT OF USE It is essential that the tool be designed to fit the clinical task for which it is intended; the context in which a tool is used is also an important consideration affecting tool effectiveness and safety. Context refers to the numerous factors that can change when using the same tool in different settings or circumstances. Various contextual factors that may affect tool use and therefore design include equipment-related factors, aspects of the task for which the tool is designed, user-related factors, patient factors, and other factors (Table 16.8). Factors related to the use location and environment are discussed in the next section. The medical setting affects tool design in specific ways. In addition to the need for sterility and the need to minimize the risk of injury to patients, task characteristics can be quite different from nonmedical settings. For example, since medical work is usually not cyclical, it may be difficult to automate many tasks. Breaks and rest periods are more difficult to provide on a set schedule. There is little control over design of the target (i.e., patients’ tissues), and it is sometimes not possible to alter the way in which a task is performed. For example, most types of surgery are done with the surgeons and scrub nurse standing. Finally, the use of other tools or equipment can result in either positive or negative interactions. TABLE 16.8 Contextual Factors to Be Considered General Category
Specific Factors
Equipment related
Workstation design Furniture Adjustability Other tools and medical devices Transfer of training from similar tools or devices Personal protective equipment (gloves, gowns, masks, face and eye shields, lead aprons) Risk of patient injury Task characteristics (task height, required force, exertions) Repetition Workload Automation Human capabilities and training Breaks, shift work, fatigue Mood, attitude, psychology Gender (hand size, strength, reach) Ethnicity (hand size, strength, reach) Handedness Individual anatomy, physiognomy, and physiology Age Gender Medical conditions Sterility, sterilization process, choice of materials Disposability and choice of materials
Task related
User related
Patient related
Other
Sources: Cancio, L.I. and Cashman, T.M., Am J Occup Ther, 53, 227–230, 1998; Morse, L.H., and Hinds, L.J. Occup Med, 8, 721–731, 1993.
TAF-ER562X_WINGR-09-0401-C016.indd 665
11/9/10 1:56:04 AM
666
FIGURE 16.20
Handbook of Human Factors in Medical Device Design
Pediatric, short-handled, and regular adult laryngoscopes, respectively.
Guideline 16.1: Essential Considerations in Handtool Design The effects of the task, the user, the device lifecycle, and the use of other equipment on medical hand tool use should be essential considerations during tool design.
An example of how context influences the use of a tool and consequently its design is the laryngoscope (Figure 16.20). Laryngoscopy is a procedure by which a breathing tube (endotracheal tube) is inserted through the mouth (or nose) into the windpipe (trachea). While both adult and pediatric patients may be cared for in the same location (i.e., the same operating rooms), the laryngoscope is used differently in these two patient populations. In the adult, the ability to elevate airway structures so as to visualize the tracheal opening typically requires the exertion of about 30 to 40 N of force, although a maximum force of nearly 80 N may be required on some occasions (Evans et al., 2003; Hastings et al.,
FIGURE 16.21 Adult laryngoscope with a handle diameter of 3 cm (~1.2 inches) and held in a power grip. The regular-sized handle permits application of higher forces (typically 30 to 50 N).
TAF-ER562X_WINGR-09-0401-C016.indd 666
FIGURE 16.22 Pediatric laryngoscope with a handle diameter of 2 cm (~0.8 inches) and held in a precision grip.
11/9/10 1:56:04 AM
Hand Tool Design
667
1996a; Rassam et al., 2005). In contrast, when performing laryngoscopy on a pediatric patient, much less force but more precision is required because of the smaller size and fragility of the tissues being manipulated. The force required for pediatric patients is about 20 N (Hastings et al., 1996a). This difference is reflected in laryngoscope design. The adult laryngoscope handle has a larger diameter, is heavier, and can be grasped by the entire hand in a power grip (Figure 16.21). The pediatric laryngoscope has a narrower handle, weighs less, and is held by the fingertips, providing more precise and gentler manipulation of airway structures (Figure 16.22).
16.2.2 LOCATION AND ENVIRONMENTAL FACTORS Medical tools are used in myriad locations, each of which represents a unique use environment (Table 16.2) (Jagger and Perry 2000). Refer to Chapter 3, “Environment of Use,” for additional information about the effects of the environment on device use. Considerations that arise as a function of location and the environment of use include the following: • • • • • •
Maneuverability and clearance Ease of access Lighting Noise Temperature Availability of resources (includes auxiliary equipment, personnel, and reference materials) • Effects of motion and vibration • Magnetic fields For example, in the operating room, laryngoscopy and intubation are commonly performed by an anesthesiologist, with good access to the patient’s head and airway, good lighting, a still and quiet setting, and with resources readily available to manage complications. In contrast, this same procedure, when performed outside the operating room, may be done by a variety of individuals (e.g., nonanesthesiologist physician or paramedic). It is not uncommon that the amount of room in which to perform laryngoscopy and intubation is limited (e.g., by room size, beds, or other personnel), and access to the patient’s head and airway may be more constrained. Illumination may be suboptimal, as may be access to resources to manage problems or complications. Other medical problems may demand simultaneous consideration (e.g., during a cardiopulmonary resuscitation), or laryngoscopy may need to be performed in a moving vehicle. Consideration of the use environment, as in this example, reveals considerable information that influences laryngoscope design, including the material used, the size and shape of the handle and blade, the brightness of the light emitted and its power source (routinely a battery, either alkaline or rechargeable), the ease of cleaning it (as opposed to using disposable laryngoscopes), device reliability (e.g., the choice between screw-in bulbs versus fiber-optic light sources), and cost (Evans et al., 2003; Hastings et al., 1996b; Rassam et al., 2005). Environmental factors include extremes of temperature and vibration and, in the case of magnetic resonance image (MRI) scanning, exposure to strong magnetic fields. The temperature may be related to the instrument itself or the area in which the work is being done. Cold temperatures can affect blood flow and nerve function. For example, the cold
TAF-ER562X_WINGR-09-0401-C016.indd 667
11/9/10 1:56:05 AM
668
Handbook of Human Factors in Medical Device Design
temperatures of the operating room have been implicated in the occurrence of carpal tunnel syndrome in anesthetists (Diaz, 2001). Vibration may be present in the environment and affect the entire body (whole-body vibration, e.g., in a moving vehicle such as a helicopter), or it may originate from a powered hand tool and affect only a hand and arm (segmental vibration) (Cacha, 1999). Vibration can be a by-product of hand tool operation or even the desired action, as is the case with plaster cast removal tools (Radwin et al., 1992). Work with handheld vibrating tools has been linked to various neurologic, vascular, and musculoskeletal disorders (Armstrong et al., 1987; Färkkilä et al., 1979; Pelmear and Taylor, 1992; Strömberg et al., 1997) and can lead to a complex of symptoms known as hand–arm vibration syndrome. Readers should refer to published sources for detailed guidance regarding tool-associated hand and arm vibration (American Conference of Government Industrial Hygienists, 2005; American National Standards Institute, 1986; International Organization for Standardization, 2001). Vibration may increase the risk of chronic tendon and nerve disorders by increasing the force exerted in repetitive manual tasks. Radwin et al. (1987) demonstrated that hand–arm vibration exposure (similar to that associated with the use of power hand tools) increased the force required to grip and operate hand tools. The required grip force increased when the hands were exposed to 40 Hz vibration during a 1-minute exertion, compared with grip forces in an equivalent task with no vibration or with vibration at 160 Hz. The transmission of vibration and its effects on the body depend on the coupling between the vibration source and the hands, vibration direction, and the vibration frequency characteristics. Energy absorbed by the hand–arm system, when exposed to sinusoidal vibration, exhibited a local maximum for absorption in the range of 50 to 150 Hz (Burstrom and Lundstrom, 1988). Guideline 16.2: Design to Withstand Abuse and Malfunction Tools should be designed to tolerate the expected environmental abuse and minimize the possibility of tool malfunctions and accidents.
Guideline 16.3: Storage Requirements Storage requirements, and the need for auxiliary equipment should be considered.
Guideline 16.4: Cleaning Tools should allow for proper cleaning, disinfection, and sterilization when required.
Guideline 16.5: Lighting Conditions The effects of ambient lighting on tool use should be considered. Tools intended for use in low-light situations should not require vision to operate.
Guideline 16.6: Tool Effects on Environment Exposure to extremes of temperature, excessive vibration, and airflow as a result of tool use should be avoided.
Guideline 16.7: Handle Temperature Tool handle temperatures should not cause excessive heating or cooling of the hand during use. Range of handle temperatures is between 17ºC and 25ºC.
TAF-ER562X_WINGR-09-0401-C016.indd 668
11/9/10 1:56:05 AM
Hand Tool Design
669
Guideline 16.8: Vibration Dampening Techniques such as vibration-attenuating handles should be used to reduce the transmission of vibration to the hand (Helander, 1995).
Guideline 16.9: Design for Magnetic Fields Tools intended for use in magnetic fields, such as the MRI scanner, should not be ferromagnetic. Tool function should not be affected by magnetic fields, nor should the tool be attracted by the magnetic field, as it could act as a projectile.
16.2.3 THE END EFFECTOR (TOOL INTERACTIONS WITH ANATOMY) The end effector is that part of a tool that interacts with the target, the item on which the tool acts. In the medical setting, this is usually patients’ tissues. The type of end effector used depends primarily on the clinical task to be performed (e.g., cutting, dissecting, grasping) (see Table 16.4). Note that end-effector design should consider not only the clinical target but also other factors, such as the type of work surface (e.g., vertical vs. horizontal) and its elevation, the horizontal distance to the target (Ulin et al., 1993), the type of tissue or material being worked on, other task-related requirements, and personal protective equipment worn by the user. Some general guidelines for the end effector follow. Guideline 16.10: Effects Limited to Target The end effector should be atraumatic to surrounding tissues so that while performing its function, it does not inadvertently damage surrounding tissues.
Guideline 16.11: Minimize Inadvertent Tissue Damage The end effector should not inadvertently damage the target tissues.
Guideline 16.12: Secure Connection to Tool Body The end effector should be secured to the tool’s main body so that it cannot become loose or dislodged.
16.2.3.1 Grasping Instruments Many medical tools are used to grasp objects. Grasping tools are used to do the following: • Grasp (e.g., needles or tissues; when grasping needles, the tool is commonly known as a needle driver) • Supply (feed) materials or tissues, such as suture materials or gauze sponges • Tie (e.g., after placement of a suture by a needle) The duration of effort will vary with the task and tool design. The longer the duration, the greater the likelihood of muscle fatigue. Grasping typically has the shortest and tying the longest duration of effort. Guideline 16.13: Tissue Traction The end effector of grasping tools should be able to provide sufficient traction on tissues to allow them to be securely held, thereby reducing the grip force needed at the handle.
TAF-ER562X_WINGR-09-0401-C016.indd 669
11/9/10 1:56:05 AM
670
Handbook of Human Factors in Medical Device Design
Guideline 16.14: End-Effector Indentations Where possible, indentations for needle placement should be provided to prevent twisting of the suture needle. This may be accomplished by the use of inserts, coatings, or milling (Patkin, 1977).
16.2.3.2 Force Output from Tool A hand tool is often a force multiplier. There is a trade-off between the force generated and precision. Precision tasks require light, easy-to-use instruments, typically held in a pinch grip. Such tools are not used to exert significant forces. Higher-force manual tasks require larger instruments that are heavier and are generally held in a power grip. Excessive grasping forces decrease haptic feedback and sensitivity to tissue condition or damage and, for precision tasks, increase the amount of tremor. Guideline 16.15: Input–Output Force Relationship The tool’s force output should be balanced relative to the hand force input. This provides comfortable grasping forces while avoiding injury to tissues due to excessive force.
Guideline 16.16: Optimal Handle for Force Delivery For tasks where significant force must be exerted, the handle should allow a power grip, with optimal grip span (handle separation) (see Figures 16.34 and 16.35) and surface texture.
Guideline 16.17: Optimal Handle for Precision Use For tasks where significant precision must be applied, the handle should be sufficiently small to allow a precision (pinch) grip (see handle diameter guidelines below).
Guideline 16.18: Use of External Power Source For tasks requiring both precision and force, an external source of power should be considered. This will allow the tool rather than the hand to generate the force. The operator can then use small hand muscles to guide and position the tool more precisely.
16.2.4 CONSIDERATIONS FOR THE WHOLE TOOL The most basic requirement of any tool is that it should allow the user to effectively perform the intended function(s). The main objective is to have a good fit between four system elements: the user, the tool, the environment, and the patient. Keep in mind that fitting the tool to the user refers not only to the hand, arm, and shoulder but also the user’s posture(s) as well as work capacity and skill level. Some general guidelines for the whole tool follow. Guideline 16.19: Tool Size Tools should be properly sized to the user’s body dimensions and accommodate differences in handedness, strength, and work capacity (i.e., allow for the expected range of differences between users).
TAF-ER562X_WINGR-09-0401-C016.indd 670
11/9/10 1:56:05 AM
Hand Tool Design
671
Guideline 16.20: Size and Forces Tools should be large enough to provide leverage yet withstand the forces applied. A tool that is too small must be held by the fingers and cannot be held securely. The user will not be able to maintain a firm grasp on the tool when large forces must be applied.
Guideline 16.21: Grip Diameter If lifting or grasp points are provided, the fingers should be able to wrap at least 270 degrees around the surface of the grip, as the tool will be more comfortable to hold and better secured against sudden shifts in position (Fraser, 1980; Greenberg and Chaffin, 1978).
Guideline 16.22: Grip Shape Any form of shaping or contouring of the handle to a specific hand (e.g., ridges, valleys, fluting, indentations, and so on) is undesirable and should be avoided, as these may not fit the range of hand sizes or allow a left-handed person to grip a right-handed tool (Fraser, 1980; Helander, 1995).
Guideline 16.23: Maintainability Product life cycle should be considered where critical for ease of use or safety, as properly maintained tools are less likely to cause injury (see Chapter 15, “Device Life Cycle”). For example, for the design of forceps, tips that are kept sharp will reduce the required grip force.
Specific areas to consider for the whole tool’s design include tool weight, center of gravity, and safety of use as well as the following handle characteristics: • • • • • • •
Angle Shape Length Diameter Cross section Material Texture
16.2.4.1 Handle Angle Angulation should take into account the grasp and the axis of function. The line of transmitted force passes through the fingers, the carpal bones, the radius, and up the arm (Figure 16.23). The axis around which the hand operates is that of the pointing index finger. Ideally, the tool handle will be angled so that neutral hand and arm positioning can be maintained, as this maximizes strength, which is preferred (Dababneh and Waters 1999). Guideline 16.24: Preserve Neutral Postures The design should help to preserve neutral postures by angulating the tool handle. Follow the design maxim “Bend the tool, not the wrist” (Helander, 1995; Mital and Karwowski, 1991).
Guideline 16.25: Angle Aligned with Force Bent or angled handles should be aligned so that force is exerted in same direction as the forearm’s movement while keeping the arm as close to the torso as possible (Greenberg and Chaffin, 1978; Woodson et al., 1992).
TAF-ER562X_WINGR-09-0401-C016.indd 671
11/9/10 1:56:05 AM
672
Handbook of Human Factors in Medical Device Design
FIGURE 16.23 Laparoscopic instrument held to illustrate the force passing in a line up through the finger, wrist, and forearm.
Guideline 16.26: Fit Grasp Axis Angled (i.e., pistol-grip) handles should fit the axis of the grasp and not require the clinician to assume awkward wrist, arm, or shoulder postures during anticipated use. One common recommendation is for the handle to be at about 80 degrees to the horizontal (Figure 16.24).
16.2.4.2 Handle Shape Shape can influence holding, grasping, and positioning a tool as well as tactile feedback during a tool’s use, but handle shape is not necessarily a significant predictor of how much exertion the user will have to provide. The preferred shape (e.g., angled vs. in-line) is related to characteristics of the target (e.g., operating room table height) and the task (Table 16.9) (Quick et al., 2003; Ulin et al., 1993). Several handle shapes are currently in common use, including pistol-grip and angled handles, in-line (axial), cylindrical (or nearly so), ball, and T-shaped (Figures 16.25, 16.26, and 16.27). Designers should consider that finger loops (used for multiple fingers) and finger rings (used for single fingers) are intended for fine manipulation. Therefore, if a large grasping force is required (e.g., to use a harmonic scalpel), the handle controls should preferably not include finger loops or rings. Ring-handled and other flat tools may also be more difficult to retrieve from a flat surface, which can prolong performance of a task.
FIGURE 16.24 Pistol-grip (angled) handle with the handle at roughly 80 degrees to the horizontal.
TAF-ER562X_WINGR-09-0401-C016.indd 672
11/9/10 1:56:05 AM
Hand Tool Design
673
TABLE 16.9 Recommended Handle Type as a Function of Work Surface and Target Height Work Surface
Below the Waist
Target Height at the Elbow
NR* Pistol
Vertical Horizontal
Pistol In-line
Midchest or Above NR In-line
NR, not recommended Note: These recommendations have not been established for and should not be applied to laparoscopic instruments.
*
Surgical instruments may have finger loops, such as scissors, or be squeezed between the fingers, such as forceps. Tools used for precision tasks commonly have cylindrical handles, such as for dental or microsurgical instruments. Laparoscopic instruments exhibit a wide variety of handle types and are discussed separately in the section “Special Considerations.” Guideline 16.27: Distribute Forces Tools that will be gripped in the palm should have a shape that distributes forces over as large a force-bearing area as possible, to minimize high pressure being exerted against the fingers or palm (Patkin, 1967; Tichauer and Gage, 1977). Surfaces that will be grasped should be rounded to a radius of at least 3 mm, although a 6- to 9-mm radius is preferred (Greenberg and Chaffin, 1978).
Guideline 16.28: Cylindrical Handles Cylindrical handles should be used if torque is to be applied (e.g., turning a screwdriver), although the grip may be more secure if the cylinder is somewhat flattened (Patkin, 2001).
Guideline 16.29: In-Line Handles for Precision In-line (including cylindrical) handles should be used when precision is required.
FIGURE 16.25
Pistol-grip handle.
TAF-ER562X_WINGR-09-0401-C016.indd 673
11/9/10 1:56:06 AM
674
FIGURE 16.26
Handbook of Human Factors in Medical Device Design
Retractor with an angled handle.
Guideline 16.30: Pistol Grip Handles Pistol-grip handles should be used when high forces are required to activate controls.
Guideline 16.31: Tapering The pistol-grip handle should taper toward the bottom because of foreshortening of the fingers (Woodson et al., 1992).
Guideline 16.32: Ball Handles Spherical (ball) handles should be used for nonpowered tools when the combined application of force and torque is required.
Guideline 16.33: Finger Ring Size Finger rings should be at least 30 mm long and 24 mm wide (van Veelen et al., 2001).
16.2.4.3 Handle Length The handle’s length will determine, in part, the force that the user can exert as well as the stability of the tool in the hand. If a handle is too short, it will be grasped by less than the entire hand, and the user will not be able to exert maximal force. It may also result in pressure points in the palm. Excessive handle length may adversely affect tool functionality.
FIGURE 16.27 Spinal surgery instrument set with interchangeable handles, showing in-line (axial), cylindrical, ball, and T-shaped handles.
TAF-ER562X_WINGR-09-0401-C016.indd 674
11/9/10 1:56:06 AM
Hand Tool Design
675
Guideline 16.34: Handle Lengh for Largest Hands The handle length should accommodate hand width dimensions of the largest-size hands (e.g., the 95th percentile) for the given user population. Several authors recommend a minimum handle length of at least 10 cm for nonmedical hand tools (Cacha, 1999; Helander, 1995; Mital and Karwowski, 1991).
Guideline 16.35: Length for Power Grips For a power grip, the minimum handle length should extend across the entire breadth of the palm to provide for proper grasping by the hand or fingers. With this handle length, grip force is not compromised because stability is provided by resting on the palm or another part of the hand. Helander (1995) recommends a minimum handle length of 12.5 cm for nonmedical hand tools. If substantial leverage is needed (e.g., for a hammer or wrench), the handle may need to be longer.
Guideline 16.36: Length for Precision Grips Smaller handles held in an external precision grip should extend to the apex of the thumb cleft (Patkin, 1977). Tools held in an internal precision grip should be long enough to extend past the palm but not so long as to contact the wrist (Mital and Karwowski, 1991).
There are numerous examples of instruments that do not satisfy these handle length guidelines. In discussions with surgeons and other medical tool users, some find the shorter-handled tools harder to use, while others prefer them, again underscoring the need to conduct studies of the intended user population and how a tool will be used in a specific context (Patkin, 1981). 16.2.4.4 Handle Diameter (Cross-Sectional Size) The handle’s diameter influences the magnitude of grip force. Handles that are intended for a power grip but are too small (less than about 2.5 cm in diameter) may require excessive force to grasp, while those that are too large (more than approximately 4 cm) may be difficult to adequately grasp. Ideally, a variety of handle sizes will be available to fit hands of varying sizes; however, this is often not possible. Handles for precision tools can likewise be either too small, in which case it will be difficult to grasp the tool, or too large, in which case control will be lost. Guideline 16.37: Maximize Surface Contact Handle diameter should be large enough to maximize hand and finger surface contact when held with a power grip. The recommendations vary between authors (see Table 16.10).
Guideline 16.38: Torque Proportional to Diameter When using a power grip to apply torque, larger-diameter handles should be provided (within the recommended range). The torque produced increases as the handle diameter increases (Mital and Karwowski, 1991).
Guideline 16.39: Diameter for Fine Manipulation For fine manipulation tasks, such as in microsurgery or dentistry, different authors recommend different handle diameters between 0.5 and 1.6 cm (see Table 16.11).
TAF-ER562X_WINGR-09-0401-C016.indd 675
11/9/10 1:56:07 AM
676
Handbook of Human Factors in Medical Device Design
TABLE 16.10 Recommendations for Handle Diameter for General Use (Power Grip) Handles Reference
Recommendation
Mital and Karwowski (1991) Woodson et al. (1992) Helander (1995) Cacha (1999) Patkin (2001)
2.5–5.0 cm 2.5–3.8 cm ≤5 cm 3–4.5 cm (for a power grip) 3–4 cm (for maximum power)
Guideline 16.40: Diameter for Precision Rotation Precision tools that require rotation by the fingers should be sufficiently small in diameter to better maintain tool control. Patkin (1977) recommends a range of 0.5 to 1 cm. A larger diameter will result in separation of the fingertips and loss of control. If smaller, the tool will not be held securely when held by the thumb and first two fingers (Figure 16.28).
16.2.4.5 Handle Cross-Sectional Shape As with length and diameter, the handle’s cross-sectional shape will affect grip force and also comfort. In general, cylindrical, conic, or oval cross sections are preferred. Even when a handle is in the form of a stirrup, a T-shape, or an L-shape, the part held by the hand will commonly be in the form of a cylinder or cone (Fraser, 1980). Modifications of these shapes, such as the hexagonal, triangular, or flattened cylindrical handle, may provide improved grip as well as better alignment feedback of the end effector in relation to the handle (Helander, 1995). Guideline 16.41: Avoid Pinch and Pressure Points Pinch points, sharp edges, or ridges should be avoided, as these may exert excessive pressures on the hand (Mital and Karwowski, 1991; Woodson et al., 1992). Edges should be rounded to a radius of 0.8 mm and corners to 1.6 mm (Fraser, 1980). If rounding is not feasible, then consider covering the handle with a plastic or rubber overlay (note that if sterility is a concern, other materials may have to be used for the overlay).
TABLE 16.11 Recommendations for Handle Diameter for Fine Manipulation Reference Patkin (1977) Mital and Karwowski (1991) Helander (1995) Cacha (1999)
TAF-ER562X_WINGR-09-0401-C016.indd 676
Recommendation 0.5–1.2 cm 0.6–1.3 cm 0.8–1.3 cm 0.8–1.6 cm
11/9/10 1:56:07 AM
Hand Tool Design
FIGURE 16.28
677
Cylindrical precision instrument held between three fingers.
Guideline 16.42: Hammer Tool Handle Shape For hammer-type tools, semiflat sides or an oval shape should be used, as they will make it easier to guide the stroke and control lateral positioning (Woodson et al., 1992).
Guideline 16.43: Shape to Maximize Torque For production of torque, noncylindrical handles (e.g., oval, knurled, flattened) should be used, as they increase the amount of torque that can be produced relative to a cylindrical handle (Mital and Karwowski, 1991).
Guideline 16.44: Pommels and Flanges Retractors and similar tools should have a pommel (a knob on the handle) or flange to prevent the hand from slipping either forward onto the tool, or backward, off the handle (Figure 16.29) (Brearly and Watson, 1983; Patkin, 1967).
Guideline 16.45: Shape for Precision A cylindrical handle should be used for fine manipulation when using precision instruments that require gripping by the fingertips. However, a flat handle may be simpler and less expensive (Patkin, 1977).
FIGURE 16.29 Retractor with a pommel on the handle to prevent the instrument from slipping out of the hand when axial forces are applied.
TAF-ER562X_WINGR-09-0401-C016.indd 677
11/9/10 1:56:07 AM
678
Handbook of Human Factors in Medical Device Design
16.2.4.6 Handle Material While many materials can be used for medical tools, specific design constraints may limit material selection. For example, wooden handles are difficult to sterilize and cannot be used on surgical instruments. The following guidelines can help with selection of handle materials for hand tools. Guideline 16.46: Cleaning and Sterilization Handle materials should be able to withstand methods of cleaning and sterilization (heat, gas or chemicals) as well as have the ability to ensure complete biological sterility.
Guideline 16.47: Non Porous Materials The handle material should prevent the retention of biologic products (i.e., bacteria, viruses, and other toxic agents) (Woodson et al., 1992).
Guideline 16.48: High Coefficient of Friction Handle materials should have a high coefficient of friction to prevent slip (Woodson et al., 1992). If hand slippage is a major design criterion, consider covering metal handles with another material (rubber, plastic, or leather) to improve grip and reduce grip force and slippage (as permitted by sterility requirements). Alternatively, the surface texture can be altered to improve grip (see Section 16.2.4.7).
Guideline 16.49: Minimize Vibration Transmission Handle material design should minimize vibration transmitted to the hand.
Guideline 16.50: Neutral Handle Temperature Handle materials should minimize transmission of excessively cold or hot temperatures to the hand.
Guideline 16.51: Non Conductive Handle The handle material should not conduct electricity (Woodson et al., 1992).
The many tool handles made of conductive metal are examples of current practice that is not consistent with best practice. Guideline 16.52: Minimize Weight for Precision For precision tools, materials should minimize weight while still providing sufficient strength and rigidity. The use of lightweight materials (e.g., composites) allows a larger-diameter handle (e.g., 1 cm instead of 0.5 cm), makes the tool easier to grasp, and helps reduce the force needed to hold (pinch) the instrument securely between the fingers (Patkin, 1977).
16.2.4.7 Handle Surface and Texture The finish of the tool handle affects the surface and texture. Various finishes are applied to hand tools to improve the gripping surface and prevent slippage. Surface finishes can also be used to indicate hand placement on the tool and intended points of interaction with it.
TAF-ER562X_WINGR-09-0401-C016.indd 678
11/9/10 1:56:07 AM
Hand Tool Design
679
Guideline 16.53: Rounded Handle Surfaces Rounded surfaces are preferred; sharp edges and pressure points should be avoided (Dababneh and Waters, 1999; Woodson et al., 1992).
Guideline 16.54: Avoid Glare-Inducing Surfaces Smooth, polished metal surfaces (or other glossy material) can reflect light and cause glare and should be avoided (Greenberg and Chaffin, 1978; Patkin, 1977; Woodson et al., 1992).
Guideline 16.55: Texured Surfaces Textured surfaces (milled or otherwise) should be used to provide sufficient friction for an adequate grip (Dababneh and Waters, 1999; Greenberg and Chaffin, 1978; Woodson et al., 1992). Textured as opposed to smooth-surfaced handles will improve grasp and reduce contact forces and static loading. The method used to provide texture should not produce hidden sources of contamination.
16.2.4.8 Tool Weight and Center of Gravity A tool’s weight affects how quickly the user’s upper extremity experiences muscle fatigue. One exception may be power tools, where increased weight and/or faster completion of the task may reduce the amount of force or exertion required of the operator (Woodson et al., 1992). A tool’s center of gravity is a critical design attribute affecting, in particular, reliable and accurate end-effector function. Guideline 16.56: Center of Gravity The center of gravity should be located where the tool is grasped to prevent rotation of the tool in the hand. That is, the tool’s weight should balance close to the point of support (Greenberg and Chaffin, 1978; Woodson et al., 1992).
Guideline 16.57: Maximum Tool Weight The maximum tool weight should be no more than 1.8 kg for precision tools and no more than 2.3 kg for tools held with one hand and used to exert force (Cacha, 1999; Helander, 1995). Note that these recommendations are based on guidelines intended for nonmedical tools and may be too high for many medical tools, such as precision (e.g., microsurgery) instruments. Tools that are used frequently or for long durations should generally be lighter.
Guideline 16.58: Two-Handed Grasp for Heavier Tools A two-handed grasp should be provided to reduce fatigue associated with lifting or holding tools that weigh 2.2 kg or more (Greenberg and Chaffin, 1978). This reduces the amount of exertion needed to make the tool perform as desired. Designs requiring two-handed operation should be evaluated in the context of user needs and other requirements (e.g., positioning) related to tool use. Alternatively, a balancer can be used for tools weighing more than 2.2 kg (Dababneh and Waters, 1999).
Guideline 16.59: Increased Weight for Increased Forces When a tool is intended to apply a significant force (e.g., hammer or torque wrench) or has significant forces applied to it during use (e.g., powered tools with axial forces), handle weight may need to be increased to ensure consistent operator control and reliable tool function.
TAF-ER562X_WINGR-09-0401-C016.indd 679
11/9/10 1:56:08 AM
680
Handbook of Human Factors in Medical Device Design
FIGURE 16.30 Surgical drill with cord for pneumatic power. Cords attached to hand tools can affect usability and safety. For example, tethering can alter the center of gravity, constrain tool movement, and increase the risk of loss of sterility.
A tool can also be suspended or connected to other devices (or a power source). The weight of cords or other connections can change the center of gravity such that the tool may rotate undesirably during use. Tool weight also affects the ability to position and hold the tool steady. Heavy tools will exacerbate tremor, which is particularly significant during microsurgery. Additionally, cord placement and strain relief requirements will affect tool usability. Battery power may be used to avoid the need for a power cord and the problems associated with it (e.g., entanglements) (Figures 16.30 and 16.31) (Woodson et al., 1992). Guideline 16.60: Tethering Affects Center of Gravity If a hand tool must be tethered, the center of gravity should be maintained in the palm.
Guideline 16.61: Strain Relief The cord joint and any other strain relief should be placed so as to minimize interference with the hand(s) using the tool.
FIGURE 16.31 Battery-powered surgical drill. Battery power can eliminate cords but increase tool size and weight, affecting the center of gravity.
TAF-ER562X_WINGR-09-0401-C016.indd 680
11/9/10 1:56:08 AM
Hand Tool Design
681
Guideline 16.62: Minimize Weight to Decrease Tremor Tools used for precision tasks should be light enough that the magnitude of tremor and the required gripping force are minimized (Patkin, 1981).
16.2.4.9 Safety Hand tool use must be safe for both the user and patient. This includes safeguards against accidental or other unintended operation. Guideline 16.63: Safety Indicators Safety indicators to warn of hazardous conditions should be provided.
Guideline 16.64: Provide Safety Controls Power tools should be equipped with a safety control, such as a constant-pressure switch (shuts off power when pressure is released) or control switch (e.g., a foot pedal that must be pushed to activate the tool) (Occupational Safety and Health Administration [OSHA], 2002; Woodson et al., 1992).
Guideline 16.65: Visible Hand Controls Hand controls should be visible from the operator’s normal working position.
Guideline 16.66: Critical Control Conspicuity Critical tool controls (e.g., safety switches) that require rapid identification should be readily identifiable by users during tool use, for example, by color and shape coding (Woodson et al., 1992).
Guideline 16.67: Safety Guards Guards should be placed around moving or other parts of power tools to protect the user and others from moving parts and flying particles (OSHA, 2002; Woodson et al., 1992).
Guideline 16.68: Electrical Safety Contact surfaces should be insulated to prevent electrical shock (Woodson et al., 1992). Electric tools should have three-wire cords with a ground, double insulation, and/or lowvoltage electrical power to minimize the risk of shock (OSHA, 2002).
Guideline 16.69: Minimize Noise Noise generated by the tool during its operation should not exceed 85 dBA over 8 hours of exposure (Mital and Karwowski, 1991). Lower sound levels may be necessary, depending on the use environment.
Guideline 16.70: Prevent Inadvertent Disconnection Pneumatic tools (i.e., tools powered by compressed gas) should be securely fastened to the gas hose with either a short wire or a positive locking device to prevent inadvertent disconnection. Otherwise, the tool and/or hose can whip about and produce damage or injury (OSHA, 2002; Woodson et al., 1992).
TAF-ER562X_WINGR-09-0401-C016.indd 681
11/9/10 1:56:09 AM
682
Handbook of Human Factors in Medical Device Design
Guideline 16.71: Tool Location Identification Tools should have a means of signaling their location if there is a risk that they could be left inside a body. Radio-frequency identification tags have been proposed as one way to accomplish this (Macario et al., 2006; Schwaitzberg, 2006).
16.2.5 USER CHARACTERISTICS AND RELATED DESIGN CONSIDERATIONS Medical tools are used by a wide variety of individuals, each having different attributes with regard to anthropometry, strength, and other personal factors. The same tool may be used in different settings, and the user may differ significantly depending on the setting. For example, in the hospital, a bandage scissors may be used by physicians, nurses, technicians, and therapists. Other settings, such as the ambulance (emergency medical technicians), clinic (physicians and nurses), or the home (patients, family caregivers, and home health care providers), involve other types of users. These individuals may vary considerably in their physical characteristics (e.g., hand size; see Chapter 4, “Anthropometry and Biomechanics”), knowledge, and skills and training and how they will use the tool(s). Some user characteristics may be interrelated (e.g., gender, hand size, and strength; see Chapter 4, “Anthropometry and Biomechanics,” for more detail). Also, because the same tool may be used by different users, designers should consider the characteristics of all potential user types. Gender is a particularly important consideration, as a large proportion of tool users in health care are female. Studies of nonmedical domains have found that some equipment was unsuitable for use by at least 10% of the female workforce (Fraser, 1980). Reasons for tool unsuitability might include weight (too heavy), grip span (too far apart), size (need two hands to grasp), effort (too hard to squeeze), or fit (poor). Additional user-related factors that may affect tool use include users’ medical conditions (e.g., neurologic or arthritic conditions) and psychosocial factors. 16.2.5.1 Posture The effects of body posture and position during tool use are mediated by both physical (i.e., force and exertion) and temporal considerations, such as duration and frequency. Ideal body position is rarely maintained for more than brief periods. Extreme postures and nonneutral positions will affect comfort, stress, and fatigue; affect the efficiency with which a task is done; and exacerbate any problems associated with the use of force, duration of effort, and frequency of repetition (Buffington et al., 2006). Tool designs that induce or promote awkward wrist and shoulder positions may lead to compensatory neck, trunk, and lower-extremity positions that can themselves be uncomfortable or promote user injuries. The effects of repetition and duration of exposure (e.g., shift length) may be partially mitigated by rest periods. Inadequate recovery during longer shifts or an unbalanced work–rest cycle may increase the risk of discomfort or injury. For example, the frequency with which sonographers perform ultrasound examinations influences the likelihood of their developing work-related musculoskeletal disorders (Smith et al., 1997). Guideline 16.72: Avoid Extreme or Awkward Postures Tools should be designed to be used with the hand, upper extremity, and body in a neutral and comfortable position (Dababneh and Waters, 1999), and not require extreme or awkward finger, hand, arm, neck, or back positions.
TAF-ER562X_WINGR-09-0401-C016.indd 682
11/9/10 1:56:09 AM
Hand Tool Design
683
FIGURE 16.32 Needle driver with ratchet mechanism that allows setting a fixed tool position, avoiding the need to squeeze the tool for prolonged periods.
Guideline 16.73: Avoid Sustained Positions or Postures The need to maintain the same position or posture for prolonged periods of time should be avoided. For example, some tools have ratchets to avoid having to squeeze them for extended periods of time (Figure 16.32).
16.2.5.2 Shoulder Shoulder strength and fatigue time vary with the degree of abduction. Strength is relatively constant for abduction up to 90 degrees, but the time to fatigue decreases rapidly when the shoulder is abducted more than 30 degrees (Salvendy, 1997). Similarly, the time to fatigue decreases as the shoulder is flexed, whether to reach forward or up. Hence, even though strength may be adequate in flexion and abduction, with any degree of flexion and with abduction beyond 30 degrees, muscles will fatigue more rapidly (Salvendy, 1997). For example, shoulder discomfort is still commonly seen in dental hygienists, as they must hold their arms in both flexion and abduction while working, despite improvements in hand tool design improvements (e.g., better and lighter materials allow larger handles that are easier to grasp) (Murphy, 1998). In some cases, elbow supports significantly decrease fatigue and allow larger degrees of flexion and abduction to be tolerated. Guideline 16.74: Avoid Sustained Shoulder Abduction The need for sustained abduction of the shoulder during tool use should be minimized.
Guideline 16.75: Avoid Shoulder Flexion or Extension The need for shoulder flexion or extension during tool use should be minimized.
Guideline 16.76: Elbow Supports If shoulder positions outside the recommended range are expected, elbow supports or other design solutions should be provided to minimize neck, shoulder, and upper-arm fatigue.
Guideline 16.77: Avoid Use with Arms Elevated Tool length should not require the user to work with the arms above midchest height, as this has been associated with muscle fatigue and shoulder disorders (Ulin et al., 1993).
TAF-ER562X_WINGR-09-0401-C016.indd 683
11/9/10 1:56:09 AM
684
Handbook of Human Factors in Medical Device Design
16.2.5.3 Elbow Strength associated with supination has been found to be about 20% to 30% greater than the strength to pronate (Nordin and Frankel, 2001). That is, the arm is stronger when performing a supination movement, such as turning a screwdriver clockwise (i.e., tightening a screw) with the right hand, than when performing a pronation movement, such as turning a screwdriver clockwise with the left hand. Average strength is about 30% greater with the elbow flexed as opposed to extended. Males are, on average, about 40% stronger than females in elbow strength testing. There is increased flexion strength with the elbow at 90 degrees, and decreased forces on the elbow are seen in this position. Tools that require elbow positions other than 90 degrees flexion likely will require more physical effort to operate. Finally, as is true for the shoulder, even small weights can dramatically increase the muscle exertion required to maintain arm position, thus increasing muscle fatigue (Nordin and Frankel, 2001). Guideline 16.78: Elbow Flexion Preferred Tool design should support the preferred elbow position of about 90 degrees of flexion when loads are supported in the forearm.
The preferred elbow position may be altered by the user for a variety of reasons. For example, aging surgeons with presbyopia (farsightedness that typically occurs in middle and old age) and with the operating room table at normal height (where elbow angle would be approximately 90 degrees) would not be able to focus their vision on the surgical target. By lowering the table height, with resulting elbow positions greater than 90 degrees, the older surgeon is able to focus and still utilize his or her instruments. The slight mechanical disadvantage is outweighed by the need to clearly see the operative field. 16.2.5.4 Wrist and Hand Most of the power-producing muscles for the hand are located in the forearm. Force is transmitted to the hand by long tendons that pass through the wrist. The result is a small and compact hand that can either generate substantial force or perform delicate, precise tasks. Stresses on tendons as they travel through the wrist can increase the risk of a variety of musculoskeletal disorders, such as tendonitis and carpal tunnel syndrome. Wrist deviations from neutral decrease the volume of the carpal tunnel, increase tendon friction, decrease the grip force that can be generated, and require the muscles to work harder. Guideline 16.79: Maintain Neutral Wrist Position The need for wrist deviations from the neutral position should be minimized. Ideally, the hand and forearm longitudinal axes should be aligned.
Guideline 16.80: Minimize Twisting Hand Motions The need to use extreme twisting hand or wrist motions should be minimized.
Guideline 16.81: Avoid Force in Pinch Grip The need for forceful pinch (precision) grips (i.e., the need to exert excessive force in a pinch grip when performing precision tasks, such as during microsurgery) will increase hand tremor and should be avoided.
TAF-ER562X_WINGR-09-0401-C016.indd 684
11/9/10 1:56:09 AM
Hand Tool Design
685
16.2.6 GRIP While the shape of the tool will influence and suggest a particular type of grip, ultimately actual tool use will determine how the tool is held. For example, it is normally more convenient to hold a cylindrical handle in a power grip, but if the tool is to be used for a precision task, it will be held in a precision grasp. Similarly, a tool handle that would normally be held in a precision grip will instead be held in a power grip if its use requires a forceful exertion. The type of grip may vary depending on different uses of the same tool. For example, a linear stapler is typically positioned using a precision grip, which is then shifted to a power grip to close and fire the device (Figure 16.33). An alternative design might preclude this mid-task grip change. Grip strength is affected by gender, wrist position, grip opening (span), the number of fingers utilized, the hand that is used, and whether gloves are being worn. Grip strength also relates to the force requirements as well as the frequency and duration of effort. 16.2.6.1 Gender Fraser (1980) reports that American males have grip strength of 42 to 60 kg, whereas female British and American workers have a grip of 25 to 33 kg. Possible reasons for reduced grip strength in female users, aside from being inherently less strong, are that they have smaller hands with smaller grip spans (see Section 16.2.6.3). 16.2.6.2 Wrist Position When analyzed as a function of wrist position, grip strength is greatest at 20 degrees of extension and approximately 20 degrees of ulnar deviation; hence, that is the usual position for a power grip (Nordin and Frankel, 2001). Grip strength is lowest when the wrist has more than 40 degrees of flexion. Guideline 16.82: Minimize Wrist Deviation Tools should be designed to minimize the need for hand and wrist positions other than slight (<20 degrees) extension and slight (<20 degrees) ulnar deviation.
16.2.6.3 Grip Span Grip span describes the hand opening needed to grasp and squeeze a pair of handles together so as to actuate or control a tool (Figure 16.34). There are two aspects of grip span: (1) the
FIGURE 16.33 In-line DST stapler is held in a precision grip to position the device. After positioning, the grip is shifted to a power grip to close and fire the stapler. (Copyright © United States Surgical, a division of Tyco Healthcare Group LP. All rights reserved. Reprinted with permission.)
TAF-ER562X_WINGR-09-0401-C016.indd 685
11/9/10 1:56:09 AM
686
Handbook of Human Factors in Medical Device Design
Maximum grip forces
kg lb 54 120
36 80
Grip axis
27 60
5th percentile male
18 40
50th percentile female
9 20 0
FIGURE 16.34 Grip span during use of a surgical instrument.
50th percentile male
45 100
0
5th percentile female 0 0
1 2 3 4 25 51 76 101 Grip axis opening ( a ) Grip strength
5 in 127 mm
FIGURE 16.35 Grip force exerted as a function of grip opening. (Salvendy, G. Handbook of Human Factors and Ergonomics (2nd ed.), Reprinted from John Wiley & Sons, Inc. With permission.)
ability to open the hand widely enough to hold an object and (2) the strength to squeeze against resistance as the tool is used (i.e., the handles are squeezed together). As shown in Figure 16.35, grip strength is maximized within a narrow range of grip openings. Grip openings outside this range will limit the gripping force that can be applied. The ideal grip opening will depend on deviations from the neutral position. Deviations in wrist position away from neutral also alter the volume of the carpal canal, with a concomitant increase in tendon friction and increased risk of carpal tunnel disease. The measurement of grip span on pistol-grip configurations extends from the back of the handle across to the location of the middle (third) finger (Figure 16.36). One-size-fits-all spans may not adequately accommodate small-handed individuals (e.g., Asian women). If the span is too large for the hand, one-handed operation may be impossible, and the tool’s position may be more difficult to maintain. An example of this is the surgical stapler, when used by female surgeons with small hands (Berguer and Hreljac, 2004) Guideline 16.83: Larger Grip Span if Higher Grip Forces Grip span should be larger if high grip forces are required to grasp and hold an object. Greenberg and Chaffin (1978) and Cacha (1999) recommend that this grip span should be 6.3 to 8.9 cm. The maximum strength is exerted when a handle opening is about 7.6 cm (Figure 16.35).
Guideline 16.84: Reduce Squeeze Grip Spans To squeeze a two-handled instrument, the grip span should be reduced further (Figures 16.34 and 16.35).
Guideline 16.85: Maximum Span for Holding To merely hold an object, an absolute maximum span of about 13 cm is recommended (Greenberg and Chaffin, 1978), although <10 cm would be preferable.
Guideline 16.86: Maximum Closing Grip Force The required grip force to close a two-handled instrument should not exceed 88 N (22 pounds). This is the maximum grip force that can be produced by 95% of the female
TAF-ER562X_WINGR-09-0401-C016.indd 686
11/9/10 1:56:09 AM
Hand Tool Design
687
FIGURE 16.36 The measurement location of grip span for a tool with a pistol grip (line extending from back of handle to trigger). This device also has dual trigger controls; one that closes the jaws and the other that opens (releases) them. (Courtesy of Atricure, Inc. With permission.)
population (Figure 16.35) (Mital and Karwowski, 1991). However, smaller maximum grip forces, roughly less than 44 N, are preferable.
Guideline 16.87: Maximum Grip Force for Repetitive Pinching A closing force less than 10 N is recommended when an object is repetitively pinched between the fingers. This is roughly 20% of the weakest operator’s maximum pinch grip strength (Helander, 1995).
Guideline 16.88: Microsurgical Closing Forces Closing forces for microsurgery forceps should range between 40 and 100 g (Patkin, 1977).
16.2.6.4 Number of Fingers Utilized Grip strength also depends on the number of fingers utilized. For example, if only a single-finger trigger is used to activate the instrument, the forces applied are limited to the capabilities of that finger, and the user may be at risk for injuries such as trigger finger if sustained forces are required. Guideline 16.89: Adjust Grip Size for Fingers Used The size of the grip (length, diameter, and so on) should be adequate for the number of fingers that will be needed to hold or operate the tool.
Guideline 16.90: Finger Number Depends on Forces The number of fingers required to activate a control or apply forces should be appropriate to the magnitude, frequency, and duration of force application.
Higher forces, frequency, and duration of activation would indicate the need for multifinger design. For example, adult laryngoscopes are provided with both a standard length handle and a short version (Figures 16.37 and 16.38). Despite the force advantage provided by the standard (longer) length handle and the possibility of pressure points in the palm
TAF-ER562X_WINGR-09-0401-C016.indd 687
11/9/10 1:56:10 AM
688
Handbook of Human Factors in Medical Device Design
FIGURE 16.37 Regular (adult) handle laryngoscope grip.
FIGURE 16.38 Short-handled laryngoscope is used on obese or large-breasted patients despite its force disadvantage over the regular handle.
when using the short handle, the shorter handle is used during laryngoscopy on obese individuals or women with large breasts because the longer handle may preclude easy insertion of the laryngoscope into the mouth. The thinner pediatric laryngoscope handle allows use of a grip that facilitates greater fine motor control since lower forces are required (Figure 16.22). 16.2.6.5 Handedness Humans prefer to use their dominant hand for tool use, especially for precision tasks. Approximately 10% of the U.S. population is left handed (Fraser, 1980). Handle design should accommodate both left- and right-handed workers whenever feasible. It is understood that some tasks cannot be as effectively performed with the left hand, however. Using a screwdriver is an example. Given that supination is stronger than pronation and screws are provided only with right-handed threads, a left-handed person using a screwdriver with the left hand in pronation to drive the screw forward will fatigue more quickly, and the work will be performed less effectively and efficiently. Guideline 16.91: Ambidextrous Tool Use Tools should be designed so they can be used safely and effectively with either hand. For example, shaping for the fingers may limit use by left-handed operators and should be avoided.
Guideline 16.92: Operate Controls With Either Hand Controls should be placed such that they can be operated equally well with either hand (unless clearly inappropriate to do so) (Fraser, 1980).
16.2.6.6 Gloves and Other Personal Protective Equipment In many instances, medical personnel wear gloves that may decrease haptic feedback or increase force requirements for device control. Glove use interferes with grasping ability and may decrease grip strength by as much as 10% to 20% (Mital and Karwowski, 1991), with subsequent risk of slippage and overgripping and possibly increased task duration (Dababneh
TAF-ER562X_WINGR-09-0401-C016.indd 688
11/9/10 1:56:11 AM
Hand Tool Design
689
and Waters, 1999). Tool use may be compromised further if improperly sized gloves are worn. Gloves that are too small may result in numbness, while gloves that are too large may reduce grip force or result in slippage. Other personal protective equipment (e.g., masks, face shields, gowns, lead aprons) may interfere with vision, dexterity, comfort, and movement. Guideline 16.93: Consider Effects of Protective Gear The effects of wearing personal protective equipment on vision, dexterity, comfort, or other aspects of performance during tool use should be considered. For example, gloves that are too loose can decrease grip force and increase slippage, while gloves that are too small can produce numbness of the fingers.
16.2.6.7 Grip Force Forces acting on the handle include the force related to the interaction at the target (patient), the force to grasp the handle, and the force to activate any mechanisms involved in the tool’s use. These forces depend on the task requirements and the instrument’s mechanical constraints. The force that must be applied during tool use has both a physical and a temporal aspect. The physical aspect is related mainly to the magnitude and direction of force. The contact area determines the pressure applied to the hand as a result of the application of grip force. Most often this pressure should be applied at the fat pads of the hands or fingers and avoid the creases. It has been shown that greater pressures are associated with a greater risk of user discomfort and injury. The temporal aspect relates to how the peak or average force is exerted over time. Force may be exerted as an impulse or in a gradually increasing or sustained manner. The frequency with which a force is exerted is referred to as the repetition rate. As a rule, the more sustained the application of force or the more frequently it must be applied, the greater the risk of muscle fatigue or injury (Iridiastati and Nussbaum, 2006). Guideline 16.94: Grip Design Affected by Multiple Factors Grip design should consider the amount of force exerted by an individual, the user’s range of motion, and the number of repetitions an individual can endure when using a tool.
Guideline 16.95: Avoid Pressure on Inadequately Protected Surfaces Pressure over inadequately protected hand surfaces (e.g., where bones are thinly covered) should be avoided. Pressure that is transferred to the hand during work should be in the parts of the hand where its effects are least felt, namely, the palm at the thenar and hypothenar emminences.
Guideline 16.96: Avoid Deep Finger Recesses Avoid using deep finger recesses (>0.3 cm), especially when high grip forces are required, as finger anthropometry is so variable. Poorly designed finger recesses can cause undesirable compression of the fingers or palms. Similar effects can occur if a tool has a high curvature or short handle, especially when high or repetitive forces are exerted (Fraser, 1980).
Guideline 16.97: Curvature of Handle The force-bearing area of the handle should span the length of the palm and have a curvature of no greater than 1.3 cm over its entire length (see Chapter 4, “Anthropometry and Biomechanics,” for hand dimensions) (Fraser, 1980).
TAF-ER562X_WINGR-09-0401-C016.indd 689
11/9/10 1:56:11 AM
690
Handbook of Human Factors in Medical Device Design
16.2.6.8 Frequency (Repetition) and Duration of Effort The clinical application will define the force needed, the number of repetitions needed, and the precision required. Recommendations regarding the amount of force that should be exerted often do not consider the effects of duration or repetition, even though these can significantly affect performance (Figure 16.39). For example, it is usually recommended that the force exerted should not exceed 30% of the maximal voluntary contraction (MVC). However, even very short low force efforts (less than 30% MVC) performed frequently (e.g., greater than 15 per minute) can produce fatigue when sustained for 1 to 2 hours (Iridiastati and Nussbaum 2006). It is also difficult to maintain intermittent static efforts of 40% to 50% MVC over a 1- to 2-hour period, even when the exertion is of short duration (Salvendy, 1997). Static efforts of only 15% MVC cannot be sustained indefinitely, although exertions of less than 8% MVC require little if any recovery time (Salvendy, 1997). These examples assume ideal conditions. Ultimately, concerns about force, frequency, and duration need to be balanced with other clinical, technical, and ergonomic considerations. If a task requires significant effort, the number of people capable of doing it will be limited, and risk will be somewhat controlled by user selection. If, on the other hand, the task does not require much effort, then many people can do it without complaint. If the task also involves significant repetition or sustained effort (long duration), the occurrence of fatigue suggests that the job or task (including the tools used) should be redesigned. Task frequency may be classified as high (>50), medium, or low (15 or less). Medium to high repetition rates, especially with high force requirements, can result in musculoskeletal disorders (Dababneh and Waters, 1999). Guideline 16.98: Avoid Prolonged or Repetitive Muscle Use Avoid the need for prolonged or repetitive exertions of any muscle group. When such motions are necessary, consider the use of a powered tool. Alternatively, a ratchet or locking mechanism can be used to maintain prolonged force.
Guideline 16.99: Frequency-Force Tradeoff
% of maximum isometric strength
The frequency of tool use should be balanced against the forces required. Frequent use of a tool should require exertion of smaller forces, while infrequent use (e.g., one time) may allow forces that exceed typically recommended limits. Fatigue curve for isometric muscle work
120 100 80 60 40 20 0 0
100 200 300 Time to fatigue, seconds
400
FIGURE 16.39 Muscle contraction over time (endurance). As the required isometric muscle strength to use a tool increases, the time to fatigue decreases exponentially. (Salvendy, G. Handbook of Human Factors and Ergonomics (2nd ed.), Reprinted from John Wiley & Sons, Inc. With permission.)
TAF-ER562X_WINGR-09-0401-C016.indd 690
11/9/10 1:56:11 AM
Hand Tool Design
691
Guideline 16.100: Maximum Sustained Force Tool-facilitated tasks should require no more than 30% of MVC when force is sustained for more than 2 minutes (although no more than 10% MVC would be preferable) (Iridiastati and Nussbaum, 2006).
Guideline 16.101: Maximum Repetitive Force A maximum of 15% MVC is recommended for repetitive handgrip work (Greenberg and Chaffin, 1978). Note that 15% MVC is still a significant amount of force, and ideally the force required will be much less than this.
16.2.7 CONTROL TYPE AND PLACEMENT Some general guidelines for hand tool controls are listed below. For further general discussion of this topic, refer to Chapter 7, “Controls.” Guideline 16.102: Understandable Control Activation Users should be able to readily understand how to activate controls.
Guideline 16.103: Control Location and Feedback Controls should be logically placed on the handle and provide the necessary tactile feedback and precision required of the clinical task.
Guideline 16.104: Minimize Need for Hand Position Changes Minimize the need for the user to change hand position or grip type during tool use while activating the controls or with instrument rotation.
Guideline 16.105: Avoid Inadvertent Operation Protection from inadvertent control activation should be provided.
Guideline 16.106: Control Accessibility Adequate access to controls should be provided without compromising the user’s ability to maintain control of the device.
Guideline 16.107: Control Simplicity Avoid control complexity. Too many controls should not be located close together (Woodson et al., 1992).
Guideline 16.108: Intuitive Use A tool’s shape and design should indicate where users should place their hands and communicate how the device should be held as well as how to use the controls (Woodson et al., 1992).
Physicians and other health care workers become comfortable with the tools with which they have trained. Consequently, designers must carefully consider deviations from first-generation designs, as it becomes increasingly difficult to modify learned behaviors. For example, if physicians have traditionally used a particular type of control for a specific
TAF-ER562X_WINGR-09-0401-C016.indd 691
11/9/10 1:56:11 AM
692
Handbook of Human Factors in Medical Device Design
function, it will be more difficult for an alternative design to be successful. The longer a specific device configuration is in the marketplace, the more users will be reluctant to adopt a considerably altered design. Guideline 16.109: End Effector Control Consistency End effectors should be controlled consistently, in accordance with the expectations of the users.
Guideline 16.110: Consonance with User Expectations Control designs should be consistent with users’ expectations based on prior experience. When a control will violate users’ expectations and there is a risk of negative transfer of training, the benefits of the new control design should be compelling and should be explicitly tested.
16.2.7.1 Triggers Triggers are levers that are used most often for highly repetitive actions and actions that involve gross movement, that mimic end-effector actions, or that require high input forces. Figure 16.40 shows a device that has dual triggers, one that closes the jaws and one that releases them. Spacing of triggers is important for multiple trigger configurations. Each trigger should provide enough space to accommodate the index finger with the secondary trigger still within its reach. Some triggering devices permit a single operation and some a repeated operation (e.g., the button to activate an electrocautery pencil can be pressed repeatedly but can also provide continuous cautery for as long as it is pressed). Still other triggers may permit continuous operation with a single activation of the trigger, for example, a surgical power tool such as a saw, drill, or laser. Guideline 16.111: Use Triggers for Continuous Powered Operation Triggers are the preferred control when using power tools in continuous operation.
Guideline 16.112: Trigger Length The trigger length should be sufficient to permit at least two fingers to actuate the trigger (e.g., at least 5 cm long) (Dababneh and Waters, 1999) but not exceeding 8.9 cm in length (Cacha, 1999).
FIGURE 16.40 permission.)
Device with a dual trigger configuration. (Courtesy of AtriCure, Inc. With
TAF-ER562X_WINGR-09-0401-C016.indd 692
11/9/10 1:56:11 AM
Hand Tool Design
693
Guideline 16.113: Optimal Finger Position The trigger should be located so that the midfinger can actuate the trigger and use of the fingertips to activate the trigger can be avoided (Dababneh and Waters, 1999). Helander (1995) recommends using the distal phalange for precision work and the thumb for power.
Guideline 16.114: Multiple Triggers For instruments with multiple triggers, all triggers should be within reach of the fingers, and sufficient space should be provided for the fingers to reach every trigger (see Figure 16.40).
Guideline 16.115: Avoid Finger Grooves Avoid finger grooves on triggers, as hand sizes may vary (Dababneh and Waters, 1999).
Guideline 16.116: Avoid Pinch or Rub Points A small extension at the top of the trigger or equivalent design feature should be provided so the finger does not rub against the tool or get pinched during trigger activation or release (Dababneh and Waters, 1999).
Guideline 16.117: Maximum Trigger Force The force required to actuate a single finger trigger should not exceed 5 N (Dababneh and Waters, 1999).
Guideline 16.118: Minimize Trigger Force The mechanism used to transfer force from the handle to the tip should minimize the amount of force the user must exert (see the section “Special Considerations” for a discussion of this issue as it applies to laparoscopic tools).
Guideline 16.119: Trigger Travel Distance The trigger’s travel distance should be 0.6 to 1.8 cm (Dababneh and Waters, 1999).
Guideline 16.120: Trigger Position When Pressed When fully pressed, the trigger should not be flush with the handle nor protrude more than 0.6 cm (Dababneh and Waters, 1999). If the trigger is flush, then gloves could be pinched.
Guideline 16.121: Discrete or Continuous Activation Triggers should be operable with either discrete or continuous motions, as determined by specific clinical needs.
Guideline 16.122: Trigger Locks Avoid the need to continuously activate (squeeze) triggers on power tools. Trigger locks should be used as necessary.
16.2.7.2 Slide Controls Slide controls can vary in overall dimension. For example, slide controls could span the entire hand piece, allowing access for both right- and left-handed operators, or they could be implemented as simple sliders on the handles of precision grip tools.
TAF-ER562X_WINGR-09-0401-C016.indd 693
11/9/10 1:56:12 AM
694
Handbook of Human Factors in Medical Device Design
Guideline 16.123: Slide Control Height The minimum height for slide controls placed on top of handles is 0.3 cm (Fraser, 1980).
Guideline 16.124: Slide Displacement Displacement of a slide control should be 0.6 to 1.6 cm (Woodson et al., 1992).
Guideline 16.125: Slide Control Force The force to operate a slide control should not exceed 280 gm (Woodson, et al., 1992).
Guideline 16.126: Slide Control Texture and Shape The surface texture and shape (e.g., concavity) of the control should provide sufficient purchase for fingers to adequately move the control.
Guideline 16.127: Slide Control Location The location of the control should allow access and activation by either the thumb or index finger but without affecting how the instrument is held. Alternatively, the control may be placed where the other hand can operate it (Woodson et al., 1992).
16.2.7.3 Push Button Push buttons on hand tools are generally activated by the thumb or index finger. Depending on postural considerations, location, and the finger used to push the button, varying force outputs may be accommodated. For example, if a tool’s push button is placed on the end of an in-line grip and activated by the thumb, a greater force can be applied than would be provided on the side of a pistol grip to be activated by a finger. Regardless, lower force requirements for pushing the button are typically preferred. Guideline 16.128: Push Button Displacement Push-button controls should have a minimum displacement of 0.3 cm (Fraser, 1980).
Guideline 16.129: Push Button Height Push-button height should be at least 0.3 cm (Woodson et al., 1992).
Guideline 16.130: Push Button Forces The push-button control force range should be 3 to 10 N (Fraser, 1980). Forces below 3 N may be acceptable. However, if the force to actuate a push button is too low, the user will not receive adequate tactile feedback that the button was pushed.
16.2.7.4 Rotation Rotary controls are generally used to act on the end effector or as a crimp mechanism (wheel) that adjusts flow (e.g., of fluids). For rotary controls that act on the end effector, it is important to recognize the transition from the handle to the control. Rotary controls are often fluted. The fluted parts of the control need to be flush with the handle, while the projections extend away from the handle sufficiently for the user to be able to manipulate the control (Figure 16.41). In both instances, surface texture should be considered a priority to prevent slippage.
TAF-ER562X_WINGR-09-0401-C016.indd 694
11/9/10 1:56:12 AM
Hand Tool Design
FIGURE 16.41
695
A fluted rotary control that is flush with the handle.
Guideline 16.131: Rotary Control Maximum Force The torque required for rotary controls actuated by a single hand should not exceed 0.6 Nm.
Guideline 16.132: Rotary Control Surface Size The control surface of a rotary control should provide adequate purchase for manipulation, extending a minimum of 0.3 cm beyond the handle housing.
16.2.7.5 Sensory Feedback When using a tool, it is important that users receive tactile, visual, and/or auditory feedback about the status of the tool and the task. Examples of tactile feedback include the sensing of pressure, an impact, or changes in texture. Visual feedback may be provided by text printed on a tool (e.g., gradations on a syringe that allow for injection of precise amounts) or illumination of colored lights (e.g., green “on” light). An example of auditory feedback is the tone that is emitted by an electrocautery unit to indicate that it has been activated. Guideline 16.133: Sensory Feedback Essential The tool should provide sensory feedback to the user during use (e.g., pressure, impact shock, texture, temperature) (Mital and Karwowski, 1991).
Guideline 16.134: Provide Tactile Feedback Tactile feedback should be provided where possible.
Guideline 16.135: Visible Scale Contrast and Readability Visual scales should have good contrast and readability (see Chapter 8, “Visual Displays,” and Chapter 13, “Signs, Symbols, and Markings, ” for guidelines about fonts, font size, contrast, the use of pictorial symbols, and so on).
Guideline 16.136: Keep Handle Clear of Markings Avoid placing visual scales on the handle, where the markings can be obscured by the user’s hand or require the user to look away from the target.
TAF-ER562X_WINGR-09-0401-C016.indd 695
11/9/10 1:56:12 AM
696
Handbook of Human Factors in Medical Device Design
Guideline 16.137: Adequate Volume of Auditory Feedback Auditory feedback should be loud enough to be heard over the expected level of ambient noise.
Guideline 16.138: Obvious Indication of Current Mode When more than one operating mode is possible, the tool should provide an obvious indication of the current mode of operation. For example, on electrocautery machines, the different modes (Cut vs. Coag vs. Blend) are indicated by different auditory tones.
Guideline 16.139: Feedback About Discrete Control Positions Discrete control positions should be indicated by tactile and auditory feedback (e.g., a clicking detent) (Patkin, 2001).
16.3 SPECIAL CONSIDERATIONS 16.3.1 LAPAROSCOPIC SURGERY Laparoscopic surgery is an example of minimally invasive surgery that employs several small incisions through which ports (trocars) are placed, allowing a video-endoscope (commonly referred to as a camera) and specialized instruments to be inserted into the abdominal cavity. The design of instruments that are safe and effective has become increasingly important because (1) laparoscopic surgery imposes limits on the surgeon’s vision and ability to manipulate tissue (2) the number and complexity of laparoscopic surgical procedures (and instruments) is steadily increasing, and (3) there is an increasing interest in the ergonomics of surgical work (Berguer, 1997; van Veelen et al., 2003). Laparoscopic surgery presents two primary challenges not associated with open surgical procedures: • The operative field is viewed through a monocular (usually) lens system with the image displayed on a two-dimensional video monitor placed next to the patient. As a result, depth perception is lost, and there is a disassociation between the surgeon’s line of sight and the surgical field. The surgeon is looking at a video monitor at eye level while manipulating without looking directly at surgical instruments held at waist level. • Long, slender, rigid instruments are passed through ports that serve as fulcrums and are inserted through the abdominal wall. This results in mirroring of hand and arm movements (e.g., hand moves left, and instrument tip moves right) as well as negative and positive scaling of the surgeon’s hand movements (i.e., a large hand movement is needed for a small movement of the instrument tip, or, less commonly, a small hand movement results in a large movement of the instrument tip) (Figure 16.42). The trocars’ fixed positions and the effects of scaling may require the surgeon’s arm and hand to traverse large external arcs to accomplish small internal movements of the end effector, with resultant wrist flexion and ulnar deviation. Consequently, the surgeon exerts significant upper-extremity effort that results in muscle fatigue (Berguer et al., 2001). In addition, laparoscopic instruments generally lack the maneuverability, force advantages, precision, and force feedback of open surgery instruments (Table 16.12)
TAF-ER562X_WINGR-09-0401-C016.indd 696
11/9/10 1:56:12 AM
Hand Tool Design
697
Abdominal cavity
FIGURE 16.42 Schematic of the mirroring and scaling effects associated with laparoscopic instrument use. The large arrow pointing to the left represents the surgeon’s hand movement. The small arrow pointing to the right represents the smaller movement (scaling), in the opposite direction (mirroring), of the instrument inside the abdomen.
(Berguer et al., 2003). For example, studies have shown that three to five times more grip force is required for laparoscopic grasping compared to open grasping. This is likely due to inefficient force transfer by the laparoscopic instrument as well as decreased force generation from nonneutral hand and wrist positions (Berguer, 1998). For laparoscopic suturing, Emam et al. (2001) found that the grip force was increased 2.5 times relative to that of open suturing. TABLE 16.12 Problems Associated with Current Laparoscopic Instrument Designs Problem
Consequence
Only 4 degrees of freedom of movement (loss of the “wrist” articulation) versus 6 degrees of freedom with the human hand Historical use of “double-ring” handle designs
Difficulty using instruments to easily accomplish tasks. Awkward hand, wrist, arm, and shoulder positions may be necessary. Neither a power nor a precision grip is accommodated well. Sharp edges and narrow pressure surfaces on handles Pressure points cause sore hands when held for extended periods. Cables and connectors required Instruments are hard to use, weighty, and produce obstructions and entanglements. Friction between the instrument shaft and entry port Fine movements are more difficult. Tactile feedback is limited associated with forces (trocar) applied to tissues by the end effectors. Friction within the instrument mechanical linkage Greater force required by the hand to achieve a tip force equal to that of conventional surgical instruments.
TAF-ER562X_WINGR-09-0401-C016.indd 697
11/9/10 1:56:13 AM
698
Handbook of Human Factors in Medical Device Design
Laparoscopic instruments can be grouped into four functional categories: tissue manipulation, electrosurgery, suction and irrigation, and tissue suturing. These can be ranked according to how difficult they are to use and in terms of their effects on the user. Of these, the first group, which includes graspers, dissectors, and needle holders, by far causes the most discomfort (van Veelen et al., 2001). The combined effects of viewing on a remote monitor and unwieldy instrument design result in surgical tasks that are significantly more difficult to perform, longer operating times, and higher levels of physical and mental stress for the surgeon (Berguer, 1997). These ergonomic problems frequently result in awkward positioning and movements of surgeons’ hands, arms, and trunk during surgery, thereby causing musculoskeletal pain, finger numbness, and eventually physical disability. Thus, small design inadequacies in laparoscopic instruments can produce substantial problems for the user. Important aspects of laparoscopic instrument design include the handle, internal mechanics, the need for fixed insertion points, and contextual factors. 16.3.1.1 Handles Two main handle types are seen in standard laparoscopic instruments: either an angled (pistol-grip) handle (Figure 16.43) or an axial (in-line) grip handle (Figure 16.44). Either type of handle can have any of several configurations, such as finger rings, finger loops, cylindrical, shanks, and so on. The angled (pistol-grip) handle (usually with finger loops or rings) is the most common. This handle is based on historical designs but is problematic when used in modern laparoscopic surgery. The straight-line alignment of the handle and the shaft requires the surgeon to abduct the arms and place the wrists in a flexed, supinated, and ulnar-deviated position (Figure 16.45) (Berguer et al., 2001). Thus, the surgeon’s wrist adopts a “crabhand” position of grasping that is neither a power nor a precision grip. This undesirable posture decreases maximal hand forces and results in discomfort and muscle fatigue in the surgeon’s upper extremities. This grip requires users to alternate between forceful and delicate grasping; in fact, it is not common for the surgeon to use a combination power/ precision grip. The postures assumed with current laparoscopic instruments do not allow either forceful or delicate grasping to be performed with ease. Nonetheless, these handles may be used with either hand, are economical, and are easy to clean. The poor ergonomics of laparoscopy are magnified if the operating table height cannot be adjusted properly to accommodate the longer instrument lengths relative to those
FIGURE 16.43
Angled laparoscopic tool handle with a finger loop.
TAF-ER562X_WINGR-09-0401-C016.indd 698
11/9/10 1:56:13 AM
Hand Tool Design
FIGURE 16.44 handle.
699
In-line laparoscopic tool handles: one with finger rings, the other with a shank
of conventional surgery. An axial (in-line) grip allows for a more conventional power grip position but may result in even greater ulnar deviation, which increases the muscle force required for grasping (Berguer, 1998). In some circumstances, the axial handle can be gripped in a “pencil” grasp and used for fine manipulation. However, the mechanical inefficiency of most instruments makes this difficult. A number of alternative handle designs have been proposed and have undergone limited usability studies (Emam et al., 2001; Matern and Waller, 1999; van Veelen et al., 1999, 2001). However, there is no clear evidence to support the superiority of one handle type over another (Moont, 2003; Uchal et al, 2002), although surgeons have traditionally expressed a preference for axial-type handles for suturing. 16.3.1.2 Internal Mechanics It is important that laparoscopic instruments are designed with low internal friction and consequently low activation forces. This minimizes surgeons’ physical effort while maximizing the force feedback they receive from tissues. Some experimental instruments have novel linkage mechanisms and even haptic feedback, but these are not yet in common use. 16.3.1.3 Fixed Insertion Points The need to place instruments through entry ports means that rigid laparoscopic instruments have only 4 degrees of freedom (i.e., they cannot flex internally in a horizontal or vertical direction). This limitation severely restricts the surgeon’s ability to manipulate tissues
FIGURE 16.45 “Crab-hand” position (i.e., wrist in flexed, supinated, ulnar-deviated position) when grasping a laparoscopic instrument with a pistol-grip handle.
TAF-ER562X_WINGR-09-0401-C016.indd 699
11/9/10 1:56:13 AM
700
Handbook of Human Factors in Medical Device Design
in hard-to-reach places. Some curved or “curveable” instruments have been designed, including flexible ports, but again these have not gained widespread acceptance. Newer computer-actuated telemanipulators and some novel mechanical instruments can provide 6 degrees of freedom of internal movement. However, these are expensive and not yet widely available. 16.3.1.4 Contextual Factors Laparoscopic surgery is usually performed in a darkened room, with several video monitors displaying the internal surgical field. As a result, visual cues are diminished, and it can be hard to distinguish one instrument from another. Multiple tubes, cables, and connectors for the video system, gas insufflation, power delivery, and suction/irrigation are required for laparoscopic procedures. The operative field can become very crowded, with tubes and cables easily becoming tangled. Instrument exchanges are time consuming during laparoscopic surgery, and the need for them should be minimized (Matern and Waller, 1999). Beyond easy identification of the instruments, this process can be aided by the use of multifunction instruments. Such instruments can carry out several different tasks without the need for an instrument exchange. Examples include cautery hooks with irrigation and suction capacities and coagulating shears that also divide the coagulated tissues. However, with this multifunctionality comes increased cost and complexity, decreased usability, and increased risk of errors (e.g., activating the wrong control function). 16.3.1.5 Guidelines for Laparoscopic Instrument Design It has been asserted that the use of general guidelines is insufficient to ensure good laparoscopic instrument design (van Veelen et al., 2003). While the previous general guidelines are still applicable, the following additional guidelines are specifically for laparoscopic tool design. Guideline 16.140: Ambidextrous Suturing Handles for laparoscopic needle holders should permit both right- and left-handed suturing (Nesbakken, 2004).
Guideline 16.141: Needle Holder Width Handles for laparoscopic needle holders should be at least 1 cm wide to prevent areas of pressure on the hand (van Veelen et al., 2003).
Guideline 16.142: Laparoscopic Handle Length The length of in-line handles should not exceed 17 cm (van Veelen et al., 2003).
Guideline 16.143: Laparoscopic Handle Grip Opening The instrument’s grip opening should not exceed 4.1 cm (Nesbakken, 2004).
Guideline 16.144: Laparoscopic Handle Closing Force The closing force for laparoscopic handles should not exceed 15 N (van Veelen et al., 2003).
TAF-ER562X_WINGR-09-0401-C016.indd 700
11/9/10 1:56:14 AM
Hand Tool Design
701
Guideline 16.145: Trigger and Finger Loop Lengths Trigger length or finger loop length should be at least 9.3 cm, adequate for four fingers (van Veelen et al., 2003).
Guideline 16.146: Low Friction Knobs Knobs should be able to rotate without appreciable friction (van Veelen et al., 2001).
Guideline 16.47: Minimize Repetitive Motion Risks Laparoscopic tools that require frequent opening and closing should minimize the effect of repetitive motion activity (Quick et al., 2003).
Guideline 16.148: Internally Flexible or Articulating Internally flexible or articulating instruments that can provide greater maneuverability should be considered (Dankelman et al., 2005).
Guideline 16.149: Tube and Cable Management Attached tubes and cables should be easily routed or fixed in position.
Guideline 16.150: Connector Positioning Connectors should be easy to use and be readily positioned out of the way of the surgeon while operating.
Guideline 16.151: Provision of Storage or Sheathing Sheathing or storage should be provided for the instrument(s) when not in use.
Guideline 16.152: Easy to Use Multifunctionality Multifunctionality should be incorporated into instrument design to minimize instrument exchanges, but the resulting instrument and its controls should be easy to use, and its use should not introduce additional high-risk use errors.
16.3.2 MINIMALLY INVASIVE CATHETER-BASED PROCEDURES The first angioplasty was performed in 1964 by Dr. Charles Dotter at the Oregon Health and Science University. Since that time, intravascular catheter-based interventions have become common, with hundreds of thousands performed annually. There are a wide variety of vascular and other image-guided procedures, ranging from very common procedures, such as the insertion of vascular access and pressure measurement catheters for use in intensive care and surgery, to the use of less common but more sophisticated intravascular catheters used for various therapeutic interventions, such as embolization of intracranial aneurysms, insertion of stents and other indwelling items, ablation of undesired tissue, chemotherapy for treatment of cancer, and so on.* * Although the discussion focuses on intravascular catheter-based procedures, noncatheter items, such as spinal cord stimulator or pacemaker electrodes, are also placed using similar imaging techniques. Although not covered in this chapter, the principles discussed here apply to these other hand-guided devices.
TAF-ER562X_WINGR-09-0401-C016.indd 701
11/9/10 1:56:14 AM
702
Handbook of Human Factors in Medical Device Design
Since they are manipulated by a physician’s hand during insertion, positioning, and use, such catheters readily meet the criteria for a “hand tool.” The handle may be considered to be the catheter hub and connectors, by which the catheter position is manipulated and substances are injected or removed. The end effector is the tip of the catheter, where the orifices, sensors, and other devices are placed. Four main functions of a catheter can be distinguished: • Therapeutic manipulation, such as ablation of tissue and angioplasty • Sensing, such as temperature, pressure, and oxygen content (saturation) • Delivery of materials or energy, such as fluids, drugs, or contrast; electrical, thermal, laser, or other forms of energy; or embolization materials or stents • Removal of material, such as aspiration of blood or air or retrieval of inserted devices In their simplest form, catheters are long, thin, flexible tubes inserted for delivery of fluids and other substances or to allow removal of blood samples or other substances (Figure 16.46). More sophisticated catheters also have tips that deliver various forms of energy or sensors for measuring various physiologic parameters. Catheters are also used to insert indwelling devices. One thing that nearly all catheter-based procedures have in common is the use of various imaging methods, such as ultrasound, X-rays, computed tomography, or MRI scanning to guide the insertion and positioning of these devices. As such, catheter use shares a characteristic of laparoscopic surgery: viewing the procedure on a monitor without actually observing the manipulations being done by the hand. In all cases, there must be a means of directing the catheter to its desired position. In most cases, the catheter is manipulated (steered) by hand (Figure 16.47). The ease with which a catheter is directed is markedly affected by the integration of the catheter’s shape and the choice of materials for catheter wall design. In some cases, the catheter can be directed by using a modification of the tip, such as the addition of a balloon (e.g., a pulmonary artery catheter). A guide wire may also be used to add stiffness and functional directionality to the catheter. Broadly speaking, the problems of advancing the catheter can be divided into two primary categories: inadequate feedback about catheter orientation and position, and inadequate controllability as the catheter is advanced. The current method for advancement
FIGURE 16.46
Catheter used in minimally invasive, catheter-based procedures.
TAF-ER562X_WINGR-09-0401-C016.indd 702
11/9/10 1:56:14 AM
Hand Tool Design
FIGURE 16.47
703
Catheter being manipulated by hand.
of a catheter under imaging guidance is to push and turn the catheter (or guide wire) at its proximal or hub end, outside the patient, in order to control tip position. The most important information on the tip’s position is obtained from the two-dimensional image of the catheter or guide wire on a remote screen. In addition to this image, the person manipulating the catheter receives haptic feedback regarding tip position and direction. Curves can be placed on the catheter or guide wire tip, which can then be turned left or right by applying torque at the insertion site. However, as catheter flexibility and distance from the insertion site increase, the user receives less haptic feedback regarding the tip’s movement in response to control actions, and it can become quite difficult to maneuver the catheter tip into the desired location. It is not uncommon for multiple catheters to be placed and used during a single procedure, as each catheter is generally used for a single purpose (e.g., injection of contrast, insertion of stents, ablation of tissue). Thus, the ease with which such catheters can be positioned strongly affects the length of the procedure, risk of complications, and radiation exposure. The incidence of work-related discomfort or injury to the user associated with catheter use is unknown. Patient injuries (e.g., damage to blood vessels) may occur, in some cases related to excessive catheter stiffness, choice of an inappropriate catheter size, or overly aggressive manipulation. Increased stiffness improves the ability to “steer” the catheter but concurrently increases the risk of vascular injury. Alternatively, if the catheter is made softer and more flexible and thereby less traumatic, it becomes more difficult to maneuver. With current materials and designs, maneuverability and safety must be traded off against each other. Ultimately, however, patient injuries may be related to inadequate tactile feedback and uncertainties regarding tip position with two-dimensional imaging. Other concerns related to catheter use may also be design-related. Flow-guided catheters use an inflated balloon to guide a pulmonary artery catheter, for example, from the internal jugular vein through the heart (right ventricle) and into a pulmonary artery. The balloon is filled with air, and if it ruptures, whether by defect or overfilling, it is potentially lethal. A variety of design solutions have been proposed (Westenskow and Silva, 1993) and implemented, but balloon ruptures still occur. Injection of drugs or fluids through a catheter by a powered device may produce vascular injuries. Catheter-related bloodstream infections are still quite common, especially when catheters are left in for longer periods, and may be affected by the choice of catheter materials and catheter design. Blood clots have a tendency to form at catheter tips, with the incidence similarly related to the duration of catheter use as well as catheter design. Other risks to the patient include disconnections; misconnections; broken, fractured, or sheared catheters; toxic materials or infected matter; immune responses; and accumulation of tissue debris.
TAF-ER562X_WINGR-09-0401-C016.indd 703
11/9/10 1:56:14 AM
704
Handbook of Human Factors in Medical Device Design
Because exposure to X-rays is dangerous for those involved, it is important to minimize radiation exposure. Exposure time is related to the time it takes to position the catheter and perform the procedure. Any design or technique that promotes catheter manipulation or steering may reduce radiation exposure. One possible solution that may further reduce or even eliminate the need for radiation exposure is the development of magnetic-based instrument guidance systems, a technology that is still in its infancy. Also, ultrasound is increasingly being used to guide needle- and catheter-based percutaneous procedures. Many new sensor technologies for diagnostic and measurement purposes are being developed. However, because of a catheter’s small diameter and potentially long length, there are design constraints, for example, related to catheter end-effector size, size and weight limitations at the insertion site, power consumption (and possible heating of tissues), and material biocompatibility. In summary, this section identifies some specific issues with catheter design and use, but because of the relative newness of these devices and procedures, HF design guidelines specifically for endovascular catheter design have not yet been formulated. There are few studies or data on which to base design recommendations. Current designs have usability issues that constrain their functionality and decrease their safety. Further research and development is clearly indicated.
16.4 WHAT TO DO WHEN GUIDELINES ARE NOT AVAILABLE There are many examples where data to guide medical hand tool design do not exist, and thus recommendations are made on the basis of either nonmedical hand tool references or the authors’ collective experience. Where no guidance is available, there are several ways to obtain the necessary information. For example, both the FDA and the Emergency Care Research Institute (ECRI) maintain medical device reporting systems that may have useful information based on prior events (although these data will be mostly about what not to do). Alternatively, one must engage a knowledgeable HF consultant and/or collect the required data one needs to guide design. General approaches to the process of incorporating HF into medical device design and general aspects of HF in health care are described in other chapters and elsewhere (ANSI/AAMI HE-74-2001; Carayon, 2007). With regard to device evaluation, there are considerations that are more specific to medical hand tool design. At a minimum, user performance, timing, and effectiveness should be assessed via direct observations and questionnaires. Checklists for handle design and for laparoscopic instruments have been proposed (Habes and Baron, 2000; Patkin, 2001). Static and dynamic force analysis can be helpful in mapping human input forces (e.g., based on available human capability charts) to the required output (end-effector operation). Biomechanical modeling may be used. Various means of instrumenting tool handles to measure force, torque, and other physical parameters have been described (Bucx et al., 1992a; McCoy et al., 1995; McGorry, 2001). EMG data, force sensors, and goniometers (joint angle measuring devices) can measure the actual muscle contractions, instrument forces, and joint angles required when using a hand tool. Additionally, the use of grip sensors applied directly to the hand can allow designers to measure and objectively evaluate the static and dynamic pressures and forces involved in gripping and grasping objects. Such systems provide detailed pressure profiles, forces, and graphical displays for quantitative analysis of various grip applications and should improve
TAF-ER562X_WINGR-09-0401-C016.indd 704
11/9/10 1:56:15 AM
Hand Tool Design
705
decision making with regard to overall ergonomic fit. Such data can also serve as a benchmark against which multiple grip types can be compared. However, there is no substitute for direct observation. Observing surgeons or others using medical hand tools and obtaining detailed information about their experience with the tool’s design can be invaluable. Videotaping allows for detailed analysis of use issues, and when actual use on patients cannot be observed, simulation can be used. Current patient care simulators have achieved a sufficient degree of realism that actual instrument use and medical tasks can be effectively evaluated and design issues identified.
16.5 CASE STUDIES The following two case studies describe applications of HF to specific medical hand tool design so as to illustrate the above HF design principles. These examples involve real manufacturers’ products; however, neither an endorsement nor a condemnation of any device is intended by their inclusion.
16.5.1 CASE EXAMPLE 1: ACCIDENTAL NEEDLE PUNCTURES WITH CATHETER USE Accidental needle puncture is an acute injury that carries with it the risk of infection. It is the most common occupational injury in health care workers that results in a visit to the emergency room. A significant fraction of needle punctures occur during intravenous (IV) catheter insertion. Many solutions to reduce the incidence of needlestick injuries have been tried, including educational efforts and process interventions (e.g., “don’t recap the needle”). Recently, a variety of devices have appeared on the market to address this safety issue, including hand tools designed to facilitate the safe insertion of an intravenous catheter. Such efforts have been successful, with wide acceptance of devices and a large decrease in the occurrence of needlestick injuries (Rivers et al., 2003). The basic (nonsafety) IV catheter insertion assembly (Figure 16.48) facilitates the task of inserting an IV catheter (a length of tubing inserted through the skin and into a blood vessel, usually a peripheral vein). The end effector is the needle, used first to enter the vein and then as a guide (or stent) for catheter insertion. The handle consists of the hub, made of plastic, and is held by the person inserting the IV catheter. In addition to providing a way to hold and control the needle’s movements, the hub has a small chamber into which blood flows on needle entry into the vein. This visible “flash” of blood provides visual feedback confirming needle entry into the vein. The catheter, which fits over the needle, has a connector that permits connection to IV tubing.
FIGURE 16.48 Basic nonsafety intravenous catheter assembly. After insertion of a nonsafety intravenous catheter, the sharp needle tip, now contaminated with the patient’s blood, is still exposed.
TAF-ER562X_WINGR-09-0401-C016.indd 705
11/9/10 1:56:15 AM
706
Handbook of Human Factors in Medical Device Design
To use this device, the needle is passed through the skin into a vein. Since the tip of the catheter must be close to but not overlap the needle bevel, when the “flash” of blood is seen, the needle must be advanced slightly further to ensure that the catheter tip, not just the needle tip, lies inside the vein. The catheter is then advanced over the needle and into the vein, the needle withdrawn, and the catheter connected to IV tubing. The catheter and tubing are held in place by taping or in, a few cases, suturing the connector and tubing to the skin. The catheter must slide easily off the needle, be closely adherent to the needle so as to pass easily without catching and bending or causing tissue damage, and be long enough to not accidentally come out of the vein. Once catheter insertion is complete, the needle is removed from the catheter and placed nearby prior to its disposal in a secure biohazard container (sharps box). It is during this postinsertion period that the risk of accidental puncture is greatest. A variety of “safety” catheter insertion devices are now available, each of which uses a different method to provide protection against needlestick injuries. These newer devices include all the previous elements of the traditional assembly but also incorporate a mechanism to effectively remove the needle from the field, for example, by retracting the needle into the hub. In the example shown in Figure 16.49, the retraction mechanism is actuated by pressing the white button at the spot (top middle) where the hub meets the needle. Several HF considerations increased the usability of this device. The button is well sized, its location makes it readily accessible to a fingertip, little force is required to actuate it, and it operates smoothly and reliably. When retracted, the needle tip is completely encased, making it virtually impossible for a stick to occur. Theoretically, a design for safety strategy reduces the dependence on behavior and auxiliary supplies (e.g., sharps boxes) and should provide a marked reduction in the frequency of needlestick injuries. Of course, the new device must still perform its basic job as well as the older designs—a safer design that is more difficult to use or less efficacious will result in work-arounds or avoidance of use. Reasons for decrements in performance include changes in the materials (e.g., different types of plastics), the size or shape of the device, poorly fitting catheter/needle assemblies, premature or unintended needle retraction, or poor visual or tactile feedback (e.g., inability to see a blood flashback). Engineered safety devices have proven to be the most effective means of reducing needlestick injury (Tan et al., 2001). The use of safety IV catheter systems has decreased the incidence of needlestick injury by as much as 85%, although the rate varies with the device design and its acceptance by health care providers (Tan et al., 2001). The introduction of these devices have resulted in other changes (unintended consequences) associated with its use. If the needle is retracted accidentally or from lack of experience by pushing the button prematurely, the catheter will be impossible to advance into the vein. More care
FIGURE 16.49 Intravenous catheter designed to reduce the likelihood of an accidental needle puncture. On insertion of the catheter, the sharp needle tip either retracts or is blunted via a mechanical mechanism.
TAF-ER562X_WINGR-09-0401-C016.indd 706
11/9/10 1:56:15 AM
Hand Tool Design
707
TABLE 16.13 Human Factors Issues to Be Considered for Intravenous Catheter Designs Issues
Considerations
Ease of use
Length of device Shape Bulkiness Hub visibility Blood “flash” during needle insertion Continued blood flow during catheter insertion Feel of the “pop” and how the catheter slides off the needle Activation of the safety mechanism (avoidance of premature activation) Sharpness of needle tip Ease of sliding catheter off of needle and into vein Ease of connection Security of connection Smoothness of passage through tissue Reaction by tissues to foreign materials
Feedback during use
Ease of insertion
Connecting to the catheter Choice of material
must therefore be taken to avoid premature retraction of the needle, a problem not seen with the older style. Another difference noted by one of the authors is the ease with which the blood flash is (or is not) observed, possibly as a function of the gloves used during catheter insertion. Use of purple nonlatex gloves appears to interfere with visualization of the blood flash, a problem not seen with the nonsafety devices. The reasons for this are unclear but may have to do with choice of materials and design of the catheter hub. In some cases, these new devices may be more difficult to use, thereby resulting in an increased number of “blown” veins and multiple needlesticks for patients. Issues that need to be addressed by the designer of intravenous catheters are listed in Table 16.13.
16.5.2 CASE EXAMPLE 2: THE HARMONIC SCALPEL The harmonic scalpel exemplifies a special case of laparoscopic instrument design. While the device was an engineering achievement and a major clinical advance its introduction and use were impeded by poor handle design. Two design issues contributed to the device’s poor usability: visualization of the tip (end effector) and the handle’s internal mechanics. Tip visualization is rarely an issue anymore, but handle mechanics continue to be problematic. In response to the risk of using electrocautery inside the abdominal cavity during laparoscopy (i.e., explosion risk), a technology to coagulate, seal, and divide tissues effectively using ultrasonic vibration was developed, avoiding the use of electrical current altogether. Initial versions of harmonic scalpel used a pistol-type handle attached to a large shaft that extended from the patient to above the surgeon’s waist. The handle was made of hard
TAF-ER562X_WINGR-09-0401-C016.indd 707
11/9/10 1:56:15 AM
708
FIGURE 16.50
Handbook of Human Factors in Medical Device Design
Harmonic scalpel with finger ring, finger loop, and curved tip.
plastic, the handle edges were rather sharp, and it required a power grip to squeeze and hold. Typically, a finger ring was provided for the thumb and a finger loop for the other fingers (Figure 16.50). The instrument shaft was straight and 10 mm in diameter, along with a straight tip (a single-action jaw, like a grasper). While dissection with cautery instruments such as fine graspers and scissors involved intermittent opening and closing of the instruments, dissection with the harmonic scalpel required grasping tissue with some force and holding it for longer periods (2 to 5 seconds) until it sealed and divided the tissues. Thus, the actual use of the harmonic scalpel was very different from standard electrocautery, even if the surgical result was similar. The original design of the harmonic scalpel required substantial sustained force on the handle to divide tissues, resulting in pain in the surgeon’s hands due to the hard and sharp edges of the handle and the lack of any ratchet mechanism to relieve the sustained grip and resultant tissue pressure. Moreover, because the endoscope was looking along the same axis as the harmonic scalpel’s shaft, the 10-mm shaft in conjunction with the straight tip made visualization of the tip during use very difficult. These designs contributed to patient injuries and chronic musculoskeletal discomfort and injuries of surgeons. More recent versions have incorporated curved tips and 5-mm shafts. The narrower shafts and curved tips have resolved the visualization problem and improved tip ease of use. Smoother, broader handles and slightly decreased grasp force requirements have improved their usability and, to some extent, reduced the risk of patient or surgeon injury. However, the force required to squeeze and hold the instrument remains excessive, and overall usability can still be improved. This example illustrates how changing the underlying technology used to accomplish a medical task (from electrocautery to ultrasonic) can substantially alter the performance of the actual task and require a redesign not only of the end-effector but also of the tool’s user interface (i.e., handle and controls).
REFERENCES American Conference of Government Industrial Hygienists. (2005). Threshold Limit Values for Chemical and Physical Agents: WholeBody Vibration-TLV Hand-Arm Vibration-TLV. Cincinnati: American Conference of Government Industrial Hygienists. American National Standards Institute. (1986). Guide for the Measurement and Evaluation of Human Exposure to Vibration Transmitted to the Hand. ANSI S3.34-1986. New York: American National Standards Institute.
TAF-ER562X_WINGR-09-0401-C016.indd 708
11/9/10 1:56:15 AM
Hand Tool Design
709
American National Standards Institute. (2000). Sterilization of Health Care Products—General Requirements for Characterization of a Sterilizing Agent and the Development, Validation, and Routine Control of a Sterilization Process for Medical Devices. ANSI/AAMI/ISO 14397. Arlington, VA: Association for the Advancement of Medical Instrumentation. American National Standards Institute/Association for the Advancement of Medical Instrumentation (ANSI/AAMI). (2001). Human Factors Design Process for Medical Devices. ANSI/AAMI HE-74-2001. Arlington, VA: Association for the Advancement of Medical Instrumentation. Anderson, R. M., and Romfh, R. F. (1980). Technique in the Use of Surgical Tools. New York: Appleton-Century-Crofts. Armstrong, T. J., Fine, L. J., Radwin, R. G., and Silverstein, B. S. (1987). Ergonomics and the effects of vibration in hand-intensive work. Scandinavian Journal of Work, Environment and Health, 13, 286–289. Babar-Craig, H., Banfield, G., and Knight, J. (2003). Prevalence of back and neck pain amongst ENT consultants: National survey. Journal of Laryngology and Otology, 117, 979–982. Baber, C. (2003). Cognition and Tool Use: Forms of Engagement in Human and Animal Use of Tools. New York: Taylor & Francis. Berguer, R. (1997). The application of ergonomics to general surgeons’ working environment. Reviews on Environmental Health, 12, 99–106. Berguer, R. (1998). Surgical technology and the ergonomics of laparoscopic instruments. Surgical Endoscopy, 12, 458–462. Berguer, R., Chen, J., and Smith, W. D. (2003). A comparison of the physical effort required for laparoscopic and open surgical techniques. Archives of Surgery, 138, 967–970. Berguer, R., Forkey, D. L., and Smith, W. D. (1999). Ergonomic problems associated with laparoscopic surgery. Surgical Endoscopy, 13, 466–468. Berguer, R., Forkey, D. L., and Smith, W. D. (2001). The effect of laparoscopic instrument working angle on surgeons’ upper extremity workload. Surgical Endoscopy, 15, 1027–1029. Berguer, R. and Hreljac, A. (2004). The relationship between hand size and difficulty using surgical instruments: A survey of 726 laparoscopic surgeons. Surgical Endoscopy, 18, 508–512. Bernard, B. P. (1997). Musculoskeletal Disorders (MSDs) and Workplace Factors: A Critical Review of Epidemiologic Evidence for Work-Related Musculoskeletal Disorders of the Neck, Upper Extremity and Low Back. Cincinnati: National Institute for Occupational Safety and Health, DHHS (NIOSH) Publication No. 97-141. Berthelet, A. and Chavaillon, J. (1993). The Use of Tools by Human and Non-Human Primates. New York: Oxford University Press. Brearly, S. and Watson, H. (1983). Towards an efficient retractor handle: An ergonomic study. Annals of the Royal College of Surgeons of England, 65, 382–384. Bucx, M. J. L., Scheck, P. A. E., Van Geel, R. T. M., Den Ouden, A. H., and Niesing, R. (1992a). Measurement of forces during laryngoscopy. Anaesthesia, 47, 348–351. Bucx, M. J. L., Snijders, C. J., van der Vegt, M. H., Holstein, J. D., and Stijnen, T. (1997). Reshaping the Macintosh blade using biomechanical modelling. Anaesthesia, 52, 662–667. Bucx, M. J. L., van Geel, R. T. M., Scheck, P. A. E., Stijnen, T., and Erdmann, W. (1992b). Forces applied during laryngoscopy and their relationship with patient characteristics. Anaesthesia, 47, 601–603. Bucx, M. J., van Geel, R. T., Wegener, J. T., Robers, C., and Stijnen, T. (1995). Does experience influence the forces exerted on maxillary incisors during laryngoscopy? A manikin study using the Macintosh laryngoscope. Canadian Journal of Anesthesia, 42, 144–149. Buffington, C. W., MacMurdo, S. D., and Ryan, C. M. (2006). Body position affects manual dexterity. Anesthesia and Analgesia, 102, 1879–1883. Burda, D. (1995). Hospitals rank high as health hazard for employees. Modern Healthcare, 25, 24. Burstrom, L. and Lundstrom, R. (1988). Mechanical energy absorption in human hand-arm exposed to sinusoidal vibration. International Archives of Occupational and Environmental Health, 61, 213–216.
TAF-ER562X_WINGR-09-0401-C016.indd 709
11/9/10 1:56:15 AM
710
Handbook of Human Factors in Medical Device Design
Buschbacher, R. (1994). Overuse syndromes among endoscopists. Endoscopy, 26, 539–544. Cacha, C. A. (1999). Ergonomics and Safety in Hand Tool Design. Boca Raton, FL: CRC Press. Cancio, L. I., and Cashman, T. M. (1998). Self-reported cumulative trauma symptoms among hospital employees: Analysis of an upper-extremity symptom survey. American Journal of Occupational Therapy, 53, 227–230. Carayon, P. (Ed.). (2007). Handbook of Human Factors and Ergonomics in Health Care and Patient Safety. Mahwah, NJ: Lawrence Erlbaum Associates. Carswell, C. M. (2004). Comment from the editor. Ergonomics in Design, 12, 3. Cherniack, M. G. and Mohr, S. (1994). Raynaud’s phenomenon associated with the use of pneumatically powered surgical instruments. Journal of Hand Surgery, 19A, 1008–1015. Dababneh, A., Lowe, B., Krieg, E., Kong, Y.-K., and Waters, T. (2004). A checklist for the ergonomic evaluation of nonpowered hand tools. Journal of Occupational and Environmental Hygiene, 1, D135–D145. Dababneh, A. and Waters, T. (1999). The ergonomic use of hand tools: Guidelines for the practitioner. Applied Occupational and Environmental Hygiene, 14, 208–215. Dankelman, J., Grimbergen, C. A., and Stassen, H. G. (Eds.). (2005). Engineering for Patient Safety: Issues in Minimally Invasive Procedures. Mahwah, NJ: Lawrence Erlbaum Associates. Diaz, J. H. (2001). Carpal tunnel syndrome in female nurse anesthetists versus operating room nurses: Prevalence, laterality, and impact of handedness. Anesthesia and Analgesia, 93, 975–980. Emam, T. A., Frank, T. G., Hanna, G. B., and Cuschieri, A. (2001). Influence of handle design on the surgeon’s upper limb movements, muscle recruitment, and fatigue during endoscopic suturing. Surgical Endoscopy, 15, 667–672. Evanoff, B. A., Bohr, P. C., and Wolf, L. D. (1999). Effects of a participatory ergonomics team among hospital orderlies. American Journal of Industrial Medicine, 35, 358–365. Evans, A., Vaughan, R. S., Hall, J. E., Mecklenburgh, J., and Wilkes, A. R. (2003). A comparison of the forces exerted during laryngoscopy using disposable and non-disposable laryngoscope blades. Anaesthesia, 58, 869–873. Färkkilä, M., Pyykkö, I., Korhonen, O., and Starck, J. (1979). Hand grip forces during chain saw operation and vibration white finger in lumberjacks. British Journal of Industrial Medicine, 36, 336–341. Food and Drug Administration. (1983). Application of the Device Good Manufacturing Practice Regulation to the Manufacture of Sterile Devices. Washington, DC: Food and Drug Administration. Available: http://www.fda.gov/cdrh/ode/267.pdf. Fraser, T. M. (1980). Ergonomic Principles in the Design of Hand Tools. Geneva: International Labour Office. Frazier, L. M., Thomann, W. R., and Jackson, G. W. (1995). Occupational hazards in the hospital, doctor’s office, and other health care facilities. North Carolina Medical Journal, 56, 189–195. Gawande, A. A., Studdert, D. M., Orav, E. J., Brennan, T. A., and Zinner, M. J. (2003a). Risk factors for retained instruments and sponges after surgery. New England Journal of Medicine, 348, 229–235. Gawande, A. A., Zinner, M. J., Studdert, D. M., and Brennan T. A. (2003b). Analysis of errors reported by surgeons at three teaching hospitals. Surgery, 133, 614–621. Greenberg, I. M. (1981). Self-retaining retractor and handrest system for neurosurgery. Neurosurgery, 8, 205–208. Greenberg, L. and Chaffin, D. B. (1978). Workers and Their Tools: A Guide to the Ergonomic Design of Hand Tools and Small Presses. Midland, MI: Pendell Publishing. Habes, D. J. and Baron, S. (2000). Ergonomic evaluation of antenatal ultrasound testing procedures. Applied Occupational and Environmental Hygiene, 15, 521–528. Halford, V., and Birch, I. (2005). The perceived causes of hand pain in podiatrists. Australasian Journal of Podiatric Medicine, 39, 57–63.
TAF-ER562X_WINGR-09-0401-C016.indd 710
11/9/10 1:56:15 AM
Hand Tool Design
711
Hastings, R. H., Hon, E. D., Nghiem, C., and Wahrenbrock, E. A. (1996a). Force, torque, and stress relaxation with direct laryngoscopy. Anesthesia and Analgesia, 82, 456–461. Hastings, R. H., Hon, E. D., Nghiem, C., and Wahrenbrock, E. A. (1996b). Force and torque vary between laryngoscopists and laryngoscope blades. Anesthesia and Analgesia, 82, 462–468. Hefflin, B. J., Gross, T. P., and Schroeder, T. J. (2004). Estimates of medical device-associated adverse events from emergency departments. American Journal of Preventive Medicine, 27, 246–253. Helander, M. (1995). A guide to the Ergonomics of Manufacturing. Bristol, PA: Taylor & Francis. Herington, T. N. and Morse, L. H. (1995). Occupational Injuries: Evaluation, Management, and Prevention. St. Louis, MO: Mosby. Hirschowitz, B. I. (1994). The cost of doing business: Occupational hazards for endoscopists. Endoscopy, 26, 559–561. Horkey, J. and King, P. (2004). Ergonomic recommendations and their role in cardiac sonography. Work, 22, 207–218. Institute of Medicine. (2000). To Err Is Human: Building a Safer Health System. Washington, DC: National Academies Press. International Organization for Standardization. (1980). Tools: Handbook on International Standards for Tools. Geneva: International Organization for Standardization. International Organization for Standardization. (2001). Mechanical Vibration-Measurement and Evaluation of Human Exposure to Hand Transmitted Vibration–Part 1: General Guidelines (ISO 5349). Geneva: International Organization for Standardization. Iridiastati, H. and Nussbaum, M. A. (2006). Muscular fatigue and endurance during intermittent static efforts: Effects of contraction level, duty cycle and cycle time. Human Factors, 48, 710–720. Jagger, J. and Perry, J. (2000). Does clinical setting affect injury risk? Nursing2000, 35. Kaplan, M. C. (1981). Task and tool design: It’s a “man’s world.” Occupational Health and Safety, 50, 29–34. Kerk, C. J. (1998). Ergonomics. Occupational Medicine, 13, 583–598. Keselman, A., Patel, V. L., Johnson, T. R., and Zhang, J. (2003). Institutional decision-making to select patient care devices: Identifying venues to promote patient safety. Journal of Biomedical Informatics, 36, 31–44. Kroemer, K. H. E. (1999). Assessment of human muscle strength for engineering purposes: A review of the basics. Ergonomics, 42, 74–93. Lalumandier, J. A. and McPhee, S. D. (2001). Prevalence and risk factors of hand problems and carpal tunnel syndrome among dental hygienists. Journal of Dental Hygiene, 75, 130–134. Levy, B. S. and Wegman, D. H. (1995). Occupational Health: Recognizing and Preventing WorkRelated Disease (3rd ed.). Boston: Little, Brown. Macario, A., Morris, D., and Morris, S. (2006). Initial clinical evaluation of a handheld device for detecting retained surgical gauze sponges using radiofrequency identification technology. Archives of Surgery, 141, 659–662. Magnavita, N., Bevilacqua, L., Mirk, P., Fileni, A., and Castellino, N. (1999). Work-related musculoskeletal complaints in sonologists. Journal of Occupational and Environmental Medicine, 41, 981–988. Majeed, A. W., Jacob, G., Reed, M. W. R., and Johnson, A. G. (1993). Laparoscopist’s thumb: An occupational hazard. Archives of Surgery, 128, 357. Matern, U. and Waller, P. (1999). Instruments for minimally invasive surgery: Principles of ergonomic handles. Surgical Endoscopy, 13, 174–182. McCoy, E. P., Austin, B. A., Mirakhur, R. K., and Wong, K. C. (1995). A new device for measuring and recording the forces applied during laryngoscopy. Anaesthesia, 50, 139–143. McCunney, R. J. (1994). A Practical Approach to Occupational and Environmental Medicine (2nd ed.). Boston: Little, Brown.
TAF-ER562X_WINGR-09-0401-C016.indd 711
11/9/10 1:56:15 AM
712
Handbook of Human Factors in Medical Device Design
McGorry, R. W. (2001). A system for the measurement of grip forces and applied moments during hand tool use. Applied Ergonomics, 32, 271–279. Mercer, R. B., Marcella, C. P., Carney, D. K., and McDonald, R. W. (1997). Occupational health hazards to the ultrasonographer and their possible prevention. Journal of the American Society of Echocardiographers, 10, 363–366. Mital, A. and Karwowski, W. (1991). Workspace, Equipment and Tool Design. Amsterdam: Elsevier. Moont, M. (2003). In-line vs. pistol-grip handles in a laparoscopic simulator. Surgical Endoscopy, 17, 1865. Morse, L. H. and Hinds, L. J. (1993). Women and ergonomics. Occupational Medicine, 8, 721–731. Murphy, D. C. (1998). Ergonomics and the Dental Care Worker. Washington, DC: United Book Press. National Research Council, Institute of Medicine. (2001). Musculoskeletal Disorders and the WorkPlace: Low Back and Upper Extremities. Washington, DC: National Academies Press. Nesbakken, R. (2004). Ergonomic Challenges in Handle Design: The Challenges of Laparoscopic Surgical Instruments. Retrieved March 28, 2007, from http://www.ivt.ntnu.no/ipd/fag/ PD9/2004/artikkel/Nesbakken.pdf. Nordin, M. and Frankel, V. H. (2001). Basic Biomechanics of the Musculoskeletal System (3rd ed.). Philadelphia: Lippincott Williams & Wilkins. Occupational Safety and Health Administration. (2002). Hand and Power Tools (OSHA 3080, 2002 [Revised]). Washington, DC: Occupational Safety and Health Administration. Available: http://www.osha.gov/publications/osha3080.html Öhman, U., Bylund, P-O., and Björnstig, U. (2002). Impairing injuries among medical personnel. Western Journal of Nursing Research, 24, 788–802. Orr, G. B. (1997). Ergonomics programs for health care organizations. Occupational Medicine, 12, 687–700. Ostlie, D. J. and Holcomb, G. W. (2003). The use of stab incisions for instrument access in laparoscopic operations. Journal of Pediatric Surgery, 38, 1837–1840. Palmer, K. T., Griffin, M. J., Syddall, H., et al. (2001). Risk of hand-arm vibration syndrome according to occupation and sources of exposure to hand-transmitted vibration: A national survey. American Journal of Industrial Medicine, 39, 389–396. Patkin, M. (1967). Ergonomic aspects of surgical dexterity. Medical Journal of Australia, 2, 775–777. Patkin, M. (1971). Ergonomics of diathermy forceps design. Medical Journal of Australia, 2, 657–660. Patkin, M. (1977). Ergonomics applied to the practice of microsurgery. Australian and New Zealand Journal of Surgery, 47, 320–329. Patkin, M. (1980). Handles for Deaver retractors. Medical Journal of Australia, 2, 686–687. Patkin, M. (1981). Ergonomics in microsurgery. Australian and New Zealand Journal of Obstetrics and Gynaecology, 21, 134–136. Patkin, M. (2001). A check-list for handle design. Ergonomics Australia On-Line, 15 (supplement). Retrieved December 21, 2006, from http://ergonomics.uq.edu.au/eaol/handle.pdf. Pelmear, P. L. and Taylor, W. (1992). Clinical picture (vascular, neurological, and musculoskeletal). In P. L. Pelmear, W. Taylor, and D. E. Wasserman (Eds.), Hand-Arm Vibration: A Comprehensive Guide for Occupational Health Professionals. New York: Van Nostrand Reinhold. Pelmear, P. L. and Leong, D. (2000). Review of occupational standards and guidelines for hand-arm (segmental) vibration syndrome (HAVS). Applied Occupational and Environmental Hygiene, 15, 291–302. Quick, N. E., Gillette, J. C., Shapiro, R., et al. (2003). The effect of using laparoscopic instruments on muscle activation patterns during minimally invasive surgical training procedures. Surgical Endoscopy, 17, 462–465. Radwin, R. G., Armstrong, T. J., and Chaffin, D. B. (1987). Power hand tool vibration effects on grip exertions. Ergonomics, 30, 833–855.
TAF-ER562X_WINGR-09-0401-C016.indd 712
11/9/10 1:56:16 AM
Hand Tool Design
713
Radwin, R. G., Armstrong, T. J., and VanBergeijk, E. (1992). Hand and arm vibration and work related disorders of the upper limbs. In P. L. Pelmear, W. Taylor, and D. E. Wasserman (Eds.), Hand-Arm Vibration: A Comprehensive Guide for Occupational Health Professionals. New York: Van Nostrand Reinhold. Radwin, R. G. and Haney, J. T. (1996). An Ergonomics Guide to Hand Tools. Fairfax, VA: American Industrial Hygiene Association. Rassam, S., Wilkes, A. R., Hall, J. E., and Mecklenburgh, J. S. (2005). A comparison of 20 laryngoscope blades using an intubating manikin: Visual analogue scores and forces exerted during laryngoscopy. Anaesthesia, 60, 384–394. Rempel, D. M., Harrison, R. J., and Barnhart, S. (1992). Work-related cumulative trauma disorders of the upper extremity. Journal of the American Medical Association, 267, 838–842. Rivers, D. L., Aday, L. A., Frankowski, R. F., et al. (2003). Predictors of nurses’ acceptance of an intravenous catheter safety device. Nursing Research, 52, 249–255. Sackier, J. M. and Berci, G. (1992). A laparoscopic hazard for the surgeon. British Journal of Surgery, 79, 713. Salvendy, G. (1997). Handbook of Human Factors and Ergonomics (2nd ed.). New York: John Wiley and Sons. Samore, M. H., Evans, R. S., Lassen, A., et al. (2004). Surveillance of medical device-related hazards and adverse events in hospitalized patients. Journal of the American Medical Association, 291, 325–334. Schoenfeld, A. (1998). Ultrasonographer’s wrist—An occupational hazard. Ultrasound in Obstetrics and Gynecology, 11, 313–316. Schwaitzberg, S. D. (2006). The emergence of radiofrequency identification tags: Applications in surgery. Surgical Endoscopy, 20, 1315–1319. Sesto, M. E., Radwin, R. G., Block, W. F., and Best, T. M. (2005). Anatomical and mechanical changes following repetitive eccentric exertions. Clinical Biomechanics, 20, 41–49. Siegel, J. H., Kasmin, F. E., and Cohen, S. A. (1994). Health hazards and endoscopy: The known and newly experienced—A personal report. Endoscopy, 26, 545–548. Simmons, J. C. (2001). Creating a safer workplace to provide quality care. The Quality Letter, 13(4), 2–11. Small, S. D. (2004). Medical device-associated safety and risk: Surveillance and stratagems. Journal of the American Medical Association, 291, 367–370. Smith, A. C., Wolf, J. G., Xie, G.-Y., and Smith, M. D. (1997). Musculoskeletal pain in cardiac ultrasonographers: Results of a random survey. Journal of the American Society of Echocardiography, 10, 357–362. Smith, E. J. and Smith, Y. R. (1983). Smiths’ Reference and Illustrated Guide to Surgical Instruments. Philadelphia: J. B. Lippincott. Stone, R. and McCloy, R. (2004). Ergonomics in medicine and surgery. British Medical Journal, 328, 1115–1118. Stout, N. A. (1992). Occupational injuries and fatalities among health care workers in the United States. Scandinavian Journal of Work, Environment and Health, 18(Suppl. 2), 88–89. Strazdins, L. and Bammer, G. (2004). Women, work and musculoskeletal health. Social Science and Medicine, 58, 997–1005. Strömberg, T., Dahlin, L. B., Brun, A., and Lundborg, G. (1997). Structural nerve changes at wrist level in workers exposed to vibration. Occupational and Environmental Medicine, 54, 307–311. Tan, L., Hawk, J. C., and Sterling, M. L. (2001). Report of the Council on Scientific Affairs: Preventing needlestick injuries in health care settings. Archives of Internal Medicine, 161, 929–936. Tang, B., Hanna, G. B., Joice, P., and Cuschieri, A. (2004). Identification and categorization of technical errors by observational clinical human reliability assessment (OCHRA) during laparoscopic cholecystectomy. Archives of Surgery, 139, 1215–1220.
TAF-ER562X_WINGR-09-0401-C016.indd 713
11/9/10 1:56:16 AM
714
Handbook of Human Factors in Medical Device Design
Thompson, C. J. S. (1942). The History and Evolution of Surgical Instruments. New York: Schuman’s. Tichauer, E. R. and Gage, H. (1977). Ergonomic principles basic to hand tool design. American Industrial Hygiene Association Journal, 38, 622–634. Uchal, M., Brogger, J., Rukas, R., Karlsen, B., and Bergamaschi, R. (2002). In-line vs pistol grip handles in a laparoscopic simulator. Surgical Endoscopy, 16, 1771–1773. Ulin, S. S., Armstrong, T. J., Snook, S. H., and Franzblau, A. (1993). Effect of tool shape and work location on perceived exertion for work on horizontal surfaces. American Industrial Hygiene Association Journal, 54, 383–391. van Veelen, M. A. and Meijer, D. W. (1999). Ergonomics and design of laparoscopic instruments: Results of a survey among laparoscopic surgeons. Journal of Laparoendoscopic and Advanced Surgical Techniques, 9, 481–489. van Veelen, M. A., Meijer, D. W., Goossens, R. H. M., and Snijders, C. J. (2001). New ergonomic design criteria for handles of laparoscopic dissection forceps. Journal of Laparoendoscopic and Advanced Surgical Techniques, 11, 17–26. van Veelen, M. A., Meijer, D. W., Uijttewall, I., et al. (2003). Improvement of the laparoscopic needle holder based on new ergonomic guidelines. Surgical Endoscopy, 17, 699–703. Verma, G. R. (2004). Pressure sore and digital neuropraxia of the thumb in laparoscopic cholecystectomy. Surgical Laparoscopy Endoscopy and Percutaneous Techniques, 14, 178–179. Violante, F., Armstrong, T., and Kilbom, A. (2000). Occupational Ergonomics: Work Related Musculoskeletal Disorders of the Upper Limb and Back. New York: Taylor & Francis. Wauben, L. S. G. L., van Veelen, M. A., Gossot, D., et al. (2006). Application of ergonomic guidelines during minimally invasive surgery: A questionnaire survey of 284 surgeons. Surgical Endoscopy, 20, 1268–1274. Westenskow, D. R. and Silva, F. H. (1993). Device to limit inflation of a pulmonary artery catheter balloon. Critical Care Medicine, 21, 1365–1368. Wilkinson, W. E., Salazar, M. K., Uhl, J. E., et al. (1992). Occupational injuries: A study of health care workers at a northwestern health science center and teaching hospital. AAOHN Journal, 40, 287–293. Woodson, W. E., Tillman, B., and Tillman, P. (1992). Human Factors Design Handbook (2nd ed.). New York: McGraw-Hill.
TAF-ER562X_WINGR-09-0401-C016.indd 714
11/9/10 1:56:16 AM
17 Mobile Medical Devices Richard Stein, BSEE; Michael E. Wiklund, MS, CHFP CONTENTS 17.1 Considerations........................................................................................................ 716 17.1.1 General Considerations ........................................................................... 716 17.1.1.1 Multiple Uses ........................................................................... 716 17.1.1.2 Multiple Users .......................................................................... 716 17.1.1.3 Multiple Use Environments ..................................................... 716 17.1.2 Special Considerations ............................................................................ 717 17.1.2.1 Transport Feasibility ................................................................ 717 17.1.2.2 Device Use While in Motion ................................................... 718 17.1.2.3 Self-Containment ..................................................................... 718 17.1.2.4 Positional Flexibility ................................................................719 17.1.2.5 Protection against Damage ......................................................719 17.1.2.6 Protection against Contamination............................................719 17.1.2.7 User Fatigue .............................................................................719 17.1.2.8 Damage Detection ...................................................................719 17.2 Design Principles ...................................................................................................720 17.2.1 Weight .....................................................................................................720 17.2.2 Moving Parts ...........................................................................................721 17.2.3 Size and Shape ........................................................................................721 17.2.4 Controls ...................................................................................................722 17.2.5 Displays ...................................................................................................724 17.2.6 Audible Signals........................................................................................725 17.2.7 Handles ....................................................................................................726 17.2.8 Power Supply ...........................................................................................728 17.2.9 Cables and Tubes .....................................................................................729 17.2.10 Accessories ..............................................................................................731 17.2.11 Work Surfaces .........................................................................................731 17.2.12 Warnings and Labels ...............................................................................732 17.2.13 Instructions for Use .................................................................................732 17.2.14 Protective Mechanisms ...........................................................................733 17.2.15 Mounting and Security Mechanisms.......................................................734 17.2.16 Brakes ......................................................................................................737 17.2.17 Wheels .....................................................................................................738 17.2.18 Storage .....................................................................................................739 17.2.19 Materials..................................................................................................740 17.2.20 Cleanability .............................................................................................740 715
TAF-ER562X_WINGR-09-0401-C017.indd 715
11/9/10 2:00:21 AM
716
Handbook of Human Factors in Medical Device Design
17.3 Case Studies ........................................................................................................... 741 17.3.1 Portable Ultrasound Device..................................................................... 741 17.3.2 Portable X-Ray Machine .........................................................................743 17.3.3 Transport Ventilator ................................................................................744 References ........................................................................................................................745 While some medical devices are designed for permanent installation in a specific environment (e.g., catheterization laboratory), others are intended to move among various use environments, both indoor and outdoor. A mobile device might move with the caregiver or with the patient. It might be small enough to fit in a lab coat pocket or large enough to fill a hallway and be motor driven on wheels. The common attributes of all mobile medical devices is that they are designed to go where they are needed, whether to monitor a patient, diagnose a medical condition, or deliver a particular therapy. This chapter highlights important human factors considerations in the design of mobile medical devices. It also presents detailed design principles derived from human factors literature as well as professional experience. Readers should note that many of the other chapters, but particularly Chapter 12, “Workstations,” and Chapter 16, “Hand Tools,” include considerable guidance that is pertinent to the design of mobile medical devices.
17.1 CONSIDERATIONS 17.1.1 GENERAL CONSIDERATIONS Mobile medical devices, while subject to most of the same general considerations as stationary devices, have some specific requirements. 17.1.1.1 Multiple Uses Many mobile medical devices are used for widely varying purposes, arguably more so than devices used in a fixed location. For example, the same portable patient monitor might be used in a physician’s office to perform an outpatient procedure, it might be used during patient transport from a postanesthesia care unit to an intensive care unit, and it might be used to monitor a patient during ambulance transport from an accident scene to an emergency department. Such diverse uses pose challenges in the design of a mobile device’s user interface. 17.1.1.2 Multiple Users Varying device uses often means various device users. Continuing with the portable patient monitor example, the users might include anesthesiologists, nurse anesthetists, critical care nurses, and paramedics. These professionals bring different perspectives and skill levels to patient monitoring and therefore pose differing user needs and preferences. In the case of mobile medical devices used in both clinical and nonclinical settings (e.g., in the home), users can range from individuals with advanced medical and technical knowledge to those with no such knowledge (i.e., laypersons), so users’ needs and preference can be even more diverse. 17.1.1.3 Multiple Use Environments As discussed above, mobile medical devices might be used in multiple environments for various purposes. These environments can pose a wide range of design requirements, such
TAF-ER562X_WINGR-09-0401-C017.indd 716
11/9/10 2:00:21 AM
Mobile Medical Devices
FIGURE 17.1
717
Cramped interior of a helicopter ambulance. (From Wiklund Research & Design.)
as excellent display readability at night in subzero temperatures as well as in climatecontrolled operating rooms with intense glare-inducing overhead illumination. A device originally designed for hospital use might end up in a community clinic or even a patient’s home where many environmental variables (e.g., noise, lighting, and power supply) and user variables (e.g., training, cognitive, and physical ability) are less well controlled. It could even be used in unusual environments, such as ambulances (see Figure 17.1), aircraft, or submarines, where routine operational tasks may be much more difficult to perform.
17.1.2 SPECIAL CONSIDERATIONS 17.1.2.1 Transport Feasibility A mobile medical device’s size, shape, weight, and robustness will dictate where and how well it can be transported. Some devices, such as a glucose meter, are sufficiently compact and ruggedly built to be slung from a belt clip and taken almost anywhere. Other devices are mobilized versions of devices that normally stay in one place, such as an X-ray machine, that need to move to incapacitated and/or immobile patients. Such devices are typically shrunk down as much as possible, placed on four wheels, and equipped with multiple handles and bumpers (see Figure 17.2). However, devices of all sizes warrant user testing to confirm their ease of transport. For instance, a handheld glucose meter might be found to make a clinician feel “weighted down,” not fit in lab coat pockets, or attach to a belt in a manner that makes it prone to screen damage (e.g., a cracked screen due to striking a door handle). An X-ray machine might not fit into an elevator. It is far better to make such discoveries in the early prototype stage of device development as opposed to after market launch.
TAF-ER562X_WINGR-09-0401-C017.indd 717
11/9/10 2:00:22 AM
718
FIGURE 17.2
Handbook of Human Factors in Medical Device Design
Portable X-ray machine is narrow enough to fit through doorways.
17.1.2.2 Device Use While in Motion A substantial proportion of mobile medical devices will move among several use environments but be operated only when stationary. However, some mobile medical devices (e.g., infusion pumps) must allow use while the device itself is in motion, such as while rolling down a hospital corridor or being carried alongside a stretcher during a field rescue. 17.1.2.3 Self-Containment Typically, a mobile medical device needs to contain all elements necessary to function at the point of care. As a result, a large proportion of mobile medical devices, ranging from blood glucose testing kits to portable scanners to defibrillators (see Figure 17.3), incorporate storage for necessary accessories and supplies.
FIGURE 17.3
Defibrillator includes a bag that contains the supplies necessary for operation.
TAF-ER562X_WINGR-09-0401-C017.indd 718
11/9/10 2:00:23 AM
Mobile Medical Devices
719
17.1.2.4 Positional Flexibility A mobile medical device might need to be positioned several different ways depending on the use environment, user, and task at hand. For example, a given device might be placed on the patient’s bed, hung on a bed rail, mounted to the wall, placed on a shelf, or even placed on the floor. This requirement calls for a user interface design that facilitates use in all the likely positions, including comfortable access to handles and controls as well as good display readability. 17.1.2.5 Protection against Damage As previously mentioned, mobile medical devices are exposed to many hazards that could render them inoperable. Collisions with obstacles or being dropped can damage sensitive components. (Readers should consult ASTM D4169-2008 and IEC 61010-1-2009 for details regarding drop testing of packaged and unpackaged devices, respectively.) Exposure to severe climate conditions, such as extreme cold, can compromise certain functions, such as the ability to scroll rapidly through menu options presented on a LCD display that responds sluggishly when cold. 17.1.2.6 Protection against Contamination Mobile medical devices, such as those used in rescues, need to withstand contamination by all sorts of materials, including dirt, moisture, and bodily fluids. For example, an emergency ventilator might be placed on a road covered with gasoline-soaked sand and an automobile accident victim’s blood and vomit. Contaminants should not compromise device operation. Furthermore, the ventilator should be easy to clean after use and not suffer appreciable damage due to contamination. 17.1.2.7 User Fatigue A person’s ability to carry or push a mobile medical device depends on many anthropomorphic and biomechanical traits (see Chapter 4, “Anthropometry and Biomechanics”) such as body size, body parts’ range of motion, and muscle strength. However, task time (which relates directly to transport distance) is also a factor. While a nurse might be able to carry a moderately heavy device 75 feet from a storage closet to a patient’s bedside, she or he might not be able to perform the task comfortably if the distance was doubled, thereby doubling the period of exertion. Design solutions to mitigate user fatigue include adding straps to devices to transfer weight from the hand through the shoulder and upper body, placing a device on a rolling cart, and adding a motorized assist. For example, several advanced care hospital beds have motorized undercarriages that facilitate movement through hallways. 17.1.2.8 Damage Detection Given that a mobile medical device might be damaged during transport, the damage and its effect on the device’s operability should be readily apparent. After all, a nonthreatening event can quickly become an emergency when a clinician finds that a critical care device is not working properly. It is rather obvious when a device’s display is cracked or an electrical lead’s bent pins prevent proper connection to a port. However, problems such as a kinked tube, jammed mechanism, or misaligned internal component might not be obvious. Accordingly, devices need to perform thorough self-checks and alert users to operational problems due to damage when it occurs, ideally in time to fix or replace the damaged device before it is used for a critical purpose.
TAF-ER562X_WINGR-09-0401-C017.indd 719
11/9/10 2:00:23 AM
720
Handbook of Human Factors in Medical Device Design
17.2 DESIGN PRINCIPLES The following guidelines should help ensure the safe and effective movement of medical devices to the necessary use locations.
17.2.1 WEIGHT A medical device’s weight can have a significant effect on its mobility. For example, it might control whether or not a device can be hand carried or will require a cart. It might determine a device’s suitability for use on rescue aircraft. Generally, lighter weight is advantageous as long as the device remains sufficiently stable in the expected use conditions. Guideline 17.1: Minimal Weight for Lifting Device weight should be minimized to facilitate lifting.
Guideline 17.2: Minimal Weight for Transport Device weight should be minimized to facilitate its movement from one use location to another as well as to and from storage locations.
Guideline 17.3: Weight for Lifting (≤3 feet) Devices to be lifted from the floor to counter height (≤3 ft/90cm) should not exceed 44 lb. (20 kg).*
Guideline 17.4: Weight for Lifting (≤5 feet) Devices to be lifted from the floor to counter height (≤5 feet/90cm) should not exceed 37 pounds (16.8 kg).†
Guideline 17.5: Carrying at Chest Height The weight of devices that are carried and/or used at chest height while in the user’s hand/ arms should be minimized.
Guideline 17.6: Carrying Short Distances Devices carried up to 33 feet (10 meters) should not weigh more than 42 lb. (19 kg).‡
Guideline 17.7: Transport of Heavy Devices Heavy devices should be placed on carts or the equivalent to facilitate transport. Extremely heavy devices, such as some hospital beds, may be motor propelled to avoid placing strain on the user(s).
Guideline 17.8: Indication of Device Weight Device weight should be indicated on a label that is visible by persons about to lift it. The label should indicate the required number of people required to safely lift the device. * MIL-STD-1472F, p. 139, table XVII. † MIL-STD-1472F, p. 139, table XVII. ‡ MIL-STD-1472F, p. 139, table XVII.
TAF-ER562X_WINGR-09-0401-C017.indd 720
11/9/10 2:00:23 AM
Mobile Medical Devices
721
Guideline 17.9: Number and Placement of Handles Devices should have a sufficient number of properly placed handles to facilitate stable lifting, where appropriate, by one or more persons (see Section 17.2.7).
Guideline 17.10: Features Support Lifting and Carrying As appropriate, devices should incorporate features that facilitate lifting and carrying, such as a shoulder strap.
Guideline 17.11: Weight Distribution for Stability Device weight should be distributed in a manner than ensures its stability when placed in all expected locations and orientations.
17.2.2 MOVING PARTS Most mobile devices have several moving parts. These parts often serve critical functions but are vulnerable to causing and incurring damage. Accordingly, care should be taken to ensure that they do not pose a hazard, can stand up to frequent and potentially rough handling, and will function properly in various use scenarios. Guideline 17.12: Safety of Moving Parts Moving parts should not place users at risk of injury. For example, extending and retracting elements should not create pinch points, and motorized components should not pose a risk of striking a distracted or inattentive user or patient.
Guideline 17.13: Avoid Damage to Moving Parts Generally, moving parts should not be positioned externally where they can strike obstacles and suffer damage. For example, a folding table should not be placed where it could be broken off from a workstation as it passes through a doorway. Alternatively, moving parts should be designed to withstand impact or possibly break away in a controlled manner and be able to be reattached easily.
Guideline 17.14: Secure Moving Parts There should be a means to secure moving parts prior to transport. For example, a device might incorporate a latching mechanism to secure a movable boom.
Guideline 17.15: Range of Motion The range of motion of moving parts should accommodate various use environments where the relationship between the device and its users can vary. For example, an anesthesia workstation that rolls among operating rooms might need to allow its breathing circuit to swing from one side of the workstation to another to suit a particular surgical case.
17.2.3 SIZE AND SHAPE A medical device’s size and shape is often dictated by its myriad components. However, it would be pointless to design a mobile diagnostic machine that is too large to pass through a hospital’s hallway. Therefore, the need for mobility should be a primary consideration
TAF-ER562X_WINGR-09-0401-C017.indd 721
11/9/10 2:00:23 AM
722
Handbook of Human Factors in Medical Device Design
FIGURE 17.4 Infant incubator components are arranged longitudinally to facilitate transport on a stretcher. (From http://www.lammersmedical.com/Img/produkt03.htm. With permission.)
when selecting and arranging components, noting that particular device shapes and sizes can substantially affect ease of handling. Guideline 17.16: Size and Shape Devices should be sized and shaped to suit the associated task but also to fit within the space available in the smallest expected use environment. In some cases, the best shape might be low and flat, while in others it might be tall and rounded.
Guideline 17.17: Compactness for Transport Generally, devices should be as compact as possible—or compactable—to facilitate transport. For example, a fairly large device might fold up to facilitate transport.
Guideline 17.18: Minimal Footprint A device’s “footprint”—the space it consumes on a tabletop or floor—should be sized to suit the intended use environments. In some cases, a device should have no footprint because of the lack of surface space, therefore requiring wall or pole mounting. In other cases, a narrow footprint might be preferable to a wide one or vice versa (see Figure 17.4).
17.2.4 CONTROLS There are many types of controls. Some will be more suitable than others for use in environments characterized by extreme climate conditions, darkness, constant vibration and
TAF-ER562X_WINGR-09-0401-C017.indd 722
11/9/10 2:00:24 AM
Mobile Medical Devices
723
FIGURE 17.5 Portable ultrasound workstation equipped with a lighted keyboard to facilitate use in dim lighting conditions.
jostling, and frequent impacts, for example, making them good design alternatives for mobile medical devices used by first responders. Designers should carefully consider a device’s potential uses when determining which types of controls will be most practical and usable. See also Chapter 7, “Controls.” Guideline 17.19: Illumination of Controls Controls and their associated labels should be illuminated if used in dim and dark lighting conditions. Conventional lighting options include spotlights and backlights (see Figure 17.5).
Guideline 17.20: Preventing Injury from Controls Controls should be shaped to prevent injury to persons who bump against them. For example, they should have rounded as opposed to sharp edges to prevent lacerations or scrapes.
Guideline 17.21: Guarding Controls Controls should be guarded against inadvertent inputs, such as bumps into objects that could turn a rotary knob to a new setting. One possible solution is to recess the controls. Another option is to place protective bars in front of the controls.
Guideline 17.22: Hand Movement Precision Controls should not require fine motor control when the use environment might not permit users to maintain a steady hand. For example, a user might need to adjust controls in the presence of substantial vibration due to air turbulence, riding on a bumpy road, or during transport of a bedridden patient to radiology.
Guideline 17.23: Immediate Multichannel Control Feedback Controls should provide immediate, multichannel (e.g., visual, audible, and tactile) feedback to ensure reliable operation in challenging use environments where one or more of the user’s sensory channels may be overloaded or masked. For example, a rotary knob incorporating distinct detents can provide useful tactile feedback when the user is not looking at the control and there is a lot of noise (e.g., if an ambulance’s alarm is blaring, and there is chatter on the rescue service radio).
TAF-ER562X_WINGR-09-0401-C017.indd 723
11/9/10 2:00:24 AM
724
Handbook of Human Factors in Medical Device Design
Guideline 17.24: Avoid Computer Mice in Mobile Devices Pointing devices used to navigate computer-based menu options and make selections should not be subject to loss or damage. In this regard, trackballs, touch pads, and touch screens have notable advantages over styli and computer mice. Specifically, a tethered computer mouse might “flop” around during transport and be vulnerable to cable damage. Conversely, the other common pointing devices are built in and protected components.
Guideline 17.25: Miniaturized Keyboards and Keypads Compact devices might warrant smaller than conventional keyboards and keypads. Generally, such keyboards should have a QWERTY key arrangement unless they must assume an unusual shape (e.g., taller than wide). Unusual keyboard sizes call for an alphabetical key arrangement that is reasonably familiar and affords reasonable throughput. That said, users are becoming more adept at entering information via small QWERTY keyboards due to their experience using mobile phones, for example. User testing is required for all nonstandard text entry designs.
Guideline 17.26: Keyboard/Keypad Protection Depending on the expected use environment, a keyboard/keypad might warrant some means of protection against fluid ingress and material contamination. For example, a keyboard might require a plastic cover. Alternatively, the keys could be molded into a single, impermeable surface.
Guideline 17.27: Controls Usable in All Conditions Controls should be operable in all expected environmental conditions, including snowy and sandy settings, as well as extreme heat, cold, and humidity. For example, push buttons should not jam if contaminated by freezing rain or sand.
Guideline 17.28: Controls Remain Accessible If a device will be used during patient transport (e.g., a portable ventilator), its controls should remain accessible to those individuals attending the patient during transport.
17.2.5 DISPLAYS Display options abound, ranging from LED readouts to index card-sized LCD displays to large CRTs. As with controls, some will be more suitable than others for specific use scenarios. Designers should carefully consider a device’s potential uses to determine which types of displays will be most practical and usable. A larger display is often more usable but might make a device more difficult to handle. See also Chapter 8, “Visual Displays.” Guideline 17.29: Legibility of Displays Displays should be legible in all expected lighting conditions, which might range from intense sunlight to total darkness.
Guideline 17.30: Automatic Adjustment of Displays Where appropriate, displays should automatically adjust to ambient lighting levels to optimize readability and minimize eye strain.
TAF-ER562X_WINGR-09-0401-C017.indd 724
11/9/10 2:00:24 AM
Mobile Medical Devices
725
Guideline 17.31: Prevent or Minimize Glare Displays should incorporate a means to prevent or minimize glare to ensure legibility.
Guideline 17.32: Impact Resistant Displays Displays should withstand moderate impacts, such as a user dropping a flashlight on the display or banging the display against a door handle when entering a room.
Guideline 17.33: Scratch Resistant Displays Displays that might be subject to physical abuse should be scratch resistant.
Guideline 17.34: Transparent Display Coverings Display coverings should not cloud or change color (i.e., yellow) over time.
Guideline 17.35: Display Dimmer Control In cases where users might wear night vision goggles, there should be a means to dim the display. So that its emitted light does not cause artificially generated images to appear over exposed. This feature would be similarly applicable to use enviornments with wide range or illumination levels (e.g., outdoors).
Guideline 17.36: Display Readability under Extreme Conditions Displays should be readable in all expected environmental conditions, including extreme heat, cold, and humidity.
Guideline 17.37: Oversized Information Displays Information (e.g., printed and displayed text and numbers) should be somewhat oversized to ensure legibility in the presence of vibrations due to air turbulence or riding on a bumpy road, for example.
Guideline 17.38: Expected Viewing Angles Displays should be readable from all expected viewing angles, which could vary widely depending on where a device is placed in a given use environment. For example, a device might be placed (or mounted) on the floor, patient bed, over-bed table, shelf, wall, or ceiling.
17.2.6 AUDIBLE SIGNALS Audible signals are a particularly effective way to communicate in a visually complex environment. Used judiciously, a specially coded audio signal can draw immediate attention and communicate critical information. Overused or poorly designed audio signals can confuse and irritate users. In the case of mobile medical devices, audio signals also have to be compatible with the intended use environments and communicate reliably to diverse users. See also Chapter 10, “Alarms.” Guideline 17.39: Volume of Audio Signals Audio signals (e.g., heartbeat tones and alarms) should be sufficiently loud and distinct to ensure detection in the expected use conditions. Chapter 3, “Environment of Use,” provides a table of ambient noise levels.
TAF-ER562X_WINGR-09-0401-C017.indd 725
11/9/10 2:00:24 AM
726
Handbook of Human Factors in Medical Device Design
Guideline 17.40: Adjustment of Audio Signals Audio signal volume should be adjustable to suit various use environments, which might range from a critical care unit to a noisy street corner or airplane flying at full throttle.
Guideline 17.41: Selective Voice Messages Voice messages should be considered for use when the user is unfamiliar with the device, must act urgently, and might not be able to read the display (i.e., illiterate or out of view). For example, voice prompting is considered an appropriate means for automated external defibrillators to direct user actions.
17.2.7 HANDLES A handle is the common earmark of a mobile medical device. After all, mobile devices are likely to be held, wheeled, and/or lifted. So, designing the right handle—one that ensures secure, comfortable, and effective device control—is paramount. The wrong handle can cause wrist strain, cause users to lose their grip on a device, and limit control over a potentially hazardous object (e.g., a 400-pound bed rolling swiftly down a crowded corridor). See also Chapter 12, “Workstations.” Guideline 17.42: Stable Grip Handle Location Handles should be placed where they ensure a stable grip. Generally, the optimal handle location on hand carried devices is in vertical alignment with the device’s center of gravity.
Guideline 17.43: Handle Dampens Swinging Motion Handles should prevent or at least dampen the swinging motion of device components.
Guideline 17.44: Simultaneous Carriage and Use If a device will be used while also being carried, its handles should be designed to enable simultaneously carrying and operating the device.
Guideline 17.45: Non-Conductivity of Handles Handles should not conduct heat or electricity.
Guideline 17.46: Handle Slipperiness Handle material or coatings should minimize slipperiness, especially when the device may be used in an environment where fluids or lubricants could get on the handle or the user’s hands.
Guideline 17.47: Gloved Use of Handles Handles should ensure a secure, comfortable grip, especially when gripped by a wet, gloved hand.
Guideline 17.48: Discomfort or Injury from Handles Handles should be devoid of pinch points and protuberances that could cause discomfort and/or injury as well as compromise the integrity of protective gloves.
TAF-ER562X_WINGR-09-0401-C017.indd 726
11/9/10 2:00:24 AM
Mobile Medical Devices
FIGURE 17.6
727
A large handle makes it easier to move this cart-mounted cardiograph.
Guideline 17.49: Handle Size Handles should accommodate the widest possible range of hand sizes (e.g., 1st-percentile female to 99th-percentile male). This range should be extended to include children if children are among the intended device users.
Guideline 17.50: Conspicuity of Handle Handles should be visually apparent (i.e., conspicuous), thereby encouraging their use and preventing users from grasping the device by delicate components (see Figure 17.6).
Guideline 17.51: Handle Enables Neutral Wrist Position Handles should enable a neutral or midrange wrist position to minimize strain that could cause fatigue or pain or possibly lead to cumulative trauma.
Guideline 17.52: Padding of Handles Where appropriate, handles should be padded to reduce point stresses on the hands. However, such padding should be designed so that it does not appreciably reduce grip stability.
Guideline 17.53: Handle Grip Style Handles should enable various grip styles rather than force users to adopt a specific grip style that might not be optimal for particularly large or small hands.
Guideline 17.54: Power Grip for Handles Handles should enable a power grip whereby the fingers and thumb are able to wrap fully around the handle and the palm is able to maintain contact with the handle surface.
TAF-ER562X_WINGR-09-0401-C017.indd 727
11/9/10 2:00:24 AM
728
Handbook of Human Factors in Medical Device Design
Guideline 17.55: Two Handles for Two Hands Generally, devices carried with two hands should have separate handles on opposite sides of the device.
Guideline 17.56: Handles for Two-Person Use If a device’s weight calls for a two-person lift, the handles should be optimally positioned to accommodate two people and distribute the load evenly between them.
Guideline 17.57: Handles for Person-to-Person Exchange In cases where a device should be handed from one person to another, the handles should be sufficiently large and/or numerous to permit secure exchanges.
17.2.8 POWER SUPPLY Many mobile medical devices require electrical power. Primary sources include AC mains (i.e., wall outlets) and batteries, although the occasional device might use other power sources, such as solar power or air pressure. Keeping a mobile medical device powered with adequate power in reserve is often crucial to patient care. For example, a portable patient monitor will need to have enough battery power to enable a patient to be transported from a critical care unit to radiology and back with plenty of time to spare. Mobile medical devices should limit the need for users to manage device power. Moreover, they should protect users from ever running out of power at the wrong moment. There are considerable trade-offs associated with powering mobile medical devices while also limiting their size and weight. For example, high-capacity batteries might not always be the best solution because of their size and weight. Guideline 17.58: Power-Up Duration The time required for the device to power up to the initialized ready state should be minimized to enable its use in urgent scenarios. Rapid power-up will enable users to transport a device to the patient’s location and begin using it almost immediately rather than waiting for the device to become fully operational.
Guideline 17.59: Battery Life Indication Battery-powered devices should indicate the remaining battery life. If the battery technology does not permit precise estimates, at least a rough approximation of battery life should be provided.
Guideline 17.60: Battery Charging Time Devices should indicate the time required and/or remaining to fully charge batteries.
Guideline 17.61: Battery Life Battery life should exceed the demands of the expected use scenarios. As such, a device that might need to run on battery power for a 3-hour procedure should have substantially more than 3 hours of available power from a fully charged battery.
Guideline 17.62: Power Saving Devices should incorporate power-saving features (e.g., screen dimming and time-out) as long as they do not adversely affect the device’s usability.
TAF-ER562X_WINGR-09-0401-C017.indd 728
11/9/10 2:00:25 AM
Mobile Medical Devices
729
Guideline 17.63: Low Battery Indication Devices should provide users with an alert when the battery is running low. Alerts should draw more attention as battery power approaches the point of depletion.
Guideline 17.64: Live Swapping of Batteries Optimally, it should be possible to swap batteries without powering down the device. If a “hot swap” is not possible, the device should immediately (or rapidly) return to the same operational state as before the battery swap.
Guideline 17.65: Battery Swap Simplicity Swapping batteries should not require special tools or knowledge.
Guideline 17.66: Nonvolatile Memory Devices should be capable of storing critical operational and clinical information in nonvolatile memory in the event of an unanticipated loss of power.
Guideline 17.67: Built-In Charger Larger, heavier devices should have a built-in charger, assuming that the added weight will not significantly affect the device’s mobility.
Guideline 17.68: Automatic Recharging Device batteries should recharge automatically when the device is plugged into AC main power.
Guideline 17.69: Power Cord Retraction Power cords should either retract automatically or wrap securely around a dedicated bracket.
Guideline 17.70: Power Switch Accessibility Power switches should be placed in a readily accessible location. Power switches placed on the side or back of a mobile medical device might be inaccessible if the device is used in cramped spaces that require several devices to be mounted side by side against a wall.
Guideline 17.71: Power Switch Guarding Power switches should be guarded against inadvertent actuation.
17.2.9 CABLES AND TUBES Some medical devices incorporate numerous cables and tubes. Keeping them all organized can be a challenge, even for experienced device users. To facilitate the organization task, some device manufacturers have developed cable and tube management mechanisms. In the case of mobile medical devices, the cable and tube management challenge extends beyond proper routing to include effective stowage and protection. See also Chapter 9, “Connections and Connectors.”
TAF-ER562X_WINGR-09-0401-C017.indd 729
11/9/10 2:00:25 AM
730
Handbook of Human Factors in Medical Device Design
Guideline 17.72: Coding of Connections and Connectors Cables, tubes, and their associated ports should be coded to preclude improper connections, particularly by users working in haste (e.g., first responders). For example, a sensor cable’s connector might be shaped, colored, and textured to ensure proper identification and insertion into the correct port.
Guideline 17.73: Connection Feedback Cables and tubes should provide visual, audible, and tactile feedback indicating complete (i.e., full and secure) attachment and detachment.
Guideline 17.74: One-Handed Operation of Connectors Cables and tubes should be designed for one-handed attachment and detachment.
Guideline 17.75: Accessibility of Connectors Commonly used cable and tube connections should be placed on the device’s front panel to ensure accessibility if the device is used in cramped spaces that require several devices to be mounted side by side against a wall (see Figure 17.7).
Guideline 17.76: Connector Storage A means should be provided to retain or stow cables and tubes during transport, thereby preventing them from catching on objects, striking the ground, and/or becoming contaminated (see also Section 17.2.18). Figures 17.3 and 17.8 show two different methods of storage.
FIGURE 17.7 Noninvasive blood pressure monitor’s cuff cable connects to a port that is conveniently located on the front panel.
TAF-ER562X_WINGR-09-0401-C017.indd 730
FIGURE 17.8 Blood pressure and vital signs monitor includes a basket to hold associated cables.
11/9/10 2:00:25 AM
Mobile Medical Devices
731
Guideline 17.77: Connector Strain Relief Cables and tubes should incorporate strain relief in cases where they might be exposed to physical stresses (i.e., sharp bending and stretching).
Guideline 17.78: Retent of Connectors Cables and tubes and their associated connections should incorporate retention mechanisms (e.g., turn and lock keying) that prevent them from being pulled out inadvertently.
17.2.10 ACCESSORIES Fixed medical devices do not have to include all necessary accessories (Figure 17.8). Accessories may be stored in local cabinets, for example. However, a mobile medical device usually needs to incorporate storage (e.g., drawers, trays, and/or pouches) for its necessary accessories. Proper storage compartments ensure that users will not begin a medical procedure only to discover that they are missing essential device-related items. It also reduces the risk that the items will be damaged or contaminated. Guideline 17.79: Accessory Availability Devices should provide the means to mount or stow all necessary accessories to ensure operability in the expected use scenarios.
Guideline 17.80: Protection of Accessories Accessories should be protected from contamination, damage, and inadvertent detachment when in use, transport, and storage.
17.2.11 WORK SURFACES Considering the paperwork associated with most medical procedures, a workspace is often a necessity. Even if paperwork has been abandoned in favor of electronic medical records and forms, work surfaces might still be necessary to facilitate procedures. For example, a diagnostic procedure might require the convenient placement of biopsy tissue containers, syringes, bandages, and other disposables. The proper sizing and placement of work surfaces can contribute substantially to a mobile medical device’s usability, particularly if alternative work surfaces are not available in the common use environments. See also Chapter 12, “Workstations.” Guideline 17.81: Work Surface Size As appropriate, devices should provide a work surface of sufficient size to facilitate associated tasks. For example, an endoscopy workstation might incorporate a folding writing surface enabling the endoscopist to write his or her report.
Guideline 17.82: Work Surface Illumination Work surfaces on devices used in dim or dark lighting conditions should be illuminated.
Guideline 17.83: Work Surface Texture Work surfaces used for completing paperwork should be relatively smooth to permit legible writing. Work surfaces for supplies and components may be textured to reduce slippage.
TAF-ER562X_WINGR-09-0401-C017.indd 731
11/9/10 2:00:25 AM
732
Handbook of Human Factors in Medical Device Design
Guideline 17.84: Writing Instrument Holder Work surfaces used for paperwork should incorporate a place to securely hold a writing instrument.
17.2.12 WARNINGS AND LABELS Some mobile medical devices are used infrequently, explaining why they might be shared among several care units and, therefore, require mobility. When users are less familiar with a device’s functions, then warnings, labels, and on-device instructions may be of greater importance, potentially preventing incorrect device use and facilitating tasks. See Chapter 13, “Signs, Symbols, and Markings? Guideline 17.85: Function Identification Label Devices should incorporate a conspicuous label indicating the device’s primary function. This will enable individuals unfamiliar with the device to identify it and determine where it belongs.
Guideline 17.86: Warning Pictograms Warnings should incorporate pictograms to communicate critical information at a glance to users who might have only seconds to look at them before responding. For example, Figure 17.9 shows pictographic instructions for device use.
Guideline 17.87: Hierarchical Labeling Scheme All ports and components should be labeled to facilitate immediate identification. Optimally, the labeling scheme should be hierarchical to eliminate redundancies and minimize the amount of labeling. For example, four connector ports could be labeled Port A, Port B, Port C, and Port D, but this has inherent redundancies (i.e., repetition of the word “Port”). In a hierarchical labeling scheme, the word “Port” would appear once over the labels A, B, C, and D. This allows for coherent labeling in small panel spaces, which are typical of many mobile medical devices.
Guideline 17.88: Label Placement Labels should be placed where they will not be blocked from view while connecting cables and tubes.
Guideline 17.89: Use of Protective Gear Warning Label As warranted, devices should warn users to wear protective gear, such as eye protection and lead-lined aprons and neck protectors.
17.2.13 INSTRUCTIONS FOR USE Generally, instructions for use are stored in bookshelves and cabinets, away from the point of use. However, users may not know where to find the instructions for use when they are needed. For mobile devices, at least those that move considerable distances away from a base location, users may need to take the instructions for use where with them. Device-embedded, computer-based instructions and/or online instructions may meet this need. That said, not all use environments provide immediate Internet access. In lieu of online instructions, requiring
TAF-ER562X_WINGR-09-0401-C017.indd 732
11/9/10 2:00:25 AM
Mobile Medical Devices
FIGURE 17.9 Automated external defibrillator presents pictographic instructions on the device’s front panel. (Courtesy of http://zoll.com/page. aspx?id=116.)
733
FIGURE 17.10 Quick reference card is attached by a chain to this intravenous infusion pump.
Internet access devices need to incorporate a space for an owner’s manual or jobs aids such as a quick reference guide or troubleshooting checklist (Figure 17.10). Guideline 17.90: Instruction Availability Devices should provide instructions for use in a manner than cannot be separated from the device. Alternative approaches include online help, inseparable quick reference cards, and written instructions on the device in a visible location (see Figure 17.9).
Guideline 17.91: Instruction Language Instructions for use should be written in the users’ language. For some mobile devices, this requirement might necessitate on-device labeling and instructions for use to be in more than one language (e.g., English and Spanish) see Chapter 19, Cross Cultural and Cross National Design.
17.2.14 PROTECTIVE MECHANISMS Given that mobile medical devices are subject to routine, albeit benign abuse, they need to incorporate myriad protective mechanisms to minimize damage and ensure device durability. Guideline 17.92: Hazard Protection Devices should incorporate mechanisms protecting them from the following kinds of events: • • • •
Impact with a door jamb or wall during transport Impact with a closing door Impact by a patient’s flailing arm or leg Collision with other equipment (e.g., crash cart)
TAF-ER562X_WINGR-09-0401-C017.indd 733
11/9/10 2:00:25 AM
734
Handbook of Human Factors in Medical Device Design
FIGURE 17.11 Laptop computer–sized ultrasound scanner features a clamshell design, allowing the keyboard and display to be protected during transport (see the related first case study at Section 17.3.1). (From SonoSite, Inc. With permission.) • • • • • •
Drops from the height of a counter or bed Splashed fluids (e.g., IV fluids, blood, vomit, urine) Compression beneath other equipment and supplies Being sat or laid on Jarring motion due to crossing a threshold or obstacle Collision with a dangling object (e.g., tethered remote control)
Potentially useful protective mechanisms include protective covers, bumpers, high-strength enclosures, shock-protection components, mounts, shock-absorbing chassis and wheels, and warning labels that identify physical hazards and proper avoidance techniques (see Figure 17.11).
17.2.15 MOUNTING AND SECURITY MECHANISMS Some care environments, such as crowded critical care units and ambulances, offer little room for mobile medical devices. Devices can be mounted on available surfaces to get them out from underfoot or to create clear work surfaces. Sometimes, there is also a need to prevent the devices from unauthorized removal and outright theft. Besides providing support and preventing loss, mounting and security mechanisms need to be easy to engage and disengage. Guideline 17.93: Provide Mounting Mechanism As appropriate, mounting mechanisms (e.g., brackets or tie-downs) should be provided to hold a device securely in place. For example, a transport ventilator should incorporate a means to be mounted securely to a helicopter’s or ambulance’s wall or ceiling (see Figure 17.12).
Guideline 17.94: Compatibility with Multiple Mountings Devices should be compatible with as many common mounting solutions as possible. For example, portable patient monitors should be suitable for mounting on bed rails, wall mounts, and IV poles (see Figures 17.13 and 17.14).
TAF-ER562X_WINGR-09-0401-C017.indd 734
11/9/10 2:00:26 AM
Mobile Medical Devices
FIGURE 17.12
735
Patient monitor is placed in bag mounted to helicopter ambulance’s side wall.
Guideline 17.95: Ability to Carry on Person Small devices should incorporate mechanisms to keep them in their desired place. For example, digital thermometers and insulin pumps might incorporate a belt clip.
Guideline 17.96: Fit in User’s Pocket Small, lightweight devices, such as glucose meters, should be sized to fit in pockets.
FIGURE 17.13 This portable patient monitor is designed to hang from a bed rail or sit on shelves.
TAF-ER562X_WINGR-09-0401-C017.indd 735
FIGURE 17.14 The space limitations in critical care environments often require IV infusion pumps to be stacked vertically on IV poles.
11/9/10 2:00:26 AM
736
Handbook of Human Factors in Medical Device Design
Guideline 17.97: Safe Mounting Mechanisms Mounting mechanisms should not pose a physical hazard to users, most notably when the device is detached or inadvertently dislodged.
Guideline 17.98: Simple Mounting Procedures Mounting procedures should not require special operational knowledge or undue strength or dexterity.
Guideline 17.99: No Mounting Tools Mounting and dismounting procedures should not require special tools or, ideally, any tools at all.
Guideline 17.100: One-Handed Mounting Lightweight devices should be able to be mounted and dismounted using one hand.
Guideline 17.101: Feedback about Mounting Mounted devices should provide strong visual, audible, and/or tactile feedback that the device is properly secured.
Guideline 17.102: Locking Mounts When appropriate, the mounting mechanism should incorporate a lock to prevent device misplacement or theft. Locking mechanisms can also decrease the risk of inadvertent detachment in moving workspaces such as the inside of an ambulance.
Guideline 17.103: Preventing Device Movement Cart-mounted devices should incorporate a means (e.g., a wire tether and lock, similar to those used to secure laptop computers) to keep them in place, particularly when safety or security is an issue.
Guideline 17.104: Preventing Unauthorized Access Device locks, such as those found on infusion pumps that deliver narcotic drugs, should be durable enough to prevent unauthorized access while being easily opened by authorized personnel when necessary.
Guideline 17.105: Indication of Use As appropriate, devices or their storage compartments should indicate when they have been used, indicating, for example, that they might need replacement, restocking, or servicing (see Figure 17.15).
Guideline 17.106: Accessibility in an Emergency Devices that must be used urgently for critical (i.e., lifesaving) purposes should not be lockable or should provide authorized users with an alternate, rapid, and reliable means of access in emergency conditions.
TAF-ER562X_WINGR-09-0401-C017.indd 736
11/9/10 2:00:27 AM
Mobile Medical Devices
737
FIGURE 17.15 This defibrillator is contained in a storage compartment that alarms when opened.
17.2.16 BRAKES As with automobiles, rolling mobile devices need good brakes. Brakes not only keep devices from rolling away from the desired location but also help keep a device motionless, which might be critical to usability and delivering proper care. Guideline 17.107: Brakes on Rolling Devices Rolling devices (e.g., carts and large machines) should incorporate brakes to keep them stationary.
Guideline 17.108: Conspicuous Brake Controls Brake controls should be visually conspicuous to enable users to find them quickly.
Guideline 17.109: Brake Control Placement Brake controls should be placed where a user can apply and release them without assuming awkward postures. For example, the device might incorporate a foot-actuated control placed at the device’s base or a hand brake located at elbow height (see Figure 17.16).
Guideline 17.110: Readily Apparent Brake Status Brake status (i.e., applied or released) should be visually or otherwise easily apparent.
Guideline 17.111: Brake Ease of Use The number of steps and time required to apply the brakes should be minimized. Optimally, applying the brake should require a single, intuitive action.
FIGURE 17.16
Foot brake is positioned for rapid access.
TAF-ER562X_WINGR-09-0401-C017.indd 737
11/9/10 2:00:27 AM
738
Handbook of Human Factors in Medical Device Design
17.2.17 WHEELS Properly designed wheels can add considerably to the ease of moving mobile medical devices, particularly heavy ones. For example, a wheel with a sufficiently large diameter enables a heavy ultrasound workstation to pass smoothly over power cords and doorway thresholds. Guideline 17.112: Wheel Size Wheels should be sized to ensure that rolling equipment can traverse obstacles, such as gaps between floors and elevator cabs, thresholds between rooms, and cables strewn on the floor.
Guideline 17.113: Smooth Rolling Wheels Wheels should be sized and made of appropriate material to enable smooth movement over the expected surfaces. For example, a heavy workstation should have wider wheels to enable smooth movement over carpeting or uneven surfaces (see Figure 17.17).
Guideline 17.114: Wheel Tracking Generally, carts should have fixed rear wheels and castered front wheels to facilitate accurate tracking and steering.
Guideline 17.115: Castered Wheels In cases where a cart must turn in a tight radius and/or rotate about its vertical axis, all wheels should be castered.
Guideline 17.116: Minimize Wheel Flutter Castered wheels should be designed to minimize flutter (swinging back and forth) and chatter.
Guideline 17.117: Minimize Wheel Noise Wheels should not make disturbing noises as they rotate. Optimally wheels should be silent when the device rolls on smooth surfaces.
Guideline 17.118: Minimize Wheel Vibration Wheels should not make disturbing vibrations as they rotate.
FIGURE 17.17 Large wheels enable this hospital bed to overcome obstacles, such as gaps between a hallway and elevator cab.
TAF-ER562X_WINGR-09-0401-C017.indd 738
11/9/10 2:00:27 AM
Mobile Medical Devices
739
Guideline 17.119: Locking Wheels In certain cases, locking casters is an appropriate means of braking. In such cases, the brake status should be visually obvious, and it should be simple to apply and release using one foot (see Section 17.2.16). Locking casters is not necessarily a desirable solution for all four wheels on a cart because of the time required to apply and release all brakes and the problems that might arise from overlooking one or more wheel brakes.
17.2.18 STORAGE It is frustrating and potentially hazardous to arrive at a patient care location without all of the necessary items. Convenient, well-placed storage helps ensure that procedures can be performed efficiently. Guideline 17.120: Provide Onboard Storage Devices should incorporate onboard storage of essential accessories and supplies, unless such storage makes the devices undesirably heavy and bulky (see Figures 17.18 and 17.19).
Guideline 17.121: Locking Storage As appropriate, storage spaces should incorporate locks to protect valuable or controlled (e.g., narcotic drugs) contents from loss or theft. The means to unlock storage compartments should be readily available to authorized individuals.
FIGURE 17.18 This mobile device incorporates a pullout basket (see bottom of cart) for accessory stowage.
TAF-ER562X_WINGR-09-0401-C017.indd 739
FIGURE 17.19 These ultrasound transducers have dedicated holders.
11/9/10 2:00:28 AM
740
Handbook of Human Factors in Medical Device Design
FIGURE 17.20 Rubberized side grips make this digital thermometer easier to hold. (From Wiklund Research & Design.)
17.2.19 MATERIALS Generally, mobile medical devices need to be more durable than fixed devices. They are exposed to a wide variety of environmental conditions that can degrade certain materials. Thus, material selection is critical to long product life. Mobile medical devices also tend to receive more and rougher handling, so it is important to choose materials that afford a secure grip. It is also important for materials to resist contamination and be easy to clean (as discussed in Section 17.2.20). Guideline 17.122: Use Durable Materials Devices should be constructed of durable materials that can tolerate the expected conditions of use.
Guideline 17.123: Resistance to Environmental Exposure Materials should not change appearance or physical properties because of exposure to various climatic conditions, including extreme heat or cold and sustained, bright sunlight. For example, protective bumper strips should not crack or discolor.
Guideline 17.124: Grip Comfort Handles and gripping surfaces should afford a comfortable grip (see Figure 17.20). For example, a device handle might be covered with a soft-feeling material to avoid pressure points, presuming that the material meets other requirements for durability and cleanability. See also Section 17.2.7.
17.2.20 CLEANABILITY Mobile medical devices, particularly those used in messy procedures or outdoors (e.g., accident sites), are routinely soiled. Such devices have to be easy to clean both for
TAF-ER562X_WINGR-09-0401-C017.indd 740
11/9/10 2:00:28 AM
Mobile Medical Devices
741
health reasons and to minimize workload, thereby increasing the likelihood of thorough cleaning. Guideline 17.125: Use of Common Cleansers Devices should be easy to clean using cleansers normally found in the use and/or storage environment.
Guideline 17.126: Detachable Components Ease Cleaning Fabric components, such as pouches used to hold accessories, should be removable to enable laundering.
Guideline 17.127: Resistance to Contamination Fabric components should not trap contaminants. For example, they should have a fine texture that shreds debris as opposed to a coarse texture (e.g., open weave) that traps debris.
Guideline 17.128: Resistance to Ground Contamination Devices placed on the ground should not pick up dirt and debris that could be transported into other care environments, such as an ambulance’s interior or emergency rooms.
17.3 CASE STUDIES The following case studies exemplify design for device mobility (i.e., portability or transportability). While the various devices differ in terms of size and method of movement, all are able to serve critical purposes by virtue of their mobility.
17.3.1 PORTABLE ULTRASOUND DEVICE The iLook Portable Ultrasound Device was designed for use in a wide range of environments, including physician’s offices, where the use of a full-size ultrasound scanner would be impractical. Rather than taking the form of a large cart, the device is sufficiently small and lightweight to enable hand carrying, thereby facilitating new uses and greater convenience. One of the manufacturer’s primary design challenges was to ensure that the display unit could be held comfortably in one hand while holding the device’s tethered transducer in the other hand for the duration of an exam (5 to 10 minutes) (see Figure 17.21). Ensuring that the handle would accommodate a wide range of hand sizes was critical to such use. A related challenge was to facilitate carrying the scanner from place to place using a “briefcase” style of grip. To meet the challenges, the design team drew heavily on anthropometric data to size the handles on several exploratory models that would be subjected to user testing. The designers found that small variations in the handle opening size (±0.2 cm) made a big difference in the users’ ability to slide their hands through the opening and brace the display unit at the desired viewing angle. Ultimately, they designed the device to accept handle pads, enabling a user/owner (e.g., an obstetrician working in her own office) to tailor the device to her particular hand size. The size and shape of the handle also enables a comfortable, briefcase grip when carrying the device but not scanning with it.
TAF-ER562X_WINGR-09-0401-C017.indd 741
11/9/10 2:00:28 AM
742
Handbook of Human Factors in Medical Device Design
FIGURE 17.21 Users can support the device with one hand while manipulating the unit’s single transducer with the other hand. (From SonoSite, Inc. With permission.)
To enhance portability (i.e., minimize size and weight), the device employs mostly touch screen–based controls as opposed to dedicated hardware controls that would have increased its size and weight (see Figure 17.22). Initially, the development team felt that the touch screen should automatically deactivate after a short idle period to prevent accidental inputs that might be more common in a handheld devices as compared to a cart-type device. However, user testing of a prototype showed that users strongly preferred an active touch screen at all times during an exam, accepting that they would need to exercise greater care not to touch it by accident. The designers also anticipated that users might prefer to use a stylus. However, user testing determined that the vast majority of clinicians would use their fingertips or
TAF-ER562X_WINGR-09-0401-C017.indd 742
11/9/10 2:00:28 AM
Mobile Medical Devices
743
FIGURE 17.22 The use of touch screen controls enabled designers to minimize the device’s size and weight. (From SonoSite, Inc. With permission.)
perhaps a capped pen rather than bothering with a stylus. User testing also determined that clinicians needed a stand in which to hold the device when it was not in use.
17.3.2 PORTABLE X-RAY MACHINE Modern portable X-ray machines allow the acquisition of high-quality digital images in primary treatment areas rather than requiring the patient to be transported to the Radiology department. This capability often speeds up diagnoses and eliminates the complexities associated with patient transport, including the temporary cessation of therapies. The Definium AMX 700 was designed to incorporate a number of features that facilitate its usability (Figure 17.23): • Overall compactness to facilitate movement by one person (e.g., the X-ray technician) • Casters that allowed the machine to be maneuvered easily among various use locations within a hospital (e.g., the emergency room and intensive care unit) • Readily accessible holders for power cords • Wireless data transmission to hospital information systems • A touch screen for rapid, intuitive interactions that eliminated the need for tethered cursor control devices • A flexible positioning arm with 270 degrees of rotation to simplify targeting • Sufficient reach to image a patient on a double bed • Controls designed to allow one-handed use • Onboard storage of a spine board (a rigid plate placed underneath the patient prior to X-ray imaging)
TAF-ER562X_WINGR-09-0401-C017.indd 743
11/9/10 2:00:29 AM
744
Handbook of Human Factors in Medical Device Design
FIGURE 17.23 Usable portable. Mobile X-ray machine. (From GE Healthcare. With permission.)
17.3.3 TRANSPORT VENTILATOR The Oxylog 3000 is an emergency and transport ventilator intended for use in diverse settings, including field locations, ambulances, and helicopters (see Figure 17.24). The designers selected a large-diameter wide handle to improve comfort during long-distance transports, such as from an emergency response vehicle to a rescue site. Critical design criteria were small footprint and minimal weight. With one hand, users could snap the device onto a variety of poles found on stretchers, trolleys, and various surfaces within transport vehicles. A built-in, swappable battery provides up to 4 hours of power. An integrated microchip accurately indicates the amount of remaining power, thereby enabling users to prevent an unexpected power loss. Wall mounts incorporated an automatic power connection mechanism that eliminated the need for users to plug the device into an AC power source and ensured that batteries were charged whenever possible. All tubes connected on the device’s left side so that it could be placed in a corner without blocking access. Large controls facilitated gloved user operation. The bright LCD display enabled use in various lighting conditions (i.e., bright sunlight and darkness) as often exists at rescue sites. A custom carrying system coupled the ventilator with a gas tank and provided a protective storage space for the large breathing tube. The device was designed to withstand a drop from
TAF-ER562X_WINGR-09-0401-C017.indd 744
11/9/10 2:00:30 AM
Mobile Medical Devices
745
FIGURE 17.24 Mobile emergency ventilator in various configurations suited to hospital and field use. (From Drägermedical. With permission.)
75 cm (29.5 inches)—a bit higher than table height—and to work in low temperatures. Its bright orange enclosure helped users locate the device.
REFERENCES American Society of Testing and Materials, Standard Practice for Performance Testing of Shipping Containers and Systems, D4169, 2008. International Eletrotechnical Commission, International Standard, International Standard, General Requirements of Medical Devices, IEC 60601-1, 2009. Lammers Medical Technology. http://www.lammersmedical.com/Img/produkt03.htm. SonoSite. http://www.sonosite.at/bilder/micromaxx.gif. U.S. Department of Defense. (1999). Human Engineering Design Criteria for Military Systems, Equipment and Facilities. MIL-STD-1472F. Washington, DC: U.S. Department of Defense.
TAF-ER562X_WINGR-09-0401-C017.indd 745
11/9/10 2:00:30 AM
TAF-ER562X_WINGR-09-0401-C017.indd 746
11/9/10 2:00:31 AM
18 Home Health Care Daryle J. Gardner-Bonneau, PhD CONTENTS 18.1 Design Considerations ...........................................................................................750 18.1.1 Flexibility of User Interfaces ...................................................................750 18.1.2 Learnability and Intuitiveness .................................................................750 18.1.3 Adjustability ............................................................................................ 751 18.1.4 Portability and Maneuverability .............................................................. 751 18.1.5 Durability ................................................................................................ 751 18.1.6 Freedom from Calibration, Maintenance, and Repair............................. 751 18.1.7 Protection from Unintended Misuse or Tampering .................................752 18.1.8 Power Requirements ................................................................................752 18.1.9 Aesthetics and Unobtrusiveness ..............................................................752 18.1.10 User Guidance and Training ...................................................................753 18.2 Design Guidelines ..................................................................................................753 18.2.1 Design Guidance Related to Sensory Capabilities and Limitations ........754 18.2.1.1 Vision .......................................................................................754 18.2.1.2 Hearing ....................................................................................755 18.2.1.3 Kinesthetic and Touch Sensitivity ...........................................757 18.2.1.4 Sense of Balance ......................................................................757 18.2.2 Design Guidance Related to Cognitive Capabilities and Limitations .....758 18.2.2.1 Attention ..................................................................................758 18.2.2.2 Information Processing ............................................................759 18.2.2.3 Memory ...................................................................................760 18.2.3 Design Guidance Related to Physical Capabilities and Limitations ....... 761 18.2.4 Design Guidance Related to the Use Environment .................................763 18.2.4.1 Device Maintenance Considerations .......................................763 18.2.4.2 Security Issues .........................................................................764 18.2.4.3 Disposable Devices Components .............................................765 18.2.5 Medical Device Training Materials and Documentation for Home Users .............................................................................................765 18.3 Case Studies ...........................................................................................................767 18.3.1 The Health Buddy® Appliance ................................................................767 18.3.2 Personal Portable Epinephrine Injector ...................................................768 Resources .........................................................................................................................769 References ........................................................................................................................769
747
TAF-ER562X_WINGR-09-0401-C018.indd 747
11/9/10 1:05:54 AM
748
Handbook of Human Factors in Medical Device Design
This chapter addresses the design of medical devices that will be used outside clinical environments (e.g., hospitals and clinics) by laypersons: lay caregivers and patients. These medical devices may be used in the home or other locations (school, mall, and so on). Until only recently, medical devices were intended for use largely by trained personnel in formal clinical environments. This allowed designers to make numerous assumptions about the environment in which the devices would be used and about the skills and the abilities of the expected users. However, medical devices originally designed for use in hospitals and clinics by trained personnel are increasingly being used outside these environments by laypersons. Lay users have different capabilities, limitations, and training than the health care professionals originally expected to use the devices. Furthermore, as hospital stays become shorter and patients are discharged sooner, the complexity of the medical devices used in the home has also increased. Dialysis equipment, infusion pumps, and apnea monitors are examples of the complex devices now used in homes. These circumstances set the stage for device usability problems and use errors that can compromise patient safety and negatively impact the quality of care. Most standards that apply to medical device user-interface design are based on research and data collected from able-bodied clinician users. These data may not be applicable to a lay user population. As an example, let’s say that a design criterion for allowable response time to a system device output needs to be established during the design process. (This example is chosen because response time is a measure known to be affected greatly by increasing age and/or physical disability, and this is often not taken into account during the design process.) The designers empirically measure response time on a sample obtained from a population of relatively young able-bodied users. If a design criterion is then set based on the 95th percentile of those measurements or even the 99th percentile of those measurements, the allowable reaction time may well be insufficient for the majority of a population of users who are over the age of 65 or have physical disabilities, as shown in Figure 18.1. This is why Rogers and her colleagues (Nichols, Rogers, and Fisk, 2003) have argued that investigators need to specify and describe the subject populations that were sampled in their research reports so that designers will not make the mistake of assuming that the study data were obtained from a representative sample of their intended user population when this often may not be the case.
Design criterion
Age = 18 to 45
Faster
Age > 65 and/or populations with disabilities
Reaction time
Slower
FIGURE 18.1 Hypothetical response time distributions for a population of users 18-45 years of age vs. a population of adults over the age of 65 and/or people with physical disabilities. Note that design criterion is sufficient for only a small percentage of the latter population (the percentage in the hash-marked area).
TAF-ER562X_WINGR-09-0401-C018.indd 748
11/9/10 1:05:55 AM
Home Health Care
749
To design medical devices effectively, one must understand both the intended user population and the environment of use. A prototypical user of home health devices might be described as a 75-year-old woman caring for her 72-year old spouse, and characterized as follows: • Doesn’t see or hear very well because of age-related changes in vision and hearing • Isn’t very strong because of age-related decreases in muscle strength • Has infirmities/disabilities of her own, such as arthritis and heart disease (more than 50% of the elderly have at least one disability that interferes with the ability to perform activities of daily living (American Association of Retired Persons [AARP] and Administration on Aging, 1999) • Has no more than a high school education • Has some problems with her memory and needs more time to learn required tasks than would a younger individual • Is under stress created by her husband’s condition and the associated changes this has created in their environment, daily routine, and living arrangements • Has limited support available should difficulties arise during the care of her spouse Another typical user might be described as a 65-year old male diabetic who lives alone, characterized as follows: • Has poor circulation and limited tactile sensation in his fingers as well as a tremor • Has a significant hearing deficit at frequencies above 200 Hz • Has a limited field of view because of increasing macular degeneration Although these users may be typical, they do not represent the full range of users of home health devices. In particular, while many of those receiving and providing home health care are older adults, there are tremendous variations in capabilities and limitations within this age-group, more so than for any other age group. Thus, it is important that designers not overgeneralize when designing for older adults and people with disabilities, the latter also being a highly variable population. All the individual characteristics described above affect users’ ability to interact with medical devices and must be considered in device design as well as in the design of devices’ labeling, documentation, and associated training materials. Failure to do so may lead to undesirable consequences, including device abandonment (despite the potential lifesaving need for a device in some cases), use errors that could cause injury to the patient or the caregiver, and lack of compliance with prescribed procedures, which could lead to device unreliability. Finally, if the demands of operation and maintenance of the medical device (i.e., the equipment burden) are too great, the patient may require transfer to a nursing home or long-term care facility to be cared for effectively. The home care environment must also be considered during device design, as it differs substantially from institutional environments. The environmental characteristics of the home that may differ from health care institutions include the following: • Electrical outlets may not be numerous or optimally placed in many homes, and electrical wiring may not comply with building codes normally enforced in health care institutions.
TAF-ER562X_WINGR-09-0401-C018.indd 749
11/9/10 1:05:55 AM
750
Handbook of Human Factors in Medical Device Design
• Ambient light levels are often lower in the home than in clinical care environments. • The home environment may be quieter, and its occupants may have an expectation of lower levels of noise, or it may be noisier (e.g., in homes with many children). • Stairs and carpeting may present a barrier to the portability and/or maneuverability of equipment. • Children and/or pets in the environment may add novel risks, including the need for additional safeguarding of devices and associated accessories (e.g., childproofing). • The home environment may not provide easy access to assistance or technical support in the case of device failure or malfunction. The challenges presented above, in addition to others such as humidity, temperature, air quality, and cleanliness, must be considered in the design of home care devices. Using any design guidance without verifying its applicability to the specific context of use could result in design solutions that may be unacceptable for home care. Thus, the design data should be specific to the intended user population, and extensive usability testing should be performed with an emphasis on those elements for which adequate design data are not available. It is also important that such testing be carried out in actual or realistically simulated home settings whenever possible. The remainder of this chapter provides guidance to assist developers in designing successful devices for the home health care market.
18.1 DESIGN CONSIDERATIONS The designer of home care devices must be sensitive to the special requirements imposed by both the user population and the use environment. Many of the principles that apply to the design of consumer products also apply to the design of medical devices for home health use.
18.1.1 FLEXIBILITY OF USER INTERFACES Because many home care devices must meet the needs of a widely varying population, devices’ user interfaces should be flexible, providing a range of input and output modes. For example, some users may have poor vision or limited reading skills and may benefit from an interface that provides for speech input and auditory output. Users who have physical limitations may also benefit from speech input or from physical input modes that do not require fine motor abilities to operate or adjust the device.
18.1.2 LEARNABILITY AND INTUITIVENESS Consumers have limited patience and may not be willing to invest a significant amount of time to learn to use a new device. The operation of home health care devices should be easy to learn and intuitive to the extent possible. Lay users will master the device more easily if its operation, look, and feel mirror other products with which they are familiar. For example, some successful glucose monitors aimed at the youth markets have user interface features similar to those of personal digital assistants (PDAs) and cell phones.
TAF-ER562X_WINGR-09-0401-C018.indd 750
11/9/10 1:05:55 AM
Home Health Care
751
18.1.3 ADJUSTABILITY Because of the highly variable capabilities and limitations of users, home care devices should allow adjustability to the maximum extent feasible. For example, visual displays should have a wide range of adjustments for brightness and color. The volume level of auditory alarms should be highly adjustable, and frequency adjustable alarms may also be warranted in some situations. Devices such as workstations should provide a broad range of height and other ergonomic adjustments.
18.1.4 PORTABILITY AND MANEUVERABILITY Home health devices may be used in many locations within or outside the home. Users may expect these devices to be highly maneuverable. Figure 18.2 shows an oxygen therapy unit that can easily be carried over the shoulder. The extent to which devices should be portable or maneuverable should be assessed early in the design process. This topic is discussed more fully in Chapter 17, “Mobile Medical Devices.”
18.1.5 DURABILITY Medical devices used outside health care institutions may be subjected to a wider range of physical stressors. These devices may be dropped, kicked, or banged or have food or beverages spilled on them. Device durability is an important design requirement.
18.1.6 FREEDOM FROM CALIBRATION, MAINTENANCE, AND REPAIR Lay users have little time, experience, or interest in device maintenance or repair. Even routine maintenance such as calibration and cleaning are much less likely to be performed by lay users.
FIGURE 18.2 the shoulder.
This oxygen therapy canister device can be carried easily and unobtrusively over
TAF-ER562X_WINGR-09-0401-C018.indd 751
11/9/10 1:05:55 AM
752
Handbook of Human Factors in Medical Device Design
Medical devices should be designed to require the very minimum of maintenance activities and no calibration or repair by the user.
18.1.7 PROTECTION FROM UNINTENDED MISUSE OR TAMPERING The home environment is less secure than institutional environments, and it is rarely possible to completely control access to home health devices. Designers must devise ways to protect both the user and the device from the consequences of unintended misuse as well as deliberate tampering (e.g., by a child). Locking mechanisms may be employed to restrict access to software, for example, and critical device controls could be “hidden” behind a panel to discourage access by unauthorized users. Design safeguards should reduce the risk of unintentional misuse. For example, if a device has multiple connections and an inadvertent misconnection can cause harm or damage, the connectors should be designed to preclude such a misconnection. Alternatively, devices could allow interchangeable connections. For example, in the case of the Health Buddy® appliance described later in this chapter, (see Section 18.3.1) there are four ports on the back of the unit to which peripheral devices (heart rate monitor, weight scale, pulse oximeter, and so on) can be attached. The unit is designed so that any peripheral can be plugged into any port.
18.1.8 POWER REQUIREMENTS Unlike hospitals, which have emergency sources of power, home environments normally do not. Designers should provide a backup power source (e.g., battery pack), especially for critical devices, along with clear procedural instructions for its use so that patients will not be placed at risk in the event of a power outage. Experience in the rehabilitation domain has shown that devices (e.g., augmentative communication devices) are often abandoned as “broken” when, in fact, their batteries were simply dead and the users either did not know how or were unable to change the batteries. Thus, designers should maximize battery life, provide clear indication of current battery charge, and facilitate easy and reliable recharging or replacement. The device should indicate when batteries need to be changed, along with simple instructions and means for replacing the batteries.
18.1.9 AESTHETICS AND UNOBTRUSIVENESS A medical device, particularly one used extensively, should be aesthetically pleasing to the user. This is especially true for devices used in public or conspicuously within the home. Many people take pride in the look and feel of their home environment. Therefore, medical devices should be as unobtrusive and attractive as possible. For example, wiring and cabling can be boxed away and hidden from view and device packaging designed to be more attractive (Figures 18.3 and 18.4). Medical devices used over an extended period of time (e.g., a wheelchair or a ventilator) can become an extension of the patient and may affect the way the person sees himself or herself. In contrast, complex medical devices can be intimidating and cause others to avoid interaction with the patient if such issues are not considered during design.
TAF-ER562X_WINGR-09-0401-C018.indd 752
11/9/10 1:05:55 AM
Home Health Care
FIGURE 18.3 This CPAP device comes in three sections, including one containing a battery that allows the unit to travel easily. The sections are stackable to form a single unit that can easily sit on a nightstand, taking up only a small amount of space.
753
FIGURE 18.4 This nebulizer is designed to be acceptable and attractive to adolescents.
18.1.10 USER GUIDANCE AND TRAINING Documentation and training materials should be easy to understand by users of varying education levels. Multimedia training materials should be considered for complex devices that will be used by a patient or caregiver. Because learning styles and capabilities differ, the use of audiovisual training materials may be more effective than extensive and complicated user manuals or instruction sheets.
18.2 DESIGN GUIDELINES A majority of patients receiving home health care services are older adults, as are the majority of lay caregivers. However, younger adults with disabilities have many of the same functional limitations that occur as a result of the aging process. Substantial data are available on the capabilities and limitations of older adults relevant to the design of medical devices used in the home. The guidelines in this section are based largely on these data and are organized according to the capabilities and limitations of this population. The subsections provide design guidance related to the following: • Sensory capabilities and limitations • Cognitive capabilities and limitations
TAF-ER562X_WINGR-09-0401-C018.indd 753
11/9/10 1:05:55 AM
754
Handbook of Human Factors in Medical Device Design
• Physical capabilities and limitations • The home use environment • Training materials and documentation for home users
18.2.1 DESIGN GUIDANCE RELATED TO SENSORY CAPABILITIES AND LIMITATIONS Sensory capabilities generally decline with age, although rates of decline differ, and the age at which decline begins differs from one sensory modality to another. 18.2.1.1 Vision A number of changes in vision occur normally as humans age: • Seven out of 10 people wear glasses by their mid-40s. • Eighty percent of adults maintain a corrected visual acuity of 20/40 or better after the age of 65. • Older adults do not adapt as quickly to lighting changes (i.e., slower dark and light adaptation) and have some decrements in depth perception. • Older adults are more sensitive to glare. • The visual field narrows from 180 degrees to 140 degrees by the age of 70 (Johnson, 1986). • The lens of the eye becomes more opaque, changing color perception, particularly for short-wavelength colors (i.e., blue, violet, and, to some extent, green) (Kashima, Trus, Unser, Edwards, and Datiles, 1993). • Light scatters more within the eye, resulting in images appearing blurred and a lessening of perceived contrast (van den Berg, 1995). • Visual processing becomes slower, and this may be related, at least in part, to the sensory changes noted above as well as to changes in cognitive capabilities. Moreover, some tasks may become more cumbersome for older adults who wear bifocals or trifocals. For example, corrective lenses may need to be removed to read small printed text or to view displays at particular distances that their lenses do not accommodate effectively. In addition to these age-related changes in vision, a number of medical conditions affect vision (e.g., diabetes) and may further degrade performance on visual tasks (Table 18.1). Guideline 18.1: Minimize Glare Work surfaces should utilize materials that minimize surface reflectance. Light sources should be shielded or diffusers utilized to minimize direct glare from displays.
Guideline 18.2: Minimize Need for Dark and Light Adaptation Maintain constant task illumination conditions to the extent possible, as older adults require more time to adapt to changes in lighting conditions.
Guideline 18.3: Provide for Display Parameter Adjustments Whenever possible, users should be able to adjust screen luminance and contrast.
TAF-ER562X_WINGR-09-0401-C018.indd 754
11/9/10 1:05:56 AM
Home Health Care
755
TABLE 18.1 Some Common Medical Conditions That Affect Vision Medical Condition
Effects
Cataracts
Poor night vision, halos around lights, day vision eventually affected Poor night vision, bind spots, loss of peripheral vision Blindness Blurred vision, loss of central vision, distorted vision, faded colors Flashes of light across the visual field, floaters Decreased vision locally or globally Dim vision, trouble seeing with one or both eyes Blindness or visual field defects Spots of light, perception of zigzag patterns Blindness in one eye, decreased vision locally Night blindness Vision loss
Glaucoma Diabetic retinopathy Macular degeneration Retinal detachment Optic neuritis Stroke Brain tumor Migraine headaches Multiple sclerosis Vitamin A deficiency HIV/AIDS
Guideline 18.4: Place Important Information Centrally Tasks and displays should be designed so that important information is presented in the center of the visual field, as older adults may not readily perceive information presented in the periphery.
Guideline 18.5: Use Discriminable Colors in Displays If users are required to recognize or discriminate displayed colors, the colors used should be readily perceivable and discriminable. Short-wavelength colors (i.e., blue and violet) should not be used if they must be discriminated. Texture or patterns should be used in conjunction with color (see Hiatt, 1987) to facilitate color discriminations.
Guideline 18.6: Make Text Easily Readable For text intended to be read at distances of 2 feet or less, font sizes of 12 point or larger should be maintained. Users should not need to remove their glasses to read displayed text. Figure 18.5 shows a blood pressure monitor display for which both the displayed values and the labels are easily readable.
Guideline 18.7: Minimize Requirements for Time-Based Responses Task designs should minimize constraints on the time allowed for responses, particularly if older adults are expected to use a device, because they require more time to process and respond to visually displayed information.
18.2.1.2 Hearing Declines in audition begin to occur after the age of 40, and by age 65, more than 50% of males and 30% of females have some age-related hearing loss (Fisk, Rogers, Charness,
TAF-ER562X_WINGR-09-0401-C018.indd 755
11/9/10 1:05:56 AM
756
FIGURE 18.5 and labels.
Handbook of Human Factors in Medical Device Design
Blood pressure monitor display with easy to read displayed measurement values
Czaja, and Sharit, 2004). Most hearing loss occurs in the high-frequency range, above 4,000 Hz. Older adults may have difficulty separating task-related auditory signals, especially speech, from background sounds or noise that should be ignored (e.g., Grose, Poth, and Peters, 1994; van Rooj and Plomp, 1992). Older adults may have more difficulty understanding synthetic speech than natural or digitized speech in auditory displays (Gulya, 1990). Some medical conditions also affect hearing, as shown in Table 18.2. Guideline 18.8: Minimize Equipment Noise Because people in home use environments will be less tolerant of additional noise, device noise (e.g., fans and motors) should be minimized to the greatest extent possible.
Guideline 18.9: Minimize Masking of Task-Relevant Auditory Information Designers should ensure that medical device noise does not mask critical auditory information, such as voice cues and alarms.
TABLE 18.2 Some Medical Conditions That Adversely Affect Hearing Medical Condition Cerebral palsy Lead poisoning Middle ear infections Tay Sachs disease Otosclerosis Méniére’s disease Herpes zoster otitis Fetal alcohol syndrome Epilepsy Carbon monoxide poisoning Brain cancer Acoustic neuroma
TAF-ER562X_WINGR-09-0401-C018.indd 756
11/9/10 1:05:56 AM
Home Health Care
757
Guideline 18.10: Use Perceivable Alarm Frequencies Alarms on medical devices should use fundamental frequencies below 2,500 Hz to increase the likelihood that older users will be able to hear them.
Guideline 18.11: Use Digitized or Recorded Speech When a device employs speech output, digitized or recorded speech (rather than synthetic speech) is preferred.
18.2.1.3 Kinesthetic and Touch Sensitivity Declines in kinesthesis and sensitivity to movement-related feedback from the vestibular organs may occur with aging and affect an individual’s ability to maintain his or her balance. However, there is a high degree of variability in kinesthesis among older adults. Although some older adults are hypersensitive to light touch because of thinner skin, the sense of touch generally decreases beyond the age of 50. Older adults are also less sensitive to pressure (Kenshalo, 1986), vibration (Gescheider, Beiles, Checkosky, Bolanowski, and Verrillo, 1994), and surface roughness (Stevens and Patterson, 1995) than younger adults. Age-related conditions, like Parkinson’s disease and diabetes, can exacerbate these effects (Sathian, Zangaladze, Green, Vitek, and DeLong, 1997). Guideline 18.12: Limit Tactile Coding of Controls Tactile coding of controls should not be used in devices intended for users with a diminished sensation of touch unless the coding can be easily detected (i.e., through major differences among control characteristics).
Guideline 18.13: Do Not Use Tactile Coding Alone As is the case for other types of coding, tactile coding should not be the sole method of conveying information but instead should be redundant with one or more other types of coding (color, size, and so on).
18.2.1.4 Sense of Balance The vestibular system enables people to maintain their balance and posture. Numerous medical conditions, some of which are correlated with age, can affect vestibular function, causing unsteadiness, loss of balance, or vertigo (Table 18.3).
TABLE 18.3
Some Medical Conditions That Adversely Affect Balance
Medical Condition
Effects
Acoustic neurinoma Amyotropic Lateral Sclerosis (ALS) Brain cancer Neurofibrom atosis Parkinson’s disease Stroke Vestibular neuronitis Cerebral palsy
Unsteadiness, vertigo Tripping, falling Lack of coordition, balance problems, stumbling Impaired balance Impaired balance Impaired balance, lack of coordination Vertigo, balance problems Impaired balance
TAF-ER562X_WINGR-09-0401-C018.indd 757
11/9/10 1:05:57 AM
758
Handbook of Human Factors in Medical Device Design
If a device requires users to maintain a fixed position (e.g., standing), users with difficulty balancing may fall. This design concern also applies to devices used by patients in clinical environments (e.g., examination tables and X-ray and mammography equipment). Winters et al. (2007) found in a survey of medical device users that weight scales (whether used at home or in clinical environments), examination tables and chairs, hospital beds, and imaging equipment pose significant accessibility and usability problems to people with disabilities that greatly limits both their access to care and their safety. In a study of congestive heart failure patients receiving telehealth services, it was not uncommon for patients to fail to weigh themselves daily as directed, in part because they could not maintain their balance on the scale without help from someone else. Guideline 18.14: Provide Support to Prevent Falls Medical devices that require use while standing or that require users to maintain a particular posture should be designed with one or more mechanisms (e.g., grab bars and/or armrests) to facilitate positioning and maintenance of balance during use.
18.2.2 DESIGN GUIDANCE RELATED TO COGNITIVE CAPABILITIES AND LIMITATIONS Cognition involves a number of abilities—attention, memory, information processing, and decision making—which may be negatively affected by age, medical conditions, and treatments (e.g., medications). Medical devices should be designed to minimize the cognitive task burden on home device users. 18.2.2.1 Attention Attention can be categorized as one of two types: selective or divided. Selective attention is the ability to focus on a particular stimulus or task, while divided attention is the ability to parcel one’s attentional resources effectively among stimuli or tasks. Most adults, even as they age, can focus on a particular task for significant periods of time without interruption. However, older adults are especially sensitive to visual distractions, such as flashing or blinking lights, and their attention can be inadvertently diverted to such events when they occur during a task. Studies indicate that performance in visual search tasks (finding a target visual element in the midst of nontarget elements) declines with age. In addition, the ability to divide attention among multiple visual tasks decreases with age because older adults require more time to change their attentional focus from one visual location to another. These difficulties increase with task complexity (Kausler, 1991). Guideline 18.15: Minimize Extraneous Visual Stimuli in Displays Visual displays for older adults should minimize the amount of extraneous visual stimuli. Such task-diverting stimuli include prominent blinking of non–task-relevant display elements and the use of animated user-interface elements not related to the task (decorative or entertaining animation). Figure 18.6 shows the simplifications of a home use model of a medical device for compression therapy.
Guideline 18.16: Use Visual or Auditory Cuing Visual and/or auditory cuing should be used to draw users’ attention to important task-related elements of a display.
TAF-ER562X_WINGR-09-0401-C018.indd 758
11/9/10 1:05:57 AM
Home Health Care
759
FIGURE 18.6 Two models of compressions therapy devices. Both have the same functions, but the displayed information and controls are much simpler for the device on the left, which is intended for use in the home.
Guideline 18.17: Minimize Requirements for Divided Attention If possible, medical device operation should not require the user to divide attention among concurrent tasks. If more than one task must be performed concurrently, users should be allowed sufficient time to switch attention among the tasks in order to perform each effectively.
Guideline 18.18: Visually Enhance Task-Relevant Information Important task-related information in medical device visual displays, documentation, and training materials should be enhanced to facilitate perception of and attention to that information. This may be done, for example, by using different colors, boldface type, and white space to distinguish task-relevant from non–task-relevant information. See Chapter 5, “Documentation,” for more information about highlighting information in text.
18.2.2.2 Information Processing Information processing is the ability to mentally manipulate information received through the senses to enable decision making and task performance. Many information processing capabilities change with age. When learning a new task, older adults tend to process information in smaller chunks than do younger adults. Older adults require additional training time to integrate what they have learned. Also, older adults do not make connections among task elements and ideas as readily as younger people unless those relationships are made explicit. Although semantic knowledge is maintained well into old age, older adults have difficulties acquiring new procedural knowledge and achieving the automaticity in task performance that occurs as a result of practice and overlearning of skills. Older adults do maintain proficiency on procedural skills they learned when they were younger but may have more difficulty inhibiting previously learned behaviors when learning a new procedure for the same activity. In addition, under stress, they are more apt to revert to previously learned skills and behaviors.
TAF-ER562X_WINGR-09-0401-C018.indd 759
11/9/10 1:05:57 AM
760
Handbook of Human Factors in Medical Device Design
While performance on simple tasks is comparable in younger and older adults, older adults exhibit cognitive slowing when faced with complex tasks, and this is exacerbated by stress or increases in task complexity. Guideline 18.19: Minimize the Need to Learn New Skills Medical device operation should make use of the intended users’ existing well-learned skills. Users should not be required to learn new skills, especially if they conflict with existing skills or knowledge. This principle is particularly important when modifying a device that is already in common use.
Guideline 18.20: Make Relationships Explicit Older users may have more difficulty inferring connections and relationships among concepts or design elements. Therefore, designers should make such relationships explicit, both in the device and in the associated training materials and documentation.
Guideline 18.21: Focus on Task-Relevant Information Device design, as well as the design of training materials, should facilitate focus of the user’s attention on task-relevant information. Tangential or extraneous information should be avoided or minimized.
Guideline 18.22: Avoid Unnecessary Response Time Requirements Device designs should not impose unnecessary time limits for task completion since information processing and response time varies among users. If response time limits are required, the limits should be verified for appropriateness through usability testing.
Guideline 18.23: Avoid Designs That Require Multitasking To the extent possible, users should need to perform only one task at a time, without having to divide their attention among tasks.
Guideline 18.24: Organize Training Materials into Small Chunks Medical device training programs and documentation designed for use by older adults should limit the pace and quantity of information presented to allow users sufficient time to comprehend the information.
18.2.2.3 Memory Many age-related deficits in cognitive performance are due to declines in short-term (i.e., working) memory. If too many items (usually more than four) are presented concurrently, some of the items are more likely to be forgotten. This limitation has design implications both for stepby-step procedures that must be performed during use of a device and for training materials. Older adults also tend to have deficits in certain types of prospective memory. They have more difficulty remembering to do something at a specific time in the future or after a certain period of time has passed. Instead, older adults perform better when tasks require event-based prospective memory (remembering to do something after a certain event has occurred, such as pressing a button after a green light comes on). In addition, a number of common medical conditions can affect memory, as shown in Table 18.4. Speech prompts and instructions can substantially reduce the memory burden on users and support those users who may have visual impairments or difficulty reading. For users
TAF-ER562X_WINGR-09-0401-C018.indd 760
11/9/10 1:05:57 AM
Home Health Care
761
TABLE 18.4 Some Medical Conditions That Adversely Affect Memory Medical Condition Alcoholism Alzheimer’s disease Thyroid disorders Brain cancer Chronic fatigue syndrome Depression HIV/AIDS Huntington’s disease Lupus Multiple sclerosis
who are deaf or who have hearing impairments, however, redundant visual instructions are needed. The effectiveness of speech instructions, especially if the device will be used in noisy environments, should be tested. Guideline 18.25: Limit Short-Term Memory Requirements The number of items that users must remember to effectively operate a device should be minimized. The device should never require users to maintain more than four items in short-term memory at one time.
Guideline 18.26: Assist Users in Remembering Their Position in Complex Procedures If users must execute lengthy procedures to operate a medical device, a means should be provided to assist them in remembering where they are in the list of steps during task execution.
Guideline 18.27: Display All Task-Relevant Procedural Information Where applicable allow users to access each step in a procedure as they execute it. All the steps in a procedure should be visually displayed simultaneously (on the device or in separate printed material, as feasible and/or practical), thus providing the entire task context to the user and minimizing the burden on short-term memory. If procedural instructions are presented via speech, the user should be provided with the option of having all or part of the instructions repeated. Figure 18.7 shows a blood pressure cuff with instructions that are physically oriented so the patient can view the instructions while putting on the cuff.
Guideline 18.28: Use Event-Based Behavioral Triggers Medical devices should employ event-based behavioral triggers (e.g., “Press Go when the green light comes on”) as opposed to time-based triggers (e.g., “Press Go in 2 minutes”). The former pose a lower memory burden, and user performance will be more reliable.
18.2.3 DESIGN GUIDANCE RELATED TO PHYSICAL CAPABILITIES AND LIMITATIONS Just as there is cognitive slowing as people age, there is a decline in physical performance that encompasses both the premotor, decision-making component of response (Panek,
TAF-ER562X_WINGR-09-0401-C018.indd 761
11/9/10 1:05:57 AM
762
Handbook of Human Factors in Medical Device Design
FIGURE 18.7
Blood pressure cuff with instructions.
Barrett, Sterns, and Alexander, 1978; Vercruyssen, 1997) and the physical act of responding. The response time for physical tasks can be up to twice as long for older adults compared with younger ones (Fisk et al., 2004). Additionally, movements become less precise with age, making tasks that require fine motor skills or significant motor dexterity more difficult (e.g., dragging and dropping display elements with a computer mouse or adjusting a highly sensitive control) (Avolio and Waldman, 1994). Conditions such as arthritis, which occurs in nearly 50% of older adults (AARP and Administration on Aging, 1999), may make movement painful and further decrease motivation to engage in tasks requiring fine motor coordination. Finally, learning complex motor skills that entail precise sequencing of motor actions or carrying out of multiple physical actions simultaneously (e.g., dragging and dropping an on-screen target using a mouse or holding down a button while simultaneously adjusting a rotary control) can be difficult for older adults (Roos, Rice, and Vandervoort, 1997). A number of other medical conditions also affect physical performance, some of which are shown in Table 18.5. Muscle strength, endurance, and tone also decrease significantly with age (e.g., Khalil, Abdel-Motoy, Diaz, Steele-Rosomoff, and Rosomoff, 1994; Spirduso and MacRae, 1990). Flexibility decreases, and the maximum forces that can be generated to perform physical tasks decrease for most muscle groups. Concerns related to motor control, dexterity, and strength apply not only to medical device operation but also to device packaging, assembly, and maintenance (see TABLE 18.5 Some Medical Conditions That Adversely Affect Physical Abilities Medical condition
Effects
Obesity Parkinson’s disease Diabetes Multiple sclerosis Muscular dystrophy Amyotrophic lateral sclerosis Stroke Cerebral palsy Tourette’s syndrome
Physical limits to joint movement; deconditioning Tremor, weakness, muscle rigidity Decreased sensation and proprioception Muscle weakness, paralysis, spasticity, or wasting Muscle atrophy Muscle atrophy Coordination difficulties, muscle atrophy Coordination difficulty, muscle rigidity or jerkiness Muscle tics
TAF-ER562X_WINGR-09-0401-C018.indd 762
11/9/10 1:05:57 AM
Home Health Care
763
Chapter 4, “Anthropometry and Biomechanics,” for additional information relevant to these issues). Guideline 18.29: Avoid Imposing Undue Response Requirements Device operation should not require the user to perform physical actions within set time limits unless all users will be able to complete the action in the time allotted. Response time limits, if necessary, should be established through user testing.
Guideline 18.30: Minimize Precision Motor Tasks Tasks that require fine motor coordination, including typing, should be minimized.
Guideline 18.31: Design for Single-Handed Use When possible, for devices that can be operated with one hand, the design should accommodate use by either hand, or separate devices should be provided for left-handed and righthanded users.
Guideline 18.32: Avoid Complex Motor Tasks Complex motor tasks should be avoided, especially those requiring precise sequencing or the performance of multiple, coordinated physical actions.
Guideline 18.33: Require Only Simple Physical Motions To the extent possible, a device should be physically operable through the use of gross (as opposed to fine) motor skills (e.g., flipping a switch rather than setting a dial).
Guideline 18.34: Limit Endurance and Strength Requirements Medical devices should avoid the need for extended or excessive physical strength (see Chapter 4, “Anthropometry and Biomechanics,” for specific age-related data on endurance and strength).
Guideline 18.35: Minimize Required Range of Motion Because many device users will have limitations in their range of motion, the range of motion required for device operation should be minimized.
Guideline 18.36: Make Packaging Secure but Accessible Packaging of devices and associated accessories should be tested to ensure that intended users can access the device safely and without device damage.
18.2.4 DESIGN GUIDANCE RELATED TO THE USE ENVIRONMENT In addition to addressing user requirements, home health care device designs must consider the conditions and environment of use, including device security, maintenance, and the use of disposable components. 18.2.4.1 Device Maintenance Considerations Device abandonment (i.e., rejection of a device by the user) is a serious problem with home medical and rehabilitative devices (Gitlin, Scheman, Landsberg, and Burgh, 1996; Philips
TAF-ER562X_WINGR-09-0401-C018.indd 763
11/9/10 1:05:57 AM
764
Handbook of Human Factors in Medical Device Design
and Zhao, 1993). Users may choose to not use a device even when it may be necessary to support their lives and well-being. The maintenance burden of a device is a common reason for abandonment. Unlike hospital users, home and public device users do not have ready access to maintenance and repair services. In addition, lay users may not have the necessary knowledge, skill, motivation, physical strength, or coordination to troubleshoot a malfunctioning device or perform complex routine maintenance. Many home use medical and rehabilitative devices are necessarily battery powered. Thus, the maintenance requirements for such devices include monitoring the status of the battery and recharging or replacing it when needed. Guideline 18.37: Minimize Maintenance Requirements Device maintenance requirements should be minimal and easily performed without the need for special tools or materials. Home health care device users are unlikely to perform extensive maintenance, repair, or calibration of medical devices.
Guideline 18.38: Facilitate Routine Device Maintenance Activities If routine device maintenance will be required, usability testing should confirm that the intended users can carry out all expected maintenance activities. Physical supports (e.g., something to hold the device stable while tasks are performed) should be provided, if needed, to complete routine maintenance activities.
Guideline 18.39: Do Not Require Users To Calibrate Equipment Equipment should be self-calibrating or calibrate automatically.
Guideline 18.40: Use Easily Replaceable Parts Devices with replaceable parts, including batteries, should use readily available parts (e.g., AA batteries).
Guideline 18.41: Replacement of Parts should be Simple Batteries and other parts that must be replaced periodically should be easily identified within the device and easy to remove and replace, as shown in Figure 18.8. If a battery or other replaceable part is small and requires fine motor coordination to remove or replace (e.g., hearing aid batteries), tools or affordances should be provided (e.g., a mechanism to hold the device securely in place while the battery is being changed).
Guideline 18.42: Clearly Indicate the Battery Status Information about battery status should be easy to see and understand, facilitating the user’s decisions as to when to recharge or replace the battery.
Guideline 18.43: Use Extended-Life Batteries Battery life should be as long possible, given weight and cost constraints.
18.2.4.2 Security Issues Devices used outside of health care facilities will pose greater security and safety threats than those used institutionally. Security threats include the possibility of access by unauthorized users and an increased likelihood of equipment damage. For example, damage
TAF-ER562X_WINGR-09-0401-C018.indd 764
11/9/10 1:05:57 AM
Home Health Care
765
FIGURE 18.8 Bottom section of this device simply flips open to reveal commonly available batteries that are relatively easy to replace.
could be caused by weather exposure or the family dog chewing on device electrical components. A balance must be struck between design requirements for safety and security and those for accessibility and usability by authorized users. This balance has not always been achieved successfully. For example, while childproof packaging has been highly effective in securing medications against use by children, it has also made them less accessible to some older adults who are unable to open the packaging without assistance. Guideline 18.44: Balance Security Needs against Accessibility and Usability Requirements Device designers should evaluate via usability testing the impact of required security features on the accessibility and usability of home health devices.
18.2.4.3 Disposable Devices Components Many home health devices contain disposable elements (e.g., glucose test strips used with a glucometer). Users may substitute one brand of disposable items for another when the preferred brand is not available or is appreciably more expensive. This can pose a safety issue if these disposable items are not interchangeable. Guideline 18.45: Use Interchangeable Disposable Components To the extent possible, disposable components should be interchangeable from one make/ model of a device to another. Compliance with standards and conventional use (e.g., consumer products) is advisable, if feasible.
18.2.5 MEDICAL DEVICE TRAINING MATERIALS AND DOCUMENTATION FOR HOME USERS Medical device manufacturers face a different set of user expectations when designing user training and documentation for home care devices. Lay users cannot be expected to master
TAF-ER562X_WINGR-09-0401-C018.indd 765
11/9/10 1:05:58 AM
766
Handbook of Human Factors in Medical Device Design
new technology, nor do they have the resources available to perform excessive maintenance and calibration. Older adults in particular may be uncomfortable with technology and have limited knowledge relevant to device use. Furthermore, initial training may occur at the hospital or soon after the patient is discharged, during a period of stress and adjustment. Since support beyond this initial period of may be limited, lay users should have ready access to training and instructional materials during device use. Finally, lay users may not be fluent in English. Therefore, it is critical that device manufacturers pay special attention to the design of training materials and documentation for home care devices (see Chapter 5, “Documentation,” for detailed information on documentation design). Guideline 18.46: Use Simple Language Training materials and documentation should be written at the eighth-grade level or lower. If a device is so complex that this guideline cannot be met, it is probably not suitable for use in the home environment.
Guideline 18.47: User Understandable Language Avoid the use of terminology, jargon, or abbreviations that lay users may not understand. Note that vocabulary comprehension declines with age (see Schaie, 2004).
Guideline 18.48: Use Large Font Sizes Text size should 12 to 16 point, and text-to-background contrast should be high (Adams and Hoffman, 1994; Hogstel, 2001). Even larger fonts may be appropriate for device materials intended for visually impaired users.
Guideline 18.49: Use Age-Appropriate Illustrations Documentation should include illustrations appropriate to the age and abilities of the target user population (Hogstel, 2001; Weinrich and Boyd, 1992).
Guideline 18.50: Use Only Validated Symbols and Icons Many older users may not be familiar with icons and symbols commonly used in software and specifically in medical devices. If symbology is considered for use in device software, labeling, documentation, or training materials, its comprehension by device users should be validated through formal testing.
Guideline 18.51: Provide Documentation in Multiple Languages When appropriate, documentation and training materials, as well as device labeling, should be provided in more than one language to accommodate users whose first language is not English. In the United States, Spanish versions of materials are often needed, and in certain areas, other languages (e.g., Vietnamese or Chinese) may also be required.
Guideline 18.52: Provide Flexible Documentation Consider providing documentation and training materials in several forms to accommodate the strengths and limitations of a diverse user population. While some users may be comfortable with written documentation, others may be served much better through videotaped or Web-based instructions.
TAF-ER562X_WINGR-09-0401-C018.indd 766
11/9/10 1:05:58 AM
Home Health Care
767
Guideline 18.53: Test Documentation for Usability Conduct formal user testing to ensure that the documentation or training materials will be readily understandable by users.
18.3 CASE STUDIES 18.3.1 THE HEALTH BUDDY® APPLIANCE This case study concerns the patient–system user interface of the Health Buddy® appliance used with the Health Hero® System. This system provides remote patient monitoring and patient education services to individuals with chronic diseases being cared for at home. It was designed to assist patients in learning to manage their diseases as well as to measure and transmit health-related data to nurses who are remotely monitoring patients’ conditions and care. The Health Hero® includes a large number of software training modules tailored to various disease conditions (e.g., diabetes, chronic heart failure, and asthma). Each module contains a list of questions concerning the management of a particular disease, a small number of which are presented to the patient at home each day through the appliance. Branching logic, keyed to the patient’s responses during the test session, is applied dynamically to determine which questions will be presented in any given daily session. The appliance also has plug-in ports for peripheral monitoring devices (thermometer, pulse oximeter, heart rate monitor, and so on). These monitors can be connected for automatic measurement with the resulting patient data transmitted to a Web site that can be accessed from a remote nurse monitoring station. The appliance (see Figure 18.9) is a small desktop unit measuring about 5 inches square that blends in with the home environment. The patient user interface is simple, and device operation requires little strength or effort. This is extremely important because users may be very ill and physically weak, often requiring assistance to complete even simple tasks. The patient is expected to use the appliance once each day by completing the disease management questions and transmitting the data to the nurses involved in the patient’s care. Each day, the display screen of the appliance illuminates and stays illuminated until the patient completes his or her session. This design feature makes use of event-related prospective memory as opposed to time-based prospective memory (i.e., the patient is prompted to use the device by virtue of the screen’s salient illumination as opposed to having to remember a particular time to complete the session). The screen goes dark at the completion of the session, relieving the patient of the burden of remembering whether they have completed their session that day.
FIGURE 18.9
Health Buddy appliance.
TAF-ER562X_WINGR-09-0401-C018.indd 767
11/9/10 1:05:58 AM
768
Handbook of Human Factors in Medical Device Design
Only four large buttons are used to input information and answer questions, and the screen display clearly indicates which buttons are appropriate for which responses. All the necessary peripherals plug easily into any of the four ports on the back of the device, so the patient needs to learn only a single action to connect monitoring devices to the appliance to obtain physiological measurements. The simple user interface that minimizes the burden on both caregivers and patients may facilitate better disease management and possibly reduce the occurrence of disease flares, emergency room visits, and rehospitalization.
18.3.2 PERSONAL PORTABLE EPINEPHRINE INJECTOR Users at risk of severe allergic responses to foods, medications, and insect venom, for example, must carry a portable device with them in case the need arises to self-administer a potentially lifesaving epinephrine injection. Such devices have been commercially available for years but have been associated with numerous usability problems. The designers’ challenge was to produce an acceptably sized device, avoiding the known shortcomings of other epinephrine pens, including (1) the potential for users to inject themselves in the hand when meaning to strike the thigh; (2) failing to activate the injection mechanism properly, thereby creating an impediment to timely drug delivery; and (3) the sharps hazard presented postinjection that may cause needlestick injuries. Other market pen injectors expected the user remove a safety tab located opposite from the needle. This was found to be counterintuitive as users believed the safety tab was protecting the needle. A card-shaped device was designed with a retractable needle system possessed advantages over conventional, barrel-shaped epinephrine pens (see Figure 18.10). Users removed a safety tab (located on the needle end of the device) and then push down on the top to activate the injection mechanism. This design encouraged proper orientation (i.e., pointing the needle toward the intended injection site) reducing the chance that users might hold the device opposite to what was intended, thereby sticking themselves. Keeping the device about the size of a credit card—albeit a bit thicker—encouraged users to keep it with them at all times (e.g., in a pants pocket, coat pocket, or handbag). This form factor was chosen because of the success of consumer devices with similar forms such as cell phones, PDAs, and MP3 players.
FIGURE 18.10 Drawings of early prototypes illustrate the simple injection steps. The final design (bottom) includes graphics to help identify and orient the device. (From Intelliject, LLC and Worrell, Inc. With permission.)
TAF-ER562X_WINGR-09-0401-C018.indd 768
11/9/10 1:05:59 AM
Home Health Care
769
Bold labeling helped users and bystanders recognize the device’s medical purpose—an important consideration if the user goes into anaphylaxis (i.e., loses consciousness). A printed arrowhead, a translucent cover, and openings on the device’s “business end” clearly indicated the device’s proper orientation. Important for a mobile medical device, the Intelliject EpiCard™ could withstand being dropped from table height and resist crushing from being sat on.
RESOURCES Burdick, D. C. and Kwon, S. (Eds.). (2004). Gerotechnology: Research and Practice in Technology and Aging. New York: Springer. Charness, N. and Holley, P. (2001). Computer interface issues for health self-care: Cognitive and perceptual constraints. In W. A. Rogers and A. D. Fisk (Eds.), Human Factors Interventions for the Health Care of Older Adults (pp. 239–254). Mahwah, NJ: Lawrence Erlbaum Associates. Charness, N., Parks, D. C., Sabel, B. A. (Eds.). (2001). Communication, Technology and Aging: Opportunities and Challenges for the Future. New York: Springer. Gardner-Bonneau, D. J. (2001). Designing medical devices for older adults. In W. A. Rogers and A. D. Fisk (Eds.), Human Factors Interventions for the Health Care of Older Adults (pp. 221–238). Mahwah, NJ: Lawrence Erlbaum Associates. Gardner-Bonneau, D. and Gosbee, J. (1997). Health care and rehabilitation. In A. D. Fisk and W. A. Rogers (Eds.), Handbook of Human Factors and the Older Adult (pp. 213–255). San Diego, CA: Academic Press. Giordano, J. L. and Deckinger, E. L. (2003, Winter). Guidelines for communicating with our most elderly. Academic Exchange Quarterly, 7(4) 136–141. Available: http://www.higher-ed.org/AEQ Morrell, R. W. and Echt, K. V. (1997). Designing written instructions for older adults: Learning to use computers. In A. D. Fisk and W. A. Rogers (Eds.), Handbook of Human Factors and the Older Adult (pp. 335–361). San Diego, CA: Academic Press. Pew, R. W. and Van Hemel, S. B. (Eds.). (2004). Technology for Adaptive Aging. Washington, DC: National Academies Press, National Research Council, Steering Committee for the Workshop in Technology for Adaptive Aging. Spratt, J. S., Hawley, R. L., and Hoye, R. E. (1997). Home Health Care: Principles and Practices. Delray Beach, FL: GR/St. Lucie Press.
REFERENCES Adams, J. M. and Hoffman, L. (1994). Implications of issues in typographical design for readability and reading satisfaction in an aging population. Experimental Aging Research, 20, 61–69. American Association of Retired Persons and Administration on Aging. (1999). A Profile of Older Americans: 1999. Washington, DC: Author. Avolio, B. J. and Waldman, D. A. (1994). Variations in cognitive, perceptual, and psychomotor abilities across the working life span: Examining the effects of race, sex, experience, education, and occupation type. Psychology and Aging, 9, 430–442. Fisk, A. D., Rogers, W. A., Charness, N., Czaja, S. J., and Sharit, J. (2004). Designing for Older Adults: Principles and Creative Human Factors Approaches. Boca Raton, FL: CRC Press. Gescheider, G. A., Beiles, E. J., Checkosky, C. M., Bolanowski, S. J., and Verrillo, R. T. (1994). The effects of aging on information-processing channels in the sense of touch: II. Temporal summation in the P channel. Somatosensory and Motor Research, 11, 359–365. Gitlin, L. N., Scheman, R. L., Landsberg, L., and Burgh, D. (1996). Factors predicting assistive device use in the home by older people following rehabilitation. Journal of Aging and Health, 8(4), 554–575.
TAF-ER562X_WINGR-09-0401-C018.indd 769
11/9/10 1:05:59 AM
770
Handbook of Human Factors in Medical Device Design
Grose, J., Poth, E., and Peters, R. (1994). Masking level differences for tones and speech in elderly listeners with relatively normal audiograms. Speech and Hearing Research, 37, 422–428. Gulya, A. J. (1990). Ear disorders. In W. Abrams and R. Berkow (Eds.), The Merck Manual of Geriatrics. Rahway, NJ: Merck & Co. Hiatt, L. G. (1987). Designing for the vision and hearing impairments of the elderly. In V. Regnier and J. Pynoos (Eds.), Housing and the Aged: Design Directives and Policy Considerations (pp. 341–371). New York: Elsevier. Hogstel, M. O. (Ed.). (2001). Gerontology: Nursing Care of the Older Adult. Albany, NY: Delmar. Johnson, C. (1986, February). Peripheral Visual Fields and Driving in an Aging Population. Paper presented at the Invitation Conference on Work, Aging and Vision, National Research Council, Washington, DC. Kashima, K., Trus, B., Unser, M., Edwards, P., and Datiles, M. (1993). Aging studies on normal lens using the Scheimpflug slit-lamp camera. Investigative Ophthalmology and Visual Science, 334, 293–326. Kausler, D. (1991). Psychology and Human Aging (2nd ed.). New York: Springer-Verlag. Kenshalo, D. (1986). Somesthetic sensitivity in young and elderly humans. Journal of Gerontology, 41, 732–742. Khalil, T. M., Abdel-Moty, E., Diaz, E. L., Steele-Rosomoff, R., and Rosomoff, H. L. (1994). Efficacy of physical restoration in the elderly. Experimental Aging Research, 20, 189–199. Nichols, T. A., Rogers, W. A., and Fisk, A. D. (2003). Do you know how old your participants are? Recognizing the importance of participant age classifications. Ergonomics in Design, 11(3), 22. Panek, P. E., Barrett, G. V., Sterns, H. L., and Alexander, R. A. (1978). Age differences in perceptual style, selective attention, and perceptual-motor reaction time. Experimental Aging Research, 4, 377–387. Philips, B. and Zhao, H. (1993). Predictors of assistive technology abandonment. Assistive Technology, 5, 36–45. Roos, M., Rice, C., and Vandervoort, A. (1997). Age-related changes in motor unit function. Muscle and Nerve, 20, 679–690. Sathian, K., Zangaladze, A., Green, J., Vitek, J., and DeLong, M. R. (1997). Tactile spatial acuity and roughness discrimination: Impairments due to aging and Parkinson’s disease. Neurology, 49, 168–177. Schaie, K. W. (2004). Developmental Influences on Adult Intelligence: The Seattle Longitudinal Study. New York: Oxford University Press. Spirduso, W. W. and MacRae, P. G. (1990). Motor performance and aging. In J. E. Birren and K. W. Schaie (Eds.), Handbook of the Psychology of Aging (3rd ed., pp. 183–200). San Diego, CA: Academic Press. Stevens, J. and Patterson, M. (1995). Dimensions of spatial acuity in the touch sense: Changes over the life span. Somatosensory and Motor Research, 12, 29–47. van den Berg, T. (1995). Analysis of intraocular stray light, especially in relation to age. Optometry and Visual Science, 72, 52–59. Van Rooj, J., & Plomp, R. (1982). Auditive and cognitive factors in speech perception by elderly users: III. Additional data and final discussion. Journal of the Acoustical Society of America, 91, 1028–1033. Vercruyssen, M. (1997). Movement control & speed of behavior. In: A. D. Fisk and W. A. Rogers (Eds.), Handbook of Human Factors and the Older Adult (pp. 55–86). San Diego, CA: Academic Press. Weinrich, S. P., and Boyd, M. (1992). Education in the Elderly. Journal of Gerontological Nursing 18(1), 15–20. Winters, J. M., Story, M. F., Barnekow, K., Kailes, J., Premo, B., Schwier, E., et al. (2007). Results of a national survey on accessibility of medical instrumentation for consumers. In J. M. Winters and M. F. Story (Eds.), Medical instrumentation: Accessibility and Usability Considerations (pp. 13–27). Boca Raton, FL: CRC Press.
TAF-ER562X_WINGR-09-0401-C018.indd 770
11/9/10 1:05:59 AM
and 19 Cross-National Cross-Cultural Design of Medical Devices Uvo Hoelscher, PhD; Long Liu, PhD; Torsten Gruchmann, Dipl-ing; Carl Pantiskas, MS CONTENTS 19.1 Design Considerations ...........................................................................................773 19.1.1 Differences among Nations and Cultures .................................................773 19.1.2 Language ...................................................................................................774 19.1.2.1 Regulatory Issues.......................................................................774 19.1.2.2 Measurement Units and Formats ...............................................774 19.2 Culture- or Nation-Specific Context of Use ...........................................................776 19.2.1 Technical Environment .............................................................................776 19.2.2 Use Environment .......................................................................................776 19.2.3 Professional Traditions and Work Organization .......................................777 19.2.4 Social Context ...........................................................................................777 19.3 Culture- or Nation-Specific User Profiles ..............................................................778 19.3.1 Demographics ...........................................................................................778 19.3.2 Anthropometric Characteristics (See also Chapter 4, “Anthropometry and Biomechanics”) ..................................................................................778 19.3.3 Preferences and Expectations ....................................................................778 19.3.4 Attention....................................................................................................779 19.3.5 Knowledge, Experience, and Educational Background ............................779 19.3.6 Interpretation of Colors and Symbols .......................................................780 19.3.7 Learning Style ...........................................................................................782 19.4 Medical Device Design for Multiple Nations or Cultures......................................782 19.4.1 Direct User-Interface Issues ......................................................................782 19.4.1.1 Hardware Design Issues ............................................................782 19.4.1.2 Interface Structure .....................................................................783 19.4.1.3 Operational Sequence ................................................................783 19.4.2 Information Presentation ...........................................................................783 19.4.2.1 Language Issues.........................................................................783 19.4.2.2 User-Interface Orientation .........................................................784 19.4.2.3 Units of Measure and Format ....................................................784 771
TAF-ER562X_WINGR-09-0401-C019.indd 771
11/9/10 1:10:55 AM
772
Handbook of Human Factors in Medical Device Design
19.4.2.4 Symbols .....................................................................................784 19.4.2.5 Color Coding .............................................................................784 19.4.3 Indirect User-Interface Design Issues .......................................................786 19.4.3.1 Functionality ..............................................................................786 19.4.3.2 Technical Features .....................................................................787 19.4.3.3 User Support ..............................................................................787 19.4.3.4 Technical Documentation ..........................................................788 19.5 Design Resources for Cross-Cultural and Cross-National Design ........................788 19.5.1 Analysis Phase Considerations ..................................................................789 19.5.2 Design Phase Considerations ....................................................................789 19.5.3 Evaluation Phase Considerations...............................................................791 19.6 Conclusions ............................................................................................................791 19.7 Case Study .............................................................................................................792 Resources .........................................................................................................................792 References ........................................................................................................................793 Cross-cultural or cross-national design of medical devices refers to the design of medical devices for international markets. These designs consider culturally specific differences that may influence the safety of medical devices in other countries. The terms “culture” and “nation” both describe the nature of the international market. Although culture is usually associated with national boundaries, a nation may contain more than one culture (e.g., the province of Quebec in Canada, where French is spoken instead of English), while different nations may exhibit similar cultures (e.g., Austria and Germany). This chapter identifies design issues that are influenced by culture- or nation-specific factors and provides information to assist in the design of medical devices for use in international markets. A process that considers the unique requirements of cross-national and cross-cultural design in the analysis, design, and evaluation phases is proposed and discussed. When this process results in specific solutions for a market segment, this is called localization. If it results in a solution suitable for all targeted market segments, this is called globalization. Globalized technology (e.g., computers) may have fewer cultural differences than do devices intended for use by the population as a whole. Ongoing economic globalization fosters the exchange of technical products worldwide. Many medical devices are now designed for a global market rather than particular national markets. Increasingly, medical devices are designed to simultaneously meet the needs of Western European, Japanese, and North American markets. A truly globalized device would also meet the needs of users in 90% of the world’s population who do not live in those three markets. As user groups become more varied, the design of medical devices needs to address global characteristics. Currently, when launched in international markets, many medical devices remain largely unchanged with respect to meeting local requirements. Yet often device features should be designed differently (e.g., the keyboard layout) for individual countries. Manufacturers commonly assume that users will adapt to the medical device through training. But training can be an ineffective way to overcome ingrained customs and expectations in a user population. At least during the transition phase, the users’ ability to operate the device may be impaired, increasing the probability of use error. Failure to meet cross-cultural and cross-national design requirements may also violate the international risk management (ISO 14971:2007) and usability (IEC 62366:2007) standards to which manufacturers
TAF-ER562X_WINGR-09-0401-C019.indd 772
11/9/10 1:10:56 AM
Cross-National and Cross-Cultural Design of Medical Devices
773
should strive to comply. Furthermore, a culture-specific design is also an important marketing issue. Users who recognize that a design does not meet their specific requirements may take their business elsewhere.
19.1 DESIGN CONSIDERATIONS A cross-cultural or cross-national design should consider both the specific factors (inputs) and design attributes (outputs). These interact with each other and change over time (Table 19.1).
19.1.1 DIFFERENCES AMONG NATIONS AND CULTURES Each individual nation has its own history, culture, level of economic development, political system, physical environment, and so on. People with different cultural backgrounds have unique values, habits, beliefs and thinking patterns. Specific cultural and national differences may include the following: • National issues—Language, regulatory requirements, unit system, and format • Context of use—Technological environment, use environment, professional traditions, and work organization • User profiles—Visual ability, cognitive styles, learning styles, values and rules, and educational background preferences The following sections briefly elaborate the design implications of some of these differences and present relevant examples for medical device designs. TABLE 19.1 A Framework of the Factors in Cross-Cultural and Cross-National Design Cultural and National Factors
Design Attributes
National issues Language Regulatory issues Units and formats Health care delivery system Context of use Technical environment Use environment Social context Professional traditions User profiles Anthropometric features System of values Preferences and expectations Knowledge and experience Interpretation of information Learning styles
Direct user-interface issues Hardware design Interface structure Operation sequence Information presentation (warnings, symbols, layout, and so on) Indirect user-interface issues Functionality Technological features User support Training needs Technical documentation
TAF-ER562X_WINGR-09-0401-C019.indd 773
Influence
Adaptation
11/9/10 1:10:56 AM
774
Handbook of Human Factors in Medical Device Design
19.1.2 LANGUAGE Language is a key challenge when designing medical devices for international markets. Since people communicate with each other in their native language, they may also expect to interact with medical devices in their native language. The hundreds of widely spoken and written languages present obstacles to effective communication, particularly with respect to medical devices. The marketplace for medical devices is far smaller than the marketplace for many consumer products. This makes it uneconomical to provide user interfaces and technical documentation in the native language of every possible user. Thus, manufacturers may choose to provide a subset of preferred languages as a compromise. Some countries require the device software and/or its accompanying documents (e.g., operating instructions, manuals) to be in their native language(s). Other nations, such as India, allow the use of English in medical devices because English is a significant secondary language and is a common language of communication among medical professionals in that country. Although many non-English-speaking user groups read and understand English well, during a critical or stressful situation, device users may be challenged to interact with the device in a secondary (non-native) language. When required to use a non-native language, use errors may be more common. Common language issues for medical device design are listed in Table 19.2. Most of these issues have been intensively studied in the computer industry, and many solutions have been developed. 19.1.2.1 Regulatory Issues National regulatory requirements affect how medical devices are required to operate, to be labeled, and to be marketed and sold. While there is an ongoing international effort to harmonize medical device regulatory requirements through the Global Harmonization Task Force, there remain substantial international differences. Such differences can create significant burdens for manufacturers trying to design and market devices in multiple countries. Unless manufacturers are careful, they may find that one nation’s requirement may make caregivers in another nation unhappy with how their medical device operates. For example, at one time medical devices used in France had to prevent users from disabling an alarm once that alarm was activated. This requirement influenced the design of medical devices for France as well as in countries where French was spoken. More insidiously, compliance with a single country’s requirements can result in a design that is more likely to induce use error in another country with different user expectations or use-shaping factors. 19.1.2.2 Measurement Units and Formats Countries use different systems of measurement. The use of the imperial units system in the United Kingdom and the United States and the International System of Units (SI) in many other countries of the world is well known. Some countries also use traditional unit systems in some product areas (e.g., Gauge, French, and Charrière for the diameter of needles or catheters). Support for configurable measurement units can create human factors issues. Recently, several patient injuries and deaths were associated with glucometers that had erroneously changed to different measurement units (e.g., deciliters vs. millimoles/liter). To mitigate this issue, some manufacturers have modified their device designs to support only a single measurement unit. This can create different use problems if not customized to the local user population.
TAF-ER562X_WINGR-09-0401-C019.indd 774
11/9/10 1:10:56 AM
Cross-National and Cross-Cultural Design of Medical Devices
775
TABLE 19.2 Differences in Language Issues among Countries Language Issues
Examples
Script
ABCD (Latin), ABXΔ (Greek), ä, ü, ö (German) 日本語 (Kanji), ぁぃぅぇぉ (Hiragana),ァィゥェォ(Katakana) 汉字 (simplified Chinese), 漢字 (traditional Chinese)
Length of words
Speed (English), Geschwindigkeit (German), 速度 (Chinese)
Grammar
Subject is followed by object and verb (Japanese and French), subject is followed by verb and object (English), or varying (German).
Spelling differences
“colour” (British English) vs. “color” (American English)
Reading/writing direction
English is written horizontally from left to right, while Hebrew is written horizontally from right to left. Arabic is written horizontally from either right to left or left to right. Chinese can be written vertically from right to left.
Alphabetical order
Å follows Z in Finnish but follows A in French
Context-dependent meanings of words
东西 (direction of east and west) or 东西 (things) Break: to break up a relationship; pause/rest in activity; stop
Different meanings for words with different character cases (uppercase/lowercase letters)
In German, “paar” (a few) and “Paar” (a pair)
Homonyms
Words with the same spelling or with different spellings but the same pronunciation have very different meanings. For example, in American English, the words “there” and “their” and “they’re” are pronounced the same way but have very different meanings.
Heteronym
Words with the same spelling but different pronunciations and different meanings. Consider, for example, the English word “tear.” When used as a verb and pronounced as [te (r)], it means “to rip.” However, when used as a noun and pronounced [ti (r)], it means a drop of fluid from the eye.
Idioms
Phrases that have different meanings if translated literally than that intended by a native speaker of that language. For example, idiomatically, to Americans “heads up” means beware of something above you.
Symbolic representations based on gestures
Symbolic representations based on gestures may mean different things in different cultures. In some cultures, holding a hand up palm outward means “stop,” whereas other cultures consider that to be a personal insult. So, a symbolic representation of a palm-out hand has very different meanings to these cultures.
e e
Other differences include varying formats for date, time, number, address, and so on, all of which can complicate the use of medical devices (Table 19.3). For example, Europe and the United States use different symbols to separate the whole part of a number from the decimal part of that number (e.g., 4,5 vs. 4.5 in Europe and the United States, respectively) and to separate segments of large numbers (e.g., 4.500.000 vs. 4,500,000). The date format of 05/09/02 is also unclear without its cultural context. In the United States, the month is typically placed first and the day second, while in Europe and other countries, the opposite is true.
TAF-ER562X_WINGR-09-0401-C019.indd 775
11/9/10 1:10:56 AM
776
Handbook of Human Factors in Medical Device Design
TABLE 19.3 Different Units and Formats among Countries Units or Formats
Examples
Units
SI system vs. English system In medical areas, e.g., various units for infusion rate: ml/h, μg/kg/min, μg/kg/h, μg/h, mg/h, mg/dl, ml/hr, ml/hour SI system vs. traditional system In blood pressure monitoring, “hPa”, mmHg, and “Torr” are all used.
Numbers
Use of comma and decimal point: 10.456,00 (Germany) vs. 10,456.00 (United States)
Time
Modal time presentations (e.g., 24 hour or 12 hour with AM or PM)
Date
23.09.1998, 09/23/1998, 23/09/1998, 1998.09.23, and so on
Paper size
DIN A4 (Germany) vs. letter size 8.5 × 11 inch (United States)
Format of post address
Name followed by address (from small to large) (Germany) vs. address (from large to small) followed by name (China)
Keyboard layout
QWERTZ (Germany), QWERTY (United States), AZERTY (France)
Age at birth
0 (United States), 1 (Korea and Bangladesh)
Some numbers may have negative connotations in certain cultures. For example, the number 13 is bad luck in the United States and Germany, while the number 4 connotes death in Japan.
19.2 CULTURE- OR NATION-SPECIFIC CONTEXT OF USE 19.2.1 TECHNICAL ENVIRONMENT The technical environment of target nations/cultures can differ in four ways: 1. The general level of usage and comfort with technology in the country 2. The technological traditions of the locale (how technology has been used) 3. The technical environment where the device is used 4. The country’s technology acceptance model The general level of acceptance of technical devices in a target market reflects that market’s ability to support and accept new or different technologies. For example, Korea tends to adopt new technologies more quickly than Ireland. The technological traditions of the target market include specific technical standards for mandatory design requirements (e.g., differing “safe” levels for radio-frequency exposure). The technological environment of use includes the presence of other devices and technologies that can influence the use of the medical device. Differences in the characteristics of the area’s power system, as well as its quality (e.g., stability of power supply), should also be considered. A device’s working life may vary by country. For example, some countries replace anesthesia machines routinely on a regular schedule, while others continue using them for 20 or more years.
19.2.2 USE ENVIRONMENT The global medical device market provides a wide variety of use environments. Macroenvironmental characteristics include climate, altitude, air conditions, transportation
TAF-ER562X_WINGR-09-0401-C019.indd 776
11/9/10 1:10:56 AM
Cross-National and Cross-Cultural Design of Medical Devices
777
and storage conditions, and so on. Microenvironmental characteristics include the cleanliness, illumination, space constraints, etc. of the location where the medical device will be used. For more information on this topic, see Chapter 3, “Environment of Use.” Unsatisfactory environmental conditions due to air and water pollution are of special concern in developing countries. Use of medical devices in areas with high humidity or air pollution may cause oxidation and cracking of electrical components and necessitate special maintenance steps. Manufacturers should also consider the local transportation conditions, especially for medical devices that may be used in transit (e.g., in an ambulance). Unexpected levels of vibration, such as in ambulances on dirt roads, can damage a medical device. For more information on this topic, see Chapter 17, “Mobile Medical Devices.” Designers should also consider differences in clothing worn by both caregivers and patients due to climate conditions and cultural conventions.
19.2.3 PROFESSIONAL TRADITIONS AND WORK ORGANIZATION The work organization (i.e., division of labor, responsibility, hierarchy, and so on) of health care differs from country to country. This may result in the same device being used by individuals with quite different levels of training and experience in different parts of the world. For example, in operating rooms throughout the world, patient monitors and anesthesia delivery systems are operated primarily by physicians (either anesthesiologists [United States]) or anaesthetists (United Kingdom) or by nurse anesthetists (United States, Africa). The same monitors are used outside the operating rooms by an even broader range of clinician users ranging from physicians to nurses and technicians. Intensive care ventilators are operated mainly by physicians in Europe and by nurses in Australia, but by respiratory therapists in the United States. There are substantial international differences in clinical workflow and workload. For example, the use of induction rooms to administer anesthesia is common in Germany but not in the United States. Another example is the practice of initiating more advanced treatments in ambulances in France. This practice is uncommon in the rest of the world. Work hours of physicians in the United States are typically substantially higher than those in Europe. Conventional medical care (i.e., widely accepted as appropriate) varies significantly among countries and should be considered when designing for international markets. For example, Chinese traditional medicine largely emphasizes restoration of body harmony through natural remedies instead of external intervention. Few medical devices are used in this diagnosis and treatment process. Medical devices may end up being used in unexpected ways in developing countries with local or traditional medical traditions.
19.2.4 SOCIAL CONTEXT Social context refers to how the people view a situation or object (i.e., interpret its meaning). Social context is a function of culture, expectations, background, upbringing, and experience. These factors will influence how medical devices are accepted and used by both clinicians and patients. Moreover, cultural factors affect how groups of people communicate and interact. Since medical devices are generally part of an interactive system involving the patient, care provider, and the care (or use) environment, the social context can be a critical design consideration.
TAF-ER562X_WINGR-09-0401-C019.indd 777
11/9/10 1:10:56 AM
778
Handbook of Human Factors in Medical Device Design
19.3 CULTURE- OR NATION-SPECIFIC USER PROFILES 19.3.1 DEMOGRAPHICS User population characteristics such as age, gender, education, and ethnic background may change over time with fluctuations in birth and mortality rates, life span, immigration, and socioeconomic developments. Since medical devices may be used for extended periods, designers should anticipate potential changes in user populations. For example, traditionally, the majority of surgeons in the United States were men, but more than 50% of medical students entering some surgical fields in recent years are women.
19.3.2 ANTHROPOMETRIC CHARACTERISTICS (SEE ALSO CHAPTER 4, “ANTHROPOMETRY AND BIOMECHANICS”) Body size, weight, and physical characteristics (e.g., reach, flexibility, and strength) differ among user populations. On average, Asian people are smaller than Westerners. Thus, anesthesia workstations designed for people who work in U.S. hospitals may be too large for Japanese anesthesia providers. The design of the workplace, control elements, and other interface components should reflect the anthropometric characteristics of target users (both caregivers and patients). This is particularly true of devices intended for use by patients in a multicultural population.
19.3.3 PREFERENCES AND EXPECTATIONS A medical device’s value derives from meeting the users’ expectations and preferences. These will vary according to users’ cultures. For example, in some cultures, integration of many functions in a small device is appreciated as high-tech and allows manufacturers to sell those devices at a premium when compared to bulkier, less integrated devices. In contrast, in China, size and weight tend to correlate with importance. Thus, an innovative compact in vitro analyzer was not successful in the Chinese market until the product was enlarged and made heavier by adding a metal plate inside. Other differences in preferences and expectations include the following: • In some cultures, users prefer to focus on a single task when they interact with a device. In others, users prefer to focus on several tasks at once. • Significant cultural differences exist in the preference for automated systems (FAA Human Factors Team, 1996). Automated systems reflect the culture in which they have been designed, but the method in which they are used reflects the culture of the user. Thus, for example, authority gradients (i.e., how rigid is the management hierarchy) may influence whether a user reprograms or disengages the automation when there is an unexpected situation (FAA Human Factors Team, 1996). • Cultures interpret the importance of time differently. For example, the Chinese view appointments more flexibly than Westerners. • People in East Asia tend to group objects and events based on the relationships between them, whereas Westerners are more likely to rely on categories. For example, the Chinese prefer to group on the basis of thematic relationships (e.g., monkey and banana), while Americans show a marked preference for grouping on the basis of membership in a common category (e.g., panda and monkey fit into the animal category) (Nisbett, 2003).
TAF-ER562X_WINGR-09-0401-C019.indd 778
11/9/10 1:10:56 AM
Cross-National and Cross-Cultural Design of Medical Devices
779
Other culturally specific conventions can significantly affect device use and the potential for use error. For example, in central Europe, the rotational directions for mechanical and electronic devices are typically different. Turning a knob clockwise reduces a setting by closing a pneumatic or hydraulic valve but increases a setting on an electronic device (e.g., the volume control). In contrast, in the United States, the rotational directions of hydraulic valves (e.g., water valves) may be different. For example, turning a knob clockwise might close a hot-water valve but open a cold-water valve. Such learned preferences and expectations are known as cultural stereotypes. Behavior associated with cultural stereotypes is called stereotypical behavior. Especially under stress (e.g., time pressure or a high-risk situation), a user may revert to stereotypical behaviour during device use despite prior training and successful usage (Maddox, 1998).
19.3.4 ATTENTION Different cultural standards also influence users’ attention. Visual objects for information presentation should be designed considering the following examples: • Users from different cultures may expect to find important information at a specific screen location. An American user who conventionally reads horizontally from left to right may expect important information to be in the upper-left quadrant of a screen. A Chinese user who reads vertically from right to left may expect important information to be in the upper-right quadrant (Aykin, 2005). • Westerners tend to have a more analytic view that focuses on salient objects and their attributes, whereas Easterners tend to have a more holistic approach that considers continuities in substances and relationships in the environment. In an experiment to identify observations of objects in a scene, it was found that Japanese participants made more references to background and environment than did American participants. Americans were more likely to pick up changes in focal, foreground objects (Nisbett, 2003).
19.3.5 KNOWLEDGE, EXPERIENCE, AND EDUCATIONAL BACKGROUND The level of training and prior experience of health care providers will affect their ability to use specific medical devices or device features. Health care practitioners may be trained to operate medical devices differently depending on the nation or culture in which they practice. For example, in China, infusion pumps are high-tech medical devices that are rarely used in health care practice. Chinese nurses who have not been trained to use infusion pumps may be somewhat afraid to do so. The use of anesthesia machines in Europe is another example. In the 1980s, many French anesthetists operated the breathing circuits of their anesthesia machines only in the semiopen mode, while German and Austrian anesthetists operated their breathing circuits mainly in the semiclosed mode. This difference significantly influenced the design and use of medical devices in these markets. The general population’s education level and its familiarity with technical products (e.g., computers) will influence the use of home medical devices.
TAF-ER562X_WINGR-09-0401-C019.indd 779
11/9/10 1:10:56 AM
780
Handbook of Human Factors in Medical Device Design
19.3.6 INTERPRETATION OF COLORS AND SYMBOLS The physiological features used to perceive visual stimuli are the same for everyone. However, people often interpret the same stimulus differently. Research has shown that certain language-independent methods of information transmission, such as colors and symbols, are culture specific. For example, Table 19.4 provides a partial list of meanings TABLE 19.4 A Nonexhaustive List of Color Symbolism by Culture Color
Location
Cultural Significance
Red
China
Symbolizes celebration and luck, used in cultural ceremonies ranging from funerals to weddings Purity (used in wedding outfits) Stop (danger) at traffic lights, high priority, high alert; Christmas color when combined with green, Love (Valentine’s Day) when combined with pink With white, signifies joy Sacred, imperial Joy or happiness, intermediate priority, alert (caution), continue but caution, stop if possible (traffic lights) Associated with immortality Associated with soap The color of Krishna Holiness Protective color Religious significance (Protestant) Inexpensive goods, Halloween (with black) Studies suggest that this is not a good color choice for packaging; green hats mean that a man’s wife is cheating on him Studies suggest that this is not a good color choice for packaging The color of Islam Religious significance (Catholic) Associated with danger
India United States
Yellow
Eastern cultures Asia Western cultures
Blue
China Colombia India (Hindus) Israel (Jewish) Middle East Orange Ireland United States Green China France India Ireland Some tropical countries United States
Go (safe) at traffic lights, environmental awareness, St. Patrick’s Day, Christmas (red and green) Purple Western cultures Royalty, bishop Gray Western cultures Decent or boring appearance Brown Colombia Discourages sales White Eastern cultures Mourning, death; in Japan, white carnations signify death United States Purity (used in weddings) Black Western cultures Mourning, death, nobility, exclusivity, elegance Saffron India (Hindus) Sacred color Pastels Korea Trust United States Spring, Easter; pale blue (baby blue) stands for an infant boy (pale pink stands for an infant girl) Note: Blue is often considered to be the safest global color.
TAF-ER562X_WINGR-09-0401-C019.indd 780
11/9/10 1:10:56 AM
Cross-National and Cross-Cultural Design of Medical Devices
(a)
(b)
781
(c)
FIGURE 19.1 These symbols or signs are not universally understood across multiple cultures and might cause use errors if incorporated into a medical device marketed in those user populations.
people intuitively associate with specific colors as part of their daily lives. In China and Japan, the color red is associated with prosperity and happiness. Using flashing red to indicate an alarm condition may conflict with traditional associations for the color red in these countries. In contrast, in Europe and the United States, people associate the color red with danger or stop and the color green with start (primarily because of how those colors are used on traffic lights). Research on the significance of different interpretations of color is quite limited. Device designers should be aware that certain colors could be interpreted differently depending on the nation or culture in which a device is used. In the absence of specific prior knowledge or experience, testing of color choices with the intended user population is recommended. Many symbols were originally designed using a specific cultural context, so cultures may interpret the same symbol differently. For example, the symbol for “directory” used in the Microsoft Windows system is an analogue to a file system in North America but might be unknown for users in other regions, such as in Asia (Figure 19.1a). The sign of a cross before an option (Figure 19.1b) indicates that option is chosen in Germany but means disapproval (or not chosen) in China. The checked box symbol (Figure 19.1c) is unambiguous for both. In medical areas, this difference in interpretation may pose a significant risk to patient safety. The “skull and crossbones” symbol, used to indicate substances toxic to humans, may not be understood in rural areas in some parts of the world (Figure 19.2). Failure to understand this symbol resulted in the deaths of many Iraqi peasants who ate seed preserved with a poisonous mercury compound. Although the seed was packaged in sacks clearly labeled with the skull and crossbones, the symbol seemed to them “nothing more than a peculiar piece of art work” (Casey, 1998).
FIGURE 19.2 peasants.
The skull-and-crossbones symbol for “danger” had no inherent meaning for Iraqi
TAF-ER562X_WINGR-09-0401-C019.indd 781
11/9/10 1:10:56 AM
782
Handbook of Human Factors in Medical Device Design
TABLE 19.5 Learning Styles of Chinese vs. German Users Objective of learning Traditions of learning Information gathering strategy Learning material
Chinese
German
Pragmatism: master only the features that are directly useful Rote learning: learning by imitation
Idealism: long-term mastery of relevant features and principles Understanding: learning by exploring Individualism: gather information from formal information sources Script and textual orientation
Collectivism: gather information from a network of relationships Pictorial and animation orientation
The general difficulties associated with designing reliable graphical symbols (or icons) for medical devices (Forcier and Weinger, 1993) are significantly amplified in a device intended for multiple medical markets. Thus, the manufacturer must carefully assess the design trade-offs between text labeling (in native language vs. nonnative language) and symbol labeling with regard to safety, usability, cost and feasibility.
19.3.7 LEARNING STYLE Learning style is unique to each individual and includes his or her ability to assimilate information visually, auditorily, textually, and kinesthetically (movements). An individual’s learning style also includes how he or she relates to educational methods and techniques, learning materials, and the role of the teacher. Personal learning style is heavily influenced by cultural background. The learning styles of the target users should be considered when designing any interactive system, especially technical documentation and information systems with a user interface. For example, Chinese users appear to prefer to learn device operations by watching others. Thus, learning material that utilizes graphical presentations, animations, or video will better support their learning processes. In contrast, German users tend to learn through exploration and from user manuals that contain detailed descriptions of operations and technical principles (Honold, 1999) (Table 19.5).
19.4 MEDICAL DEVICE DESIGN FOR MULTIPLE NATIONS OR CULTURES Device designers should consider both direct and indirect issues when creating or modifying a user interface for use in multiple nations or cultures. Direct issues include the design of the control elements, dialogue system, information presentation, and warnings. Indirect issues include functional and technological features, technical documentation, training, and user support.
19.4.1 DIRECT USER-INTERFACE ISSUES 19.4.1.1 Hardware Design Issues The user interface must fit the workspace, size of control elements, and other components to the anthropometric features of the target users. Input and output (display and printing) should facilitate the user’s interaction with the medical device. This is especially important
TAF-ER562X_WINGR-09-0401-C019.indd 782
11/9/10 1:10:56 AM
Cross-National and Cross-Cultural Design of Medical Devices
783
for languages (e.g., Greek, Russian, Chinese, Japanese) with many ideographical characters. For example, a Chinese user may prefer a writing pad rather than a keyboard to input the Chinese characters. The memory size required for the device’s character set should also be considered. 19.4.1.2 Interface Structure Interface structure refers to the framework within which a medical device’s operating functions and information presentations are organized. For many medical devices, there is more than one user type (e.g., physicians and nurses). If workflow, task completion, and responsibility vary among users, the device user-interface structure should be optimized to support all important user needs to the greatest extent possible. For example, the primary users of intensive care unit ventilators in the United States are respiratory therapists but are physicians or nurses in Europe. If an intensive care unit workstation designed for use in both the United States and Europe incorporates patient physiological monitoring and ventilation, then the structure of the interface should consider the needs and functional requirements of all groups of users. It should also consider that in the United States, several different people may interact with the integrated workstation concurrently. 19.4.1.3 Operational Sequence Language may influence a user’s preferred sequence of operations. For example, in Japanese, the object is generally identified first, followed by the action to be performed. This is probably due to the fact that in Japanese grammar, the subject is followed by the object and then a verb, for example, “I the file delete” instead of “I delete the file.” In English grammar, a subject is followed by a verb and then an object. In a typical graphical editor, users are required to first select an object in the window and then select a menu item to specify an action. Japanese users may find this sequence more natural than U.S. users. More significantly, the operational sequence is largely determined by the user’s work habits. For example, the operational sequence for setting the infusion rate of a large-volume infusion pump is quite different for nurses in Germany, compared with those in Spain, Switzerland, and the United Kingdom because the VTBI (volume to be infused) function is rarely used in Germany but is frequently used in the other countries.
19.4.2 INFORMATION PRESENTATION 19.4.2.1 Language Issues The translation of a device’s user interface into the user’s native language is the most basic adaptation of a medical device to a particular country or region. The translation should be completed by persons with experience using these medical devices and by persons with a working knowledge of the daily (idiomatic or conversational) language of the intended user population (rather than an academic knowledge). Even so, target users may still have problems understanding the translated user interface. The translation should be evaluated as part of the user interface for each target market. Several details must be considered during translation: • Display resolution—A higher and/or larger display resolution is needed to display a Chinese character than a Roman character.
TAF-ER562X_WINGR-09-0401-C019.indd 783
11/9/10 1:10:56 AM
784
Handbook of Human Factors in Medical Device Design
• Text field (character) widths—A paragraph of German text usually occupies more space than a paragraph of English text; a general rule of thumb is 25% to 30% more space for German. • Prioritized translation—If a complete translation of the user interface is unrealistic considering time and cost, at least the most critical labeling should be translated into the native language. The risk analysis should drive the selection of text to be translated, and the risk analysis must be updated to address the nontranslated parts. • Country requirements—Translation requirements may exist for the target market. For example, some countries require specific languages for devices and their accompanying documents. Other countries allow devices to be in a foreign language but require the accompanying documents to be in a specific native language. Manufacturers should also check with sales agents within the country because specific language requirements may exist in tender documents. 19.4.2.2 User-Interface Orientation A consequence of translation is reading direction as it relates to the organization and display of information in a device user interface. For example, Arabic and Hebrew require that the information content be presented from right to left, so the upper-right point is used as the origin for user-interface layout. In contrast, Chinese users perform better with menus in a vertical layout when compared to a horizontal English layout (Dong and Salvendy, 1999). To facilitate task completion and reduce the risk of use error, designers should consider reading direction when adapting the user-interface layout. However, more computer-literate user populations may have already adapted to the widespread left-to-right orientation used in Western-designed computer interfaces. 19.4.2.3 Units of Measure and Format To ensure that information is presented in the correct way for the target user, the user interface should incorporate the target market’s national formats for presentation of numbers, time, date, address, name, and so on. 19.4.2.4 Symbols The effective use of symbols alone (without accompanying text) in device user interfaces is questionable because of possible misinterpretation in certain cultures. Whenever symbols are incorporated, the manufacturer should test the symbols in all target markets. International standards specify hundreds of symbols (e.g., IEC 60878). Many manufacturers take this as permission to incorporate these symbols into device user interfaces. Unfortunately, some symbols may not be understood by a significant number of users, and comprehension varies from country to country. The Center for Ergonomics and Usability Engineering of the Muenster University of Applied Sciences evaluated the comprehensibility of some common “international” symbols using the evaluation procedure suggested by ISO 9186 (2001) (Liu and Hölscher, 2005, 2006). Comprehensibility differed, sometimes appreciably, between German and Chinese users and between symbols (Table 19.6). 19.4.2.5 Color Coding Despite efforts to standardize color coding in the medical domain, consistent use of color coding has not yet been achieved because of both cultural and national differences. For
TAF-ER562X_WINGR-09-0401-C019.indd 784
11/9/10 1:10:57 AM
Cross-National and Cross-Cultural Design of Medical Devices
785
TABLE 19.6 Comprehensibility of Some Standard Symbols* in Different Countries Comprehensibility
Description
Symbol’s IEC Number
In Germany
In China
Date of manufacturing
2497
35.0%
46.2%
Display transfer
5648
30.0%
11.5%
Do not reuse
1051
32.5%
46.2%
Zero point adjustment
0540
70.0%
0.0%
Person identification
5664
20.0%
11.5%
Manual control
0096
48.8%
61.5%
Locking
5569
67.5%
92.3%
Bell cancel
5576
100.0%
65.4%
Symbol
*Symbols with numbers below 5000 are published in ISO 7000; those with numbers from 5000 onward refer to IEC 60417.
example, the color coding for compressed gas cylinders for medical use still varies from country to country. Table 19.7 shows different “standard” colors for four medical gases in the United States, Japan, Europe, and the DACH countries (Germany, Austria, Switzerland, and Hungary). The DACH countries just recently completed a 10-year transition from their gas cylinder colors to the European ISO colors. In recognition of the potential risks of ambiguous color coding, the U.S. National Center for Patient Safety (NCPS) has recommended that other labeling modalities (e.g., text, symbols, and shape) should be the primary means of identifying gas cylinder content. The NCPS is working with the U.S. Food and Drug Administration to identify optimal labeling without use of color coding. An internationally accepted ISO standard for color coding might be a better solution. However, implementing such a standard would require a lengthy transition period with specific safety measures to minimize the risk of use errors. TABLE 19.7 Color Coding for Gas Cylinders in Different Areas
Gas Type
United States (ANSI Z535.1:1991 and CGA C-9:2004)
Japan (JIS T7111)
Europe (ISO 32)
DACH (DIN 13252, dismissed)
Green Blue Black Yellow
Green Blue Gray Yellow
White Blue Black Black and white
Blue Gray Green Yellow
Oxygen Nitrous oxide Nitrogen Air
TAF-ER562X_WINGR-09-0401-C019.indd 785
11/9/10 1:10:57 AM
786
Handbook of Human Factors in Medical Device Design
TABLE 19.8 Color Coding for AC Power Cords Lead
United States (UL1950)
Europe (DIN VDE 0293-308)
Black White Green
Brown Blue Green/yellow
Live conductor Neutral Protective earth (ground)
Inconsistent color coding used on power cords may result in hazardous situations for multicultural users (Table 19.8). The color markings on specific ECG leads are yet another example. Two sets of ECG lead labels and colors exist: the U.S. national standard (normally referred to as the AHA set) and the set defined by the IEC standard (Table 19.9). Color coding used in medical devices should be consistent and should not conflict with the meanings of colors already in use in target markets.
19.4.3 INDIRECT USER-INTERFACE DESIGN ISSUES 19.4.3.1 Functionality The range of functions available in a medical device should be adapted to the needs of the target users since the desired level of functionality may differ between nations or cultures. Typically, a device used in Western Europe has a wide range of functions so as to appeal to wider group of users. The same device may not be suitable for Chinese users who tend to emphasize device functions they need presently. An all-in-one device is overly complex for their needs and may be considered too extravagant to justify its purchase. Certain functions may become obsolete or unnecessary as medical practice evolves in individual countries. An example is the VTBI function common to many large-volume TABLE 19.9 ECG Lead Labels and Colors Specified in Different Standards ANSI/AAMI EC53 Placement Right arm Left arm Left leg Right leg 4th intercostal—right sternum 4th intercostal—left sternum Midway between V2 and V4 5th intercostal—left midclavicular Left anterior axillary line at V4 Left midaxillary line at V4 Chest
TAF-ER562X_WINGR-09-0401-C019.indd 786
Electrode Identifier RA LA LL RL V1 V2 V3 V4 V5 V6 C
IEC 60601-2-27
Color Code
Electrode Identifier
Color Code
White Black Red Green Brown/red Brown/yellow Brown/green Brown/blue Brown/orange Brown/violet Brown
R L F N C1 C2 C3 C4 C5 C6 C
Red Yellow Green Black White/red White/yellow White/green White/brown White/black White/violet White
11/9/10 1:10:57 AM
Cross-National and Cross-Cultural Design of Medical Devices
787
infusion pumps. The infusion pump’s VTBI function automatically calculates the pump’s flow rate using information the caregiver provides by setting the total amount of fluid to infuse followed by the running time of the infusion. The VTBI function, routinely used by nurses in countries including Spain, Switzerland, the United States, and the United Kingdom, is rarely used in Germany. In Germany, the clinical practice is for the caregiver to use the prescribed dose of the drug and its characteristics to directly calculate the flow rate that the user then enters into the infusion pump. User groups may prefer different elements to control the medical device. For example, nurses are frequently asked to administer a prescribed infusion therapy exactly to the decimal, while physicians often titrate the dose to obtain the required effect by increasing or decreasing it in small steps. A numeric keypad is an appropriate way for nurses to set up the pump, while up and down arrow keys may better support the physician’s need to adjust the dose. 19.4.3.2 Technical Features While not directly a human factors issue, the technical specifications of a medical device must meet the relevant regulatory requirements of the target market. For example, manufacturers must address the different limits for chassis/enclosure touch current and earth leakage current found in the U.S. national standard and the IEC standard. The U.S. national standard allows a maximum current of 300 microamperes for normal and single fault conditions for both values. The IEC standard allows a maximum current of 100 microamperes (normal conditions), 500 microamperes (single fault conditions) for chassis/enclosures touch current, and 500 microamperes (normal conditions) or 1 milliampere (single fault conditions) for earth leakage current. Other differences, such as in local power supply, radio-frequency allocation, and electromagnetic interference and compliance, also need to be considered. To ensure that medical devices function reliably wherever they are used, device design should incorporate special technical measures intended to mitigate the effects of possible environmental factors such as humidity, air pollution (e.g., acidic substance), ambient temperature, electrical supply, and so on. To the extent that differences in regulatory or technical requirements change how users may interact with the device, this becomes a user-interface issue. 19.4.3.3 User Support Because of geographical restraints, socioeconomic factors, and possible shortages of qualified service personnel available in a specific area, support services such as maintenance and repair, spare part supply, and user training may be problematic. Special measures, such as using spare parts that are readily available on the local market, should be considered to simplify the dependence on manufacturer-provided support. With widespread access to the Internet, manufacturers should consider providing online support for remote customers. However, to avoid creating new cross-cultural or cross-national communication problems, manufacturers should employ local service personnel to provide this kind of support. Suitable training programs should be provided for users in different cultures or nations according to their different learning styles, language, education, and experience. Training should be tailored to provide the relevant content and may be organized in classes, at work, or simply through tutorial programs designed around a device user interface. Training
TAF-ER562X_WINGR-09-0401-C019.indd 787
11/9/10 1:10:57 AM
788
Handbook of Human Factors in Medical Device Design
materials should be available to hospitals and caregivers and can be based on textbooks or video. Suitable instructors should be well chosen and properly trained to be able to impart effectively the essential information and skills to the target users. 19.4.3.4 Technical Documentation Challenges in designing technical documentation for international markets include language issues and the suitability of the type, format, and scope of the documents. Language issues include translating the technical documentation into the native language of the users and adjusting the language style to the reading habits of target users and service personnel. A literal (word-for-word) translation will almost always be problematic. A culturally sensitive translator should be used, and either involvement of bilingual users in the translation or rigorous user testing is a minimum requirement. Manufacturers may provide technical documentation in a multilingual form, incorporating several languages in a document. This may be a regulatory requirement in countries with more than one official language (e.g., Canada [French and English], Belgium [French, Flemish, and German], and Switzerland [French, German, and Italian]) (del Galdo and Nielsen, 1996). One useful technique to assure a high quality translation is to reverse translate the material to assure it reverts to the original (see Chapter 5, Documentation). The type and format of the documentation should also be adapted to the target user’s cultural/national needs. Manufacturers could consider providing documentation in electronic form (e.g., in CD-ROM). A quick-start manual is often valuable. The content of the technical documentation should be adjusted to the experience or training level of target users.
19.5 DESIGN RESOURCES FOR CROSS-CULTURAL AND CROSS-NATIONAL DESIGN Decision making on cross-cultural and cross-national design is influenced largely by the following factors: • How large is the expected market volume for the device? • What is the scope of the changes required to modify the design for the new market? • How complex is the user interface? • If device-related material is not translated into the local language, will users have difficulties operating the device, particularly in high-stress situations? The decision on when to start a cross-cultural and cross-national design process can be based on a cost–benefit analysis. Manufacturers have to consider their available design resources, experience, and the potential benefits gained through the design adaptations. The manufacturer’s representatives in the target markets should be able to provide adequate information. The final decision will be based on the device and the manufacturer’s own circumstances. The general development process that incorporates cultural and national differences is essentially the same as the human factors engineering process defined in national and international standards (e.g., IEC 62366, IEC 60601-1-6, and AAMI HE74). Recognition of cultural or national differences may, however, affect the characteristics of the analysis and specification (user investigation), design and realization, and design evaluation.
TAF-ER562X_WINGR-09-0401-C019.indd 788
11/9/10 1:10:57 AM
Cross-National and Cross-Cultural Design of Medical Devices
789
19.5.1 ANALYSIS PHASE CONSIDERATIONS Investigating global use requirements may be complicated by the fact that the investigators and the users may be from different cultures. Because misinterpretations may occur, manufacturers should start validating the global requirements early and continue throughout the development process. It is useful to have one or more “cultural translators”—individuals from the target culture who are also freely conversant with the investigators’ background and culture. For example, when a Western manufacturer is trying to adapt a specialized medical device for the Chinese market, it can be valuable to draw on the expertise of a Chinese-born clinician who trained in the West, is familiar with the Western version of the device, and now practices in China. The investigation should focus on differences between target cultures and the significance of those differences. These differences should be identified before any design for the target market begins. An explanatory rationale for each identified and documented difference should also be included so that designers can use this information to generate design features effectively. In the analysis phase, the hazard/risk analysis should be updated to consider worst-case use scenarios of the target market.
19.5.2 DESIGN PHASE CONSIDERATIONS Designers usually depend on prior experience to convert design requirements into specific design features. However, designers may have difficulty evaluating design alternatives with respect to global requirements since their experience in many cases is based on a different culture (their own). For some design issues (e.g., the language issues), designers will need to collaborate with native experts to attain suitable design solutions. Existing devices accepted in the target market may be a useful source of design ideas for new devices. A manufacturer’s representatives in target markets may be able to collect helpful comments and insights on design issues both through their own observations and from target users. Desirable design features can be characterized according to their similarity among different devices and across different cultures/nations. Summarizing design features systematically into groups using the categories illustrated in Figure 19.3 helps to organize the cross-cultural and cross-national design issues and manage the cost and timeline of the development effort. The possible design features can be divided into four categories. • Category I—Design features that are the same for different devices and for multiple markets. If these design features are generated and evaluated, they can be applied in the cross-cultural and cross-national design with the lowest additional costs. • Category II—Design features that are different for different devices but the same for different markets. The manufacturers could integrate these features in a device as a base to form its “core” for adaptation to the unique requirements of different markets. • Category III—Design features that are the same among devices but different for different markets. They are therefore usually taken as features for localizing a device for a target market. • Category IV—Design features that differ between devices and between different markets. These features usually cause difficulties in the cross-cultural and crossnational design process and should be avoided in the design whenever possible.
TAF-ER562X_WINGR-09-0401-C019.indd 789
11/9/10 1:10:57 AM
Different
III
IV
Same
Handbook of Human Factors in Medical Device Design
For different markets
790
I
II
Same Different For different devices FIGURE 19.3 Categories of design features for different devices and markets. Identification of the category to which a design feature belongs will inform its appropriate use in cross-cultural and cross-national design.
To control the cost of cross-cultural and cross-national design, manufacturers should document those design features that are required to meet the needs of a target market. A manufacturer’s initial investment, to identify the user requirements and generate new design adaptations, does not increase linearly with each new device or market since many of the necessary design adaptations may apply to other new markets or to other devices. In the long term, an investment in cross-cultural and cross-national design pays dividend through more safe, effective, and satisfactory use of the devices in multiple markets. Both of the two general approaches to determine how specific design features should be adapted (Figure 19.4) are based on the analysis of the market: 1. Internationalization describes a way of identifying requirements of all target markets and developing a “global base” device structure with common design features that can be customized by adding features for different markets.
Internationalization:
Localization:
One universal design which can be used in more cultures
One special design for one specific culture
A design which is free of special cultural specifics or an all-in-one design with different options for different cultures
Normally easier to implement and more robust than internationalization
Internationalization Loc. 1
Loc. 2
Loc. 3
..........
FIGURE 19.4 There are two approaches to the cross-cultural and cross-national design of medical devices: internationalization and localization.
TAF-ER562X_WINGR-09-0401-C019.indd 790
11/9/10 1:10:57 AM
Cross-National and Cross-Cultural Design of Medical Devices
791
2. Localization describes a way of developing a device with features specific to a particular market. The localization approach (development of separate devices for each specific market) is suitable for manufacturers that design medical devices for only a few markets. Manufacturers with a global array of markets usually find that internationalization is the more suitable approach (to develop a device core that can be modified for a specific market). One of these two approaches should be chosen before design begins. Attempting to modify a completed design after the fact for a new target market may require significant and costly design changes.
19.5.3 EVALUATION PHASE CONSIDERATIONS The desired design features for the target market should be evaluated as early as possible in the design process. Not all standard human factors verification methods are suitable due to the nature of cross-cultural and cross-national design. Expert reviews, cognitive walk-through, or heuristic analysis may not be sufficient if conducted by a person with a cultural background different from the target user. At a minimum, a medical device should be evaluated by local usability specialists who can point out, for example, design attributes that conflict with the expectations of the target users. Usability testing in different locales is useful but more expensive. Validation may be conducted through different methods including those described in Chapter 6, “Testing and Evaluation.” The preferred way to mitigate potential risks in a design is to validate that design in the target markets. Sometimes only small-scale usability tests are feasible because of resource limitations. However, any usability testing is better than none since even limited data can help identify usability problems in the target market. Usability testing in simulated clinical environments or in the actual use environment can be an effective way to discover residual risks. Testing in foreign markets is usually more difficult than testing in the manufacturer’s region or nation. Additional difficulties exist in selecting target cities, identifying suitable test locations, identifying and acquiring the necessary equipment, recruiting participants, conducting the test, and analyzing the results. Testing in foreign countries may also require that interpreters be selected and trained, that evaluation materials and tests be translated, and so on. However, this is an essential step in ensuring that a design meets the needs of the target users.
19.6 CONCLUSIONS As medical devices become accessible on a global scale, manufacturers need to recognize the importance of cross-cultural and cross-national considerations. Manufacturers who analyze culture- and nation-specific requirements of each market’s user groups and integrate these requirements into device design significantly increase their chances of a successful product. Cross-cultural and cross-national design addresses the fundamental issue of how far users can adapt to technology (without risk of harm to themselves or their patients) versus how far technology should adapt to users (possibly at increased development cost). From a human factors perspective, successful devices will meet the needs of users in the context of actual work. The manufacturer must make this usability/safety
TAF-ER562X_WINGR-09-0401-C019.indd 791
11/9/10 1:10:58 AM
792
Handbook of Human Factors in Medical Device Design
versus development cost trade-off explicit and well informed during cross-national and cross-cultural design decisions. Unfortunately, some markets are not large enough to be treated independently, so their unique requirements may not be considered explicitly. It is the manufacturer’s obligation, via its risk management process, to make decisions on the acceptability of the residual risks and to establish methods to collect postmarket data for feedback into the risk management process. Globalization may replace many local habits and values with internationally uniform habits and values. Many international user populations have been strongly influenced by the computer, Internet, entertainment industry, news, and literature. The need for specific cross-cultural and cross-national designs in a medical device user interface may become less critical over time in these populations. However, even in an apparently sophisticated international market, subtle cultural, language, or other differences predispose devices to unanticipated use errors if a proper cross-cultural and cross-national design process is not followed.
19.7 CASE STUDY An infusion pump from Japan designed for the local expectations of the home market was introduced into the European market. A critical design feature was the use of color coding for functional keys. In the Japanese design, the color red was assigned for the start button. This design conflicted with standard color coding in Western countries as well as international standards, in which red means stop. The pump could cause significant risks to patient safety if, for example, the user prematurely pressed the start key during pump programming, thinking that it was the control to stop the programming task. This would be more likely under stress or time pressure. A Western design that considered the cultural requirements of the target market would require a different coding scheme (for internationalization) or a different color scheme in the device marketed in the West (for localization).
RESOURCES Feigenwinter, P., Györy, L., et al. (2000). Die Umstellung der Gaskennfarben in Deutschland, Österreich, der Schweiz und Ungarn. mt-Medizintechnik, 120, 134–137. Fernandes, T. (1995). Global Interface Design: A Guide to Designing International User Interfaces. Boston: AP Professional. Hall, E. T. (1989). Beyond Culture. New York: Anchor Books. Henderson, P. W., Cote, J. A., Leong, S. M., and Schmitt, B. (2003). Building strong brands in Asia: Selecting the visual components of image to maximize brand strength. International Journal of Research in Marketing, 20, 297–313. Hofstede, G. (1997). Cultures and Organizations: Software of the Mind. New York: McGraw-Hill. IEC 60601-1-8. (2006). Medical electrical equipment—Part 1–8: General requirements for safety— Collateral standard: General requirements, tests and guidance for alarm systems in medical electrical equipment and medical electrical systems. Geneva: International Electrotechnical Commission. Karwowski, W. (Ed.), International Encyclopedia of Ergonomics and Human Factors (2nd ed., pp. 1053–1057). Boca Raton, FL: CRC Press/Taylor & Francis. Kaufman-Scarborough, C., and Lindquist, J. D. (1999). Time management and polychronicity: Comparisons, contrasts, and insights for the workplace. Journal of Managerial Psychology, Special Issue on Polychronicity, 14(3/4), 288–312.
TAF-ER562X_WINGR-09-0401-C019.indd 792
11/9/10 1:10:58 AM
Cross-National and Cross-Cultural Design of Medical Devices
793
Maletzke, G. (1996). Interkulturelle Kommunikation [Intercultural Communication]. Opladen: Westdeutscher Verlag. Nielsen, J. (1993). Usability Engineering. San Francisco: Morgan Kaufmann. Pease, A. and Pease, B. (2000). Why Men Don’t Listen and Women Can’t Read Maps: How We’re Different and What to Do About It. New York: Welcome Rain Publishers. Russo, P. and Boor, S. (1993, April 24–29). How fluent is your interface? Designing for international users. In Proceedings INTERCHI ’93 Conference on Human Factors in Computing Systems: INTERACT ’93 and CHI’93 (pp. 342–347). Amsterdam: ACM Press. Whorf, B. L. (1956). Language, thought and reality. In J. B. Carroll (Ed.), Selected Writings of Benjamin Lee Whorf (pp. 246–270). Cambridge, MA: MIT Press.
REFERENCES ANSI/AAMI EC13. (2007). Cardiac Monitors, Heart Rate Meters, and Alarms. New York. American National Standards Institute. ANSI/AAMI HE74. (2001). Human Factors Design Process for Medical Devices. Philadelphia, PA: Association for the Advancement of Medical Instrumentation. Aykin, N. (Ed.). (2005). Usability and Internationalization of Information Technology. Mahwah, NJ: Lawrence Erlbaum Associates. Casey, S. (1998). Set Phasers on Stun and Other True Tales of Design, Technology, and Human Error. Santa Barbara, CA: Aegean Publishing. CGA C-9. (2004). Standard Color Marking of Compressed Gas Containers for Medical Use. Chantilly, VA: Compressed Gas Association. del Galdo, E. M. and Nielsen, J. (Eds.). (1996). International User Interfaces. New York: Wiley. DIN VDE 0293-308. (2003). Berlin. Deutsches Institut für Normung. Dong, J. M. and Salvendy, G. (1999). Designing menus for the Chinese population: Horizontal or vertical? Behaviour and Information Technology, 18(6), 467–471. FAA Human Factors Team. (1996). The Interfaces Between Flight Crews and Modern Flight Deck Systems. Washington, DC: Federal Aviation Administration. Forcier, H. and Weinger, M. B. (1993). An evaluation of proposed graphical symbols for medical devices. Anesthesiology, 79, 625–627. Gosbee, J. and DeRosier, J. M. (2002). Oxygen (compressed gas) Cylinder Hazard Summary. Available: http://www.patientsafety.gov/O2Cylinder.html. Honold, P. (1999). Learning how to use a cellular phone: Comparison between German and Chinese users. Technical Communication, 46(2), 196–205. IEC 62366. (2007). Medical devices—Application of usability engineering to medical devices. Geneva: International Electrotechnical Committee. IEC 60601-2-27. (2005). Medical electrical equipment—Part 2-27: Particular requirements for the safety, including essential performance, of electrocardiographic monitoring equipment. Geneva: International Electrotechnical Committee. IEC 62366. (2007). Medical electrical equipment— Part 1–6: General requirements for safety— Collateral standard: Usability. Geneva: International Electrotechnical Commission. IEC 60878 TR Ed. 2.0. (2003) (draft). Graphical symbols for electrical equipment in medical practice. Geneva: International Electrotechnical Commission (62A/416/DTR). IEC 60417. (1998). Graphical symbols for use on equipment. Geneva: International Electrotechnical Committee. ISO 14971. (2000). Medical devices—Application of risk management to medical devices. Geneva: International Organization for Standardization. ISO 7000. (1989). Graphical symbols for use on equipment. Index and synopsis. Geneva: International Organization for Standardization. ISO 32. (1977). Gas cylinders for medical use—Marking for identification of content. Geneva: International Organization for Standardization.
TAF-ER562X_WINGR-09-0401-C019.indd 793
11/9/10 1:10:58 AM
794
Handbook of Human Factors in Medical Device Design
ISO 9186. (2001). Graphical symbols—Test methods for judged comprehensibility and for comprehension. Geneva: International Organization for Standardization. JIS T7111. (1993). Hose assemblies for use with medical gas systems. Tokyo: Japanese Standards Association, Tokyo. Liu, L. and Hölscher, U. (2005). Evaluation of Graphical Symbols used in Intensive Care Units (ICU): Comprehension Among Users in Different Countries. Paper presented at the 16th annual meeting of ESCTAIC, Aalborg, Denmark. Maddox, M. (Ed.). (1998). Human Factors Guide for Aviation Maintenance. Washington, DC: Office of Aviation Medicine, Federal Aviation Administration. Nisbett, R. E. (2003). The Geography of thought: How Asians and Westerners Think Differently and Why. New York: Free Press. UL1950. (1989). Standard for safety, information technology equipment including electrical business equipment. Camas, WA: Underwriters Laboratories.
TAF-ER562X_WINGR-09-0401-C019.indd 794
11/9/10 1:10:58 AM
Editors Matthew B. Weinger, MD, is the Norman Ty Smith Chair in Patient Safety and Medical Simulation. He is a professor of anesthesiology, biomedical informatics, and medical education at Vanderbilt University School of Medicine. He is the director of the Vanderbilt Center for Perioperative Research in Quality, director of the Vanderbilt Center for Medical Simulation, codirector of the Center for Improving Patient Safety, and a member of the Health Services R&D Service at the Tennessee Valley VA Healthcare System. He has been teaching and conducting research in anesthesia patient safety, human factors, simulation, and clinical decision making for almost two decades. He received a bachelor’s degree in electrical engineering and a master’s degree in biology (neurosciences) from Stanford University in 1978. He completed his MD degree at the University of California, San Diego, in 1982 and did his anesthesiology residency training at the University of California, San Francisco. He has published extensively on topics relevant to medical technology design and evaluation, including human factors, use error, user-interface design, measures of clinician performance, workload, alarms and vigilance, the nature of clinical experience, automation, clinician fatigue, and clinical decision support. He is on the editorial board of the journal Human Factors. He is cochairman of the Association for the Advancement of Medical Instrumentation’s Human Factors Committee, which is developing national standards for all medical device user interfaces.
Michael Wiklund is president of Wiklund Research & Design, Inc., a human factors research and design consulting firm. His consulting practice focuses on the development of safe, useful, usable, and appealing medical products and systems. He received his MS degree in engineering design (human factors) from Tufts University in 1984 and is a licensed professional engineer. He is a voting member of Association for the Advancement of Medical Instrumentation’s Human Factors Engineering Committee. An adjunct associate professor at Tufts University, he annually teaches a graduate course on software user-interface design. His prior publications include three books: Designing Usability into Medical Products, Medical Device and Equipment Design, and Usability in Practice. He holds several user-interface design patents and has served as a judge for both the Medical Design Excellence Awards and the International Design Awards.
795
TAF-ER562X_WINGR-09-0401-C020.indd 795
11/9/10 1:11:54 AM
796
Editors
Daryle Gardner-Bonneau, PhD, is the principal of Bonneau and Associates, a consulting firm specializing in human factors as applied to health care, telecommunications services, and over-the-phone speech technology user interfaces. She earned her PhD at The Ohio State University in human performance, is currently an Adjunct Associate Professor in the College of Health and Human Services at Western Michigan University, and is a Fellow of the Human Factors and Ergonomics Society (HFES). She serves as the Chair of the Technical Advisory Group (TAG) to the International Standards Organization (ISO) Technical Committee (TC) 159 – Ergonomics, and is actively involved in technical standards work in ergonomics for both HFES and the Association for the Advancement of Medical Instrumentation (AAMI). For the past two years, Daryle has served on a National Academy of Sciences study panel on the Role of Human Factors in Home Health Care. She is the author of more than 50 articles and book chapters on human factors and co-authored and edited Human Factors and Voice Interactive Systems. She serves on the editorial board of Ergonomics in Design and was the editor-in-chief of the International Journal of Speech Technology from 1999 to 2006.
TAF-ER562X_WINGR-09-0401-C020.indd 796
11/9/10 1:11:54 AM
Contributors W. Gary Allread is a program director for the Institute for Ergonomics at Ohio State University. He manages and conducts ergonomics research and educational programs in the physical and cognitive ergonomics domains. He also provides training and ergonomic technical assistance to a wide range of businesses and organizations, focusing on the prevention of injuries and musculoskeletal disorders in many types of occupational work settings and with a variety of consumer products. He has authored peer-reviewed research articles on various ergonomics topics and also has implemented ergonomics programs in several companies. He received his PhD and MS in industrial and systems engineering from Ohio State University, with an emphasis on industrial ergonomics and biomechanics. He previously earned a BSE in human factors engineering from Wright State University. He has been a member of the Human Factors and Ergonomics Society since 1986 and holds certified professional ergonomist credentials from the Board of Certification in Professional Ergonomics. Carla J. Alvarado received her BS from Miami University, Oxford, Ohio, and her MS in preventive medicine and epidemiology and PhD in industrial and systems engineering and human factors from the University of Wisconsin–Madison. She is a research scientist emerita at the Center for Quality and Productivity Improvement of the University of Wisconsin– Madison. Prior to her present position, she was an infection control professional for 19 years at the University of Wisconsin Hospitals and Clinics. Her publications and research areas include nosocomial infections associated with medical devices, safety culture/safety climate, work redesign, and human factors and ergonomics related to health care. She received the APIC Carole DeMille Award for outstanding career achievement in the field of infection control and health care epidemiology. She is on the editorial board of the American Journal of Infection Control and has served on the Board of Directors of the Association for Professionals in Infection Control and Epidemiology, the American Hospital Association Technical Advisory Panel on Infections in Hospitals, and the Board of Trustees of the Research Foundation for Complications Associated with Health Care. She is a past chair of the Health Care Technical Group of the Human Factors and Ergonomics Society.
797
TAF-ER562X_WINGR-09-0401-C021.indd 797
11/9/10 2:03:34 AM
798
Contributors
Larry W. Avery is a senior human factors analyst with BMT Designers and Planners in Washington, D.C. His career in human factors has spanned more than 25 years. His recent work involves developing human factors design standards for the U.S. Army, guidelines for the maritime industry, and user-interface style guides for mobile devices. He has an MS in industrial psychology from George Mason University.
Ramon Berguer holds a BS in computer science from the University of Michigan and obtained his MD from Wayne State University. He completed his general surgery residency at the University of Colorado Health Sciences Center along with a one-year Fellowship in Gastrointestinal Research. He is currently chief of surgery at Contra Costa Regional Medical Center. His academic degrees include clinical professor of surgery at the University of California, Davis, School of Medicine, and adjunct associate professor in the School of Engineering at California State University, Sacramento. He has published more than 50 peer-reviewed scientific papers as well as six book chapters in the areas of stress immunology, laparoscopic surgery, surgical ergonomics, and the prevention of sharps injuries in the operating room. He is a guest speaker at national and international surgical conferences on the topics of surgical ergonomics and sharps injury prevention. He has been a consultant in surgical ergonomics for leading medical instrument manufacturers and is presently cofounder and president of Lifeguard Surgical Systems. Richard Botney is an assistant professor of anesthesiology and perioperative medicine at the Oregon Health and Science University in Portland, Oregon. He holds a BS in electrical engineering and computer sciences from the University of California, Berkeley, and obtained his MD from Washington University. He has been chair of the university’s Patient Safety Committee, has been interested in human factors and patient safety for more than a decade, and has authored several chapters and articles on this subject. He is the co-author of a recently published text, Patient Safety in Plastic Surgery, and is actively involved in technical standards work for the Association for the Advancement of Medical Instrumentation. Pascale Carayon, PhD, is the Procter & Gamble Bascom Professor in Total Quality in the Department of Industrial and Systems Engineering and the director of the Center for Quality and Productivity Improvement at the University of Wisconsin– Madison. She received her engineering diploma from the Ecole Centrale de Paris in 1984, and her PhD in industrial engineering from the University of Wisconsin–Madison in 1988. Her research areas include systems engineering, human factors and
TAF-ER562X_WINGR-09-0401-C021.indd 798
11/9/10 2:03:35 AM
Contributors
799
ergonomics, sociotechnical engineering, and occupational health and safety. She is the North American editor for Applied Ergonomics. In July 2006, she was elected as the secretarygeneral of the International Ergonomics Association. She is a fellow of the Human Factors and Ergonomics Society. Her current research is funded by the Agency for Healthcare Research and Quality, the National Science Foundation, the Department of Defense, the National Institute on Aging, the National Institute for Occupational Safety and Health, and various foundations and businesses. She leads the Systems Engineering Initiative for Patient Safety at the University of Wisconsin–Madison (http://www2.fpm.wisc.edu/seips). Joseph F. Dyro is president of the Biomedical Resource Group in Setauket, New York, and editor of the Journal of Clinical Engineering. He is past president and a founding fellow of the American College of Clinical Engineering, a founding fellow of the American Institute of Medical and Biological Engineering, a certified clinical engineer, an administrative council member of the International Federation for Medical and Biological Engineering, and a senior member of the Institute of Electrical and Electronics Engineers. He consults on the safety and efficacy of medical devices and the management of health care technology. He is editor in chief of the Clinical Engineering Handbook. He has published extensively and has lectured worldwide on clinical engineering and medical device safety and design. He is a graduate of the Massachusetts Institute of Technology with a BS degree and earned MS and PhD degrees from the University of Pennsylvania. He has worked for ECRI and University Hospital, Stony Brook, New York, where he was director of biomedical engineering. Torsten Gruchmann is founder and managing director of Use-Lab GmbH, an independent German company that focuses on the development and optimization of usability concepts for medical devices. He has more than eight years of human factors experience, concentrated in the medical arena. In 1995, he graduated from the Muenster University of Applied Sciences in physical engineering. Since then, his work has focused on understanding users’ needs and translating them into product specifications and concept realization. He has shared his professional insights in numerous publications and presentations over the years. He is an active member of several standards committees and associations as well as a board member of the Dutch-German TIMP network. John W. Gwynne III is a senior scientist with Pacific Science and Engineering Group, a human factors research, development, and consulting firm in San Diego, California. He received his PhD in cognitive psychology from the University of New Mexico in 1988. His professional interests include document design and human error and reliability in advanced technology systems. He is a member of the Human Factors and Ergonomics Society and Sigma Xi, the Scientific Research Society.
TAF-ER562X_WINGR-09-0401-C021.indd 799
11/9/10 2:03:36 AM
800
Contributors
Uvo Hoelscher received his bachelor’s degree in chemical engineering from the Technical University of Hannover and his PhD from the Technical University of Clausthal-Zellerfeld. Until 1996, he worked in the Medical Branch of the Draeger Werk AG in Luebeck, Germany, where he developed an interest in human factors engineering. He currently teaches Biomedical Engineering and Ergonomics at the Faculty of Engineering Physics of the Muenster University of Applied Sciences. There he manages a large usability lab that focuses on ergonomics for biomedical devices. He has been involved in national and international standards writing committees for medical devices for more than 17 years. He currently works in the IEC SC 62A WG 5 “Ergonomics” and in the IEC-ISO JWG 4 “Usability” and is a member of the AAMI Human Factors Engineering Committee. He is the convener of the Technical Committee for Ergonomics of the German Association of Biomedical Engineering, organized the MEK 2005 congress on biomedical engineering and ergonomics (240 participants), and was engaged in mirroring this human factors conference through the Association for the Advancement of Medical Instrumentation. In 2008, he organized another MEK congress in Muenster, Germany. Edmond W. Israelski joined Abbott Laboratories in 2001. He is leading a cross-division team to embed best-practice human factors design methods into all of Abbott’s products as well as doing hands-on design of critical new medical devices to ensure safety and usability. He writes corporate policy and teaches for Abbott. He is convener of medical device human factors standards groups for ISO/IEC, as well as co-chair of the HFE committee for AAMI. In 1970, he joined AT&T Bell Labs and worked as a systems engineer, product manager, market researcher, industrial/organizational psychologist, and human factors engineer. In 1997, he joined Ameritech/SBC, where he was the director of human factors. In 2000, he became chief technology officer at Human Factors International, a user-interface design and consulting firm. He is a fellow of the Human Factors and Ergonomics Society. He is a boardcertified human factors professional. He received a BSEE from the New Jersey Institute of Technology, an MS in operations research from Columbia University, and a PhD in human factors from the Stevens Institute of Technology. Michael J. Kalsher is an associate professor of psychology and cognitive science at Rensselaer Polytechnic Institute. He has a PhD in psychology from Virginia Tech and MS and BS degrees in applied/experimental psychology from Montana State University. His research has focused on applying principles of behavioral science to promote safety, including increasing the use of safety belts, reducing drunk driving, and designing effective warnings. He served as chair of the Department of Cognitive Science at Rensselaer from 1997 to 2002. He is the author of numerous articles and book chapters and is coauthor of a popular introductory psychology textbook. He has presented his findings at numerous national and international professional and scientific conferences.
TAF-ER562X_WINGR-09-0401-C021.indd 800
11/9/10 2:03:36 AM
Contributors
801
Ben-Tzion Karsh, PhD, is an associate professor of industrial and systems engineering at the University of Wisconsin– Madison. His research, which has been funded by the Agency for Healthcare Research and Quality (AHRQ), the United Kingdom Department of Health, and the National Library of Medicine, focuses on the use of human factors engineering for improving patient and health care employee safety. He is a past national chair of the Health Care Technical Group of the Human Factors and Ergonomics Society, has served on grant review panels related to patient safety for the AHRQ and the Veteran’s Health Administration, and is a peer reviewer for multiple journals. David A. Kobus is director of medical systems at Pacific Science and Engineering Group, a human factors research, development, and consulting firm in San Diego, California. He has been involved in human factors research and project management for more than 20 years, having served as principal investigator or program manager on more than 30 projects evaluating human performance in complex technological systems. He has published more than 60 papers and technical reports and presented more than 65 papers at national or international conferences. He is past chair of the Medical Systems and Rehabilitation Technical Group of the Human Factors and Ergonomics Society and is a certified professional ergonomist. Long Liu is an associate professor at the School of Mechanical Engineering at the Tongji University in China. He received a bachelor’s degree in industrial design in 1989 and a master’s degree in ergonomics in 1992 from the Chongqing University in China. In 2006, he received his PhD from the University of Kaiserslautern in Germany. In 1998, he worked at the Chongqing University in China as a lecturer, teaching courses in ergonomics and industrial design as well as conducting research in the area of human factors in production systems. From 2002 to 2005, he worked at the Center for Ergonomics and Usability Engineering at the Muenster University of Applied Sciences, specializing in usability and safe use of medical devices. Michael E. Maddox is a senior scientist with HumanCentric Technologies, Inc., in Cary, North Carolina. His human factors career spans more than 25 years in various technical domains. His current work focuses on systematic approaches to patient safety, risk analysis in medical devices and health care processes, and system design and evaluation. He received his MS and PhD degrees in industrial engineering and human factors from Virginia Tech. He is a member of the Human Factors and Ergonomics Society and is a certified human factors professional.
TAF-ER562X_WINGR-09-0401-C021.indd 801
11/9/10 2:03:36 AM
802
Contributors
William H. Muto is principal of human factors engineer in the Diagnostics Division of Abbott Laboratories. He has also served as a human factors consultant to other Abbott product divisions focusing on medical product design. As a key member of the corporate human factors council, he contributes to corporate policies and procedures on human factors and assists in the development and teaching of in-house human factors courses. Prior to joining Abbott, he held senior engineering and management positions at Xerox Corporation and Texas Instruments. He is a voting member of the Association for the Advancement of Medical Instrumentation’s Human Engineering Standards Committee and is a Former director and officer of the Board of Certification for Professional Ergonomics. He is a member of the Human Factors and Ergonomics Society and is a certified professional ergonomist. He holds a PhD in industrial engineering (human factors specialty) from Virginia Tech. Carl Pantiskas received an MS in biomedical engineering from California State University, Sacramento, in 1986. From 1986 to 2006, he worked for Spacelabs Healthcare as a clinical engineer specializing in human factors and system design. Since 2006, he has worked as a senior clinical engineer at Draeger Medical. He has been a member of the Human Factors Committee of the Association for the Advancement of Medical Instrumentation (AAMI) since the late 1980s and was the producer and cochair of that committee from 1998 to 2004. He is currently is a member of the AAMI Standards Board and holds appointments as a U.S. expert member to the International Organization for Standardization/International Electrotechnical Commission (IEC) Joint Working Group on medical alarms and IEC SC62D MT22 on electromedical diagnostic and patient monitoring equipment. Mary Beth Privitera is codeveloper and faculty member in the Medical Device Innovation and Entrepreneurship Program at the University of Cincinnati. She is an assistant professor of biomedical engineering and an adjunct instructor of industrial design. An expert in the application of human factors in medical product design, she has also worked in the medical device industry since 1988. She is currently a principal of Y-NOT Design (Cincinnati), a firm specializing in design consulting for the medical device industry. Previously she held positions at Design Science Inc. (Philadelphia) and Ethicon Endo-Surgery Inc. (Cincinnati) and has consulted for a number of major medical device firms. She is currently a member of the Industrial Designers Society of America and has served on its National Educational Council. In addition, she is a member of the Association for the Advancement of Medical Instrumentation’s Human Factors Committee, the American Society of Engineering Educators, the Product Development and Management Association, and the Design Management Institute. She has been associated with more than 30 product releases, holds five patents and several
TAF-ER562X_WINGR-09-0401-C021.indd 802
11/9/10 2:03:37 AM
Contributors
803
provisional patents, and has published and lectured on a variety of topics, including collaborative design, innovation methodology, and surgical techniques. She received her BS in industrial design from the University of Cincinnati in 1985 and her master of design degree from the University of Cincinnati in 1995. Robert G. Radwin is a professor and founding chair of the Department of Biomedical Engineering at the University of Wisconsin–Madison, where he conducts research and teaches in the areas of ergonomics and human factors. He also holds a faculty appointment in the Department of Industrial Engineering and the Department of Orthopedics and Rehabilitation. He earned his PhD from the University of Michigan and studied at the Center for Ergonomics. He has received numerous honors and distinctions, including a Presidential Young Investigator Award from the National Science Foundation and a Special Emphasis Research Career Award from the National Institute for Occupational Safety and Health. He is a fellow of the Human Factors and Ergonomics Society (United States), a fellow of the Ergonomics Society (United Kingdom), a fellow of the Biomedical Engineering Society, and a fellow of the American Institute of Medical and Biological Engineers. His research fields of interest include analytical methods for measuring and assessing exposure to physical stress in the workplace, ergonomic aspects of manually operated equipment and hand tools, causes and prevention of work-related musculoskeletal disorders and peripheral neuropathies, occupational biomechanics, and rehabilitation engineering. Richard Stein is an engineer with 20 years experience at Guidant, Boston Scientific, and currently St. Jude Medical, where he is developing medical products for use in clinics, hospitals, and the home. He is a voting member of Association for the Advancement of Medical Instrumentation’s Human Factors Engineering Committee. He is winner of a Design Excellence Award in 2004 for the PARTNER Rhythm Assistant™ Inventor and patent holder of Quick-booting of pacemaker programmers and medical devices. He received his BS in electrical engineering from the University of Minnesota in 1977 and currently lectures and advises for its biomedical engineering senior design course. Stephen B. Wilcox, PhD, FIDSA, is a principal and the founder of Design Science, a Philadelphia-based consultancy that provides ethnographic research, human factors, and interaction design for product development for clients, including major medical device manufacturers. He also chairs the Human Factors Professional Interest Section of the Industrial Designers’ Society of America and serves on the Human Factors Engineering Committee of the Association for the Advancement of Medical Instrumentation. He holds a BS in psychology and anthropology from Tulane University, a PhD in experimental psychology from Pennsylvania State University, and a certificate in business administration from the Wharton School of the
TAF-ER562X_WINGR-09-0401-C021.indd 803
11/9/10 2:03:38 AM
804
Contributors
University of Pennsylvania. His recent book, with Michael Wiklund, is titled Designing Usability into Medical Products. Michael S. Wogalter is a professor of psychology at North Carolina State University. He has also worked at Rensselaer Polytechnic Institute and the University of Richmond. He has a PhD in human factors psychology from Rice University, an MA in experimental psychology from the University of South Florida, and a BA in psychology from the University of Virginia. His research focuses on hazard perception, warnings, information design, and applied cognition. He is a fellow of the Human Factors and Ergonomics Society and the International Ergonomics Association and is on the editorial boards of several journals. He has authored numerous research-based publications and has lectured at many national and international professional and scientific meetings.
TAF-ER562X_WINGR-09-0401-C021.indd 804
11/9/10 2:03:38 AM
FIGURE 1.12 An emergency stop button on a scanner is recessed to prevent inadvertent actuation. Large size, red color, symbolic label, shape, and recessing also provide redundant means of differentiating the control from others.
Normal color vision
Tritanopia
Red blind (Protanopia)
Green blind (Deuteranopia)
FIGURE 2.9 Comparison of colors seen by various users of a patient monitor with normal and deficient color vision, including tritanopia, protanopia, and deuteranopia.
TAF-ER562X_WINGR-09-0401-CI.indd 1
11/9/10 2:04:47 AM
98.6
98.6
98.6
98.6
FIGURE 11.17 Red letters on a green background (left) have a lower contrast ratio than red letters on a dark blue background (right), as illustrated when the pairs are converted to gray scale, roughly simulating how an individual with color-impaired vision is likely to perceive it.
FIGURE 12.19 An anesthesia machine’s gas flow meters (tubes) reside directly above their associated controls. The controls are also differentiated by color and tactile feel.
TAF-ER562X_WINGR-09-0401-CI.indd 2
11/9/10 2:04:51 AM
FIGURE 13.8
The use of specific-shape connectors and color coding can help reduce user error.
(a)
(b)
FIGURE 13.10 Color coding can be a useful technique for improving labeling on (a) receptacles and (b) connectors.
FIGURE 13.12 Salience refers to aspects that aid in making the label more conspicuous or prominent. In this example, the labeling is made more conspicuous through the use of color, placement, and coherent information grouping.
TAF-ER562X_WINGR-09-0401-CI.indd 3
11/9/10 2:04:52 AM
FIGURE 13.20 labeling.
The use of different background colors can help differentiate various portions of
FIGURE 13.23 Redundant coding can help ensure that users receive the information they need to operate medical devices safely.
TAF-ER562X_WINGR-09-0401-CI.indd 4
11/9/10 2:05:24 AM