Patient Safety in Health Care Management (Adavances in Health Care Management)

ADVANCES IN HEALTH CARE MANAGEMENT Series Editors: John D. Blair, Myron D. Fottler and Grant T. Savage Recent Volumes: V...

Author: Grant T. Savage

29 downloads 1344 Views 2MB 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

DOWNLOAD PDF

ADVANCES IN HEALTH CARE MANAGEMENT Series Editors: John D. Blair, Myron D. Fottler and Grant T. Savage Recent Volumes: Volume 1:

Advances in Healthcare Management, Edited by John D. Blair, Myron D. Fottler, and Grant T. Savage

Volume 2:

Advances in Healthcare Management, Edited by Myron D. Fottler, Grant T. Savage, and John D. Blair

Volume 3:

Advances in Healthcare Management, Edited by Grant T. Savage, John D. Blair, and Myron D. Fottler

Volume 4:

Bioterrorism, Preparedness, Attack and Response, Edited by J. D. Blair, M. D. Fottler, and A. C. Zapanta

Volume 5:

International Healthcare Management, Edited by Grant T. Savage, Jon A. Chilingerian, and Michael Powell

Volume 6:

Strategic Thinking and Entrepreneurial Action in the Health Care Industry, Edited by John D. Blair, Myron D. Fottler, Eric W. Ford, and G. Tyge Payne

ADVANCES IN HEALTH CARE MANAGEMENT

VOLUME 7

PATIENT SAFETY AND HEALTH CARE MANAGEMENT EDITED BY

GRANT T. SAVAGE University of Missouri, USA

ERIC W. FORD Texas Tech University, USA

United Kingdom – North America – Japan India – Malaysia – China

JAI Press is an imprint of Emerald Group Publishing Limited Howard House, Wagon Lane, Bingley BD16 1WA, UK First edition 2008 Copyright r 2008 Emerald Group Publishing Limited Reprints and permission service Contact: [email protected] No part of this book may be reproduced, stored in a retrieval system, transmitted in any form or by any means electronic, mechanical, photocopying, recording or otherwise without either the prior written permission of the publisher or a licence permitting restricted copying issued in the UK by The Copyright Licensing Agency and in the USA by The Copyright Clearance Center. No responsibility is accepted for the accuracy of information contained in the text, illustrations or advertisements. The opinions expressed in these chapters are not necessarily those of the Editor or the publisher. British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library ISBN: 978-1-84663-954-8 ISSN: 1474-8231 (Series)

Awarded in recognition of Emerald’s production department’s adherence to quality systems and processes when preparing scholarly journals for print

LIST OF CONTRIBUTORS Amy Abbott

Creighton Health Service Research Program, School of Nursing, Creighton University, USA

Donde Batten

Batten Consulting, Houston, TX, USA

Brian Binck

Department of Anesthesia & Critical Care Medicine, Alfred I. duPont Hospital for Children, Wilmington, DE, USA

James D. Bramble

Creighton Health Service Research Program, School of Pharmacy and Health Professions, Creighton University, USA

Racquel M. Calderon

Department of Respiratory Services, Totally Kids Specialty Healthcare, Loma Linda, CA, USA Department of Cardiopulmonary Sciences, School of Allied Health Professions, Loma Linda University, Loma Linda, CA, USA

Marsha Chan

St. Francis Medical Center, Lynwood, CA, USA

Bartholomew E. Clark

Pharmacy Sciences Department, Creighton Health Service Research Program, School of Pharmacy and Health Professions, Creighton University, USA

Richard A. Culbertson

Department of Health Systems Management, School of Public Health and Tropical Medicine, Tulane University, USA ix

x

LIST OF CONTRIBUTORS

Patricia R. DeLucia

Department of Psychology, Texas Tech University, TX, USA

Ellen S. Deutsch

Division of Otolaryngology, Department of Surgery, Alfred I. duPont Hospital for Children, Wilmington, DE, USA

Susan M. Distefano

Texas Children’s Hospital, Houston, TX, USA

Andjela Drincic

Creighton Health Service Research Program, School of Medicine, Creighton University, USA

Eric W. Ford

Jerry S. Rawls College of Business, Center for Innovation, Education & Research, Texas Tech University, TX, USA

Kimberly A. Galt

Creighton Health Service Research Program, School of Pharmacy and Health Professions, Creighton University, USA

Gerald Goodman

Department of Health Care Administration, College of Health Sciences, Texas Woman’s University, TX, USA

Alexia Green

Center for Patient Safety, School of Nursing, Texas Tech University Health Sciences Center, TX, USA

Julia A. Hughes

Department of Health Systems Management, School of Public Health and Tropical Medicine, Tulane University, USA

Ben-Tzion Karsh

Department of Industrial and Systems Engineering, College of Engineering, University of Wisconsin, Madison, USA

Jeff F. Lewis

Kosair Charities Pediatric Convalescent Center, Louisville, KY

xi

List of Contributors

Ann Scheck McAlearney

Division of Health Services Management and Policy, College of Public Health, Ohio State University, Columbus, OH, USA

Gina Moore

QCS and Regulatory Management, Alfred I. duPont Hospital for Children, Wilmington, DE, USA

Tammy E. Ott

Department of Psychology, Texas Tech University, TX, USA

Patrick A. Palmieri

Duke University School of Nursing and Duke Health Technology Solutions, Duke University, USA

Karen A. Paschal

Creighton Health Service Research Program, School of Pharmacy and Health Professions, Creighton University, USA

Lori T. Peterson

Jerry S. Rawls College of Business, Texas Tech University, TX, USA

Karlene H. Roberts

Walter A. Haas School of Business, University of California, Berkeley, CA, USA

Louis Rubino

Department of Health Sciences, California State University, Northridge, CA, USA

Ann M. Rule

Medical Liaison Alliance, Outreach Purdue Pharma, L. P. Stamford, CT, USA

Cynthia K. Russell

Acute & Chronic Care Department, College of Nursing, University of Tennessee Health Science Center, USA

Grant T. Savage

Health Management and Informatics Department School of Medicine, University of Missouri, USA

xii

LIST OF CONTRIBUTORS

Mark V. Siracuse

Creighton Health Service Research Program, School of Pharmacy and Health Professions Creighton University, USA

Michal Tamuz

Department of Preventive Medicine, College of Medicine, University of Tennessee Health Science Center, USA

Eric J. Thomas

The University of Texas Medical School at Houston, Houston, TX, USA

Daved W. van Stralen

Department of Pediatrics, School of Medicine, Loma Linda University, Loma Linda, CA, USA and Children’s Subacute Center, Community Hospital of San Bernardino, San Bernardino, CA, USA

Eric S. Williams

Culverhouse College of Commerce and Business Administration, Management and Marketing Department, University of Alabama, AL, USA

LIST OF REVIEWERS David A. Fleming University of Missouri

Nir Menachemi University of Alabama-Birmingham

Mark E. Frisse Vanderbilt University

Lori. T. Peterson Texas Tech University

Timothy R. Huerta Texas Tech University

Douglas S. Wakeﬁeld University of Missouri

Ann Scheck McAlearney Ohio State University

Eric S. Williams University of Alabama

xiii

PREFACE: SOME OBSERVATIONS ON PATIENT SAFETY The constitution of the United States refers to the obligation of the government to safeguard the people’s health and welfare. For those engaged in patient care, the health and safety of the patient is uppermost, but the safety of the patient is inﬂuenced by a multitude of factors. These factors include not only the actions of the providers of care and the patient’s environment, but also the patient’s own actions. Understandably, actions by caregivers can readily pose a risk to patient safety. When one considers the millions of contacts with patients for testing, for treatment, or nursing care, there are ample opportunities for mishaps to occur. Modern medicine increasingly relies on technology to protect and ensure the safety of patients. On one hand, telehealth makes oral, visual, and auditory communication feasible wherever the patient may be located. These applications not only protect the patient, but also contribute to the quality of care. One of the hallmarks of such applications is that they provide essential communication between patients and caregivers, whether the patient is receiving home care, ambulatory care, or intensive hospital care. For example, intensive care telehealth, implemented with success at the Lehigh Valley Health System, contributes materially in ensuring a safe environment for the patient. On the other hand, modern medical diagnostic tests, either invasive or non-invasive, have built-in modalities designed to prevent damage to the patient or for that matter, the caregiver. Injection needles and syringes are not only sterile; they are disposable and have safety features that protect against accidental injury. Most if not all medical instruments that depend on electric power contain automatic shut-off mechanisms. These mechanisms are necessarily an integral part of the instrument and are effective in preventing potentially lethal complications. Wherever health care is provided, the patient’s wellbeing is inﬂuenced by the environmental conditions that surround him or her. Again, modern health care relies on science and technology to enhance patient safety. For example, the design of medical facilities, their color, lighting, temperature, humidity, and noise all have an affect on the patient. Moreover, proper xv

xvi

PREFACE

ergonomics for furniture, such as beds, chairs, and examining tables, help minimize patient safety risks such pressure ulcers. A safety-related issue that deserves considerable attention is the patient’s own responsibility in safeguarding his or her safety. A major factor that impedes patient safety is non-compliance. Every day millions of people either take too much or too little of the medications that were prescribed, often with serious consequences. For example, patients, who receive more than one medication, and especially when the medications are prescribed by different and multiple physicians, are apt to have problems. Clear identiﬁcation and coordination is a must when patients take more than one medication. However, a well-informed patient can relate more intelligently and rationally with caregivers. This rapport makes for better coordination and cooperation and, ultimately, results in better outcomes. This review would not be complete without some remarks about the development of health information technology and its impact on medicine and, in turn, on patient safety. One outstanding feature is the implementation of electronic medical records, which has the potential to eliminate most of the errors that are made in the dispensing of medications. That feature alone, when properly and sufﬁciently implemented, will annually save thousands of lives. This special volume on patient safety and health care management refers to a multitude of steps, actions, and behavior that alone or in combination affect the health and safety of a patient. It is apparent that many of these measures are under an individual’s control, while others depend on application of tools and techniques that are primarily instrumental or mechanical. Either way attention to all these issues is not only important, but it also affords genuine protection and safeguard the safety of patients. In short, this volume illustrates that the provision of care is insured by close attention to the organizational structures and processes and the safety features of which they are an essential part. Leo van der Reis, M.D., Quincy Foundation for Medical Research – Charitable Trust and University of Missouri, USA

PATIENT SAFETY: STATE-OF-THE-ART IN HEALTH CARE MANAGEMENT AND FUTURE DIRECTIONS Eric W. Ford and Grant T. Savage ABSTRACT The needs for health system change and improved patient safety have been pointed out by policymakers, researchers, and managers for several decades. Patient safety is now widely accepted as being fundamental to all aspects of health care. The question motivating this special volume on patient safety is: How can the increased emphasis on patient safety among health care managers be more effectively translated into better policy and reduced clinical risk? The 12 contributions in this volume are divided into four sections: (1) theoretical perspectives on managing patient safety; (2) top management perspectives on patient safety; (3) health information technology (HIT) perspectives on patient safety; and (4) organizational behavior and change perspectives on patient safety. Patient safety is a topic that provides a fertile niche for management researchers to test existing theories and develop new ones. For example, the patient safety goals of reducing medical errors while maximizing health outcomes draws upon the tenets of evidence-based medicine (EBM), as well as the managerial theories of human relations, Patient Safety and Health Care Management Advances in Health Care Management, Volume 7, 1–14 Copyright r 2008 by Emerald Group Publishing Limited All rights of reproduction in any form reserved ISSN: 1474-8231/doi:10.1016/S1474-8231(08)07001-8

1

2

ERIC W. FORD AND GRANT T. SAVAGE

organizational culture, organizational development, organizational learning, organizational structure, quality improvement, and systems thinking. Indeed, these and other managerial theories are drawn upon and applied in different ways by the various contributors. Overall, the authors of this volume demonstrate that the future of patient safety for health care management requires health care professionals and managers who can successfully engage in multi-faceted projects that are socially and technically complex.

Patient safety is now widely accepted as being fundamental to all aspects of health care. Therefore, professionals in the policy and administration ﬁelds need to be even more vigilant to the potential risks in delivering clinical care. It is important for policymakers, administrators, and clinicians to understand a wide variety of system features that must be correctly aligned to ensure successful patient management. Further, developing common mental models among these key stakeholders will facilitate both the sharing of information and the aligning of incentives with desired outcomes. The overarching question motivating this special volume on patient safety is: How can the increased emphasis on patient safety among health care managers be more effectively translated into better policy and reduced clinical risk? Health care managers are at the fulcrum of balancing policy imperatives and practice practicalities. Therefore, conducting research from the managerial perspective that looks in both directions is necessary. Further, research teams drawn from a wide variety of disciplines should ensure frequent assessment and analysis of their ongoing theoretical and empirical work in relationship to system features that harm patients or create the potential for patient harm. Health management theories and methods will vary depending upon the researchers’ areas of expertise and the clinical practice or policy issues being studied. This special volume presents a collection of health care management articles that look at patient safety change efforts ranging from board room strategies (e.g., Rubino & Chan and Culbertson & Hughes) to discreet patient exchanges on the ﬂoor (e.g., Deutsch et al., 2008). In addition, we assess the state-of-the-art and future directions for health care management theory and research on patient safety. We believe the primary beneﬁt to a special volume is that it goes beyond the sum of its parts. Further, synergistic beneﬁts arise from the questions that are created from papers presented in juxtaposition to one another. Therefore, we frame and pose an

Patient Safety

3

initial set of questions for health care management practitioners and researchers to consider as they move forward in making health systems both safer and more effective.

ASSESSING THE STATE-OF-THE-ART IN PATIENT SAFETY RESEARCH The needs for health system change and improved patient safety have been highlighted by policymakers, researchers, and managers. As early as 1991, the Institute of Medicine (IOM) was calling for the universal adoption of electronic medical records (EMRs) to control cost and provide actionable data for quality improvement (Institute of Medicine, 1991). However, these alarms went largely unnoticed by policymakers and the public until the publication of To err is human: Building a safer health system (Kohn, Corrigan, & Donaldson, 1999) and its widespread dissemination of research estimating that 44,000–98,000 avoidable fatalities occurred annually in U.S. hospitals. Since this seminal report, researchers from numerous disciplines have brought their theories, methods, and experiences to bear on the causes of poor quality in health care.

Theories, Models, and Methods for Managing Patient Safety As an emerging ﬁeld of research, the relatively recent recognition and interest in patient safety provides a fertile niche for management researchers to test existing theories and develop new ones. An indication of the emerging nature of patient safety is the makeup of the organizations that have formed to address the topic. While the Institute for Health Improvement (www.ihi.org) began relatively early, other organizations such as those at Johns Hopkins, the Center for Innovation and Quality in Patient Care (www.hopkinsquality.com) and Vanderbilt’s Center for Better Health (www.mc.vanderbilt.edu/vcbh/) did not begin until 2002 and 2001, respectively. Each of these organizations has a clear agenda related to policy and clinical practice, but it is difﬁcult to discern any discreet programs targeting the improvement of management practices. Taken together, the papers in this volume of Advances in Health Care Management are intended to aid management researchers in formulating both individual and collaborative agendas to address the issue of patient safety.

4

ERIC W. FORD AND GRANT T. SAVAGE

The ﬁrst chapter in this volume discusses evidence-based medicine (EBM), its application to patient safety, and its limits for improving safety in health service delivery (Savage and Williams). Many initiatives to improve patient safety are based on EBM, without recognition of either its key assumptions, or its inherent limits for improving patient safety. Savage and Williams address three research questions: (1) How does EBM contribute to patient safety? (2) How and why is EBM limited in improving patient safety? and (3) How can patient safety be maximized, given the limitations of EBM? Currently, EBM contributes to patient safety both by educating clinicians on the value and use of empirical evidence for medical practice and via large-scale initiatives to improve care processes. Attempts to apply EBM to individual patient care are limited, in part, because EBM relies on biostatisical and epidemiological reasoning to assess whether a screening, diagnostic, or treatment process produces desired health outcomes for a population – not for an individual. Health care processes that are most amenable to EBM are those that can be standardized or routinized; nonroutine processes, such as diagnosing and treating a person with both acute and chronic co-morbidities are cases where EBM has limited applicability. A ﬁrst step in bridging the gap between EBM and management research is the development of models that help to identify how to ﬁt such work into larger organizational frameworks. For example, to improve patient safety, health care organizations should not rely solely on EBM, but also recognize the need to foster mindfulness within the medical professions and develop patient centric organizational systems and cultures. The second manuscript presents a model designed to help managers think about the nature of errors (Palmieri et al., 2008). The ‘Swiss Cheese’ model promotes a systems thinking approach to identifying the multiple causes that underlie the worst errors in health care. In particular, the role of organizational-level policies in retarding and promoting medical errors is critical. In many health care settings the role of policies in outcomes are often ignored because the organizational levels between the top managers and patients is populated with professionals that assume full responsibility for the patient’s care. Therefore, when a failure or error occurs, the individuals charged with executing untenable or impracticable policies are blamed. The greatest promise of the ‘Swiss Cheese’ model is that it will diminish these types of attribution errors – blaming an individual for a systemic failure – that is the current mode of operation in many health care organizations. Nurses in particular bear the brunt of fundamental attribution error because they work at the end point of the patient care system. To the extent

5

Patient Safety

that it is possible to mitigate ineffective or poor policies, front-line nurse managers take on this task. The separation of policy input, operating authority, and outcome responsibility places these professionals in untenable situations on a daily basis. The third article by Tamuz, Russell, and Thomas describes this phenomenon using a series of case studies. Drawing on interviews with 20 nurse managers from three tertiary care hospitals, their study identiﬁes key exemplars that illustrate how managers monitor nursing errors. The exemplars examine how nurse managers: (1) sent mixed messages to staff nurses about incident reporting, (2) kept two sets of books for recording errors, and (3) developed routines for classifying potentially harmful errors into non-reportable categories. These exemplars highlight two tensions: ﬁrst, applying bureaucratic rule-based standards to professional tasks; and, second, maintaining accountability for errors while also learning from them. These fundamental tensions inﬂuence organizational learning and suggest theoretical and practical research questions. While nurse managers are engaged in important forms of organizational learning to improve patient safety, they cannot address the core issue that many problems have their roots at the other end of the organizational chart – the boardroom.

Top Management Perspectives on Patient Safety The commitments necessary to build high reliability health organizations that are safe take many forms. For example, the requisite ﬁnancial investment to bring the latest information technology into the system is typically a signiﬁcant percentage of an organization’s budget. Another common problem is that facilities are often designed in ways that make workﬂows inefﬁcient and ineffective and they need to be remodeled or replaced. Because the costs of changes to an organization’s physical plant are so high, such decisions invariably require the approval of the board of directors. However, simply committing money is not enough. The biggest challenge that health care organization boards face is changing their own cultures and those of their organizations to put safety at the forefront of the care agenda. The case study by Rubino and Chan details how the Board of Directors at St. Francis Medical Center took on the task of improving patient safety. They provide a set of tools that other boards can adapt to their institutions in order to pursue similar goals. For example, they use a ‘Balanced Scorecard’ approach that is familiar to many hospital administrators and

6

ERIC W. FORD AND GRANT T. SAVAGE

board members. The use of scorecards reduces an important barrier to adoption by allowing board members to fold their patient safety efforts into existing quality assurance and improvement activities in the hospital. While such incremental steps are valuable, they do not provide a holistic theoretical framework to work from in instituting change or address the role of medical professionals in changing the delivery of care. The involvement of hospital boards in change processes is critical and there is growing pressure to hold board members accountable for organizational outcomes – including those of a clinical nature. Herein lays a dilemma, as physicians are the profession with the authority and responsibility for directing patient care. To that end, most hospitals rely on physicians to ensure the quality of care delivered. The article by Culbertson and Hughes considers this problem using the organizational structure theory ﬁrst put forth by Mintzberg (1979), which views the hospital as a case of a professional bureaucracy. Physicians, as professional staff, are thus responsible for standard setting and regulation. However, trustees are now asked to examine reports identifying physician compliance in attaining safety standards without education in the practice supporting those standards. Physician board members, whose numbers have increased in the past decade, are often sought to take the lead on interpretation of patient safety standards and results. The very public nature of patient safety reporting and its reﬂection on the reputation of the organization for which the trustee is ultimately accountable create a new level of tension and workload that challenges the dominant voluntary model of trusteeship in the U.S. health system. Culbertson and Hughes offer some advice about how board composition and duties might be conﬁgured to include physicians, but not exclude them from other policymaking activities. The roles of nurses, physicians, and boards are undoubtedly critical foci in any fully formed theory related to patient safety; however, they do not, as individual areas of research or even in combination, address the systemic nature of the problem. Health care delivery takes place in a community context. Norms and standards of care have signiﬁcant regional components that go beyond individual institutions. Many health services purchasers hope to activate these local market forces using tools such as pay-forperformance (P4P) to help improve care and control costs. The nursing shortage has already empowered that professional to activate market forces and negotiate better wage packages. Further nurses are asserting their own professional autonomy to redeﬁne inter-professional relationships with physicians and managers. The power of community action can yield remarkable changes in the way people within organizations behave.

Patient Safety

7

The last article in the Top Management Team section provides an example of a regional effort to change the way care is delivered. The Houston–Galveston region created an aggressive approach to this issue by forming an unusual coalition of business, university, and hospital leaders and using a quality-improvement approach. Batten, Goodman, and Distefano’s ﬁndings indicate that shifting the focus away from individual employee behaviors to meaningful management change had a far more profound effect that stretched across an entire community. The project has achieved over 40% participation among hospitals in the 13-county region, and it includes 50 hospitals employing approximately 15,000 registered nurses. The data that have been collected by this collaborative to date suggest that hospitals are taking action to improve outcomes by modifying their key initiatives to address the attributed causes of poor work environments. From 2004 to 2005, executives of top performing hospitals increasingly attributed successful work environment outcomes to an emphasis on management development and executive-driven initiatives, de-emphasizing speciﬁc employee behavior, process, and outcome-based initiatives. The admonition to physicians to ‘heal themselves’ may be one that health care administrators ought to take to heart when launching efforts to improve patient safety. Administrators need develop management systems that do more than create policies and track claims data. Managers need to synthesize information into actionable forms that can be used to redesign work processes at levels that cannot be readily changed through written policies.

Health Information Technology Perspectives on Patient Safety Health Information Technology (HIT) has been held out as a ‘silver bullet’ solution to all of the problems that make the U.S. system suffer from lapses in patient safety and cause medical errors to occur. In particular, mandatory universal adoption of EMRs has been suggested as a means to document encounters, coordinate care among providers, monitor compliance with clinical guidelines, and provide decision support to physicians. These are ambitious goals and the EMR products currently available in the marketplace are a quantum leap away from achieving the level of functionality necessary to realize those goals. Further, even if a particular system possesses such features, it is unclear whether it would be capable of effectively interfacing with another manufacturer’s system.

8

ERIC W. FORD AND GRANT T. SAVAGE

In recent years there have been signiﬁcant efforts to make systems interoperable and the U.S. Government formed the Certiﬁcation Commission for Healthcare Information Technology (CCHIT) to promote and coordinate this goal (Classen, Avery, & Bates, 2007). Beginning in 2006, several products were certiﬁed as being interoperable to the HL7 standard. However, even among those products, it is unclear that it is possible to effectively match patients and share medical information from one to the next. Patient identiﬁcation is particularly important for two reasons. First, it is the clinical imperative to correctly identify the person being treated. In emergency situations, having incorrect medical information such as blood type, allergies, immunizations, and pre-existing conditions can be far worse than having no information at all. The second issue is the consumer’s right to privacy. Fundamental questions around who owns and controls medical information are potentially more problematic than the technological challenges of interoperability and accurate individual identiﬁcation. Coordinating the safeguarding and sharing of health information is a governance issue. Frequently, a local market’s competitive dynamics make it nearly impossible for organizations to harmonize their policies and procedures in a way that allows for meaningful interchange. A nationwide effort to assess state and regional stakeholders’ views on the issues surrounding health information and privacy was begun in 2005 under the auspices of the Department of Health and Human Services. One goal of the program was to identify current legal and regulatory standards that needed to be harmonized so that patient information sharing within and across communities could occur without the fear of violating laws. Galt and her colleagues examine how the process played out in the state of Nebraska. They conducted an in-depth case study to explore the knowledge, understanding, and awareness of 25 health board/facility oversight managers and 20 health professional association directors about privacy and security issues important to achieving Health Information Exchange (HIE). The case analysis revealed that health board/facility oversight managers were unaware of key elements of the federal agenda; their concerns about privacy encompassed broad deﬁnitions both of what constituted a ‘health record’ and ‘regulations centeredness.’ Alternatively, health professional association leaders were keenly aware of national initiatives. Despite concerns about HIE, they supported information exchange believing that patient care quality and safety would improve. The analysis also revealed a perceptual disconnect between board/facility oversight managers and professional association leaders; however, both favored HIE.

Patient Safety

9

Licensure and facility boards at the state level are likely to have a major role in the assurance of patient protections through facility oversight and provider behavior. Similarly, professional associations are the major vehicles for post-graduate education of practicing health professionals. Their engagement is essential to maintaining health professions knowledge. States will need to understand and engage both of these key stakeholders to make substantial progress in moving the HIE agenda forward. In addition to the challenges in these efforts, one theme that emerges is the large number and transient nature of the umbrella organizations that are charged with conducting these projects. The list of acronyms labeling the organizations charged with solving the Gordian Knot-like dilemma of HIT interoperability and exchange is now legion. The American Health Information Community (AHIC) has already come and gone. Its successor, AHIC.2 – a private–public partnership – is still in the formative stages. The contract with the Research Triangle Institute (RTI) to manage the Health Information Security and Privacy Collaboration (HISPC) is near its end – but what has come of it? Will it merely be another set of recommendations that no one has the authority or wherewithal to implement? Perhaps, the most telling story is that of the Santa Barbara County Data Exchange (Brailer, 2007; Frohlich, Karp, Smith, & Sujansky, 2007; Holmquest, 2007; Miller & Miller, 2007). Although it is not chronicled in this volume, its legacy is inescapable for those hoping to build a national system for sharing patient information. As one of the ﬁrst and highest proﬁle Regional Health Information Exchanges (RHIOs), the Santa Barbara organization was held out as a model for other communities. Its leader, Dr. David Brailer, became the ﬁrst National Coordinator for Health Information Technology (ONCHIT). In the end, its peer-to-peer model collapsed because organizations could not interface their data systems and there was no sustainable business plan in place. One common element to all of the large-scale efforts has been the underlying philosophy that ‘if you build it, organizations will come.’ The separation of costs and beneﬁts can be profound in information exchanges with the providers and their parent organizations bearing the cost, while the beneﬁts accrue to others (Menachemi & Brooks, 2006). The impact of this misalignment is most keenly felt in smaller organizations that face all the ﬁxed costs of adoption, but lack the economies of scale to make the ﬁnancing work in their favor. Physicians in small practices are at the pointy-end of the HIT adoption stick (Ford, Menachemi, & Phillips, 2006). As much as any other group of

10

ERIC W. FORD AND GRANT T. SAVAGE

professionals, it is physicians’ workﬂows and pocketbooks that are likely to feel the major impact of moving to EMR systems. The paper by Bramble and his colleagues describe how physicians characterize these barriers to HIT implementation. Content analysis of qualitative interviews revealed three barrier themes: time, technology, and environment. Interviews also revealed two other major concerns; speciﬁcally, the compatibility of the HIT with the physician’s patient mix and the physician’s own attitude toward the use of HIT. The axiom that ‘time is money’ and that reduced productivity is a major concern among physicians when they consider adopting HIT are well illustrated. When the reward systems are not aligned with the policy goals of HIT advocates, the reward system will dominate decision making and the status quo will prevail. Changing individuals and organizations’ behaviors is one of the most difﬁcult tasks that managers face and lies at the heart of many patient safety programs.

Organizational Behavior and Change Perspectives on Patient Safety Efforts to change the organizations’ structures, process, and cultures to improve patient safety related outcomes are proliferating rapidly. Health organizations are drawing on LEAN (Manos, Sattler, & Alukal, 2006) and Six Sigma (Lazarus & Novicoff, 2004; Revere & Black, 2003) manufacturing principles to redesign care delivery processes. New facilities are being built to create healing and family-centered environments (Fottler, Ford, Roberts, Ford, & Spears, 2000; Towill & Christopher, 2005). While some companies are achieving amazing results, three out of four reengineering programs fail (Manganelli & Klein, 1994). Those leading reengineer efforts often make bold promises to transform organizations, but the hard part is taking the theory and putting it into practice. To change conventional thinking and traditional practices pits managers against the status quo. To overcome these barriers, not only is senior management sponsorship essential, the leadership itself needs to be exposed to external change agents and ideas to reshape their views. The article by McAlearney describes a new mechanism to allow organization leaders to gather new ideas and skills – the corporate university. As McAlearney points out, in other industries, the rise of corporate universities has been steady over the past 20 years (Anonymous, 2005). The corporate university is organized to allow health leaders to stay at their current job while gathering the skills and knowledge to implement innovations. Aside from the human resources motivation for wanting to

Patient Safety

11

develop and retain talent, corporate universities allow innovation and change to originate from within the organization. The hope is that the corporate university will foster a culture of change as a shared value among leaders. In health care organizations, OD programs can serve an important institutional function by providing a framework through which patient safety can be emphasized as an organizational priority, and patient safety training can be delivered as part of OD efforts. In addition, organizations committed to creating a patient-focused safety culture can use OD initiatives strategically to support organizational culture change efforts. McAlearney’s paper describes different approaches to including patient safety in an OD framework, drawing from both management theory and practice. Findings from three extensive qualitative studies of leadership development and corporate universities in health care provide speciﬁc examples of how health care organizations discuss patient safety improvement using this alternative approach. The article by Deutsch and her colleagues discusses a common operating procedure that is anything but standardized in most health care organizations – the patient handoff. A ‘handoff ’ occurs so that patient-speciﬁc medical information can be provided to the medical professional(s) assuming responsibility for that patient. Providing an appropriate summary supports safe, high quality, effective medical care; inadequate or incorrect information may create risk for the patient. A handoff approach was developed to facilitate this process, using the mnemonic START (S: situation; T: therapies; A: anticipated course; R: reconciliation; T: transfer). Surveys of handoffs occurring before and after introduction of the START system demonstrate that there are several areas with potential for process improvement. Contrasted with the McAlearney paper, which deals with the macro-level of organizational change, Deutsch et al.’s paper takes us to the most ﬁnite of organizational activities. Nevertheless, the lack of common techniques for ensuring that vital patient information is effectively communicated from one caregiver to the next is a weak link in the system that often breaks. Indeed, the results of such breaks create the ‘holes’ in the continuity of care processes described by Palmieri and his colleagues in the ﬁrst section of this volume. Nonetheless, there may not be one best way to conduct every handoff, since the intensive care unit (ICU) has different needs than subacute care facilities (SCFs). As van Stralen and his colleagues note, there often are not common standards within the same unit at many such facilities. Speciﬁcally, this last chapter in this compendium describes how a nursing home was transformed to a pediatric SCF. The transformation entailed not only

12

ERIC W. FORD AND GRANT T. SAVAGE

making information ﬂows more effective, but also empowering the personnel to make the SCF a high reliability organization (HRO). To obtain these goals, the health care team implemented change in four behavioral areas: (1) risk awareness and acknowledgement; (2) deﬁning care; (3) thinking and making decisions; and (4) information ﬂow. The team focused on ﬁve reliability enhancement issues that emerged from previous research on banking institutions: (1) process auditing; (2) the reward system; (3) quality degradation; (4) risk awareness and acknowledgement; and (5) command and control. Three additional HRO processes also emerged: high trust and building a high reliability culture based on values and on beliefs. The case demonstrates that HRO processes can reduce costs, improve safety, and aid in developing new markets. Key to van Stralen et al.’s ﬁndings was that every organization must tailor its processes to ﬁt its own situation. Further, organizations need managers with the skills and training to be ﬂexible and adapt to in applying HRO principles.

CONCLUDING COMMENTS: THE FUTURE OF PATIENT SAFETY Much of the rhetoric promoting patient safety in the U.S. has focused on technical solutions – such as computer-assisted physician order entry (CPOE) for prescriptions – and over-simpliﬁes the challenges facing health care managers and professionals. The authors of this volume – all of who are from the U.S. – demonstrate that the future of patient safety for health care management requires health care professionals and managers who can successfully engage in multi-faceted projects that are socially and technically complex. These challenges, more often than not, involve changing the social structures and cultures of health care organizations. Improvements in patient safety, thus, require long-term commitments from health care managers and professionals, as well as competencies in managing complexity. What should health care management researchers do to improve patient safety? Two paths are suggested based on the work published in this volume. First, we believe that health care management researchers can help clinical and managerial practitioners to improve patient safety by engaging in (a) multi-level of analysis research that evaluates organizational change efforts; (b) institutional-level research on inter-organizational and public–private collaborations; and (c) social-technical system evaluations of HIT and other

13

Patient Safety

technical implementations. These three general areas of research explore the complex and multi-faceted nature of health care organizations, and will better ensure that research ﬁndings inform both policymakers and health care managers and professionals. Second, and lastly, health care management researchers should look beyond the U.S. and its organizational and institutional landscape to investigate ways to improve patient safety. Efforts underway in Europe, Asia, and other countries represent naturally occurring experiments in patient safety. International comparative research holds the promise of illuminating new facets on the complex challenges of improving patient safety, while hastening the dissemination of best practices throughout the world.

REFERENCES Anonymous. (2005). The corporate university: Riding the third wave. Development and Learning in Organizations, 19(6), 16. Brailer, D. J. (2007). From Santa Barbara to Washington: A person’s and a nation’s journey toward portable health information. Health Affairs, 26(5), w581–w588. Classen, D. C., Avery, A. J., & Bates, D. W. (2007). Evaluation and certiﬁcation of computerized provider order entry systems. Journal of the American Medical Informatics Association, 14(1), 48–55. Ford, E. W., Menachemi, N., & Phillips, M. T. (2006). Predicting the adoption of electronic health records by physicians: When will healthcare be paperless? Journal of the American Medical Informatics Association, 13(1), 106–112. Fottler, M. D., Ford, R. C., Roberts, V., Ford, E. W., & Spears, J. D. (2000). Creating a healing environment: The importance of the service setting in the new consumer-oriented healthcare system/practitioner application. Journal of Healthcare Management, 45(2), 91. Frohlich, J., Karp, S., Smith, M. D., & Sujansky, W. (2007). Retrospective: Lessons learned from the Santa Barbara project and their implications for health information exchange. Health Affairs, 26(5), w589–w591. Holmquest, D. L. (2007). Another lesson from Santa Barbara. Health Affairs, 26(5), w592–w594. Institute of Medicine. (1991). The computer-based patient record: An essential technology for healthcare. Washington, DC: National Academy Press. Kohn, L. T., Corrigan, J. M., & Donaldson, M. S. (Eds). (1999). To err is human: Building a safer health system. Washington, DC: National Academy Press. Lazarus, I. R., & Novicoff, W. M. (2004). Six Sigma enters the healthcare mainstream. Managed Healthcare Executive, 14(1), 26. Manganelli, R. L., & Klein, M. M. (1994). A framework for reengineering (Part I). Management Review, 83(6), 10. Manos, A., Sattler, M., & Alukal, G. (2006). Make healthcare lean. Quality Progress, 39(7), 24. Menachemi, N., & Brooks, R. G. (2006). Reviewing the beneﬁts and costs of electronic health records and associated patient safety technologies. Journal of Medical Systems, 30(3), 159–168.

14

ERIC W. FORD AND GRANT T. SAVAGE

Miller, R. H., & Miller, B. S. (2007). The Santa Barbara County Care Data Exchange: What happened? Health Affairs, 26(5), w568–w580. Mintzberg, H. (1979). The structuring of organizations. Englewood Cliffs: Prentice-Hall. Revere, L., & Black, K. (2003). Integrating six sigma with total quality management: A case example for measuring medication errors. Journal of Healthcare Management, 48(6), 377. Towill, D. R., & Christopher, M. (2005). An evolutionary approach to the architecture of effective healthcare delivery systems. Journal of Health Organization and Management, 19(2), 130.

EVIDENCE-BASED MEDICINE AND PATIENT SAFETY: LIMITATIONS AND IMPLICATIONS Grant T. Savage and Eric S. Williams ABSTRACT A fundamental assumption by the Institute of Medicine (IOM) is that evidence-based medicine (EBM) improves the effectiveness of medical diagnosis and treatment and, thus, the safety of patients. However, EBM remains controversial, especially its links to patient safety. This chapter addresses three research questions: (1) How does EBM contribute to patient safety? (2) How and why is EBM limited in improving patient safety? and (3) How can patient safety be maximized, given the limitations of EBM? Currently, EMB contributes to patient safety both by educating clinicians on the value and use of empirical evidence for medical practice and via large-scale initiatives to improve care processes. Attempts to apply EBM to individual patient care are limited, in part, because EMB relies on biostatisical and epidemiological reasoning to assess whether a screening, diagnostic, or treatment process produces desired health outcomes for a general population. Health care processes that are most amenable to EBM are those that can be standardized or routinized; non-routine processes, such as diagnosing and treating a person with both acute and chronic co-morbidities, are cases where EBM has limited applicability. To improve patient safety, health care Patient Safety and Health Care Management Advances in Health Care Management, Volume 7, 17–31 Copyright r 2008 by Emerald Group Publishing Limited All rights of reproduction in any form reserved ISSN: 1474-8231/doi:10.1016/S1474-8231(08)07002-X

17

18

GRANT T. SAVAGE AND ERIC S. WILLIAMS

organizations should not rely solely on EBM, but also recognize the need to foster mindfulness within the medical professions and develop patientcentric organizational systems and cultures.

The Institute of Medicine (IOM) report, To Err is Human, made the U.S. public both aware and alarmed that medical errors are estimated to kill between 44,000 and 98,000 American each year (Kohn, Corrigan, & Donaldson, 1999). The follow up report, Crossing the Quality Chasm, advocated reforming the entire health care system – including the organization, delivery, and ﬁnancing of care – to improve quality while containing costs (Institute of Medicine, 2001a). These reports have generated a wave of funding, research, and organizational changes, including a renewed emphasis on a variety of approaches to improve both patient safety and the quality of care. A fundamental assumption by the IOM, as well as many other health care providers and health service researchers, is that evidence-based medicine (EBM) improves the effectiveness of medical diagnosis and treatment and, thus, the safety of patients (Berwick, 2002). However, EBM remains controversial, especially its connection to individual patient safety (Miles & Loughlin, 2006; Miles, Polychronis, & Grey, 2006). Hence, this chapter addresses three research questions: 1. How does EBM contribute to patient safety? 2. How and why is EBM limited in improving patient safety? 3. How can patient safety be maximized, given the limitations of EBM?

The chapter is divided into six sections. The ﬁrst section provides a deﬁnition of EBM and patient safety, and establishes the relationship between these two concepts. Section two examines the contributions of EBM to patient safety. The third section reviews the criticisms of EBM and its limitations for improving patient safety. Section four explores these limitations by examining the heterogeneity of treatment effects (HTE), uncertainty, and the tenets of statistical quality control and improvement. The ﬁfth section addresses how patient safety can be maximized given the limitations of EBM, while the sixth and concluding section focuses on the research challenges this perspective raises for health care organizations and health services researchers.

Evidence-Based Medicine

19

DEFINING EVIDENCE-BASED MEDICINE AND PATIENT SAFETY The term, evidence-based medicine (EBM), was proposed in the early 1990s by various academic physicians – ﬁrst in Canada, then the United Kingdom, and, lastly, the United States – as a new way of teaching the practice of medicine (Cohen, Stavri, & Hersh, 2004; Evidence-Based Medicine Working Group, 1992). These physicians believed that a paradigm shift was occurring, from ‘‘intuition, unsystematic clinical experience and pathophysiological rationale’’ to an emphasis on ‘‘evidence from clinical research’’ (Evidence-Based Medicine Working Group, 1992). While there are several deﬁnitions of the term, Davidoff and his colleagues (Davidoff, Haynes, Sackett, & Smith, 1995), have articulated the most comprehensive and useful one for our discussion: [E]vidence-based medicine is rooted in ﬁve linked ideas: ﬁrstly, clinical decisions should be based on the best available scientiﬁc evidence; secondly, the clinical problem – rather than habits or protocols – should determine the type of evidence to be sought; thirdly, identifying the best evidence means using epidemiological and biostatistical ways of thinking; fourthly, conclusions derived from identifying and critically appraising evidence are useful only if put into action in managing patients or making health care decisions; and, ﬁnally, performance should be constantly evaluated. (p. 1085)

Compare the above deﬁnition of EBM with the deﬁnitions of both patient safety and safe care as promulgated by the IOM. On the one hand, patient safety is assured through ‘‘[t]he prevention of harm caused by errors of commission and omission’’ (Aspden, Corrigan, Wolcott, & Erickson, 2004). On the other hand, ‘‘[s]afe care involves making evidence-based clinical decisions to maximize the health outcomes of an individual and to minimize the potential for harm. Both errors of commission and omission should be avoided’’ (Aspden et al., 2004). Clearly, the IOM deﬁnition of patient safety is deeply rooted in Western medical tradition, with the dual notions of providing beneﬁt and preventing harm echoing key passages in the Hippocratic Oath (Smith, 1996; von Staden, 1996). Also, the deﬁnition of safe care draws explicitly on EBM as the basis for both maximizing health outcomes for, and minimizing harm to, the patient. Moreover, the above deﬁnition aligns EBM with several tenets of quality improvement, including making decisions based on empirical data, using statistics to disclose variations in process outcomes, acting to improve processes, and continuous evaluation. EBM’s explicit connection to both safe care and quality

20

GRANT T. SAVAGE AND ERIC S. WILLIAMS

improvement, as we will demonstrate, has signiﬁcant implications for delimiting EBM.

CONTRIBUTIONS OF EVIDENCE-BASED MEDICINE TO PATIENT SAFETY EBM has made two large contributions to the improvement of patient safety since its introduction in the early 1990s. On one hand, EBM and its offshoots – evidence-based nursing, evidence-based practice, etc. – have become fundamental components of medical, nursing, and allied health professional education. Currently, in most academic health centers practicing Western medicine, medical, nursing, and allied health students are exposed to the best empirical evidence for diagnosis and treatment, are taught how to ﬁnd and use EBM guidelines, and are encouraged to maintain their knowledge and application of EBM once they enter practice (Gerhardt, Schoettker, Donovan, Kotagal, & Muething, 2007; McCluskey & Lovarini, 2005; McConnell, Lekan, Hebert, & Leatherwood, 2007; NailChiwetalu & Ratner, 2006; Sinclair, 2004; Slawson & Shaughnessy, 2005; Stone & Rowles, 2007; Wanvarie et al., 2006; Weberschock et al., 2005). While concerns remain about the effectiveness of the dissemination of EBM practice guidelines and the quality of the evidence for those guidelines, there has been a clear shift toward the use of EBM (Grimshaw, 2004a, 2004b, 2006). On the other hand, EBM informs many of the pay-for-performance and other projects that attempt to improve health care quality in hospitals, medical groups, and nursing homes. Leading examples in the U.S. include Medicare’s Hospital Compare (http://www.hospitalcompare.hhs.gov/) and the 5 Million Lives Campaign (http://www.ihi.org/IHI/Programs/Campaign/) sponsored by the Institute for Health Improvement. Currently, Hospital Compare includes 21 process of care measures: eight measures related to heart attack care, four measures related to heart failure care, seven measures related to pneumonia care, and two measures relates to surgical infection prevention. All of these care measures are derived from EBM. In turn, the 5 Million Lives Campaign includes multiple EBM interventions, from preventing ventilator-associated pneumonia to reducing surgical complications to preventing pressure ulcers to delivering care for acute myocardial infarction and congestive heart failure.

Evidence-Based Medicine

21

CRITICISMS OF EVIDENCE-BASED MEDICINE AND ITS LIMITATIONS FOR PATIENT SAFETY Criticisms of EBM include its (1) narrow deﬁnition of evidence, for example, elevation of randomized controlled trials; (2) reliance on empiricism; (3) lack of evidence of its effectiveness; (4) limited usefulness for individual patients; and (5) threats to physician autonomy and patient relations (Cohen et al., 2004). The ﬁrst two criticisms have their basis in disputes over the philosophy of science, while the last criticism is disputed on both political and ethical grounds. Pragmatically, the most negative of these criticisms is the lack of evidence for EBM’s effectiveness and its limited usefulness for individual patients. Indeed, few rigorous empirical studies measure the impact of EBM on patient safety and safe care (see, however, Patkar et al., 2006; Shojania, Duncan, McDonald, Wachter, & Markowitz, 2001). Thus, its effectiveness in reducing errors and improving health outcomes remains largely untested (Buetow, Upshur, Miles, & Loughlin, 2006). Moreover, since its conception, EBM has been faulted for its provider-centric focus (Bensing, 2000) and its limited beneﬁt for the individual patient (Marshall, 2006).

HETEROGENEITY OF TREATMENT EFFECTS, UNCERTAINTY, AND QUALITY IMPROVEMENT The arguments of Kravitz and his colleagues regarding the HTE provide, perhaps, the most constructive way to understand why EBM has had limited beneﬁt to individual patients (Kravitz, Duan, & Braslow, 2004). At the same time, the uncertainty associated with individual patient differences creates a common-cause source of variance that limits the applicability of EBM. From a quality improvement perspective, EBM is least applicable to nonroutine health care processes. These points are explicated in the three parts to this section of the chapter. Heterogeneity of Treatment Effects The HTE occurs when the same treatment to different patients within a population generates various outcomes. In statistical terms, HTE ‘‘is the magnitude of the variation of individual treatment effects across a population’’ (Kravitz et al., 2004, p. 664). Four dimensions of treatmenteffect heterogeneity contribute to how an individual may respond to medical

22

GRANT T. SAVAGE AND ERIC S. WILLIAMS

treatment: (1) risk of disease without treatment; (2) responsiveness to treatment; (3) vulnerability to adverse effects; and (4) utility or preferences for different outcomes. Variations in one or more of these dimensions create treatment-effect heterogeneity. HTE has important implications for clinical trials, which may be characterized as controlled experiments that test a population sample. Unless extraordinary efforts are made to test a random sample of the population stratiﬁed into statistically meaningful subgroups, a clinical trial typically tests treatments on a single subgroup within the population. Furthermore, Kravitz and his colleagues caution that the trend toward including women, minorities, and children, as well as men, in a single trial may ‘‘do nothing but ensure that the estimates for any one subgroup are unreliable due to small numbers’’ (p. 677). Given either approach, the average treatment effect reported from a clinical trial may drastically underestimate the HTE for a treatment once it is widely disseminated into practice. ‘‘When HTE is present, the modest beneﬁt ascribed to many treatments in clinical trials can be misleading because modest average effects may reﬂect a mixture of substantial beneﬁts for some, little beneﬁt for many, and harm for a few’’ (Kravitz et al., 2004, p. 662). Uncertainty While Kravitz and his colleagues, for the most part, focus on the HTE associated with random controlled trials for drugs, their arguments highlight how EBM has to deal with four types of uncertainties: risk without treatment, responsiveness to treatment, vulnerability to adverse effects, and utility for different outcomes. From a statistical quality-control perspective (Shewhart, 1939), these uncertainties are derived from differences among patients, not from the medical treatment per se. In other words, patients, as key inputs to the care process, represent a ‘‘commoncause’’ source of variation. In contrast, EBM attempts to eliminate the ‘‘special-cause’’ variation associated with different types of screenings, diagnostic tests, and treatments, as well as the variance introduced if different protocols are used for administering the same type of screening, diagnostic test, or treatment (McNeil, 2001). Quality Improvement From a quality improvement perspective, EBM works best on care processes with low to moderate uncertainty. As uncertainty within a care

23

Evidence-Based Medicine

process increases, the utility of methods that reduce special-cause variance, such as EBM, decreases. White (2004) discusses this observation in terms of EBM and patient safety: For EBM, the challenge is not just systematizing a vast and rapidly expanding knowledge base, but it is also to support its application to situations involving complex clinical judgments. In such situations, the usefulness of standardized guidelines may be limited, especially for seriously ill patients who account for a disproportionately large share of total expenditures. (p. 863)

Moreover, the limited research conducted on the success and failures of various quality initiatives has found most successes occurring when interventions are applied to processes with greater rather than lesser certainty (Easton & Jarrel, 1998; Flynn, Schroeder, & Sadao, 1995). Drawing upon the preceding studies and related research on quality improvement, Lillrank and his colleagues (Lillrank, 2002, 2003; Lillrank & Liukko, 2004) offer a useful metaphor, the quality broom, that has clear implications for EBM and patient safety. Fig. 1 shows a traditional broom with a hard thin handle, a broader middle section linking the handle and bristles, and the bristles themselves. This metaphor characterizes the three categories of care processes and their increasing level of uncertainty: standard, routine, and non-routine.

STANDARD

ROUTINE

NONROUTINE

• identical • repetition

• similar but not identical repetition

• non-repetitive

• compliance • procedures

• selection • clinical guidelines

• deviation

• error

• interpretation • intuition • failure

Level of uncertainty QUALITY SYSTEMS

QUALITY CULTURE

Fig. 1.

The Quality Broom. Source: Lillrank and Liukko (2004, p. 44).

24

GRANT T. SAVAGE AND ERIC S. WILLIAMS

Standard care processes (the thin, hard handle) have little uncertainty, and can be repeated without signiﬁcant deviation. The administration of inactivated inﬂuenza vaccines via a ﬂu shot is an example of a standard care process. In contrast, routine types of care processes (the middle section combining the handle and bristles) are bundles of standard sub-processes intermixed with patient-based variation. Routine care processes call upon methods for (a) assessing a patient’s risk of illness without treatment; (b) specifying decision rules for generating an appropriate treatment based on the patient’s responsiveness, vulnerability to side effects, and preferences; and (c) implementing a treatment to eliminate or alleviate the patient’s illness. ‘‘The essential thing in managing routine processes is not mindless, defect-free repetition (as in standard processes), but assessment and classiﬁcation of input [i.e., patients], and selection from a ﬁnite set of alternative algorithms and actions [i.e., EBM]’’ (Lillrank & Liukko, 2004, p. 41). Lastly, some health care processes are non-routine (the bristles themselves). In many such instances, the inputs provided by the patients’ symptoms are unclear and not easily diagnosed (Lillrank, 2002, 2003; Lillrank & Liukko, 2004). In other instances, given a conﬁrmed diagnosis, the efﬁcacy of various treatments for the disease may be uncertain, with no clear understanding of the possible outcomes (McNeil, 2001). Non-routine processes, Lillrank and Liukko (2004) argue, ‘‘are best managed by indirect means, such as competence, improvement and professional values, visions and missions’’ (p. 44).

IMPLICATIONS FOR MAXIMIZING PATIENT SAFETY While EBM reduces much of the ‘‘special-cause’’ variance in the clinical decision-making processes, its impact on patient safety is limited by the ‘‘common-cause’’ variance represented by patient differences. For routine health care processes, Fig. 1 suggests that a mixture of EMB and clinical judgment is necessary to prevent the errors that result from improper diagnoses or the selection of an improper treatments or therapies. For nonroutine care processes, both clinical judgment and organizational culture come to the foreground, while EBM recedes to the background. Interestingly, sound clinical judgment is best supported by a strong culture of safety and quality – developed through health professionals’ clinical training (Bosk, 1979) and reinforced by organizational and institutional values

25

Evidence-Based Medicine

(Nieva & Sorra, 2003). Indeed, the second IOM report, Crossing the Quality Chasm, lays out a compelling vision for reorganizing the health care delivery system and improving patient safety and care quality (Institute of Medicine, 2001b). The aspects of this vision most relevant for maximizing patient safety vis-a`-vis EBM involve two key activities: (1) sharpening physicians’ abilities to operate in an increasingly complex and interdependent health system, and (2) sharpening health care organizations’ ability to support physicians’ decision-making and handling of routine and non-routine care processes.

Sharpening the Physician Sharpening the physician as an instrument of clinical judgment requires changing the models of medical education and continuing medical education (CME). The current model of medical education focuses on biomedical education and practice, and has its roots in the Flexner Report of 1910 (Flexner, 1972, c1910). Given the changes called for in the IOM report (Institute of Medicine, 2001b), two substantial changes to medical education may be in order. The ﬁrst involves multidisciplinary training. If physicians are to work effectively within integrated care teams to manage non-routine care, they must have some exposure to other health professionals during their clinical training. Physicians-in-training should work with students from nursing, OT, PT, health care management, etc., on a variety of case studies, applied projects, and/or simulations. Ideally, such multidisciplinary training should begin early in medical school and take place across the four years of medical training. Ultimately, a new model for integrated health care education may emerge that includes medicine, nursing, health care management, and various therapeutic disciplines. The second direction involves training physicians to work effectively with issues beyond the clinical practice of medicine. Given the increasingly complex, technological, and multidisciplinary nature of health care, physicians-in-training should be exposed to coursework and practical experience in such diverse topics as quality improvement tools and methods, health care management, and health informatics. Ideally, these courses should be integrated into the four-year curriculum of medical school. However, there may be a sufﬁcient amount of coursework to require an additional year of medical school, as well as post-graduate fellowships for those physicians most interested in applying the tools of quality improvement, health informatics, and management.

26

GRANT T. SAVAGE AND ERIC S. WILLIAMS

Physician education, however, does not end with medical school and residency; rather, it moves into the professionally controlled, fairly unstructured system of CME. Given the necessities of sharpening the ability of physicians to operate in a highly integrated system, it would seem necessary to revamp substantially the current system of CME. While CME might continue to be controlled by professional boards and societies, both IOM reports suggest that CME needs to be more rigorous and systematic. The model of training and retraining used by airlines has been suggested (Gaba, 2000). As applied to physicians, such training may involve periodic training and evaluation to maintain licensure or board certiﬁcation. Such a system would also help to remove physicians who become less competent through age, stress, or other conditions. We believe the key to gaining physician acceptance of such a system would be to allow the continuation of professional control over licensure and board certiﬁcation.

Sharpening the Organization To sharpen the abilities of health care organizations, the IOM report suggests a number of system changes (Institute of Medicine, 2001b). Here, the focus is on two ways of improving the organization’s capacity to support clinicians and improve patient safety: (1) creating learning organizations, and (2) developing high-reliability organizations.

Creating Learning Organizations Physicians are highly skilled and expensive workers. If they are engaged in standard care processes, they typically are being underutilized. Depending upon their degree of specialization, physicians contribute the most value to routine and, especially, non-routine care processes. The key is to provide an organizational context that supports and allows them to add the maximum amount of value. One of the elements in improving physician’s ability to manage nonroutine processes is to improve an organization’s capacity for problem solving and learning. The learning organization framework (Argyris & Schon, 1978; Senge, 1990) contains substantial insight into this process of learning at two levels (Senge, 1990). Single-loop learning involves incremental advances in existing practices. For example, most qualityimprovement programs operate by making incremental improvements in

Evidence-Based Medicine

27

existing practices. Double-loop learning emerges when organizations examine their mental models and their underlying assumptions inherent in their care processes. Such an approach to quality improvement involves radically reengineering routine care processes (Hammer & Champy, 1993). A learning organization emerges both as an organization comes to understand each type of learning and as continual learning becomes part of the organization’s culture and operations.

Developing High-Reliability Organizations In discussing the current state of high-reliability organization theory, Weick, Sutcliffe, and Obstfeld (1999) suggest that the principle factors include a ‘‘strategic prioritization of safety, careful attention to design and procedures, a limited degree of trial-and-error learning, redundancy, decentralized decision-making, continuous training often through simulation, and strong cultures that create broad vigilance for and responsiveness to potential accidents.’’ Many of these factors have found their way into health care applications. For example, redundancy has been used in surgery to avoid wrong-site surgeries. Surgeons will mark the limb or area to be operated on and this will be veriﬁed by the patient and others. Strong organizational cultures, particularly safety cultures, have also gained substantial credibility in the patient safety literature (Nieva & Sorra, 2003). However, while there are numerous success stories, Resar (2006) suggests that we need to learn ‘‘to walk before running in creating high-reliability organizations.’’

SUMMARY AND CONCLUSIONS We addressed three research questions in this chapter, ﬁrst asking, ‘‘How does EBM contribute to patient safety?’’ We showed that EMB contributes to patient safety, both by educating clinicians on the value and use of empirical evidence for medical practice and via large-scale initiatives to improve care processes. Next, we addressed the question, ‘‘How and why is EBM limited in improving patient safety?’’ While there are ﬁve basic criticisms of EBM, we focused on its pragmatic shortcomings, especially its difﬁculty in application for individual patients. On one hand, EBM reduces much of the ‘‘special-cause’’ variance in the clinical decision-making processes; on the other hand, its impact is limited by the ‘‘common-cause’’ variance represented by patient differences. Hence, EBM works best on care processes with low to moderate uncertainty.

28

GRANT T. SAVAGE AND ERIC S. WILLIAMS

As uncertainty within a care process increases, the utility of methods that reduce special-cause variance, such as EBM, decreases. Finally, we asked, ‘‘How can patient safety be maximized, given the limitations of EBM?’’ Health care processes that are most amenable to EBM are those that can be standardized or routinized; for non-routine processes, such as diagnosing and treating a person with both acute and chronic co-morbidities, EBM has limited applicability. To improve patient safety, health care organizations should not rely solely on EBM, but also recognize the need to foster mindfulness within the medical professions and to develop patient-centric systems within high reliability and learning organization cultures. The recognition that health care organizations adopting EBM should also engage in strengthening an organizational culture of patient safety highlights a potential arena for research. High-reliability health care organizations focus on developing methods to translate effective nonroutine care processes into routine care processes. This transition is akin to the research and development process. As a complement to EBM, health service researchers should focus on creating systems for assessing nonroutine care process innovations that reduce patient-based variation. Nonetheless, in health care, the diffusion of innovations is notoriously slow (Balas & Boren, 2000). Thus, an applied research opportunity is developing systems for training and diffusing both EBM and complementary medical innovations, once they are deemed to be safe and effective. While preparing this manuscript, we came to recognize that clinicians and those working to improve patient safety and care quality via EBM see most care processes in different ways. In essence, they view quality improvement through different lenses. Physicians, attuned to the medical uncertainties embodied by the individual patient, see most care processes as non-routine. In contrast, health service researchers, viewing population health through the lenses of epidemiology and biostatistics, see most care processes as routine. This work shows both to be right y and wrong.

REFERENCES Argyris, C., & Schon, D. A. (1978). Organizational learning. Reading, MA: Addison-Wesley. Aspden, P., Corrigan, J. M., Wolcott, J., & Erickson, S. M. (Eds). (2004). Patient safety: Achieving a new standard of care. Washington, DC: National Academies Press. Balas, E. A., & Boren, S. A. (2000). Managing clinical knowledge for health care improvement. Yearbook of Medical Informatics, 65–70.

Evidence-Based Medicine

29

Bensing, J. (2000). Bridging the gap. The separate worlds of evidence-based medicine and patient-centered medicine. Patient Education and Counseling, 39(1), 17–25. Berwick, D. M. (2002). A user’s manual for the IOM’s ‘Quality Chasm’ report. Health Affairs, 21(3), 80–90. Bosk, C. L. (1979). Forgive and remember: Managing medical failure. Chicago: University of Chicago Press. Buetow, S., Upshur, R., Miles, A., & Loughlin, M. (2006). Taking stock of evidence-based medicine: Opportunities for its continuing evolution. Journal of Evaluation in Clinical Practice, 12(4), 399–404. Cohen, A. M., Stavri, P. Z., & Hersh, W. R. (2004). A categorization and analysis of the criticisms of evidence-based medicine. International Journal of Medical Informatics, 73(1), 35–43. Davidoff, F., Haynes, B., Sackett, D., & Smith, R. (1995). Evidence based medicine. British Medical Journal, 310(6987), 1085–1086. Easton, G. S., & Jarrel, S. (1998). The effects of total quality management on corporate performance: An empirical investigation. Journal of Business, 71, 253–307. Evidence-Based Medicine Working Group. (1992). Evidence-based medicine: A new approach to teaching the practice of medicine. JAMA, 268(17), 2420–2425. Flexner, A. (1972, c1910). Medical education in the United States and Canada: A report to the Carnegie Foundation for the advancement of teaching. New York: Arno Press. Flynn, B. B., Schroeder, R. G., & Sadao, S. (1995). The impact of quality management practices on performance and competitive advantage. Decision Science, 11, 611–629. Gaba, D. M. (2000). Structural and organizational issues in patient safety: A comparison of health care to other high-hazard industries. California Management Review, 43(1), 83–102. Gerhardt, W. E., Schoettker, P. J., Donovan, E. F., Kotagal, U. R., & Muething, S. E. (2007). Putting evidence-based clinical practice guidelines into practice: An academic pediatric center’s experience. Joint Commission Journal on Quality and Patient Safety/Joint Commission Resources, 33(4), 226–235. Grimshaw, J. (2004a). Implementing clinical guidelines: Current evidence and future implications. Journal of Continuing Education in the Health Professions, 24(Suppl 1), S31–S37. Grimshaw, J. M. (2004b). Effectiveness and efﬁciency of guideline dissemination and implementation strategies. Health Technology Assessment, 8(6)iii-1-72. Grimshaw, J. (2006). Toward evidence-based quality improvement. Evidence (and its limitations) of the effectiveness of guideline dissemination and implementation strategies 1966–1998. Journal of General Internal Medicine, 21(Suppl 2), S14–S20. Hammer, M., & Champy, J. (1993). Reengineering the corporation : A manifesto for business revolution (1st ed.). New York, NY: HarperBusiness. Institute of Medicine. (2001a). Crossing the quality chasm: A new health system for the 21st Century. Washington, DC: National Academy Press. Institute of Medicine. (2001b). Crossing the quality chasm: A new health system for the 21st Century. Washington, DC: National Academy Press. Kohn, L. T., Corrigan, J. M., & Donaldson, M. S. (1999). To err is human: Building a safer health system. Washington, DC: National Academy Press. Kravitz, R. L., Duan, N., & Braslow, J. (2004). Evidence-based medicine, heterogeneity of treatment effects, and the trouble with averages. The Milbank Quarterly, 82(4), 661–687.

30

GRANT T. SAVAGE AND ERIC S. WILLIAMS

Lillrank, P. (2002). The broom and nonroutine processes: A metaphor for understanding variability in organizations. Knowledge and Process Management, 9(3), 143–148. Lillrank, P. (2003). The quality of standard, routine, and nonroutine processes. Organizational Studies, 34(1), 215–233. Lillrank, P., & Liukko, M. (2004). Standard, routine and non-routine processes in health care. International Journal of Health Care Quality Assurance, 17(1), 39–46. Marshall, J. C. (2006). Surgical decision-making: Integrating evidence, inference, and experience. The Surgical Clinics of North America, 86(1), 201–215, xii. McCluskey, A., & Lovarini, M. (2005). Providing education on evidence-based practice improved knowledge but did not change behaviour: A before and after study. BMC Medical Education, 5, 40. McConnell, E. S., Lekan, D., Hebert, C., & Leatherwood, L. (2007). Academic-practice partnerships to promote evidence-based practice in long-term care: Oral hygiene care practices as an exemplar. Nursing Outlook, 55(2), 95–105. McNeil, B. J. (2001). Hidden barriers to improvement in the quality of care. New England Journal of Medicine, 345(22), 1612–1620. Miles, A., & Loughlin, M. (2006). Continuing the evidence-based health care debate in 2006. The progress and price of EBM. Journal of Evaluation in Clinical Practice, 12(4), 385–398. Miles, A., Polychronis, A., & Grey, J. E. (2006). The evidence-based health care debate – 2006. Where are we now? Journal of Evaluation in Clinical Practice, 12(3), 239–247. Nail-Chiwetalu, B. J., & Ratner, N. B. (2006). Information literacy for speech-language pathologists: A key to evidence-based practice. Language, Speech, and Hearing Services in Schools, 37(3), 157–167. Nieva, V. F., & Sorra, J. (2003). Safety culture assessment: A tool for improving patient safety in healthcare organizations. Quality and Safety in Health Care, 12(Supplement II), i17–i23. Patkar, V., Hurt, C., Steele, R., Love, S., Purushotham, A., Williams, M., Thomson, R., & Fox, J. (2006). Evidence-based guidelines and decision support services: A discussion and evaluation in triple assessment of suspected breast cancer. British Journal of Cancer, 95(11), 1490–1496. Resar, R. K. (2006). Making noncatastrophic health care processes reliable: Learning to walk before running in creating high-reliability organizations. Health Services Research, 41(4 Pt 2), 1677–1689. Senge, P. (1990). The ﬁfth discipline: The art and practice of the learning organization. New York: Doubleday/Currency. Shewhart, W. A. (1939). Statistical method from the viewpoint of quality control. Washington, DC: The Graduate School of the Department of Agriculture. Shojania, K. G., Duncan, B. W., McDonald, K. M., Wachter, R. M., & Markowitz, A. J. (2001). Making health care safer: A critical analysis of patient safety practices. Evidence Report/Technology Assessment no. 43. Rockville, MD: Agency for Healthcare Research and Quality. Sinclair, S. (2004). Evidence-based medicine: A new ritual in medical teaching. British Medical Bulletin, 69(1), 179–196. Slawson, D. C., & Shaughnessy, A. F. (2005). Teaching evidence-based medicine: Should we be teaching information management instead? Academic Medicine, 80(7), 685–689.

Evidence-Based Medicine

31

Smith, D. C. (1996). The Hippocratic Oath and modern medicine. Journal of the History of Medicine and Allied Sciences, 51(4), 484–500. Stone, C., & Rowles, C. J. (2007). Nursing students can help support evidence-based practice on clinical nursing units. Journal of Nursing Management, 15(3), 367–370. von Staden, H. (1996). ‘‘In a pure and holy way’’: Personal and professional conduct in the Hippocratic Oath? Journal of the History of Medicine and Allied Sciences, 51(4), 404–437. Wanvarie, S., Sathapatayavongs, B., Sirinavin, S., Ingsathit, A., Ungkanont, A., & Sirinan, C. (2006). Evidence-based medicine in clinical curriculum. Annals of the Academy of Medicine, Singapore, 35(9), 615–618. Weberschock, T. B., Ginn, T. C., Reinhold, J., Strametz, R., Krug, D., Bergold, M., & Schulze, J. (2005). Change in knowledge and skills of year 3 undergraduates in evidence-based medicine seminars. Medical Education, 39(7), 665–671. Weick, K. E., Sutcliffe, K. M., & Obstfeld, D. (1999). Organizing for high reliability: Processes of collective mindfulness. Research in Organizational Behavior, 21, 81–123. White, W. D. (2004). Reason, rationalization, and professionalism in the era of managed care. Journal of Health Politics, Policy and Law, 29(4–5), 853–868; discussion 1005–1019.

THE ANATOMY AND PHYSIOLOGY OF ERROR IN ADVERSE HEALTH CARE EVENTS Patrick A. Palmieri, Patricia R. DeLucia, Lori T. Peterson, Tammy E. Ott and Alexia Green ABSTRACT Recent reports by the Institute of Medicine (IOM) signal a substantial yet unrealized deﬁcit in patient safety innovation and improvement. With the aim of reducing this dilemma, we provide an introductory account of clinical error resulting from poorly designed systems by reviewing the relevant health care, management, psychology, and organizational accident sciences literature. First, we discuss the concept of health care error and describe two approaches to analyze error proliferation and causation. Next, by applying transdisciplinary evidence and knowledge to health care, we detail the attributes fundamental to constructing safer health care systems as embedded components within the complex adaptive environment. Then, the Health Care Error Proliferation Model explains the sequence of events typically leading to adverse outcomes, emphasizing the role that organizational and external cultures contribute to error identiﬁcation, prevention, mitigation, and defense construction. Subsequently, we discuss the critical contribution health care leaders can make to address error as they strive to position their institution as a high Patient Safety and Health Care Management Advances in Health Care Management, Volume 7, 33–68 Copyright r 2008 by Emerald Group Publishing Limited All rights of reproduction in any form reserved ISSN: 1474-8231/doi:10.1016/S1474-8231(08)07003-1

33

34

PATRICK A. PALMIERI ET AL.

reliability organization (HRO). Finally, we conclude that the future of patient safety depends on health care leaders adopting a system philosophy of error management, investigation, mitigation, and prevention. This change is accomplished when leaders apply the basic organizational accident and health care safety principles within their respective organizations.

INTRODUCTION The Institute of Medicine (IOM) established the contemporary starting line for the national patient safety movement with the seminal report, To Err is Human: Building a Safer Health System. The IOM estimated 98,000 patients die annually in American hospitals as a result of medical errors (Kohn, Corrigan, & Donaldson, 2000). Subsequent research using additional resources indicated that the number of preventable deaths was closer to 200,000 each year (Zhan & Miller, 2003). Based on their analysis, the IOM depicted the overall state of health care as a system that frequently harms and routinely fails to deliver the appropriate standard of care (Davis et al., 2002). In the years following the IOM report, tremendous public and political pressures motivated health care organizations to identify and attempt to reduce adverse events (Berta & Baker, 2004; Wachter, 2004). Indeed, patient safety advocates point to the IOM report as ‘‘galvanizing a dramatically expanded level of conversation and concern about patient injuries in health care both in the United States and abroad’’ (Leape & Berwick, 2005, p. 2384). Patient safety arrived at the point of being ‘‘a national problem that became increasingly difﬁcult for providers to ignore’’ (Devers, Pham, & Lui, 2004, p. 103). Regulatory (e.g. CMS), accreditation (e.g. Joint Commission), and quality-improvement organizations (e.g. NCQA) develop and advocate patient safety standards primarily derived from expert panel recommendations and opinions with the majority lacking evidentiary knowledge (Agency for Healthcare Research and Quality, 2001; Institute of Medicine, 2001). Although health care organizations acknowledged the importance of To Err is Human, the majority of leaders ‘‘expressed different levels of commitment to patient safety’’ (Devers et al., 2004, p. 111) reﬂected by increasing patient morbidity and mortality related to adverse events (Institute of Medicine, 2004; Zhan & Miller, 2003). Devers and colleagues

35

Error in Health Care

characterized the impact of the patient safety movement on improving health care systems as ‘‘occurring relatively slowly and incrementally’’ (2004, p. 114). By most accounts, the patient safety movement requires transformation in the areas of error identiﬁcation, process improvement, and cultural renovation to defend patients from harm (Committee on the Future of Rural Healthcare, 2005; Institute of Medicine, 2001, 2003, 2004, 2007a, 2007b; Kohn et al., 2000).

Reducing Errors The purpose of this chapter is to describe the current state of the often unreliable health care system and to provide an introductory account of error resulting from poorly designed systems. In doing so, we review the relevant health care, management, psychology, and organizational accident science literature and synthesize an error framework from a transdisciplinary perspective. First, we discuss the concept of error and the complexities associated with errors within the complex adaptive health care system. Second, we describe two approaches to analyze error causation and discuss the associated implications. Next, we summarize the Swiss Cheese Model of adverse events (Reason, 1990, 2000), advocating modiﬁcations to emphasize both the impact of the complex adaptive system on health care professionals and the role error serves to distract leaders and clinicians from system improvements. Through this transdisciplinary approach, we present our Health Care Error Proliferation Model. This adaptation of the Swiss Cheese Model improves and updates the applicability of the general structural elements speciﬁcally to health care. Finally, we discuss the critical role health care leadership serves to proactively address error causation. Ultimately, the emphasis on error reduction and system defense strategies will lead to a minimal number of adverse events, an attribute associated with high reliability organizations (HROs) and industries.

PHYSIOLOGY OF ERROR In this section, we review two approaches to dissecting the root cause of error: the person approach and the system approach. These two approaches represent distinct philosophies of error causation which lead to divergent methods for managing liability and reducing errors.

36

PATRICK A. PALMIERI ET AL.

Person Approach The person approach advocates identifying the culpable party as the root cause of an adverse event (Reason, 1998). Historically, this health care error investigation process focuses on the ‘‘who did it’’ instead of the ‘‘why did it happen’’ (Kohn et al., 2000; Rasmussen, 1999; Reason, 2000). The person approach is commonly preferred because ‘‘blaming individuals is emotionally more satisfying than targeting institutions’’ (Reason, 2000, p. 70). A leading rationale to explicate this person focus is fundamental attribution theory. Attribution theory is a body of knowledge about how people explain things by either internal or external factors (Heider, 1958; Weiner, 1972). Generally, attribution theory describes the process whereby leaders seek: (1) to understand the cause of an event (Heider, 1958), (2) assess responsibility for the outcomes (Martinko & Gardner, 1982), and (3) appraise the personal attributes of the involved parties (Weiner, 1995b). These three attributes allow leaders to rapidly form determinations to describe employee behavior and performance (Harvey & Weary, 1985). Attribution theory explains leaders’ and managers’ thinking processes as related to conclusions about the development and causation of an event (Martinko, Douglas, & Harvey, 2006; Martinko & Thomson, 1998). Essential to attribution theory is the method by which leaders form opinions about worker’s performance (Martinko & Thomson, 1998; Weiner, 1995a). Attribution is synonymous with explaining ‘‘why did something occur’’ (Green & Mitchell, 1979) in the context of the complex adaptive environment. As such, attributing error to a particular clinician is quick, easy, and efﬁcient for leaders to formulate immediate conclusions. Although swift in generating answers, the accuracy of error attribution is troublesome within complex systems given the plethora of variables, processes, and functions (Dorner, 1996). In addition, speed matters while accuracy is less important following an accident because attributing error to the clinician protects the organization from needing to admit culpability during the mitigation phase of the risk management process. With further scrutiny, however, we discover the majority of errors leading to patient harm are not related to incompetent or substandard clinician care (Cook & Woods, 1994; Kohn et al., 2000; Reason, Carthey, & de Leval, 2001). Rather, errors more frequently reﬂect an inability of clinicians and other health care workers to cope with multiple gaps produced by system complexity (Dorner, 1996; Wiegmann & Shappell, 1999; Woods & Cook, 2002). Aviation is an industry that has achieved a remarkably low error rate. Reports attribute 80–90% of all errors to system complexity (Wiegmann &

Error in Health Care

37

Shappell, 1999). Experts hypothesize that this complexity is similar for the health care industry (Helmreich, 2000; Helmreich & Davies, 2004; Kohn et al., 2000). In other words, the root causes of error are not prominently located at individual level but tend to be a system property requiring causeand-effect analysis to elucidate ‘‘the causes’’ (Gano, 2003; Reason, 2000). Often, these system ﬂaws, including obvious and dangerous conditions, remain undiscovered, hidden, or invisible until a sentinel event results. Described simply as an ‘‘error with sad consequences’’ (Cherns, 1962), most accidents result in the course of a complex chain reaction with a triggering cause or causes. In health care, a serious accident is called a sentinel event. Sentinel events are characterized as the unanticipated adverse event resulting in an outcome or potential outcome of serious injury or death not related to a patient’s expected illness trajectory (Joint Commission Resources, 2007). By supporting the person approach, clinical professionals, health care leaders, professional boards, and even the public frequently consider practitioners involved in errors to be ‘‘at fault’’ for neglecting to protect the patient. In fact, the National Council of State Boards of Nursing (NCSBN) could further exacerbate the assignment of blame by the person approach in describing the types and sources of nursing error through an inductive process called ‘‘practice breakdowns’’ (National Council of State Boards of Nursing, 2007). One universal consequence of adopting the person approach to error management is the culture of fear it engenders (Kohn et al., 2000; Reason, 2000; Rosenthal, 1994). The term ‘‘practice breakdowns’’ is deﬁned as ‘‘the disruption or absence of any of the aspects of good practice’’ (Benner et al., 2006, p. 53). By categorizing error in terms of ‘‘practice breakdowns,’’ without speciﬁcally describing or deﬁning what the disruption to, or the absence of, good practice means, error will continue to be adjudicated by examining performance closest to the adverse event, at the clinician practice level, as separate and removed components from the greater complex adaptive system. Perrow (1984) rejected the term ‘‘operator breakdown’’ in describing an accident because this blames the worker. The IOM (Kohn et al., 2000, p. 43) emphasizes the ‘‘individual provider issue’’ rather than failures in the process of providing care in a complex system as problematic to improving health care delivery. In short, the term ‘‘practice breakdowns’’ suggests a ‘‘clinician fault’’ connotation and may slow down the improvement to health care. Furthermore, public blame for clinician error and the subsequent punishment for adverse events due to error are expected by the majority of health care leaders regardless of the root causes (Vincent, 1997). The result of this punitive culture is that doctors and nurses hide ‘‘practice

38

PATRICK A. PALMIERI ET AL.

breakdown’’ situations (Kohn et al., 2000; Lawton & Parker, 2002) and mistakes to avoid reprisal and punishment (Gibson & Singh, 2003; Kohn et al., 2000). Speaking in opposition to the person approach, Gibson and Singh (2003, p. 24) stated, ‘‘When a health care professional reports a medical error, they suffer intimidation. They lose their standing, their status and are ostracized. An impenetrable culture of silence has developedy’’ Similarly, this person approach was illustrated by a recent sentinel event at a Midwestern hospital. An experienced labor and delivery nurse mistakenly delivered an epidural drug to a young late-term pregnant patient through the intravenous line (Institute for Safe Medication Practices, 2006). Unfortunately, the patient died as a direct result of this critical sentinel event error at the sharp-end, near the bedside. However, closer analysis revealed that the fatal nursing error was probably only one in a cascade of events plagued with ‘‘hidden’’ or contributory errors. Typically observed in medication delivery systems, errors originate proximal to system level and cascade through the process distally to clinical practice (Smetzer & Cohen, 2006). Due to the collective inability to recognize the nursing error in the context of a complex adaptive system with abundant dormant conditions, the nurse was subjected to signiﬁcant public blame creating humiliation and hardship as well as criminal prosecution (Institute for Safe Medication Practices, 2006; State of Wisconsin, 2006). The nurse faced serious punishment despite the presence of other signiﬁcant contributory factors (Institute for Safe Medication Practices, 2006). Numerous professional and quality-improvement organizations reacted with position statements objecting to the miscarriage of justice (Institute for Safe Medication Practices, 2006; Wisconsin Hospital Association, 2006; Wisconsin Medical Society, 2006), which is not in the spirit of To Err is Human (Institute for Safe Medication Practices, 2006; Kohn et al., 2000; Wisconsin Medical Society, 2006). In fact, the continual application of blame, punishment, and shame to address health care error facilitates the cultural evolution of learned helplessness (Seligman, 1965; Garber & Seligman, 1980; Seligman, Maier, & Geer, 1968). Learned helplessness is an evolutionary process where clinicians become passive professionals as a consequence of repeated punishment for errors that make success unlikely even following organizational change (Abramson, Garber, & Seligman, 1980; Martinko & Gardner, 1982). Repeated punishment for clinical errors may be internalized (Garber & Seligman, 1980) as a result of practitioners frequently witnessing the adjudication of active error via ‘‘blame and shame’’ with chastisement and character assassination. Outside the organization walls, sensationalized media reports only aggravate this dilemma. As a result, clinicians may develop a lack

39

Error in Health Care

of self-conﬁdence in their abilities to perform without punishment for mistakes resulting in deterioration of performance (Peterson, Maier, & Seligman, 1993). Fashioned by this sharp system quality, the condition may be exacerbated by quasi-legal entities, such as the NCSBN, distinguishing error using the person approach and viewing error as a professional practice characteristics as opposed to a common system attribute.

System Approach The system approach stands in contrast to the person approach in the philosophy related to error and adverse event adjudication, investigation, and mitigation. According to Reason, the system approach to error management unearths concealed (latent) errors (2000) and exposes vulnerability through intensive system evaluation (Reason et al., 2001), while discounting the visible (active) errors caused by being human. The system approach advocates that while individual practitioners ‘‘must be responsible for the quality of their work, more errors will be eliminated by focusing on systems than on individuals’’ (Leape et al., 1995, p. 40). Consequently, this approach relies on investigative techniques and transdisciplinary analysis of both latent and active errors as threats to the system (Helmreich, 2000). Systems thinking and proactive process improvement on the part of health care organizations remains a signiﬁcant opportunity for advancement (Amalberti, Auroy, Berwick, & Barach, 2005). The vast majority of errors that contribute to accidents results from poorly designed systems (Cook, Render, & Woods, 2000; Rasmussen, 1990; Reason, 2000) rather than from the carelessness or neglect of professionals working within the environment (Cook & Woods, 1994; Helmreich & Davies, 2004). The IOM recognized this approach in the report, Crossing the Quality Chasm (2001, p. 4) stating, ‘‘Trying harder will not work. Changing systems of care will.’’ The responsiveness of clinicians and hospital leaders to the numerous patient safety calls have been slow to emerge (Leape & Berwick, 2005). In response to the IOM’s call for comprehensive system transformation, Millenson (2003, p. 104) identiﬁed a barrier related to system improvement, stating ‘‘the IOM’s focus on ‘system’ improvement ignores the repeated refusal by physicians and hospital leaders to adopt [better] systems’’ in their effort to improve patient safety. Most clinicians believe they are already working to improve the system when they are not. This is supported by Devers et al. (2004, p. 111) categorization of physicians as ‘‘barrier[s] failing to buy into the magnitude of the [safety] problem.’’

40

PATRICK A. PALMIERI ET AL.

Deeply embedded but masked features, or latent factors (Dorner, 1996; Perrow, 1984; Reason, 2000) are the primary leadership focus subsequent to an adverse event (Reason, 1990). Although no system can completely eliminate error (Perrow, 1984; Rasmussen, 1990), the system approach is an instrumental philosophy to tackle potential hazards aimed at reducing risk, increasing reliability, and improving quality (Kohn et al., 2000; Reason, 2000; Smetzer & Cohen, 1998). HROs are prime examples of the system approach (Reason, 2000; Roberts, 2002) put into organizational practice. These industries are considered to be high risk, such as airlines and nuclear power plants, whose system design, organizational culture, and leadership commitment to safety facilitate highly reliable, low accident, systems (Roberts, 1990; Weick & Sutcliffe, 2001). The achievement and maintenance of exceptionally low operational process variation is the key distinguishing feature among HROs (Roberts, 1990). Although health care is a high-risk industry (Helmreich & Davies, 2004; Kohn et al., 2000), by most accounts there is excessive process variation (Institute of Medicine, 2004; Kohn et al., 2000; Reason, 2000) and unacceptably large quantities of adverse events (Institute of Medicine, 2004; Kohn et al., 2000; Zhan & Miller, 2003). As such, the majority of hospitals operate as low reliability organizations. Removing ‘‘blame and shame’’ from the equation may encourage health care professionals to embrace the system approach and participate in error reporting for the expressed purpose of system improvement. For example, one study found that 41% of physicians are discouraged from or not encouraged to report medical errors (Blendon et al., 2001). Shifting to a system philosophy diminishes the appropriateness of the current physician hierarchy emphasizing the assumption of personal responsibility and accountability for failures (Helmreich, 2000). Understanding the properties of HROs, the system philosophy is the ideal approach to address error within the complex adaptive health care system. Weick and Sutcliffe (2001) describe ﬁve ‘‘mindful’’ organizational attributes summarizing the typical HRO system approach for safe and effective operations. Reliable organizations characteristically: (1) stay preoccupied with reducing failures; (2) stand reluctant to simplify errors; (3) maintain a heightened sensitivity to latent failures; (4) remain committed to learning from failures; and (5) defer decisions to experts at various levels. These elements are often cited and discussed by the IOM to call upon the health care industry to accept the system approach to error prevention and investigation (Institute of Medicine, 2003, 2004, 2007b; Kohn et al., 2000).

41

Error in Health Care

ANATOMY OF ERROR With technological advances signaling the new millennium of health care, we must concede that extraordinary damage results from ordinary errors (Chiles, 2002). Unmistakably, medical errors are a signiﬁcant reality (Kohn et al., 2000; Rasmussen, 1999; Reason, 2000) and a frequent product of poorly constructed health care delivery systems (Helmreich, 2000; Kohn et al., 2000; Reason et al., 2001). David Eddy, a nationally recognized physician patient safety expert, concisely summarized the impact modern demands have created with technological improvement and knowledge generation by stating, ‘‘the complexity of modern medicine exceeds the inherent limitations of the unaided human mind’’ (Millenson, 1997, p. 75). In an evaluation of accidents within the context of a system, Reason and Hobbs (2003, p. 39) believe errors result from ‘‘the failure of planned actions to achieve their desired goal, where this occurs without some unforeseeable or chance interventions.’’ Reason et al. (2001) describe this failure as a type of ‘‘vulnerable system syndrome’’ and emphasize early identiﬁcation, evidence-based treatment, and prevention strategies to address process and system failures. To facilitate meaningful discussion, errors ﬁrst need to be deﬁned and accurately described (Benner et al., 2006; Wiegmann & Shappell, 2003). Broadly stated, there are two types of errors – latent and active (Reason, 1990, 2000). Both latent and active errors work to disrupt systems damaging both patients and clinicians by contributing to adverse events and poor outcomes. As systems fail in a ‘‘step-by-step’’ fashion analogous to the cracking of metal under intense stress (Chiles, 2002), the evolution of a sentinel event is a system of contributory fractures. To explain this complex system of layers, fractures, and pressures, we offer the Health Care Error Proliferation Model illustrated in Fig. 1. This model, incorporates several important elements of Reason’s work, depicts the health care system segregated into defensive layers within the complex adaptive system as well as part of the global health care environment. Organization leaders are positioned at the blunt-end, while the clinician works closest to patient bedsides and resides at the sharp-end. The holes in each layer provide opportunity for error to manifest when health care professionals are unable to defend these system gaps at various organizational levels. Vigilant clinician error defenses are analogous to systematic survival techniques developed through experiential learning and reﬂection. Frequently, these defenses derive from localized workarounds, described as clinical improvisation (Hanley & Fenton, 2007), purposed to

42

PATRICK A. PALMIERI ET AL.

Important Concepts: o Holes in any layer increases vulnerability of entire system • Size of hole proportional to significance of vulnerability o Virtually impossible to eliminate all holes o Important to understand whole system versus fragments • Continuously monitor the health of whole system o Error closest to the patient is the sharpest, furthest away the bluntest

Fig. 1. Health Care Error Proliferation Model. Source: Based on Concepts from Reason (1990, 1998). Artistic and Graphic Assistance by Larry Reising.

repair gaps produced by actions, changes, and adjustments fashioned at higher defensive layers.

Latent Conditions and Errors The least visible but most frequent type of error can be described as latent (Rasmussen, 1990, 1999; Reason, 1990, 2000). Latent errors can be deﬁned as those less apparent failures of processes that are precursors to the

43

Error in Health Care

occurrence of errors or permit them to cause patient harm. Latent conditions exist distant from the delivery of patient care. For example, commonly observed latent conditions include operational glitches caused by poorly designed surgical time-out procedures, faults created by ﬂawed patient identiﬁcation policies, and inadequate resources allocation such as stafﬁng, equipment, and supplies. Problematic or latent conditions may remain dormant for extended periods of time (Reason, 1990) as errors do not manifest themselves until the right circumstances arise (Reason, 2000; Wiegmann & Shappell, 2003). Latent conditions have been characterized as situations placed in the system as part of its design or caused by actions taken by decision makers removed from the direct provision of care (Sasou & Reason, 1999). ‘‘Latent conditions are to technological organizations what resident pathogens are to the human body’’ (Reason, 1998, p. 10). Latent errors are literally ‘‘accidents waiting to happen’’ (Whittingham, 2004) in the absence of attention and treatment. As such, latent conditions are present as hidden malfunctions in organizational systems (Reason, 1990) contributing to the occurrence of adverse clinical events that are potentially harmful to patients (Rasmussen, 1990; Reason, 1998, 2000). Latent errors provide early and perhaps repetitive warnings of imminent accidents of consequence (Reason, 1998). Thus, in relation to health care systems, latent error identiﬁcation and intercession provides an important adverse event prevention strategy (Cook & O’Connor, 2005; Cook & Woods, 1994; Reason et al., 2001).

Active Conditions and Errors Active errors are actual breaches in the system defenses arising from the dormant latent conditions energized by some type of stimuli (Reason, 1990). Revealed through immediate feedback (Reason, 1998), active errors are usually related to clinical tasks manifesting out of patient care activities such as administering an intravenous antibiotic or performing a knee replacement surgery. Often, active errors result when latent strategic and operational decisions made at the highest organizational levels (Reason, 1998; Reason et al., 2001) shape a potentially dangerous setting for clinicians to chance. The active errors act like small holes in a water container. Occasionally, improvisation is utilized such that moving a single ﬁnger over a hole offers a temporary workaround and water stops leaking (Hanley & Fenton, 2007). Over time, however, more holes will likely materialize and soon the professionals can no longer mange the workaround. Although, the circumstance

44

PATRICK A. PALMIERI ET AL.

necessitating a workaround can temporarily alleviate local symptoms from a systemic failure, this localized improvisation can worsen or potentate the overall decline in a complex system. As with other industries, health care professionals can either manage uexpected events ineffectually (Weick & Sutcliffe, 2001) as if covering a hole with a ﬁnger, leading to future accidents or health care professionals can pull the system into correction adverting future debacles. Similar to other industries, sentinel-like events are virtually impossible to predict and even more difﬁcult to scientiﬁcally study (Reason & Mycielska, 1982). Even though active errors appear more common due to the immediate result, latent errors are actually more prevalent (Reason, 1998) but less distinguishable without diligent surveillance activities (Lawton & Parker, 2002), such as robust occurrence reporting, failure mode effect analysis (FMEA), or root cause analysis (RCA). Hence, focused efforts to correct active errors provide little beneﬁt to the system going forward (Rasmussen, 1990; Reason et al., 2001), especially from the future patient’s perspective. While most health care organizations ﬁnd surveillance activities difﬁcult to master, those with a sensitivity to operations and a demonstrated concern for discovering the unexpected prove successful in receiving active reporting of discrepancies (Weick & Sutcliffe, 2001). The practitioner’s perspective for error and error causation makes a difference in the robustness of surveillance. Through providing practitioners with clear guidance for detecting, discussing, and reporting error, as well as frequently soliciting feedback, system issues begin to illuminate for easier identiﬁcation. This perspective is an important aspect to identify early issues related to two situations which might arise when clinicians become aware of an error but may choose not report. The ﬁrst situation is the ‘‘error of judgment and planning’’ when clinician performance progresses as intended but the overall plan is ﬂawed leading to an error. The second situation is the ‘‘action not as planned’’ that describes poor clinical execution that occurs despite the presence of a good plan (Reason & Mycielska, 1982). When patient outcome is not impacted by either of these situations, clinicians may not report the error even though the chances of recurrence are signiﬁcant. There are several explanations suggested as to the reason why clinicians elect not to report near-errors or actual errors where patient harm was adverted. The clinician perception about what information should be reported is discussed in the literature. For example, when an active error causes injury practitioners generally believe it most appropriate to report ‘‘what happened’’ but not necessarily ‘‘how it happened’’ (Espin, Lingard, Baker, & Regehr, 2006). Surveillance activities that highlight active errors,

Error in Health Care

45

or the ‘‘what happened’’ have been favored to process analysis for the ‘‘how it happened’’ leading to forfeited improvement prospects. This active error emphasis perpetuates system volatility (Reason et al., 2001). Attention to correcting latent errors directly correlates to error reduction and systemic improvement (Institute of Medicine, 2004; Kohn et al., 2000; Reason, 1990, 2000) frequently attributed to reliable system processes (Weick & Sutcliffe, 2001).

The Proximity of Errors to the Adverse Event: Blunt and Sharp-Ends Latent and active errors are described, in part, by the system element they arise from, as well as, by their proximity to an adverse event. As illustrated in Fig. 1, latent errors usually remain distal to the adverse event site (Reason, 2000; Reason et al., 2001). Within the complex health care system, distinct failure causation, and event prevention opportunities exist at varying levels of the organization (represented in Fig. 1 as the layers of defenses). There are four defensive layers in the model. As these layers are discussed in the following section, it is important to remember that each layer may contain multiple sublayers and other complex attributes not speciﬁcally discussed in this chapter. At the ﬁrst macroscopic layer (layer 1), leaders make organizational decisions about policy, procedure, and clinical function. These decisions potentially change, create, and/or eliminate holes at different levels within the complex adaptive system. Then, the second layer (layer 2) represents the supervisory and management role in the context of localized operations. These managers direct and organize the localized operations with varied strategies. Next, layer 3 represents the zone where policy, procedural, and environmental imbalances impact clinical practice. The resulting practices may contribute to the systematic error trajectory and interactions with system defenses. Finally the last defensive layer, layer 4, is the proverbial ‘‘rubber meeting the road’’ layer where unsafe clinical acts can result or lead to adverse events. At this microscopic level, clinicians work at the sharp-end while often not realizing they are protecting the patient from the unintended consequences created by multiple failsafe breakdowns. The large triangle in Fig. 1 represents the complex adaptive health care system. Complex adaptive systems characteristically demonstrate selforganization as diverse agents interact spontaneously in nonlinear relationships (Anderson, Issel, & McDaniel, 2003; Cilliers, 1998), where professionals act as information processors (Cilliers, 1998; McDaniel &

46

PATRICK A. PALMIERI ET AL.

Driebe, 2001) and co-evolve with the environment (Casti, 1997). Health care professionals function in the system as diverse actors within the complex environment utilizing different methods to process information (Coleman, 1999) and solve systemic problems within and across organizational layers (McDaniel & Driebe, 2001). Self-organization emerges as clinicians to adjust, revise, and rearrange their behavior and practice to manage the changing internal and external environmental system demands (Anderson et al., 2003; Cilliers, 1998) utilizing experiential knowledge (McDaniel & Driebe, 2001) and improvisation (Hanley & Fenton, 2007). Health care environments supporting and accepting self-organization produce better patient outcomes (Anderson et al., 2003). As such, care delivery system selforganization attributes, represented by the triangle in Fig. 1, impact the recognition, mitigation, and prevention of latent and active errors. Next, the large square reﬂects the health care environment including the inﬂuences and forces outside the organization such as regulatory boards, consumers, payers, legislators, and others. The impact of payment systems, litigation trends, and evidence-based practice changes all contribute to external complexity that impacts individual health care organizations (Davis et al., 2002; Kohn et al., 2000). Although we describe some of these features within this chapter, the speciﬁc application of this external environment model attribute is better suited for another discussion. A particularly valuable and integral aspect of our model is the attention to both the blunt-end and the sharp-end of error causation. Latent errors tend to reside closest to the triangle’s blunt-end (left side) and represent the organizational level attributes. Errors manifesting at the clinician level develop as active errors or possibly near-events. A near-miss event is the nomenclature to describe those errors coming close to injuring a patient with harm adverted by a ‘‘last minute’’ defensive action close to the bedside. At the other end of the triangle and adjacent to the accident, active errors intimately link to an adverse event at or closest to the patient (Cook & Woods, 1994). The apex or pointed-end of the triangle resides closest to the accident. If the triangle were metal, the pointed-end would be sharp; hence, the term ‘‘sharp-end’’ is used to describe active errors. Also, the term sharpend error is metaphorically analogous to the blame and the punishment frequently exacted on the health care professionals for human error (Kohn et al., 2000; Reason, 2000). The speciﬁc relationship of the sharp-end and the blunt-end concepts to adverse events and the defensive layers will be further developed in the sections to follow. Active and latent errors are impacted and inﬂuenced, directly and indirectly, by the internal defensive layers as well as the entire complex adaptive health care system.

Error in Health Care

47

HEALTH CARE ERROR PROLIFERATION MODEL In this section, we adapt the Swiss Cheese Model (Reason, 1990) to health care organizations in which we call the Health Care Error Proliferation Model (Fig. 1). The Swiss Cheese Model, likens the complex adaptive system to multiple hole infested slices of Swiss cheese positioned side-by-side (Reason, 1990, 2000). The cheese slices are dubbed defensive layers to describe their role and function as the system location outﬁtted with features capable of intercepting and deﬂecting hazards. The layers represent discrete locations or organizational levels potentially populated with errors permitting error progression. The four layers include: (1) organizational leadership, (2) risky supervision, (3) situations for unsafe practices, and (4) unsafe performance. The Swiss Cheese Model portrays hospitals as having multiple operational defensive layers outﬁtted with essential elements necessary to maintain key defensive barricades (Cook & O’Connor, 2005; Reason, 2000). By examining the defensive layers attributes, prospective locales of failure, the etiology of accidents might be revealed (Leape et al., 1995). Experts have discussed the importance of examining these layers within the context of the complex adaptive health care system (Kohn et al., 2000; Wiegmann & Shappell, 2003) as illustrated in our Health Care Error Proliferation Model. The contribution that the complex system provides to error was suggested in a 2001 survey in which 76% of nurses indicated that their inability to deliver safe health care was attributed to impediments created by unsafe working conditions (American Nurses Association, 2006). This data reﬂects the general presence of unsafe working conditions in health care facilities across the nation. There is probably an operational disconnect at many organizational levels and between multiple disciplines created by the expansive system complexity while under intense ﬁnancial, human, and political pressures. The holes in the cheese represent actual areas where potential breakdowns or gaps in the defensive layers permit hazardous error progression through the system (Cook et al., 2000; Reason, 1990). These holes continuously change size and shift locations (Reason, 2000). When gaps at each layer sequentially align, the system fails. In other words, sentinel events occur when multiple individual, yet small faults, come togather to create the circumstances for a hole alignment sufﬁcient to produce total system failure (Cook & O’Connor, 2005). This unobstructed hole alignment occurs infrequently. In the vast majority of situations, health care clinicians control these holes and defend the system integrity (Cook et al., 2000) to

48

PATRICK A. PALMIERI ET AL.

stop error progression. Only when the practitioner is unable to anticipate, detect, and impede hazards from passing through the holes will an adverse event manifest (Reason, 1998). However, given the substantial number of activities involved in delivering patient care (Institute of Medicine, 2004), coupled with the heavy workloads of clinicians (Joint Commission on the Accreditation of Healthcare Organizations, 2007), individual holes or gaps within speciﬁc layers represent the normal consequence of complexity and operational demands. The alignment of the holes to create an adverse or sentinel event is analogous to the concept of the ‘‘perfect storm.’’ Using the concepts illustrated in Fig. 1, we discuss in the following sections the organization, leader, manager, and clinician contribution to strengthening and protecting the defensive layers along the continuum from the bluntend to the sharp-end of adverse events. When an accident manifests, the presence of latent factors may be revealed through careful and unbiased examination of the system (Farrokh, Vang, & Laskey, 2007) by methodically evaluating each defensive layer (Reason, 2000). Hazards blocked within or at any one of the ﬁrst three defensive layers is termed a latent error (Reason, 2000), while failure blocked at the fourth layer, virtually at the patient’s bedside, is termed a near-event (Cook et al., 2000; Cook & Woods, 1994; Reason, 2000). The culmination of error proliferation is heavily dependent not only on the layers of interaction but interdependent to the general culture of both the health care system internally, at the organization level, and externally, at the professional, legal, and delivery system levels.

Layer 1: Organizational Leadership The ﬁrst layer, most distant from the adverse event, is leadership at the highest organizational level. This level of defense is the most challenging to alter in health care organizations (Nolan, Resar, Haraden, & Grifﬁn, 2004) because the current ‘‘blame and shame’’ person approach is frequently derived from an attribution like process (Reason, 1998; Sasou & Reason, 1999). Leaders often ascribe responsibility for faulty processes or adverse events as resulting from the lack of effort (Reason, 2000), inability (Vincent, 2003), incompetence (Rasmussen, 1990, 1999), and absence of vigilance (Reason & Hobbs, 2003) to the clinical professional. As such, attribution for poor performance, or ‘‘practice breakdowns’’ results in actions aimed at clinicians instead of the decisions produced by leaders. Effective leadership is necessary to maintain safe systems (Joint Commission Resources, 2007). Leaders are accountable for ‘‘engineering a

Error in Health Care

49

just culture’’ as a critical aspect to providing safe patient care (Reason, 2000). The Institute of Medicine (2004) speaks to three vital aspects of a leader’s responsibility to shape their institutional culture of safety. First, leaders need to recognize and accept that the majority of errors are created by the system that they cultivate and direct (Institute of Medicine, 2004; Reason, 2000). Second, the role of leadership should include the perceptible daily support for practitioners (Institute of Medicine, 2004; Weick & Sutcliffe, 2001). Third, leaders should sincerely embrace and inculcate continuous organizational learning (Hofmann & Stetzer, 1998; Institute of Medicine, 2004). In combination, these three IOM recommended qualities not only create efﬁcient organizations but they also help reduce latent and active errors. Accordingly, the Institute of Medicine (2004) suggests several focus areas for strengthening the organization. These measures include: adopting evidence-based and transformational leadership practices, increasing workforce capability, enhancing workplace design to be error resistant, and creating a sustainable safety culture as necessary elements for vigorous safety defenses. Developing a robust error reporting system or incorporating prospective process reviews by quality teams are just two examples of leaders positively inﬂuencing organizational acceptance of a patient safety culture. In order to better understand the behavior of individuals, leaders must attempt to determine what subordinates are thinking about situations (Pfeffer, 1977). Green and Mitchell (1979) developed a model to study attributional leader traits. Leader behavior is a consequence of interpretation relevant to subordinate performance. In attempting to understand how subordinate performance affects leader reaction, it is vital to determine the cause of subordinate performance whether good or poor. There are four causes that may or may not be within the subordinate’s control: competence, effort, chance, and other uncontrollable external causes. Both competence and effort are internal causes of performance, while chance and other uncontrollable causes are outside the subordinate’s control (Green & Mitchell, 1979). Causality is attributed more to the subordinate than to the situation if the subordinate has had a history of poor performance. Additionally, if the effects of the poor performance have severe outcomes, it could be determined that the subordinate was at fault (Mitchell & Wood, 1979). In this situation, the leader would focus more on some remedial action towards the subordinate, rather than on the situation (Bass, 1990). Continued punishment may lead to a downward spiral in performance leading to learned helplessness which would be dealt with even more punitively if a

50

PATRICK A. PALMIERI ET AL.

subordinate was previously viewed as a poor performer (James & White, 1983). Unfortunately, this punitive punishment can lead to even greater declines in subordinate performance (Peterson, 1985) and possibly greater system instability. Developing and supporting a culture of excellence begins at the leadership level by embracing a preoccupation with preventing failure (Weick & Sutcliffe, 2001). Chiles (2002, p. 15) describes a leader’s preoccupation with system cracks as the ‘‘chess master spend[ing] more time thinking about the board from his opponent’s perspective than he does from his own.’’ Leaders ought to seek assistance from clinicians and other health care professionals with expertise in detecting and extracting the subtle signals of impending issues from the constant noise of routine workﬂow prior to initiating systematic change. These experts can provide the feedback necessary to proactively identify and correct system abnormalities with minimal disturbance given that the best intended system changes can be quite disruptive to the institution.

Layer 2: Risky Supervision The second defensive layer is management, speciﬁcally those supervisors at various organizational levels directly responsible to the leadership team for system metrics and outcomes. When management is ineffective in correcting known gaps or holes, patient welfare is endangered. Thus, management involvement is critical in maintaining a defense against adverse events (Reason, 1990). Risky supervision arises when management decisions lead to circumventing or not enforcing written policies and procedures (Reason & Hobbs, 2003), overlooking known issues, and the absence of corrective actions (Leduc, Rash, & Manning, 2005) as employees engage in potentially consequential unsafe activity. Practitioner deviations from written policies and procedures signiﬁcantly contribute to adverse events (Vaughn, 1996). As such, management’s enforcement of policies is critical to the safe functioning of any organization. Workload pressures created by understafﬁng or overly stressful work assignments can lead to poor outcomes (Aiken et al., 2001; Aiken, Clarke, Sloane, Sochalski, & Silber, 2002; Stone et al., 2007). Performance pressures caused by any number of issues creating a difﬁcult and error-prone workplace are examples of risky supervision. In 2002, the Joint Commission reported the lack of nursing staff contributed to nearly 25% of the unanticipated issues resulting in patient death or injury (Joint Commission on Accreditation of

Error in Health Care

51

Healthcare Organizations, 2007). Tucker and Spear (2006) reported inadequate nursing task times, multiple unplanned changes in task, and frequent interruptions mid-task throughout a typical shift. These reported issues may result in harm to patients, especially in complex care situations such as the intensive care or perioperative environments. In addition, unpaid overtime at the end of the shift is a well known phenomenon (Tucker & Spear, 2006) with nurses reporting on average greater than one hour per shift (Rogers, Wang, Scott, Aiken, & Dinges, 2004; Tucker & Spear, 2006). The intensity of work coupled with the inability of nurses to complete their work within a scheduled shift indicates poor job design and inadequate support, possibly at the level of immediate supervision. Also, these work pressures are not unique to nursing since other health care professional face similar conditions (Institute of Medicine, 2004; Kohn et al., 2000). In short, managers, both clinical and administrative, are directly responsible for organizing, implementing, and evaluating policies and procedures to guide the safe supervision and management of the complex adaptive care delivery system. Considering system design, establishment of boundaries, and expectations under which clinicians practice, management can foster safer performance through more supportive organizational norms and values (Institute of Medicine, 2004; Kohn et al., 2000; Rasmussen, 1999; Reason et al., 2001). Supervisors should foster a safe environment by actively identifying and addressing deﬁciencies among practitioners, equipment, processes, and/or training. Safe practices include correcting issues, utilizing progressive discipline and on-the-spot performance coaching immediately addressing unacceptable clinical behavior while promoting adherence to policies and procedures. When considering the situations related to unsafe practices, supervisory level staff are integral to maintaining the sensitivity to the operations and commitment to the resiliency of the organization (Weick & Sutcliffe, 2001). Imperfect environmental conditions represented in the next layer, preconditions for unsafe practices can be revealed by vigilant supervisory and management participation. Thus, situational awareness of their respective environment in relation to the ‘‘big picture’’ is vital to managers’ impact on the next layer.

Layer 3: Situations for Unsafe Practices The third layer represents precursors to practices that are not safe. Careful attention to error substance at the situations for unsafe acts level is

52

PATRICK A. PALMIERI ET AL.

necessary to protect the system and the clinicians from mistaken error attribution. Substandard conditions and practices are two situational circumstances responsible for the presence of unsafe acts in practice (Reason, 1998). In fact, these two synergistic conditions each fuel the other as situations increasing the likelihood of an unsafe act (Reason, 1990). The inability of local systems to provide practitioners with reliable support, including information, equipment, and supplies, has been described as ‘‘operational failures’’ (Tucker, 2004) which create substandard conditions for practitioners attempting to execute their patient care responsibilities. Tucker and Spear (2006) observed that nurses experience an average of 8.4 operational failures per shift. An example of a serious, but often stealth, operational failure situation is the ‘‘down-time’’ associated with an automated pharmaceutical dispensing system interface that links the dispensing machine to the pharmacy. When this operational failure occurs, the dispensing machine is unable to proﬁle each patient’s personal medication schedule. When a patient speciﬁc medication proﬁle is absent a notable defensive hole is created. This issue increases the probability for error as the dispensing machine’s entire pharmaceutical contents are readily available and the selection is not limited to only those medications speciﬁcally ordered for each patient. Notably, a clinician is able to remove the wrong medication or the wrong medication dose from the dispensing unit, an act that is less likely when the pharmacymachine interface is properly functioning. As such, this situation creates the precondition for a potentially serious event as delivering the wrong medication or dosage in this example results in the occurrence of an unsafe practice that may lead to patient harm or death. Other failures, or errors, at the clinician level result from difﬁculties in task partitioning exempliﬁed by postponing a task until missing supplies arrive due to operational failures (Tucker & Spear, 2006) acting as a precondition to actualize clinician error. With task partitioning, clinicians experience work interruptions (such as not being able to administer an urgent or stat medication) linked to an unsafe precondition (reoccurring delays in pharmacy providing stat medications to the clinical unit) thereby leading to substandard practice (omission of medication delivery, possibly due to memory, or deterioration in patient condition, due to time). These examples illustrate the complexity of error at the clinician level that might be considered substandard practice or even a ‘‘practice breakdowns’’ (Benner et al., 2006). System focused organizations highlight the reluctance to accept simpliﬁcations like ‘‘practice breakdowns’’ in the pursuit of error causation. Informed organizations use strategies such as tracing

53

Error in Health Care

preconditions through the complex adaptive system, layer by layer, prior to the clinical practice level, seeking to determine why the event manifested versus rather than who was responsible (Weick & Sutcliffe, 2001). In a concerted effort to universally improve the situations for safer clinical practices, many agencies and organizations advocated best practices and suggested system attributes purposed at addressing errors related to substandard practices. The Joint Commission, through the creation of the National Patient Safety Goals (NPSG), and the Institute for Safe Medication Practices (ISMP), through medication delivery recommendations have positively impacted health care in emphasizing the importance of patient safety. For example, sound-alike, look-alike drugs (SALAD) and high-alert medications have been identiﬁed as areas prone to compromise patient safety, especially when clinician vigilance is altered by other complex adaptive attributes, such as fatigue and workload. Strategies have been adopted to address issues related to substandard medication practices caused by unsafe situations. Recommendations can quickly become accreditation standards such as removing dangerous highly concentration drugs like potassium chloride from patient care units, adopting formal patient identiﬁcation requirements, and mandating surgical instrument counts as an essential step at the end of surgical procedures. On the horizon, situations related to using automated pharmaceutical dispensing devices, incorporating bar-scanning technologies into the medication delivery process, and implementing electronic medical records all promise to reduce opportunities for unsafe practice.

Layer 4: Unsafe Performance The fourth layer represents unsafe practices which are usually linked to the health care professional’s performance at the ‘‘sharp-end’’ of care leading to an adverse event. In the simplest form, unsafe practices consist of errors and violations (Reason, 2000). Errors frequently involve perceptual (Wiegmann & Shappell, 2001), decision making (Norman, 1988; Perrow, 1984; Rasmussen, 1990), and/or skill-based lapses (Norman, 1988; Reason, 1990). Unsafe practices are the tip of the adverse event iceberg (Reason, 1998) as the number of these acts tends to be rather large but discrete (Reason, 1990, 1998) and usually are the most noticeable following a sentinel event (Perrow, 1984; Reason, 1998, 2000; Wiegmann & Shappell, 2003). Although the error and sentinel event numbers have yet to be

54

PATRICK A. PALMIERI ET AL.

accurately quantiﬁed, the magnitude of unsafe practices is put into perspective by considering the IOM’s (2007b) ﬁnding that on average, each hospitalized patient experiences one medication error per day. Violations are errors that are more serious in nature but occur less frequently (Reason, 1990). Wiegmann and Shappell (1997) describe violations as the willful disregard of policies, procedures, and rules governing health care professionals at both the organizational and regulatory levels. Violations are willful, therefore, missing the sincere intention to enhance a problematic situation through improvisation. Reason (1998) attributes violations to issues related to motivation, attitude, and culture. These acts amount to deviations from safe operating practices (Vincent, Taylor-Adams, & Stanhope, 1998) even possibly resulting from a lack of training or education (Institute of Medicine, 2004). Examples of clearly willful clinician violations amounting to ‘‘practice breakdowns,’’ are purposely falsifying documents, failing to execute life-saving procedures when observably necessary, and stating untruthfully that actions, events, or treatments were completed. At this layer, clinicians develop capabilities to detect, contain, and report unsafe situations or errors. Expert clinicians maintain their distance from the sharp-end of an adverse event with heightened situational improvisation (Hanley & Fenton, 2007) to generate a protective barrier despite numerous complex care management issues (Kohn et al., 2000; Vincent, 2003). This protection essentially results from experience and heightened familiarity with improvisation to mitigate issues created by lack of clear protocols for a situation, high stress clinical scenarios, and fatigue from heavy workloads and long working hours (Hanley & Fenton, 2007; Helmreich & Davies, 2004; Vincent, 2003). Left to novice clinician hands, improvisation may lead to unfavorable consequences, as important experiential knowledge has not developed. It is suggested that error is the symptom of a practice breakdown (Benner et al., 2006). The complexity embedded within the situation when an error is unsuccessfully defended by an expert clinician at the sharp-end is substantial. Speciﬁcally, incorporating improvisation as a normal practice attribute might render the system better capable of defending from error. Health care professionals should be expected to draw upon their prior experiences (Weick, 1993) in order to positively impact troublesome situations with the resources at hand (Hanley & Fenton, 2007) even if it requires them to not follow policies and procedures. When system opportunity for failure exceeds the clinician’s ability to workaround and minimize error, the risk for a system meltdown is substantial (Chiles, 2002;

Error in Health Care

55

Kohn et al., 2000). Expert health care professionals when found at or near the sharp-end of a sentinel event signals the likely presence of a system overload rendering ineffective the expert’s improvisation skills. Weick and Sutcliffe suggest that organizations embrace ‘‘the reluctance to accept error simpliﬁcations responsible for creating more complete and nuanced pictures’’ as an important HRO characteristic (2001, p. 11). There are times when occurrences considered to be violations are actually issues related to motivation, attitude, and culture (Reason, 1998). Some acts amount to deviations from safe operating practices (Vincent et al., 1998) while others probably result from a lack of training or education (Institute of Medicine, 2004). Some might argue leaders and managers responsible for safe stafﬁng and correcting hazards are responsible for the commission of a violation when patients are harmed due to known but uncorrected issues (Reason, 1990, 1998). Willful practice violations assume the presence of an informed and trained professional with an unobstructed understanding of the policies and procedures (Amalberti, Vincenct, Auroy, & de Saint Maurice, 2006). The professional must be working in an environment where the speciﬁc policy and/or procedure is enforced, and the professional knowingly and deliberately disregarded the rule or rules (Reason, 1998). However, health care education and training speciﬁc to patient safety is lacking in many curriculums (Institute of Medicine, 2003) and it is an important aspect to view adverse events in the context of error versus willful violation. As such, the signiﬁcant issue facing organizations, even society in general, is they do not recognize errors as an unavoidable aspect of limitations in human performance (Reason et al., 2001). Despite the negative impact upon the complex adaptive system when human error manifests, the overall impact of excellent patient care produced by wellintended clinicians should outweigh the consequence caused by unintentional error.

Defending the Patient from System Errors Although problem solving and short-term ﬁxes correct errors that arise in the normal course of task completion (Tucker & Edmondson, 2003), these ﬁxes are not typically reported as clinicians improvise and adapt in the normal course of work. Consider the situation created when important medications are unavailable to nurses, yet the patient’s condition requires availability. The result is a clear gap in the medication delivery system.

56

PATRICK A. PALMIERI ET AL.

By improvising, the clinician could borrow the same medication from a different patient’s medication bin (which could be considered a violation of policy). Time permitting, the nurse could leave the busy unit in order to walk to the pharmacy. This ﬁlls the existing gap, or need, but may create yet another if another patient deteriorates while the nurse is off of the unit. It is important to recognize that eliminating active failures stemming from unsafe acts prevents only one adverse event (Reason, 2000) and does not construct better defenses for removing system gaps (Cook et al., 2000). Therefore, examining and adjusting the system at the organizational and managerial layers to ‘‘manage error’’ is a more effective method for eliminating a larger variety of latent errors (Amalberti et al., 2006), subsequently limiting the manifestation of active errors (Reason, 1998). To prevent adverse events that result directly from unsafe acts requires blocking holes or trapping hazards that result from within the clinician’s controlled defensive layer. The most effective method to mitigate any errorprone process is to create and support an occurrence or incident reporting system (Cook & O’Connor, 2005; Department of Veterans Affairs, 2002). Sasou and Reason (1999, p. 3) state, ‘‘failures to detect are more frequent than failures to indicate and correct.’’ However, leaders must support the removal of blame and punishment if error reporting and detection is to become routine in practice (Hofmann & Stetzer, 1998; Kohn et al., 2000). In an effort to protect the system and the patient from error, it is perhaps even more important to review the organizational process when near-events occur, those situations where an adverse event is avoided at the very last moment. Contrary to contemporary patient safety knowledge, professional and quasi-legal bodies appear less concerned with the category of error resulting in the near-events or nearly harmful to patients instead favoring those situations resulting in actual patient injury. This focus might be explained by the regulatory body predisposition to allow the consequences of an accident to ‘‘color and even distort’’ the perception of the system attributes and the actions of the clinicians leading up to the event (Reason & Mycielska, 1982). These almost accidents, or near-events, present signiﬁcant evidence of an approaching sentinel event like the tremor felt prior to an impending earthquake. Reason (2000) considers near misses as free lessons that previews or represents potential and preventable future adverse events. Organizations must be alert as well as vigilant in identifying these near misses in order to correct, change, or modify policies, procedures, and practices to prevent future errors that could result in a sentinel event.

Error in Health Care

57

THE ROLE OF HEALTH CARE LEADERSHIP IN PATIENT SAFETY Since its formalization in 2000, the patient safety movement gained both national attention and increased resources dedicated to creating safer health care delivery systems. In 2003, the Joint Commission established the initial NPSG. These initiatives helped patient safety advocates to actively transform health care based on an understanding from the organizational accident and human factors sciences. This transdisciplinary knowledge transfer assists leaders in understanding the path to develop health care into a high reliability or high risk but low accident industry (Institute of Medicine, 2003; Kohn et al., 2000). Developing and maintaining the optimal patient safety culture while concurrently managing the overall organization performance metrics is challenging (Joyce, Boaden, & Esmail, 2005; Reason & Hobbs, 2003). In a 2004 survey, hospital administrators ranked ﬁnancial challenges at the top of the list of their 11 priorities while patient safety ranked 10th and patient satisfaction ranked 11th (American College of Healthcare Executives, 2005). Over the last decade, health care organizations attempted to eliminate error through the zero-deﬁcit culture (Institute of Medicine, 2004) with modest success (Nolan et al., 2004; Roberts, 2002). As a direct result of the ﬂawed human approach to error management, health care professionals hide errors in order to protect themselves from reprisal and punishment (Kohn et al., 2000). Unreported latent errors exist partly due to the organizational inertia (Palmieri, Godkin, & Green, 2007) created by fear (Kohn et al., 2000; Millenson, 2003), and remain free in the environment until a system meltdown results in an adverse event (Reason, 1998; Reason et al., 2001; Vaughn, 1996, 1999). While the NCSBN seeks to establish the attentiveness and surveillance standard of nursing practice associated with watchful vigilance in preventing the failure to rescue (Benner et al., 2006), health care leaders are familiar with clinician inability to effectively monitor patients within the complex adaptive system due to a variety of pressures. The increased complexity of the typical hospital unit coupled with reduced nursing hours due to ﬁnancial, human resource, and simple supply constraints and increased physician workload is problematic to effective and safe hospital operations. As a result, health care leaders should recognize that patient safety experts are unable to uniformly deﬁne a realistic ‘‘one case ﬁts all’’ threshold for system error versus professional neglect in the contemporary health care system.

58

PATRICK A. PALMIERI ET AL.

Professional and governing organizations for pharmacy, medicine, and nursing play pivotal roles in providing support to all constituents in advocating the system approach versus their traditional person approach to event investigation and adjudication. Health care leaders are well positioned to support clinical professionals and improve patient safety by vigorously advocating that error causation investigations related to adverse events embrace the system approach. Some organizations remain attached to the human approach as evidenced by statements such as, ‘‘when nursing practice falls below this minimal acceptable threshold sanctions are necessary to protect the public’’ (Benner et al., 2006, p. 59). The differentiation between willful and accidental error is an essential point for consideration in any RCA investigation. Learning about near-events and latent errors necessitates an open, honest, and sincere organizational culture. The Institute of Medicine (2004, p. 226) states, ‘‘although near-miss events are much more common than adverse events – as much as 7 to 100 times more frequent – reporting systems for such events are much less common.’’ In order to turn the potential lemon of near misses in the lemonade of error prevention, leaders should consider deﬁning what a near miss is in their organization, talk openly about near-events when they occur, and preach near-events are positive signs of a system with functioning safeguards despite vulnerability (Weick & Sutcliffe, 2001). Preoccupied with probing their potential and actual failures related to nearly adverting disaster (Weick & Sutcliffe, 2001), HROs realize these nearly missed events create future success for all parties. Leaders should also recognize that the vast majority of patient safety knowledge, including the NPSGs, are based on expert opinions, recommendations, and experiences as well as anecdotal knowledge (Agency for Healthcare Research and Quality, 2001). The Agency for Healthcare Research and Quality (AHRQ) analyzed the current state of affairs for patient safety practices of which 79 practices were reviewed for the strength of research evidence supporting current recommendations. After review, only 11 patient safety practices were correlated with a high level of reliable evidence. These 11 practices were clinical in nature and did not include crew resource management, computerized physician order entry, simulation use, or bar-coding technology. In addition, AHRQ found the literature is insufﬁcient to make rational decisions about organizational nursing environments. A number of recommended practices with longstanding success outside of health care were not included in the analysis because of the general weakness of health care evidence. These practices included incident reporting, application of human factors principles, utilization of

Error in Health Care

59

communication models, and promoting cultures of safety. Further research, especially for practices drawn from industries outside of health care, is needed to ﬁll ‘‘the substantial gaps in the evidentiary base’’ of patient safety practices (Agency for Healthcare Research and Quality, 2001).

Health Care Leaders and High Reliability Organizations Through ‘‘problemistic search’’ (Cyert & March, 1963, p. 120) managers and leaders scan the environment for quick solutions to immediate problems regardless of complexity. Reason (2000, p. 770) states, ‘‘most managers y attribute human unreliability to unwanted variability and strive to eliminate it.’’ Effective error management includes decreasing the incident of critical error (Reason, 2000) and developing improved systems capable of containing dangerous error (Reason et al., 2001). When management demonstrates positive organizational safety practices, subordinates’ will accept safety as an essential job responsibility considerably improves (Dunbar, 1975; Katz-Navon, Naveh, & Stern, 2005; Kohn et al., 2000). This approach, reﬂective of HROs, focuses attention to learning and awareness at all levels of the organization (Reason, 1997), including the individual, department, and leadership. Preoccupied with probing their potential and actual failures related to nearly adverting disaster, HROs frequently avoid sentinel events (Weick & Sutcliffe, 2001) as these nearly missed events create future successes from a single realization. In fact, an organization should develop a policy speciﬁcally endorsing the system philosophy to error with emphasis on the non-punitive consequences related to unintentional errors (Helmreich & Davies, 2004). HROs exhibit resilient systems and superior hazard intelligence (Reason, 2000) where practitioners employ positive adaptive behaviors (Mallak, 1998) to immediately defend the patient from potentially harmful situations therefore sustaining the reliability of the process (Carroll & Rudolph, 2006). Linking the actual error to organizational factors is difﬁcult (Wiegmann & Shappell, 2001) as system failure results from the misguided dynamics and imperfect interactions of the care delivery system (Institute of Medicine, 2004). Recognizing the importance of developing safe and reliable health care delivery systems (Institute of Medicine, 2004) necessitates active involvement by multiple health care stakeholders such as practitioners, administrators, regulators, and even patients (Katz-Navon et al., 2005; Leonard, Frankel, & Simmonds, 2004). The reliability of an organization is a measurement of process capability (Reason et al., 2001) within a complex

60

PATRICK A. PALMIERI ET AL.

system (Cook & O’Connor, 2005) under the stresses of ordinary and unordinary conditions (Berwick & Nolan, 2006). An elevated level of organizational vigilance is reﬂective of HROs as they generally support a cultural predisposition that prevents the materialization of adverse events (Weick & Sutcliffe, 2001). Hence, the investigatory work for organizations embracing the system approach deemphasizes active errors in order to scrutinize system malfunctions (Reason, 1990) in the quest to discover latent conditions and rejuvenate processes (Rapala & Kerfoot, 2005). Rigid hierarchies create a vulnerability to error and leaders should ﬁrst seek to elimination error at their own defensive layer. In doing so, leaders prevent the culmination with lower level errors by deferring decisions to those with the expertise (skills, knowledge, and experience) consequently improving the overall system (Kohn et al., 2000; Reason, 2000; Weick & Sutcliffe, 2001). Leaders focused on eliminating the majority of common error from their respective organization should recognize and understand the elements of Reason’s work and our subsequent adoption to the Health Care Error Proliferation Model. Low incidences of accidents are observed and reported in HROs, most commonly found in aerospace (Nolan et al., 2004) and nuclear generation industries (Cook, Woods, & Miller, 1998). HROs experience fewer accidents then other organizations (Reason, 2000) as HROs recognize the importance of limiting human variability in their pursuit of safety (Roberts, 2002). In addition, HROs remain vigilant in recognizing the possibility of failures (Reason, 2000; Weick & Sutcliffe, 2001) and discover solutions in the work of human factor and organizational accident experts (Helmreich & Davies, 2004; Kohn et al., 2000). The increased diligence in examining systems for error is important. In recent literature, the concept of technological iatrogenesis (Palmieri & Peterson, 2008) and the resulting e-iatrogenic error (Campbell, Sittig, Ash, Guappone, & Dykstra, 2007; Weiner, Kfuri, Chan, & Fowles, 2007) has been suggested to be a signiﬁcant future contributor to health care adverse events. As the health care technology enterprise rapidly expands, the likelihood of new and previously unseen error typologies impacting systems of care is probable (Harrison, Koppel, & Bar-Lev, 2007; Palmieri, Peterson, & Ford, 2007). This new error formation presents signiﬁcant challenges for the system as Lyons et al. (2005) found that administrators, physicians, and nurses hold different beliefs about the universal assistance and the speciﬁc obstacles relevant to the overall information system as a clinical beneﬁt. The most important distinguishing feature of HROs is their collective preoccupation with identifying and eliminating system failure (Reason,

61

Error in Health Care

2000; Weick & Sutcliffe, 2001). The notion that HRO’s are preoccupied with proactively identifying system ﬂaws as an effective methodology for organizational improvement is not universally accepted, however. Some health care experts suggest ‘‘a systems approach is based upon a post hoc analysis and redesign of a system based upon unsafe performance’’ (Benner et al., 2006, p. 51). As the proactive FEMA increases in popularity with continued organizational interest in developing HRO attributes, the system approach to preventing failures will be universally recognized. Health care leaders remain ideally positioned to make the system approach a reality by transforming their institutions into HROs.

CONCLUSION The Health Care Error Proliferation Model is based on the premise that health care organizations struggle to effectively integrate existing research, novel approaches, practice innovations, and new knowledge about patient safety into their complex care delivery systems (Gray, 2001; Kohn et al., 2000; Nolan et al., 2004). Health care clinicians and leaders are beginning to recognize the probable creation of more serious problems should the prevailing ‘‘punitive cultures and a focus on ‘bad apples’ instead of poorly constructed systems’’ be permitted to continue to impedes progress (Devers et al., 2004, p. 114). We reviewed a diverse body of knowledge and provided a transdisciplinary analysis of health care errors and events. Then, we discussed the relevant literature pertaining to the conceptualization of errors, approaches to identifying the causes of errors, and the utility of the Swiss Cheese Model of adverse events in reconsidering the existing health care leadership approach for developing organizational cultures of safety. Finally, we discussed the critical role of health care leadership in building defenses against adverse events by demonstrating the attributes of HROs. We conclude that within the discipline of health care leadership, the value of learning and applying basic organizational accident principles (Leape et al., 1998) to the complex adaptive system of health care organizations is paramount to preserving patient safety (Helmreich & Davies, 2004; Reason, 2000). Embracing the system philosophy to error management is vitally important for health care leaders desiring to achieve HRO status. Health care professionals should support the patient safety movement by focusing on systems versus people as the fundamental strategy for developing into a high reliability industry.

62

PATRICK A. PALMIERI ET AL.

We agree with Bennis (1989) who argued that organizational cultures are created, supported, and sustained through leadership role modeling of acceptable organizational conduct. Leaders create and sustain organizational safety by: (1) role modeling the acceptable behaviors; (2) embracing the system approach to managing critical organizational incidents; (3) rewarding desired and productive behaviors; and (4) eliminating the use of blame, shame, and punishment to address professional error. Understanding the Health Care Error Proliferation Model as a strategic overview for identifying latent and active error will improve the complex health care delivery systems. The responsibility for organization improvement rests in the capable hands of health care leaders to promote discovery, motivate and advocate transformation, and protect the integrity of systems and people in the pursuit of patient safety. As leaders continue to emphasize support for the system approach and push to reduce the person ‘‘blame and shame’’ approach, errors will decline. As errors decline, hospitals will continue their progress to demonstrate the characteristics associated with HROs. In the future the incredible number of patients who die each year due to health care errors will decline and the entire system will successfully improve the quality and safety of patient care.

ACKNOWLEDGMENTS We would like to acknowledge the assistance of Larry Reising for his thorough content review as an expert in the area of industrial and aviation safety, to Ruth Anderson for her assistance with editing aspects related to complex adaptive systems, and to Barbara Cherry and Rodney Hicks for their extensive and helpful editorial suggestions and comments.

REFERENCES Abramson, L. Y., Garber, J., & Seligman, M. E. P. (1980). Learned helplessness in humans: An attributional analysis. New York: Academic Press. Agency for Healthcare Research and Quality. (2001). Making health care safer: A critical analysis of patient safety practices, AHRQ Publication No. 01-E058. Rockville, MD: Agency for Healthcare Research and Quality. Aiken, L. H., Clarke, S. P., Sloane, D. M., Sochalaski, J. A., Busse, R., Clarke, H., Giovanneti, P., Junt, J., Rafferty, A. M., & Shamian, J. (2001). Nurse reports on hospital care in ﬁve countries. Health Affairs, 20(3), 43–53. Aiken, L. H., Clarke, S. P., Sloane, D. M., Sochalski, J. A., & Silber, J. H. (2002). Hospital nurse stafﬁng and patient mortality, nurse burnout and job dissatisfaction. Journal of the American Medical Association, 288(16), 1987–1993.

Error in Health Care

63

Amalberti, R., Auroy, Y., Berwick, D., & Barach, P. (2005). Five system barriers to achieving ultrasafe health care. Annals of Internal Medicine, 142(9), 756–764. Amalberti, R., Vincenct, C., Auroy, Y., & de Saint Maurice, G. (2006). Violations of migrations in health care: A framework for understanding and management. Quality and Safety in Health Care, 15(S1), 66–71. American College of Healthcare Executives. (2005). Top issues confronting hospitals: 2004; Available at http://www.ache.org/pugs/research/ceoissues; January 8. American Nurses Association. (2006). The American nurses association comments on the Wisconsin department of justice decision to pursue criminal charges against an RN in Wisconsin; Available at http://nursingworld.org/ethics/wis11-20-06.pdf; December 10. Anderson, R. A., Issel, M. L., & McDaniel, R. R. (2003). Nursing homes as complex adaptive systems: Relationship between management practice and resident outcomes. Nursing Research, 52(1), 12–21. Bass, B. M. (1990). Bass and Stogdill’s handbook of leadership: Theory, research and managerial applications (3rd ed.). New York: The Free Press. Benner, P., Malloch, K., Sheets, V., Bitz, K., Emrich, L., Thomas, M. B., Bowen, K., Scott, K., Patterson, L., Schwed, K., & Farrel, M. (2006). TERCAP: Creating a national database on nursing errors. Harvard Health Policy Review, 7(1), 48–63. Berta, W. B., & Baker, R. (2004). Factors that impact the transfer and retention of best practices for reducing error in hospitals. Health Care Management Review, 29(2), 90–97. Berwick, D. & Nolan, T. (2006). High reliability healthcare; Available at http://www.ihi.org/ihi/ topics/reliability/reliabilitygeneral/emergingcontent/highreliabilityhealthcarepresentation. htm; November 22. Blendon, R. J., Schoen, C., Donelan, K., Osborn, R., DesRoches, C. M., Scoles, K., Davis, K., Binns, K., & Zapert, K. (2001). Physicians’ views on quality of care: A ﬁve-country comparison. Health Affairs, 20(3), 233–243. Campbell, E. M., Sittig, D. F., Ash, J. S., Guappone, K. P., & Dykstra, R. H. (2007). In reply to: e-Iatrogenesis: The most critical consequence of CPOE and other HIT. Journal of the American Medical Informatics Association, 14(3), 389–390. Carroll, J. S., & Rudolph, J. W. (2006). Design of high reliability organizations in health care. Quality and Safety in Health Care, 15(Suppl. 1), i4–i9. Casti, J. L. (1997). Would-be worlds. New York: Wiley. Cherns, A. B. (1962). Accidents at work. In: A. T. Welford, M. Argyle, D. V. Glass & J. N. Morris (Eds), Oriblems and methods of study. London: Routledge and Kegan Paul. Chiles, J. R. (2002). Inviting disaster: Lessons from the edge of technology. New York: HarperCollins Publishers. Cilliers, P. (1998). Complexity and post modernism: Understanding complex systems. New York: Routledge. Coleman, H. J. (1999). What enables self-organizing behavior in business. Emergence, 1(1), 33–48. Committee on the Future of Rural Healthcare. (2005). Quality through collaboration: The future of rural healthcare. Washington, DC: National Academies Press. Cook, R. I., & O’Connor, M. F. (2005). Thinking about accidents and systems. In: H. Manasse & K. Thompson (Eds), Improving medication safety. Mathesda, MD: American Society for Health-System Pharmacists. Cook, R. I., Render, M., & Woods, D. D. (2000). Gaps in the continuity of care and progress on patient safety. British Medical Journal, 320(7237), 791–794.

64

PATRICK A. PALMIERI ET AL.

Cook, R. I., & Woods, D. D. (1994). Operating at the sharp end: The complexity of human error. In: M. S. Bogner (Ed.), Human error in medicine (pp. 285–310). Hinsdale, NJ: Lawrence Erlbaum. Cook, R. I., Woods, D. D., & Miller, C. (1998). A tale of two stories: Contrasting views of patient safety, Report from a workshop on assembling the scientiﬁc basis for progress on patient safety: 1–86. Chicago: National Health Care Safety Council of the National Patient Safety Foundation at the AMA. Cyert, R., & March, J. G. (1963). A behavioral theory of the ﬁrm. Englewood Cliffs, NJ: Prentice-Hall. Davis, K., Schoenbaum, S. C., Collins, K. S., Tenney, K., Hughes, D. L., & Audet, A. J. (2002). Room for improvement: Patients report on the quality of their health care. New York: The Commonwealth Fund. Department of Veterans Affairs. (2002). The Veterans Health Administration national patient safety improvement handbook. Washington, DC: U.S. Department of Veterans Affairs. Devers, K. J., Pham, H. H., & Lui, G. (2004). What is driving hospitals’ patient-safety efforts? Health Affairs, 23(3), 103–115. Dorner, D. (1996). The logic of failure: Recognizing and avoiding error in complex situations. New York: Metropolitan Books. Dunbar, R. L. M. (1975). Manager’s inﬂuence on subordinates’ thinking about safety. Academy of Management Journal, 18(2), 364–369. Espin, S., Lingard, L., Baker, G. R., & Regehr, G. (2006). Persistence of unsafe practice in everyday work: An exploration of organizational and psychological factors constraining safety in the operating room. Quality and Safety in Health Care, 15(3), 165–170. Farrokh, F., Vang, J., & Laskey, K. (2007). Root cause analysis. In: F. Alemi & D. H. Gustafson (Eds), Decision analysis for healthcare managers. Chicago: Health Administration Press. Gano, D. L. (2003). Apollo root cause analysis: A new way of thinking (2nd ed.). Yakima, WA: Apollonian Publications. Garber, J., & Seligman, M. E. P. (1980). Human helplessness: Theory and applications. New York: Academic Press. Gibson, R., & Singh, J. P. (2003). Wall of silence: The untold story of the medical mistakes that kill and injure millions of Americans. Washington, DC: Lifeline Press. Gray, J. A. M. (2001). Evidence-based healthcare: How to make health policy and management decisions. London: Churchill Livingstone. Green, S. G., & Mitchell, T. R. (1979). Attributional processes of leaders in leader-member interaction. Organizational Behavior and Human Performance, 23, 429–458. Hanley, M. A., & Fenton, M. V. (2007). Exploring improvisation in nursing. Journal of Holistic Nursing, 25(2), 126–133. Harrison, M. I., Koppel, R., & Bar-Lev, S. (2007). Unintended consequences of information technologies in health care: An interactive sociotechnical analysis. Journal of the American Medical Informatics Association, 14(5), 542–549. Harvey, J. H., & Weary, G. (1985). Attribution: Basic issues and applications. San Diego, CA: Academic Press. Heider, F. (1958). The psychology of interpersonal relations. New York: Wiley. Helmreich, R. L. (2000). On error management: Lessons from aviation. British Medical Journal, 320, 781–785.

Error in Health Care

65

Helmreich, R. L., & Davies, J. M. (2004). Culture, threat, and error: Lessons from aviation. Canadian Journal of Anesthesia, 51(6), R1–R6. Hofmann, D. A., & Stetzer, A. (1998). The role of safety climate and communication in accident interpretation: Implications for learning from negative events. Academy of Management Journal, 41(6), 644–657. Institute for Safe Medication Practices. (2006). Since when it is a crime to be human? Available at http://www.ismp.org/pressroom/viewpoints/julie.asp; November 30. Institute of Medicine. (2001). Crossing the quality chasm: A new health system for the 21st century. Washington, DC: National Academy Press. Institute of Medicine. (2003). Health professions education: A bridge to quality. Washington, DC: National Academy Press. Institute of Medicine. (2004). Patient safety: Achieving a new standard for care. Washington, DC: National Academy Press. Institute of Medicine. (2007a). Frequently asked questions; Available at http://www.iom.edu/ CMS/6008.aspx; 8 June. Institute of Medicine. (2007b). Preventing medication errors: Quality chasm series. Washington, DC: National Academies Press. James, L. R., & White, J. F. (1983). Cross situational speciﬁcity in manager’s perceptions of subordinate performance, attributions and leader behavior. Personnel Psychology, 36, 809–856. Joint Commission on Accreditation of Healthcare Organizations. (2007). Health care at the crossroads: Strategies for addressing the nursing crisis; Available at http://www. jointcommission.org/publicpolicy/nurse_stafﬁng.htm; 18 March. Joint Commission Resources. (2007). Front line of defense: The role of nurses in preventing sentinel events (2nd ed). Oakbrook Terrace, IL: Joint Commission Resources. Joyce, P., Boaden, R., & Esmail, A. (2005). Managing risk: A taxonomy of error in health policy. Health Care Analysis, 13(4), 337–346. Katz-Navon, T., Naveh, E., & Stern, Z. (2005). Safety climate health care organizations: A multidimensional approach. Academy of Management Journal, 48(6), 1075–1089. Kohn, L. T., Corrigan, J. M., & Donaldson, M. S. (Eds). (2000). To err is human: Building a safer health system. Washington, DC: National Academy Press. Lawton, R., & Parker, D. (2002). Barriers to incident reporting in a healthcare system. Quality and Safety in Health Care, 11(1), 15–18. Leape, L. L., Bates, D. W., Cullen, D. J., Cooper, J., Demonaco, H. J., Gallivan, T. R. H., Ives, J., Laird, N., Laffel, G., Nemeskal, R., Peterson, L. A., Porter, K., Servi, D., Shea, B. F., Small, S. D., Sweitzer, B. J., Thompson, B. T., & van der Vliet, M. (1995). Systems analysis of adverse drug events. ADE prevention study group. Journal of the American Medical Association, 274(1), 35–43. Leape, L. L., & Berwick, D. M. (2005). Five years after ‘‘To err is human’’: What have we learned? Journal of the American Medical Association, 293(19), 2384–2390. Leduc, P. A., Rash, C. E., & Manning, M. S. (2005). Human factors in UAV accidents. Special Operations Technology, 3(8). Available at http://www.special-operations-technology. com/article.cfm?DocID=1275. Accessed on May 18, 2006. Leonard, M. L., Frankel, A., & Simmonds, T. (2004). Achieving safe and reliable healthcare: Strategies and solutions. Chicago: Health Administration Press. Lyons, S. S., Tripp-Reinmer, T., Sorofman, B. A., Dewitt, J. E., Bootsmiller, B. J., Vaughn, T. E., & Doebbeling, B. N. (2005). Va queri informatics paper: Information technology for

66

PATRICK A. PALMIERI ET AL.

clinical guideline implementation: Perceptions of multidisciplinary stakeholders. Journal of the American Medical Informatics Association, 12(10), 64–71. Mallak, L. (1998). Putting organizational resilience to work. Industrial Management, 40(6), 8–13. Martinko, M. J., Douglas, S. C., & Harvey, P. (2006). Attribution theory in industrial and organizational psychology: A review. In: G. P. Hodgkinson & K. J. Ford (Eds), International review of industrial and organizational psychology (Vol. 21, pp. 127–187). Chicester: Wiley. Martinko, M. J., & Gardner, W. L. (1982). Learned helplessness: An alternative explanation for performance deﬁcits. Academy of Management Review, 7(2), 195–204. Martinko, M. J., & Thomson, N. F. (1998). A synthesis and extension of the Weiner and Kelley attribution models. Basic and Applied Social Psychology, 20(4), 271–284. McDaniel, R. R., & Driebe, D. J. (2001). Complexity science and health care management. Advances In Health Care Management, 2, 11–36. Millenson, M. L. (1997). Demanding medical excellence. Chicago: University of Chicago Press. Millenson, M. L. (2003). The silence. Health Affairs, 22(2), 103–112. Mitchell, T. R., & Wood, R. E. (1979). An empirical test of attributional model of leaders’ responses to poor performance. Paper presented at the Symposium on Leadership, Duke University, Durham, NC. National Council of State Boards of Nursing. (2007). Practice and discipline: TERCAP; Available at https://www.ncsbn.org/441.htm; April 18. Nolan, T., Resar, R., Haraden, C., & Grifﬁn, F. A. (2004). Innovation series: Improving the reliability of health care. Cambridge, MA: Institute for Healthcare Improvement. Norman, D. (1988). The design of everyday things. New York: Doubleday. Palmieri, P. A., Godkin, L., & Green, A. (2007). Organizational inertia: Patient safety movement slows as organizations struggle with cultural transformation. Texas: Tech University. Palmieri, P. A., & Peterson, L. T. (2008). Technological iatrogenesis: An expansion of the medical nemesis framework to improve modern healthcare organization performance. Proceedings of the Annual Meeting of the Western Academy of Management. March 26–29, 2008. Oakland, California. Palmieri, P. A., Peterson, L. T., & Ford, E. W. (2007). Technological iatrogenesis: New risks necessitate heightened management awareness. Journal of Healthcare Risk Management, 27(4), 19–24. Perrow, C. (1984). Normal accidents: Living with high risk systems. New York: Basic Books. Peterson, C., Maier, S., & Seligman, M. E. P. (1993). Learned helplessness: A theory for the age of personal control. New York: Oxford University Press. Peterson, M. F. (1985). Paradigm struggles in leadership research: Progress in the 1980s. Paper presented at the Academy of Management, San Diego, CA. Pfeffer, J. (1977). The ambiguity of leadership. Academy of Management Review, 2, 104–112. Rapala, K., & Kerfoot, K. M. (2005). From metaphor to model: The Clarian safe passage program. Nursing Economic, 23(4), 201–204. Rasmussen, J. (1990). The role of error in organizing behavior. Ergonomics, 33, 1185–1199. Rasmussen, J. (1999). The concept of human error: Is it useful for the design of safe systems in health care? In: C. Vincent & B. deMoll (Eds), Risk and safety in medicine (pp. 31–47). London: Elsevier. Reason, J. (2000). Human error: Models and management. British Medical Journal, 320(7237), 768–770. Reason, J. T. (1990). Human error. New York: Cambridge University Press.

Error in Health Care

67

Reason, J. T. (1997). Managing the rosks of organizational accidents. Aldershot: Ashgate Publishing. Reason, J. T. (1998). Managing the risks of organizational accidents. Aldershot, England: Ashgate. Reason, J. T., Carthey, J., & de Leval, M. R. (2001). Diagnosing ‘‘Vulnerable system syndrome’’: An essential prerequisite to effective risk management. Quality in Health Care, 10(S2), 21–25. Reason, J. T., & Hobbs, A. (2003). Managing maintenance error: A practical guide. Aldershot, England: Ashgate. Reason, J. T., & Mycielska, K. (1982). Absent-minded? The psychology of mental laspes and everyday errors. Englewood Cliffs, NJ: Prentice-Hall Inc. Roberts, K. (1990). Some characteristics of one type of high reliability organization. Organization Science, 1(2), 160–176. Roberts, K. H. (2002). High reliability systems. Report on the institute of medicine committee on data standards for patient safety on September 23, 2003. Rogers, A., Wang, W. T., Scott, L. D., Aiken, L. H., & Dinges, D. F. (2004). The working house of hospital staff nurses and patient safety. Health Affairs, 23(4), 202–212. Rosenthal, N. M. (1994). The incompetent doctor: Behind closed doors. London: Open University Press. Sasou, K., & Reason, J. T. (1999). Team errors: Deﬁnitions and taxonomy. Reliability Engineering and System Safety, 65(1), 1–9. Seligman, M. E. P., Maier, S. F., & Geer, J. (1968). The alleviation of learned helplessness in dogs. Journal of Abnormal Psychology, 73, 256–262. Smetzer, J. L., & Cohen, M. R. (1998). Lessons from Denver medication error/criminal negligence case: Look beyond blaming individuals. Hospital Pharmacy, 33, 640–656. Smetzer, J. L., & Cohen, M. R. (2006). Lessons from Denver. In: P. Aspden, J. Wolcott, J. L. Bootman & L. R. Cronenwett (Eds), Preventing medication errors (pp. 43–104). Washington, DC: National Academy Press. State of Wisconsin. (2006). Criminal complaint: State of Wisconsin versus Julie Thao. Wisconsin: Circiut court of Dane County. Stone, P. W., Mooney-Kane, C., Larson, E. L., Hran, T., Glance, L. G., Zawanziger, J., & Dick, A. W. (2007). Nurse working conditions and patient safety outcomes. Medical Care, 45, 571–578. Tucker, A. L. (2004). The impact of organizational failures on hospital nurses and their patients. Journal of Operations Management, 22(2), 151–169. Tucker, A. L., & Edmondson, A. (2003). Why hospitals don’t learn from hailures: Organizational and psychological dynamics that inhibit system change. California Management Review, 45(2), 1–18. Tucker, A. L., & Spear, S. J. (2006). Operational failures and interruptions in hospital nursing. Health Services Research, 41(3), 643–662. Vaughn, D. (1996). The challenger launch decision. Chicago: Chicago University Press. Vaughn, D. (1999). The dark side of organizations: Mistake, misconduct, and disaster. In: J. Hagan & K. S. Cook (Eds), Annual review of sociology (pp. 271–305). Palo Alto, CA: Annual Reviews. Vincent, C. (1997). Risk, safety and the dark side of quality. British Medical Journal, 314, 1175–1176. Vincent, C. (2003). Understanding and responding to adverse events. New England Journal of Medicine, 348(11), 1051–1056.

68

PATRICK A. PALMIERI ET AL.

Vincent, C., Taylor-Adams, S., & Stanhope, N. (1998). Framework for analyzing risk and safety in clinical medicine. British Medical Journal, 316(7138), 1154–1157. Wachter, R. M. (2004). The end of the beginning: Patient safety ﬁve years after ‘‘To err is human’’ [web exclusive]. Health Affairs, W4, 534–545. Weick, K. E. (1993). The collapse of sense-making in organizations: The Mann Gulch disaster. Administrative Science Quarterly, 38, 628–652. Weick, K. E., & Sutcliffe, K. M. (2001). Managing the unexpected: Assuring high performance in a range of complexity. San Francisco: Jossey-Bass. Weiner, B. (1972). Theories of motivation. Chicago: Markham Publishing Company. Weiner, B. (1995a). Judgment of responsibility: A foundation for a theory of social conduct. New York: Guilford Press. Weiner, B. (1995b). Attribution theory in organizational behavior: A relationship of mutual beneﬁt. In: M. J. Martinko (Ed.), Attributional theory in organizational perspective (pp. 3–6). Del Ray Beach, FL: St. Lucie Press. Weiner, J. P., Kfuri, T., Chan, K., & Fowles, J. B. (2007). ‘‘e-Iatrogenesis’’: The most critical unintended consequence of CPOE and other HIT. Journal of American Medical Informatics Association, 14(3), 387–389. Whittingham, R. B. (2004). The blame machine: Why human error causes accidents. Amsterdam: Elsevier Butterworth-Heinemann. Wiegmann, D. A., & Shappell, S. A. (1997). Human factor analysis of post-accident data: Applying theoretical taxonomies of human error. The International Journal of Aviation Psychology, 7(4), 67–81. Wiegmann, D. A., & Shappell, S. A. (1999). Human error and crew resource management failures in naval aviation mishaps: A review of U.S. Naval safety center data, 1990–1996. Aviation Space Environmental Medicine, 70(12), 1147–1151. Wiegmann, D. A., & Shappell, S. A. (2001). Human error perspectives in aviation. The International Journal of Aviation Psychology, 11(4), 341–357. Wiegmann, D. A., & Shappell, S. A. (2003). A human error approach to aviation accident analysis: The human factors analysis and classiﬁcation system. Aldershot: Ashgate Publishing Company. Wisconsin Hospital Association. (2006). Hospital association statement regarding legal actions against nurse; Available at http://www.wha.org/newscenter/pdf/nr11-2-06crimchargestmt. pdf; 10 November. Wisconsin Medical Society. (2006). Position statment regarding attorney general charges ﬁles today; Available at http://www.wisconsinmedicalsociety.org/member_resources/insider/ ﬁles/reaction_to_ag_changes.pdf; 7 November. Woods, D. D., & Cook, R. I. (2002). Nine steps to move forward from error. Cognitive Technology and Work, 4(2), 137–144. Zhan, C., & Miller, M. R. (2003). Excess length of stay, charges, and mortality attributable to medical injuries during hospitalization. Journal of the American Medical Association, 290, 1868–1874.

PROMOTING PATIENT SAFETY BY MONITORING ERRORS: A VIEW FROM THE MIDDLE Michal Tamuz, Cynthia K. Russell and Eric J. Thomas ABSTRACT Hospital nurse managers are in the middle. Their supervisors expect that they will monitor and discipline nurses who commit errors, while also asking them to create a culture that fosters reporting of errors. Their staff nurses expect the managers to support them after errors occur. Drawing on interviews with 20 nurse managers from three tertiary care hospitals, the study identiﬁes key exemplars that illustrate how managers monitor nursing errors. The exemplars examine how nurse managers: (1) sent mixed messages to staff nurses about incident reporting, (2) kept two sets of books for recording errors, and (3) developed routines for classifying potentially harmful errors into non-reportable categories. These exemplars highlight two tensions: the application of bureaucratic rule-based standards to professional tasks, and maintaining accountability for errors while also learning from them. We discuss how these fundamental tensions inﬂuence organizational learning and suggest theoretical and practical research questions and a conceptual framework.

Patient Safety and Health Care Management Advances in Health Care Management, Volume 7, 69–99 Copyright r 2008 by Emerald Group Publishing Limited All rights of reproduction in any form reserved ISSN: 1474-8231/doi:10.1016/S1474-8231(08)07004-3

69

70

MICHAL TAMUZ ET AL.

INTRODUCTION The inﬂuential Institute of Medicine (IOM) report on medical errors recommends that health care organizations ‘‘implement mechanisms of feedback and learning from error’’ (Kohn, Corrigan, & Donaldson, 2000, p. 181). Researchers demonstrate how accountability measures reduce frontline health care providers’ willingness to disclose potential problems and medical errors – information necessary for learning. They call for enacting a non-punitive environment (e.g., Kohn et al., 2000) or a ‘‘just culture’’ (Marx, 2001, 2003) in hospitals. Nurse managers are in the middle. Their supervisors expect that they will monitor and discipline nurses who commit errors, while also asking them to create a culture that fosters reporting of errors. Their staff nurses expect the managers to support them after errors occur and to understand that errors are often due to circumstances beyond their control. But from their positions in the middle, nurse managers also play a pivotal role in gathering information about and learning from medical errors. Drawing on interviews with nurse managers from three hospitals, we present three exemplars that illustrate how nurse managers cope with the demands of monitoring nursing errors and we explore the implications of these coping strategies for organizational learning. Understanding the roles of nurse managers is important because they can inﬂuence the availability of information about errors to other parts of the organization. Nurses are expected to report errors directly to their managers and managers also discover events that nurses do not report. Nurse managers can then pass on information to relevant parts of the hospital or sequester it. Finally, the tension between ensuring accountability and fostering learning is generalizable to many safety-reporting systems (Tamuz, 2001), and examining the roles of nurse managers may also inform our understanding of safety monitoring in high-hazard industries.

Nursing Roles Nurses play key roles in enabling the hospital to learn from its experience with errors and in maintaining accountability for mistakes. Nurses are well positioned to observe and receive feedback about ‘‘small failures’’ in the hospital (e.g., Edmondson, 2004) because they serve in a liaison role; they coordinate between the patient and other diagnostic and treatment units in the hospital (e.g., pharmacy, laboratory, and imaging). Moreover, nurses

Promoting Patient Safety

71

usually serve as the last safety barrier before patients receive medications. This position in an organization is often referred to as the ‘‘sharp end,’’ and leaders frequently blame those working at the sharp end when something goes wrong (Cook & Woods, 1994). Nurses are in a position to observe and be involved in medical errors, but they may be reluctant to report them. Although many hospitals have means for staff nurses to report errors independent of their managers, survey research indicates that nurses consistently attribute their failure to report errors to fear of negative repercussions. They are wary of being subject to disciplinary action and punitive responses by management (Wakeﬁeld et al., 1999; Osborne, Blais, & Hayes, 1999; Wolf, Serembus, Smetzer, Cohen, & Cohen, 2000). Wakeﬁeld et al. (1999) also found that nurses attributed their non-reporting to disagreement about whether an error occurred, the effort required to submit reports, and concerns about appearing incompetent. However, nurses surveyed expressed their intent to report errors if they were life threatening to the patient (Walker & Lowe, 1998) or if the patient had been injured, even at a minimal level of injury (Throckmorton & Etchegaray, 2007). The relationship between nurse managers and the nurses they supervise also inﬂuences whether nurses will call attention to medical errors and threats to patient safety. Edmondson (1996, 1999) found that nurse managers could be instrumental in encouraging staff nurses to report errors and problems; managers can create an environment of psychological safety that fosters nurses’ error reporting. In a survey of hospital nurses, Vogus and Sutcliffe (2007) also found that nurses’ trust in their managers was important. However, trust interacted with other safety-promoting activities to reduce medication error reporting when nurses also described a high level of safety-related activities in their unit. These studies illustrate the difﬁculties of untangling error reporting behavior and the actual underlying error rate. The relationship between nurses and their managers can also affect informal reporting of potential threats to patient safety. Individual nurses demonstrate their professional skills by problem solving. When they devise work arounds to solve immediate problems, they rarely inform nurse managers about the underlying issues; and thus, the lack of information hinders organizational learning (Tucker, Edmondson, & Spear, 2002; Tucker & Edmondson, 2003). The roles of nurse managers differ from those of staff nurses. Nurse managers act as a communication and administrative conduit between the staff nurses and the health care providers and administrators outside the patient care unit and thus, can be instrumental in maintaining accountability and promoting learning. Researchers have extensively studied the role of

72

MICHAL TAMUZ ET AL.

middle level managers in business (e.g., Balogun, 2006; Fenton-O’Creevy, 1998; Kanter, 1982; Mair, 2005) and in health care organizations (e.g., Ashmos, Huonker, & McDaniel, 1998; Brewer & Lok, 1995; Carney, 2004). Although considerable research has focused on the role of nursing middle management in hospitals, more research is needed on the roles of middle level nurse managers in promoting (or hindering) organizational learning.

Organizational Learning in Hospitals Research on organizational learning in hospitals and related health care organizations has focused on how organizations learn (or fail to learn) from an array of events that vary in their outcomes. Researchers discuss organizational learning from small failures (e.g., Edmondson, 2004; Sitkin, 1992), near misses (e.g., Callum et al., 2001), and preventable patient deaths. Some of these events resulted from errors, while others did not. Yet in other cases, health care providers did not recognize that their actions contributed to adverse outcomes (e.g., Weick & Sutcliffe, 2003). Hospitals formally designate decision-making forums for organizational learning from preventable injuries, near injuries, and deaths. These include physicians’ Morbidity and Mortality sessions (e.g., Lipshitz & Popper, 2000) and hospital-wide Root Cause Analysis meetings (e.g., Bagian et al., 2002). Nurses and nurse managers routinely participate in Root Cause Analysis sessions; however, they do not have a tradition of reviewing signiﬁcant deaths and injuries, comparable to the physicians. Nurses engage in organizational learning activities at the unit-level and these activities can be distributed unevenly among various patient care units in a single hospital (e.g., Edmondson, 2004). In the current study, organizational learning is deﬁned as a process in which decision makers weigh the organization’s experience as a basis for changing the routines that will guide future behavior (Levitt & March, 1988; March, 1999). This deﬁnition emphasizes that learning is a process, but it does not necessarily result in change or improvement.

METHODS Using qualitative research methods (Miles & Huberman, 1994), we examined interviews with nurse managers who worked in four critical care units located in three tertiary care teaching hospitals. The research design

73

Promoting Patient Safety

used semi-structured interviews supplemented by workplace observations of health care providers, allowing us to focus on the details of the hospital. We contextualized our research design primarily through rich description (Rousseau & Fried, 2001). We concentrated on examining the medication process from distinct professional viewpoints; in this study the focus is on nurse managers. This allowed us to gather rich, nuanced descriptions (e.g., Weick, 2007) of the methods that hospitals used to monitor nurses’ medication errors, as seen from the nurse managers’ perspectives. Speciﬁcally, we identiﬁed key exemplars that illustrate the role of nurse managers in monitoring and assessing nurses’ errors. The purpose of discussing these exemplars is to raise theoretical and practical questions that can guide further research.

Study Sample This study is part of a research project that examines how three hospitals learn from their experience how to improve medication safety. The larger research project focuses on drug-related errors because adverse events (i.e., involving patient harm) resulting from medications are some of the most common and costly errors (Aspden, Wolcott, Bootman, & Cronenwett, 2007; Bates et al., 1995; Thomas et al., 2000). The investigators used three criteria to select hospitals and sub-units within them (the hospital pharmacies and four critical care patient units). The three hospitals had a high volume of medication usage and a high potential for adverse events due to the complexity of their patients. The hospitals also had implemented medication safety programs, including reporting systems, indicating their efforts to reduce errors. The pseudonyms for the three hospitals are West Hospital (WH), South Hospital (SH), and North Hospital (NH). The larger research project was based on interviews with study participants who worked at different levels in the organizational hierarchy and represented diverse professional groups. The study was designed to interview a purposeful sample of key hospital administrators and middle managers and a ‘‘purposeful random sample’’ (Patton, 2002, p. 141) of health care providers, including nurses, physicians, and pharmacists. The objective was to reduce bias and enhance the credibility of the small sample, not to generalize from it. The larger research included 341 participants, representing 86 participants in WH, 98 in SH, and 157 in NH. In the three hospitals, 18 people declined to participate and two interviews were excluded due to recording malfunctions.

74

MICHAL TAMUZ ET AL.

The study reported here focuses on interviews with a purposeful sample of 20 nursing middle managers from four critical care patient units in three hospitals. The sample was designed to include all of the managers who were responsible for supervising nurses in the four critical care units. The study is based on interviews with 6 middle managers from a unit in WH, 3 from a unit in SH, and 11 from two units in NH, including 6 responsible for one critical care unit and 5 responsible for another. All of the managers who met the purposeful sample criteria were interviewed and no one declined to participate. The nurse managers served in middle management positions between the top hospital administrators, such as the Vice President of Nursing, and the frontline staff nurses, who were almost all registered nurses. The highest level middle management positions were Directors of Nursing, who were responsible for several patient care units (including the critical care unit we studied), and who reported to a hospital vice president. The second level middle management positions were unit managers who were responsible for the nursing staff in one critical care unit. The third level positions were assistant managers who reported to the unit manager and worked as frontline supervisors on each shift, in a speciﬁc critical care unit. To de-identify the study participants, we use the term ‘‘nurse manager’’ when referring to all three levels of middle management nursing positions. The study reported here does not include ‘‘charge nurses’’ (i.e., experienced lead nurses) because many of the registered nurses rotated positions, acting alternatively as charge nurses or staff nurses.

Data Collection and Analysis We used a general interview guide (Patton, 2002) in which we asked participants to describe hospital reporting systems and programs for monitoring medication safety as well as to discuss their own experiences with medication problems. Interviews with the nurse managers averaged about 90 min, ranging in length from approximately 1–6 h (across multiple interview sessions). Nurse managers who gave longer interviews were not over-represented because the study did not calculate the frequency of medication errors or other related events. We examined how the nurse managers described their roles in identifying, responding to, and making decisions about possible medication errors. Interviews were audio-recorded, transcribed, and checked for accuracy of transcription. Field notes,

Promoting Patient Safety

75

document review, and observations of routine activities supplemented the interviews. In the study reported here, we conducted a ‘‘framework analysis’’ of the interviews with nurse managers to identify the nurse managers’ roles in and their methods of making decisions about monitoring and assessing nursing errors (Green & Thorogood, 2004). Speciﬁcally, we sought to identify key exemplars that could illustrate these roles and decision-making processes. Applying framework analysis, two research team members familiarized themselves with the data by reading the transcripts and discussing ﬁeld notes. They constructed a thematic framework by focusing on a priori study objectives and previous research as well as themes that emerged from the data. Speciﬁcally, they identiﬁed speciﬁc issues that the study participants raised (e.g., mixed messages) and took note of patterns in the activities that the study participants described (e.g., decision-making guidelines). The thematic framework was also based on key concepts derived from previous research: production pressures, accountability, and learning. Research ﬁndings suggest that staff nurses refrain from disclosing errors because of fears of reprimand (i.e., accountability) and lack of time (i.e., production pressures) (e.g., Wakeﬁeld, et al., 1999). Unit-level nurse managers’ actions can intensify (or alleviate) the nurses’ concerns and thus, inﬂuence the disclosure of error data necessary for organizational learning (e.g., Edmondson, 1996, 1999). The investigators used the thematic framework to select and analyze excerpts from the transcripts that illustrated error monitoring. The research team also looked for counter-examples that contradicted the coding categories in the thematic framework. The team analyzed the data by comparing within and across the exemplars. Two researchers grouped the data (i.e., transcript excerpts) into exemplars and looked for common and distinctive themes in each of the exemplars. A third team member reviewed the preliminary analysis and interpretation of the exemplars to assess whether the data supported the interpretation and to assure deidentiﬁcation. The purpose of the data analysis was to reﬁne and generate research questions and propose a conceptual framework to guide future studies. To protect the conﬁdentiality of the data, the investigators removed details that could identify the research site and study participants. Pseudonyms disguise the names of the hospitals, patient care units, and the speciﬁc position titles. We also use a pseudonym for the External Review Committee (ERC) and do not specify the regulatory agency that enacted

76

MICHAL TAMUZ ET AL.

the committee. Although the nurses and nurse managers included men and women, in the results, male pronouns are used to refer to all study participants. In excerpts from the interview transcripts, the text in brackets was changed to de-identify participants or added to clarify the contents. The Institutional Review Boards designated by the hospitals and the universities approved the study.

RESULTS Error-Monitoring Systems In the three study hospitals, nurse managers were responsible for holding the staff nurses accountable for their errors. The emphasis on maintaining accountability in nursing came from multiple sources. The hospital expected nurse managers to ensure that nurses reported errors through the hospital’s voluntary incident reporting system. Nursing norms guided staff nurses to report errors to their charge nurse – usually an experienced lead nurse who assumed administrative duties for a shift in a particular unit. After consulting with (or being directed by) the charge nurse, the staff nurse might ﬁle a report with the hospital incident reporting system. Consultation with the charge nurse was customary, but not required. In each of the hospitals, a nurse could ﬁle an incident report directly to the reporting system. Indeed in SH, health care providers could ﬁle reports anonymously. Nurse managers were also expected to follow the mandates of two disciplinary systems: one system based on professional norms and standards and another separate due process system established by the hospital and common to many bureaucracies (Edelman, 1990). At the hospital level, the human resource department implemented the customary due process measures to assure a fair and equitable disciplinary system. For example, before dismissing an employee, the hospital must demonstrate that it provided adequate verbal and written warnings. This system not only applies to deviations from standard hospital routines (e.g., tardiness), but also to nurses’ medication errors. State-level regulators also implemented an ERC designed to protect patient safety by holding nurses accountable for their mistakes. Hospitals were required to report a nurse to the ERC if he made three minor errors or one major error. According to the ERC classiﬁcation system, the category of minor errors included medication errors in which the patient

77

Promoting Patient Safety

was unharmed. Before the hospital reported a nurse to the ERC, the hospital convened an Internal Review Committee (IRC). The committee, comprised of representatives of hospital nurses, assessed the particular circumstances of the case and decided whether the nurse should be referred to the ERC. Although the nurse managers were bound by sometimes overlapping sets of bureaucratic rules, the hospital enabled them to use professional discretion when applying these rules. Nurse managers described how they could choose one or more of the following options: record the event in the nurse’s personnel ﬁle, relay the report to the hospital incident reporting system, report the event to the hospital disciplinary system, or initiate an internal review by the IRC for possible referral to the ERC. Each of the following three exemplars illustrates how nurse managers monitored errors, while buffering and balancing the pressures to hold nurses accountable for their actions. The ﬁrst exemplar focuses on the strategies that nurse managers used to deal with hospital incident reporting requirements; the second examines how some of these managers kept two sets of books; and the third considers how managers responded to pressures from the ERC.

Exemplars This section includes three exemplars based on the interviews with nurse managers. Each exemplar begins with a description of the error-monitoring practice from the perspective of the nurse managers; the exemplar concludes with the investigators’ analyses of the themes related to production pressures, accountability, and learning, where applicable, as well as the description of the roles of the nurse managers.

Exemplar 1: Promoting Incident Reporting The ﬁrst exemplar focuses on managers’ strategies for responding to hospital incident reporting requirements. The nurse managers developed idiosyncratic criteria for classifying which medication errors constituted reportable incidents. Some managers inadvertently generated mixed messages about whether nurses could be disciplined for their mistakes, while others ﬁltered the incident reports that were relayed from the nursing unit to the hospital incident reporting system.

78

MICHAL TAMUZ ET AL.

Mixed Messages Nurse managers expressed their awareness of the staff nurses’ concerns about reporting their own errors or the errors of fellow nurses. They also addressed the efforts by hospital administrators to allay the nurses’ fears and explain the hospital’s non-punitive policy towards making errors. A nurse manager vividly described this situation in WH. They’re [the nurses] scared. I mean, because, when you make an incidentywhen you do something wrong, y that really should be written up; but you know what? You think to yourself, ‘‘Gosh! I’m really scared. You know, it didn’t do anything to the [patient], nobody knows, I’m just not gonna do it. I’m scared. Maybe I could get ﬁred y’’ So they don’t want to report themselves, either.

According to the managers, the nurses’ fear of reporting also extended to reporting their nursing co-workers: They don’t want to get the person in trouble; they don’t want to send the person to [Internal Review Committee] or think a person could potentially be terminated. You know, the only thing, and I’m not trying to be mean, but it’s ignorance. That is why they would not write somebody up.

As the interview continued, the manager explained further. He described the institution’s challenges in encouraging nurses to document nursing errors ofﬁcially so that the hospital could understand the full range of errors. Because, if they truly understood what the process was, they would know that it has nothing to do [with punitive actions] y because, the hospital has tried over and over again. We’ve given in-services; we’ve had Risk Management come y to try to get over to them that the incident report is not anything punitive. I mean, occasionally, once you follow up, and if there is a trend, there may be a punitive thing that does happen to the nurse.

This illustrates how, during the lengthy interviews, nurse managers contradicted themselves in describing the potential punitive consequences of nurses reporting mistakes. Indeed, nurse managers explained that they usually initiated disciplinary measures if a nurse repeated similar mistakes. One nurse manager was acutely aware of the mixed message that he and other health care providers had been sending to the nurses: It’s kind of double–talk or something, because then one side is saying it’s not punitive, but the other side you’re saying, ‘‘Well, they can get counseling, so that can get referred to [the IRC]’’ y It does sometimes not make sense, because there is situations where it could become punitive.

Promoting Patient Safety

79

These comments suggest that the nurse manager inadvertently sent mixed messages to the staff nurses regarding the punitive response to incident reporting. The hospitals publicly espoused a non-punitive policy towards incident reporting. But if nurses repeated similar mistakes, they could be subject to internal and external disciplinary procedures. These disciplinary measures gave nurses disincentives for reporting their own errors and those of other nurses. They also created the conditions for sending mixed messages to the staff nurses. On one hand, nurse managers instructed nurses that all incidents should be reported so that the hospital could have a better understanding of the range of incidents occurring. On the other, the managers also noted that there was an unspoken range of acceptable and unacceptable frequency and types of incidents that could jeopardize a nurse’s career. Filtering. The managers indirectly ﬁltered the events that staff nurses reported to the hospital by excluding those errors that the managers consider to be non-reportable. Nurse managers developed their own idiosyncratic criteria for which errors should be classiﬁed as reportable or non-reportable incidents. The managers’ various rules of thumb for classifying reportable incidents illustrate how nurse managers reinterpreted the hospital guidelines for incident reporting. Based on the managers’ descriptions of reportable incidents, the investigators described the criteria as including the potential and actual severity of the patient injury, speciﬁc types of medication errors, how quickly the error was detected and corrected, and the involvement of others in the event. Nurse managers considered the potential harm to the patient. An NH manager stated, ‘‘It depends on the severity.’’ An SH manager clariﬁed further, ‘‘There are classes of mistakes that I don’t bat an eye at. They’re too low risk to be of any real concern to me.’’ However, the managers’ classiﬁcation criteria also extended beyond concerns about the potential harm to the patient. One of the NH managers focused on speciﬁc infractions that constituted a reportable incident: ‘‘If they omitted some medication [y] and it’s in the order, then they have to make an incident report that it was omitted.’’ An SH manager focused on how quickly the nurse noticed and corrected their mistake: If you hang a brand new bag, and you look at it and then you say, ‘‘Oh my goodness, this is the wrong concentration, let me change this real quick in the pump.’’ No, they wouldn’t [report] that. If it’s something that is not rectiﬁed pretty much immediately, and that they didn’t catch immediately, then they [ﬁle an incident report about] it.

80

MICHAL TAMUZ ET AL.

An SH manager focused on the involvement of other health care providers in the error. If nurses on another shift were involved in an error, he classiﬁed the event as a reportable incident: ‘‘If they [the nurses] come in and it’s a change of shift issue and they ﬁnd something that happened on the prior shift, they have to [ﬁle an incident report on] that.’’ Similarly, if physicians were informed about the error, it is a reportable incident: ‘‘If it’s a situation where intervention was required, that physicians had to be notiﬁed, then absolutely they must [ﬁle an incident report on] that.’’ In this case, physician involvement also signals that the event was serious. Furthermore, an NH manager used patient involvement as an event classiﬁcation criterion: Because a nurse would come to me and would say, ‘‘Gosh I did this’’ and they really feel bad about it then I say, ‘‘Then we need to do [an incident report]. It’s not to get you in trouble but we need to in case something happens, especially if the patient knows.’’ If the patient knows that a mistake was made, we need to do [an incident report].

All of these examples reﬂect the managers’ expectations of what constitutes a reportable incident. They are also consistent with the staff nurses’ accounts that they often consult with a nursing supervisor, before submitting an incident report. Once ofﬁcially documented, not all incident reports were relayed from the speciﬁc nursing unit to the hospital reporting system. A nurse manager described how a nursing management team directly ﬁltered an incident report: I know a nurse who wrote himself up, because he had a delay on a [hazardous drug] dose. He was supposed to give it like 9 o’clock in the morning and he didn’t give it until like 4 o’clock in the afternoon, and he wrote himself up. I was like, ‘‘What are you doing? Why did you [report] yourself on that?’’ He was like, ‘‘Well, because I was late with it; and it was wrong, and if something happens to that patient, if he gets a DVT [deep vein thrombosis] or something, you know, it’s my responsibility.’’ I am like, ‘‘Okay, you are just a straight up [person], you know.’’

When asked what would happen to the incident report written by that nurse, the manager indicated that it would be given to the nursing management team; they would make the ultimate call on whether to send it on to the hospital’s incident reporting system. The manager noted, however, that it would likely go no further than the team and that it would not go to the hospital’s reporting system. In contrast, nurse managers in the same hospital stated that if an incident was ﬁled on someone in their unit by another

Promoting Patient Safety

81

department or if a nurse from their unit ﬁled an incident on someone in another department, then the incident ‘‘has to go through the proper channels.’’ Analysis In the ﬁrst exemplar, production pressures did not appear to be salient. However, accountability and learning issues were signiﬁcant. The nurse managers also engaged in buffering and balancing roles. Accountability The case of the mixed messages highlights the nurse managers’ responsibility to monitor their staff nurses’ performance. The hospitals declared a nonpunitive policy towards incident reporting, but hospital administrators continued to hold the nurse managers accountable for insuring that their staff followed the rules. Nurse managers developed informal guidelines for specifying the conditions under which they would hold staff nurses accountable for reportable incidents. As could be expected, the managers were more likely to hold nurses accountable if the patient actually was harmed or could have been harmed under slightly different circumstances. Public disclosure of the error also seemed to increase the likelihood that nurse managers would hold nurses accountable for an error and require that they submit an incident report. Learning This exemplar underscores several possible impediments to learning. When nurses hear mixed messages about ﬁling an incident report, they might continue to be fearful of reporting their own errors and those of other nurses. Thus, the hospital would have fewer cases from which to learn. Furthermore, because the ofﬁcial hospital guidelines for reporting incidents differed from the nurse managers’ idiosyncratic reporting criteria, it could limit the ability of nurses to learn when incident reporting was required. When the nurse managers directly or indirectly ﬁltered event reports, it reduced the hospital’s capacity for organizational learning. Because the nurses and nurse managers did not report all relevant errors, it limited the amount of data available to the hospital from which it could observe patterns or draw conclusions. Nurse Manager Roles The use of mixed messages clearly illustrates the nurse manager’s role in balancing contradictory hospital-level policies. On one hand, the hospital

82

MICHAL TAMUZ ET AL.

maintains a non-punitive policy towards incident reporting, while on the other, hospital administrators expect that nurse managers will enforce compliance with hospital rules and procedures. Nurse managers translated these policies into an informal guideline in which a nurse usually would be disciplined after repeating the same mistake – but not as the result of an isolated error. Nurse managers saw themselves as protecting staff nurses when they ﬁltered out what they perceived as non-hazardous infractions. In doing so, the nurse managers buffered their staff from the possible future intensiﬁcation of disciplinary measures. If a nurse conscientiously reported every error he thought to be important, he would strengthen his local professional reputation within the patient care unit, but could undermine his ofﬁcial standing in the hospital. In the future, if he were to be involved in an error that caused harm to a patient, his history of previous mistakes might increase the severity of the hospital’s disciplinary measures. The nurse managers also served as informal consultants to the staff nurses. Although the nurses were not required to obtain approval by the charge nurse or a nurse manager before ﬁling a report, the staff nurses often conferred with them before submitting an incident report. This informal consulting also occurred in SH even though health care providers had the option of ﬁling anonymous reports. Consider that the incident reporting systems in the three hospitals were not computerized at the time of the interviews. Computerized incident reporting may have reduced the opportunities for nurse management teams directly to ﬁlter out the incident reports they perceived as inconsequential. However, the electronic ﬁling of incident reports would be unlikely to eliminate the informal consulting between staff nurses and nurse managers, when deciding whether an event warrants submitting a formal report.

Exemplar 2: Keeping Two Sets of Records The second exemplar examines how nurse managers kept two sets of error records – a formal one, available to hospital administrators, and an informal one for the nurse managers’ usage. The formal, ofﬁcial ﬁle contained a subset of a nurse’s errors and related disciplinary measures; the informal ﬁle included a complete record of all of the nurse’s (reported) mistakes. Nurse managers developed idiosyncratic guidelines for determining which errors should be placed in the ofﬁcial permanent record or the informal one.

Promoting Patient Safety

83

Members of the nurse management teams in SH and WH spontaneously described remarkably similar processes for keeping two sets of records. Nurse managers could decide when and how to discipline nurses for making medication errors. As a nurse manager at WH explained: I don’t think that every time something happens with a medication error that I have to write them up. I think I have the option to, you know, choose how I’m going to handle the situation.

An SH nurse manager shared the view that ‘‘every medication error is different.’’ However, nurses and managers in all three hospitals consistently stated that in the case of patient harm, nurses must report the injury to their supervisors and their supervisors must formally record the injury. ‘‘There’s some things that you have no option, that you would have to write them [the nurses] up if there was adverse reaction to the patient or something,’’ a WH nurse manager explained. The nurse managers developed guidelines or rules of thumb for deciding whether or not they would take formal disciplinary measures (i.e., by ‘‘writing-up’’ a nurse). Nurse managers in different hospitals developed comparable guidelines. For example, a WH nurse manager speciﬁed one of his guidelines: I would not write somebody up if it was like a circumstance-type thing. I mean, obviously, if. I guess I might have to write somebody up if it was like a major thing, and then it caused adverse reaction to the [patient], I would have no option.

A nurse manager in SH also took into account the potential danger to the patient and was wary of errors ‘‘that can affect a patient so drastically; they can kill someone.’’ On the other hand, the same nurse manager was also attentive to the circumstances under which the error occurred: So you didn’t tell me you gave your medication an hour late, big deal. I think that’s going overboard to tell me that you gave something an hour late. Sometimes we’re three hours late in there. There’s just no getting to it. You can’t get to it and you’ve got to prioritize.

When assessing an error, nurse managers also took into consideration conditions that were beyond the control of the nurse: That’s when you would write in your investigation that this was a totally unreasonable assignment, or it was easy and then it just fell apart, and we didn’t have enough staff. There was nothing they could do; they did the best job they could do.

84

MICHAL TAMUZ ET AL.

When deciding how to record an error, the nurse managers emphasized the importance of taking into consideration not only the patient outcomes, but also the circumstances that could have contributed to a nursing error. Some managers explained that the perceived severity of the infraction inﬂuenced their decision whether or not to include a report to the nurse’s permanent ﬁle. For example, if the nurse manager considered the nurse’s infraction to be inconsequential, then he did not record the counseling session in the nurse’s ﬁle. To illustrate, an SH manager recalled his conversation with a nurse: [Nurse:] ‘‘I gave my [medication] one hour late; I can’t believe I missed it.’’ [Manager:] ‘‘Please try and write yourself a note, let yourself know what times all your meds are due y’’ That’s not going in their permanent ﬁle.

However, if the error could potentially harm the patient, this same SH manager would not only talk with the nurse, but also record the event in the nurse’s ﬁle. It depends on what happened. Sometimes I just call them in and talk to them and tell them that this is the situation, this is what happened. It depends. If they gave 40 of [hazardous medication] instead of 20 of [hazardous medication], and they realized it [y] There was no intervention required with the patient, everything was okay, then it just goes in their ﬁle as, ‘‘This was a medication error.’’ It’s like a verbal warning and they have to be conscientious of that. And then usually you ﬁnd that they’re much more careful.

These examples show how nurse managers differentiated between formally choosing to record a warning in a nurse’s ﬁle and using informal methods to record a nurse’s errors. When nurse managers monitored mistakes that did not warrant a formal warning, they devised informal methods to remember whom they had coached or counseled about mistakes. The nurse managers, however, differed in the informal strategies they used. One nurse manager explained that his management team developed a spreadsheet to keep an informal record of all of the nurses’ errors: If I looked at it [the event] and I saw, I thought it was extremely unreasonable, I would not write them up. I would probably put it on the spreadsheet, which the spreadsheet is just ours; nobody sees that, I don’t send it out to anybody, that’s just for the management team so we can just glance at it and see: ‘‘Is there people that we need to be following up with?’’ that kind of stuff.

Promoting Patient Safety

85

The nurse managers also used informal methods to keep track of counseling sessions, as a WH nurse manager explained: ‘‘You might write it on a paper towel and put it in their [the nurse’s] ﬁle, so it’s an unofﬁcial thing.’’ Indeed, some managers described relying on their memories: ‘‘If something happens and say it’s little bitty tiny things, and, ‘Oops I caught this, Oops I caught this’ – and it happens several times in a row, sure you remember that.’’ They referred to their memories to keep track of positive events (i.e., nurses catching something before it became a problem) as well as more negative ones (i.e., nurses who had made minor errors that did not warrant an incident report). These examples illustrate how the nurse managers kept two sets of books regarding medication errors. They kept an informal set of records comprised of all of the nurse’s (reported) medication errors, including apparently insigniﬁcant infractions. They also formally assembled a list of counseling sessions and written warnings about potentially signiﬁcant errors in the nurse’s permanent ﬁle. The nurse managers were concerned about the possible consequences of recording all of the nurse’s errors in his permanent ﬁle. An SH nurse manager noted that the formal documentation of all medication errors could discourage nurses from reporting important, potentially dangerous events: If you have a ﬁle full of verbal warnings because of little bitty incidents that have happened, I think that would be discouraging to them [the nurses] that later they might not want to report something that should actually be reported.

Furthermore, a nurse’s conscientious error reporting could also undermine his professional reputation and misrepresent his history of previous errors, should the nurse be reported to the IRC: Their ﬁle goes to [the IRC]. So if there’s other incidents where they’ve made medication errors in their ﬁle, that’s reported in the [IRC]. And so they [the IRC] know: Okay, this is their fourth time that they’ve had a medication error; or this is their ﬁrst.

In case the nurse was involved in additional errors, the nurse managers and hospital administrators would use the ofﬁcial ﬁle to establish the nurse’s history. Thus, this nurse manager was aware of the long-term implications of formally recording an error. Analysis In this section, we examine the managers’ concerns with production pressures, accountability, and learning. The nurse managers’ roles are also considered.

86

MICHAL TAMUZ ET AL.

Production Pressures This exemplar illustrates how nurse managers balance the demands for stringent rule enforcement and accountability with the production pressures on staff nurses caring for patients. The managers’ concern for production pressures is reﬂected in their attentiveness to the circumstances under which nurses made errors. Consider the SH nurse manager who explained that nurses might not be able to administer drugs on schedule because ‘‘You can’t get to it and you’ve got to prioritize.’’ When deciding how to respond to a nurse’s error, the managers took into consideration the extent to which work conditions contributed to the error. Thus, if the error occurred due to circumstances beyond the nurse’s control, it would inﬂuence the manager’s decision whether or not to record it ofﬁcially. Accountability Keeping two sets of records clearly involves accountability issues. But it also raises issues of accountability to whom? Managers seem almost always to hold a nurse accountable to himself – the managers consider it important to inform the nurses about the mistakes they have made. However, the managers have developed ﬁnely calibrated guidelines for deciding whether a nurse should be held accountable only to the nursing unit’s management team, or also to the hospital, and possibly to the ERC. Learning The nurse managers did not explicitly discuss organizational learning, however, they emphasized that it was important to let a nurse know that he made a mistake, implicitly, so that he could learn from his experience. Therefore, an individual nurse could learn from his error, regardless of whether it was recorded in the ofﬁcial ﬁle. By keeping informal records, the nurse managers not only kept a complete history of an individual nurse’s reported errors, but also could discover problems that were common to other nurses. Indeed, several managers mentioned keeping a running tab in their memory of mistakes that had occurred in the unit. When they noticed that different nurses made the same mistake, the pattern of mistakes called their attention to the problem, and they began investigating whether there was a system problem. Thus, the ability of the management team to learn from their experience within the patient care unit did not appear to be hampered by keeping two sets of records.

Promoting Patient Safety

87

The practice of keeping two sets of records limited how the hospital learned from nursing mistakes. The hospital could learn from nursing mistakes that were reported to the incident reporting system or that the nurse managers discussed with risk management or other hospital administrators. If the nurse management team kept informal tabs on mistakes within the unit, this information was not available to other nursing units or at the hospital level. Thus, the lack of information could hinder the discovery of emerging error patterns, especially in events that rarely occurred in one patient care unit, but that could crop up across the hospital.

Nurse Manager Roles The investigators observed that the nurse managers fulﬁlled three key roles. First, the nurse managers developed decision-making guidelines for classifying which events should be handled formally or informally. The guidelines included one set of criteria focused on the patient and the second, on the nurse and other health care providers. On one hand, the managers assessed the actual and potential harm to the patient and whether the error required that providers give the patient additional treatment. On the other, they considered mitigating circumstances, such as the nurse’s workload and whether the nurse noticed and corrected his own error. Second, they buffered the nurses from hospital’s disciplinary procedures. If managers had ofﬁcially recorded all of the nurse’s benign mistakes, their ofﬁcial ﬁles would be full of warnings. In case a nurse was later involved in an adverse drug event, in which a patient suffered a preventable injury from a medication error, then this history of warnings might prompt the hospital to take severe disciplinary action. Third, the nurse managers balanced multiple, and sometimes conﬂicting goals espoused by the hospital administrators: to ensure that employees followed hospital and professional procedures, to hold employees accountable for their mistakes, and to refrain from blaming health care providers for system failures. By keeping dual records, nurse managers could comply with these different objectives by addressing them one at a time. To illustrate, they could maintain accountability when they classiﬁed events as reportable and warranting a formal disciplinary notice. They could refrain from blame when they informally recorded an error and fostered the nurse’s learning. They could address system failures when they tracked patterns in informal records of circumstance-based errors and thus, promoted organizational learning within the patient care unit.

88

MICHAL TAMUZ ET AL.

Exemplar 3: The External Review Committee In the third exemplar, nurse managers responded to pressures from an ERC. Cases of serious patient harm were invariably reported to the ERC. For less harmful mishaps, the managers developed routines for classifying mishaps into reportable or non-reportable categories. If the nurse managers considered an event reportable, they were required to report it to the hospital’s IRC for possible referral to the ERC. When managers classiﬁed a signiﬁcant patient safety-related event into a non-reportable category, they did not report the event to the IRC or ERC. Tamuz and Thomas (2006), in their previous analysis of the ERC, described this situation as ‘‘classifying away dangers.’’ The following discussion draws upon Tamuz and Thomas’s description of the ERC, while adding additional detail about how nurses responded to it and providing a more extensive analysis. The ERC regulations speciﬁed that hospitals report every time a nurse makes a major error or a series of three minor errors; however, they did not provide detailed instructions for implementing the regulations. Nurse management teams could exercise discretion in determining which situations were reportable, as a nurse manager explained: Because I don’t think there’s anything really clearly spelled out. I mean, there’s not a policy that says, ‘‘If this happens, this is a minor, you must do this; if this happens, it’s a major, you must do this.’’ So it’s kind of left up to you as a leadership team to decide how are you gonna handle these situations and what are you gonna do? (Tamuz & Thomas, 2006, p. 931).

Nurse managers met as a team to discuss how to respond to particular errors. A nurse manager explained, ‘‘Usually we discuss what’s the best way to handle it, depending y I mean, every [incident] is different.’’ A different management team followed a similar routine, as a manager noted: ‘‘Most of the time when we have incidents like this, the managers and I will discuss them.’’ Thus, nurse management teams could decide how to classify and respond to each error report. The managers not only assessed the circumstances surrounding the event, but also considered the nurses who were involved. ‘‘It depends on the circumstances surrounding it. It depends on the signiﬁcance of the error. It depends on a lot of things, a lot of factors,’’ a manager noted. A manager in a different hospital described how the team worked together ‘‘to protect this nurse and educate this nurse and never have this happen again.’’ Consider an instance where a nurse administered the wrong drug intravenously and after repeated medical interventions, the patient recovered from the error.

Promoting Patient Safety

89

The management team gave the nurse a verbal warning, assigned him to complete a study plan, and required that he sign-off particular medication orders with another nurse for six months. Thus, they reported the event through the hospital disciplinary system (e.g., verbal warning), but not to the ERC. A manager emphasized, ‘‘Everybody deserves one chance’’ and this nurse ‘‘is a very conscientious person.’’ This example illustrates that the management teams could exercise considerable discretion in deciding how to maintain accountability. Nurse management teams followed guidelines for classifying events as reportable (or non-reportable) to the ERC; however, the teams developed different classiﬁcation criteria. To illustrate, NH team guidelines concentrated on nursing process and SH guidelines were based on system issues. An NH management team did not consider an error as reportable to the ERC – if the nurse followed the appropriate process. For example, a manager recounted how a nurse gave a patient the wrong dose of medication and the patient was unharmed: She had juxtaposed the numbers. This nurse had charted as she went. She looked at this, she printed, she highlighted [y.] She had gone through all her steps, but she had made a mistake (Tamuz & Thomas, 2006, p. 932).

The nurse manager continued to explain how he queried the nurse to decided whether he was following accepted procedures: ‘‘Tell me what you were thinking. Talk me through your process.’’ The stuff where they [the nurses] have logical process and it came to a bad outcome – they’re comfortable that they will get support. If their process was sound (Tamuz & Thomas, 2006, p. 932).

Thus, if a nurse went through the correct process and it resulted in a benign error, the NH management team classiﬁed the event as non-reportable to the ERC. An SH nurse management team considered errors resulting from system issues as non-reportable to the ERC. However, if the team attributed an error to an individual nurse’s lack of skill or knowledge, than they classiﬁed the event as reportable. Therefore, the team investigated events to decide if they were caused by a nurse’s inadequacy (e.g., a lack of knowledge or skill) or by a system failure (e.g., a pneumatic tube malfunction or understafﬁng). To decide whether an error could be attributed to system problems, team members asked: ‘‘If any other nurse had been in this situation, would he have made the same mistake?’’ If the team concluded that anybody could do this and it could happen again, then they classiﬁed the mistake ‘‘as a system problem’’ and ‘‘not an individual problem.’’ Because the managers

90

MICHAL TAMUZ ET AL.

concluded that the error resulted from a system problem, they did not consider it reportable to the ERC. Analysis The nurse managers’ methods of classifying away events relates to issues of production pressures, accountability, and learning. It also illustrates their roles in balancing conﬂicting objectives. Production Pressures The nurse managers mentioned production pressures as a possible cause of system-based errors. For example, if there were an insufﬁcient number of nurses on duty, it may have contributed to errors. However, the context leads the investigators to speculate that production pressures also might underlie the reasoning behind some of the managers’ reluctance to report a ‘‘conscientious nurse’’ to the ERC. There was a nation-wide nursing shortage in the US and these nurse managers may have been concerned about retaining their ‘‘good nurses.’’ Accountability Maintaining accountability is a central theme in this exemplar. Maintaining accountability seems like it should be objective and rule-based, but it was subjective in practice – based in part on nurse managers’ estimates of who was a ‘‘good nurse.’’ Furthermore, nurse managers sought to hold the staff nurses accountable for their mistakes, but without having to report them to the ERC, unless it was unavoidable. The nurse managers also had to maintain their own personal accountability. As an NH nurse manager emphasized, he was required to report nurses to the ERC and must carry out his responsibility; ERC reporting was mandatory. However, in another NH unit, the nurse managers rarely classiﬁed an event as reportable to the ERC. The staff nurses in this unit usually did not mention the ERC during the research interviews, and when the interviewer directly questioned some nurses about it, they did not express concerns. These examples suggest that by classifying away errors as non-reportable, nurse managers technically could follow the rules and maintain their own responsibility as managers. Learning Classifying away potentially dangerous events did not prevent learning within a nursing unit, but it also did not promote organizational learning. Some nurse managers fostered learning, while others did not. In both NH and SH, even if the nurse managers did not classify the event as an error

91

Promoting Patient Safety

according to ERC criteria, the data remained available to them. The nurse and the nursing unit could learn from the event and, as SH managers described, change nursing practices within the unit. Although classifying away dangers did not preclude learning within the management team, the relevant information usually was not shared with other nursing units in the hospital, limiting the potential for organizational learning at the hospital level. Nurse Manager Roles Nurse managers clearly were buffering the staff nurses from pressures from the ERC. For example, some nurse managers thought it was unrealistic for the ERC regulations to stipulate that every medication error, regardless of potential severity, constituted a minor error; and if a nurse made three minor errors, he should be reported to the ERC. If this regulation had been followed precisely, one manager ventured, then he would have to report almost all the nurses to the ERC. The ERC example also illustrates how nurse managers aimed to achieve multiple hospital objectives. They balanced the hospital goal of assuring that employees follow hospital procedures and the goal of maintaining a non-punitive environment for health care providers who report their mistakes. The nurse managers also confronted the dual objectives of protecting patient safety, while protecting nurses from being blamed for errors that primarily were caused by hospital shortcomings. In addition, the state-level regulators expected nurse managers to meet their obligations to refer a nurse to the ERC if he made three minor errors in a year. To avoid breaking the rules, it appears that the nurse managers reinterpreted how to apply them. They did not fail to report nursing errors, but simply classiﬁed these errors in non-reportable categories.

DISCUSSION Nurse managers are charged with fulﬁlling multiple, and sometimes conﬂicting objectives. They are expected to hold staff nurses accountable for medical errors and also learn from the nurses’ mistakes. These objectives come into conﬂict when the same data are used as a basis for disciplining nurses and as a source of learning. The conﬂict is also expressed in the difference between the espoused goals of maintaining a non-punitive environment and the expectations that managers will discipline nurses who repeatedly make mistakes.

92

MICHAL TAMUZ ET AL.

Nurse managers grappled with this tension in different ways. The ﬁrst exemplar focuses on the incident reporting system. It illustrates how some nurse managers sent mixed messages to the staff nurses. Other managers directly and indirectly ﬁltered the events that were reported. In both exemplars two and three, the nurse managers developed classiﬁcation systems for evaluating each situation and determining whether it was a reportable event. Based on the event classiﬁcations, the nurse managers enacted routines for handling the situations. In the second exemplar, managers created idiosyncratic classiﬁcation schemes to determine whether an event was reportable to the hospital disciplinary system. The classiﬁcation criteria focused primarily on whether the patient was harmed, whether additional medical intervention was required, and whether the nurse noticed his own mistake. Based on these criteria, managers handled the event formally or informally, resulting in two sets of records. In the third exemplar, nurse management teams constructed classiﬁcation guidelines to determine whether an event was potentially reportable to the ERC. Events that were considered non-reportable were handled within the unit; they were not reported to the hospital-level IRC that could determine if the event warranted reporting to the ERC. When nurse managers classiﬁed away potential threats to patient safety as non-reportable, they did not preclude learning within the unit. However, they limited the potential for organizational learning at the hospital level. Similarly, these exemplars highlight two critical tensions that inﬂuence nurse managers’ efforts to promote patient safety: (1) applying bureaucratic rule-based standards to control professionals’ behavior; and (2) using the same information as a basis for maintaining accountability and learning.

Bureaucratic Rules and Professional Tasks Relatively little attention has been focused on the consequences of and conﬂicts inherent in using bureaucratic rule enforcement to guide and control professional nursing behavior. The lack of attention is especially puzzling because of the extensive social science literature on this topic (e.g., Blau & Scott, 1962; Hall, 1968; Scott, 1965, 1969). Researchers have long recognized the consequences of subjecting professionals to bureaucratic controls. Furthermore, similar issues were addressed in classic studies of alternative means of assessing employee performance in organizations, such as monitoring performance by rule compliance, processes, or outcome measures (e.g., Scott, 1977).

Promoting Patient Safety

93

This raises research questions related to both theory and practice. Perhaps some of the difﬁculties of determining ‘‘what is an error’’ should not be attributed to disagreements over semantics and standards, but rather to the inappropriate use of rules to guide professional activities. Medication administration, while seemingly straightforward, depends on multiple contingencies – such as, patient condition, type of medication, and triaging needs (i.e., determining priorities among multiple critically ill patients). Perhaps the ineffectiveness of using simple rules to guide contingent nursing tasks can contribute to the explanation of why nurses redeﬁne medication errors based on the situations in which they occurred (Baker, 1997), and the nurse managers in this study generated their own criteria for classifying reportable incidents. Furthermore, it raises questions about the possibility of specifying and proposing conditions under which nurse managers protect nurses by reclassifying errors. Practice-related research questions might include: Are there alternative methods of guiding and monitoring nursing behavior – in particular in the administration of medications? Is it time to take a fresh look at the traditional ‘‘ﬁve rights of medicine administration’’ – and not just by adding a sixth (Wilson & DiVito-Thomas, 2004)? In this study, we focus on registered nurses in critical care units of tertiary care hospitals. However, nursing researchers could explore whether there should be gradated methods of guiding nursing practice, depending on nurses’ levels of education, training, and experience.

Data for Maintaining Accountability and Learning The three exemplars illustrate some of the difﬁculties that occur when organizations use the same error data for the purposes of disciplining employee behavior and learning how to improve the organization’s performance (Tamuz, 2001). As organizational researchers have long known (e.g., Lawler & Rhode, 1976), the organization gives individual employees disincentives to reveal information about errors, when information from their self-disclosures can be used to judge their performance. The disincentives for self-reporting errors and for disclosing co-workers’ errors are pervasive in organizations and can partially account for the widespread underreporting of patient safety-related incidents in hospitals (e.g., Kohn et al., 2000). If a hospital uses error data for disciplining employees, the interviews suggest that there might be a complicated set of countervailing pushes and

94

MICHAL TAMUZ ET AL.

pulls that can alternatively foster and hinder error reporting. For example, if another health care provider or the patient knows about an error, a nurse might report an error because failure to report it could damage his professional reputation, given the strong professional norms for taking individual responsibility for mistakes (e.g., Witt, 2007). Alternatively, nurses may be reluctant to reveal their errors, for fear of being reported to the ERC. However, when nurse managers buffer staff nurses from the ERC, such concerns may be less salient. Furthermore, nurse managers may also contribute to underreporting by classifying away errors into non-reportable categories (Tamuz, Thomas, & Franchois, 2004). The ERC exemplar also contributes to the discussion on how nurse managers inﬂuence error reporting by staff nurses (Edmondson, 1996,1999; Vogus & Sutcliffe, 2007). It suggests that managers may build trust, in part, by buffering the staff nurses from a punitive regulatory environment, and thus, foster the nurses’ disclosure of errors to their direct supervisors. We propose a conceptual framework that reﬂects the inﬂuence of these fundamental tensions on event classiﬁcation and how the nurse managers chose to respond to them. The tension between maintaining accountability and promoting learning intensiﬁes when the same data are used not only as a basis for learning, but also as an indicator for assessing the performance of individuals. Nurse managers delivered mixed messages when the espoused objectives conﬂicted with actual behavior. The dual use of data for disciplinary and learning purposes generated disincentives for nurses to report errors and incentives for nurse managers to carefully classify events into reportable and non-reportable categories. Furthermore, based on these exemplars, we posit an interaction effect in nursing because the bureaucratic rules used to assess performance are often ill suited to reﬂect the professional judgment that nurses use to assess complex patients, prioritize their care, and cope with unexpected workplace conditions. Thus, it is difﬁcult to classify reportable events simply as deviations from procedures, without specifying the context in which the procedures were performed. Hence, these two fundamental tensions can interact to inﬂuence event classiﬁcation. The classiﬁcation of events, in turn, inﬂuences the choice of organizational response routines. The choice of routines can inﬂuence the information about errors that managers report to the hospital and thus, affect the hospital’s ability to learn from its experience. To illustrate, in exemplar two, the classiﬁcation of events as reportable or non-reportable to the hospital disciplinary system resulted in formal or informal recordkeeping responses. In exemplar three, the classiﬁcation of errors as

95

Promoting Patient Safety

reportable or non-reportable to the ERC resulted respectively in hospitallevel or unit-level responses. When managers classiﬁed errors as nonreportable, they reduced the error data available to the hospital and thus, constrained the capacity for organizational learning at the hospital level. Further research could explore how the organizational capacity to learn from errors is shaped by these two underlying tensions and the strategies that managers developed for coping with them.

Practical Applications The study results underscore the widely accepted (and often ignored) recommendations that data from voluntary reporting systems should not be used as a proxy for the ‘‘true’’ event rate. Error and adverse event rates derived from these systems are biased and cannot be used to measure progress in safety or compare hospitals or clinical areas (Thomas & Petersen, 2003; Pronovost, Berenholtz, & Needham, 2007). For instance, the study results document why and how nurse managers inﬂuence whether an event is reported. These inﬂuences are not related to the true rate of events. When event data are used as safety indicators to compare across units and hospitals, it compounds the disincentives for disclosure and confounds the hospital’s ability to learn from experience. The study underscores that even when hospitals state that they want to learn, it is difﬁcult to foster open reporting. The results might also encourage hospital leaders to re-examine how they apply bureaucratic rules and standards to nursing care, and to try to dissociate disciplinary and learning processes.

STUDY LIMITATIONS This is an exploratory study. It is limited by its sample and study design. The study is based on interviews with 20 nurse managers from four critical care units in three tertiary care hospitals. We do not claim to generalize from this small sample, but rather through rich description we attempt to gain insights into the roles of nurse managers in promoting patient safety by monitoring nursing errors. We propose a conceptual framework that can be examined through further research. Furthermore, we present three exemplars that are related and not necessarily mutually exclusive. They reﬂect the nurse managers’ experience with multiple, partially overlapping systems for monitoring medication

96

MICHAL TAMUZ ET AL.

errors. Because we inferred the nurse managers’ decision-making processes from the examples they provided, we cannot be certain that event classiﬁcation criteria described in the ﬁrst exemplar, for instance, refer exclusively to incident reports.

CONCLUSION We present three exemplars of how nurse managers maintained external demands for accountability, while buffering the staff nurses from these pressures. First, they inﬂuenced staff nurses’ reporting to the hospital voluntary incident reporting system by inadvertently sending mixed messages to the nurses about the consequences of incident reporting. Nurse managers also indirectly and directly ﬁltered the incidents that were reported. Second, nurse managers kept two sets of error records. Third, they classiﬁed potentially harmful events into non-reportable categories. In practice, these methods enabled nurse managers to hold nurses accountable for their performance. However, these methods tended to reduce reporting to the hospital and the ERC, and thus, reduced the information available to the hospital for learning from its mistakes. By examining error monitoring from the point of view of nurse managers, we explore how the nurse managers juggled multiple objectives. On one hand, they pursued the objective of ensuring that nurses followed hospital procedures and they held nurses accountable if the nurses inadvertently varied from the procedures. On the other, the nurse managers sought to protect the staff nurses from inappropriately harsh discipline. The nurse managers also grappled with applying rule-based performance measures that were ill suited to assess the complexities of tasks requiring professional judgment. This exploratory study raises research questions regarding theory and practice. Further research on these questions might eventually enable hospitals simultaneously to maintain compliance with their procedures, while learning from their mistakes.

REFERENCES Ashmos, D. P., Huonker, J. W., & McDaniel, R. R., Jr. (1998). Participation as a complicating mechanism: The effect of clinical professional and middle manager participation on hospital performance. Health Care Management Review, 23(4), 7–20.

Promoting Patient Safety

97

Aspden, P., Wolcott, J. A., Bootman, J. L., & Cronenwett, L. R. (Eds). (2007). Preventing medication errors. Washington, DC: The National Academies Press. Bagian, J. P., Gosbee, J., Lee, C. Z., Williams, L., McKnight, S. D., & Mannos, D. M. (2002). The veterans affairs root cause analysis system in action. Joint Commission Journal of Quality Improvement, 28(10), 531–545. Baker, H. M. (1997). Rules outside the rules for administration of medication: A study in New South Wales, Australia. IMAGE: Journal of Nursing Scholarship, 29(2), 155–158. Balogun, J. (2006). Managing change: Steering a course between intended strategies and unanticipated outcomes. Long Range Planning, 39(1), 29–49. Bates, D. W., Cullen, D. J., Laird, N., Petersen, L. A., Small, S. D., Servi, D., Laffel, G., Sweitzer, B. J., Shea, B. F., & Hallisey, R. (1995). Incidence of adverse drug events and potential adverse drug events. Implications for prevention. ADE prevention study group. JAMA, 274(1), 29–34. Blau, P. M., & Scott, W. R. (1962). Formal organizations. San Francisco: Chandler. Brewer, A. M., & Lok, P. (1995). Managerial strategy and nursing commitment in Australian hospitals. Journal Advanced Nursing, 21(4), 789–799. Callum, J. L., Kaplan, H. S., Merkley, L. L., Pinkerton, P. H., Rabin Fastman, B., Romans, R. A., Coovadia, A. S., & Reis, M. D. (2001). Reporting of near-miss events for transfusion medicine: Improving transfusion safety. Transfusion, 41, 1204–1211. Carney, M. (2004). Perceptions of professional clinicians and non-clinicians on their involvement in strategic planning in health care management: Implications for interdisciplinary involvement. Nursing Health Science, 6(4), 321–328. Cook, R. I., & Woods, D. D. (1994). Operating at the sharp end: The complexity of human error. In: M. S. Bogner (Ed.), Human error in medicine. Hillsdale, NJ: Lawrence Erlbaum. Edelman, L. B. (1990). Legal environments and organizational governance: The expansion of due process in the American workplace. American Journal of Sociology, 95(6), 1401–1440. Edmondson, A. C. (1996). Learning from mistakes is easier said than done: Group and organizational inﬂuences on the detection and correction of human error. Journal of Applied Behavioral Science, 32, 5–32. Edmondson, A. C. (1999). Psychological safety and learning behavior in work teams. Administrative Science Quarterly, 44(4), 350–383. Edmondson, A. C. (2004). Learning from failure in health care: Frequent opportunities, pervasive barriers. Quality and Safety in Health Care, 13, ii3–ii9. Fenton-O’Creevy, M. (1998). Employee involvement and the middle manager: Evidence from a survey of organizations. Journal of Organizational Behavior, 19(1), 67–84. Green, J., & Thorogood, N. (2004). Qualitative methods for health research. Thousand Oaks, CA: Sage Publications. Hall, R. H. (1968). Professionalization and bureaucratization. American Sociological Review, 33, 92–104. Kanter, R. M. (1982). The middle manager as innovator. Harvard Business Review, 60(4), 95–105. Kohn, L. T., Corrigan, J. M., & Donaldson, M. S. (Eds). (2000). To err is human: Building a safer health system. Washington, DC: National Academy Press, Institute of Medicine. Lawler, E., & Rhode, J. (1976). Information and control in organizations. Santa Monica: Goodyear. Levitt, B., & March, J. G. (1988). Organizational Learning. Annual Review of Sociology, 14, 319–340.

98

MICHAL TAMUZ ET AL.

Lipshitz, R., & Popper, M. (2000). Organizational learning in a hospital. Journal of Applied Behavioral Science, 36(3), 345–361. Mair, J. (2005). Exploring the determinants of unit performance – the role of middle managers in stimulating proﬁt growth. Group and Organization Management, 30(3), 263–288. March, J. G. (1999). The pursuit of organizational intelligence. Malden, MA: Blackwell. Marx, D. (2001). Patient safety and the ‘‘just culture’’: A primer for health care executives. New York, NY: Columbia University. Available at: http://www.mers tm.net/support/ Marx_Primer.pdf Marx, D. (2003). How building a ‘just culture’ helps an organization learn from errors. OR Manager, 19(5), 1, 14–5, 20. Miles, M. B., & Huberman, A. M. (1994). Qualitative data analysis: An expanded sourcebook. Thousand Oaks, CA: Sage Publications. Osborne, J., Blais, K., & Hayes, J. S. (1999). Nurses’ perceptions: When is it a medication error? Journal of Nursing Administration, 29(4), 33–38. Patton, M. Q. (2002). Qualitative research and evaluation methods (3rd ed.). Thousand Oaks, CA: Sage Publications. Pronovost, P. J., Berenholtz, S. M., & Needham, D. M. (2007). A framework for health care organizations to develop and evaluate a safety scorecard. JAMA, 298, 2063–2065. Rousseau, D. M., & Fried, Y. (2001). Location, location, location: Contextualizing organizational research. Journal of Organizational Behavior, 22, 1–13. Scott, W. R. (1965). Reactions to supervision in a heteronomous professional organization. Administrative Science Quarterly, 10, 65–81. Scott, W. R. (1969). Professional employees in a bureaucratic structure: Social work. In: A. Etzioni (Ed.), The semi-professions and their organization (pp. 82–144). New York: Free Press. Scott, W. R. (1977). Effectiveness of organizational effectiveness studies. In: P. S. Goodman & J. M. Pennings (Eds), New perspectives on organizational effectiveness (pp. 63–95). San Francisco: Jossey-Bass. Sitkin, S. B. (1992). Learning through failure: The strategy of small losses. Research in Organizational Behavior, 14, 231–266. Tamuz, M. (2001). Learning disabilities for regulators: The perils of organizational learning in the air transportation industry. Administration and Society, 33(3), 276–302. Tamuz, M., & Thomas, E. J. (2006). Classifying and interpreting threats to patient safety in hospitals: Insights from aviation. Journal of Organizational Behavior, 27, 919–940. Tamuz, M., Thomas, E. J., & Franchois, K. (2004). Deﬁning and classifying medical error: Lessons for reporting systems. Quality and Safety in Health Care, 13, 13–20. Thomas, E. J., & Petersen, L. A. (2003). Measuring errors and adverse events in healthcare. Journal of General Internal Medicine, 18, 61–67. Thomas, E. J., Studdert, D. M., Burstin, H. R., Orav, E. J., Zeena, T., Williams, E. J., Howard, K. M., Weiler, P. C., & Brennan, T. A. (2000). Incidence and types of adverse events and negligent care in Utah and Colorado in 1992. Medical Care, 38(3), 261–271. Throckmorton, T., & Etchegaray, J. (2007). Factors affecting incident reporting by registered nurses: The relationship of perceptions of the environment for reporting errors, knowledge of the Nursing Practice Act, and demographics on intent to report errors. Journal of PeriAnesthesia Nursing, 22(6), 400–412.

Promoting Patient Safety

99

Tucker, A., & Edmondson, A. (2003). Why hospitals don’t learn from failures: Organizational and psychological dynamics that inhibit system change. California Management Review, 45(2), 55–72. Tucker, A. L., Edmondson, A. C., & Spear, S. J. (2002). When problem solving prevents organizational learning. Journal of Organizational Change Management, 15(2), 122–137. Vogus, T. J., & Sutcliffe, K. M. (2007). The impact of safety organizing, trusted leadership, and care pathways on reported medication errors in hospital nursing units. Medical Care, 45(10), 997–1002. Wakeﬁeld, D. S., Wakeﬁeld, B. J., Uden-Holman, T., Borders, T., Blegen, M., & Vaughn, T. (1999). Understanding why medication administration errors may not be reported. American Journal of Medical Quality, 14(2), 81–88. Walker, S. B., & Lowe, M. J. (1998). Nurses’ views on reporting medication incidents. International Journal of Nursing Practice, 4(2), 97–102. Weick, K. E. (2007). The generative properties of richness. Academy of Management Journal, 50(1), 14–19. Weick, K. E., & Sutcliffe, K. M. (2003). Hospitals as cultures of entrapment: A re-analysis of the Bristol Royal Inﬁrmary. California Management Review, 45(2), 73–84. Wilson, D., & DiVito-Thomas, P. (2004). The sixth right of medication administration: Right response. Nurse Educator, 29(4), 131–132. Witt, C. L. (2007). Holding ourselves accountable. Advances in Neonatal Care, 7(2), 57–58. Wolf, Z. R., Serembus, J. F., Smetzer, J., Cohen, H., & Cohen, M. (2000). Responses and concerns of healthcare providers to medication errors. Clinical Nurse Specialist, 14(6), 278–287.

QUALITY AND PATIENT SAFETY FROM THE TOP: A CASE STUDY OF ST. FRANCIS MEDICAL CENTER GOVERNING BOARD’S CALL TO ACTION Louis Rubino and Marsha Chan ABSTRACT The Institute for Healthcare Improvement (IHI) has broadened their campaign focus to include protecting hospital patients from ﬁve million incidents of medical harm through 2008. A critical component of this campaign is the engagement of governance in the process, noting evidence of better patient outcomes for hospitals with governing boards that spend at least 25% of their time on quality and safety. St. Francis Medical Center (SFMC), a 384-bed hospital in Southeast Los Angeles serving a high number of uninsured and underinsured patients and a population characterized by signiﬁcant poverty, has initiated through a top-down approach, an aggressive plan to improve the care at its facilities through a call to action by its board of directors. In this article innovative methods are shared, tools are provided, and the initial positive results achieved are reported which show how a cultural change is occurring regarding

Patient Safety and Health Care Management Advances in Health Care Management, Volume 7, 103–126 Copyright r 2008 by Emerald Group Publishing Limited All rights of reproduction in any form reserved ISSN: 1474-8231/doi:10.1016/S1474-8231(08)07005-5

103

104

LOUIS RUBINO AND MARSHA CHAN

quality and patient safety (QPS) at this hospital’s organizational and delivery system level.

A CULTURE OF QUALITY AND PATIENT SAFETY The United States health care system must improve the quality of care. A recent RAND national study found quality deﬁcits in the acute as well as the preventative and chronic levels of care (Goldman & McGlynn, 2005). Hospital governing boards should see quality of care as one of their top priorities. With the recent oversight emphasis placed on ﬁnancial management and managerial ethics, due to the issues with such companies as Enron, WorldCom, and health care organizations Columbia/HCA and Tenet Healthcare, governance attention has been diverted from core operations. Dr. Donald Berwick, President and CEO of the Institute of Healthcare Improvement (IHI) stated in a recent roundtable interview, ‘‘The majority of American hospitals or health systems still haven’t centered on improvement as a strategy. We have serious governance gaps. The board of trustees are not yet feeling the stewardship of quality of care as a fundamental board duty y’’ (Molpus et al., 2006, p. 1). While not synonymous, quality health care and patient safety go hand-inhand. Without a commitment to quality, an organization would struggle to implement the proper patient safety practices. Similarly, without a concerted patient safety effort, care cannot be of the highest quality. Within the last decade, agencies have heralded the importance of both quality and patient safety (QPS) (see Table 1). Hospitals governing bodies have quality oversight as one of their core responsibilities (Grifﬁth & White, 2006). Much is discussed at board meetings about QPS but sometimes the conversations rest on retrospective reviews of problematic peer-reviewed cases, malpractice claims, incident reports, and root cause analysis of sentinel events. In fact, many experts cite the ‘‘passive’’ role governing bodies play in overseeing quality of patient care (National Quality Forum, 2005). Lister (2003) states that board members are unsure how their board can effectively review and monitor QPS data. CEOs also voice frustration that their boards ‘‘are supportive but passive, don’t really understand clinical information, get mired in the details and can’t see the big picture’’ (Bader, 2006, p. 64). Even with the best of systems and tools in place, one challenge trumps all others; changing the culture of the organization. This is ﬁnally being

Agency

Institute for Healthcare Improvement Founded

Institute of Medicine

American Hospital Association

Leapfrog Group

Institute of Medicine

Hospital Quality Alliance National Quality Forum

JCAHOa

JCAHOa

Year

1991

1999

2000

2000

2001

2002

2002

2003

2003

National patient safety goals

Serious reportable events Leadership standards for patient safety

Public accessibility

Crossing the quality chasm

Getting health care right

Committee on governance

To err is human

Improving healthcare worldwide

Initiative

Patient identiﬁcation, communication, surgery, medication use, critical alarms, and infusion pumps

Six aims: safe, effective, patient-centered, timely, efﬁcient, and equitable Endorsed national voluntary consensus standards Need to publicly report occurrence of 27 ‘‘never’’ events Need more strategic direction by ‘‘leaders of the organization’’

Aim for no needless deaths, no needless pain/suffering, no helplessness in those served/ serving, no unwanted waiting, no waste 98,000 people die in U.S. hospitals due to preventable medical error Need for direct quality measurement and error reporting Mobilizing health purchasers can reduce errors, save lives, and save costs

Finding

Leadership’s Call

Adopt standards established by experts Coordinate event reporting for public accountability Ensure implementation of integrated patient safety program Align safety improvement efforts with others in health care industry

Reduce preventable medical mistakes, quality, and affordability through initiating ‘‘leaps’’ Create safety as a system property

Understand importance of system and processes ‘‘Approach the door, Open the door, Ask the right questions’’

To collaborate, set the vision, motivate, innovate, and drive change to get results

Table 1. Chronology of Quality and Patient Safety Initiatives.

Quality and Patient Safety 105

Call to responsibility

International patient safety center

Institute for Healthcare Improvement

National Quality Forum

JCAHOa

Institute for Healthcare Improvement

2004

2004

2005

2006

Improve quality of care through ﬁnancial rewards and transparency of performance data Adoption of six strategies to reduce preventable deaths in U.S. hospitals Outlined role of the hospital trustee in quality improvement Mission to continuously improve patient safety in all health care settings Board’s active involvement will lead to improvements (getting boards on board)

Finding

Spend at least 25% of meeting time on quality and safety issues

Focus on improvements to clinical delivery systems to improve patient outcomes and reimbursement Commitment to implement strategies and report mortality data Develop quality literacy, oversee participation and performance Continuous learning and focus on patient safety

Leadership’s Call

Sources: Adapted from Schyve (2003), DeLashmutt et al. (2003), www.leapfroggroup.org, www.qualityforum.org; www.cms.hhs.gov, www.IHI.org., www.JointCommission.org a From 1988–2006, the Joint Commission was known as the Joint Commission on Accreditation of Healthcare Organizations (JCAHO).

Protecting ﬁve million lives from harm

100,000 Lives Campaign

Medicare pay for performance

Centers for Medicare and Medicaid Services

2003

Initiative

Agency

Year

Table 1. (Continued )

106 LOUIS RUBINO AND MARSHA CHAN

Quality and Patient Safety

107

recognized by today’s hospitals (Bader & O’Malley, 2006). Hospital leadership can only be successful in improving quality when it is truly committed to a culture of patient safety. As this is not often the case, all trustees should declare health care quality and safety are not at the level they should be.

LEADERSHIP RESPONSIBILITY FOR CULTURE CHANGE Governing boards might feel vulnerable though if they make this declaration. Yet, they must acknowledge the need to improve QPS if they are determined to change the culture of their hospitals. A veil of secrecy is held over poor quality and a sense of protectionism may be displayed not only for the hospital but also for speciﬁc physicians, other clinicians, and even some clinical departments. Maintaining compliance with regulatory requirements is frequently found to drive patient safety initiatives rather than hospitals striving for the intrinsic and real rewards stemming from improved quality of care (Devers, Hoangmai, & Liu, 2004). A recent study by the Agency for Healthcare Research and Quality (AHRQ) demonstrated improvements to patient safety have a positive ﬁnancial beneﬁt for hospitals based on their Medicare payment history (Zhan, Friedman, Mosso, & Pronovost, 2006). Taking a proactive approach in addressing quality issues is still the exception rather than the norm. Hospitals should recognize that effective quality improvement efforts will result in beneﬁts beyond ﬁnancial upswings to positively affect the overall performance of the institution. Little research has been conducted regarding the role of hospital governance in quality efforts (Joshi & Hines, 2006). Some evidence is present that hospital governing boards adopting a proactive philosophy and actively engaging in governance work are more likely to perform well overall, using Solucient Center for Health Improvement’s 100 top hospitals for correlation (Prybil, 2006). Descriptive studies of hospital governing boards actually acting speciﬁcally on patient quality and safety issues are needed as best practice models. A hospital governing board taking this approach through active engagement in QPS improvement is that of St. Francis Medical Center (SFMC), a 384-bed acute care facility in Lynwood, California. This nonproﬁt hospital serves a high number of patients who lack health insurance and access to primary health care

108

LOUIS RUBINO AND MARSHA CHAN

services, and thus faces ﬁnancial constraints similar to many other American hospitals (Reinhardt, 2000). Yet, guided by its value of compassionate service, this health care ministry’s mission to serve the poor and sick is led by a governing body that states QPS is its number one priority.

INSTITUTE FOR HEALTHCARE IMPROVEMENT Five years after the Institute of Medicine’s (IOM) report, ‘‘To Err is Human,’’ (IOM, 2000) there is greater acknowledgment of patient safety issues but few infrastructure changes have been implemented to measurably decrease medical errors (Wachter, 2004). One organization which has mobilized health care leaders to address the problems identiﬁed in the IOM’s landmark report is the IHI. Over 3,000 hospitals adopted the IHI’s initial six strategies to reduce needless deaths as part of its ‘‘100,000 Lives Campaign,’’ but many still waiver and are not participating (Gosﬁeld & Reinertsen, 2005). This prompted the IHI to launch a new campaign, ‘‘Protecting 5 Million Lives from Harm Campaign.’’ The campaign calls for six additional interventions intended to advance patient safety, one speciﬁc to leadership. The IHI recognizes the key role governance plays in processing a hospital’s agenda for enhanced QPS by challenging boards to set and prioritize goals for error reduction and monitor the progress (Robeznieks, 2006). The IHI recommends hospital boards ‘‘get on board’’ by deﬁning and initiating new processes for hospital board directors so that they can become more effective in accelerating the improvement of care at their facilities (www.IHI.org/ihi/programs/campaign). A Framework for Leadership of Improvement has been presented by IHI and provides direction through its steps to (1) set direction; (2) establish the foundation; (3) build will; (4) generate ideas; and (5) execute change (IHI, 2006a, 2006b); as illustrated in Fig. 1. Although the quality of an organization’s core business should be of primary concern, a recent article citing the top 10 nonproﬁt governance trends did not address quality oversight (Peregrine, 2007). The IHI’s advocacy for the necessity of hospitals to recognize the need for change beyond clinical interventions does not come without support. There is evidence that board involvement can lead to improvements in care quality. Outcomes are better in hospitals where (1) the board spends more than 25% of its time on quality and safety; (2) the board receives a formal quality measurement report; (3) bases the senior executives’ compensation in part on quality improvement performance; and (4) there is a high level of

Fig. 1.

• Build Relationships • Develop Future Leaders

• Use Model for Improvement for Design and Redesign • Review and Guide Key Initiatives • Spread Ideas • Communicate Results • Sustain Improved Levels of Performance

5. Execute Change

Make the future attractive

PULL

IHI’s Framework for Leadership of Improvement (Reproduced with Permission from IHI).

• Prepare Personally • Choose and Align the Senior Team

2. Establish the Foundation

• Understand Organization as a System • Read and Scan Widely, Learning from Other Industries & Disciplines • Benchmark to Find Ideas • Listen to Patients • Invest in Research & Development • Manage Knowledge

4. Generate Ideas

Make the status quo uncomfortable

• Reframe Operating Values • Build Improvement Capability

• Plan for Improvement • Set Aims/Allocate Resources • Measure System Performance • Provide Encouragement • Make Financial Linkages • Learn Subject Matter

3. Build Will

PUSH

1. Set Direction: Mission, Vision and Strategy

Quality and Patient Safety 109

110

Table 2.

LOUIS RUBINO AND MARSHA CHAN

IHI’s Six Strategies for a Hospital Board to Improve Quality and Patient Safety (from www.IHI.org).

Set aims: Make an explicit, public commitment to measurable improvement Seek data and personal stories: Audit at least 20 randomly chosen patient charts for all types and levels of injury, and conduct a ‘‘deep dive’’ investigation of one major incident, including interviewing the affected patient, family, and staff Establish and monitor system-level measures: Track organization-wide progress by installing and overseeing crucial system-level metrics of clinical quality, such as medical harm per 1,000 patient days or risk-adjusted mortality rates over time Change the environment, policies, and culture: Require respect, communication, disclosure, transparency, resolution, and all the elements of an organization fully committed to quality and safety Encourage learning, starting with yourself: Identify the capabilities and achievements of the best hospital boards and apply that standard to yourself and all staff Establish accountability: Set the agenda for improvement by linking executive performance and compensation

interaction between the board and the medical staff on quality strategy (Vaughn et al., 2006). Through IHI’s call for hospital boards’ involvement, six speciﬁc strategies (see Table 2) were published for boards to adopt in an effort to improve QPS (www.IHI.org). The IHI took the unusual step to instruct boards to gather data on sentinel events and actually learn the stories by putting a ‘‘human face’’ on the data. Boards are encouraged to set a goal of interacting with at least one patient (or family member of a patient) who sustained injury from a sentinel event at their institution within the last year. This strategy proves effective as shown in the PBS documentary, ‘‘Remaking American Medicine: Health care for the 21st Century.’’ This documentary is the story of Mrs. Sorrel King, whose 18 month old daughter’s death as a result of medical errors sparked a partnership with Johns Hopkins Medical Center, to improve the quality of care and safety for other patients (Public Broadcasting Service, 2006). The IHI recognizes the unique role that hospital governing boards and senior leaders have in patient safety citing this role critical and nondelegable (Botwinick, Bisognano, & Haraden, 2006). This leadership role is to establish the value system in the organization; set strategic goals for activities to be undertaken; align efforts within the organization to achieve these goals; provide resources for the creation, spread, and sustainability of effective systems; remove obstacles to improvements for clinicians and staff; and require adherence to known practices that will promote patient safety.

Quality and Patient Safety

111

IHI believes the culture of the organization will begin to change as these processes focus on what happened rather than who is to blame.

BARRIERS TO EFFECTIVE BOARD QUALITY OVERSIGHT Many have called for leadership to place QPS at the forefront of the hospital’s operations (Spath, 2002; Mohr, Abelson, & Barach, 2002; Gautam, 2005), yet this request is unfulﬁlled. There are numerous reasons expressed for this situation. First, a board may be comprised of any combination of hospital managers, community members, and practitioners. Some board members may not possess the health care knowledge supporting a complete understanding about the QPS matters discussed at the board level. A hospital governing board comprised of many non-clinical members may not be attuned to matters in the same way as a committee of only practicing professionals. Therefore, these non-clinical board members may not completely engage in deliberations as the board examines important QPS issues. It is not necessarily the board’s indifference about quality matters but their diminishing concern as one is farther removed from the actual situation. As a result, leadership may be less aware of the patient safety issues than the front-line workers (Singer, Gaba, Geppert, Sinaiko, & Park, 2003). As one goes further up the organizational chart, individuals tend to have a higher perception of the organization’s quality. For example, boards perceive quality is better at their hospitals than their CEOs (Sandrick, 2007). This blindness may easily affect the priority level the governing board places on quality and safety measures. At times, a governing body delegates their quality oversight responsibility to the medical staff, acknowledging they have the clinical expertise for medical care evaluation. However, the perceived disincentives in peer review and the excessive demands on physicians’ time may impede their leading a true quality improvement effort (Marren, 2004). Research has demonstrated that active staff physician involvement in governance can result in a signiﬁcant positive effect on board activity in quality improvement (Weiner, Alexander, & Shortell, 1996). Active physicians therefore are a good addition to boards to push the quality agenda, but the board should not rely solely on them for overseeing the quality at their facilities.

112

LOUIS RUBINO AND MARSHA CHAN

When physicians are involved, care must be given to ensure the appropriate information is transferred for the board to make effective decisions regarding their quality oversight. Yet, communication is often poor between administration and medical staff thus not allowing for effective information ﬂow. Measures by management are needed to ensure the correct information ﬂow is occurring. One way to have this concerted effort is by establishing a QPS subcommittee of the board which has active physician membership. Finally, organizations may not have the infrastructure in place to fully assess and evaluate QPS. This takes a commitment by leadership to allocate the appropriate resources in personnel, systems and information technology to collect, display and analyze, in an easy to understand manner, the data needed for its determination. The size or type of the hospital should not be an issue when initiating a culture change. Large systems have embraced a commitment to quality and safety as demonstrated by the best practices of U.S. Department of Veterans Affairs (VA). Leaders at the VA made a public commitment to address its system of past failures and to establish a culture of safety (Beyea, 2002). The VA developed patient safety centers of inquiry that individually research a different aspect of safety and identify problems in patient care processes. The VA has provided monetary and recognition awards to employees who design and implement solutions or solve patient safety issues, In addition, the VA has incorporated patient safety as part of the performance measurement system, instituted a mandatory and voluntary reporting system, and implemented speciﬁc safety programs throughout the system.

BALANCED SCOREBOARD APPROACH To apply the IHI suggested practice of board engagement in QPS, a hospital framework needs to be selected to assess care being provided and to monitor its improvement. The Malcolm Baldrige Criteria (Fisher, 1996) and Six Sigma (Christianson, Warrick, Howard, & Vollum, 2005) are two modern approaches hospitals use to bolster their quality programs and to ensure alignment with their strategic goals. Spath (2007a) recognizes that measurement has advanced quality of patient care. She further expresses the need for senior leaders to actively support an ever-evolving structure that can tame the ‘‘measurement monster’’ with sufﬁcient ﬂexibility to meet current and future requirements.

Quality and Patient Safety

113

One popular method in use, to summarize volumes of quality data into a well-organized format and highlight important focus areas, is the balanced scorecard or dashboards. Nebraska Medical Center in Omaha supported the use of balanced scorecards indicating it allows the board to be more effective by holding its management more accountable for outcomes. ‘‘Using the balanced scorecard forced us to focus on what we were doing and reprioritize what areas of the hospital we should be focusing on’’ (Meyers, 2004, p. 14). Introduced by Kaplan and Norton (1992), the balanced scorecard approach is a rational planning model which can guide a hospital board to improve patient safety and quality. Experience from industries proves using the balanced scorecard can be an effective method to drive the process of change (Kaplan & Norton, 1993). Information from the report can be obtained by high-level leaders who can then direct initiatives for improvement from the top-down. Recent research asserts this high-level engagement trickles down to front-line staff and cements care quality as an organization-wide priority (Advisory Board, 2006). This does not mean that board members are micro-managing the process. One speciﬁc research study of 562 hospitals suggested that boards in top-performing hospitals take explicit actions that are signiﬁcantly correlated with high organization-wide performance measures (Lockee, 2006). These hospitals work with medical staff and executives to set standards using traditional board oversight powers to achieve goals rather than involving themselves in implementation and execution. This is the key approach taken by SFMC.

ST. FRANCIS MEDICAL CENTER: A CASE STUDY The Daughters of Charity Health System (DCHS) is a regional health care system of six hospitals and medical centers spanning the California coast from the Bay Area to Los Angeles. The health care ministry exists to support the Mission of its sponsors, the Daughters of Charity of St. Vincent de Paul, by providing holistic, spiritually centered care to the sick and the poor. DCHS was formed on January 1, 2002. Facilities within DCHS include O’Connor Hospital, San Jose; Seton Medical Center, Daly City; Seton Coastside, Moss Beach; SFMC, Lynwood; Saint Louise Regional Hospital, Gilroy; and St. Vincent Medical Center, Los Angeles. Recognizing the need to formulate a system-wide strategy to ensure the delivery of the highest quality and safest patient care at each of its six facilities, the DCHS Board appointed a Quality Task Force charged with developing a QPS strategy. The strategy, driven by the DCHS Mission, Vision, and Values,

114

LOUIS RUBINO AND MARSHA CHAN

The Daughters of Charity health System quality and patient safety efforts are guided by our Mission and Values to serve the sick and poor and emphasizes our commitment to “comprehensive, excellent health care that is compassionate and attentive to the whole person: body, mind and spirit”. We live this commitment by striving to provide: • Caring and compassionate service, respectfully and ethically delivered by competent individuals • Error free processes, systems and technologies and a state of the art environment that is safe • Patient outcomes that are consistently excellent and validated by objective, national comparative data Guided by the DCHS Value of “Inventiveness to Infinity”, the quality and patient safety efforts are system-wide and enhance the local health ministries (LHM) and our system’s evolving clinical interdependence through: • Establishing system-wide priorities for excellence that are jointly pursued by each LHM • Trending and comparing hospital data and outcomes (unblended) • Providing clinicians with opportunities for sharing expertise and resources across the system • Sharing organizational and operational knowledge and resources (i.e. replicating best practices) • Embracing a common performance improvement model (PDCA) across all LHM • Learning from experiences, practices and outcomes – both positive and negative – in an environment committed to ongoing excellence and individual and corporate responsibility and accountability, and • Celebrating our success together as a unified system

Fig. 2.

DCHS Quality and Patient Safety Statement (2004).

was detailed in the QPS Statement (Fig. 2). From the strategy, the Task Force developed a series of recommendations adopted in 2004 by the DCHS Board of Directors. This governing body called for implementation of the recommendations over the subsequent two years, including the creation of a board committee dedicated to QPS endeavors at each of the hospitals. From a recent extensive survey conducted by the Health Research and Educational Trust (HRET), 52% of hospital governing boards had a quality improvement committee demonstrating this focused subgroup has not yet reached its full potential (Margolin, Hawkins, Alexander, & Prybil, 2006). In 2006, the SFMC QPS Committee of the Board of Directors was formed.

BOARD QUALITY AND PATIENT SAFETY LITERACY One of the ﬁrst commitments of the Board’s Quality and Patient Safety Committee was to self-educate about key quality and patient safety

Quality and Patient Safety

115

initiatives, metrics, the medical staff credentialing and re-appointment process, and patient satisfaction. The members recognized the need to identify best practices including the value of learning about strategies that have been successful elsewhere. Seeking to identify and adopt best practices, representatives of the board and QPS visited St. Mary’s Hospital and Medical Center in Grand Junction, Colorado, a recognized top performer in the Hospital Quality Incentive (HQI) Demonstration Project. This delegation of highest-level leadership met with St. Mary’s executive team and key medical staff leaders, and nursing and quality leaders. The St. Mary’s Hospital and Medical Center team provided an overview of their medical staff credentialing and peer review process, the medical staff and quality infrastructure, and their approach to clinical quality improvement including patient satisfaction. The St. Mary’s approach to the HQI Demonstration Project and the IHI’s 100,000 Lives Campaign were presented and the St. Francis representatives attended a surgical care infection prevention (SCIP) team meeting, interacted with the rapid response team leaders and clinical unit directors throughout the visit. The site visit was an invaluable experience for the SFMC team, providing the QPS and leadership team with greater understanding about the level of commitment, accountability, and focus required to effect signiﬁcant and sustainable improvement. The experience and best practices from St. Mary’s Hospital and Medical Center are continually referenced by the SFMC QPS and used as a benchmark for the organization.

QUALITY AND PATIENT SAFETY COMMITTEE The QPS Committee is comprised of seven (7) voting members: four (4) members of the SFMC Governing Board of Directors (a physician, a health care administration university professor, and two Daughters of Charity), a community representative, and the Chief and Chief elect of the Medical Staff. The hospital President/CEO and the system Vice President of Quality serve as ex-ofﬁcio members to the committee which is further supported by other hospital leaders (administration, nursing, quality, patient satisfaction, risk, and medical staff services). On a monthly basis, the committee reviews recommendations made and actions taken by the Medical Executive Committee (MEC), such as credentialing, peer review, reappointment, and policies and procedures. The quality scorecard includes data reﬂecting the hospital’s clinical quality and process performance in Joint Commission

116

LOUIS RUBINO AND MARSHA CHAN

Core Measures, Centers for Medicare and Medicaid Services (CMS) HQI Demonstration Project, patient safety, and the 100,000 Lives campaign initiatives (see Fig. 3). It is reviewed along with any adverse/sentinel events and patient throughput metrics. The committee reviews patient satisfaction data, risk reports and environment of care metrics on a quarterly basis, and licensing and accreditation issues on an as needed basis. The reporting frequency is depicted in Table 3. The committee establishes oversight and accountability for the results, setting the direction for continuously improving quality and safety. On a monthly basis, the QPS chairman provides the SFMC Board of Directors with a summary report of progress made and recommendations for board action, as necessary.

CLINICAL UNIT/DEPARTMENT ROUNDS Through a unique hands-on approach, members of the QPS initiated rounds to clinical units/departments prior to each committee meeting in order to gain ﬁrst-hand knowledge of the patient delivery system. Walk-around rounds are recommended to positively impact QPS (Spath, 2007b; Bisognano, Lloyd, & Schummers, 2006). The visits are pre-announced to the unit manager/director, who is invited to provide the board members with an introductory overview and tour of their unit. While rounding, the board members interact with staff, physicians, and patients on the unit. Staff members are asked to share any quality of care and/or patient safety concerns and to discuss their department’s resource wish list. The unit visits are intended to evaluate compliance of clinical quality processes, accreditation and licensing, appropriateness of care, and medical record documentation. Examples of issues addressed during board rounds are listed in Table 4. Furthermore, the rounds provide the committee members the opportunity to validate information they have received at previous QPS meetings and to identify additional issues warranting further review. A summary report of the unit visit and ﬁndings is then reported to the whole committee. These rounds serve as a springboard as the QPS committee identiﬁes additional improvement areas. The rounds are an effective tool to create real dialogue between the board and hospital staff, further reﬂecting the board’s level of engagement in the provision of high quality and safe patient care.

117

Quality and Patient Safety

Fig. 3.

Quality Scorecard.

118

LOUIS RUBINO AND MARSHA CHAN

Table 3.

Reporting Frequency.

Topic

Monthly Quarterly Annually

Medical staff peer review activities Medical staff appointments, credentialing, and reappointment Quality scorecard Clinical quality Patient safety Patient throughput Patient satisfaction Risk management Environment of care Human resource reports Stafﬁng effectiveness Competency Associate retention Associate satisfaction Annual plans Plan for the provision of patient care Plan for improving organizational performance Patient safety Risk management Management of human resources Policies and procedures Licensing and accreditation

As Needed

O O O

O O O O

O

O O

Peer Review While the majority of peer review cases are reviewed and scored for care variations at medical staff department meetings, it is not unusual for some cases to require repeated reviews. Prior to the formation of the QPS, delays in rendering a peer review decision were often present. In order to support timely adjudication, the board requested the MEC to ensure reviews are completed 90–120 days following the initial screening to the end assignment of a peer review score. Peer review timeliness was regularly reported as a metric to each department. With increased focus, the past six months reﬂect 100% compliance in case reviews being scored. In addition, the QPS Committee of the board emphasized the need for the medical staff to be self critical and thorough in the peer review process. If reviews and peer review scorings are not reﬂective of the care rendered and resultant patient outcome, QPS recommends the board return the case to

119

Quality and Patient Safety

Table 4.

Examples of Issues Evaluated During Board Rounds.

System/Process Evaluated

Thrombosis risk screening National patient safety goals Excess length of stay Legibility Fall precautions Medication reconciliation Patient throughput Restraint use Immunization screening Patient satisfaction Advance health care directive and informed consent

Clinical Quality Processes

Accreditation and Licensing

Appropriateness of Care

Medical Record Documentation

O

O

O

O

O O

O O O

O O

O O O O

O O O

O O O O O O O

O O O

the MEC to reconsider the score assignment. From October 2005, the board rejected the peer review decision in six cases and sent each back to the MEC for a more thorough and critical review. In the period after the QPS was formed, the peer review scores are more reﬂective of the actual care rendered. These board level actions served to reinforce medical staff accountability in the peer review process. The thoroughness of the reviews and the overall timeliness has improved.

MEDICAL STAFF CREDENTIALING AND RE-APPOINTMENT The medical staff utilizes a tiered approach to categorize medical staff appointment and reappointment applications. The medical staff has designated tiers, or categories, which reﬂect the physician’s record from the standpoint of professional liability settlements/judgments, disciplinary action/licensure status, references, and veriﬁcation of information for initial appointment (Fig. 4). For reappointment, additional data reﬂecting the physician’s case volume as well as quality performance is reviewed. Medical record documentation

120

LOUIS RUBINO AND MARSHA CHAN

Medical Staff Appointment & Reappointment Categories Appointment: Level One (Fast Track) Reappointment: Level One Practitioner has:

• • • • • • •

Satisfactory references, Satisfied all criteria for membership and privileges requested, No record of malpractice payments within the past 5 years, No disciplinary actions, no licensure restrictions, No problems verifying information, No indications of investigations or potential problems, Information returned in a timely manner and contains nothing that suggests the practitioner is anything other highly qualified in all areas.

Practitioner meets all criteria for Level One Appointment, plus:

• • • • •

•

Practitioner is not requesting new privileges, or satisfies all criteria for any new privileges requested, No record of malpractice payments within the past 2 years, CME is sufficient in volume and relates to privileges, Practitioner meets all criteria and is currently competent to perform privileges requested, Practitioner’s specific profile indicates that performance has been satisfactory in all areas (clinical practice, behavior, etc.) and is absent problematic trends or patterns, No problems have been identified regarding the practitioner’s ability to perform the privileges requested.

Appointment: Level Two

Reappointment: Level Two

•

•

Applicant fails to meet one or more of the criteria identified in Level One above, however, after careful review by the Credentials Committee and department, has been recommended for appointment.

Practitioner fails to meet one or more of the criteria identified in Level One Appointment or Level One Reappointment above, however, after careful review by the Credentials Committee and department, has been recommended for a two year reappointment.

Appointment: Level Three

Reappointment: Level Three

•

•

Applicant fails to meet one or more of the criteria identified in Level One above and, after careful review by the Credentials Committee and department, has NOT been recommended for APPOINTMENT.

Practitioner fails to meet one or more of the criteria identified in Level One Appointment or Level One Reappointment above, however, after careful review by the Credentials Committee and department, has been recommended for a limited reappointment. Limited reappointment is for a time period of less than two years.

Reappointment: Level Four

•

Fig. 4.

Practitioner fails to meet one or more of the criteria identified in Level One Appointment or Level One Reappointment above, and after careful review by the Credentials Committee and department, has NOT been recommended for REAPPOINTMENT. Note – Level Four Reappointment decisions based on medical disciplinary cause or reason must be forwarded to the Board for final action.

Medical Staff Appointment and Reappointment Categories.

121

Quality and Patient Safety Reappointment Criteria

Medical Record Suspension

Level I *

Level II *

Level III *

0 – 15 days

16 – 60 days

> 60 days

2

3–5

6 or more

(cumulative days)

Illegible documentation (# incidents referred to Medical Records or P&T Committee for action)

Malpractice Settlements

0

Monetary Trend Settlement Included in Peer Review, handled on a case by case basis.

Complaints (# patient or family complaints/ # Inpatient Discharges)

Avoidable Days – Admission Criteria Not Met Avoidable Days – Physician Delay

0

Medi-Cal Denied Days

0

Medication Use Pharmacy Interventions Compliance

0

75 – 100%

10% of total days for that doctor 10% of total days for that doctor 10% of total days for that doctor

20% of total days for that doctor 20% of total days for that doctor 20% of total days for that doctor

50 – 74%

<49%

(# compliant/# reviewed)

Adverse Drug Reactions – Preventable (# preventable/# reactions)

Blood Use

75 – 100%

50 – 74%

<49%

No more than one inappropriate transfusion.

No more than two inappropriate transfusions.

Three or more inappropriate transfusions.

Meeting Attendance (requirement of 25% attendance at department >25% 0 – 24% Letter of reprimand and general staff meetings) * Key to Reappointment Levels: Level I = Practitioner meets all criteria, will be reappointed for a two year period. Level II = Practitioner fails to meet one or more Level I criteria, however, will be reappointed for a two year period. Level III = Practitioner fails to meet one or more Level I or Level II criteria, may be reappointed for a limited time period of less than 2 years, to be determined by the department, with approval by the MEC and Board of Directors, on a case by case basis, depending upon the number and types of infractions.

Fig. 5.

Reappointment Criteria.

(delinquency, legibility, and clinical pertinence), peer review volume and scoring, medication use reviews, preventable adverse events, patient complaints, mortality rate, blood use, utilization, as well as applicable key clinical process indicators (i.e. aspirin and beta blocker use for acute myocardial infarction) are summarized in each physician’s report. The practitioner’s performance is categorized into tiers, based upon criteria established by the medical staff (Fig. 5). Historically, the information has been reviewed at the time of reappointment and physicians have been reappointed for a full two-year term, regardless of performance. The process of granting limited reappointment, while delineated in the Medical Staff Rules and Regulations, has not been utilized. However, in the most recent reappointment period, the QPS

Fig. 6.

Physician Reappointment Report Sample.

122 LOUIS RUBINO AND MARSHA CHAN

123

Quality and Patient Safety

Committee recommended the MEC reappoint seven (7) physicians for a limited, one-year period based upon their performance data consistent with Level III reappointment criteria. As a result, the medical staff has expressed a renewed interest in the criteria used to critically evaluate the physician reappointments. An example of a sample physician reappointment report is depicted in Fig. 6.

CONCLUSION SFMC is not unique in having the board be more engaged in quality improvement and patient safety efforts. Hospital governing boards increasingly recognize their need for involvement in strategic issues and lend expertise to improve operations (Thrall, 2007). Even though SFMC’s board quality initiative is relatively young, positive outcomes have been produced. Recently, the hospital successfully completed its three-year Joint Commission survey. President and CEO Gerald T. Kozai stated, ‘‘the survey team acknowledged SFMC for our high quality standards, and they praised us for our best practices in several areas y’’ (Kozai, 2006, p. 1) giving facility associates positive feedback reﬂective of the progress. The board continues the effort to drive QPS improvements. This includes asking the medical staff and administration to consider adoption of ICU intensivists and addressing bed capacity demand issues. The hospital board continues to assess the culture of the organization via an ongoing evaluation process derived from associates’ and physicians’ feedback that identiﬁes their perceptions relative to patient safety, prioritization, reporting, fear of retribution, and ongoing improvement efforts. These, along with the other performance-improvement monitoring activities already mentioned, are altering the infrastructure at this hospital to support the board’s number one commitment to quality improvement and enhanced patient safety. For other hospitals, SFMC can serve as an example of board engagement of quality. Jim Conway, the Senior Vice President of IHI, in a recent interview said, ‘‘when you look at literature on change, it’s clear that engaged leadership always plays an essential role. When I began working with IHI, I had an opportunity to visit many hospitals, and I heard over and over again that governing boards face increased expectations today, and ﬁnd themselves ill-trained to perform their new roles’’ (Zablocki, 2007, p. 1). It is up to hospital governance leadership to investigate evidencebased management principles and explore how best they might get engaged.

124

LOUIS RUBINO AND MARSHA CHAN

The literature pertaining to governance for QPS is rapidly expanding and provides much useful information to assist hospital board members (Clough & Nash, 2007). To make it even easier, IHI has recently come out with a ‘‘how-to guide’’ for boards to get started to improve their quality of care (IHI, 2007). The Joint Commission also has recently published their own guide to ‘‘Getting the Board on Board’’ (Joint Commission Resources, 2007). Only from this top-level concern, direction and action, will hospitals be able to begin to cultivate and build a culture of QPS at their facilities.

ACKNOWLEDGMENT The authors wish to acknowledge the valuable assistance of Ms. Julie Doan, a graduate student at California State University, Northridge, in researching some of the referenced articles.

REFERENCES Advisory Board. (2006). Hospital boards: Challenged to expand oversight of care quality, patient safety. Clinical Strategy Watch, 4(11), 2–3. Bader, B. (2006). Quality and patient safety: Engaging your board to take the lead. Healthcare Executive, 21(2), 64, 66–67. Bader, B., & O’Malley, S. (2006). 7 things your board can do to improve quality and safety. Great Boards, 6(7), 2–6. Beyea, S. C. (2002). Creating a culture of safety. AORN Journal, 76(1), 163–166. Bisognano, M. A., Lloyd, R. C., & Schummers, D. (2006). 10 powerful ideas for improving patient care: Book 3. Chicago: Health Administration Press. Botwinick, L., Bisognano, M., & Haraden, C. (2006). Leadership guide to patient safety. IHI Innovation Series white paper. Cambridge, MA: Institute for Healthcare Improvement (available on www.IHI.org). Christianson, J. B., Warrick, L. H., Howard, R., & Vollum, J. (2005). Deploying six sigma in a health care system as a work in progress. Joint Commission Journal of Quality Patient Safety, 31(11), 603–613. Clough, J., & Nash, D. J. (2007). Health care governance for quality and safety: The new agenda. American Journal of Medical Quality, 22(3), 203–213. DeLashmutt, S., Albertalli, L. B., Beck, C., McHenry, L., Rheault, L., & Robbins, K. (2003). Opening doors to patient safety: A board checklist. Trustee, 56(1), 31–32. Devers, K. J., Hoangmai, H. P., & Liu, G. (2004). What is driving hospitals’ patient-safety efforts? Health Affairs, 23(2), 103–115. Fisher, D. (1996). The Baldrige workbook for healthcare. New York: Quality Resources. Gautam, K. S. (2005). A call for board leadership on quality in hospitals. Quality Management in Health Care, 14(1), 18–30.

Quality and Patient Safety

125

Goldman, D. P., & McGlynn, E. A. (2005). U.S. health care: Facts about cost, access and quality. Santa Monica, CA: The Rand Corporation. Gosﬁeld, A. G., & Reinertsen, J. L. (2005). The 100,000 Lives Campaign. Crystallizing Standards of Care for Hospitals, 24(6), 1560–1570. Grifﬁth, J. R., & White, K. R. (2006). The well-managed healthcare organization (6th Ed.). Chicago: Health Administration Press. Institute for Healthcare Improvement (IHI). (2006a). A framework for leadership of improvement. February (available on www.IHI.org). Institute for Healthcare Improvement (IHI). (2006b) (available at www.IHI.org/ihi/programs/ campaign). Institute for Healthcare Improvement (IHI). (2007). Getting started kit: Governance leadership how-to guide (available on www.IHI.org). Institute of Medicine (IOM). (2000). To err is human: Building a safer health system. Washington, DC: National Academy Press. Joint Commission Resources. (2007). Getting the board on board: What your board needs to know about quality and patient safety. Chicago: Joint Commission on Accreditation of Healthcare Organizations. Joshi, M. S., & Hines, S. C. (2006). Getting the board on board: Engaging hospital boards in quality and safety. Journal on Quality and Patient Safety, 32(4), 179–187. Kaplan, R. S., & Norton, D. P. (1992). The balanced scorecard – measures that drive performance. Harvard Business Review, 70(1), 71–79. Kaplan, R. S., & Norton, D. P. (1993). Putting the balanced scorecard to work. Harvard Business Review, 71(5), 134–147. Kozai, G. (2006). I will. Circle of excellence: The spirit of SFMC. December 1, p. 3. Lister, E. D. (2003). How can boards fulﬁll their quality oversight role? Trustee, 56(5), 22–23. Lockee, C. (Ed.) (2006). Quality. San Diego: The Governance Institute. Margolin, F. S., Hawkins, S., Alexander, J. A., & Prybil, L. (2006). Hospital governance: Initial summary report of 2005 survey of CEOs and Board Chairs. Chicago: Health Research and Educational Trust. Marren, J. P. (2004). The trustee’s responsibility for quality care. Trustee, 57(7), 26, 28. Meyers, S. (2004). Data in, safety out. Trustee, 57(7), 12–14, 19. Mohr, J. J., Abelson, H. T., & Barach, P. (2002). Creating effective leadership for improving patient safety. Quality Management in Health Care, 11(1), 69–78. Molpus, J., Berwick, D. M., O’Kane, M., Pickrell, M., & Roper, W. (2006). The leadership of quality (available at http://www.healthleadersmedia.com). National Quality Forum. (2005). Hospital governing boards and quality of care: A call to responsibility. Trustee, 58(3), 15–18. Peregrine, M. W. (2007). 2007s top 10 governance trends. Trustee, 60(1), 34–36. Prybil, L. D. (2006). Size, composition, and culture of high-performing hospital boards. American Journal of Medical Quality, 21(4), 224–229. Public Broadcasting Service. (2006). Silent killer, remaking American medicine: Health care for the 21st century. PBS Home Video. Reinhardt, U. E. (2000). The economics of proﬁt and not-for-proﬁt hospitals. Health Affairs, 19(6), 178–186. Robeznieks, A. (2006). IHI’s ambition rises for ‘07. Modern Healthcare, 36(50), 8–9. Sandrick, S. (2007). Enhancing the board’s role in quality. Trustee, 60(1), 20–24.

126

LOUIS RUBINO AND MARSHA CHAN

Schyve, P. M. (2003). What you can do: The trustee, patient safety, and JCAHO. Trustee, 56(2), 19–21. Singer, S. J., Gaba, D. M., Geppert, J. J., Sinaiko, A. D., & Park, K. C. (2003). The culture of safety: Results of an organization-wide survey of 15 California hospitals. Quality and Safety in Health Care, 12, 112–118. Spath, P. (2002). Patient safety: Just old wine in a new bottle? Hospital Peer Review, 27(2), 25–28. Spath, P. (2007a). Taming the Measurement monster. In strategies for taming the quality measurement monster. Frontiers of Health Services Management, 23(4), 3–14. Spath, P. (2007b). Sustainable patient safety. hospitalconnect.com, January 17, Hospitals and health networks online (available at www.hhnmag.com). Thrall, T. H. (2007). Building a bolder board – and learning to like it. Hospital and Health Networks, 81(1), 61–62, 64. Vaughn, T., Koepke, J. D., Kroch, E., Lehrman, W., Sinha, S., & Levey, S. (2006). Engagement of leadership in quality improvement initiatives: Executive quality improvement survey results. Journal of Patient Safety, 2(1), 2–9. Wachter, R. M. (2004). The end of the beginning: Patient safety ﬁve years after ‘To Err is Human’’. Health Affairs, 23, w534–w545 (published online November 30, 10.1377/ hlthaff.w4.534). Weiner, B. J., Alexander, J. A., & Shortell, S. (1996). Leadership for quality improvement in health care: Empirical evidence on hospital boards, managers and physicians. Medical Care Research and Review, 53(4), 397–416. Zablocki, E. (2007). IHI calls on boards to lead quality and safety: An interview with Jim Conway, Senior Vice President, Institute for Healthcare Improvement. Great Boards, 7(2), 1–5. Zhan, C., Friedman, B., Mosso, A., & Pronovost, P. (2006). Medicare payment for selected adverse events: Building the business case for investing in patient safety. Health Affairs, 25(5), 1386–1393.

THE TRUSTEE AND PATIENT SAFETY: REDEFINING BOUNDARIES Richard A. Culbertson and Julia A. Hughes ABSTRACT The voluntary hospital trustee has traditionally seen issues of medical care, including those of patient safety, as falling within the delegated sphere of the medical staff. This customary distancing of the trustee from direct involvement in patient safety issues is now challenged by unprecedented scrutiny of hospital safety results through voluntary disclosure or mandatory public reporting. This new climate, fostered by the Institute of Medicine’s To Err is Human and the Institute for Healthcare Improvement’s 100,000 Lives campaign, has complicated the role of the trustee in satisfying the traditional ‘‘prudent person’’ test for meeting ﬁduciary obligation as the trustee’s breadth of involvement expands. Viewed theoretically, Mintzberg models the hospital as a case of a professional bureaucracy, in which the professional staff is responsible for standard setting and regulation. This traditional role of the professional staff is potentially assumed by others lacking technical background. Trustees are now asked to examine reports identifying physician compliance in attaining safety standards without education in the practice supporting those standards. Physician board members, whose numbers have increased in the past decade, are often sought to take the Patient Safety and Health Care Management Advances in Health Care Management, Volume 7, 127–148 Copyright r 2008 by Emerald Group Publishing Limited All rights of reproduction in any form reserved ISSN: 1474-8231/doi:10.1016/S1474-8231(08)07006-7

127

128

RICHARD A. CULBERTSON AND JULIA A. HUGHES

lead on interpretation of patient safety standards and results. The very public nature of patient safety reporting and its reﬂection on the reputation of the organization for which the trustee is ultimately accountable create a new level of tension and workload that challenges the dominant voluntary model of trusteeship in the United States health system.

The governance of hospitals in the United States has traditionally relied on the voluntary contributions of community leaders to oversee operation of what Drucker (1980) has referred to as the most complex organizational structure of contemporary society. The historic view of trusteeship has emphasized the stewardship of the trustee in preserving the ﬁnancial and physical assets of the organization so that it might continue as a going concern. In this approach, trustees are involved in the setting of the broad policy objectives of the organization, with responsibility for the effective operation of the facility delegated by the board to a chief executive ofﬁcer. A further delegation of authority for the generation of actual health services to individual patients within the hospital was made to the organized medical staff (Friedson, 1970). The process of peer review was designed to assure that errors resulting in threats to patient safety in the facility could be identiﬁed and corrected in a timely manner (Bosk, 2003). It was deemed adequate to assure that these mechanisms of quasi-autonomous medical staff review of care delivery exist to fulﬁll the test of the trustee acting as a ‘‘prudent person’’ in assessing the overall performance of the organization. This historic compromise of hospital organization, referred to by Drucker (1980) as the ‘‘dual-headed monster’’, has rapidly evaporated in the face of challenges to the quality of services provided by hospitals. Legal decisions dating from 1965 have established the ultimate culpability of trustees for adverse events in their facility. Yet even this unsettling development could be viewed as occurring within the ﬁscal comfort zone of the trustee in that the perceived negative impact on the organization was basically ﬁnancial. The traditional role of the trustee has been fundamentally challenged by the markedly increased attention to issues of quality at the trustee level. While The Joint Commission has argued over time for greater governing board involvement in quality issues, there has been a reluctance to embrace this expanded role due to the perceived professional and technical nature of many quality questions. The patient safety movement has made it impossible for trustees to fully delegate quality matters, as vigilance at the

The Trustee and Patient Safety

129

highest levels of the organization is demanded by consumer groups and public agencies. However, this is often a brave new world for which lay trustees often feel unprepared and which in many instances may be deferred to medical colleagues on the board who are present by virtue of their trusteeship rather than professional standing. The result is a difﬁcult environment for trustees to attempt to immerse themselves more fully in issues of patient safety, since the demands of the standard ﬁscal ﬁduciary and philanthropic roles of the trustee are also increasing (Alexander & Lee, 2006). The purpose of this chapter is to explain (1) how hospital governing boards have become accountable for patient safety, (2) the organizational and institutional difﬁculties they face in meeting this responsibility, and (3) how they might best prepare for these responsibilities.

PATIENT SAFETY CONCERNS INTENSIFY Voluntary hospitals once enjoyed high levels of public trust built on the presumption that hospitals and the medical professionals that practiced in them could be relied on to at a minimum ‘‘do no harm’’ to those they served. However, like other once-venerated societal institutions such as public schools and religious institutions, hospitals now ﬁnd themselves challenged by increasing public skepticism and stakeholder scrutiny (Institute for Healthcare Improvement, 2006c). The Institute of Medicine’s To Err Is Human (Kohn, Corrigan, & Donaldson, 2000) and its subsequent Crossing the Quality Chasm (Committee on Quality of Health Care in America, 2001) brought patient safety issues to the forefront of national debate and laid out recommendations for fundamental redesign of the American healthcare system based on evidence-based decision making and accountability for results. As a result of these rising concerns about quality and safety, hospitals now ﬁnd their patient safety records under public scrutiny. In addition to internet-accessible ‘‘report cards’’ detailing hospital quality and safety records (Consumer Reports.org, 2003; HealthGrades, 2007; Solucient, 2006), government and accrediting agency programs spotlight patient safety issues. The Centers for Medicare and Medicaid (CMS) has introduced ‘‘pay for performance or P4P’’ Hospital Quality Initiative, which links hospital reporting on 10 quality measures to payments hospitals receive for Medicare discharges (Centers for Medicare and Medicaid Services, 2005). Hospitals must report this data to receive the full payment update amount for their Medicare DRGs. CMS’s Hospital Compare program (2007) has

130

RICHARD A. CULBERTSON AND JULIA A. HUGHES

expanded the potential impact of this information directly to the general population as it provides consumers with information on mortality rates and process of care measures. The Joint Commission now has a distinct set of goals concerning patient safety which hospitals are expected to implement with performance assessed throughout the accreditation cycle (The Joint Commission, 2007). In addition to government and accrediting agency initiatives, hospitals face pressure to participate in voluntary safety reporting and improvement programs. A high-proﬁle example is the Institute for Healthcare Improvement (IHI) 100,000 Lives Campaign. Launched in December 2004, this voluntary effort to improve patient safety targeted six common patient safety issues. Within the ﬁrst year, IHI reported that more than 3,100 hospitals had enrolled in the program with a signiﬁcant proportion voluntarily sending IHI their quality improvement effort data (Institute for Healthcare Improvement, 2006b). Based on the 100,000 Lives success, IHI has developed an even more ambitious initiative, 5 Million Lives Campaign, which aims to enroll more than 4,000 hospitals and speciﬁcally identiﬁes Boards of Trustees as a new target audience through the ‘‘Get Boards on Board’’ initiative (Institute for Healthcare Improvement, 2006a, 2006c). Don Berwick, president and CEO of IHI, states: I think we should declare 2007 ‘‘the Year of Governance’’ and start to put back on the table of the boards not just a request, but an absolute sense of obligation, that learning who does better and then doing at least that well is central to proper stewardship of healthcare. Until leaders own that problem, I don’t think spread is going to happen. The buck stops in the board room. (The Joint Commission, 2006)

The pressure for board involvement in patient safety issues also comes from other quality and governance experts. A 2006 study by The Governance Institute and Solucient found a statistically signiﬁcant relationship between board activities and hospital performance (Becker, 2006; The Governance Institute, 2006). In a survey of national healthcare leaders, quality and safety issues were identiﬁed as one of the seven most important issues boards will face in 2007 (Zablocki & Bader, 2006). Dr. David Nash, chairman of the Department of Health Policy at Jefferson Medical College and chairman of Catholic Healthcare Partners board quality committee says ‘‘We have an epidemic of medical errors, and 50% of patients in the U.S. don’t get the care they out to, based on the evidence. Not that all of this is the board’s responsibility, but an awful big chunk is’’ (Bader & O’Malley, 2006, p. 2). These trends signal that the demands for board accountability for patient safety issues will likely only escalate in the future.

The Trustee and Patient Safety

131

HISTORICAL CONTEXT The role of the trustee in patient safety is a relatively recent development in healthcare governance in the United States. Starr cites the eminent sociologist Charles Perrow in suggesting that authority in American hospitals has passed successively from trustees to physicians to administrators based on the changing technology and needs of the hospital (Perrow). In the early development of hospitals through the mid-twentieth century, Starr writes that ‘‘The domination of the trustees was rooted in the need for capital investment and community acceptance’’ (Starr, 1982, p. 178). As the ascendance of the physician gained momentum, the hospital was viewed as a ‘‘doctors workshop’’ in which physicians would have exclusive ability to oversee their professional activities (Starr, 1982). Rosenberg observes in his history of American hospitals that as early as 1902 the Visiting Committee of Johns Hopkins Hospital adopted as policy ‘‘that their ‘true object’ was to become familiar with the workings of the hospital, ‘that their interest and supervising care be manifested to those here engaged,’ but that speciﬁc suggestions ‘should be made with caution, and are rather to be avoided than specially sought’’’ (Rosenberg, 1987, p. 267). In other words, the Board would desire to be aware of clinical activity within the hospital, but would not presume to intervene – a hands-off approach that predominated discussions of direct Board involvement in patient care issues. Schulz and Johnson undertook an extensive literature review on studies of the work of Boards in American hospitals in 1976. In reporting on Hickey’s 1972 study of Directors ranking of the importance of the functions of hospital boards of directors, they note that only one of ten functions (‘‘maintain qualiﬁed medical staff’’, ranked sixth) pertains in any way to medical care within the institution (Schulz & Johnson, 1976). Similarly, Kaluzny and Veney (1972) found signiﬁcant board involvement pertaining to medical staff issues only in the area of ‘‘appointment of medical staff members’’. Interestingly, while the administrators polled rated trustee participation in even this narrow role as ‘‘considerable’’, trustees reported it as ‘‘some or none’’ (Kaluzny & Veney, 1972). These circa 1970 studies indicate that the role of the Board was judged to be one of stewardship of the assets of the organization, with a heavy emphasis on protection of ﬁscal and capital assets of the organization. The very name trustee, which is commonly used to describe the role of governing board members that would be typically referred to as ‘‘directors’’ in a for-proﬁt organization, carries a conservative and protective connotation.

132

RICHARD A. CULBERTSON AND JULIA A. HUGHES

The role of the Board of Trustees, it is argued, is one of preservation and transmission of these assets to future generations. Within this understanding of the role, then, the primary obligation of the trustee is one of ﬁscal oversight and promotion of philanthropy to conserve the ﬁnancial viability of the organization (Culbertson, 2004). One area of Board concern that has historically transcended ﬁnancial considerations should also be emphasized. This is the obligation of board members to preserve the reputation or good name of the facility (Starr, 1982, p. 155). In contemporary marketing language, this concept might be identiﬁed as protecting the integrity of the ‘‘brand’’ of the institution. This duty has remained constant throughout the evolution of American hospitals, and as we will see later, provides an opening for the reassertion of trustee involvement in considerations of patient safety.

Legal Challenges to Trustee Oversight of Patient Safety As previously noted, the Board has consistently shouldered the responsibility of appointment of physicians to the medical staff. The role of the Board in assessing the competence of medical staff members is a delegated power to the organized medical staff. However, the Board is not shielded from acts detrimental to patient safety by virtue of the act of delegation as it retains ultimate accountability for authorized acts of its medical staff. The courts have provided clear direction about the Board’s responsibilities through a series of legal decisions. A landmark decision was handed down in 1965 in the pivotal Darling v. Charleston Community Memorial Hospital ruling, in which an Illinois court determined that the governing body of the hospital ‘‘has a duty to establish a mechanism for the medical staff to evaluate, counsel, and, when necessary, take action against an unreasonable risk of harm to a patient arising from the patient’s treatment by a physician’’ (Pozgar, 2007, p. 162). The hospital had attempted to argue that liability for the patient’s adverse outcome should be limited to the attending physician, as the hospital was not engaged in the practice of medicine. The court found this argument to be inadequate in addressing responsibility for patient safety. The precedent established by the Darling case has been upheld in two subsequent important state court rulings. In the 1981 Wisconsin case, Johnson v. Misericordia Community Hospital, a malpractice suit was brought against the hospital and its malpractice insurer for failure to properly review orthopedic privileges granted to a physician. The Wisconsin Court of

The Trustee and Patient Safety

133

Appeals, Pozgar notes, found that ‘‘The measure of quality and the degree of quality control exercised in a hospital are the direct responsibilities of the medical staff’’, and that if the hospital had exercised ordinary prudence, it would not have appointed this physician (Pozgar, 2007, p. 161). A similar ruling was delivered in a 1994 Georgia case, Candler General Hospital Inc. v. Persaud, surrounding the death of a patient after a negligently performed cholecystectomy. Pozgar summarizes that ‘‘The court found that a hospital has a direct and independent responsibility to its patients to take reasonable steps to ensure that physicians using hospital facilities are qualiﬁed for privileges granted. The hospital owed a duty to the plaintiffs’ decedent to act in good faith and with reasonable care to ensure that the surgeon was qualiﬁed to practice the procedure that he was granted privileges to perform’’ (Pozgar, 2007, p. 161). The courts have clearly indicated responsibility of the governing body of the healthcare organization for the safety of those seeking care under its auspices. The main emphasis of the two more contemporary rulings following the seminal Darling decision has been the assurance of practitioner competence through the credentialing mechanism (Darling v. Charleston Community Memorial Hospital, 1965). This mechanism has been, as noted earlier, an historic responsibility of the governing board upon the recommendation of the organized medical staff. There exists, however, a history of notions of shared governance in these areas with the organized medical staff of the institution that sometimes makes responsibility for patient safety ambiguous on a practical operational level if not in an ultimate legal sense.

The Joint Commission’s Role in Patient Safety The Joint Commission has played a central role in deﬁning the role of the board in issues of quality and safety since its establishment in 1951 (Longo & Avant, 1994). The initial focus of the Joint Commission centered on creation and maintenance of an effective process of credentialing for which the board would ultimately be held accountable. However, the process through which credentialing would take place would be under the delegated authority of the quasi-autonomous organized medical staff (Orlikoff, 1994). Processes for the granting of medical staff privileges, organizational staff credentialing, appointment, and evaluation were governed by the bylaws of the medical staff, which were ultimately approved by the board (Longo & Avant, 1994). Patient safety as an

134

RICHARD A. CULBERTSON AND JULIA A. HUGHES

identiﬁable issue had not yet emerged as a distinct concern in the trustee literature, and the prevailing philosophy of the latter half of the twentieth century was that patient safety could be assumed to follow from proper pursuit of the credentialing process, with the resultant elimination of unqualiﬁed and incompetent practitioners (Berwick, James, & Coye, 2003). This faith in process and structure did not please all critics, who argued that the link to the third of Donabedian’s time honored principles of quality, outcome, was not assured (Donabedian, 1982). The Joint Commission initially recommended that information pertaining to medical care within the organization be advanced to the full board through the institution of the Joint Conference Committee, consisting of representatives of the board, the medical staff executive committee, the CEO, and possibly a medical director or full time chief of staff (Schulz & Johnson, 1983). The institution of the Joint Conference Committee was endorsed by the Joint Commission in 1965 and included in its standards for the organization of the medical staff (McGibony, 1969). This intermediary structure for communication between the board and medical staff, though endorsed by The Joint Commission at the time, was found by its Associate Director, Martin Stone, in 1971 to be not ‘‘overly impressive’’ as a liaison device (Schulz & Johnson, 1983). Johnson and Johnson (1986) identify the problem directly as follows: For a number of years the Joint Commission on Accreditation of Hospitals [i.e., The Joint Commission] has included in its standards the idea that the medical staff of a hospital would be self-governing. Interpretations of what this means have varied considerably depending on the viewpoint of the interpreter. The problem arises in the organizational context when a subordinate post of the organization, the medical staff, can be subordinate to a governing authority in a hospital and at the same time be self-governing – a contradiction in terms. To an organizational purist, it is either subject to a governing board or it is not. If it is subject to a board, then the subordinate body is responsible for carrying out the dictates of a superior body in the organization, and there is no room for a concept such as self-governance. (Johnson & Johnson, 1986, p. 257)

In this remarkably foresighted article, the authors argue that in the mind of trustees ‘‘clinical decisions of individual physicians are best judged by other physicians and that trustees do not have the necessary knowledge to decide such matters y’’ (Johnson & Johnson, 1986, p. 258). The ultimate responsibility of the medical staff is to ‘‘assure patient safety’’, and yet it is a shared obligation with the governing body. It is the pessimistic conclusion of Johnson and Johnson that this model has failed in the area of patient safety, saying that ‘‘self-governance has failed as an instrument

The Trustee and Patient Safety

135

for warning governing authorities of impending disasters’’ (Johnson & Johnson, 1986, p. 267). Most recently, The Joint Commission has adopted an increasingly directive position regarding the role of the board in patient safety. Note the signiﬁcant revisions to policy adopted effective with the organization’s 2004 Hospital Accreditation Standards. Patient safety as a speciﬁc consideration in the 2003 version of the standards is mentioned several times, prominently as Standard LD.5.3 ‘‘Leaders ensure that patient safety issues are given a high priority and addressed when processes, functions, or services are designed or redesigned’’ (JCAHO, 2003, p. 211). An expansion of leader involvement in patient safety occurs in the 2004 version of the standards, as a total of four standards that deal explicitly with patient safety are introduced (JCAHO, 2004, Standards LD.4.40, LD.4.50, LD.4.60, and LD.4.70). Most noteworthy is the mandate in Standard LD.4.40, element 8, that requires the board establish a patient safety program to report, ‘‘at least annually, to the hospital’s governance or authority on system or process failures and actions taken to improve safety, both proactively and in response to actual occurrences’’ (JCAHO, 2004, p. 225). Patient safety is assessed on a continuum of occurrences ranging from no harm incidents that might pose future danger (such as a ‘‘slip’’ that does not result in a fall) to ‘‘sentinel events’’ with serious adverse results (JCAHO, 2004). In addition, The Joint Commission further endorses the adoption of clinical practice guidelines by the organization as a way of protecting patient safety. Standard LD.5.30 requires that ‘‘Appropriate leaders, practitioners, and healthcare professionals in the hospital review and approve clinical practice guidelines selected for implementation’’ (JCAHO, 2004, p. 227).

Public Reporting and Patient Safety The board has a deﬁnite stake in maintaining and promoting the good reputation and image of the healthcare organization. A board that might otherwise look upon The Joint Commission’s recommendation of adoption of practice guidelines as merely optional may believe otherwise when public reporting of patient safety indicators becomes the norm. The public awareness of patient safety as a public policy issue has a relatively brief but rapidly developing history, prompted by the Institute of Medicine’s 2000 report To Err is Human and embellished by two subsequent white papers focused on patient safety (Kohn et al., 2000). Initially these reports were viewed as calling attention to the broader issue of patient safety and

136

RICHARD A. CULBERTSON AND JULIA A. HUGHES

advancing recommendations for the hospital ﬁeld in general, as well as steps that might be taken at the facility level to improve safety. More recently, the call for transparency in reporting of safety-related events has resulted in the consideration or adoption of state reporting laws that feature reporting on identiﬁed facilities. Minnesota has been at the forefront of this movement, passing its Adverse Health Events Reporting Law in 2003 that requires public reporting by facility of 27 identiﬁed serious reportable events (Mn. US Statutes 144 sections 144.706 through 144.7069). At present, mandatory reporting is generally restricted to conditions or events of life-threatening consequence, but the scope of such reporting can be expanded at the discretion of an individual state legislature. Wood and Nash (2005) have proﬁled extreme variation in state reporting systems, but advocate the expansion of mandatory state medical error-reporting systems as good public policy. Certainly a board member must be concerned for the preservation of the reputation of the hospital one serves as public attention and discussion of safety-related issues broadens.

HOSPITALS AS BUREAUCRACIES Mintzberg (1981) argues that organizations function most effectively when their design ﬁts with their circumstances, i.e. the type of work they do, their size, their production technology, and the industry conditions in which they operate. In his seminal work on organizational structure, Mintzberg identiﬁed ﬁve building blocks which together make up the framework of organizations: the organization’s strategic apex or its senior management, its operating core or component where the basic work of the organization is accomplished, its middle management or middle line, and two types of staff support, systems design specialists in the technostructure and the general support staff who provide services to the rest of the organization (Mintzberg, 1989). Using this typology, Mintzberg categorizes ﬁve ‘‘natural conﬁgurations’’ that organizations may assume, which are characterized by distinctive arrangements of the ﬁve basic organizational components. These conﬁgurations, the simple structure, the machine bureaucracy, the professional organization or bureaucracy, the divisionalized form, and the adhocracy, vary in how and by whom the work of the organization is controlled and coordinated (Mintzberg, 1989). Of greatest interest to this analysis are the machine and professional bureaucracies.

The Trustee and Patient Safety

137

The popular image of the machine bureaucracy is captured in Charlie Chaplin’s 1936 ﬁlm Modern Times which shows relentless assembly lines that dehumanize employees with boring and repetitive work and draconian managerial control systems. However, the machine bureaucracy is a highly effective structure when the work of the organization is simple and highly standardized, and the environment is relatively simple and stable (Mintzberg, 1989). Such organizations develop large technostructures to design standardized routines for its repetitive work processes. Decision making and control remain centralized and concentrated at top in the organization’s strategic apex. Machine bureaucracies often develop when an organization is subject to tight external controls, since centralized decision making supports the strict standardization required to assure compliance. In contrast, hospitals, like universities, engineering companies, and accounting ﬁrms, are most often organized as professional bureaucracies (Mintzberg, 1989). While the work processes of a machine bureaucracy are both simple and standardized, the work of professional organizations is at once highly complex, e.g. heart transplant surgery, but at the same time relatively stable or standardized, e.g. the essential steps are the same for each heart transplant. In contrast to the centrally controlled machine bureaucracy, control in the professional bureaucracy is decentralized to the operating core, because the complex core businesses processes of the organization are dependent on highly skilled specialists who must be given a great deal of autonomy to direct their own work. In the case of the hospital, those at the strategic apex, the senior administrators and boards of trustees, cede much decision-making control to physicians, whose judgments determine the success of the hospital’s core business processes. The work of administrators and boards is to facilitate the work of professionals, which they do by ensuring sufﬁcient staff support, settling jurisdictional disputes, securing ﬁnancial resources, managing outside inﬂuences, setting broad strategic direction, etc. (Mintzberg & Quinn, 1998). Work in the professional bureaucracy is coordinated by the standardization of professionals’ skills, a direct contrast to the machine bureaucracy’s standardization of work processes. Physicians receive years of formal training in medical school, followed by more years of on-the-job training in their residencies and fellowships, before they are deemed to have mastered their professional skill set. Not surprisingly, in many cases, the professional may demonstrate more loyalty to the educational and professional associations that train, license and police their professional behaviors than to the organizations in which they practice those skills.

138

RICHARD A. CULBERTSON AND JULIA A. HUGHES

Unlike machine bureaucracies, professional bureaucracies do not thrive in situations open to the imposition of external controls from outside the professional realm. To respond to external controls, the organization’s administrative structure must exert authority to mandate compliance to the imposed standards, thus disrupting the organization’s decentralized power structure and undermining the autonomy of the professionals. The attempt to formalize complex work via externally imposed standards, whether rules, regulations, or performance measures, pushes the organization toward the machine bureaucracy and puts the success of the operating core in jeopardy (Mintzberg, 1981). As Cohen, Eustis, and Gribbins (2003) have observed, real power in a professional bureaucracy resides in the clinical staff, and to change this culture is a daunting undertaking requiring adept application of leadership skills. Medical care and surgical procedures, under the control of physicians, form the core work of the hospital. Since issues of patient safety are clearly linked to the functioning of these core processes, in a successfully functioning professional bureaucracy these issues should naturally fall under the purview of the physicians controlling the operating core. Assignment of accountability to the hospital board for meeting patient safety standards, whether by government or accrediting agency ﬁat, is a clear example of external controls being imposed on the professional bureaucracy. This shift of accountability up to the strategic apex has indeed been experienced by physicians as a threat to their autonomy and may pose a threat to the work of the hospital as an effective professional organization.

THE PROBLEM OF CLINICAL AUTONOMY OF PHYSICIANS Professional autonomy has been regarded as the deﬁning attribute of a true professional since its identiﬁcation by Freidson in 1970. Wolinsky has stated that autonomy is the keystone of a theory of an independent profession, stating that ‘‘All other characteristics of a profession ﬂow from it. Thus, autonomy is the acid test of professional status’’ (Wolinsky, 1988, p. 44). Autonomy is nothing more or less than the right of the profession to control its own work (Freidson, 1970, p. 71). The clinical dimension of autonomy provides for the independent exercise of professional judgment within parameters deﬁned by the profession itself and scrutinized solely by one’s professional peers (Freidson, 1970).

The Trustee and Patient Safety

139

Nonetheless, Bosk notes in his study of academic surgeons that clinical autonomy can lead to varying interpretations even with the medical profession. He writes that ‘‘The range of claims made in the name of an individual’s clinical autonomy was quite broad. It included: the ability to overrule courses of treatment indicated by the scientiﬁc literature based on clinical authority y’’ (Bosk, 2003, p. 241). Hospital boards have been quite reluctant to impinge in any way on the autonomy of providers. A study published in 2003 in the Journal of the American Medical Association found that clinical autonomy is the primary driver of physician career satisfaction based on a survey of 12,000 primary care and specialist physicians in the United States (Landon, Reschovsky, & Blumenthal, 2003). Measures of clinical autonomy, including the ability to determine one’s number of hours worked and to order services for patients without external review, were the biggest predictors of physician satisfaction. These issues exceed income in importance to the surveyed physicians in determining satisfaction with work life. As issues of provider satisfaction are clearly one of the historic concerns of board members, there is generally considerable caution exercised by boards in appearing to reduce the autonomy of providers.

Peer Review as a Speciﬁc Instance of Physician Autonomy The exclusive right of a profession to review the work of its members and to recommend discipline if required is one of the traditional attributes of a profession. For hospitals, this right entails that only the members of the medical profession possess the technical expertise to critique and judge the work of other physicians, and that this will take place within the peer review process of the organized medical staff. During the 20th century, even the federal government in establishing the system of Professional Standards Review Organizations in 1972 and their successor Professional Review Organizations was reluctant to deviate from the peer review model (Light, 1997). However, as early as 1970, Freidson questioned the efﬁcacy of this system in protecting the public interest, writing that ‘‘The question here is whether the profession, left to its own devices by its autonomy, can organize the work of its members in such a way that the public can be assured of a reasonably high standard of performance on the part of every professionaly’’ (Freidson, 1970, p. 365). He concludes that the profession must open itself to external review by trustees and outside agencies, for he

140

RICHARD A. CULBERTSON AND JULIA A. HUGHES

writes that ‘‘The experience of the creators [physicians] is a function of the perspective they have gained by virtue of being in an especially protected, autonomous position in the social structure, a position which systematically discounts the experience and evaluation of the laymen outside’’ (Freidson, 1970, pp. 379–380). The patient safety initiative has provided a great incentive to expand the scope of review of the board to areas that might previously been considered as exclusively within the purview of the medical staff. The recognition by The Joint Commission of broader leadership obligations in patient safety for the board has blurred the former role distinctions that existed as a working compromise between the board and the medical staff for approximately a half century. How this information is reported to the board and what is actually done with it in setting policy becomes the crucial question.

A Case Vignette: The Conservative Orthopedic Surgeon The range of information presented to a hospital board represents a distillation of information deemed essential for the attention of the board in making policy. Standard practice is to caution any hospital board that its role is not one of micromanagement of the operations of the organization, but in praxis this is a determination that often results from trial and error. One prominently recommended tool for hospital boards to assess organizational performance is the ‘‘dashboard’’ of select key performance indicators (Center for Healthcare Governance, 2007). The use of dashboards as a best practice is ‘‘not micromanagement by the board members. It is a methodology for identifying and measuring those priority objectives established by the organization’s policymakers’’ (National Center for Healthcare Leadership, 2006, p. 12). Making speciﬁc application to patient safety of this best practice recommendation presents an interesting dilemma. For the hospital board, this may be best depicted in a case example. A clear patient safety objective identiﬁed in the seminal 2000 Institute of Medicine report is the reduction of adverse events related to infection acquired during hospitalization (Kohn et al., 2000). Infections are cited by the study authors as the second leading cause of adverse patient events, and a major target for patient safety improvement. A prominent recommendation to reduce surgical infections is the administration of antibiotics preoperatively as one key indicator and for

The Trustee and Patient Safety

141

the discontinuation of their administration within an established time postoperatively. Speciﬁcally, for hip replacement surgery, a major orthopedic procedure, antibiotics are to be administered 1 h prior to incision and are to be discontinued within 24 h postoperatively (Centers for Medicare and Medicaid Services, 2003). The primary focus of these recommendations is one of patient safety through infection prevention, and results are reported to the hospital board’s committee on quality and subsequently reviewed by the full board. At a board meeting, it is noted that the dashboard indicator for the discontinuation of antibiotics 24 h postoperatively is consistently registering ‘‘red’’, indicating signiﬁcant deviation from best practice guidelines. The board committee, comprised of board members (both physicians and nonphysicians), other health professionals, and medical staff ofﬁcers as well as the appointed medical director, puzzles over this evident departure from best practice in a genuine desire to improve patient safety. Another concern is that the accumulation of a substantial number of red indicators may reﬂect poorly on the hospital if these are publicly reported and the reputation of the hospital is compromised. As a result of these concerns, the medical director is asked to investigate and report back to the committee. At the next meeting, the medical director reports on his encounter with one of the operating physicians. The orthopedic surgeon responded that he too was greatly concerned with patient safety, as he felt any conscientious physician would be in keeping with the ethical obligations of the profession. In fact, he believed himself to be so committed to the cause of patient safety that he sustained administration of antibiotics for 48 h postoperatively in order to be on ‘‘the safe side’’. When the medical director suggested this might be excessive, the orthopedic surgeon responded that he had never had a patient develop an infection postoperatively and saw no reason to change his practice. To do otherwise, he argued, would be an infringement upon his clinical judgment exercised in the patient’s best interest and a violation of his clinical autonomy. Evidence-based medicine guidelines would argue that the orthopedic surgeon is incorrect in his practice, but the board is reluctant to impose this conclusion. Given that patient safety is the ultimate objective of all parties, who is to prevail in a disagreement? The board is the ﬁnal legal authority in all matters pertaining to practice in the hospital, yet board members are by custom aware of the limits to their own technical knowledge and are reluctant to impose speciﬁc practices in a less than clear cut circumstance.

142

RICHARD A. CULBERTSON AND JULIA A. HUGHES

INVOLVEMENT OF PHYSICIANS ON BOARDS OF TRUSTEES Although almost all boards receive education to improve their competency in quality and patient safety issues, many hospital CEOs recognize that education cannot transform even the most well-intentioned lay trustee into a quality expert capable of leading the board’s clinical quality agenda (Bader, 2006). To insure the proper levels of expertise, most recommendations for improving board involvement in quality and patient safety call for including knowledgeable quality experts in board membership (Bader & O’Malley, 2006; Institute for Healthcare Improvement, 2006a). Patient safety ‘‘literacy’’ is regarded as a high educational priority for hospital leaders and boards, especially as this promotes improved systems of care and the assessment of these systems (Clarke, Lerner, & Marella, 2007). Although some boards have turned to quality or customer service experts in other industries, more commonly boards include physician trustees to gain in-depth knowledge of clinical issues (GreatBoards.org, 2004; The Walker Company, 2004). Current recommendations for physician membership on hospital boards range from 15% to 25%, with the caveat that IRS regulations stipulate that no more than 49% of a tax exempt hospital’s board should be made up of insiders, such as physicians, who have privileges at the facility (GreatBoards.org, 2004). The Governance Institute’s, 2000 survey showed the median hospital board size was 12 members, with two being medical staff members. Only 18% of surveyed hospitals reported having no physician members on their board of trustees (The Governance Institute, 2000). A more recent study found the hospital boards comprised, on average, 13 members, of which 2.5 were physicians, or roughly 20% of the overall composition (Margolin, Hawkins, Alexander, & Prybil, 2006). This physician presence is a dramatic departure from the historic practice of deliberately excluding physicians from board membership. MacEachern, a physician himself, argued in his early text on hospital management that the inclusion of physicians created the risk of a class of board members who would promote the self-interest of their professional group rather than the collective good of the hospital enterprise (MacEachern, 1946). By the early 1970s, this notion had given way to one promoting physician inclusion on the board, either as representatives of the medical staff by virtue of ofﬁce, or as physicians selected in the same manner as non-physician members by the board itself (Schulz & Johnson, 1976). The key question is whether the level of board involvement in patient safety issues concerns the full board or is reserved to the medical staff and its

The Trustee and Patient Safety

143

self-governance processes. Will the board as a whole become more involved in patient safety in light of increased public awareness and greater governmental and accrediting pressure to address these issues, or will it continue the tradition of delegation to professional experts? Gribbin suggests that ‘‘boards often feel that quality and safety are nebulous concepts at best’’ (Gribbin, 2007, p. 36). The Institute for Healthcare Improvement (2007) suggests two possible responses to this problem that involve the total board. One of these is formal board education, perhaps leading to trustee certiﬁcation, that will elevate all board members to a common level of proﬁciency. It should be noted, however, that the program introduced in Tennessee requires only 6 h/year of training in all areas of board work. The second response is the disciplined analysis of data as more safety data results from enhanced reporting requirements. Better patient safety will result when board oversight and review of patient safety information takes place, although a signiﬁcant role in framing and interpreting these results for the board’s consideration remains.

CONCLUSION: PATIENT SAFETY AS AN INCREASED ROLE FOR THE BOARD It is the authors’ experience that patient safety considerations have captured the attention of governing board members with an immediacy that has not been true in the longer history of quality improvement initiatives. The Governance Institute reports that 64.9% of boards are involved in setting the quality agenda for the organization (Becker, 2006). Moreover, 85% of board chair respondents surveyed by the Health Research Education Trust state that their boards routinely review safety indicators in the hospital (Margolin et al., 2006). Perhaps this is a reﬂection of the effectiveness of public awareness campaigns such as the 100,000 Lives initiative that have given a more human face to patient safety than was the case in more abstract discussions of quality. Board members of all backgrounds do understand adverse events such as medication errors, falls, and infections, and wish out of compassion as well as reputation and ﬁnancial stewardship to prevent their occurrence. The historic formula for board membership selection has been the three Ws of governance: work, wisdom, and wealth (Taylor, Chait, & Holland, 1996). Individual trustees have generally opted to work within a comfort zone reﬂecting their personal and professional background when accepting

144

RICHARD A. CULBERTSON AND JULIA A. HUGHES

appointment to the hospital board and its committees. The dramatically increased attention devoted to patient safety coincides with signiﬁcantly increased responsibilities for board members of publicly traded organizations resulting from the enactment of the Sarbanes-Oxley Act of 2002 (Orlikoff, 2005). The ﬁduciary standards of this law that apply to audit and ﬁnancial reporting are being extended as an expectation to not-for-proﬁt boards by bond rating groups and state agencies, although not-for-proﬁts are not directly subject to this particular law (O’Hare, 2002). Philanthropy as a board obligation has also received renewed attention as revenue streams from other traditional operating sources are threatened by competitive or policy forces (Tyler & Biggs, 2005). As a result of the intensiﬁcation of these competing obligations, the board may have more incentive to look to physician members of the board for guidance and direction on patient safety issues. This may appear to be a paradox in that the issue is one of general concern to board members, and is regarded as having signiﬁcant importance in those surveys reviewed in this chapter. The question for future consideration is whether board members will invest sufﬁcient time and attention to education in patient safety issues, which are admittedly outside the normal competence of most board members. The alternative is to refer these issues to a professionally dominated subset of the board and rely on their counsel. This latter strategy could lead to a concentration of physician and professional members in safety and quality committee assignments, while moving them away from other valuable areas of board work, such as ﬁnance and strategic planning, in which their input would also be highly beneﬁcial to the entire organization. A future research strategy to test these concluding observations could involve an analysis of board committee assignments of board members by professional background and personal attributes. In addition to professional data on each member, one might examine how committee memberships were obtained (voluntary or assigned) and the degree of satisfaction of the individual board member with them. The board members might also be asked to self-assess their contribution in each selected area. Did the board members regard their contribution as one that draws upon their particular expertise or on general knowledge? A further area for inquiry is the role of the board in promoting an overall culture of patient safety as opposed to its involvement in technical aspects of performance management and evaluation. The Blue Ridge Academic Health Group observes that ‘‘‘Culture Eats Strategy for Lunch.’ It is not enough to plan and/or engineer organizational changes. Unless the goal and practice of

145

The Trustee and Patient Safety

quality becomes part of the organizational culture, quality initiatives are not likely to succeed’’ (Blue Ridge Academic Health Group, 2006, p. 10). Weiner, Shortell, and Alexander (1997) suggest a particular opportunity exists for contribution by active staff physicians participating in governance in advancing clinical process improvements through total quality management. Notable efforts to study the effort to promote patient safety in speciﬁc clinical settings have recently been published. Longo, Hewett, Ge, Schubert, and Kiely (2007) have analyzed key organizational characteristics that predict greater implementation of patient safety systems. They report a strong association between Joint Commission accreditation and patient safety goal attainment. The involvement of the governing board, however, is not examined as one of the variables employed in the study. Safety indicators and the process of involving the board could be factored into future studies as a contributing variable. Winokur and Beauregard’s (2005) assessment of creation of a culture of patient safety at William Beaumont Hospital uses a case-based analytic approach. Case analyses are another model for inquiry that might be used to examine board involvement in the creation of a culture of patient safety. Patient safety presents a signiﬁcant challenge for the contemporary board that boards recognize, but are now obligated to evolve new strategies to integrate into board work. Traditional divisions of authority between board and medical staff are being continually renegotiated, and the result of this redeﬁnition will impact the future of patient safety initiatives, as well as the governance process itself.

REFERENCES Alexander, J. A., & Lee, S. D. (2006). Does governance matter? Board conﬁguration and performance in not-for-proﬁt hospitals. The Milbank Quarterly, 84(4), 733. Bader, B. S. (2006). Quality and patient safety: Engaging your board to take the lead [Electronic Version]. Healthcare Executive, 21, 64–67. Retrieved on February 8, 2007. Bader, B. S., & O’Malley, S. (2006). Boardroom brieﬁng: 7 things your board can do to improve quality and patient safety [Electronic Version]. Great Boards, VI, 1–5. Retrieved on February 8, 2007. Becker, C. (2006). From the top down: Boards play key role in clinical improvement. Modern Healthcare, 36(24 Supplement), 24–26. Berwick, D., James, B., & Coye, M. (2003). Connections between quality measurement and improvement. Medical Care, 41(1), I-30–I-38.

146

RICHARD A. CULBERTSON AND JULIA A. HUGHES

Blue Ridge Academic Health Group. (2006). Health care quality and safety in the Academic Medical Center. Report 11. The Robert W. Woodruff Health Sciences Center at Emory University. Atlanta, GA. Bosk, C. (2003). Forgive and remember (Second edition). Chicago, IL: University of Chicago Press. Center for Healthcare Governance. (2007). Building an exceptional board: Effective practices for healthcare governance. Chicago, IL: Center for Healthcare Governance. Centers for Medicare and Medicaid Services. (2003). Premier hospital quality incentive demonstration clinical areas and quality measures. Retrieved on February 23, 2007 from http://www.cms.hhs.gov/HospitalQualityInits/downloads/HospitalPremierMeasures.pdf Centers for Medicare and Medicaid Services. (2005). Medicare ‘‘pay for performance (P4P)’’ Initiatives. Retrieved on February 14, 2007, from http://www.cms.hhs.gov/apps/media/ press/release.asp?Counter ¼ 1343 Centers for Medicare and Medicaid Services. (2007). Hospital compare: A quality tool for adults, including people with Medicare. Retrieved on January 28, 2008 from http://www. hospitalcompare.hhs.gov Chaplin, C. S. (Producer and Director). (1936). Modern Times [Motion Picture] USA: Charles Chaplin Productions. Clarke, J., Lerner, J., & Marella, W. (2007). The role for leaders of health care organizations in patient safety. American Journal of Medical Quality, 22(5), 311–318. Cohen, M., Eustis, M., & Gribbins, R. (2003). Changing the culture of patient safety: Leadership’s role in health care quality improvement. Joint Commission Journal on Quality and Safety, 29(7), 329–335. Committee on Quality of Health Care in America. (2001). Crossing the quality chasm: A new health system for the 21st century. Washington, DC: National Academy Press. Consumer Reports.org. (2003). Hospital report cards. Retrieved on February 14, 2007, from http://www.consumerreports.org/cro/health-ﬁtness/health-care/hospitals-how-safe103/hospital-report-cards/index.htm Culbertson, R. (2004). Boards of directors. In: M. Stahl (Ed.), Encyclopedia of health care management (pp. 46–48). Thousand Oaks, CA: Sage. Darling v. Charleston Community Memorial Hospital. (1965). 33 Illinois 2d 326, 211 N.E. 253, cert. denied 383 U.S. 946. Donabedian, A. (1982). The criteria and standards of quality. Ann Arbor, MI: Health Administration Press. Drucker, P. (1980). Managing in turbulent times. New York: Harper and Row. Freidson, E. (1970). Profession of medicine. New York: Harper and Row. GreatBoards.org. (2004). FAQ about governance: How many physicians should a hospital board have? [Electronic Version]. Retrieved on February 7, 2007 from http://www. greatboards.org/faq/indes.asp?topic ¼ 9 Gribbin, J. (2007). Quality and patient safety: Governance at the crossroads. Trustee, 60(8), 34–36. HealthGrades. (2007). HealthGrades ﬁfth annual hospital quality and clinical excellence study [Electronic Version]. Retrieved on February 14, 2007. Institute for Healthcare Improvement. (2006a). How to guide: Governance leadership. Cambridge, MA: IHI. Institute for Healthcare Improvement. (2006b). Introduction to the 5 million lives campaign [PowerPoint Presentation].

The Trustee and Patient Safety

147

Institute for Healthcare Improvement. (2006c). Protecting 5 million lives from harm [Electronic Version] from www.ihi.org Institute for Healthcare Improvement. (2007). In good company: The gathering force of board engagement with quality and safety. July 5, Electronic Version from www.ihi.org Johnson, E., & Johnson, R. (1986). Hospitals under ﬁre. Rockville, MD: Aspen Publishers. Joint Commission on Accreditation of Healthcare Organizations (JCAHO). (2003). Hospital accreditation standards. Oak Brook Terrace, IL: Joint Commission. Joint Commission on Accreditation of Healthcare Organizations (JCAHO). (2004). Hospital accreditation standards. Oak Brook Terrace, IL: Joint Commission. Kaluzny, A. D., & Veney, J. E. (1972). Who inﬂuences decisions in the hospital? Not even the administrator really knows. Modern Hospital, 119(6), 52–53. Kohn, L. T., Corrigan, J. M., & Donaldson, M. S. (Eds). (2000). To err is human: Building a safer health system. Washington, DC: National Academy Press. Landon, B., Reschovsky, J., & Blumenthal, D. (2003). Changes in career satisfaction among primary care and specialist physicians, 1997–2001. Journal of the American Medical Association, 289(4), 442–449. Light, D. (1997). The restructuring of the American healthcare system. In: T. Litman & L. Robins (Eds), Health policy and politics (3rd ed.). New York: Delmar. Longo, D., & Avant, D. (1994). In: R. Taylor & S. Taylor (Eds), Managing Quality. The AUPHA Manual of Health Services Management (pp. 45–59). Rockville, MD: Aspen Publishers. Longo, D., Hewett, J., Ge, B., Schubert, S., & Kiely, R. (2007). Hospital patient safety: Characteristics of best-performing hospitals. Journal of Healthcare Management, 52(3), 188–204, discussion 204–205. MacEachern, M. (1946). Hospital organization and management. Chicago, IL: Physicians Record Company. Margolin, F., Hawkins, S., Alexander, J., & Prybil, L. (2006). Hospital governance: Initial summary report of 2005 survey of CEOs and board chairs. Chicago, IL: Health Research and Educational Trust. McGibony, J. (1969). Principles of hospital administration (2nd ed). New York: G. P. Putnam’s Sons. Mintzberg, H. (1981). Organizational design: Fashion or ﬁt? Harvard Business Review, 59(1), 103–116. Mintzberg, H. (1989). Mintzberg on management: Inside our strange world of organizations. New York: The Free Press. Mintzberg, H., & Quinn, J. B. (1998). Readings in the strategy process. Upper Saddle River, NJ: Prentice Hall. Mn. US Statutes. 144 Section 144.706 through 144.7069. National Center for Healthcare Leadership. (2006). Best practices for measuring the effectiveness of nonproﬁt healthcare boards. Bulletin of the National Center for Healthcare Leadership, December, 1–16. O’Hare, P. (2002). Sarbanes-Oxley raises red ﬂag for not-for-proﬁts. HealthCare Financial Management, 56(10), 42–44. Orlikoff, J. (1994). Governance. In: R. Taylor & S. Taylor (Eds), The AUPHA manual of health services management (pp. 160–181). Rockville, MD: Aspen Publishers. Orlikoff, J. (2005). Building better boards in the new era of accountability. Frontiers of Health Services Management, 21(3), 3–12.

148

RICHARD A. CULBERTSON AND JULIA A. HUGHES

Pozgar, G. (2007). Legal aspects of health care administration (10th ed). Sudbury, MA: Jones and Bartlett. Rosenberg, C. (1987). The care of strangers; the rise of America’s hospital system. New York: Basic Books. Schulz, R., & Johnson, A. (1976). Management of hospitals. New York: McGraw-Hill. Schulz, R., & Johnson, A. (1983). Management of hospitals (2nd ed). New York: McGraw-Hill. Solucient. (2006). 100 top hospitals. Retrieved on February 15, 2007, from http://www. 100tophospitals.com/ Starr, P. (1982). The social transformation of American medicine. New York: Basic Books. Taylor, B., Chait, R., & Holland, T. (1996). The new work of the nonproﬁt board. Harvard Business Review, 74(5), 4–11. The Governance Institute. (2000). Value Added Governance. San Diego, CA: The Governance Institute. The Governance Institute. (2006). Quality [Electronic Version]. Retrieved on February 15, 2007 from http://www.governanceinstitute.com/_coreModules/content/contentDisplay.aspx? contentID ¼ 713 The Joint Commission. (2006). An interview with Don Berwick. Joint Commission Journal on Quality Improvement, 32(12), 666. The Joint Commission. (2007). 2007 National Patient Safety Goals, Critical Access Hospital Program. Retrieved on February 15, 2007, from http://www.jointcommission.org/NR/ rdonlyres/1AD5F8C0-CB2D-46F0-8052-C2AD126C1377/0/07_CAH_NPSGs.pdf The Walker Company. (2004). Board/medical staff alignment [Electronic Version]. BoardBrief, 1–3. Retrieved on February 15, 2007. Tyler, L., & Biggs, E. (2005). Getting a grip on governance. Frontiers of Health Services Management, 21(3), 37–42. Weiner, B., Shortell, S., & Alexander, J. (1997). Promoting clinical involvement in hospital quality improvement efforts: The effects of top management, board, and physician leadership. Health Services Research, 32(4), 491–510. Winokur, S., & Beauregard, K. (2005). Patient safety: Mindful, meaningful, and fulﬁlling. Frontiers of Health Services Management, 22(1), 17–32. Wolinsky, F. (1988). The professional dominance perspective, revisited. The Milbank Quarterly, 66(2), 33–47. Wood, K., & Nash, D. (2005). Mandatory state-based error-reporting systems: Current and future prospects. American Journal of Medical Quality, 20(6), 297–303. Zablocki, E., & Bader, B. S. (2006). Seven for ‘07: Key issues for boards in the new year [Electronic Version]. Great Boards, VI, 1–7. Retrieved on February 8, 2007.

PROTECTING THE PATIENT: COLLABORATING TO ACHIEVE THE IDEAL HOSPITAL WORK ENVIRONMENT Donde Batten, Gerald Goodman and Susan M. Distefano ABSTRACT Research suggests that improving hospital work environments and solving the nursing shortage are critical to improving patient care. The Houston– Galveston region created an aggressive approach to this issue by forming an unusual coalition of business, university, and hospital leaders and using a quality improvement approach. Four years later, the project has achieved over 40% participation among hospitals in the 13-county region and includes 50 hospitals employing approximately 15,000 registered nurses. The data that have been collected by this collaborative to date suggests that hospitals are taking action to improve outcomes by modifying their key initiatives to address the attributed causes of poor work environments. From 2004 to 2005, executives of top performing hospitals increasingly attributed successful work environment outcomes to an emphasis on management development and executive-driven initiatives, de-emphasizing speciﬁc employee behavior, process, and outcome-based initiatives. Patient Safety and Health Care Management Advances in Health Care Management, Volume 7, 149–161 Copyright r 2008 by Emerald Group Publishing Limited All rights of reproduction in any form reserved ISSN: 1474-8231/doi:10.1016/S1474-8231(08)07007-9

149

150

DONDE BATTEN ET AL.

The literature suggests that there are two important and interrelated issues facing hospital executives today, the shortage of registered nurses and patient safety. The existing and growing shortage of registered nurses has a substantial impact on patient safety. Many reasons are given for the shortage of registered nurses, including an increased market demand for nurses, the changing demographics of the U.S. population, a continuing change in the job opportunities for women, and the work environment in the acute care hospital setting. Registered nurses constitute an around-the-clock surveillance system in hospitals for early detection and prompt intervention when patients’ conditions deteriorate (Aiken, Clarke, Sloane, Sochalski, & Silber, 2002). The effectiveness of such surveillance is inﬂuenced, in part, on nurse availability. Adequate nurse stafﬁng is seen as important in explaining variations in hospital mortality. While technology may provide some efﬁciency beneﬁts related to nursing care, keeping nursing units staffed with already trained registered nurses provides a continuity that increases the likelihood that patient safety and overall hospital quality can be maintained at an adequate level. In 2004, Texas had only 624.5 registered nurses (RNs) per 100,000 population as compared to the national rate of 825 RNs per 100,000 population (Health Professions Resource Center, 2005). The state’s population is estimated to grow 25% from 2000 to 2010 (Texas State Data Center, 2006) and the Texas Workforce Commission estimates that demand for nurses is expected to increase 27.8% during the same period. The shortage of nurses among acute care hospitals in the Houston–Galveston region of Texas reached a critical level in 2000, when RN vacancy rates were so high that ambulances were being diverted from hospital to hospital due to nursing shortages in emergency rooms. Bidding wars erupted among neighboring hospitals, which used large sign-on bonuses to lure RNs from one hospital to another. Rather than continue the failing strategy of competition for RNs, local hospital and business leaders in the Houston–Galveston region of Texas began to discuss an aggressive, inter-institutional collaboration to address the nursing shortage. Hospitals, along with area business leaders and nursing schools, formed a collaborative with the Gulf Coast Workforce Board (serving a 13-county region including and surrounding Houston, Texas) and the Greater Houston Partnership (the region’s largest chamber of commerce) to form a coordinated action plan to reduce RN turnover and vacancy rates in the Houston–Galveston region. This chapter reports on participating hospitals’ efforts to improve the work environment on a

Protecting the Patient

151

region-wide basis through the implementation of a shared best practices and outcome measurement methodology, targeting reducing RN turnover and improving patient care and safety.

THE EFFECTS OF A NEGATIVE WORK ENVIRONMENT Hospital executives face mounting pressure to improve patient care, improve staff safety and reduce errors, deal with nursing shortages, and attain reasonable ﬁnancial margins. Although it is tempting to reduce nursing staff in this situation, there is a growing body of evidence showing that various nursing-related factors strongly impact patient outcomes. The National Quality Forum (NQF) used ‘‘nurse-sensitive’’ care to describe health care quality inﬂuenced by nursing. In 2004, the Nursing Steering Committee endorsed 15 voluntary consensus standards for nursing-sensitive care, the ﬁrst set of national standardized performance measures to assess the extent to which nurses in acute care hospitals contribute to patient safety, health care quality, and a professional work environment. In naming these measures, the NQF recognized that signiﬁcant gaps exist in terms of scientiﬁc evidence and research (NQF, 2004). Research evidence describing the relationships between nursing factors and patient outcomes is increasing. Much of the research examines the impact of nurse stafﬁng on patient outcomes. The overall picture shows the complexity of the issues. For example, research has shown that higher levels of inpatient RN stafﬁng are associated with lower rates of adverse outcomes in both medical and surgical patients (Needleman, Buerhaus, Mattke, Stewart, & Zelevinsky, 2002) and that in hospitals with high patient-to-nurse ratios, surgical patients experience higher risk-adjusted 30-day mortality and failure-to-rescue rates (Aiken et al., 2002). Cost estimates showed that raising the proportion of RNs without changing hours worked might not reduce hospital costs related to nurse wages, but patients treated in these hospitals would likely beneﬁt from reduced lengthof-stay and fewer adverse outcomes (Needleman, Buerhaus, Stewart, Zelevinsky, & Mattke, 2006). Increases in the number of patients per nurse have been shown to increase 30-day mortality and failure-to-rescue among adult surgical patients, as well as a greater likelihood of burnout and job dissatisfaction (Aiken et al., 2002). Other research has shown that the use of extended work shifts and overtime signiﬁcantly increased errors when nurses

152

DONDE BATTEN ET AL.

worked longer than 12 h, worked overtime, or worked more than 40 h/week. In fact, risks begin to increase when shift durations exceeded 8.5 h (Rogers, Hwang, Scott, Aiken, & Dinges, 2004). The Institute of Medicine has recommended that working 60 h/week or less be the threshold for minimizing errors related to fatigue (Institute of Medicine, 2004). The problem has become so recognized that media reports occur frequently. A 2000 Chicago Tribune three-part series on hospital safety included an analysis of three million medical records and concluded that a harried work environment compromises patient welfare, often leading to injury and death (Berens, 2000). Decisions about stafﬁng affect the work environment of nurses and may adversely affect job satisfaction and turnover (Blegen, Goode, & Reed, 1998; Brooks, 2000). The negative aspects of work environments in hospitals, both in terms of structural and teamwork-related factors are well documented, especially with respect to their impact on nurses (Kimball & O’Neil, 2002). A 2002 Advisory Board study of drivers of nurse job satisfaction and turnover reported that more than one-half of nurses were only ‘‘somewhat satisﬁed’’ with their jobs and that more than one-fourth of nurses are ‘‘dissatisﬁed’’ with their jobs. More than one-half of the more than 1,600 nurses surveyed reported that their satisfaction had decreased over the past two years. The chance that a nurse would contemplate leaving their current employment more than doubled when satisfaction drops from ‘‘very satisﬁed’’ to ‘‘somewhat satisﬁed.’’ The study found that turnover can increase when any of the following factors are dissatisfying to nurses: total compensation, scheduling options, intensity of work, professional growth opportunities, competence of clinical staff, support services, effectiveness of direct manager, participation in decisionmaking, and recognition. However, since most nurses indicated that they still believed in a long-term career in nursing, it is likely that hospitals can improve retention by improving these facets of the work environment (The Advisory Board, 2000). Low morale and burnout are not unique to the United States; most nurses do not attribute these problems to physicians or peer nurses’ clinical competency, instead they assign blame to stafﬁng adequacy and workforce management issues (Aiken et al., 2001). From a more positive perspective, a recent case study of nurses in an academic medical center found that characteristics of the practice setting, such as nurse managers, job characteristics, management style, and service quality emphasis, are critical for developing satisfaction, improving organization commitment, reducing job tension, and reducing turnover (Smith, Hood, Waldman, & Smith, 2005). One structural modeling study in a Magnet hospital found

153

Protecting the Patient

that several characteristics of the work environment (leadership, teamwork, staff expertise, staff stability, and resources) were linked to patient outcomes (Capuano, Bokovoy, Hitchings, & Houser, 2005).

THE INTERVENTION The Houston–Galveston region has recognized the importance of collecting longitudinal data on hospital work environments for more than ﬁve years. In 2000, the Houston–Galveston Council (the Workforce Investment Board serving the 13-county region including and surrounding Houston, Texas) joined in partnership with the Greater Houston Partnership (the region’s largest chamber of commerce) to address the shortage of nurses and retention issues. In collaboration with area hospitals and nursing schools, they formed the Gulf Coast Health Services Steering Committee. Area hospitals encompass all types of missions: community hospitals, academic medical centers, state/county hospitals, VA hospitals, children’s hospitals, charitable entities, and various specialty hospitals. The Gulf Coast region consists of counties facing a substantial current nurse shortage and large expected annual growth rates for nursing jobs. Started as a means to ﬁnd solutions to the nursing shortage, the committee has formed four workgroups in order to market career opportunities in health care, work with government to increase funding for nursing education, enhance educational capacity through collaborations between nursing schools and hospitals, and improve the work environment in area hospitals. Substantial outcomes have been achieved in all workgroups. One of these workgroups, the Work Environment Workgroup (WEW) began working in 2000 to develop a strategy for improving nurse retention. This volunteer group, consisting of human resources and nursing executives from regional hospitals as well as academicians and consultants, has signiﬁcant experience in hospital administration, hospital outcomes measurement, benchmarking, and survey design. By 2002, they had developed a four-part plan for improvement taking into account the diversity of hospitals in the region: (1) the development of an ideal hospital work environment model including ﬁve key dimensions: professionalism, interpersonal communications and teamwork, staff training and development, environmental factors, and economic considerations (see Table 1); (2) the collection of nurse-sensitive work environment measurement data and examples of practices that have successfully improved work environments, (3) the publication of regional norms and top performance levels as

154

DONDE BATTEN ET AL.

Table 1. Dimensions of an Ideal Hospital Work Environment. Dimension Professionalism

Interpersonal communications and teamwork

Staff training and development

Environmental factors

Economic considerations

Supporting Elements Found in Ideal Hospitals Core values include professionalism Board supports professionalism Senior leadership participates in establishing professionalism as a value Part of executives’ pay is contingent on professionalism Behavior is measured and rewarded based on well-known professional standards and expectations Selection, evaluation, and credentialing of physicians and staff-incorporate professionalism Professionalism is incorporated into the organizational structure Team training creates an environment of professionalism Professionalism is demonstrated in medical error reporting Open environment facilitates communication Shared decisionmaking is practiced Communication and team behavior competencies are demonstrated Hospital is recognized as one of best places to work in the region A strong commitment to lifelong learning is demonstrated Staff pursuit of career development is supported Career counseling is available on-site Ongoing, comprehensive mentoring programs are provided Senior leaders and managers receive ongoing training Training supports diverse work environments Training supports business issues Staff instructors are provided for nursing schools Hospital collaborates with providers of education for career and professional development Cutting edge technology is used for delivery of training Staff regularly participate in decisionmaking Factors affecting retention are identiﬁed A safe work environment is created Decisions about staff workload are data-driven Schedules support balanced work and home life Preceptors are dedicated Work is designed to be meaningful and rewarding Patients and staff have positive perceptions of staff Accommodations are provided for older patient care professionals A fun work atmosphere is created Staff receive recognition for excellent work Total compensation is high for point-of-service jobs Patient care total compensation is superior to less stressful work Compensation and work environment create positive career options, increasing retention Compensation is based on demonstrated core competencies

155

Protecting the Patient

well as the preparation of conﬁdential, hospital-speciﬁc benchmarking reports; and (4) the initiation of an annual regional conference to recognize and share practices that have successfully improved the work environments in regional hospitals. The WEW members and their colleagues developed the Ideal Hospital dimensions based on their experience working in hospital environments and conducting work environment research. The dimensions’ descriptive format was designed to help other managers and executives understand the process of implementing dimension concepts in the pragmatic hospital setting. The goal was to provide information valuable to each type of hospital without being prescriptive. In 2003, the WEW received funding from the Division of Nursing, Bureau of Health Professions, Health Resources Services Administration under the Nurse Education, Practice and Retention program and the Ideal Hospital project began. The WEW now has a four-year history of participating in outcome measurement, benchmarking, and best practice sharing. Work environment surveys collect outcome data and are reviewed for potential revision each year. Table 2 shows that participation in the project has grown, despite the devastating effects of hurricanes in 2005. Nearly half of the hospitals in the region support and participate in this type of research. This is an achievement, given the growth and changes in hospitals and hospital systems in the region during the last four years. A diverse cross-section of hospitals has participated in the project: for proﬁt, non-proﬁt and government; urban and rural; general acute care, LTAC, and specialty; large and small, with bed sizes ranging from less than 100 to over 500 beds. Hospitals participating in 2003–2006 represented from 60% to 71% of the hospital beds in the region and employed from 13,000þ to 15,000þ RNs. Five Magnet hospitals (which have received the Magnet designation by the American Nursing Credentialing Center) have participated in the project since its inception.

Table 2.

Participation in the Ideal Hospital Project.

Number of hospitals participating in the project Number of hospitals in the region Percent of regional hospitals participating in the project (%)

2003

2004

2005

2006

45 91 49

49 95 51

43 101 42

50 121 41

156

DONDE BATTEN ET AL.

PROJECT OUTCOMES Turnover and Patient Safety The collaborative has collected data over a three-year period on nurse turnover as well as for nurse critical measures including medication error rate, ICU infection rate, ventilator pneumonia rate, and surgical site infection rate. Hospitals were asked to use prescribed formulas to calculate these measures for data collected between October 1 and September 30 of the following year. Nurse (RN) turnover is a classic work environment measure. Hospitals reported data only for full-time hospital RNs, not including research entities, academic organizations, or other subsidiary organizations. They excluded contract or agency staff members not on the hospital’s payroll. Hospitals reported headcounts at the beginning and end of the year, not FTEs. Terminations were measured as the number of individuals deleted from the hospital’s payroll for any reason (voluntary or involuntary) including transfers to another hospital in a multi-hospital system. The turnover formula used was the number of terminations divided by average headcount. Data reported are median data across participating hospitals. From 2003 to 2005, the percentage of participating hospitals reporting this data has increased from 73% to 91%. Table 3 shows that nurse turnover decreased 3.5% between 2003 and 2005. The region’s 2002 nurse turnover rate was 23.38%. The region’s 2005 nurse turnover rate (14.26%) is a full percentage point lower than the Texas average hospital RN turnover rate of 15.6% (Texas Department of State Health Services, 2004). We are not able to claim that our project activities caused decreases in the nurse turnover. We must consider that there are probable reporting errors given that data were collected from over 40 hospitals. A proposal for a more aggressive data analysis approach is being developed. However, we are able say that because of a collaborative Table 3.

Changes in Turnover Rates from 2003 to 2005.

Nurse turnover rate (terminations per average headcount)

2003 Median

2004 Median

2005 Median

14.78

14.56

14.26

157

Protecting the Patient

representing 40–50 health care provider organizations and approximately 15,000 RNs in the same geographic area, bidding wars for RNs among area hospitals have ended, nurse turnover has been reduced, and best practices are being shared. Data on patient outcomes related to nurse stafﬁng continues to be collected from the hospitals in the collaborative as part of our ongoing effort to validate the relationship between the hospital work environment and reduced nurse turnover, and improved patient outcomes.

Attributions of Causality In 2004 and 2005, executives from hospitals that were identiﬁed as top performers on selected measures were asked to indicate what practices were keys to successful outcomes. Top performance was identiﬁed in terms of these indicators: all employee turnover, RN turnover, all employee vacancy rate, RN vacancy rate, and percentage of RN staff giving the highest rating of satisfaction at their hospital. Patient safety ratings were not used because of the lower reporting rate. The threshold for top performance was meeting or exceeding a score equal to or better than the top quartile performance score established by the previous year’s survey. Sixty percent of the hospitals met the threshold on at least one measure in 2005 and were designated top performers, a large increase from 2004, when 39% of hospitals met at least one of these criteria and were designated top performers. In 2004, one hospital-designated executive from each top performing hospital (either the chief executive ofﬁcer, the chief nursing ofﬁcer, or the chief human resources ofﬁcer) was asked to select or write in the top three initiatives that they believed were the keys to their hospital’s success regarding their top performance. In 2005, the same process was used except that executives from all hospitals were asked to identify initiatives related to successful outcomes and were not told if their hospitals were top performers. Executives could select responses from a list of 13 items developed by the WEW in 2004 as being representative of initiatives that regional hospitals were considering to improve their work environments. They could also write in any response. In 2004, 95% of the top performing hospitals responded to this question (18 hospitals); in 2005, 73% of the top performing hospitals responded to this question (19 hospitals). In 2005,

158

DONDE BATTEN ET AL.

85% of the hospitals that were not designated top performers responded to the question (10 hospitals). Occasionally, a hospital executive selected less than three or more than three initiatives (especially the ﬁrst year of the study in 2004); all of these data were included since no ranking was available to determine which were the top three initiatives. Not shown are the write-in responses. In 2005, there was only one write-in response: teamwork. In 2004, seven hospitals provided nine write-in responses: four hospitals indicated ‘‘our values,’’ and one hospital each wrote in ﬁve speciﬁc responses concerning location, wound care, culture, and/or management staff. Table 4 shows the percentage of hospitals that selected each initiative by top performers in 2004 and 2005 and by non-top performers in 2005. It also shows the percentage change in response by top performers from 2004 to 2005. The initiatives are listed in order of frequency of selection in 2004. All of the initiatives listed in the survey were selected by at least one top performing hospital in the 2004 survey, but in 2005, no top performing hospital selected employee scheduling. In 2004, the most commonly selected initiative was professional behavior by employee and physicians, with 60% of hospitals selecting it as a key initiative. It is interesting to note that the 2004 survey occurred 6 months after that year’s annual conference. The conference plenary speaker described a hospital intervention, focused on increasing staff professionalism, which led to improved work environment outcomes; also tools for implementing professionalism interventions were provided. Thus, it appears that this information had an impact on attending nursing and human resources executives and managers. In 2005, the most commonly selected initiative (by both top and non-top performers) was management development, with 58% of top performers and 70% of non-top performers selecting that key initiative. This survey also occurred 6 months after that year’s annual conference in which the project director reported ﬁndings from the 2004 survey showing that the training hours for supervisors and managers were proportionately low compared to other training hours. Perhaps many executives decided that management development was an under-utilized lever in creating and sustaining ideal hospital work environments and took action. In any case, the median number of training hours per supervisory and management employee increased 191% from 2004 to 2005. In general, among top performers, from 2004 to 2005, there was a de-emphasis from speciﬁc employee behavior, process, and outcome-based initiatives to management development and executive initiatives.

Percentage of Hospitals Selecting Initiatives as Keys to Successful Performance.

Initiatives

Top Performing Hospitals 2004 (%)

Top Performing Hospitals 2005 (%)

Top Performing Hospital (Percent Change 2004–2005, %)

Non-Top Performing Hospitals 2005 (%)

61

37

39

10

50 44 39 33

37 16 0 5

26 67 100 85

40 10 30 0

33 33 28 28

26 16 58 16

21 51 þ107 43

50 40 70 2

28

42

þ50

10

11

10

9

0

Senior management emphasizes professional behavior by employees and physicians Employees are involved in decisionmaking on their units Employee safety is a priority Employee scheduling supports work life balance Shared decisionmaking is practiced between staff and physicians A fun work atmosphere is created Employee career development is supported Management development is provided Employees use the latest technology to learn and to treat patients An executive-led initiative is stimulating signiﬁcant culture improvements Effective physician-employee conﬂict resolution programs are in place and utilized Compensation includes ﬁnancial incentives tied speciﬁcally to professional behavior Non-ﬁnancial recognition programs are implemented for staff

11

10

9

0

11

16

þ45

20

N

18

19

Protecting the Patient

Table 4.

10

159

160

DONDE BATTEN ET AL.

CONCLUSION This project has achieved substantial success in building collaboration among disparate hospitals in the region and improving performance indicators related to the work environment and patient outcomes. The usefulness of this project in improving work environments is not limited to large regions. The same ideas can be used to affect positive changes at any level within a single organization or across organizations. The project’s success was made possible by the inclusion of these design features: (1) engaging non-traditional stakeholders (local area businesses awakened to the nursing shortage as a quality of life issue for all residents) at the project inception, creating a sense of urgency and enabling the acceptance of a quality improvement philosophy, (2) creating a model that can be adapted for various organizations and/or unique cultures (large or small, urban or rural, academic or community-based), and (3) sharing outcome data and recommended practices in a non-competitive, nonpunitive environment. Hospital administrators are asked to improve work environments, hospitals outcomes, and patient outcomes, largely without substantive data to support their decisions. This project provides participating hospitals with conﬁdential reports showing how they perform on outcome measures relative to regional data. The project also provides access to regional hospital colleagues and recommended practices through the annual conference. Based on increased participation in the project and individual hospitals reporting improved performance outcomes, this approach appears to have merit. Finally, the key initiative survey shows that administrators are very responsive to shifting emphasis from one initiative to another as they seek to improve their work environments and outcomes. They also shift their causal attributions for that success as they make changes and review outcomes. Most often researchers deride these decisions as following fads or fashions. However, this study shows that executives are interested in quality improvement methods and benchmarking data to ﬁne-tune their strategies.

ACKNOWLEDGMENTS The project is supported by funds from the Division of Nursing (DN), Bureau of Health Professions (BHPr), Health Resources Services Administration (HRSA), Department of Health and Human Services (DHHS)

161

Protecting the Patient

under Grant #D66HP01391 – Nurse Education, Practice and Retention: Enhancing Patient Care Delivery Systems. The information or content and conclusions are those of the author and should not be construed as the ofﬁcial position or policy of, nor should any endorsements be inferred by the DN, BHPr, DHHS or the U.S. Government.

REFERENCES Aiken, L. H., Clarke, S. P., Sloane, D. M., Sochalski, J., Busse, R., Clarke, H., et al. (2001). Nurses’ report on hospital care in ﬁve countries. Health Affairs, 20(3), 43–53. Aiken, L. H., Clarke, S. P., Sloane, D. M., Sochalski, J., & Silber, J. H. (2002). Hospital nurse stafﬁng and patient mortality, nurse burnout, and job dissatisfaction. Journal of the American Medical Association, 288(16), 1987–1993. Berens, M. J. (2000). Nursing mistakes kill, injure thousands. Chicago Tribune, September 10, 2000 Blegen, M. A., Goode, C. J., & Reed, L. (1998). Nurse stafﬁng and patient outcomes. Nursing Research, 47(1), 43–50. Brooks, I. (2000). Nurse retention: Moderating the ill-effects of shift work. Human Resources Management Journal, 10(4), 16–51. Capuano, T., Bokovoy, J., Hitchings, K., & Houser, J. (2005). Use of a validated model to evaluate the impact of the work environment on outcomes at a magnet hospital. Health Care Management Review, 30(3), 229–236. Health Professions Resource Center. (2005). Supply Trends Among Licensed Health Professions – Texas 1980–2004. (E-publication #25-11847). Retrieved on January 20, 2007, from www.dshs.state.tx.us/chs/hprc/05trends.pdf Institute of Medicine. (2004). Keeping patients safe: Transforming the work environment of nurses. Washington, DC: The National Academies Press. Kimball, B., & O’Neil, E. (2002). Health care’s human crisis: The American nursing shortage. Princeton, NJ: Robert Wood Johnson Foundation. National Quality Forum. (2004). National voluntary consensus standards for nursing-sensitive care: An initial performance measure set. A consensus report. Washington, DC: NQF. Needleman, J., Buerhaus, P. I., Mattke, S., Stewart, M., & Zelevinsky, K. (2002). Nurse stafﬁng levels and the quality of care in hospitals. New England Journal of Medicine, 346(22), 1715–1722. Needleman, J., Buerhaus, P. I., Stewart, M., Zelevinsky, K., & Mattke, S. (2006). Nurse stafﬁng in hospitals: Is there a business case for quality? Health Affairs, 25(1), 204–211. Nursing Executive Center. The Advisory Board. (2000). The nurse perspective: Drivers of nurse job satisfaction and turnover. Washington, DC: The Advisory Board. Rogers, A. E., Hwang, W., Scott, L. D., Aiken, L. H., & Dinges, D. F. (2004). The working hours of hospital staff nurses and patient safety. Health Affairs, 23(4), 202–212. Smith, H. L., Hood, J. N., Waldman, J. D., & Smith, V. L. (2005). Creating a favorable practice environment for nurses. Journal of Nursing Administration, 35(12), 525–532. Texas State Data Center. (2006). Summary of Trends in Texas State Data Center 2006 Population Projections. Retrieved on January 20, 2007, from http://txsdc.utsa.edu/tpepp/ 2006projections/summary/

PRIVACY, SECURITY AND THE NATIONAL HEALTH INFORMATION NETWORK: A MIXED METHODS CASE STUDY OF STATE-LEVEL STAKEHOLDER AWARENESS Kimberly A. Galt, Karen A. Paschal, Amy Abbott, Andjela Drincic, Mark V. Siracuse, James D. Bramble and Ann M. Rule ABSTRACT This mixed methods multiple case study examines the knowledge, understanding, and awareness of 25 health board/facility oversight managers and 20 health professional association directors about privacy and security issues important to achieving health information exchange (HIE) in the state of Nebraska. Within case analyses revealed that health board/facility oversight managers were unaware of key elements of the federal agenda; their concerns about privacy encompassed broad deﬁnitions both of what constituted a ‘‘health record’’ and ‘‘regulations centeredness.’’ Alternatively, health professional association leaders were Patient Safety and Health Care Management Advances in Health Care Management, Volume 7, 165–189 Copyright r 2008 by Emerald Group Publishing Limited All rights of reproduction in any form reserved ISSN: 1474-8231/doi:10.1016/S1474-8231(08)07008-0

165

166

KIMBERLY A. GALT ET AL.

keenly aware of national initiatives. Despite concerns about HIE, they supported information exchange believing that patient care quality and safety would improve. Cross-case analyses revealed a perceptual disconnect between board/facility oversight managers and professional association leaders; however, both favored HIE. Understanding statelevel stakeholder perceptions helps us further understand our progress toward achieving the national health information interoperability goal. There is an ongoing need to assure adequate patient privacy protection. Licensure and facility boards at the state level are likely to have a major role in the assurance of patient protections through facility oversight and provider behavior. The need for these boards to take an active role in oversight of patient rights and protections is imminent. Similarly, professional associations are the major vehicles for post-graduate education of practicing health professionals. Their engagement is essential to maintaining health professions knowledge. States will need to understand and engage both of these key stakeholders to make substantial progress in moving the HIE agenda forward.

In 2004, the President of the United States issued Executive Order 13335, solidifying the promotion of HIE through conceptualizing a national health infrastructure network (NHIN) of technology (Incentives, 2004). The purpose of the order is to increase efﬁciency, reduce medical errors, and improve quality of care through HIE for patients and care providers in both the public and private sectors. The core vehicle of the health information content in the NHIN is the electronic health record (EHR) and the order targets the year 2014 for all Americans to have exchangeable electronic health records. In 2005, a U.S. Department of Health and Human Services (DHHS) leadership guidance group on strategic NHIN decisions recognized that widespread health information technology (HIT) adoption and HIE could not succeed without buy-in from the health consumer public. Clear themes emerged from the public in response to a published request for information (RFI) about whether and how an NHIN could be developed (Osborn, 2005). The public indicated that an NHIN should be a decentralized architecture with open standards and policies, have a governance entity composed of public and private stakeholders to oversee this, and reﬂect the interests of all stakeholders. The public wants methods to accurately verify patient identities and to address discordant inter- and intra-state laws regarding HIE. They recognized that incentives may be needed to accelerate the deployment and adoption of an NHIN, and that

Health Information Network

167

existing technologies, federal leadership, and certiﬁcation of EHRs will all be critical enablers of an NHIN. The public identiﬁed that a key challenge will be the provision of sufﬁcient safeguards to protect the privacy of personal health information. Without this, exchange was not likely to succeed. In early 2006, the state of Nebraska formed a working group to study these issues called the Health Information Security and Privacy Committee (HISPC). This gave rise to a series of questions about the awareness and engagement of key stakeholders in the state about the NHIN in order to address issues of security and privacy in ways responsive to the protection of the privacy of personal health information. The research reported here was motivated by the need to address these key questions in order for the state of Nebraska to progress and succeed at HIE. An investment in HIT is required throughout the vast array of health care provider practices and settings in order to achieve HIE. Despite the accumulating evidence that intelligent investment in HIT reduces errors, improves health care efﬁciency, and reduces waste, the adoption rate of HIT, speciﬁcally use of EHRs, remains quite low among providers of care nationally (Jha et al., 2006; States Getting Connected, 2006; Chaudhry et al., 2006). Research in the state of Nebraska indicates a similar response rate (Galt et al., 2007). The successful integration and standardization of HIT to be achieved by so many individuals, communities, organizations, government and private sector stakeholders is an immense undertaking, requiring macro and micro level alignments of several factors (see Table 1; Klein, 2006; Senate HELP Committee, 2005). As major areas of need have been identiﬁed, the federal government has responded with policy, legislation, and ﬁnancial practices to assist in forming structure and removing barriers. A timeline of major events and practices is provided in Table 2. The government has reached an informed understanding from the public that security and privacy issues are a substantive area that needed nationwide exploration, problem identiﬁcation, and solution generation. Nation wide and federal initiatives have been intensely focused on security and the privacy concerns for the last two years. One of these is the Connecting for Health Common Framework, developed by a public–private collaborative made up of more than 100 organizations (Anon, 2007). This initiative has sketched a vision for a nationwide HIE of decentralized networks united by a common framework of policies and standards that put patient privacy ﬁrst. The DHHS also awarded a contract to RTI International, and working with the National Governors Association Center for Best Practices, formed the Health Information Security and Privacy Collaboration (HISPC). This

168

KIMBERLY A. GALT ET AL.

Table 1. Macro- and Micro Level Factors Required to Align for Successful National Health Information Network. 1) Infrastructure and technology readiness An uninterrupted local/regional/national technology infrastructure for information exchange must exist National standards must be adopted for products, systems and individuals that promote and assure technical interoperability Functional local and clinical micro system technology networks must be in place 2) Economic Start up investment made by individuals and organizations in the private and public sectors Sustainable economic models are developed for the individuals, private and public organizations 3) Health Information Technology applications Functional applications for data capture must be in place Functional applications for decision support must be in place 4) User/consumer acceptance A critical number of people view the goal as a national imperative Consumer acceptance and participation must occur Health professionals and patients who use these applications and systems must be both willing and able to adopt new and changing technologies and applications 5) Legal/privacy concerns Privacy, security, professional, and business laws and regulations are not in conﬂict Organizations, health care providers, and patients are willing to participate in electronic health information exchange with an informed understanding of risks to privacy, security, and liability

contract includes health care stakeholders, including consumers, within and across 34 states and territories are in the process of assessing variations in organization-level business policies and state laws that affect electronic HIE. This is a cross-sectional look at the major areas that health care stakeholders have identiﬁed as presenting organizational level business practice challenges, policy and legal challenges related to privacy and security of electronic HIE. An interim assessment of variation across the participating states and U.S. territories of this project has been conducted (Dimitropoulos, 2006). Despite these efforts, in mid-June 2007, the Government Accountability Ofﬁce (GAO) released a report indicating that although efforts are continuing, a comprehensive privacy approach is still needed for a national strategy, and a sustainable method of tracking and monitoring actions and progress is needed within each state (Koontz & Melvin, 2007). Recognizing that health care is local, it is informative to the research problem to understand recent events in each state’s progress on the national agenda for adoption of

Health Information Network

Table 2. 2001 2003

2004

2005

2005

2005

2006 2006

2006

2006

a

169

Chronology of Federal Initiatives Related to Health Information Exchange.

Patient Safety Task Force created by DHHS with representatives from AHRQ, CDC, FDA, and CMS. Served as foundation for future relationships related to HIEa Medicare Modernization Act mandates that prescribers of Medicare-covered drugs conform to uniform e-prescribing standards no later than April 1, 2009. Prescriptions must be generated and transmitted electronically for Medicare beneﬁciaries receiving covered Medicare drugs Publication of ‘‘Strategic Framework: The Decade of Health Information Technology: Delivering Consumer-centric and Information-rich Health Care’’ by DHHS outlining an approach toward nationwide implementation of interoperable EHRS The DHHS formed the Health Information Technology Leadership Panel. This panel was formed to help the country understand how IT has transformed other industries, and how it can transform the health care industry. The panel recommended that the Federal government should act as a leader, catalyst, and convener of the nation’s HIT effort Secretary of HHS formed the American Health Information Community (AHIC), with members from the public and private sectors, to facilitate the transition to interoperable electronic health systems. The Community provides input and recommendations to the Secretary of Health and Human Services on use of common standards and how interoperability among HIT systems can be achieved while assuring that the privacy and security of those records are protected. The DHHS issued contracts to assist the AHIC in: (1) standards harmonization, (2) compliance certiﬁcation, (3) NHIN architecture, and (4) security and privacy Reduction in ﬁnancial barriers attempted through policy. CMS created exception to the ‘‘physician self-referral’’ law allowing hospitals/health care organizations to furnish hardware, software, and related training services to physicians for e-prescribing and EHRs. OIG proposed safe harbors for arrangements involving the donation of technology for e-prescribing and EHRs HL7 & NCPDP (2007). DHHS adopts national standards for e-prescribing of Medicare Part D drugs DHHS awarded a contract to the Certiﬁcation Commission for Health Information Technology (CCHIT) to develop criteria and evaluation processes for certifying EHRs The Health Information Security and Privacy Collaboration was formed through a DHHS contract awarded to RTI International and the National Governors Association Center for Best Practices to assess variations in policies and statutes that affect electronic health information exchange and propose solutions to address the variation DHHS awarded contract to American National Standards Institute to convene the Health Information Technology Standards Panel (HITSP). The HITSP is developing and implementing a harmonization process for achieving a widely accepted and useful set of HIT standards that will support interoperability among health care software applications, particularly EHRs (Terry, 2006; Leavitt & Gallagher, 2006). This was followed by contract awards totaling $18.6 million to four consortia of health care and HIT organizations to develop prototype architectures for the NHIN (Klein, 2006) President Bush signed an executive order to help increase the transparency of America’s health care system. The order directs federal agencies that administer or sponsor federal health insurance programs to adopt health IT standards and use health IT systems that conform

Agency for Healthcare Research and Quality (AHRQ), Centers for Disease Control and Prevention (CDC), Food and Drug Administration (FDA), and Centers for Medicare & Medicaid Services (CMS).

170

KIMBERLY A. GALT ET AL.

interoperable HIE (States Getting Connected, 2006). This study was conducted at the request of the HISPC in the state of Nebraska. The committee met each month for six months. Several stakeholder groups were identiﬁed as having a role in the vision of HIE. As members of the committee discussed the importance of different stakeholder groups, it became apparent that for both the health, licensure and certiﬁcation boards within Nebraska Health and Human Services, and the health professions associations boards, little information was known. The Nebraska HISPC asked researchers to explore these two stakeholder groups since they are inﬂuential leaders in both the public and private sector and are inﬂuential to their respective constituencies with respect to both education and implementation of change. No published literature on these stakeholder groups was found describing their understanding and the views that these stakeholder groups have on the subject of privacy and security of patient information as it relates to the eventual interoperability of health records. The purpose of this study is to describe, explore, and further explain how well we are translating the federal government’s HIT agenda to the state and local level by examining the awareness of two key stakeholder groups in the state of Nebraska. As we prepare to address these issues many questions arise. What do key stakeholders know about the NHIN? How has this national agenda been perceived on a regional and state level by stakeholders? What are key stakeholder concerns about patients’ rights to privacy and provider’s obligations to steward this while acting in the best interest of patients’ health care? How will security and privacy of health information issues be dealt with to make this goal achievable, and yet assure our citizens their rights? Will we be faced with changing our essential rights to privacy? Understanding state-level stakeholder perceptions of these will help us to further understand our progress toward achieving the health information interoperability goal of the nation. By studying key stakeholder awareness and knowledge in the state of Nebraska, we are providing new knowledge to further inform the gap that exists in states such as Nebraska that did not receive funding to participate in the larger cross-state HISPC initiatives. We are also contributing new knowledge about stakeholder groups within one state at a deeper level than a cross-sectional examination will provide. This is one of the ﬁrst studies to our knowledge that explores stakeholder awareness and knowledge related to the security and privacy issues of shared national concern directly with these key stakeholders using research-based methodologies. This work is useful to both national and state stakeholders who are working to establish continuity and collaboration between the federal government, state government, patients,

Health Information Network

171

and providers. Other state stakeholder groups may beneﬁt from what we have learned, and how we have gone about this exploration within the state.

THE RESEARCH STUDY A mixed methods descriptive multi-case study approach was used to explore the issues faced by the stakeholders (Creswell & Plano-Clark, 2007; Yin, 2003). Mixed methods research focuses on collecting, analyzing, and mixing both quantitative and qualitative data in a single study or series of studies. The central premise of mixed methods research is that the use of quantitative and qualitative approaches in combination provides a better understanding of research problems than either approach alone (Creswell & Plano-Clark, 2007). The rapid and aggressive series of national events outlined in Table 2 provides context for this study. This study incorporated the assumptions of an emerging design in both the paradigm traditions of constructivism and pragmatism (Guba & Lincoln, 1988; Creswell, 2003). This is a context-dependent inquiry that employs inductive data analysis. Two case studies were purposefully selected following the critical case approach. This approach permits logical generalizations and maximum application of information to other cases (Miles & Huberman, 1994). Each case is a bounded system representing a four-week period in December 2006. In qualitative methodology, a case may be an event, a process, a program, or several people (Stake, 1995). The ﬁrst is a case study of the managers of the 25 health, licensure and certiﬁcation boards in Nebraska Health and Human Services to determine their knowledge and awareness of this movement and how these groups collectively view what health record interoperability looks like from their respective roles. The second case study examines these same issues amongst individuals who are on the professional association boards that represent individuals involved in health care delivery and services in the state. Purposeful sampling of the health professions associations elected board leadership was conducted. The leadership was approached and asked to identify a minimum of two participants: one being an action-oriented leader and the second being someone who actively maintained a practice in the discipline represented by the association. More individuals could participate; however, the criteria used to select the participants were intended to maximize variation within the purposeful sample to provide us with individuals who would represent different perspectives on the central phenomenon.

172

KIMBERLY A. GALT ET AL.

Data were collected in multiple ways. Table 3 provides a detailed data source and collection matrix for both cases. The primary source of data for case 1 were responses from a self-administered questionnaire consisting of both closed- and open-ended questions of all of the managers of the health/ licensure/ certiﬁcation boards in the State of Nebraska, and managers of the facilities oversight boards responsible for hospitals, ambulatory surgical centers, rural health clinics, laboratories, hospice care, home health Table 3.

Data Sources and Collection Matrix for Multi-Case Study.

Participant Sources

Case Study 1 Practitioner Licensure/Certiﬁcation Board Managersa Facility Oversight Board Managersb HISPC Committee Members Case Study 2 Professional Association Board Ofﬁcersc Professional Association Board Membersc HISPC Committee Members a

Surveys

Focus Groups

Yes

Yes

Yes

Yes

Observations

Documents

Interviews

Yes

Yes

Yes

Yes

Yes

Yes

Yes Yes

Yes

Advanced Practice Registered Nurses, Alcohol and Drug Counseling , Athletic Training, Audiology & Speech Language Pathology, Chiropractic, Cosmetology, Dental & Dental Hygiene, Emergency Medical Services, Environmental Health Specialists, Funeral Directing & Embalming, Hearing Aid Instrument Dispensers and Fitters, Massage Therapy, Medical Nutrition Therapy, Medicine and Surgery, Mental Health Practice, Nursing, Occupational Therapy, Optometry, Pharmacy, Physical Therapy, Podiatry, Psychology, Respiratory Care Practice, Veterinary Medicine. b Ambulatory Surgical Centers, Assisted Living Facilities, Centers for Developmental Disabilities, Home Health Agencies, Hospice, Hospitals, Intermediate Care Facility for the Mentally Retarded, Laboratories, Long Term Care Facilities, Mental Health Centers/ Substance Abuse Treatment Centers, Nursing Home Administration, Outpatient Rehabilitation Agencies, Rural Health Clinics. c Nebraska Academy of Family Physicians, Nebraska Chiropractic Physicians Association, Nebraska Dental Association, Nebraska Emergency Medical Services Association, Nebraska Emergency Nurses Association, Nebraska Environmental Health Association, Nebraska Funeral Directors Association, Nebraska Health Care Association, Nebraska Hospice & Palliative Care Association, Nebraska Hospital Association, Nebraska Medical Association, Nebraska Nurses Association, Nebraska Occupational Therapy Association, Nebraska Optometric Association, Nebraska Pharmacists Association, Nebraska Physical Therapy Association, Nebraska Psychological Association, Nebraska Speech-Language-Hearing Association, Public Health Association of Nebraska, State Association of City & County Health Directors.

Health Information Network

173

agencies, outpatient rehabilitation agencies, and nursing home administration. The purpose of the survey was to learn the present level of awareness of board managers about the movement toward health record interoperability, and their impressions about the present knowledge of their own board members in this area. The overall knowledge of board managers, and their impressions about the knowledge of the professions they work with most closely, was explored. Questions on the survey are shown in Table 4. In addition, the Director of the Ofﬁce of Rural Health held two focus group meetings with the board managers to discuss the questions on the survey and answer any participant questions. Focus groups were a practical way of interviewing and facilitating conversation in case study 1 because these participants worked in the same building. The researchers also directly observed the interactions of members of the monthly state HISPC and reviewed minutes from these meetings to determine their perceptions about the knowledge and awareness of the case study participants. The primary source of data for case study 2 was board members of professional organizations. An initial exploration of the individuals overall knowledge and their impressions about their respective profession’s knowledge of the issues discussed above was similarly performed using the slightly modiﬁed self-administered questionnaire that consisted of both closed- and open-ended questions. The health professions organizations do not work in the same location. Consequently, participants in case study 2 from health professions associations were invited to have a telephone conversation with a researcher if they had interest in further discussing an issue. Observations from the state HISPC meetings and reviews of minutes from these meetings were also used. Data from these cases were analyzed through quantitative description and qualitative analysis. The quantitative description and display of these ﬁndings are published in technical reports used by the state of Nebraska (Galt & Berens, 2007a, 2007b). The qualitative analysis was conducted by ﬁrst transcribing responses to the open-ended questions into an analysis transcript. The transcripts were read and initial open codes formed. These codes were then reorganized into categories. The categories were compared for patterns and themes identiﬁed. Each case was written to interpret and make sense of the ﬁndings and naturalistic generalizations of the data made. Each case was then written to provide an in-depth picture of the ﬁndings. Both cases were then examined for cross-case analysis themes. Assertions for practices related to the boards and the professions they oversee were posed. The case studies also reveal areas requiring further description and study.

174

KIMBERLY A. GALT ET AL.

Table 4. Questions on Professional Board Leader Survey. 1. How informed generally is your organization’s leadership (ofﬁcers/board members) about this federal initiative? (Brief response.) 2. How informed generally are the individuals who practice your profession about this federal initiative? (Brief response.) 3. Please identify the topics that your profession’s practice act addresses. Check all that apply: ____ there is not a practice act that governs the professions I represent ____ a professional’s responsibility toward privacy and conﬁdentiality of patient information ____ a professional’s responsibility for sharing or release of patient/client health information records to individuals other than the patient/client ____ describe any other areas related to privacy and/or conﬁdentiality that your practice act addresses 4. Are there any barriers in the laws, rules and regulations governing your profession to exchanging health information between patients, providers and organizations? ___ No, skip to question 6 ___ Yes – brieﬂy describe: 5. If you identiﬁed any barriers, what recommendations would you make to overcome them? (brieﬂy describe) 6. What barriers do you personally see to sharing electronic health information records amongst all providers of care and the patient? (Brieﬂy describe) 7. Has your organization’s board recently discussed any of the following practice issues? Check all that apply: ____ adoption of electronic health records in practice ____ sharing patient health information about your patient if shared access and exchange of health information becomes available through electronic health records, ____ accessing patient health information that has been created by other health providers if shared access and exchange of health information becomes available through electronic health records. 8. Describe any concerns you may have about assuring privacy of health information if electronic health records become available through shared access? 9. Would you be willing to share your patients’ electronic health information data with others (e.g., physicians and other practitioner ofﬁces, hospitals, etc.)? ___ ___ ___ ___ ___

No ___ Yes, what data? (check Active conditions/diagnoses Physical examination ﬁndings Plan of care Progress or consultation notes

all that apply) ___ Hospitalizations ___ Medication history ___ Active prescriptions ___ Laboratory data

___ Radiologic information ___ Immunizations ___ Allergy information

10. What content areas that would be part of a comprehensive electronic health record would be helpful to you to have access to in the provision of care for your patients/clients? (check all that apply): ___ None would be particularly helpful to me ___ Active conditions/diagnoses ___ Hospitalizations

___ Radiologic information

175

Health Information Network

Table 4. (Continued ) ___ Physical examination ﬁndings ___ Plan of care ___ Progress or consultation notes

___ Medication history ___ Active prescriptions ___ Laboratory data

___ Immunizations ___ Allergy information

11. Would you share the patient data you receive from other practitioners with your patient if your patient requested it? ___ No ___ Yes 12. Besides the patient, are there any circumstances when someone else should be allowed to have a patient’s electronic health records without getting expressed permission from the patient? Check all that apply and describe the circumstances: ___ ___ ___ ___ ___ ___

No one else The patient’s family Any health provider the patients goes to for care The payer of health services (e.g. insurance) The federal, state or county government Other (describe)

13. Who, beside the patient, should have authority to give someone else access to a patient’s health information? (check all that apply and describe the circumstances): ___ ___ ___ ___ ___ ___

No one else The patient’s family Any health provider the patients goes to for care The payer of health services (e.g. insurance) The federal, state or county government Other (describe)

13. Having an electronic health record in your professional practice environment would: ___ Have a ( positive negative neutral) impact on the quality of my health care delivery ___ Have a ( positive negative neutral ) impact on the safety of health care delivery Please brieﬂy explain: Sharing an electronic health record between various types of health care providers (e.g., pharmacies, laboratories, hospitals, etc.) would: ___ Have a (positive negative neutral) impact on the quality of my health care delivery ___ Have a (positive negative neutral) impact on the safety of health care delivery 16. What risks do you personally see to the patient with sharing of electronic health information records (describe)? 17. What risks do you personally see for yourself with sharing of patients’ electronic health information records (describe)? 18. Patients keep track of their personal health information in several different ways. Please identify the ways your patients document personal health information (check all that apply): ___ List of medications on a paper record ___ List of conditions or diseases on a paper record ___ Patient carries an electronic device (like a PDA) with a PHR

176

KIMBERLY A. GALT ET AL.

Table 4. (Continued ) ___ Patient carries a SmartCard that contains personal health information records ___ Patients access a website and provide you with permission to access it 19. What is your experience with patients who have PHRs? (check all that apply): ___ ___ ___ ___ ___ ___ ___

I do not know that any of my patients use PHRs Some of my patients use PHRs I provide patients with assistance in using their PHRs A co-worker in my practice provides patients with assistance using their PHRs I am able to upload the patient PHR information into my computer system I am able to transfer information from my computer system to my patient’s PHR I am interested in integrating a PHR system into my current computerized record system

20. Are you familiar with Regional Health Information Organizations (RHIO)? ___ No (An RHIO is an electronic network for exchanging patient health information among providers) ___ Yes 21. I believe an RHIO will (select all that apply): ___ Improve health care service quality and safety ___ Facilitate a simpler patient visit by allowing ofﬁce staff to view patient records to learn of insurance coverage, deductibles, eligibility for services, procedures, and other items ___ Improve providers abilities to evaluate a patient’s continuity of care ___ Eliminate repeating patients’ histories and documenting allergies ___ Provide practitioner access to view previous lab results ___ Provide practitioner access to view patients prescription medicines they have had ﬁlled at a pharmacy ___ Improve practitioner ability to provide medication use information to other providers Please add any ﬁnal comments below. Your comments will help ensure that the quality of health care for all Nebraskans continues to improve. Gender: __ Male __ Female Age: _______ What is your profession? ______________________ Are you in active practice? __ Yes__ No What is your specialty? _________________________ How long have you been in active practice? _____ (years) Where do you practice? ______ Urban ______ Rural ______ Suburban

Establishing validity in a mixed methods research design is challenging. We have accepted the deﬁnition of validity in this mixed methods design as the ability to draw meaningful and accurate conclusions from all data in the study. Data were analyzed by one researcher and peer reviewed by a second researcher, both familiar with the qualitative data analysis approaches and knowledgeable about the subject matter. The threats to validity were minimized in the data collection process by collecting both quantitative and

177

Health Information Network

qualitative samples from the same population, and using unobtrusive data collection procedures (written survey) as the primary data source. During analysis, validity threats were reduced by asking the same questions of both the qualitative and quantitative data (Creswell, 2007). A comparison of the ﬁndings of the case studies to be addressed within our state and its’ communities to the HISPC 34 state/territory interim report were similar, providing one form of external validity (Dimitropoulos, 2006).

CASE STUDY 1 The 24 health, licensure and certiﬁcation boards and managers of facilities oversight have a mission to assure the public that health-related practices provided by individuals, facilities, and programs are safe, of acceptable quality, and that the cost of expanded services is justiﬁed by the need. This mission is accomplished through licensing and certifying individuals to provide health care services, inspecting or surveying health care and child care facilities, and reviewing capital expansions proposed for nursing homes and the sale of hospitals. The Nebraska Health and Human Services Department of Regulation and Licensure establishes and maintains regulations for public protection and provides regulatory education and consultation. Of particular relevance is the role of these boards with consumers, families, and the general public. Boards have the responsibility to disseminate information to these constituents, especially as it relates to factors affecting provider decisions regarding care and services and user and family rights and responsibilities.

CASE STUDY 1 THEMES A within case analysis of the responses to the written research questions was coded for themes. A separate analysis of the notes taken from the small focus groups was performed. A cross-comparison of these two data sources revealed the following themes (Galt & Berens, 2007a, 2007b).

Lack of Awareness Health board and facility oversight managers are generally not aware or have not discussed electronic health record exchange, have had recent discussions about privacy of health information, and do not have an

178

KIMBERLY A. GALT ET AL.

awareness of the concept of a Regional Health Information Organization (RHIO) or NHIN. These managers are generally unaware of the health information interoperability and record exchange initiative in the state. They further indicate that the practitioner groups they oversee have not had these conversations or they are unsure if this has taken place. It is striking that almost all of these managers represent groups which have some health information that would be part of a patient/client/consumer comprehensive health record or information of importance to public. The exception to this ﬁnding is the Emergency Medical Services group which has discussed the barriers regarding transferring health records to other providers and facilities. However, many have not had these discussions, reinforcing the need for education on the subject.

Conﬁdentiality and Privacy Concerns are Regulations Centered Almost all practitioners have speciﬁc rules or regulations that address issues of privacy, conﬁdentiality, sharing or releasing records created as a result of providing care or service to individuals, patients, clients or consumers. Board and oversight managers report that maintaining conﬁdentiality of all information obtained from the patient/client/consumer is a consistent and common requirement, rule or regulation throughout the practitioner environment. Audiology and Speech-Language Pathology also have an interesting requirement: transfer of care must occur to another health care professional and it is unprofessional conduct to not facilitate this when needed. It may be worth further exploration about how this will be achieved while still maintaining conﬁdentiality. A few board managers indicate that there have been some recent discussions about assuring the privacy of health information. However, overall, most are not.1 The Mental Health Practice board has recently sponsored a clinical records continuing education workshop where issues such as internet counseling, telephone conversations, and written records were discussed. In general, facilities oversight is unsure of the engagement of board members providing more evidence that the education of board members is needed. Managers described they are most concerned with abiding by state and federal rules or regulations governing the security and privacy of health records. Board managers are in need of information and education about HIE and interoperability, given their responsibilities and concerns. These board managers need to be prepared to address how current rules and regulations affect and are affected by this initiative.

179

Health Information Network

The Meaning of ‘‘Health Record’’ is Broad The scope of what constitutes a ‘‘record’’ is broad, as each provider has varying content which is of most importance to his own work. The healthrelated data primary to one type of provider may be unique and essential to the provider in their work with the patient. Some data elements overlap, such as, diagnoses or conditions. Several board managers identiﬁed the patient medications as a speciﬁc data element. Even funeral directors have ofﬁcial health record information.

CASE STUDY 2 The second case is the exploration of the state’s health professional association leadership. The health professions organizations that provided input are listed in Table 3. These organizations are important to the widespread dissemination and adoption of change related to safety and quality practices in health care. These organizations also inﬂuence and sometimes direct the culture formation of a profession within communities. In a state such as Nebraska, they have a substantial inﬂuence within a profession because the professional association is often the main method of networking.

CASE STUDY 2 THEMES A within case analysis of the responses to the written research questions was coded for themes. A separate analysis of the telephone conversations with some individuals was performed. A cross-comparison of these two data sources revealed the following themes.

Awareness Leaders in these associations are aware of the federal initiative to achieve health information interoperability. Those representing health care organizations, such as hospitals, home care, long-term care, and city and county health departments had the greatest level of awareness, describing detailed and explicit knowledge about the executive order and overall NHIN concept. However, the associations are generally not familiar with RHIO.

180

KIMBERLY A. GALT ET AL.

Most professions have practice acts that address both: (1) a professional’s responsibility toward privacy and conﬁdentiality of patient information and (2) a professional’s responsibility for sharing release of patient/client health information records to individuals other than the patient/client. The boards are actively discussing the adoption of health records in practice, sharing, and access to health information generated by other providers and patients. From the ‘‘professional practitioner’s eyes,’’ patients keep track of health information on paper. However, early signs of HIT use by patients are present. One association board member described a patient that he sees who uses a ‘‘Smartcard’’ to track and exchange health information and a few reported patients who access websites for their personal health information. There is some experience that professionals are beginning to have with personal health records (PHRs). One individual reports assisting his patients with how to use the PHR. Similar to the associations, most professionals were not aware of what constitutes an RHIO.

Barriers to Health Information Exchange are both Personal and Professional Explicit and speciﬁc barriers emerged that need to be addressed before HIE can successfully occur. The overarching perception is that the Health Insurance Portability and Accountability Act (HIPAA) of 1996 greatly restricts information exchange. Examples were cited such as: Cannot discuss the patient’s information with family without permission’’ y and ‘‘Many people still do not understand how to apply HIPAA regulations and guidelines. They use unnecessary restrictions, which does create a barrier when trying to provide the best patient care possible. No clear answer as to what is covered on all the levels of conﬁdentiality, HIPAA, state law and others sometimes creates wall to exchange information either in who can get the information or what can be received. There are even requirements in HIPAA about HIPAA compliant providers working with nonHIPAA providers.

Another practitioner indicated that many health professionals are fearful of violating rights and statutes. Other barriers were identiﬁed around lack of technology available throughout all organizations and practices, and lack of standardization and system interoperability. Solutions to barriers involved either changing legislation or providing education.

Health Information Network

181

Risks are Present with Health Information Exchange There was a strong presence of concern for patient risks. These risks were described as: (1) unnecessary access to information others do not need, (2) violations of privacy, (3) increased medical errors due to interpretation problems (information used the wrong way), and (4) getting patient’s records confused. ‘‘We need to protect the patient – what information can be used against them? This is a way in the rural area to achieve appropriate patient care. However the concern is that this will still be misused.’’ Only people who need to know the information about a patient should see it; not everyone in the ofﬁce or facility. ‘‘There is already too much personal information on the Internet.’’ Several people identiﬁed a personal risk to professionals themselves. There is an element of mistrust that the technology and systems will be dependable. Concerns about assuring privacy of health information centered on data security, on what content speciﬁcally is at risk, and on having up-to-date information when it is accessed. Record Sharing is Embraced – but with Caution Professionals will share the information they obtained from other practitioners with their patients, citing respect for the patient as the primary owner of the information. However, there are boundaries to this. One person pointed out that he would not feel comfortable releasing information that he did not have the background to explain. Another indicated he would do so within HIPAA compliance regulations. The professions have a wide range of beliefs about who else should be allowed to have a patient’s EHRs without getting express permission from the patient. Most commonly they identiﬁed any health provider the patient goes to for care. Fewer indicated these options: no one else, the family, the payer of health services, and the government. Most professions indicated no one else should have the authority to give someone else access to the patient’s health information besides the patient. I think it should be up to the patient how much of their medical history is available. They should be able to set preferences as to what other physicians, ER’s, & hospitals can access. I think this could be very helpful for physician’s if we had more open access, especially to immunizations, allergies, medication lists, lab results, & X-ray reports y. The state could really help with getting the immunization registry active statewide & accessible to all physicians.

Many components of the health record are considered helpful across the wide range of professionals. No one indicated that he would not ﬁnd sharing

182

KIMBERLY A. GALT ET AL.

records useful. The great majority of respondents would be willing to share their patient’s electronic health information with other providers if the patient gave permission. Most indicated they would be willing to share a wide range of data. This included most of the data elements found in a patients health record such as active conditions/diagnoses, physical examinations ﬁndings, plan of care, progress or consultation notes, hospitalizations, medication history, active prescriptions, laboratory data, radiologic information, immunizations, and allergy information. A couple of individuals indicated that emergency medical personnel should be able to access EHRs without patient permission if the patient is unable to give permission and in circumstances when the patient’s safety or well being may be compromised.

Quality and Safety for Patients will Improve with Electronic Health Records Overwhelmingly, the professions believe that having an electronic health record in their own practice environment will improve both the overall safety and quality of health care. They just as strongly believe that sharing EHRs will do the same. When they became familiar with the concept of an RHIO, the professionals believed RHIOs would provide the following positive beneﬁts: Improved health care service quality and safety. Facilitation of a simpler patient visit by allowing ofﬁce staff to view patient records to learn of insurance coverage, deductibles, eligibility for services, procedures, and other items. Improved provider ability to evaluate a patients’ continuity of care. Elimination of repetitive patient histories and allergy documentation. Practitioner access to view previous lab results. Practitioner access to view patients prescription medicines they have had ﬁlled at a pharmacy. Improved practitioner ability to provide medication use information to other providers.

CROSS-CASE ANALYSIS THEMES Themes were compared across the two cases to develop emergent themes representing the similarities and differences between the two stakeholder groups.

Health Information Network

183

Disconnect between Federal, State-Level Public Oversight, and Professional Associations The lack of awareness of the health board managers about the federal initiatives is a major disconnect in the area of security and privacy. Health board managers focused on HIPAA knowledge and did not provide any indication of awareness of the aggressive federal movement toward HIE and interoperability. Professional association leaders were far more aware about determining how they were going to deal with this knowledge. The professional associations generally knew about the NHIN and national standards for privacy and security, for example. The implications are that public protections about conﬁdentiality, security, and privacy are being stewarded by the care providers, and far less so by the public boards at the present time. Engaging the public boards is an important step toward having rights of privacy protected.

Those Who Provide Care Know The engagement of practitioners who are knowledgeable about the federal initiative to develop health information interoperability and who serve as leaders in the state professional associations are at various stages of exploration, evaluation, and use. These organizations offer a largely untapped resource to develop and advance sound security and privacy practices as the HIE effort proceeds. However, there is still a lack of knowledge among both health board managers and professional associations about developing technical infrastructures, most commonly identiﬁed as RHIOs. It seems like health professionals and patients will deal with the development of sound privacy and security practices, with oversight from health boards.

Cooperation about Data Exchange and Interoperability is Professionally Acceptable The common belief between the health board managers and professional association leaders is that if the patient gives permission or requests it, the patient’s health information can be shared. Removal of the patients’ authority over their medical records is not supported by the participants in our case studies. A strong message about how effective interoperability must work is that the patient is the gatekeeper to access.

184

KIMBERLY A. GALT ET AL.

Limitations There are limitations to this mixed methods case study approach. Case studies are not developed as a methodology intended to overtly develop a generalizable ﬁnding for the nation. Rather, case studies are employed often to understand issues within context and lessons learned. The ﬁndings in Nebraska regarding health board and facility oversight managers not being aware about the national agenda for electronic health record exchange or RHIOs, or discussing privacy of health information only from an HIPAA regulatory viewpoint, are not likely generalizable across the nation. It would be our premise that the level of awareness about RHIOs varied greatly from state to state, with the states of California and Indiana, leading in the implementation and testing of RHIOs. However, most states are not as advanced as these two, with a high degree of variation most likely observed if we were to look across the nation and compare (Robeznieks, 2006). In this sense, our ﬁndings may be of use to some states, particularly those that did not receive funding to participate in the federally managed multi-state and territory HISPC project. We also did not examine the voluntary board members that report to the professional licensure and facility boards. Nor did we examine representative general members of the professional associations. We certainly may have discovered awareness and knowledge beyond that of what was revealed in our case studies report. However, the value of our case study is the ﬁnding that there is a lack of connection being made by the key stakeholder leaders between the federal agenda and the state constituents. This is a critical aspect to implementing HIE across the nation. These ﬁndings indicate there is a need to focus effort on these key individuals.

DISCUSSION Boards at the state level are likely to have a major role in the assurance of patient protections through facility oversight and provider behavior. The need for these boards to take an active role in oversight of patient rights and protections is imminent. The evolving role of health, certiﬁcation and licensure boards has become important as the HIT movement continues. Similarly, the process of HIT product certiﬁcation using recently developed national standards is important. State and local policymakers have been responsible for development and standardization of the HIE focused on maintaining the security and privacy

Health Information Network

185

of personal health information (Health Information Exchange, 2006). The lack of awareness in our case study stakeholders is remarkable given the various HIT activities throughout Nebraska. Speciﬁcally, there are two active RHIO projects funded by AHRQ at the present time in Nebraska. The state has also developed and maintained the Nebraska Telehealth Network with eight hub T-line connections to 101 sites, including hospitals throughout the state providing two-way telecommunications (Nebraska, 2007). Others include the private coalition called the Nebraska Health Information group that was formed to facilitate the development of a technical infrastructure for HIE. These projects were supported by AHRQ Demonstration Project grants and two major research grants in HIT and patient safety. Nationally there have been several competitive grants and various regional demonstration contracts awarded to address HIT adoption and HIE initiatives that focus on overcoming issues related to security and privacy. Numerous partnerships have also been developed to tackle this challenge, including those with RTI International, Inc. (RTI), the Vanderbilt Center for Better Health, the California Healthcare Foundation (CHCF), SAFEHEALTH in Massachusetts, Health Care Solutions Group, and the National Opinion Research Center (NORC) at the University of Chicago. RTI, a nonproﬁt, private corporation that partnered with AHRQ and subcontracts with multiple states, manages a project to address both privacy and security policy affecting HIE among states (State Privacy, 2006). RTI was established as part of the National Health Information Security and Privacy Collaboration and, in May 2006, 33 states and Puerto Rico had subcontracted with them. Through this grant, these states and territories were awarded money to study barriers and solutions to privacy and security of the interoperability of HIE. The goal of this effort was for each of these states to do research and bring back the results of their various projects so that comparing and contrasting, as well as collaboration of the ﬁndings, could be done. In California, the governor ordered that eHealth Action Forum develop a comprehensive state policy agenda for HIT by July 2007. The CHCF was required to release recommendations to this forum on how to contribute to the State’s action plan (California, 2006). SAFEHEALTH is another such example of a state’s project to build and operate an RHIO in Central Massachusetts. At Vanderbilt University’s Center for Better Health, they have completed individual projects and have also joined forces with some of health care’s biggest names throughout the United States to form The Health Care Solutions Group. They worked in conjunction with the NORC

186

KIMBERLY A. GALT ET AL.

at the University of Chicago to establish and operate the National Health Information Technology Resource Center. The Health Care Solutions Group is a blend of experts in the health care ﬁeld, policymakers, and various stakeholder groups’ representatives (Osborn, 2005). Although some recommendations and state-level HIE materials have been developed, each community and state must engage as many stakeholders as they can in order to make decisions on what information can be exchanged, how it can be exchanged, by whom, and under what guidelines. Patients and consumers must feel like their information is secure and remains private while improving the quality and efﬁciency of health care. These stakeholders must also assist in proposing solutions and helping implement activities that address security and privacy questions with HIE. One of the federal solutions is to establish product and training standards that grant certiﬁcation, such as the CCHIT described in Table 2. State and local efforts are essential to assist in improving the way that health care is delivered to patients across the United States. According to eHealth Initiatives August 2006 report, 35 states were either in the planning or implementation phase of HIT-related efforts. Quality, safety, and rising costs of health care are the biggest drivers of these various state efforts. Numerous bills (many of which have been passed into law) and legislative activities (e.g. committees, councils, task forces, and health reform plans) emerged in 2005 and 2006. In mid-2007, state solution and implementation plans began to emerge. State subcontractors formed collaborative workgroups to deﬁne this preliminary landscape. The DHHS presented evaluation of e-prescribing pilot tests to Congress and announced the ﬁnal e-prescribing standards. After implementation of EHRs and HIE is complete, there will be an ongoing need to assure adequate patient privacy protection. Boards at the state level are likely to have a major role in the assurance of patient protections through facility oversight and provider behavior. The need for these boards to take an active role in oversight of patient rights and protections is imminent. As with many things, the bigger picture must be addressed. Although it is commendable that all of these activities are in process, the major challenge is to engage state and local stakeholders to facilitate and participate in this process. Although we cannot generalize our ﬁndings about Nebraska to others, these case studies may serve as lessons learned and reﬂect similar ﬁndings in other states. Moreover, these case studies reﬂect the need for health care managers at the Board and professional association levels to take an active role in leading the effort of their constituents to further implement the use of HIE. Their leadership

187

Health Information Network

roles may be critical in the resolution of patient privacy, security, and conﬁdentiality issues that must occur at the provider- and patient level to reach the quality and safety beneﬁts that are believed to be achievable through HIE. It is likely that much education is needed to achieve knowledge and awareness of stakeholders at the state level before such objectives can be reached.

NOTE 1. Includes Social Work, Professional Counseling and Marriage and Family Therapy.

REFERENCES About CCHIT (Certiﬁcation Commission for Healthcare Information Technology). Certiﬁcation Commission for Healthcare Information Technology. Retrieved on January 26, 2007 from http://www.cchit.org Anon (2006). California can lead the way in health information technology – Recommendations to Governor Schwarzenegger’s eHealth Action Forum. Oakland, CA: California Health Care Foundation. Retrieved on January 26, 2007 from www.chcf.org/topics/view. cfm?itemID ¼ 125646 Anon (2007). The connecting for health common framework: Resources for implementing private and secure health information exchange. Retrieved on June 2007 from http:// www.connectingforhealth.org/commonframework/ Chaudhry, B., Wang, J., Wu, S., Maglione, M., Mojica, W., Roth, E., Morton, S. C., & Shekelle, P. G. (2006). Systematic review: Impact of health information technology on quality, efﬁciency, and costs of medical care. Annals of Internal Medicine, 144(10), 742–752. Creswell, J. W. (2003). Research design: Qualitative, quantitative, and mixed methods approaches (2nd ed). Thousand Oaks, CA: Sage. Creswell, J. W. (2007). Qualitative inquiry and research design: Choosing among the ﬁve traditions (2nd ed). Thousand Oaks, CA: Sage. Creswell, J. W., & Plano-Clark, V. L. (2007). Designing and conducting mixed methods research. Thousand Oaks, CA: Sage. Dimitropoulos, L. L. (2006). Interim assessment of variation: Privacy and security solutions for interoperable health information exchange. December 29, RTI Project No. 0209825.000.004.002. RTI International, Chicago, IL. Galt, K. A., & Berens, D. (Eds.) (2007a). Report 1: Health Information Security and Privacy Committee (HISPC) State of Nebraska. Retrieved on January 25, 2007 from http:// chrp.creighton.edu Galt, K. A., & Berens, D. (Eds.) (2007b). Report 2: Health Information Security and Privacy Committee (HISPC) State of Nebraska. Retrieved on January 25, 2007 from http:// chrp.creighton.edu

188

KIMBERLY A. GALT ET AL.

Galt, K. A., Johnson, S., & the EHRNebraska Research Team: Abbott, A., Anderson, D., Bramble, J. D., Drincic, A., Fuji, K. T., Glover, D., Kasha, T., Mahlman, D., Paschal, K., Siracuse, M., & Rule, A. (2007). How many physician practices have adopted electronic health records in Nebraska? An update on the Nebraska Medical Association EHRNebraska Project. The Nebraska Medical Association. Published at http:// chrp.creighton.edu, accessed on June 20, 2007. Guba, E., & Lincoln, Y. (1988). Do inquiry paradigms imply inquiry methodologies? In: D. M. Fetterman (Ed.), Qualitative approaches to evaluation in education. New York: Praeger, 1988. Health Information Exchange (2006). Health information exchange policy issues. United States Department of Health & Human Services Agency for Healthcare Research and Quality. Retrieved on January 25, 2007 from healthit.ahrq.gov/portal/server.pt?open ¼ 514&objID ¼ 5554&mode ¼ 2&holderDisplayURL ¼ http://prodportallb.ahrq.gov:7087/ publishedcontent/publish/communities/k_o/knowledge_library/key_topics/health_brieﬁng_ 04052006112504/health_information_exchange_policy_issues.html HL7, NCPDP Collaborate on Mapping Project Efforts to Help Streamline Implementation of e-Prescribing. Retrieved on January 20, 2007 from www.hl7.org/documentcenter/public/ pressreleases/20051209.pdf Incentives (2004). Incentives for the use of health information technology and establishing the position of the national health information technology coordinator/Title 3 – The President. Executive Order No. 13335, 69 C.F.R. 24059. Jha, A. K., Ferris T. G., Donelan K., DesRoches C., Shields A., Rosenbaum S., Blumenthal, D. (2006). How common are electronic health records in the United States? A summary of the evidence. Health Affairs, 25, w496–w507, published online November/December 2006;10.1377/hltaff.25.w496 Klein, S. (2006). Issue of the month: Who has $400 billion to build a national health information network? Retrieved on September 13, 2006 from http://www.commonwealthfund.org/publications/publications_show.htm?doc_id ¼ 294918#issue Koontz, L. D., & Melvin, V. C. (2007). Testimony Before the Subcommittee on Information Policy, Census, and National Archives Committee on Oversight and Government Reform U.S. House of Representatives. Health Information Technology: Efforts Continue but Comprehensive Privacy Approach Needed for National Strategy. Document No. GAO-07-988T. www.gao.gov/cgi-bin/getrpt?GAO-07-988T, accessed on June 20, 2007. Leavitt, M., & Gallagher, L. (2006). The EHR seal of approval: CCHIT introduces product certiﬁcation to spur EHR adoption. Journal of the American Medical Informatics Association, 77(5), 26–30. Nebraska (2007). Nebraska Statewide Telehealth Network. http://www.netelehealth.org/ Accessed on November 2007. Osborn, D. (2005). Health care industry leaders launch ‘‘solutions group’’ to address U.S. health care challenges. Health Care Solutions Group. Retrieved on January 25, 2007 from http://www.healthcaresolutionsgroup.net/news/05oct25_hcsg_launch_pr.pdf Robeznieks, A. (2006). Let the testing begin. Tensions remain as EHR-system certiﬁcation process gets under way. Modern Healthcare, 15;36(20), 30. Senate HELP Committee (2005). Senate HELP Committee approves joint measure to expand use of health information technology. Kaiser Daily Health Policy Report. Retrieved

Health Information Network

189

on September 26, 2006 from www.kaisernetwork.org/daily_reports/print_report.cfm? DR_ID ¼ 31544&dr_cat ¼ 3 Stake, R. (1995). The art of case study research. Thousand Oaks, CA: Sage. State Privacy (2006). State privacy and security subcontract opportunities announced under expanded HHS contract with RTI. United States Department of Health & Human Services. Retrieved on January 25, 2007 from www.hhs.gov/news/press/2006pres/ 20060523.html States Getting Connected (2006). States getting connected: State policy-makers drive improvements in healthcare quality and safety through IT. Washington, DC: eHealth Initiative and Foundation. Retrieved on January 25, 2007 from http:// www.ehealthinitiative.org/ Terry, K. (2006). What is EHR certiﬁcation? Medical Economics, Retrieved on January 25, 2007 from http://medicaleconomics.modernmedicine.com/memag/ManagedþCareþ%26þ MedicareþPartþD//What-is-EHR-certiﬁcation/ArticleStandard/Article/detail/361457 Yin, R. K. (2003). Case study research: Design and Method (3rd ed). Thousand Oaks, CA: Sage.

EXAMINING BARRIERS TO HEALTH INFORMATION TECHNOLOGY ADOPTION James D. Bramble, Mark V. Siracuse, Kimberly A. Galt, Ann M. Rule, Bartholomew E. Clark and Karen A. Paschal ABSTRACT Results of a previous study showed that use of health information technology (HIT) signiﬁcantly reduced potential medication prescribing errors. However, the results also revealed a less than 100% rate of HIT adoption by primary care physicians. The current study reports on personal interviews with participating physicians that explored the barriers they faced when attempting to fully adopt a particular HIT. Content analysis of qualitative interviews revealed three barrier themes: time, technology, and environment. Interviews also revealed two other areas of concern; speciﬁcally, the compatibility of the HIT with the physician’s patient mix and the physician’s own attitude toward the use of HIT. A theoretical model of technology acceptance and use is used to discuss and further explain the data derived from the physician interviews. With a better understanding of these issues, health care administrators can develop successful strategies for adoption of HIT across their health care organizations. Patient Safety and Health Care Management Advances in Health Care Management, Volume 7, 191–209 Copyright r 2008 by Emerald Group Publishing Limited All rights of reproduction in any form reserved ISSN: 1474-8231/doi:10.1016/S1474-8231(08)07009-2

191

192

JAMES D. BRAMBLE ET AL.

The oft-cited Institute of Medicine (IOM) report brought the public spotlight to a myriad of patient safety problems faced across the United States (Kohn, Corrigan, & Donaldson, 2000). Following this report, increased activity in the health care community focused on ﬁnding methods to reduce medical errors and improve the safety of individuals seeking health care. In the reports, ‘‘To Err is Human’’ and ‘‘Crossing the Quality Chasm’’, one area of particular interest is the expanding use of health information technologies (HITs) aimed at error reduction (Kohn et al., 2000; IOM, 2001). Various forms of HIT are currently used in the practice of medicine and help in numerous ways to ensure safe practices. The HIT discussed in this paper is the use of personal digital assistants (PDAs) as a tool for generating electronic prescriptions; speciﬁcally, in the outpatient environment. The majority of medication prescribing originates in the outpatient arena. According to an Annual Report of the National Association of Chain Drug Stores Foundation (2005), 3.4 billion prescriptions were dispensed by US pharmacies in 2003. One of the most commonly identiﬁed prescription characteristics with signiﬁcant potential for error is that of illegibility (Brodell, Helms, & KrishnaRao, 1997; Cohen, 1999; Vitillo & Lesar, 1991). The inability to correctly read the prescription is a serious concern. The American Medical Association reported in 1997 that the second most prevalent and expensive claim in professional liability cases was the misinterpretation of physicians’ prescriptions (Cabral, 1997). The combination of high volume and high error potential provides a great opportunity to examine the impact of HIT use – not only in terms of reducing medication prescribing errors; but, even more importantly, in determining how thoroughly HIT is adopted and what barriers to adoption are encountered when HIT is incorporated into the physician ofﬁce workﬂow. As part of the current study, it was found that using PDAs in the prescribing process decreases the number of potential prescribing errors. Using a PDA reduced all errors of omission; highlighted, perhaps, by a large improvement in the reduction of dosage form omissions. Notably, overall legibility was improved. Illegible prescriptions decreased from 9.1% to 2.7%. However, not all physicians chose to adopt and use the PDA; thus, limiting the potential patient safety impact. Barriers to adoption experienced by individual users of technology must be identiﬁed and addressed. This chapter contributes to better understanding what perceived barriers inhibit the adoption of PDAs in the prescribing process.

193

Barriers to HIT

PDAS: A PATIENT SAFETY TOOL Physician clinical decision making is often an inefﬁcient process due to the delay in accessing evidence needed to make individualized patient care decisions. Thirty percent of all medical errors have been attributed to lack of access to critical information (Lesar, Briceland, & Stein, 1997). Many respected, credible information sources are now available as electronic media for use on PDA devices, and future trends are clearly to move all expert published information resources to electronic access (Arnone et al., 1998). The PDA offers one solution to bringing this information to the point-of-care; thus, increasing the capability of health care professionals to improve both the quality and safety of decision making because appropriate information resources are readily available (Galt, Rule, Houghton, Young, & Remington, 2005a). Personal digital assistants are also useful in improving accuracy of patient identiﬁcation and in patient-speciﬁc medical information access for use in diagnosis and care decisions (Overdyk, Haynes, & Arvanitis, 1999; Pettis et al., 1999). In addition to being an information source, the PDA is an HIT tool that can generate prescriptions. Prescriptions are easily sent to a printer in the ofﬁce or transmitted electronically to pharmacies. This functionality is a high priority for improving patient safety, especially in the ambulatory care area. A primary rationale is the high proportion of ambulatory care medication errors that could be reduced or eliminated through the use of this HIT (Galt et al., 2002; Galt, 2002; Buckley, 2002). Once electronic prescription documents are generated, in some geographic locations application service providers can transmit these documents directly to local pharmacies. When electronic transmission is not available, computergenerated typeface prescriptions are produced using the PDA’s wireless infrared capability to transmit the prescription to an ofﬁce printer. Computer-generated prescriptions reduce prescription legibility problems, omission of critical information on prescriptions, and the use of error-prone abbreviations (Galt, 2002; Buckley, 2002; Galt et al., 2005a). Currently, systems in place to protect patients from prescribing problems include double-checks through other responsible health care professionals (i.e., nurses and pharmacists) at the time of care delivery. Although PDAs do not eliminate the need for this step, the impact of immediate information access to physicians at the point-of-care and greater legibility of the prescription should reduce the interactions between physicians and these professionals related to correcting prescribing errors.

194

JAMES D. BRAMBLE ET AL.

In a current study, 226 of 553 Nebraska physicians (40.9%) reported using a PDA as part of patient care activities. Only a small percentage, however, reported generating prescriptions with their PDAs (Siracuse et al., 2008). This would seem to indicate that PDAs are seen by those that use them to be more useful as information sources or note-taking instruments than as devices to actually generate prescriptions.

PDAS: POTENTIAL IMPACT ON THE ORGANIZATION While PDAs seemingly have many potential advantages, they also may be associated with some potential problems or additional requirements. It has been noted that the process of integrating and implementing HIT into practice is complex (Hackbarth & Milgate, 2005). To help better understand this relationship we employ the use of Donabedian’s (1966) classic structure-processoutcome model. If the outcome of interest is the successful adoption of HIT, the model then suggests that physicians, group practice administrators, and others must address both structure and process issues that the HIT will affect. There are a number of potential structural factors that may be necessary for HIT to operate fully as intended. Many applications on these devices need to regularly communicate with a web server, often requiring cradlebased synchronization for exchanging data with a server (Elson, 2001). Also, generation of hard copy prescription documents requires printing devices compatible with the ofﬁce technology environment and with the PDA. Finally, ongoing maintenance of these technology systems is essential to successful incorporation into practice. In terms of process variables, the selection of a PDA and subsequent patient care applications produce changes in how the organization (or ofﬁce) functions. These changes include alterations in the routine of physician–patient encounters and the manner in which all other workﬂow occurs. Further, such changes in care routines may necessitate formal training for ofﬁce employees. Such a seemingly small technical change can have substantial impact on individuals, organizations, and systems alike. Accessing evidence to improve medical decision making at the point-of-care, and improving prescription legibility are steps critical to improving medication-related safety. Bringing evidence directly to physicians through PDAs should enhance performance. However, technologies do create demands on ‘‘operators’’. Even the introduction of low-level technology such as a PDA, is

195

Barriers to HIT

likely to encounter barriers to acceptance and use due to human factors principles that must be addressed to account for the ergonomics and utility of the human–machine interface (Cook & Woods, 1994; Van Cott, 1994). Both individual adaptation and adoption of a technology are required for a technology to be systematically incorporated into use. Adaptation requires successful learning and application by the user, as well as the physical incorporation of the technology into the ofﬁce environment itself (Weinger, Pantiskas, Wiklund, & Carstensen, 1998; Galt et al., 2005b). While information technologies may reduce some errors, they may also generate new problems. The net effect is not entirely predictable, so it is vital to study the overall impact of these technologies. The aims of this research are to identify perceived barriers to PDA use in clinical practice as perceived by physicians and to identify the successful strategies employed by these physicians to overcome these barriers. The implications of these ﬁndings are then discussed in the context of a theoretical model that was developed from quantitative data analyses of this research and explores the adoption of PDAs by primary care physicians.

METHODS Study Design The current study is part of a larger mixed methods Agency for Healthcare Research and Quality (AHRQ) funded project that employed a concurrent triangulation approach (Creswell, 2003). In this approach, we used separate quantitative and qualitative methods to explore the impact of PDAs as an HIT prescribing tool in physician ofﬁce-based practices. In this approach both quantitative and qualitative data were collected concurrently. Data collected from both of these methods are integrated for the purpose of analysis. This research strategy employs both methods to conﬁrm, crossvalidate, and corroborate the ﬁndings discussed below (Fig. 1). The quantitative approach to this study examined the impact of PDA use on potential errors through a randomized controlled trial while the qualitative approach focused on the impact of PDAs on physicians. Further details and subsequent ﬁndings of the quantitative methods are discussed in detail elsewhere (Galt et al., 2005b). For the qualitative approach, we conducted interviews for the primary purpose of identifying barriers that physicians faced when using the PDA in their practices. Additionally, the strategies employed to overcome these barriers were explored during the interview.

196

JAMES D. BRAMBLE ET AL.

QUALITATIVE

QUANTITATIVE

Data Collection

Data Collection

Data Analysis

Data Analysis

Results Compared, Integrated, and Interpreted

Fig. 1. Concurrent Triangulation Method. Note: Both Quantitative and Qualitative Data Collection and Analyses Occur Concurrently. Source: Creswell (2003). X O2

O1

O3

O4

Where, X =

Introduction of the intervention.

O1 =

Baseline criteria-based performance of PDA use assessment using direct observation by a field researcher and subject interview information documented. All intervention group physicians were required to perform all case simulations competently (accurate and complete prescriptions).

O2 =

Criteria-based subject self-assessment of PDA use and performance 4–7 days post intervention. All intervention group physicians were required to perform competently.

O3 =

Semi-structured interview of subject about barriers and solutions to performance at 2–4 weeks post introduction of the intervention.

O4 =

Direct observation and follow-up interview with subjects upon completion of intervention.

Fig. 2.

Timeline for Data Collection.

The timeline for the various parts of this study is depicted in the diagram on Fig. 2, where X represents the introduction of the intervention (i.e., the PDA) and O1 is the baseline criteria-based performance of PDA use assessment using direct observation by a ﬁeld researcher and subject

197

Barriers to HIT

interview information documented. All intervention group physicians were required to perform all case simulations competently (accurate and complete prescriptions). Three additional data collection points occurred following the introduction of the PDA and baseline performance measures. The responses used for this report come from O3.

Study Subjects The study subjects are 39 primary care physicians who agreed to participate in this ofﬁce-based ﬁeld study to determine the natural adaptation and adoption behaviors related to PDA use in practice. These subjects were recruited from the larger patient safety ﬁeld study that was conducted in 31 primary care ofﬁces (Galt et al., 2005b). The average age of the participating physicians was 42 years with a minimum age of 31 and maximum age of 79. One-fourth of the sample was female with 85% of the physicians in family medicine practices and 15% with internal medicine practices. The physicians in the sample worked in ofﬁces that ranged from one to seven practitioners with an average of 2.1 per site. Subjects were eligible for participation if they had limited or no prior use of a PDA.

Study Intervention Subjects received PDAs (Sony Clie’ PEG-SJ30, Sony Corporation) with a drug information software suite (Lexi-Comps) and prescription writing software. The prescription writing software generated prescriptions transmittable via infrared to a local ofﬁce printer. The subjects received an initial training program in their ofﬁce-based practice settings. This training program consisted of demonstrating and educating the participant about the device itself, use of the local infrared printing function, and the software applications installed on the PDA. Using a clinical case approach, each subject had to respond to drug information questions using the PDA software as the information source. Additionally, each study subject generated a printed prescription from verbal orders given by the trainer using the software application installed on the PDA. Subjects then had to successfully demonstrate their ability to use the PDA device as well as drug information and software applications. The ﬁeld researcher trainer assessed each physician’s performance using a set of objective criteria developed previously until 100% of the performance-based criteria were met.

198

JAMES D. BRAMBLE ET AL.

Data Collection We employed a multi-perspective, multi-method evaluation process (Bates, 2000) using both quantitative and qualitative methods. Key areas were identiﬁed and addressed during direct observation, self assessments, and structured interviews. Periodic assessment of participants’ abilities to use the PDA and corresponding software took place throughout the study. These assessments included a self-assessment instrument at 3–7 days post training, a structured interview 2–4 weeks after PDA introduction, and again after completion of 500 prescriptions. The purpose of the interview was to determine barriers encountered and strategies used to overcome these barriers in practice. We conducted on-site observations using the observer-as-participant approach (Morse & Field, 1995). The semi-structured individual interviews were conducted face-to-face with physicians in their ofﬁce settings, using principles of good interviewing techniques (Morse & Field, 1995; LeCompte, Schensul, Weeks, & Singer, 1999; LeCompte & Schensul, 1999). The interviews were audio recorded and converted to written transcripts for subsequent analysis.

Data Analysis We conducted a qualitative analysis of each interview transcript. The interview content consisted of responses to seven open-ended questions (see Table 1) intended to elicit subject experiences with barriers to hand-held PDA use and subject descriptions of how they overcame these barriers. Table 1. Open-Ended Questions Assessing Perceived Barriers and Solutions to PDA Use. 1. Would you tell me how you have incorporated the use of the PDA into your practice? 2. What have you tried to do with your PDA? 3. What has worked? 4. What has not worked? 5. Are there situations that you can recall that have stopped you from using the PDA? 6. When you have encountered difﬁculty with using the PDA what have you done to solve it? 7. Are there any barriers in your immediate work environment that you have encountered when trying to use your PDA?

199

Barriers to HIT

Content analysis was employed to extract categories of data from the text of the subject interviews. These interview texts were read independently by ﬁve reviewers who were experts in the prescribing process. These experts ﬁrst discussed generally the categories of information that emerged naturally from the text, in order to provide a rich and detailed understanding of the content of the interviews. Open-coding, the process of breaking down, examining, comparing, conceptualizing, and categorizing data was used. Theme analysis was conducted using the constant comparative method to develop a theme pattern from the coded data. The theme analysis resulted in organizing observations into categories of barriers and strategies used to overcome barriers (Morse & Field, 1995). Information from the informal observations at 4–6 weeks post PDA implementation was then compared against the themes that arose from the interview data to verify the results of the analysis and further explain the ﬁndings. Two methods of veriﬁcation were used: (1) veriﬁcation of the ﬁndings with the physicians who participated in the interviews and (2) veriﬁcation against existing published literature for similar ﬁndings. Human factors and technology acceptance theory were employed in this stage of analysis (Davis, Bagozzi, & Warshaw, 1989; Davis, 1986).

RESULTS Distinct themes emerged with regard to both the barriers experienced by physicians and the solutions that were employed to overcome or, at minimum, cope with those barriers. Reported barriers fell into three major categories: (1) time, (2) technology, and (3) environment. Additionally, two speciﬁc human factor issues emerged from the interview comments: (1) patient-centeredness and (2) personal views (see Tables 2 and 3). These categories are consistent with the technology-implementation systems model described by Karsh (2004). The most reported barrier to successful use of the PDA was time. Time barriers were contextual to two distinct and important areas: (1) time related to the prescription generation which was often tied to the physician’s workload and (2) time related to learning. Many physicians reported that writing the prescription on the PDA was time consuming and took longer than when using a prescription pad. This became a greater problem when they had many patients to see or multiple prescriptions for one patient. Additionally, the time it took to print the hard copy was seen as a barrier as

200

JAMES D. BRAMBLE ET AL.

Table 2. Barrier Examples Identiﬁed through Physician Interviews. Physician Identiﬁed Barriers

Related Model Constructs

Example Interview Comments

Time

Ease of use

Overwhelmed by my workload y limited the use of my PDA It was a good device for 15–18 patients per day, no good for 20–25 patients per day Not good for multiple medications

Technology

Perceived usefulness Ease of use

Missing drugs in the drug list of the prescription software and name in the systems Printing time is the biggest barrier With regard to the prescription writing software there is an inadequate drug lists y all drugs should be there

Environment

Ease of use

The printer is in a bad location I need a printer in every room I move but the printer doesn’t

Note: These types of comments were made by a half of participating users.

Table 3.

Human Factor Issues Identiﬁed through Physician Interviews.

Physician Identiﬁed Human Factor Issues

Related Model Constructs

Example Interview Comments

Patient centeredness

Attitude toward use Perceived usefulness

Patient feedback was negative Did not write prescriptions in Spanish Did not have speciﬁc drugs

Personal views

Attitude toward use

Not too excited about the PDA Not a PDA user

well. One physician reported that, due to the time delay, the patient told him he would come back the next day to pick up the prescription. Physicians also reported that they felt they needed more time to learn how to use the PDA and software, but that their schedules really did not give them the opportunity to do so. The interviews indicated that ‘‘learning on the job’’ only worked when physicians were not busy. There was a general consensus that physicians felt their workload was a signiﬁcant factor in their ability to successfully use the PDA on a consistent basis.

Barriers to HIT

201

Technology barriers included both computer hardware and software problems. Hardware problems centered on the PDA itself or the printer used to generate the electronic prescriptions. Problems with the PDA ranged from troubles using the Grafﬁtis writing system to failing to recharge the unit. Printer problems were generally limited to prescribers perceiving the printing function taking too long to produce a hard copy of the prescription. Software barriers, on the other hand, were more abundant and were primarily associated with the prescription generation application software. Physicians reported that it was both cumbersome and time consuming to enter information using Grafﬁtis. The only systematic entry process using Grafﬁtis was the entry of the patient’s name. The remainder of the information was entered into the PDA using a touch screen pull down menu. Additionally, some drugs were not in the index, which required a unique entry, and prescribing multiple drugs was more time consuming than traditional methods of prescription writing. We also found some barriers related to the working environment. Most of the clinics were not able to have a printer in the examination room – a limitation caused by size and availability of work surface for printer placement. Thus, physicians would place printers on carts in hallway locations so they were close by and available for generating hard copies of the prescriptions. This created a minor, but repetitive, systematic inefﬁciency in workﬂow which physicians perceived as a time constraint. In addition to the three barriers just discussed, interview data revealed two human factor constructs worth noting. Both of these factors are interrelated to at least one of the three barriers identiﬁed. These factors include patientcenteredness and personal views. Patient-centeredness focused on speciﬁc clinical concerns and how the technology addressed those concerns. For example, such concerns may include physician perceptions of how patients felt about their physician using this technology or even explicit patient reaction to the use of the HIT. The second factor, personal views, focused on the physician’s attitude toward HITs generally, and, speciﬁcally toward the PDA in the current study. Interviews also revealed that some subjects were ‘‘techno-terriﬁed’’ individuals who just did not want to use the new technology. These physicians reported they were not users, had no prior experience with technology, and had no interest in the technology. Solutions to these barriers reported by physicians included solutions the individual could employ and systematic solutions that went beyond the individual (see Table 3). Individual solutions included self-helping behaviors such as self-development, seeking necessary assistance, or changing their individual workﬂow.

202

JAMES D. BRAMBLE ET AL.

For example, one physician’s solution to the limited number of printing devices was to invite the patient into his ofﬁce to ﬁnish their discussion while the prescription was printed. In addition to self-helping solutions, some physicians chose to only partially adapt to the PDA technology. These physicians would choose to use only the technology application that beneﬁted them the most or caused the least amount of disruption in their workﬂow. This selective use of applications varied on a day-by-day basis. Often this included just using the drug information software. A ﬁnal individual solution was avoidance. Physicians who felt the technology was impeding their patient care and did not have the time to solve the problem would often report they gave up on the PDA and returned to the traditional method of generating prescriptions. The second major category of solutions included system solutions. These solutions went beyond the individual user and required the system (e.g., the clinic) to incorporate needed changes to overcome the barrier. Solutions that fell into this category included structural changes and changes in the system’s workﬂow. For example, many physicians reported they needed more printers – an investment that must be made by the organization. Additionally, they felt they needed more time to learn and some continuing education on use of the PDA so they could maximize its efﬁciency and beneﬁts. Many physicians felt they did not have the time to fully learn the system, especially as their clinic became busy. Their overall views expressed the need for the physician’s organizations to help them get ‘‘up to speed’’ on using their PDA.

DISCUSSION Quantitative results of this study showed an overwhelmingly positive pre-topost PDA intervention improvement in the reduction of potential prescription errors generated (Galt et al., 2005b). However, the data also revealed that adoption of this particular HIT was less than 100%. The results just presented provide some indication as to why. To aid in our discussion of these results we revisit a model developed by Galt, Siracuse, Rule, Clark, and Taylor (2005c) based on the quantitative work of this study design. As shown in Fig. 3, three of the constructs inﬂuencing actual use of HIT include perceived usefulness, perceived ease of use, and attitude toward use. As just reported, the results of the qualitative part of the study offer greater insight and more in-depth explanations for these constructs. Perceived ease of use is explained in the model as having no human factor limitations in the incorporation of the HIT. The model indicates that the

203

Barriers to HIT

Perceived Usefulness (accurate Rxs)

Perceived Usefulness (increased efficacy)

Attitude toward use

Attitude toward use (emotional)

(learning related)

Perceived Usefulness

Fig. 3.

Intent to Use

Use

External Factors

Theoretical Model of Physician Adoption and Use of a PDA HIT.

fewer the limitations in incorporating PDA technology the more likely it is that physicians will use the PDA as a prescribing tool. Data from physician interviews show that perceived ease of use is explained by their self-reported technological and environmental barriers. For example, physicians reported printers were too slow in generating a hard copy prescription. Additionally, the location of the printer was not convenient. Such self-reported barriers provide good examples of human factor limitations that reduce the level of adoption. Perceived usefulness is the belief that PDAs will increase efﬁciency. In the physician interviews, it became clear that physicians wanted a tool that would help them become more efﬁcient, or, at minimum, maintain their current efﬁciencies. When physicians report that the electronic prescribing software did not have a complete list of drugs for their patient base, thus reducing its usefulness for their practice, the likelihood of full adoption lessens. We grouped these comments under technology barriers and factors associated with patient-centeredness. Thus, a major aspect of perceived usefulness is related to the degree the HIT ‘‘ﬁts’’ the physician’s practice. Attitude toward use is the reaction of the user toward the HIT. This reaction is sometimes mitigated by the reactions or opinions of others. Thus, we categorized these comments not only as personal views barriers, but also

204

Table 4.

JAMES D. BRAMBLE ET AL.

Solution Examples Identiﬁed through Physician Interviews.

Individual solutions Practice and use outside the ofﬁce is needed to improve my efﬁciency Extra effort to learn outside the job is needed Drug information only I am too busy to seek a solution y I gave up and went to using old ways System solutions Ongoing training needed I want another printer More time is needed prior to PDA use

considered these issues as related to patient-centeredness. Both provide examples that help explain the ‘‘attitude toward use’’ construct. Some physicians report negative feedback from their patients regarding their use of technology while other physicians lack enthusiasm toward using the PDA. Both of these instances are negatively related to actual use. The overarching issue that tied all barriers together was the relationship between the encountered barrier and the physician’s ability to stay on schedule in seeing their scheduled patients. Overall, both barrier and solution comments related to physicians being patient-centered in their adoption of the PDA as a prescribing tool. That is, when the PDA inhibited patient care and seeing the requisite scheduled patients for the day, physicians changed behavior in a manner consistent with the continuance of acceptable patient care such that the solutions employed allowed them to maintain seeing their patients on schedule. Both individual and system-level solutions were generated by the physicians (Table 4). Individual solutions often had the characteristics of immediate implementation, allowing the user to ‘‘move on’’ with their work while incorporating the technology as much, or as little, as possible. These solutions were characterized by individuals taking responsibility for ‘‘owning’’ the system and ﬁnding the necessary time to learn. The solutions are matched to the effort expended to successfully incorporate the new technology into daily work. Alternatively, system solutions are those solutions identiﬁed as needed, but beyond the individual’s ability to implement. These solutions require too much time or interfere with the work individual physicians need to do (e.g., care for patients). Solutions in this category need more resources and investment than available at the individual level. These solutions are

Barriers to HIT

205

characterized by the individual’s view that accomplishment of the needed solution is outside of the individual’s responsibility or ability. In these cases, often the immediate, individual solution was to give up indeﬁnitely until the system ‘‘ﬁxes’’ the problem. Returning to Donabedian’s (1966) structure-process-outcome model helps us further understand the various levels at which barriers are encountered, and what solutions might best be implemented to overcome those barriers. The model suggests that physicians and physician clinics must address both structural variables, such as adequate number of printers appropriately located, and process variables, such as the ability to write on the PDA and generate prescriptions in a timely manner within an efﬁcient workﬂow, for successful adoption to occur. Most structural barriers we identiﬁed were associated with the technology matching the needs of the practice environment. Process barriers most often involved how individuals use the device and, more speciﬁcally, how the environment was prepared or able to accommodate the changes in physical workﬂow induced by the new technology. To maximize the use of HIT in the prescribing process and thus take full advantage of its ability to reduce errors and improve quality in medication prescribing, organizations must examine and analyze their infrastructure. This examination and analysis should be used to determine whether the infrastructure is designed to support the integration of HIT, and to ﬁnd solutions where support of these practices and systems is lacking. This research demonstrates that though some individuals ﬁnd ways to work ‘‘through’’ their problems (including giving up), the whole micro-system must be evaluated in order to successfully adopt and incorporate HIT that has a positive inﬂuence on patient safety and quality of care. While the importance of a proper infrastructure is a key component in the successful adoption of HIT, we also recognize that systems consist of not just the infrastructure but the interplay between individuals and the environment around them. Each component of the system (both individuals and the infrastructure) either directly or indirectly affects the other. Thus, successful adoption will also need to examine and potentially address the human side of change management. Adoption is likely to take place if users ﬁnd value in this newer way, and solutions to overcome barriers are integrated in practice. Improved safety through both access to needed scientiﬁc information and generation of typeface prescriptions will only take place if there is successful technology adoption by all users.

206

JAMES D. BRAMBLE ET AL.

IMPLICATIONS FOR HEALTH CARE MANAGEMENT This research helps administrators understand the constructs related to adoption of HIT and the barriers that physicians may encounter. A better understanding of these issues will help health care administrators develop successful strategies for adoption of HIT across their health care organization. Health care managers who take the time and devote the resources necessary to address the three major barriers prior to HIT implementation will likely proﬁt from greater adoption and use. For example, using a technology that matches the needs of patients with the needs of the physicians should reduce or eliminate not only the barriers to adoption of that technology, but also address the human factor issues of patientcenteredness and personal attitudes described in this study. In addition to ensuring that the technology meets the need, the health care manager must set up a patient care environment that can accept the HIT while simultaneously allowing health care workers to take full advantage of those functions for which the HIT was designed. By reducing these barriers, administrators will also indirectly address many of the time barriers associated with the HIT. Recall that time barriers existed on two levels: (1) time related to the prescription generation and (2) time related to the learning how to use the HIT. A better environmental ﬁt and technology match that reduces or eliminates the prescription generation time barrier will further a more seamless adoption of HIT and thus allow physicians to maintain their current patient workloads. However, we note that time to train and learn HIT still represents a major hurdle that must be addressed. In addition to addressing barriers related to the organization’s supporting infrastructure, addressing individual concerns and obtaining physician support and buy-in to the adoption of HIT are also important. Successful strategies may include identiﬁcation of effective physician leaders who can champion the adoption process, development of training modules that are tailored to speciﬁc learning styles and circumstances, and providing constructive and useful feedback that encourages continued use. Another strategy is to position the adoption of HIT as a proactive process designed to prevent errors and promote safer practices. The key is that physicians view HIT tools as patient-focused and intended to increase patient safety. Repeated educational interventions, combined with listening, understanding and adjusting, when necessary, may be needed for complete adoption to take place. The key is to move forward with the best possible tools for

207

Barriers to HIT

improving safety in the patient care system until the use of HIT becomes a standard part of patient care. Training also needs to address the many other stakeholders that operate within most medical ofﬁces such as nursing staff and ofﬁce managers. Since all must work together to provide appropriate patient care, it is reasonable to assume that if any of the stakeholders do not adopt the HIT, physicians may not choose to use the HIT because they perceive this intraofﬁce incongruence as yet another barrier. As health care managers focus their efforts on changing the physical barriers identiﬁed by this research they will be able to inﬂuence the human factor constructs previously identiﬁed. When time, technology, and environmental barriers are minimized, we postulate that individual views on patient-centeredness and/or attitudes toward use should become more positive and thus enhance HIT adoption. Future research about barriers and solutions should continue to include the perspectives of individual users.

ACKNOWLEDGMENT This study was supported in part by Grant Number R18HS11808-1, Agency for Healthcare Research and Quality.

REFERENCES Arnone, G., Bianchi, A., ella-Pietra, B., Sernicola, R., Sparacino, E., & Vitolo, R. (1998). Easy medic: An internet application for the general practitioner. Journal of Telemedicine and Telecare, 4(Suppl. 1), 93–94. Bates, D. W. (2000). Using information technology to reduce rates of medication errors in hospitals. British Medical Journal, 320, 788–790. Brodell, R. T., Helms, S. E., KrishnaRao, I., & Bredle, D. L. (1997). Prescription errors: Legibility and drug name confusion. Archives of Family Medicine, 6, 296–298. Buckley, M. (2002). Improving drug prescribing practices in the outpatient setting: A market analysis. California HealthCare Foundation (Report downloaded from http://www.chcf. org/documents/healthit/ImprovingDrugPrescribing.pdf). Cabral, J. D. (1997). Poor physician penmanship. JAMA, 278, 116–117. Cohen, M. R. (1999). Medication errors. Washington, DC: American Pharmaceutical Association. Cook, R., & Woods, D. (1994). Operating at the sharp end: The complexity of human errors. In: M. Bogner (Ed.), Human error in medicine. Hillsdale, NJ: Lawrence Erlbaum Associates. Creswell, J. W. (2003). Research design: Qualitative, quantitative, and mixed methods approaches (2nd ed). Thousand Oaks, CA: Sage Publications.

208

JAMES D. BRAMBLE ET AL.

Davis, F. D. (1986). A technology acceptance model for empirically testing new end-user information systems: Theory and results. Doctoral Dissertation, Sloan School of Management, Boston, MA: Massachusetts Institute of Technology. Davis, F. D., Bagozzi, R. P., & Warshaw, P. R. (1989). User acceptance of computer technology: A comparison of two theoretical models. Management Science, 35(8), 982–1003. Donabedian, A. (1966). Evaluating the quality of medical care. Milbank Memorial Fund Quarterly, 44(pt. 2), 166–206. Elson, B. (2001). Electronic prescribing in ambulatory care: A market primer and implications for managed care pharmacy. Journal of Managed Care Pharmacy, 7(2), 115–120. Galt, K. A. (2002). Medication errors in ambulatory care. Topics in Health Information Management, 23(2), 34–46. Galt, K. A., Rich, E. C., Young, W. W., Markert, R. J., Barr, C., Houghton, B., Taylor, W., Rule, A. M., & Bramble, J. D. (2002). Impact of hand-held technologies on medication errors in primary care. Topics in Health Information Management, 23(2), 71–81. Galt, K. A., Rule, A. M., Houghton, B., Young, D. O., & Remington, G. (2005a). Personal digital assistant-based drug information sources: Potential to improve medication safety. Journal of the Medical Library Association, 93(2), 229–236. Galt, K. A., Rule, A. M., Taylor, W., Siracuse, M. V., Bramble, J. D., Rich, E. C., Young, W. W., Clark, B. E., & Houghton, B. (2005b). Impact of personal digital assistant devices on medication safety in primary care. In: Advances in patient safety: From research to implementation (Vol. 1–1V). Rockville, MD: Agency for Healthcare Research and Quality. Galt, K. A., Siracuse, M. V., Rule, A. M., Clark, B. E., & Taylor, W. (2005c). Physician use of hand-held computers for drug information and prescribing. In: Advances in patient safety: From research to implementation (Vol. 4, pp. 93–108). Rockville, MD: Agency for Healthcare Research and Quality. Hackbarth, G., & Milgate, K. (2005). Using incentives to drive physician adoption of health information technology. Health Affairs, 24(5), 1147–1149. Institute of Medicine. (2001). Crossing the quality chasm: A new health system for the 21st century. Washington, DC: National Academy Press. Karsh, B. (2004). Beyond usability for patient safety: Designing effective technology implementation systems. British Medical Journal: Quality and Safety in Health Care, 13(5), 388–394. Kohn, L. T., Corrigan, J. M., & Donaldson, M. S. (2000). To err is human: Building a safer health system. A report of the Committee on Quality of Health Care in America, Institute of Medicine. Washington, DC: National Academy Press. LeCompte, M. D., & Schensul, J. J. (1999). Analyzing and interpreting ethnographic data (Vol. 5). Walnut Creek, CA: Altamira Press. LeCompte, M. D., Schensul, J. J., Weeks, M. R., & Singer, M. (1999). Researcher roles and research partnerships (Vol. 6). Walnut Creek, CA: Altamira Press. Lesar, T. S., Briceland, L., & Stein, D. S. (1997). Factors related to errors in medication prescribing. Journal of the American Medical Association, 277, 312–317. Morse, J. M., & Field, P. A. (1995). Qualitative research methods for health professionals, 2nd ed. Thousand Oaks, CA: SAGE Publications. National Association of Chain Drug Stores Foundation. (2005). Annual report 2005: Research, education, charity. Alexandria, VA: NACDS Foundation.

Barriers to HIT

209

Overdyk, F. J., Haynes, G. R., & Arvanitis, P. J. (1999). Patient-borne memory device facilitates ‘‘point of care’’ data access. MD Computing, 16(3), 60–63. Pettis, K. S., Savona, M. R., Leibrandt, P. N., Maynard, C., Lawson, W. T., Gates, K. B., & Wagner, G. S. (1999). Evaluation of the efﬁcacy of hand held computer screens for cardiologists’ interpretations of 12-lead electrocardiograms. American Heart Journal, 138(4, Pt. 1), 765–770. Siracuse, M. V., Galt, K. A., Bramble, J. D., Drincic, A., Paschal, K. A., Abbott, A. A, Fuji, K. T., & Kasha, T. A. (2008). Adoption of electronic health records and ePrescribing in Nebraska and South Dakota. Journal of the American Pharmacists Association, 48(2), p. 271. Van Cott, H. (1994). Human errors: Their causes and reductions. In: M. Bogner (Ed.), Human error in medicine. Hillsdale, NJ: Lawrence Erlbaum Associates. Vitillo, J. A., & Lesar, T. S. (1991). Preventing medication prescribing errors. Annals of Pharmacotherapy, 25, 1388. Weinger, M. B., Pantiskas, C., Wiklund, M., & Carstensen, P. (1998). Incorporating human factors into the design of medical devices. Journal of the American Medical Association, 280(17), 1484.

IMPROVING PATIENT SAFETY THROUGH ORGANIZATIONAL DEVELOPMENT: CONSIDERING THE OPPORTUNITIES Ann Scheck McAlearney ABSTRACT Increased attention to improve patient safety in healthcare has challenged healthcare managers to consider innovative approaches to meet this need. Organizational development (OD) programs have been used in both health services and other industries to address organizational training and development requirements, and can provide focused, timely, and effective education and training to a broad spectrum of program participants. In healthcare organizations, OD programs can serve an important institutional function by providing a framework through which patient safety can be emphasized as an organizational priority, and patient safety training can be delivered as part of OD efforts. In addition, organizations committed to creating a patient-focused safety culture can use OD initiatives strategically to support organizational culture change efforts. This chapter describes different approaches to including patient safety in an OD framework, drawing from both management theory and practice. Findings from three extensive qualitative studies of leadership development and corporate universities in healthcare provide speciﬁc examples of Patient Safety and Health Care Management Advances in Health Care Management, Volume 7, 213–239 Copyright r 2008 by Emerald Group Publishing Limited All rights of reproduction in any form reserved ISSN: 1474-8231/doi:10.1016/S1474-8231(08)07010-9

213

214

ANN SCHECK MCALEARNEY

how healthcare organizations discuss patient safety improvement using this alternative approach. Considering the concepts and ﬁndings described in this chapter can help healthcare organizations make strides toward positive changes in organizational culture that will promote patient safety on the organizational agenda.

BACKGROUND As the healthcare industry is faced with a critical need to improve patient safety and reduce medical errors (e.g., Kohn, Corrigan, & Donaldson, 1999), healthcare managers are challenged to respond (e.g., Joint Commission on Accreditation of Healthcare Organizations (JCAHO), 2007a). Issues such as the safety of medication administration (Kohn et al., 1999; Taxis & Barber, 2003; O’Hare, Gallagher, & Shields, 1995; Cousins, Sabatier, Begue, Schmitt, & Hoppe-Tichy, 2005; Husch et al., 2005; Kaushal et al., 2001; Leape & Bates, 1995; Leape et al., 1991; Bobb et al., 2004), the potential of health information technologies to reduce medical errors (Bates et al., 1998, 1999; McAlearney, Chisolm, Schweikhart, Medow, & Kelleher, 2007), and the potential to reduce hospital infection rates through better hand washing techniques (Gawande, 2004, JCAHO, 2007b) are three examples of patient safety challenges highlighted in the recent literature. Yet patient safety is only one of a myriad of competing organizational priorities healthcare leaders are charged to address in their efforts to best serve patients, communities, and other stakeholders. Persistent and rapid change within the nearly $2 trillion industry (Catlin, Cowan, Hefﬂer, & Washington, 2007) bombards healthcare managers with daily demands such as needing to contain costs, address inequities and disparities in care, identify and avoid management mistakes, and retain and develop good employees, among other concerns (e.g., Kerr, McGlynn, Adams, Keesey, & Asch, 2004; McGlynn et al., 2003; Smedley, Institute of Medicine (IOM), Stith & Nelson, 2002; Fulmer & Goldsmith, 2001; Kouzes & Posner, 1995; Ramanujam & Rousseau, 2004; Russell & Greenspan, 2005; Jones, 2005; Hofmann, 2005; Hofmann & Perry, 2005). While a clear imperative exists to focus on patient safety improvement, especially in the context of improving the quality of care, it is less clear how to execute this agenda in practice. There is tremendous potential to improve patient safety in healthcare by considering opportunities in organizational development (OD). In this chapter I discuss how an OD framework can be used to focus efforts to improve patient safety, serving as both a conceptual umbrella and structure

Improving Patient Safety

215

through which patient safety education and training can be delivered. First, I brieﬂy describe the conceptual issues around this topic, focusing on the areas of OD theory and considering how a development approach could be applied to patient safety improvement. Second, I describe the research methods associated with three extensive qualitative studies of leadership development and corporate universities in healthcare that have framed the concepts and ideas presented in this chapter. Third, I discuss my research ﬁndings as answers to questions about how healthcare organizations can use an OD framework to improve patient safety. Finally, I present a conceptual model of how an organizational commitment to development can impact patient safety. This research contributes to the literature in various ways. First, scant research in OD has been focused on the healthcare industry, and the opportunity to focus OD efforts on a particular organizational priority such as patient safety has not been addressed. This chapter presents both a rationale for proceeding along those lines, and relevant examples of how focused OD efforts can be effective. Second, given the glaring needs for OD and employee training in the healthcare industry beyond what might be provided in formal educational institutions, the paucity of research in this area is striking. This research helps to inform our knowledge of education and development opportunities that can be provided in practice. Third, these studies attempt to investigate the issues surrounding OD and leadership development from both organizational and individual perspectives. By asking questions about organizations’ practices and individuals’ experiences, these studies move beyond single points of view to consider both the context and practices of leadership development and OD. Finally, this chapter contributes to our understanding of how OD programs can help healthcare organizations better support their commitments to and investments in their employees, recognizing that options for accommodating focused OD strategies can be incorporated in virtually all healthcare organizations. Findings from this chapter should be of use to a variety of audiences including academics, consultants, and OD practitioners interested in assisting healthcare organizations with their patient safety improvement efforts.

CONCEPTUAL FRAMEWORK Conceptualization of the opportunities to consider patient safety improvements in the context of OD requires brief examination of answers to three

216

ANN SCHECK MCALEARNEY

key questions: (1) What are organizational and leadership development? (2) What do we know about these approaches in healthcare? and (3) How might general OD approaches be adapted to promote patient safety? In the following section, each of these areas will be discussed in greater detail.

Organizational and Leadership Development OD is rooted in behavioral sciences and is described as focusing on interventions designed for organization improvement (French & Bell, 1978). This concept is formally explained as follows: y organization development is a long-range effort to improve an organization’s problem-solving and renewal processes, particularly through a more effective and collaborative management of organizational culture – with special emphasis on the culture of formal work teams – with the assistance of a change agent, or catalyst, and the use of the theory and technology of applied behavioral science, including action research. (French & Bell, 1978, p. 14)

Comprised of both concepts and techniques that are applied to improve organizational effectiveness and the well-being of individuals working in organizations (French, Bell, & Zawacki, 1989), OD is considered as differentiated from other educational interventions because of this focus on the functioning of organizations (French & Bell, 1978). Within OD, leadership development practices are described as educational processes that are designed to improve individuals’ leadership capabilities. Leadership development programs often draw from traditions in management training, and can focus on both improvement of managerial skills and individual job performance (Burke & Day, 1986; Day, Zaccaro, & Halpin, 2004; Giber, Carter, & Goldsmith, 2000; Tichy, 1999); in addition, they can have measurable effects on both organizational climate (Moxnes & Eilertsen, 1991) and organizational culture (Schein, 1985). Often synonymous with the concept of management development, leadership development practices include interventions that focus on improving organizational culture or enhancing leaders’ effectiveness by facilitating learning (Gray & Snell, 1985). Examples of such interventions include practices such as 360-degree feedback, skill-based training, focused job assignments, developmental relationships including mentoring and coaching, and action learning (McCall, Lombardo, & Morrison, 1998; McCauley, Moxley, & VanVelson, 1998; Revans, 1980).

Improving Patient Safety

217

Organizational and Leadership Development in Healthcare Until recently, organizational and leadership development have received relatively little attention in U.S. healthcare organizations. Generally, healthcare organizations are considered to lag behind other industries with respect to leadership development (McAlearney, 2006), despite concern about an impending leadership crisis (Mecklenburg, 2001; Institute for the Future, 2000; Schneller, 1997). Further, while leadership development is of current interest among healthcare organizations and associations (e.g., efforts by the National Center for Healthcare Leadership and the American College of Healthcare Executives), the more general framework of OD has received very little attention in both research and industry practice. Healthcare organizations spend substantial amounts on OD through employee education, training, and leadership development efforts with scant understanding of whether they receive any returns on their investments. In addition, most of these organizations miss valuable opportunities to develop employees according to their strategic priorities. Thus, the opportunity to deliver targeted OD programs related to patient safety improvement offers an attractive option by which healthcare organizations can strategically deploy both ﬁnancial and human resources.

Adapting Organizational and Leadership Development Approaches to Patient Safety Given the formal deﬁnitions and descriptions of OD, such approaches would provide an appropriate framework for patient safety improvement efforts in four ways. First, the focus of OD on ‘‘organization improvement’’ is well aligned with patient safety improvement goals. Second, the notion of organizational development as a long-range effort focused on both problemsolving and renewal processes with an emphasis on organizational culture change also translates into patient safety improvement objectives, notably those that highlight the importance of developing a ‘‘patient safety culture.’’ Third, the emphasis in OD on formal work teams and their culture is reﬂected in many patient safety improvement initiatives using wellestablished frameworks such as Six Sigma (Harry & Schroeder, 2000). Finally, the importance of behavioral interventions and action research is consistent with employee and leadership development efforts which similarly emphasize behavior change and promote active and action learning methods in attempts to improve patient safety.

218

ANN SCHECK MCALEARNEY

As an example of how an organization development framework can be used to guide patient safety improvement efforts, leadership development programs are particularly appropriate. In this literature, four general approaches to leadership development have been outlined: (1) development of the individual leader; (2) socializing organizational vision and values; (3) focusing on strategic leadership initiatives; and (4) promotion of action learning (Conger & Benjamin, 1999). In Table 1 I list the features of these four approaches and then provide examples describing how each can be applied to patient safety improvement in healthcare. For example, the organization selecting an approach that focuses on individual leader development would work carefully to customize program activities and on-the-job experiences around patient safety to maximize learning and development of that particular individual. It is likely that developmental relationships with formal mentors or supervisors would be leveraged to tailor the particular patient safety improvement training and messages that are appropriate for that individual. In contrast, for the organization concerned about socializing company vision and values in its leadership development approach, patient safety training might become part of organization-wide efforts to emphasize patient safety and patientcentered care as part of the values of the organization. The focus on patient safety improvement might be emphasized in metrics measuring organizational performance (e.g., a balanced scorecard format) in order to ensure that organization members were socialized to understand the priority and primacy of patient safety. Speciﬁc application of these four leadership development approaches to patient safety improvement efforts in a given organization would necessarily depend upon the particular organizational goals and the audience being targeted for development in that organization. In the preceding section I explored three areas important in the conceptualization of patient safety improvement as an organization development opportunity. Next I will describe my research studies of leadership development in healthcare organizations to provide a context for uncovering opportunities to improve patient safety using an organizational development framework.

METHODS Summary of Research Over the past four years I have conducted 200 interviews with participants in three major qualitative studies. Characteristics of participants across the

219

Improving Patient Safety

Table 1. Approach

A Patient Safety Focus for Four General Leadership Development Approaches. Features of Effective Programs (Conger & Benjamin, 1999)

Adaptation to Patient Safety Framework

Developing the individual leader

Build around a well-deﬁned model Use clear criteria for participant selection Include pre-course preparation Use personalized feedback Apply multiple learning methods Conduct extended learning periods, multiple sessions Put organizational systems in place

Deﬁne program goals and program model around patient safety agenda Tailor programs to participants as appropriate (e.g., physicians versus nurses versus administrators) Provide background information about patient safety goals, medical errors, quality improvement strategies Link program activities to on-thejob experiences and challenges in patient safety Provide individualized feedback about adverse events, near misses, and so forth, where appropriate

Socializing company vision and values

Carefully select program participants Use organizational needs assessment to tailor program content, design Apply a well-articulated vision and philosophy Use practicing leaders to provide instruction Develop theories into explicit models, desired attributes, behaviors Design program around participant exchange, structured interaction Utilize a systems approach to emphasize integration, support Consider evaluation and accountability for results

Determine if program will be organization-wide or targeted to a subset within the organization Perform organizational audit to understand organizational culture and climate around patient safety Provide a consistent message about patient safety improvement, grounded in organizational values Consider opportunities to discuss speciﬁc adverse events or near misses for group learning opportunities Build culture of patient safety and transparency of error reporting Develop evaluation metrics to track activities and improvements

Strategic leadership initiatives

Strategic framework drives program content Organization’s needs are assessed up front Programs structured around speciﬁc learning objectives, for executive cohorts

Clarify organization’s strategic priorities and explicit goals for patient safety improvements Use qualitative assessment and quantitative measurement to establish baseline from which to improve patient safety

220

ANN SCHECK MCALEARNEY

Table 1. (Continued ) Approach

Action learning

Features of Effective Programs (Conger & Benjamin, 1999)

Adaptation to Patient Safety Framework

Curricula designed to permit collective dialogue between units, across levels Trained facilitators help provide critical process assistance Learning experiences cascade across levels Active feedback mechanisms

Deﬁne speciﬁc patient safety improvement objectives for individual program participants and executive cohorts Ensure curricula is designed to include perspectives and issues across the organization Provide ongoing feedback through program monitoring and evaluation

Learning projects are carefully selected Objectives and outcomes are clearly deﬁned Multiple opportunities provided for reﬂective learning Senior management is actively involved Facilitation and coaching provided by experts

Select action learning projects with high-priority patient safety improvement goals Deﬁne project objectives and outcomes with respect to measurable improvements in patient safety, reduction in adverse events, improvement in organizational culture, and so forth Ensure leadership commitment and engagement by providing forum for regular communication and feedback about project progress and results

Source: Adapted from Conger & Benjamin (1999).

three studies are shown in Table 2. The ﬁrst, ‘‘Leadership Development in Health Care: Practical Strategies for Health Care Organizations,’’ was designed to investigate the content and process of healthcare leadership development programs nationwide. In this study, a total of 160 key informant interviews were conducted in 2003–2004. Informants included recognized experts in healthcare leadership, purposely selected through a snowball sampling approach, and focused study of healthcare provider organizations and vendors of healthcare leadership development programs with reputable leadership development programs. A second study, ‘‘Leadership Development and Succession Management in Healthcare Organizations,’’ was initiated in 2004 to extend the ﬁrst study and explore emerging topics such as succession planning and strategic

221

Improving Patient Safety

Table 2. Study Participants across Three Studies. Description Experts interviewed

Organizational informants interviewed

Number (%)

Association leaders University faculty Industry consultants

17 (42%) 14 (34%) 10 (24%)

Total

41

Executive-level informant Director-level informant Manager-level informant Program participant

49 58 35 17

Total

159

Total key informants

200

(31%) (36%) (22%) (11%)

leadership program planning in greater depth. As part of this research, 10 key informant interviews were held between 2004 and 2006 with leaders of healthcare provider organizations concerned about the topics of succession management and leadership development. The third study, ‘‘Corporate Universities in Healthcare Organizations: Exploring the Evidence,’’ was conceived as an exploratory investigation of the use and reach of corporate universities, or focused leadership training and development programs, in healthcare provider organizations. Begun in 2006, 30 key informant interviews have been held, to date, focused on the role and opportunities for such initiatives. Research Design For all interviews, I used standard, semi-structured interview guides (three versions matching the three individual studies), including open-ended questions to guide the interviews, and probing questions to obtain additional information (Miles & Huberman, 1994; McCracken, 1988). Basic interview domains including topics such as program history and rationale, program development barriers and facilitators, and future opportunities and plans were similar across the three studies, with later interviews including newer study-focused questions about succession planning and the speciﬁc concepts associated with corporate universities. Each of the original interview guides was pilot tested with leaders in local healthcare provider organizations before use with the target key informants. Interviews lasted from 15–90 min, with an average duration of 45 min.

222

ANN SCHECK MCALEARNEY

A qualitative research design (Maxwell, 1996) was consistent with the exploratory objectives of each of these studies, and enabled me to investigate different issues as they emerged around the general research topics. This qualitative design was appropriate because of the paucity of current literature on these different topics which left wide gaps in general understanding, and because I suspected that the various key informants’ perspectives about leadership development were multidimensional (Miles & Huberman, 1994). Further, through both in-person and telephone interviews with a variety of key informants across the country, I was able to obtain rich information about the different facets of leadership development in healthcare organizations, including multiple issues which would have been extremely difﬁcult to examine using quantitative methods (Crabtree & Miller, 1999; Miles & Huberman, 1994). Throughout all three studies, no potential informant contacted refused to participate in the study. Human subjects approval was obtained through the Institutional Review Board of The Ohio State University, and all participants were assured that their voluntary participation would remain anonymous. Analyses A predominant majority of the interviews were audiotaped and transcribed, and extensive ﬁeld notes were collected in the several cases where taping was infeasible. I used the constant comparative method of qualitative data analysis (Glaser & Strauss, 1967) and common approaches to code the data (Constas, 1992; Miles & Huberman, 1994) from my interviews. Through all three studies, I have used a grounded theory approach (Glaser & Strauss, 1967; Strauss & Corbin, 1998), taking advantage of opportunities to discuss ﬁndings with research associates, professional colleagues, and new key informants as each study progressed. This iterative process has permitted exploration of new themes and concepts in subsequent interviews and studies. I used the qualitative data analysis software Atlas.ti (version 4.2) (Scientiﬁc Software Development, 1998) to support much of these analyses. Research Question Guiding this Chapter For the development of the concepts and ideas reported in this chapter, I analyzed the data from these studies with the aim of ﬁnding opportunities to improve patient safety. My general research question was framed, ‘‘How

223

Improving Patient Safety

can organizational and leadership development efforts be used to improve patient safety?’’ To explore this issue in my data, I reviewed interview transcripts from the ﬁrst two studies using a search function for the keywords ‘‘safety’’ and ‘‘culture.’’ In addition, I was able to structure interviews from my ongoing studies to include speciﬁc questions about patient safety and culture in the context of leadership development and corporate universities. Additional conversations with professional colleagues and an ongoing review of the literature have helped me to conceptualize, validate, compare, and extend my ﬁndings, where appropriate (Glaser & Strauss, 1967).

RESULTS Improving Patient Safety through Organizational Development Considering the ﬁndings from these studies in the context of opportunities to improve patient safety, I focus this discussion of results around three speciﬁc, organizing questions: (1) What does an integrated OD program include to emphasize patient safety improvement? (2) What program characteristics are critical to successful program development? and (3) How can organizations move to develop a patient safety culture? Answers to these questions provide information that can help organizations increase their likelihood of success at efforts to improve patient safety. Structuring Organizational Development to Emphasize Patient Safety An effective OD program has broad organizational reach, spanning the organization and touching both employees and afﬁliated professionals. From human resources practices to employee training, an OD umbrella can ensure emphasis on organizational improvement, and align focus with strategic organizational priorities. In the context of patient safety improvement, this OD framework is particularly appropriate. Speciﬁcally, an OD framework can tie together six common areas of organizational concern to emphasize patient safety improvement: (1) Employee Development Programs; (2) Strategic Priority Planning; (3) Human Resources Programs; (4) Employee Education and Training; (5) Clinician Education and Training; and (6) Patient Education. Thus, in practice, the OD framework can help to organize patient safety education, whether emphasized and delivered through leadership development programs that are part of employee development, information

224

ANN SCHECK MCALEARNEY

technology training through employee training and education, or continuing education (CE) programs about safety measurement for physicians. Even patient education programs can include a safety and improvement focus, such as by providing patient education about communicating with physicians, or by offering classes to improve health literacy in an attempt to reduce gaps in patient understanding. Healthcare institutions that have developed and employed OD programs report that a focus on strategic priorities is facilitated by this integrated framework. For instance, in one organization they had developed an ‘‘Academy’’ that provided an umbrella structure for different development programs. This organized structure for development thus facilitated delivery of all education, training, and development activities throughout the organization by utilizing common communications and delivery methods. While a minority of healthcare organizations assert that their OD activities are sophisticated enough to be considered integrated, most organizations have investments in employee development, training, and education which can be leveraged to support initiatives in patient safety improvement. For instance, many hospitals and health systems have some form of organization-wide meetings, and the focus of these meetings can be shifted to emphasize patient safety. One organization that had done this reported how ‘‘the last one was about safety. There was a very powerful speaker that spoke. It was a former patient who had several powerful messages that everyone found very moving,’’ and the explicit examples provided helped meeting participants focus their attention on targeted steps to improve patient safety in order to eliminate problems such as those described by the patient speaker. In another organization an informant explained, ‘‘We have three, 2-hour meetings each year, run by the CEO, and these are educational for our employees. They are updates and focus education on patient safety, ﬁnance, 360-degree leaders, etc. The purpose is for managers to learn and have education and leadership from the leadership team.’’ These meetings can thus be tailored to organizational priorities, with a higher priority for patient safety receiving correspondingly more attention organization-wide. Even in organizations without centralized opportunities for employee development and education, opportunities exist to increase emphasis on patient safety improvement. Examples of these types of opportunities that can be taken at either the department level or organization-wide are shown in Table 3. In practice, the key to integrating these efforts toward patient safety improvement is consistent communication, messages, and focus. Considering an OD framework can thus help organizations create consistency and redundancy in their messages, as well as in their investment decisions to

Improving Patient Safety

Table 3.

225

Immediate Opportunities to Increase Focus on Patient Safety.

Department-level opportunities Book/article club (e.g., assign IOM report and discuss) Encourage investigation into issues surrounding safety and errors (e.g., reporting, culture, sharing information, communication) Hold monthly brainstorming sessions How to reduce medical errors in a particular clinical area Ideas from current literature Crazy opportunities and innovations Brown bag lunches with internal or external experts Make debrieﬁngs after programs and conferences mandatory Require attendees at external conferences to spend time reﬂecting on lessons learned from programs that can be directly applied to institution-based patient safety improvements Organization-wide opportunities Direct employees to educational and training opportunities Hold ‘‘executive chats’’ – Encourage ‘‘real,’’ meaningful conversations between executives and staff about the issues of patient safety and medical errors in both structured and unstructured contexts Monthly ‘‘meet the leader’’ talks – CEO and senior executives give brief lectures about patient safety and culture; establish link between patient safety and each executive role Job shadowing, especially a program pairing clinicians with administrators to learn more about clinical processes, medication administration, etc. Debrief and discuss actual medical errors and adverse events with broad audience – take the time, provide safety, encourage transparency Use of recognition for risk taking, new ideas to improve patient safety culture Require administrators to spend time with physicians, in physician meetings, in clinical areas, and so forth

support employee development. Ideally, this approach can help organizations work toward the development of more ‘‘system-ness,’’ as recommended by the Institute of Medicine (2001), as they try to solve complicated patient safety problems through improvement of healthcare management. Critical Characteristics of Successful Organizational Development Programs Successful OD programs, whether focused on patient safety improvement speciﬁcally or on organizational improvement as a whole, share several characteristics in common. These critical characteristics are listed in Table 4, and include considerations of program structure, politics, inclusiveness, competencies, and evaluation. First, program ﬂexibility is essential, especially given rapid changes within the healthcare industry. A successful OD program will enable introduction

226

ANN SCHECK MCALEARNEY

Table 4.

Characteristics of Successful Organizational Development Programs.

1. Flexible program structure Permits introduction of new topics Incorporates of new trainers/faculty/educators Programs can be stratiﬁed by target audience 2. Program connected to senior executive leadership Guided by organizational priorities Program results shared with leadership on regular basis 3. Inclusiveness of programs Participation of all levels of employees possible Web access and on-line learning options incorporated/planned 4. Competence and capacity of internal department Ability to tailor program offerings Capacity to work with many people, different groups, organizational politics Understanding of adult learning principles and challenges Linkage to clinicians, clinical impact, physician education Capacity to deliver multiple methods of program delivery (e.g., experiential learning, webbased, case study, simulations) 5. Measurement and evaluation Linked to strategic priorities and expectations of organization Capacity to target speciﬁc metrics and measures with OD programs (e.g., JCAHO, CMS Core Measures) Understanding of opportunity to deﬁne program value and track multiple measures for different stakeholders (e.g., deﬁning and building business case for organization)

of new topics, and tailoring of programs or program focus to include an emphasis such as patient safety. As one respondent explained, ‘‘We try to ﬁnd out who is best at it. We had to ﬁgure out what would work in our culture,’’ and then they delivered the program they developed that was most appropriate for the given audience. With program ﬂexibility, organizations can be as responsive as need be, whether to demands from employees, physicians, patients, regulators, or the market. Another critical characteristic of organizational programs is a strong connection to senior leadership. As one informant noted, ‘‘Executive engagement. You have to have the top person very engaged in the process in terms of deﬁning what is important for the leadership in that organization and then holding people accountable. So executive engagement and accountability [are critical].’’ Programs that are guided by organizational priorities and regularly monitored through tactics such as requirements for periodic reports and presentations to organizational leadership are more likely to be recognized and sustained for the contributions they make to organizational improvement.

Improving Patient Safety

227

A third critical characteristic of OD programs is their inclusiveness. Considering an emphasis such as patient safety, this likely involves employees at all levels, in addition to physicians who are associated with but not employed by the organization. One informant explained how they attempted to reach across organizational lines to ensure program inclusiveness by incorporating technology: ‘‘For our patient safety, we have put all of that on-line. And everyone had to go through that – even if you don’t touch patients.’’ However, recognizing that clinicians may not have easy access to computers or free time to complete programs is another issue that must be addressed in development programs. Successful programs determine feasible alternatives for the multiple audiences targeted, and establish regular feedback mechanisms to ensure that programs are attaining their goals and objectives. Clinical audiences may be offered programs tied to continuing medical education (CME) credits or other CE credits, thereby making achievement of speciﬁc program goals and objectives explicit. Fourth among these critical characteristics is the collective competence and capability of the employees working in the organizational department accountable for OD. This particular group of individuals must have the capacity to work with a myriad of different people, and negotiate the often complex organizational politics associated with healthcare institutions. Further, successful OD professionals recognize the demands of their audiences of adult learners, and are able to customize programs to participants’ needs. As one respondent explained, ‘‘The education we create for each of our employees has to be focused on adult learning. Some are 30-year employees. It [the program] must be comprehensive and have the educational pieces you need.’’ Finally, successful OD approaches will build evaluation into their programs to ensure that goals and objectives are established, measured, and reached. One key informant noted how, ‘‘After everyone has been through it, you have to assess it. Are our indicators getting better? If you don’t assess the results of that, then you don’t get the full picture.’’ Determining which indicators and metrics matter to the organization is ideally part of the program development process in which goals and objectives are established, for both the short- and long-term. One interviewee noted how, ‘‘For safety we are taking the Joint Commission and National Quality Forum’s indicators y and orienting ourselves around that.’’ Alternative metrics that have been suggested and could be adapted for measurement of patient safety improvement are shown in Table 5. Considering these characteristics in combination, the successful program will be ﬂexible enough to respond appropriately to the demands of senior

228

ANN SCHECK MCALEARNEY

Table 5. Examples of Alternative Metrics by which to Evaluate Patient Safety Improvement through Organizational Development Programs. Program metrics

Evaluation of individual patient safety programs Participation rates Program quality, satisfaction with programs Participant evaluation Knowledge, skills, behavior of participants Evaluations by subordinates, peers, superiors

Organizational metrics

Quality of work life Perception of organizational culture around patient safety Comfort with patient safety in organization Employee satisfaction with organization Willingness to refer friend/relative for care Willingness to see organizational provider Financial metrics Amount spent on program as % of total expense Value (deﬁned by program) for expense Patient satisfaction With care, safety, providers Willingness to refer friend/relative for care Achievement of key strategic goals of organization National patient safety goals CMS Core Measures Institute for Healthcare Improvement campaigns Additional patient safety, quality improvement initiatives

Clinical Metrics (Institute of Medicine Aims for Improvement – IOM 2001)

Safe Standardized mortality rate Adverse drug events per 1,000 doses administered Effective Lost work days per employee Growth in market share Patient safety statistics Perceptions of patient safety in culture Patient-centered Patient satisfaction Patient education statistics Timely Process: programs delivered in time frame Process: responsiveness to reported errors Efﬁcient Cost per adjusted admission Operating margin: cash from operations Equitable Disparities in error rates Disparities in utilization

Intangibles

Conﬁdence about patient safety Employee morale

Improving Patient Safety

229

leaders and the marketplace, tailor program offerings to multiple audiences in need of training and development, and deliver high-quality programs, measured on the basis of achievement of accepted goals and objectives. In practice, an OD practitioner might describe this program as did one of our informants: ‘‘We look at it as supporting the managers who are supporting the employees who are then supporting the patients.’’ Thus, by linking OD program goals to patients and clinical outcomes, a program will maximize its likelihood of both achieving its objectives, and making important improvements in patient safety. Developing a Patient Safety Culture As previously described, an OD program presents an organizing structure for investments in education and development of employees. Yet success in patient safety improvement is increasing acknowledged as requiring changes in organizational culture to ensure that correct behaviors occur and are institutionalized (e.g., Smith & Kaluzny, 1986) or ‘‘hard-wired’’ (e.g., Studer, 2003). Such changes are notably difﬁcult (e.g., Schein, 1985; Kotter, 1996; Deal & Kennedy, 1982), but, an OD framework can help organize and deliver interventions that can facilitate the change process. As one organizational leader noted, ‘‘If you want to shape a culture or how you do what you do, there has to be an investment in the formation and development and education and training of leadership.’’ Another explained how they try to, ‘‘create a culture that the patient is why we exist and that everything should surround that and that the workforce is your biggest tool to accomplish that goal.’’ Thus, in practice, the organization that decides to make patient safety improvement primary among other strategic priorities will invest in the education and development required to train employees and physicians in the new expectations and behaviors. Examples of how an OD approach can help shift culture toward patient safety improvement are provided in Table 6. Making progress toward patient safety improvement in healthcare is particularly difﬁcult because of the nature of the problem: patient safety issues involve errors, and neither admitting nor discussing these mistakes is comfortable for most individuals (e.g., Rosenthal & Sutcliffe, 2002). In healthcare organizations, encouraging attention to these issues is often described as an effort to become more ‘‘transparent’’ about what has occurred or might have occurred around patient care by emphasizing data and providing safety around communications about errors. As one informant described, ‘‘We are trying to create a culture where it is transparent.’’ Yet, as another interviewee explained, ‘‘That is a really

230

Table 6.

ANN SCHECK MCALEARNEY

Helping Shift Organizational Emphasis Toward Patient Safety Improvement.

Organizational Development Approach

Potential Action

Change expectations for managers

Build responsibility for patient safety into every manager’s job description; hold accountable Expect managers to take time to train and develop their employees in a patient safety culture Create expectation for application of new skills, new ideas and follow-up after any developmental program

Enhance human resources practices

Evaluate and leverage existing expertise of HR in organizational development to focus education and development modules, courses, and programs on patient safety improvement Challenge HR to initiate in-house patient safety improvement program using examples from other organizations Begin to shift HR model from policies and procedures to organizational development, if development focus is absent

Focus on senior leadership Consider initiating action learning projects focused on patient development safety improvement Require regular reports from patient safety and improvement teams, projects, and other activities to senior leadership Consider developing patient safety dashboard of indicators and training senior leaders to address these metrics

different culture for us. We have developed it to be about learning, not about punishment.’’ In practice, this may involve focusing on patientcentered care, as pursued by a different organization: ‘‘We put the patient in the center and we actually believe that we’re oriented to the patient as opposed to doing the work so it’s easier for us.’’ But making this transition involves considerable effort. Organizations that succeed in this transformation often succeed by following a strategy that, as one informant explained, is ‘‘a multi-pronged solution to patient safety issues – cultural, philosophical, practical, and operational.’’ An OD framework can be useful by providing multiple points of emphasis for these ‘‘many prongs’’ of the strategy. Then, in executing the strategy, an organization can follow concrete operational steps, as suggested in Table 7, in order to ensure OD programs have the critical characteristics previously listed that will maximize the likelihood of program success. Ideally, the result of incorporating an OD approach is to help the entire organization move forward toward shared goals of patient safety

231

Improving Patient Safety

Table 7.

Six Steps to Progress toward Patient Safety Improvement.

Step 1: Get CEO, Board involved and committed

Align patient safety program goals with organizational strategies, vision, mission, and values Without senior management support and commitment, limit program development activities

Step 2: Designate executive ofﬁcer to lead

Establish a formal plan and revise periodically Deﬁne program accountability

Step 3: Assess organizational needs and goals

Perform organizational audit to determine needs of organization and its employees; consider qualitative and quantitative assessment of organization on basis of culture and climate around patient safety

Time: Allow time for employees to be trained and Step 4: Invest in patient safety-focused develop in new patient safety behaviors organizational development programs Money: Begin investment and plan for growth over time Efﬁciency: Recognize and plan for potential productivity losses and gains as employees progress along the learning curve and attempt to implement principles and practices learned; adjust operations accordingly Attention: Because ‘‘the currency of leadership is attention,’’ if leadership pays attention and reinforces the patient safety message, the message to the organization is clear Value: Make clear that organizational and employee development are valued; participants are respected and not resented; acknowledge value of human capital in the organization and potential of better-educated and trained employees to improve culture and climate of organization Flexibility: Enable possibilities for program modiﬁcations, developmental experiences for participants, etc. Step 5: Establish measurement mechanisms

Take baseline measurements Tie program results to unit level Include systematic assessment and review of participants Include both short- and long-term metrics

232

ANN SCHECK MCALEARNEY

Table 7. (Continued ) Step 6: Work to create a transparent patient safety culture

Build ability to take risks, to fail Foster coaching and mentoring Ensure organizational ﬂexibility – job assignments, rotations Encourage openness to new ideas Foster open, direct, honest communications Guarantee profound respect for people, their development Emphasize the value of human capital Ensure that participants are respected, not resented Focus on patient safety, error reduction, quality of care Create emphasis on service, hospitality, and respect throughout the organization: to patients; to physicians; to employees; to community

improvement, within an organizational culture that values patient safety. A picture of success at one organization was described as follows: ‘‘We have patient safety red rules. For example, if you are a patient and you have a red band, I have to check and make sure that your arm band matches the medication given. We educate everyone in clinical and non-clinical areas. This is who we are: we are keeping our patients safe.’’ By equating ‘‘who we are’’ with ‘‘keeping our patients safe,’’ this individual provides a strong statement about the patient safety culture she believes exists within this organization. Further, by establishing a link between this safety culture and organization-wide education, she is providing an example of how development activities can promote needed culture change. Though this type of linkage may not yet be common, the potential certainly exists to use OD as an organizing framework. By focusing on opportunities for organizational improvement, organizational interventions can then be developed to improve patient safety. As one individual described how this framework works in his organization, the culture can be used to encourage thoughtful reﬂection about improvement opportunities: ‘‘What are we doing right? What are we doing wrong? How can we embed it throughout the organization?’’ Thus an OD framework that helps facilitate a shift toward a more patient safety-focused, improvement-oriented culture can be a powerful tool for an organization challenged to improve patient safety and the quality of healthcare provided.

Improving Patient Safety

233

A CONCEPTUAL MODEL From these data and considerations, I propose a conceptual model depicting the organizational inﬂuences and impacts of an OD approach to patient safety improvement (Fig. 1). In this model an organization’s commitment and/or ability to commit to patient-safety focused development programs is inﬂuenced by three major organizational factors: strategy, structure, and culture. As a result, the level of organizational commitment to this approach will constrain decisions about program design and delivery, thereby inﬂuencing the depth and breadth of programs offered. Finally, program design decisions will impact patient safety and the success of these programs; especially as design limits program reach, scope, and goals. Further, changes in any of the organizational factors of strategy, structure, or culture may modify the organization’s commitment to this approach, or thus potentially inﬂuence program design decisions and, as a result, program impacts on patient safety. Consider a healthcare organization that places a high value on patientcentered care and has recently made a major investment in health information technology (HIT) as an approach to improving patient safety. In this organization, the HIT investment decision was made based on an assessment of strategic priorities, and greatly improves the potential capabilities and coordination of information systems, as well as the use of data around the organization, assuming that implementation is successful. If this organization had a high level of commitment to an OD approach to patient safety improvement, both human and ﬁnancial resources would be made available to facilitate the implementation of the HIT, to train users of the technologies and the data, and to educate employees about the importance of this HIT in improving patient safety. A well-placed program director would be designated to deploy and direct these resources, and this individual would be accountable for both employee learning and use of the new technology, and for expected improvements in patient safety over time. Programs would then be designed and tailored to the needs of the different clinicians and nonclinicians affected by this HIT and the data it generates, and metrics would be established by which to evaluate the success of the education and development programs, as well as the satisfaction of employees, and potentially patients, with the new technology. Finally, the impact of this organizational commitment on patient safety would be measurable, with improvements such as reductions in adverse events, enhancement of organizational culture, and increases in the aforementioned patient and employee satisfaction. In contrast, the organization with low commitment to an OD approach to

ANN SCHECK MCALEARNEY

234

Organizational Strategy • Urgency of focus on patient safety • Time horizon for organizational decisions • Opportunities to link development to other strategic initiatives (e.g., feasibility of action learning projects)

Organizational Structure • Placement of leadership development, OD function • Centralization of decision making about development resource allocation • Information systems capabilities and coordination

• • • •

Organizational Culture Transparency of culture around medical errors Senior executive support, attention Value of patient-centered care Ease of collaboration and coordination among clinicians, non-clinicians

Commitment to Organizational Development Approach Resource availability for program development and delivery Position and power of program director(s) Expectations and accountability for program participation Sustainability of commitment to programs

• • • •

• • •

Program Design Decisions Breadth and depth of program offerings based on assessment of organizational needs Differentiated program development for different audiences (e.g., for physicians versus for administrators) Metrics to assess programs

• • • •

Impact on Patient Safety Improvement in metrics related to patient safety Increase in patient satisfaction Improvement in patient safety culture and climate Increase in employee satisfaction with organization, positive organizational culture

Fig. 1. Conceptual Model Depicting Inﬂuences on and Impacts of Organizational Commitment to Organizational Development Approach to Patient Safety Improvement.

patient safety improvement might have to rely upon disjointed technology implementation processes, uncoordinated measurement and evaluation, resulting in scant impact on organizational culture, and little ability to link education and development activities to patient safety improvements.

235

Improving Patient Safety

DISCUSSION While efforts are underway to both measure patient safety culture (e.g., Agency for Healthcare Research and Quality’s safety culture assessment) and determine ways to provide higher-quality or more ‘‘perfect’’ care (e.g., Robert Wood Johnson Foundation’s ‘‘Pursuing Perfection’’), few initiatives have moved beyond speculation about the important roles of leadership, organizational culture, and systems factors to determine how to make these changes. This chapter has presented and discussed ideas to improve patient safety that are grounded in an OD framework, and provided guidance about how to move forward with these concepts. Additional research is clearly needed to both validate this conceptual model and attempt to formally assess the impact of individual OD programs on patient safety outcomes. However, most patient safety improvement initiatives could likely be enhanced by considering opportunities to broaden their reach and improve their likelihood of success by linking such initiatives to organized developmental strategies. Speciﬁcally, by taking advantage of the capabilities and competencies of most healthcare organizations’ training, education, and development departments, education and learning experts can ensure that patient safety programs and initiatives are properly targeted, delivered, and measured. Moreover, in organizations without this capability, I would advocate that such opportunities provide strong impetus to develop or strengthen this area. Organizational change is admittedly difﬁcult, and changing organizational culture to include patient safety is perhaps one of the more challenging organizational change initiatives recently introduced in U.S. healthcare organizations. As such a change can be deﬁned to include all individuals associated with the organization – (clinicians and non-clinicians, employed and nonemployed), – an organization-wide effort is required. Thus an OD approach offers a logical framework by which to organize and facilitate the required changes.

CONCLUSION As this chapter has discussed, healthcare organizations can use OD as a conceptual umbrella for training and development, helping to improve and extend patient safety education and training efforts. Similarly, organizations committed to changing organizational culture to create a patient safetyfocused culture can use development initiatives strategically to support these

236

ANN SCHECK MCALEARNEY

culture change efforts. It is hoped that the concepts and ﬁndings described in this chapter can help healthcare organizations make strides toward positive changes in organizational culture and climate that can effectively improve patient safety in healthcare.

ACKNOWLEDGMENTS The research reported in this chapter has been supported by grants from the Center for Health Management Research, the American College of Healthcare Executives, and the Division of Health Services Management and Policy at The Ohio State University. I greatly appreciate the help of all study participants, as well as the research assistance provided by Katrina Buchholtz, Sarah Hoshaw, Viktorya Pelts, Mindy Marcum Slenn, Stacy Baker, Jordan Harmon, and Justin Barlow, all afﬁliated with The Ohio State University during the studies. In addition, I am grateful to the editors of this volume and to my anonymous reviewers for their suggestions to help improve this chapter.

REFERENCES Bates, D., Leape, L., Cullen, D., Laird, N., Petersen, L. A., Teich, J. M., Burdick, E., Hickey, M., Kleeﬁeld, S., Shea, B., Vander Vliet, M., & Seger, D. L. (1998). Effect of computerized physician order entry and a team intervention on prevention of serious medication errors. Journal of the American Medical Association, 280, 1311–1316. Bates, D., Teich, J., Lee, J., Seger, D., Kuperman, G. J., Ma’Luf, N., Boyle, D., & Leape, L. (1999). The impact of computerized physician order entry on medication error prevention. Journal of the American Medical Informatics Association, 6, 313–321. Bobb, A., Gleason, K., Husch, M., Feinglass, J., Yarnold, P. R., & Noskin, G. A. (2004). The epidemiology of prescribing errors: The potential impact of computerized prescriber order entry. Archives of Internal Medicine, 164(7), 785–792. Burke, J. M., & Day, R. R. (1986). A cumulative study of the effectiveness of managerial training. Journal of Applied Psychology, 71, 232–245. Catlin, A., Cowan, C., Hefﬂer, S., & Washington, B. (2007). National health spending in 2005: The slowdown continues. Health Affairs, 26(1), 142–153. Conger, J., & Benjamin, B. (1999). Building leaders: How successful companies develop the next generation. San Francisco: Jossey-Bass. Constas, M. (1992). Qualitative analysis as a public event: The documentation of category development procedures. American Education Research Journal, 29, 253–266. Cousins, D. H., Sabatier, B., Begue, D., Schmitt, C., & Hoppe-Tichy, T. (2005). Medication errors in intravenous drug preparation and administration: A multicentre audit in the UK, Germany, and France. Quality & Safety in Health Care, 14(3), 190–195. Crabtree, F., & Miller, W. (1999). Doing qualitative research. Thousand Oaks, CA: Sage.

Improving Patient Safety

237

Day, D. V., Zaccaro, S. J., & Halpin, S. M. (2004). Leader development for transforming organizations. Growing leaders for tomorrow. Mahwah, NJ: Lawrence Erlbaum Associates. Deal, T. E., & Kennedy, A. A. (1982). Corporate cultures: The rites and rituals of corporate life. Reading, MA: Addison Wesley Publishers. French, W. L., & Bell, C. H. (1978). Organization development: Behavioral science interventions for organization improvement (2nd ed). Englewood Cliffs, NJ: Prentice-Hall. French, W. L., Bell, C. H., & Zawacki, R. A. (1989). Organization gevelopment: Theory, practice, and research (3rd Edition). Homewood, IL: Richard Irwin, Inc. Fulmer, R. M., & Goldsmith, M. (2001). The leadership investment. How the world’s best organizations gain strategic advantage through leadership development. New York: AMACOM. Gawande, A. (2004). On washing hands. New England Journal of Medicine, 350(13), 1283–1286. Giber, D., Carter, L., & Goldsmith, M. (2000). Best practices in leadership development handbook: Case studies, instruments, training. San Francisco: Jossey-Bass/Pfeiffer. Glaser, B., & Strauss, A. (1967). The discovery of grounded theory: Strategies for qualitative research. New York: Aldine de Gruyter. Gray, H., & Snell, R. (1985). Towards effective practice where management development is a recent concern. Leadership and Organizational Development Journal, 7, 21–26. Harry, M., & Schroeder, R. (2000). Six sigma. New York: Doubleday Publishers. Hofmann, P. B. (2005). Acknowledging and examining management mistakes. In: P. B. Hofmann & F. Perry (Eds), Management mistakes in healthcare: Identiﬁcation, correction, and prevention (Chapter 1, pp. 3–27). Cambridge: Cambridge University Press. Hofmann, P. B., & Perry, F. (Eds). (2005). Management mistakes in healthcare: Identiﬁcation, correction, and prevention. Cambridge: Cambridge University Press. Husch, M., Sullivan, C., Rooney, D., Barnard, C., Fotis, M., Clarke, J., & Noskin, G. (2005). Insights from the sharp end of intravenous medication errors: Implications for infusion pump technology. Quality & Safety in Health Care, 14, 80–86. Institute for the Future. (2000). Health and health care 2010: The forecast, the challenge. San Francisco, CA: Jossey-Bass. Institute of Medicine. (2001). Crossing the quality chasm: A new health system for the 21st century. Washington, DC: National Academy Press. Joint Commission on Accreditation of Healthcare Organizations (2007a). National patient safety goals. Available at: www.jointcommission.org/PatientSafety/NationalPatient SafetyGoals/. Accessed on February 22, 2007. Joint Commission on Accreditation of Healthcare Organizations (2007b). The Joint Commission seeks examples of hand hygiene measurement methods. www.jointcommission.org/ AccreditationPrograms/hand_hygiene.htm. Accessed on February 22, 2007. Jones, W. J. (2005). Identifying, classifying, and disclosing mistakes. In: P. B. Hofmann & F. Perry (Eds), Management mistakes in healthcare: Identiﬁcation, correction, and prevention (Chapter 3, pp. 40–73). Cambridge: Cambridge University Press. Kaushal, R., Bates, D. W., Landrigan, C., McKenna, K. J., Clapp, M. D., Federico, F., & Goldmann, D. A. (2001). Medication errors and adverse drug events in pediatric inpatients. Journal of the American Medical Association, 285, 2114–2120. Kerr, E. A., McGlynn, E. A., Adams, J., Keesey, J., & Asch, S. M. (2004). Proﬁling the quality of care in twelve communities: Results from the CQI study. Health Affairs, 23(3), 247–256.

238

ANN SCHECK MCALEARNEY

Kohn, L., Corrigan, J., & Donaldson, M. (Eds). (1999). To err is human: Building a safer health system. Washington, DC: National Academy Press. Kotter, J. (1996). Leading change. Boston: Harvard Business School Press. Kouzes, J. M., & Posner, B. Z. (1995). The leadership challenge (2nd ed). San Francisco: Jossey-Bass. Leape, L. L., & Bates, D. W. (1995). Systems analysis of adverse drug events. Journal of the American Medical Association, 274(1), 35–43. Leape, L. L., Brennan, T. A., Laird, N. M., Lawthers, A. G., Localio, A. R., Barnes, B. A., Hebert, L., Newhouse, J. P., Weiler, P. C., & 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. Maxwell, J. (1996). Qualitative research design. Thousand Oaks, CA: Sage. McAlearney, A. S. (2006). Leadership development in healthcare organizations: A qualitative study. Journal of Organizational Behavior, 27(7), 967–982. McAlearney, A. S., Chisolm, D., Schweikhart, S. B., Medow, M. A., & Kelleher, K. (2007). The story behind the story: Physician skepticism about relying on information technologies to reduce medical errors. International Journal of Medical Informatics, 76, 836–842. McCall, M. W., Lombardo, M. M., & Morrison, A. M. (1998). The lessons of experience: How successful executives develop on the job. Lexington, MA: Lexington Press. McCauley, C. D., Moxley, R. S., & VanVelson, E. (1998). The center for creative leadership handbook of leadership development. San Francisco: Jossey-Bass. McCracken, G. (1988). The long interview. Thousand Oaks, CA: Sage. McGlynn, E. A., Asch, S. M., Adams, J., Keesey, J., Hicks, J., DeCristofaro, A., & Kerr, E. A. (2003). The quality of healthcare delivered to adults in the United States. New England Journal of Medicine, 348(26), 2635–2645. Mecklenburg, G. (2001). Career performance: How are we doing? Journal of Healthcare Management, 46(1), 8–13. Miles, M., & Huberman, A. (1994). Qualitative data analysis. Thousand Oaks, CA: Sage. Moxnes, P., & Eilertsen, D. (1991). The inﬂuence of management training upon organizational climate: An exploratory study. Journal of Organizational Behavior, 12, 399–411. O’Hare, M. C. B., Gallagher, T., & Shields, M. D. (1995). Errors in the administration of intravenous drugs. British Medical Journal, 310, 1536–1537. Ramanujam, R., & Rousseau, D. M. (2004). Organizational behavior in healthcare – the challenges are organizational, not just clinical. Journal of Organizational Behavior, 25, 667–669. Revans, R. W. (1980). Action learning. London: Blond and Briggs. Rosenthal, M. M., & Sutcliffe, K. M. (Eds). (2002). Medical error: What do we know? What do we do? San Francisco, CA: Jossey-Bass Publishers. Russell, J. A., & Greenspan, B. (2005). Correcting and preventing management mistakes. In: P. B. Hofmann & F. Perry (Eds), Management mistakes in healthcare: Identiﬁcation, correction, and prevention (Chapter 5, pp. 84–102). Cambridge: Cambridge University Press. Schein, E. (1985). Organizational culture and leadership. A dynamic view. San Francisco: Jossey-Bass. Schneller, E. S. (1997). Accountability for health care: A white paper on leadership and management for the U.S. health care system. Health Care Management Review, 22(1), 38–48. Scientiﬁc Software Development. (1998). Atlas.ti (4.2 ed). Berlin: Scientiﬁc Software Development.

Improving Patient Safety

239

Smedley, B., Institute of Medicine, Stith, A. Y., & Nelson, A. R. (2002). Unequal treatment: Confronting racial and ethnic disparities in health. Washington, DC: National Academies Press. Smith, D., & Kaluzny, A. (1986). The white labyrinth. Chicago, IL: Health Administration Press. Strauss, A., & Corbin, J. (1998). Basics of qualitative research: Techniques and procedures for developing grounded theory. Thousand Oaks, CA: Sage. Studer, Q. (2003). Hardwiring excellence. Gulf Breeze, FL: Firestarter Publishing. Taxis, K., & Barber, N. (2003). An ethnographic study of incidence and severity of intravenous drug errors. British Medical Journal, 326, 684–687. Tichy, N. M. (1999). The leadership engine. How winning companies build leaders at every level. New York: Harper Business.

S*T*A*R*T: A GREAT HANDOFF – AN APPROACH TO EFFECTIVE MEDICAL COMMUNICATION IN A HIGH-RISK ENVIRONMENT$ Ellen S. Deutsch, Brian Binck, Gina Moore and Ben-Tzion Karsh ABSTRACT When responsibility for an individual patient is transferred within or between groups of medical professionals, a ‘‘handoff’’ occurs so that patient-speciﬁc medical information can be provided to the medical professional(s) assuming responsibility for that patient. Providing an appropriate summary supports safe, high-quality, effective medical care; inadequate or incorrect information may create risk for the patient. A handoff approach was developed to facilitate this process, using the mnemonic ‘‘S*T*A*R*T’’ (S: situation; T: therapies; A: anticipated course; R: reconciliation; T: transfer). Surveys of handoffs occurring before and after introduction of the S*T*A*R*T system document areas with potential for process improvement.

$

The authors received no ﬁnancial support for this study.

Patient Safety and Health Care Management Advances in Health Care Management, Volume 7, 241–258 Copyright r 2008 by Emerald Group Publishing Limited All rights of reproduction in any form reserved ISSN: 1474-8231/doi:10.1016/S1474-8231(08)07011-0

241

242

ELLEN S. DEUTSCH ET AL.

INTRODUCTION ‘‘Handoff’’ is a colloquial term for a contemporaneous process that is expected to occur when responsibility for an individual patient is transferred, so that patient-speciﬁc medical information can be provided to the medical professional and/or the professional team assuming responsibility for that patient. The handoff may occur between nurses at change of shift, between physicians beginning or ending ‘‘call’’ responsibilities, and between other health care professionals who take turns in their care for patients. Handoffs may also occur between physicians, nurses, and others when patient care is transitioned between different specialties or different locations or to different levels of acuity such as into or out of the intensive care unit (ICU), nursing home, etc. The information provided is intended to support continuity, quality, and safety, and generally occurs from physician-to-physician, nurse-to-nurse, etc. This transfer of responsibility may be temporary or long term. ‘‘Sign-outs’’ comprise a particular subset of handoffs, wherein the responsibility for a group of patients is transferred from one team of residents to another, often occurring as a change in work shift rather than a change in the level or location of care required for an individual patient. The number of signouts affecting an individual patient may increase because of changes in resident work hour rules (Frank, Lawler, Jackson, Steinberg, & Lawless, 2005a).

Risks During Handoffs During transfers of care, providing an appropriate summary of the patient’s medical condition and expected medical course is intended to facilitate efﬁcient, effective handoffs. However, studies examining the handoff process have documented that it is variable, unstructured, and prone to error (Arora, Johnson, Lovinger, Humphrey, & Meltzer, 2005; Frank, Lawless, & Steinberg et al., 2005b; Solet, Norvell, Rutan, Rutan, & Frankel, 2005); that it ‘‘often happens in a remarkably haphazard manner’’ (Volpp & Grande, 2003); and that handoff communication failures can have adverse clinical outcomes for patients (Horwitz, Krumholz, Green, & Huot, 2006). Handoffs are not often taught in a didactic manner and variability occurs in their preparation, content, and methods (Arora et al., 2005; Solet et al., 2005). The sign-out process has been described as a ‘‘precarious exchange’’ (Mukherjee, 2004), with inaccurate information documented on sign-out sheets maintained by residents using word processing or database programs

243

S*T*A*R*T: A Great Handoff

(Frank et al., 2005a, 2005b). Arora et al. (2005) conducted interviews of interns using a critical incident technique to identify and describe near misses and adverse events resulting from poor sign-outs and to evaluate their peers’ sign-outs. Reported error categories included content omission, such as omission of active medical problems, medications, and test results, and failure-prone communication processes, such as multiple sign-outs. The most frequent suggestion elicited was incorporating verbal communication of anticipated problems. In our study hospital, statistics directly correlating shortcomings of the handoff process and serious events are not available; a review of statistics addressing the broader category of recent serious events suggests that communication problems are the most common root cause (Fig. 1). Likewise, a review of recent patient safety assessments obtained through root cause analyses and through surveys and in-depth interviews of employees, ‘‘consumers,’’ and ‘‘teams’’ includes communication as the most consistently involved theme (Fig. 2). Handoff quality affects both caregivers and patients. Problematic handoffs may lead to frustration, anxiety, and uncertainty among caregivers, and, worse, caregivers may not even know they do not have complete information. This can decrease satisfaction and conﬁdence among patients and can lead to poor treatment decisions that may compromise the quality and safety of care (Arora et al., 2005; Hohenhaus, Powel, & Hohenhaus, 2006). Written 16% Cognitive 39%

Tangible 45%

Fig. 1. Relative Distribution of Cognitive, Tangible, and Written Factors Contributing to Serious Events at Our Hospital, 2002–2005. Communication was the Largest Single Cognitive Factor; Other Cognitive Factors Included Knowledge, Behavior, etc. Tangible Factors Included Orders, Equipment, Medication. Written Factors Included Consents, Documentation, and Policies and Procedures.

244

ELLEN S. DEUTSCH ET AL.

Themes identified from various patient safety assessments

Safety Actual Patients & Behaviors Employees Families Events Team

COMMUNICATION CULTURE ENVIRONMENT EQUIPMENT KNOWLEDGE MEDICATION SAFETY POLICY & PROCEDURES TEAMWORK TREATMENT/CARE

OTHER:

balancing safety and productivity

patient time pressure identification

Fig. 2. Patient Safety Concerns Identiﬁed from Various Patient Safety Assessments and Surveys, 2002–2005. Note that Each Survey or Interview Assessment Varied in Safety Content. While Direct Comparisons Across them Cannot be Made, Communication is a Common Theme.

Lessons from Other High-Risk Professions Health care is sometimes described as ‘‘managed risk,’’ and strategies to improve the risk/beneﬁt ratio are valued. Other high-risk professions, such as aviation, aeronautics, and the nuclear power industry, have addressed the potential for errors related to the handoff of information during shift change. Their lessons learned may have cautious application in the ﬁeld of medicine. Patterson and Woods (2001) studied shift change processes in aeronautics and found that the process depended on both ‘‘updates’’ and prior knowledge. Event recognition, prioritization, and communication were different from that provided in the traditional automated logs. Controllers rarely communicated exact data values but described events at a ‘‘higher level based on a complex combination of multiple parameters, not all of which would independently be viewed to be out of normal ranges for most situations and not all of which occurred simultaneously in a discrete fashion.’’ Not all of the nominal events were included in the update, although off-nominal and anomalous events were. Logs were principally used to document past information, whereas much of the handover content related to future events, activities, analyses, and decisions to facilitate preparation and anticipation by the incoming controller.

S*T*A*R*T: A Great Handoff

245

Regulatory Considerations Along with caregivers, regulatory agencies have recognized the risks involved in handoffs. Statistics from the Joint Commission on Accreditation of Healthcare Organizations (JCAHO): (now known as Joint Commission [JC]) suggest that approximately 2/3 of sentinel events include communication failures among their root causes (JCAHO, no date). Their National Patient Safety Goals include improving ‘‘the effectiveness of communication among caregivers,’’ with an additional requirement in 2006 to ‘‘implement a standardized approach to ‘hand off’ communication, including an opportunity to ask and respond to questions’’ (JCAHO, 2006a, 2006b, 2007). Implementation expectations include using a process that is interactive, is up-to-date, limits interruptions, and includes an opportunity for the receiver to review relevant historical data. There exists a ﬁeld of research on communication that can be used to guide the development of better handoff processes.

Information and Communication Theory Although unique challenges exist in patient care situations, principles of information and communication theory have relevance to this specialized form of medical communication. In the mid-20th century, Claude E. Shannon developed a unidirectional Mathematical Theory of Communication, which states that all communication involves 3 steps: coding a message at its source, transmitting the message through a communications channel, and decoding the message at its destination. In addition, he addressed uncertainty or unpredictability as a stimulus and justiﬁcation for communication, as well as the principles of ‘‘noise,’’ data compression, information redundancy, and error reduction. Although his theory was mathematical, these principles have ramiﬁcations for medical content and context (Shannon, 1948). Subsequent to this, Schramm (1954) developed a Circular Model, which expands the communication process to include a feedback system: an encoder sends a message to a decoder, which interprets the message and then encodes a message back to the sender; the sender decodes and interprets that message, and a continuing process of interaction occurs (Fig. 3). Later in the 20th century, Charles Berger addressed the use of communication as a tool to cope with the human need to reduce uncertainty. His Uncertainty Reduction Theory describes how communication strategies are used in social situations to decrease high degrees of uncertainty that interfere with behavior prediction and increase anxiety (Berger & Bradac, 1985).

246

ELLEN S. DEUTSCH ET AL.

MESSAGE

ENCODER

DECODER

INTERPRETER

INTERPRETER

DECODER

ENCODER

MESSAGE

Fig. 3. Osgood and Schramm’s Circular Model: Communication Includes Feedback and is a Continuing Process of Interaction. Used with Permission from Underwood (2005). Osgood and Schramm Circular Model. http://www. ccms-infobase.com. Date Accessed 07/04/2008.

In light of the literature documenting handoff problems, our hospital’s own data showing communication challenges and the 2006 National Patient Safety Goals, our hospital chose to develop a handoff process grounded in communication and organizational risk management theories to improve patient safety.

METHODS Algorithm Development The Clinical Safety Committee at our hospital created a multidisciplinary ‘‘Effective Handoffs Team’’ comprising physicians, nurses, respiratory therapists, and transport personnel as well as facilitators from the Safety Department and an Executive Sponsor. Using a consensus process, the committee developed an organizational approach to handoffs and an approach to educating medical professionals. Existing handoff processes and literature were reviewed. Among the various processes identiﬁed, SBAR (Situation, Background, Assessment,

S*T*A*R*T: A Great Handoff

247

Recommendation) seemed to be the relevant mnemonic most widely known, although it was not initially designed for use in the handoff process. SBAR has been used to strengthen the process of informing a physician that a patient’s condition requires increased attention. It does not inherently support interaction, because it does not incorporate reconciliation of either the patient’s expected course or the mutual understanding of the handoff giver and receiver. In addition, although the recommendation component could be limited to a recommendation that the patient be reevaluated, in practice this key component may be interpreted concretely, leading to inappropriate recommendations for intervention beyond the experience of the nurse or the physician. Although a rule-based checklist process would sometimes remind handoff participants to incorporate appropriate information, the committee was concerned that a checklist process could also include requests for information not relevant to an individual handoff, thereby distracting participants from recognizing additional important information, and could also provide a false sense of security by failing to include less common but potentially important information relevant to an individual handoff. For example, candidacy for subacute bacterial endocarditis (SBE) prophylaxis, the possibility of unrecognized pregnancy, termination of parental rights involving a child, and a history of colonization with methicillin-resistant staphylococcus aureus (MRSA) might be important information for a particular patient, but including a checklist reminder of all of the possible medical considerations affecting an individual patient would lead to an unwieldy handoff document. Therefore, the process was designed as a higher level ‘‘knowledge-based’’ interaction (Rasmussen, 1983; Reason, 2004) so that handoff content would be determined by using ‘‘situational awareness’’ (Endsley, 1995) supporting participants in the use of ‘‘professional standards,’’ ‘‘clinical judgment,’’ and ‘‘critical thinking.’’ Although it is important to include sufﬁcient information in a handoff, in the current information age, with our capability to access large volumes of data that are not always optimally organized or presented, we were also cognizant of the potential risks of incorporating too much information: ‘‘We failed to anticipate Pearl Harbor not for want of the relevant materials, but because of a plethora of irrelevant ones’’ (attributed to Roberta Wohlstetter). The intention of process design included avoiding complex checklists and nonproductive paperwork and facilitating, rather than distracting from, patient care. The committee developed the handoff process incorporating the principles outlined in Fig. 4. Additional design considerations included accommodating tertiary care center circumstances, in which a patient’s medical condition could range

248

ELLEN S. DEUTSCH ET AL. 1. 2. 3.

4. 5. 6. 7.

8.

Fig. 4.

Avoid limitations of a checklist approach Support knowledge-based approach Include discussion of expectations and parameters beyond which re-evaluation and possibly intervention are indicated Incorporate interaction between handoff giver and receiver Reconcile clinical expectations of patient’s progress Reconcile informational expectations of the handoff giver and receiver Incorporate a broad view of the patient’s expected course, including eventual patient care goals and disposition Titrate handoff content to address the various levels of experience of the caregivers and complexity of the medical circumstances

Handoff Principles Incorporated into S*T*A*R*T Design.

from relatively straightforward and routine to very complex and unusual. As a teaching institution, our approach should be suitable for professionals with various levels of training and knowledge, including developing professionals who are learning standards and habits of care. A ﬂexible process should be adaptable to the end user, supporting thorough, systematic handoffs important to novices, while allowing experts to ‘‘solve problems and make clinical decisions using fewer items of information’’ (Hawkins et al., 2004).

Design Considerations and Theory A handoff process was developed that generated the mnemonic ‘‘S*T*A*R*T.’’ For each letter in the mnemonic, the following information is suggested but not necessarily required; conversely, more detail should be added as appropriate for speciﬁc clinical circumstances. S

Situation, background: Current complaint or problem with succinct summary of background History and physical examination Allergies Medications

S*T*A*R*T: A Great Handoff

T

A

R

T

249

Isolation PMH PSH Social, etc. Therapies: Update on treatment, interventions, and course Procedures Interventions and their results Additional data Tests, evaluations, consults Laboratory tests pending and completed EKGs, audiograms, etc. Radiographs Findings Anticipated course: Assessment of current status and anticipated course Issues: considerations that may require additional attention or intervention Risks: potential complications or adverse outcomes and their management May be organized by organ system or other format Short or long term Reconciliation: Discussion and adjustments of plan May include physicians, nurses, respiratory therapy, family, ethics committee, other Plans, orders Short and long term Transfer: To another level of care or service, or to home or chronic care facility

Fig. 5 demonstrates the poster version of the mnemonic; a ‘‘pocket prompt’’ was also created. In addition, factors potentially affecting the level of detail suitable in particular circumstances are addressed as follows (Fig. 6): Handoff conversation requires increased level of detail based on: Increased medical complexity of patient Decreased educational level/experience of person handing off and person receiving information Decreased team experience with each other Decreased team experience with this patient Decreased team experience with this medical circumstance.

250

ELLEN S. DEUTSCH ET AL.

Fig. 5.

S*T*A*R*T Mnemonic, Poster Version.

251

S*T*A*R*T: A Great Handoff

S*T*A*R*T • • • • •

Situation, background Therapies Anticipated course Reconciliation Transfer

Include MORE detail based on: Increased medical complexity of patient Decreased educational level/experience of person handing off or person receiving information Decreased team experience with -- this medical circumstance -- this patient

Situational Scaling

-- each other

Fig. 6. Illustration of Increasing Level of Detail Appropriate for Handoffs Relative to the Education and Experience of the Patient Care Teams and the Medical Complexity of the Patient.

This structure reinforces a shared mental model, or social construct, including knowledge of what information to expect and recognition of missing components. This may reduce uncertainty and provide a reassuring expectation of content for various medical circumstances and for participants at various educational levels. The ‘‘anticipated course’’ component addresses the suggestion elicited by Arora et al. (2005) to verbally communicate anticipated problems. The ‘‘reconciliation’’ component prompts both the handoff giver and receiver to consider whether the patient’s medical course is proceeding appropriately, as well as providing a predictable opportunity to clarify information.

Survey of Handoff Quality A one-page survey of the handoff process was created, including basic group identiﬁers for the handoff giver and receiver and three sections of questions (Fig. 7). In the ﬁrst section, handoff quality was rated as poor to good based on information accuracy and completeness and the opportunity to ask

252

ELLEN S. DEUTSCH ET AL. EFFECTIVE HANDOFF COMMUNICATION STAFF SURVEY – PICU Pre-Implementation

INSTRUCTIONS FOR USE: Please complete the following survey after you have received report and assessed the patient. MRN: ____________________ Patient Name: ________________________ Date: ________________ Patient location prior to transfer: ____________________ after transfer: __________________________ Provider type giving report (circle one):

Physician Nurse

RT Therapy Srvs

Provider type receiving report (circle one): Physician Nurse RT

Therapy Srvs

SW

Other

SW

Other

How would you rate the following elements of the patient report (handoff) you received? 1 = Very Poor 5 = Very Good O O O

Very Poor

1

2

Very Good

3

4

5

Accuracy of information Completeness of information Opportunity to ask questions about the information What would have improved the accuracy or completeness of the report?

Other Comments:

Did the patient report (handoff) contain the following information? YES NO NA COMMENTS Patient identifiers (Name; DOB; Age) Allergies Pertinent Diagnosis/Diagnoses Code/DNR status Significant Past Medical History Current Medications Respiratory therapy (Ventilator; Treatments) IV therapy/Nutrition Equipment needs Planned diagnostics/tests Goals for the day (“To Do List”)

Please place completed survey in box provided at each nurses station

Fig. 7.

Effective Handoff Communications Staff Survey: Pre-PICU Implementation.

S*T*A*R*T: A Great Handoff

253

questions. In the middle section, an open-ended query asked about suggestions to improve the accuracy and completeness of the handoff report, and the last section addressed the presence or absence of speciﬁc content items such as patient identiﬁers, medical conditions and information, planned diagnostic and testing procedures, and patient care goals.

Presentation to Handoff Givers and Receivers After discussion with medical leadership, the S*T*A*R*T process was introduced sequentially to selected groups. The ﬁrst group comprised handoffs from operating room (OR) or post-anesthesia care unit (PACU) providers to pediatric intensive care unit (PICU) providers, on the assumption that this handoff process is clear-cut, frequent, standardized, repetitive, and likely of high quality. The initial phase of this study addressed OR or PACU to PICU handoffs; by institutional policy, these are attending-to-attending transactions. Although not required, the usual nature of the process is a group-to-group interaction. Unbeknownst to handoff givers, a pre-implementation handoff survey was completed by handoff receivers, evaluating handoffs on a convenience basis. Following this, the S*T*A*R*T process was explained, and then post-implementation surveys were provided. Post-implementation survey forms were identical to pre-implementation surveys except that two questions were added to evaluate whether the S*T*A*R*T process had been used and whether its use was initiated by the handoff giver or receiver. At that point, simultaneous S*T*A*R*T implementation was begun on a typical ‘‘ﬂoor’’ unit, again beginning with a pre-implementation survey.

Challenges and the Process of Seeking Acceptance The concept of using the S*T*A*R*T structure for guidance in articulating a meaningful handoff with appropriate content was not always readily understood, because descriptions of potential handoff content were sometimes mistaken for a template or checklist of required data elements. Some physicians expressed resentment when they thought they were being told what must be included in each handoff. There was also some disagreement about what would comprise the minimum amount of appropriate information. Clariﬁcation of attempted resuscitation status seemed to be the most controversial element, because

254

ELLEN S. DEUTSCH ET AL.

some professionals felt this was relevant for every patient, and others thought that full attempt at resuscitation would be the normal situation and only deviations required explicit deﬁnition. We found that some speciﬁc groups, such as Transport and Day Medicine, had recognized a need and developed their own customized processes to ensure effective handoffs; we reviewed some of these approaches and found them amenable to integration into the S*T*A*R*T structure, providing a standardized approach while supporting local needs deﬁned by sharp end users. In addition, JCAHO does not require a speciﬁc procedure or form, but several ‘‘sharp end’’ users requested a ‘‘prompt,’’ so a laminated card with a design similar to, but smaller than, that of the poster was created and distributed.

RESULTS During the month preceding explanation of the S*T*A*R*T process to handoff givers from the OR, PICU physicians, nurses, and other health care providers receiving handoffs were asked to complete a survey about the quality of handoffs from the OR or PACU. Twenty-eight surveys completed on a convenience basis were received. Subjective impressions of handoff quality were generally positive with good or very good ratings for accuracy 60% of the time (15 of 25 responses) and completeness and opportunity for questions each 54% (14 of 26 responses, Fig. 8). Objective measurements suggested limited objective information was being presented during signoff (Fig. 9); however, caution must be used in interpreting whether the process was sufﬁcient for individual patient care. ‘‘Binary checklists may lead to less valid scores and conclusions’’ (Hawkins et al., 2004) when evaluating complex processes. Several informational sessions were held, addressing the theory and process of the S*T*A*R*T approach to handoffs with physicians and nurses involved as both handoff givers and receivers. Thirteen postimplementation surveys were completed on a convenience basis; S*T*A*R*T was used in 11 (85%) of these handoffs and was initiated by the giver 7 times (54%) and the receiver 3 times (23%; unknown ¼ 3, 23%). There was improvement in each of the subjective and objective measures (Figs. 10 and 11), and 7 of the 11 measured objective parameters were included in at least 80% of the handoffs.

255

S*T*A*R*T: A Great Handoff

14

accuracy

12

completeness

10

opportunity for question

8 6 4 2 0 1

2

accuracy 3

4

5

Fig. 8. Graph of Subjective PICU Pre-Implementation Data, with Subjective Ratings of OR or PACU to PICU Handoff Content on a 5-Point Likert Scale (5 is the Highest Score, n=28).

Identifiers

90

Allergies

80

Pertinent Dx

70

Sig PMH

60

Code s tatus

50

Meds

40

Res p.

30

IV/nutrit

20

Equip

10

Plan/tes ts

0 percent

Goals

Fig. 9. Graph of Objective PICU Pre-Implementation Data, Addressing Objective Content of OR or PACU to PICU Handoff Content (n=28). None of the Parameters Measured Achieved 80% Inclusion.

256

ELLEN S. DEUTSCH ET AL.

4.5 4 3.5 3 2.5 pre post

2 1.5 1 0.5 0 accuracy

completeness

opportunity to ask questions

Fig. 10. Comparison of Subjective Assessments of Handoff Content before and after S*T*A*R*T Education and Implementation, Demonstrating Improvement in Each Parameter.

100% 80% 60%

pre, n=28

40%

post, n=13

20%

Ide

ntif iers Alle rgie s Dia gno ses Sig PM H Me Cod ds es t... Res p. IV/n utri t Equ ip Pla n/te sts Go als S*T *A* R*T

0%

Fig. 11. Comparison of Objective Components of Handoff Content before and after S*T*A*R*T Education and Implementation, Demonstrating Improvement in Each Parameter, with 7 of these Components Included in at least 80% of the Handoffs Surveyed.

DISCUSSION The S*T*A*R*T standardized approach to handoffs provides shared language and expectations and a mutually understood mental model and social construct as well as support for bidirectional communication. This

257

S*T*A*R*T: A Great Handoff

approach was designed to communicate relevant medical information as a knowledge-based process, rather than a checklist method, supporting situational awareness, professionalism, and thoughtful safe care. The pilot implementation of the S*T*A*R*T process resulted in improvement in measured subjective and objective components of handoffs. Further assessment is needed to determine if this improvement is related to education rather than the S*T*A*R*T process itself, whether this improvement is sustained, and whether it can be applied to additional patient care settings. Limitations of this study include small numbers, lack of a validated instrument, and convenience sampling.

CONCLUSIONS Implementation of the S*T*A*R*T approach to handoffs has improved the quality of handoffs of patients being transferred to the PICU. Assessment of additional pre- and post-implementation data will provide information about the acceptability and effectiveness of this approach as well as information that may support process improvement.

ACKNOWLEDGMENTS We would like to recognize the following for their additional help: Loretta Consiglio-Ward, RN, Safety, BJ Clark, MD, Executive Sponsor, Jesse Harris, Transport, Keith Fishlock, RN, Day Medicine, Marguerite Fritsch, RN, NICU, Richard Ermak, RT, Respiratory, Trudy Loge, RN, ED, Jamie Bustard, RN, Patient Care Unit 3E, Tara Bluzard, RN, Patient Care Unit 2A.

REFERENCES Arora, V., Johnson, J., Lovinger, D., Humphrey, H. J., & Meltzer, D. O. (2005). Education and training: Communication failures in patient sign-out and suggestions for improvement – a critical incident analysis. Qual Saf Health Care, 14, 401–407. Berger, C. R., & Bradac, J. J. (1985). Language and social knowledge: The social psychology of language – uncertainty in interpersonal relations. London: Edward Arnold. Endsley, M. R. (1995). Toward a theory of situation awareness in dynamic systems. Human Factors, 37, 32–64. Frank, G., Lawler, L. A., Jackson, A. A., Steinberg, T. H., & Lawless, S. T. (2005a). Resident miscommunication: Accuracy of the resident sign-out sheet. Journal of Healthcare Quality, 27W2–10–W2–14.

258

ELLEN S. DEUTSCH ET AL.

Frank, G., Lawless, S. T., & Steinberg, T. H. (2005b). Improving physician communication through an automated, integrated sign-out system. Journal of Healthcare Information Management, 19, 68–74. Hawkins, R., Gaglione, M. M., LaDuca, T., Leung, C., Sample, L., Gliva-McConvey, G., Liston, W., De Champlain, A., & Ciccone, A. (2004). Assessment of patient management skills and clinical skills of practising doctors using computer-based case simulations and standardised patients. Medical Education, 38, 958–968. Hohenhaus, S., Powel, S., & Hohenhaus, J. T. (2006). Enhancing patient safety during handoffs: Standardized communication and teamwork using the ‘‘SBAR’’ method. American Journal of Nursing, 106, 72A–72C. Horwitz, L. I., Krumholz, H. M., Green, M. L., & Huot, S. J. (2006). Transfers of patient care between house staff on internal medicine wards: a national survey. Archives of Internal Medicine, 166, 1173–1177. Joint Commission on Accreditation of Healthcare Organizations (JCAHO). (2006a). FAQs for the 2006 National Patient Safety Goals. Retrieved on September 1, from http:// www.jointcommission.org/NR/rdonlyres/25E48E23-6946-43E4-916C-65E116960FD5/ 0/06_npsg_faq2.pdf Joint Commission on Accreditation of Healthcare Organizations (JCAHO). (2006b). Root causes of sentinel events. Retrieved on September 1, from http://www.jointcommission.org/NR/rdonlyres/FA465646-5F5F-4543-AC8F-E8AF6571E372/0/root_cause_se.jpg Joint Commission on Accreditation of Healthcare Organizations (JCAHO). (2007). 2007 hospital/critical access hospital National Patient Safety Goals. Retrieved on September 1, 2006, from http://www.jointcommission.org/PatientSafety/NationalPatientSafetyGoals/ 07_hap_cah_npsgs.htm Mukherjee, S. (2004). Becoming a physician: a precarious exchange. New England Journal of Medicine, 351, 1822–1824. Patterson, E. S., & Woods, D. D. (2001). Shift changes, updates, and the on-call architecture in space shuttle mission control. Comput Support Coop Work, 10, 317–346. Rasmussen, J. (1983). Skills, rules, and knowledge – signals, signs, and symbols, and other distinctions in human-performance models. IEEE Transactions on Systems, Man and Cybernetics, 13, 257–266. Reason, J. (2004). Managing the risks of organizational accidents. Hants, England: Ashgate Publishing Limited. Schramm, W. (1954). How communication works. In: W. Schramm (Ed.), The process and effects of mass communication. Urbana, IL: University of Illinois Press. Shannon, C. E. (1948). A mathematical theory of communication. The Bell System Technical Journal, 27, 379–423, 623–656. Solet, D. J., Norvell, J. M., Rutan, G. H., Rutan, G. H., & Frankel, R. M. (2005). Lost in translation: Challenges and opportunities in physician-to-physician communication during patient handoffs. Academic Medicine, 80, 1094–1099. Volpp, K. G. M., & Grande, D. (2003). Residents’ suggestions for reducing errors in teaching hospitals. New England Journal of Medicine, 348, 851–855.

CHANGING A PEDIATRIC SUB-ACUTE FACILITY TO INCREASE SAFETY AND RELIABILITY Daved W. van Stralen, Racquel M. Calderon, Jeff F. Lewis and Karlene H. Roberts ABSTRACT This chapter describes the efforts of a team of health care workers to make a sub-acute health care facility (SCF) serving profoundly damaged children into a high reliability organization (HRO). To obtain this goal, the health care team implemented change in four behavioral areas: (1) risk awareness and acknowledgment; (2) deﬁning care; (3) how to think and make decisions; and (4) information ﬂow. The team focused on ﬁve reliability enhancement issues that emerged from previous research on banking institutions: (1) process auditing; (2) the reward system; (3) quality degradation; (4) risk awareness and acknowledgment; and (5) command and control. These HRO processes emerged from the change effort. Three additional HRO processes also emerged: high trust, and building a high reliability culture based on values and on beliefs. This case demonstrates that HRO processes can reduce costs, improve safety, and aid in developing new markets. Other experiences in Patient Safety and Health Care Management Advances in Health Care Management, Volume 7, 259–282 Copyright r 2008 by Emerald Group Publishing Limited All rights of reproduction in any form reserved ISSN: 1474-8231/doi:10.1016/S1474-8231(08)07012-2

259

260

DAVED W. VAN STRALEN ET AL.

implementing high reliability processes show that each organization must tailor make processes to its own situation (e.g. BP, U.S. Chemical Safety and Hazards Board, Federal Aviation Administration, U.S. Navy Aviation Program, and Kaiser Permanente Health Care System). Just as in the ﬂexibility called for in organizing for high reliability operations, ﬂexibility is called for in deciding which HRO processes work in speciﬁc situations.

For a number of years researchers and practitioners have been interested in understanding how organizations which must operate nearly ﬂawlessly, because errors in them can result in catastrophic consequences, do so. Those organizations in this set which succeed were labeled High Reliability Organizations (HROs) (Weick, 1987; Rochlin, La Porte, & Roberts, 1987). Roberts (1990) deﬁnes an HRO as an organization conducting relatively error-free operations, over a long period of time, and making consistently good decisions resulting in high quality and reliability operations. There are many examples of organizations which should be HROs in the health care industry. For example, recently Cedars Sinai Hospital, a hospital usually lauded for its clinical quality, injected actor Dennis Quaid’s infant twins with doses of Heparin more than 1,000 times larger than normal. Assuming health care HROs perform as well as early research showed other industries could perform (e.g. Weick & Roberts, 1993; La Porte & Consolini, 1991; Eisenhardt, 1993), and for the reasons described in that research, one question remains: How does one create an HRO? Or, if the organization already exists, how does one transform it from an ordinary organization into an HRO? This chapter attempts to answer the second question. It does so through a case study led by the chapter’s ﬁrst author. For the last few years researchers and practitioners interested in implementing HRO concepts have met in workshops, ﬁrst in Southern California and then in Europe. The questions above were the center of discussion in all the workshops. Today there are various attempts at HRO implementation running across a variety of industries including health care (e.g. education, ﬁnance, military, commercial aviation, the military, and NASA). One of the problems with implementation is that there are now a plethora of constructs about behavioral processes in HROs. HRO researchers say implementation programs must be tailor made to the settings in which they are applied. But they have yet to sort out which processes are best suited to which situations. This study did not review the growing literature on HROs

Changing a Pediatric Sub-Acute Facility

261

and select from this literature a set of processes. Instead it went back to an early set of processes developed by Carolyn Libuser (1994) for the banking industry. It did this because many managers said these processes can be implemented and some have successfully implemented them. In addition, an assessment device based on the Libuser model is available. Thus, an organization can track how it is doing in maintaining high reliability (Gaba, Singer, Sinaiko, Bowen, & Ciavarelli, 2003). Then, too, other processes identiﬁed in the HRO literature seem more difﬁcult to implement (e.g. sensemaking and improvisation). The Libuser study provided an initial conceptual lens to guide this effort. However, several additional high reliability principles emerged from the implementation process. The study is important in that it highlights the ability to implement a set of processes in an organization that needed to become an HRO, and it points out some of the difﬁculties in doing so. First we will describe Libuser’s guiding principles, then the setting, the behaviors focused on for change, the outcomes, and, ﬁnally, we provide some conclusions from this activity.

THE GUIDING LENS Libuser’s model consists of ﬁve processes: (1) process auditing, (2) reward system, (3) quality degradation, (4) perception of risk, and (5) command and control. Each of these processes is described below.

Process Auditing An established system of ongoing checks designed to spot expected as well as unexpected safety problems. Safety drills and equipment testing are included in this category. Follow-ups on problems revealed by prior audits are also important.

Reward System The reward system is the payoff an individual or organization receives for behaving one way or another. Reward systems in organizations tend to have powerful inﬂuences on behavior of people in them. Inter-organizational reward systems also inﬂuence the behavior of organizations.

262

DAVED W. VAN STRALEN ET AL.

Quality Degradation Organizations must struggle to avoid degrading quality. Usually some referent system is perceived to be the gold standard in this area and all other organizations attempt to reach the quality levels of the referent. Perception of Risk There are two elements of risk perception: (a) whether the organization knows risk exists, and (b) the extent to which steps are taken to acknowledge and minimize it. Command and Control Libuser borrows this notion from previous research and highlights four key elements. 1. Migrating decision making. The person with the most expertise, not the one with the most chevrons, makes the decision. 2. Redundancy. People and hardware provide back up systems. 3. Senior managers who can see the big picture. Managers do not micromanage. 4. Formal rules and procedures. The existence of hierarchy, but not bureaucracy.

THE SETTING In 1995 a pediatric nursing home that cares for severely damaged children received poor ratings from the state licensing agency and had a poor reputation in the local medical community. The state banned new admissions for several years and applied a high level of scrutiny to the nursing home’s routine reports. It also reduced the number of beds for this level of service to about half of the facility’s capacity (from about 60–30 beds). This ban lasted from mid-1995 well into 1998. During this period, the state also refused to increase payment rates to offset increased costs. The medical community (physicians and nurses) and the local Emergency Medical Services (emergency medical services, ﬁre, and paramedic services)

Changing a Pediatric Sub-Acute Facility

263

believed the nursing home used poorly trained staff members and provided substandard care. The nursing home found itself in a destructive situation because its reputation precluded it from seeking quality staff members to assist in a turn-around; the admissions ban and refusal to increase reimbursement did not allow for consultants or major changes in the program; and adequate time for a change in trajectory was uncertain before events led to closure. A medical director, contracted through a nearby medical school, brought a strategy for change based on the Libuser research ﬁndings. These processes are the subject of this study. About the same time these problems occurred, the nursing home changed its license to a pediatric sub-acute care facility (SCF). Children with chronic illness increasingly contribute to the census of pediatric intensive care units (PICUs) (Briassoulis, Filippou, Natsi, Mavrikiou, & Hatzis, 2004), while children with chronic ventilator needs have increased by nearly threefold in 15 years in one state (Graham, Fleegler, & Robinson, 2007). Sub-acute facilities reduce the burden that long-term ventilator patients place on the resources of an ICU (Lindsay, Bijwadia, Schauer, & Rozich, 2004), and provide care for medically complex, non-communicative, severely disabled children who suffer from severe central nervous system (CNS) diseases such as profound mental retardation, severe developmental delay, and persistent vegetative state. These children typically have limited mobility, which leads to further complications such as scoliosis, contractures, decubitus ulcers, osteoporosis, and fragility fractures. The state regulatory agency deﬁnes sub-acute care (for purposes of Medicaid reimbursement) as dependence on two or more medical technologies. For example, many children in sub-acute facilities have both tracheostomy and gastrostomy tubes. Importantly, their reliance on technology and their medical fragility can result in sudden physiologic destabilization and death. The facility was staffed by certiﬁed nursing assistants (CNAs) and licensed vocational nurses (LVNs), with one supervising registered nurse (RN). Several respiratory care practitioners (RCPs) provided respiratory care and managed the mechanical ventilators, which were the type used in private homes. The facility desired to improve the level of care provided to residents, rehabilitate its image to licensing agencies and medical professionals, expand its market, and increase the level of clinical services offered to residents. Without these changes, the facility would continue to struggle ﬁnancially, and possibly fail.

264

DAVED W. VAN STRALEN ET AL.

THE CHANGE To implement change, the administrative team focused on four areas of behavior: (1) risk awareness and acknowledgment, (2) deﬁning care, (3) how to think and make decisions, and (4) information ﬂow. In varying degrees, Libuser’s ﬁve processes (process auditing, reward system, quality degradation, perception of risk, and command and control) inform each of these four areas of behavior.

Risk Awareness and Acknowledgment Hospitals’ risk proﬁles are quantitatively and qualitatively different from nursing homes. Hospital risks arise from acute or critical illness that can lead to rapid physiological destabilization affecting the respiratory, cardiovascular, or neurological systems. Nursing home risks are generally due to complications of care such as falls, dehydration, or aspiration of oral secretions or gastric contents into the airway. Time dependence for action in hospital operations is in minutes (critical care) or hours (acute care); in contrast, time dependence is in hours or days for the typical nursing home. As a result, nursing home work cultures rarely incorporate processes for clinically engaging and responding to dynamic physiological dysfunctions. Instead, at the earliest sign of such dysfunction, a nursing home will typically transfer the client to an acute care hospital, usually through the emergency medical services 911 system. Interestingly, these different time dependencies are symbolically captured by the way these institutions refer to their primary customers. Hospitals refer to patients who receive treatment, while nursing homes refer to residents who reside at the facility. Differences in hospitals’ versus nursing homes’ time demands and dependencies lead to staff self-selection paradigms. Staff members elect to work in acute care hospitals for the greater variety of patient situations, increased responsibility for care decisions, and short-term relationships with patients. Hospitals generally use RNs to provide care. Staff members who migrate to nursing homes generally prefer low-tempo work conditions and building long term, connected relationships with residents. Sub-acute facilities may be viewed as a hybrid of intensive care, acute hospital care, and nursing home care where residents with chronic, stable illness reside, but may abruptly deteriorate. Given the prevailing nursing home culture, it is counter-intuitive that the typical long-term care staff members would become proﬁcient at the tasks required to care for such children.

Changing a Pediatric Sub-Acute Facility

265

Hence, without differently trained and motivated staff members, subacute nursing home children would quickly lose their chances at life. Educational programs do not prepare nursing home medical caregivers to engage in unexpected high-risk problems. As noted previously, when faced with a sudden medical emergency, the typical nursing home caregiver calls for emergency services through the 911 system. Moreover, when faced with a vague or ambiguous medical condition, nursing home staff members will call the physician for a ‘‘change of condition,’’ which usually leads to the physician referring the client to the 911 system without further treatment. The ﬁrst objective, then, was to have staff members identify the high-risk nature of the environment by identifying their internalized beliefs. The medical director began asking staff members if the children were in danger. Invariably, a CNA, LVN, or RCP answered, ‘‘No this is a nursing home.’’ After a series of such answers, the medical director invited the staff to a picnic in the parking lot for several hours. No one accepted the invitation because they all believed that if the children were left alone, one of them would die. In subsequent discussions, each staff member identiﬁed several ways children would die if vigilant care was not present. Staff members began identifying the more obvious causes such as tracheostomy tube dislodgement, falls from the bed, and fever. With time, they began to discuss these obvious risks in terms of early heralds of events, such as the active child who may pull at the tracheostomy tubing or the child who, with strengthening arms, has the strength to pull to a sitting position and may now fall over the rail. These discussions became the basis for developing the facility as a HRO. Conversation was now opened to discuss both client risks and the expertise the staff had developed to identify early subtle signals and engage in problem solving. Early attention to problems allowed treatment when a child’s disease was more amenable to therapy, with fewer and less severe complications. Staff members began to understand that they provided highrisk care in a medically austere environment, with special characteristics. To elaborate, in nursing homes, staff members have inherent difﬁculties interpreting responses from non-communicative residents. They are taught to use physiologic ﬁndings rather than any diagnosis, as complications frequently develop from the synergy between diagnoses or between physiologic systems. Collaboration among staff members helped interpret these signs. For example, the LVN might call an emergency for a respiratory problem. After evaluation, the RCP might recommend a rectal suppository to assist the client’s bowel movement. After the bowel movement the client returned

266

DAVED W. VAN STRALEN ET AL.

to the pre-emergency state. The RCP identiﬁed a non-pulmonary cause of respiratory distress caused by the Val Salva maneuver required for the bowel movement. Upward abdominal pressure on the weak diaphragm compressed the chest cavity which decreased chest volume and airway compliance causing patient-ventilator asynchrony. Collaboration between the RCP for evaluation of the respiratory system and the LVN for treatment of the gastrointestinal system, along with recognition by both members that one specialty’s problem might appear in the other specialty’s area, solved the problem and further increased risk awareness of the interactions among physiological systems. Once risks are identiﬁed, staff and the organization’s leadership must acknowledge them through policies, procedures, and education. State regulation does not allow the facility to use protocols for treatment, as protocols are plans for carrying out a patient’s treatment regimen before contacting a physician. This limits the facility’s planning and places more reliance on staff member judgment and actions. Because of limited resources, staff could not rely on technology to assess the patient or provide care. Clinical assessment at the bedside was an important factor for treating patients. This involved use of and trust in staff members assessment skills each day to learn to identify normal signs for the patients, to predict the direction of change (deterioration or improvement), and further signs to look for in such situations. Staff members relied heavily on identiﬁcation of response to therapy, followed by interpretation of those responses. For staff members, the greatest threat to identifying and engaging risk is unrecognized fear. The physiologic fear responses are the classic adrenalinemediated ﬁght and ﬂight responses in the sympathetic nervous system and the cortisol-mediated freeze response (Kalin, 1993). Unrecognized fear can lead to the coning of attention during an emergency, to rapid and unthinking reactions to unfolding events, or to unthinking inaction as events occur. Coning of attention occurs when a caregiver maintains tight attention to one aspect of a problem. Importantly, this is not the same as maintaining focus, which allows for the processing of new information. Staff members were taught to recognize these fear responses in themselves and others. Fight manifests as anger, and staff members learned to treat the angry parent, outside caregiver, or employee as afraid and then try to identify what triggered the fear response. This involved giving people an action or easily obtained objective to help them bring a sense of control to the situation. For most staff members, ﬂight manifests itself as plausible avoidance, such as ﬁlling out unneeded forms or evaluating a stable patient

Changing a Pediatric Sub-Acute Facility

267

to avoid participating in the emergency. Mutual support and easy actions that one could quickly complete helped bring in or return the individual to the team. Freeze responses often manifest as confusion or inability to recall information. Staff members were instructed to return to a previously accomplished action, which helped clear the fog and bring staff members back to the situation. All of these processes facilitated staff members’ appropriate awareness of risk. In addition, staff members and supervisors were taught situational awareness (Endsley, 1995) where both the bedside caregiver and the clinical manager see the big picture. Situational awareness allows the leader to capture migrating decisions and integrate them into actions. The leader of the emergency does not micromanage. To avoid micro management, the supervisors stood back during STAT (statim – a medical emergency situation requiring immediate action) responses to observe both the medical condition of the client and the behavior of the caregivers. The charge nurse or designated leader of the STAT team leads by managing the team and asking questions. The right questions come from observing the client, the responses from caregivers, and the functioning of the team. The charge nurse does not assume the role of bedside nurse, but oversees and manages the interaction between the caregivers and the client. To develop continued attention to risk awareness, managers shared true, personal stories of patients who appeared to be doing ﬁne, yet suddenly and rapidly deteriorated or developed profuse bleeding. These deteriorations could occur despite the absence of early detectable signs or symptoms. Stories about children who pull their own tracheostomy tubes out, lose their airway, and approach permanent irreversible damage were also used to provide examples of how staff members needed to be aware of and acknowledge client risks.

Deﬁning Care Developing and using technologies to maintain life for profoundly disabled children is in a nascent state. In contrast, general pediatricians provide care to handicapped children with straightforward technologies, such as tracheostomy and gastrostomy tubes. As the complexities of a child’s disability increase, so do the number of potential complications. For example, ventilation dramatically increases both technological complexity and residents’ risk to die. As the degree of disability and dependence on

268

DAVED W. VAN STRALEN ET AL.

technology increases, the interactions between managing technological complexity and the attendant risks to residents can become deadly. The medical director came to the facility without experience in the nursing home ﬁeld or in long-term care for the profoundly disabled. The executive group wanted growth in census by direct marketing to physicians who have the target patient in their practice or to acute care hospitals. A conﬂict developed between deﬁning characteristics that described children who would beneﬁt from facility care and the type of care the facility could offer. For successful marketing, the facility should provide care that families or other facilities cannot. The facility also had to identify the sources of dissatisfaction from the state and methods to address the problems in an efﬁcient and effective manner. It seemed that, before the facility could improve care in the judgment of the state, care had to be deﬁned and beneﬁts to the child described. The deﬁned care given by the facility staff member might not be what it appears at ﬁrst view. The new medical director recalled a question an experienced ﬁre ﬁghter asked his ﬁreﬁghters, ‘‘What do we do in the ﬁre department?’’ After ﬁreﬁghters made various attempts to answer, all related to ﬁre suppression and rescue work, he answered, ‘‘We solve problems citizens cannot or will not solve themselves.’’ For sub-acute care, when asked, ‘‘What do we do?’’ we had to search deeper for the answer than ‘‘provide nursing home care to profoundly disabled children.’’ The answer came, not from ‘‘Who do we treat?’’ or ‘‘What do we do?’’ but from ‘‘What do we say?’’ and ‘‘What is the response to our actions?’’ Clear, unambiguous descriptions and commands were needed in a culture that readily used slang and jargon as a part of belonging and vagueness and ambiguity as self protection. In high-risk environments where people can die, this use of obtuse terminology leads to deadly incidents. When the physician was off-site and an occasional unstable situation arose, decisions had to migrate to the person with the expertise to decide, sometimes up the chain of command, but more often down the chain of command to the bedside caregiver. Staff members learned to articulate the situation in a clear, concise, objective manner and without slang or jargon. Observation and interpretation were separated, as the medical director and staff members discussed care in post-emergency critiques or on clinical rounds. They identiﬁed straightforward interventions that could be applied immediately and were within the scope of practice of the caregiver. An important aspect of this process was identifying when an intervention failed, possible contributions to failure, and means for identifying successful interventions.

Changing a Pediatric Sub-Acute Facility

269

Deﬁning care also was effected through staff members’ efforts to improve safety and by their identiﬁcation of threatening procedures and effective treatments. Across a series of actions, the facility’s objectives advanced from use of ventilators to calm children to use of ventilators to enhance their lives. One winter, a hospital PICU had a full census and returned a child to the facility for weaning from the ventilator, which was accomplished. Within the month, the same hospital, because of a full census, could not accept a client in acute respiratory failure. In both cases, the sub-acute facility provided ventilator services without the laboratory or pharmacy services found in a hospital. Subsequent to that incident, and through discussion during client rounds, the facility managers and the medical director set as an over-arching objective to keep the technology in the background to enable the child to live. This practice, too, became a deﬁnition of care. Continued difﬁculties in transferring children dependent on ventilators to the PICU led the care team to develop new models of ventilator use for these residents. The team developed the model of adjusting the ventilator to calm the child rather than using drugs for that purpose. Asynchrony between the child and ventilators was considered a medical emergency that should receive immediate attention by the RCP. This occurred through ventilation by hand, with a self-inﬂating resuscitator bag using high rates but shallow tidal volumes or low respiratory rates and large tidal volumes. Once the RCP achieved a calm child, ventilator settings were adjusted and the medical director was notiﬁed and discussed the situation. Once it became clear that this approach produced ventilator synchrony, a search developed for speciﬁc interventions that most calmed children on ventilators. RCPs adjusted ventilator pressure for visible chest expansion. During patient agitation episodes, RCPs found that ventilator inspiratory times faster than those used in the PICU produced longer periods between agitation spells. Of all the interventions, the one that worked reliably and with greater permanence was increasing respiratory rates. As a standard practice, intensive care physicians prefer ventilator rates below 20 breaths per minute (bpm) to ensure safety from stacking breaths (too rapid respirations that lead to incomplete exhalation and chest hyper-expansion) and subsequent hypoventilation, agitation, and possible pneumothorax that could result in death. Also, lower rates reduce the risk of apnea if the ventilator becomes disconnected from the patient. Because child-ventilator asynchrony continued to occur and the team could not be sure why or if higher rates kept the child calm, the team held

270

DAVED W. VAN STRALEN ET AL.

long discussions during client care rounds with all staff. It seemed children responded best to ventilator rates between 20–30 bpm. After two months discussing these counter-intuitive ﬁndings that higher ventilator rates calmed children, everyone noticed that the children were more awake and alert, and began to smile, play, and laugh. The clinical care team concluded that, despite normal blood gas ﬁndings, these children had the sensation of suffocation when ventilator rates were below 20 bpm, which prevented them from crying, smiling, or laughing. The institutional objective became one of using ventilators for relief of suffocation and to produce calm, smiling children. This change in the model occurred because of decision migration and deference to expertise, the expert being the bedside caregivers. While developing this evolving model of sub-acute care, the management team began a program to demonstrate to the state, through the peer review process, that the facility provided quality medical care. Presentation of research material directed the clinical team to articulate their work and the processes they used and opened their approach to a limited form of peer review. Within one year, with support from the local School of Public Health and the facility administrative staff members, the facility presented more than 20 research posters at several national conferences and one international conference. Articulating ideas without hidden assumptions; questioning actions and assumptions; submission of ideas to peer review and criticism; and participation of bedside staff members in improving care all, allowed staff members and outsiders to see that the facility could improve care. Along the way, all staff members further deﬁned what care meant. Critical to the introduction of intensive care techniques to a nursing home was the support and insight of the general pediatricians on staff. With openmindedness, the general pediatricians incorporated intensive care techniques into general pediatrics, such as the clinical identiﬁcation of hypovolemia using tachycardia, prolonged capillary reﬁll, and cool limbs. The general pediatricians also contributed ideas on how to use the ventilators for development issues, such as learning to walk while attached to the ventilator. This care developed from interaction among intensive care physicians, general pediatricians, nursing, respiratory care, and administrative personnel. The facility care teams now follow a pyramid of care, with technology supporting residents at the bottom of the pyramid. By making the technology invisible, the bedside caregivers can address medical issues. This invisibility allows the children to grow, thrive, smile, play, and laugh.

Changing a Pediatric Sub-Acute Facility

271

How to Think and Make Decisions Medical culture does not easily allow decision migration down the hierarchy to those with less medical education. One way this can happen is through the use of protocols, preplanned medical treatments, approved beforehand for use within the scope of practice of allied health practitioners. This works well with deterministic medical situations where the diagnosis or clinical ﬁndings determine an intervention and treatment will not have serious complications. A consequence of this medical culture is a lack of knowledge and experience about how to make independent decisions, particularly in situations in which limited facts are available. Importantly, physicians are not always in the sub-acute facility and a nurse practitioner or physician’s assistant works only a routine workweek, leaving staff members working both weekends and night shifts responsible for emergency decisions. As the level of service increased for complex ventilator problems, the facility managers found that staff members (CNAs and LVNs) did not have knowledge about how to make emergency decisions. Staff members also refused to make decisions because they believed, with good reason, that supervisors and administrators would hold them accountable for undesired or bad outcomes. To help overcome this resistance, the medical director and RCP manager began a program to teach decision making while on clinical rounds. They understood that developing thinking and decision making is necessary to manage ventilator care without a physician on site. For example, the medical director elicited solutions to problems during clinical rounds, and regardless of the answer, the physician made it ﬁt either by adding necessary facts or presenting reasons that the answer appeared right. He taught the phrase, ‘‘Every decision is the right decision, one that I would make.’’ In addition, ﬁne-tuning after each clinical decision was made without the presence of the physician. In a sub-acute facility, a major impediment for staff members is reserving the use of emergency decision-making techniques only for emergencies. The problem, as noted previously, is that of identifying an emergency in its early, latent phase. The guiding principle became, ‘‘What you do everyday is what you do in an emergency.’’ For this purpose, John Boyd’s OODA Loop (Hammond, 2001; Coram, 2002) was a helpful aid in structuring the decision process. Col. Boyd developed the OODA loop in response to increased US losses of aircraft in aerial combat during the war in Vietnam. It increases the speed of decision cycles for ﬁghter pilots to outmaneuver their opponents. It operates with minimal or incomplete information about the situation.

272

DAVED W. VAN STRALEN ET AL.

The OODA Loop is an iterative set of decision processes of observing, orienting, deciding, and acting. Within the sub-acute setting, observing entails: acquiring sufﬁcient knowledge for a clinical decision and no more by perceiving the unfolding of clinical circumstances; incorporating outside information; and drawing on professional training and experience for implicit guidance and control. Next, orienting involves the synthesis and analysis of the cultural beliefs and genetic heritage of the client caregiver; of previous experience with the client and others; of new information about the client or setting; of the medical facility’s values, beliefs, and behaviors; and of human factors and performance decrements. Then, deciding requires staff members to hypothesize about the client’s condition and how best to respond to it. Lastly, acting tests this hypothesis. Because the OODA Loop is iterative, the staff member then observes the results of acting, and so on. The OODA Loop helps staff members engage in an emergent, problemsolving interaction with the environment. Feedback occurs from the Decide and Act functions to the Observe function. Feed forward occurs from the Observe to Orient and from Orient to Decide functions. For example, the OODA Loop allows the RCP to identify interventions for initiation of mechanical ventilation. These interventions have the objective of calming the client, while ensuring good chest expansion without breath stacking. The OODA loop model allows rapid interventions to mitigate threat in time-dependent situations. Actions begin without dependence on unavailable resources or loss of time. Disadvantages include its counter-intuitive nature and seemingly high risk. It requires distributed decision making, where all members have this knowledge and skill and it runs counter to the medical culture in which the physician is the central decision maker. This program did not come easily to the RCPs, who were not accustomed to presenting a patient to a physician for discussion, making decisions in public, or discussing events that could go wrong. For example, after clinical rounds were ﬁnished, RCPs became upset that their suggestions were not heard. They believed they should not say anything during clinical rounds if their suggestions were not accepted. There came a time when the RCPs made a pact that no one would make suggestions during clinical rounds, but only provide the information the physician requested and answer his questions. The respiratory manager observed this tension within the group and began work with the RCP staff with particular focus on the change in behavior, where they previously made suggestions and participated in discussions. The respiratory manager worked with staff on an individual and group basis to identify how different ideas can achieve the same goal. In a group

Changing a Pediatric Sub-Acute Facility

273

setting, the RCPs became more observant that each had a different method for reaching similar outcomes. Afterwards, the respiratory manager explained that this is what the physician had taught during clinical rounds. Some of the more insecure RCP staff members were hesitant to enter the discussion on clinical rounds; however, with individual attention and support beforehand from the physician, they would join the discussion, and began making signiﬁcant contributions to care. Those who were not forthcoming in taking credit for their ideas were singled out and told that their ideas were valuable. This was a slow process from no suggestions or participation in clinical rounds toward more in-depth discussion of what each individual and the team could do to improve.

Information Flow Communication up and down the chain of command was initially vague and ambiguous. To ensure multi-directional information ﬂow, the medical director encouraged the use of articulate, objective, but succinct presentations. Bi-directional communication consists of concise requests for information and instructive material from up the hierarchy and clear expressions of what is observed in response to therapy from down the hierarchy. Because they are at the bedside, facility staff members identify early heralds of deterioration, interpret the ﬁndings in context, and translate those ﬁndings to other staff members and physicians. These staff members learned the importance of accurately capturing data by using the data. For example, during client visits, the physician used all available records, particularly CNA and RCP records. Because the latter records had not been previously reviewed, they often included shortcuts. When staff members observed physicians or managers reviewing bedside records, the comprehensive quality of these records increased. Routine clinical discussions pointed to the ‘‘ignorance in medicine,’’ and staff came to realize much of what is assumed in medicine is either not certain or not known. For example, physicians cannot explain why some bacteria that infect lungs do not infect connective tissue. This openness by the medical director and other physicians facilitated discussions about uncertainty, which both led to a research program and developed a learning environment. Because of this environment, managers and physicians also admitted uncertainty or saying ‘‘I don’t know,’’ which lead to further evaluation of a client’s clinical situation. Staff members now had an

274

DAVED W. VAN STRALEN ET AL.

important part in client care. During clinical rounds, staff members also discussed the danger of not saying, ‘‘I don’t know.’’ After repeated use, it became easy to say and brought all staff members into evaluations and discussions. Decisions would sometimes need to be made based on the caregiver’s assessment before a conﬁrmatory examination by the physician. After the examination, reﬁnements in the decision would bring the decision closer to what it should be. Non-emergency decisions were discussed before execution to help the caregiver learn to think and decide, a major contribution to a high trust environment. Information ﬂow in this manner produced less sense of isolation for staff members when working difﬁcult situations. With the focus on individual accomplishments, staff members became more open and willing to ask questions about improvements they could make, wanting to learn different approaches. Criticism directed toward any staff member became a sign of system failure. The facility willingly supported staff members who showed interest in growth, giving opportunities to use newly learned tools and advance their education. It is now common for CNAs to become LVNs and LVNs to become RNs while remaining in the facility’s system. A major difﬁculty for information ﬂow to enable decision making came from the methods medical caregivers use for presentation of a patient. Nurses tend to present the patient system-by-system with problems and treatments discussed each step of the way. They presented the respiratory system problem, evaluation, and treatments, and then moved to the circulatory system. RCPs discussed the respiratory system by chronic or active processes and the ordered treatments or those they were requesting. Physicians presented patients in a systemic manner of all information ﬁrst followed by an assessment or diagnosis and ending in a plan. This included subjective ﬁndings communicated by the patient, and objective ﬁndings of the physical examination, laboratory, and radiological ﬁndings. A plan was developed only after all information was identiﬁed and discussed. Conﬂict in presentations occurred when bedside staff reported incomplete ﬁndings and offered suggestions or requested treatment orders. Recognition of this paved the way for appropriate use of each model: the RCP activeprocess approach during an emergency, the nursing systemic approach when an acute problem developed, and the physician review-all-data approach for longer-term plans. An example of rapid information ﬂow the facility strived to emulate the ﬁre team arriving at a ﬁre. The ﬁrst arriving unit provides a rapid and accurate evaluation of the ﬁre called a ‘‘size up.’’ Similarly, in the 1970s,

Changing a Pediatric Sub-Acute Facility

275

before algorithms and protocols were used, paramedics functioned as the ‘‘eyes, ears, and hands’’ of the physician and presented an articulate, objective, succinct patient evaluation. The facility clinical team developed a similar presentation style through the use of clinical ventilator rounds with the RCP presenting the patient and developing a plan for treatment. All discussions adhered rigorously to an articulate, succinct, objective presentation. The physician always let the ﬁnal plan come from the RCP or LVN. The team then learned to identify what is important, necessary, and sufﬁcient to make a decision.

BUILDING AN HRO Most sub-acute facility’s problems with the state licensing agency occurred from uncommon but high-risk events whereas problems with the medical reputation developed from frequent, low-impact encounters between the referral hospital staff or emergency medical service providers, and the facility patients and staff members. The existing medical culture impeded improvements in facility medical care and repair of its reputation. The dominant medical culture represents a deterministic system where success comes from proper diagnosis and the application of indicated therapies. The deterministic medical system model uses central command applied through a vertical hierarchy, limited questioning of authority, belief that authorities have solutions, and obedience. The sub-acute facility was often in an unfolding and stochastic setting, in which much is unknown and things constantly change. This required culture change from the predominant medical and nursing home cultures to a more ﬂexible and less brittle culture. The ﬁnal question is whether the processes in the Libuser model are reﬂected in the four areas of change focused on by the change agents. We expand our analysis by asking whether additional or different processes evolved in the course of this intervention.

Process Auditing The numerous discussions that characterized this effort offer opportunities for audits. For example, the staff discussed risk to children. That helped them measure, in some informal way, risks to children. Recognition of fear responses reduces coning and allows people broader views of their situations. Broader views allow people to see more of the total picture

276

DAVED W. VAN STRALEN ET AL.

from which they can develop an accurate audit scheme. The continuous search for deﬁnitions of care is an audit process. Driving decision making down the hierarchy creates potentials for redundancy and offers the checks and balances one needs in an audit system. Finally intense review of bedside records makes people create more accurate records and improves the audit system.

Reward System Rewards and punishments guide behaviors in individuals and organizations. The clinical management team focused on rewards rather than punishments because of their greater impact on behavior (e.g. Kerr, 1975). By focusing on inclusiveness and internalization of values, the team worked to develop a setting in which appropriate rewards were attached to desired behaviors. For example, having greater inﬂuence in choosing therapies rewards those who openly discuss their thought processes.

Quality Degradation To avoid quality degradation, the management team regularly and openly discussed failures and near misses with focus on the thought processes used by bedside care members. The leaders studied small failures as clues to emerging or evolving large system failures. Caregivers constantly discussed whether individuals or teams missed something important. These discussions occurred in anticipation of events (proactive), and during review of care (retroactive). The leaders found the greatest utility of discussions during the event with interactive, real-time risk assessment.

Risk Awareness and Risk Acknowledgment Risk awareness alone does not lead to reliability; it must be followed by acknowledging that something needs to be done. Clinical discussions helped staff members link risk with clinical interventions. Risk lies in the situation either as probability or possibility. The medical director used the word possibility when working with vague risk and great threat. Education focused on early heralds of deterioration, physiologic time-course of diseases, and therapies available in the facility or the Emergency

Changing a Pediatric Sub-Acute Facility

277

Department and Intensive Care Unit. Strong responses to these weak signals facilitated engagement of the problem when interventions are most effective and have the least number of side effects or complications. Early heralds of deterioration were problematic as they tend to be vague and difﬁcult to interpret. For example, hypoxemia may indicate thick secretions that will respond to a ﬂuid bolus through the gastrostomy tube or it could indicate early pneumonia. Bleeding through the tracheostomy tube may indicate problems such as trauma from suctioning, tracheitis, papilomatosis, bronchiectasis, or fatal innominate artery hemorrhage. Early heralds also tend to be ambiguous, not clearly indicating which system has the pathology. Nasal ﬂaring, tachypnea, and tachycardia develop from early respiratory disease and have also indicated positional pain relieved by repositioning, or gastrointestinal disease such as ileus or constipation. Transfer by emergency services occurs when a disease state accelerates or the remaining therapies available to a nursing home are limited. As long as the child responds to interventions in a timely manner the facility did not refer the child to the acute care hospital. Non-responders (those patients who were not responding to interventions) were transferred. The team always thought and observed ahead of the problem to identify and know their limits.

Command and Control Decision migration. Decisions are hypotheses one acts on. In uncertain environments the individual with the most up-to-date knowledge of the circumstances is frequently the person at the environmental interface, yet the person with the greatest knowledge of principles and concepts may be the individual higher in the hierarchy and physically away from the situation. Use of decision processes such as the OODA Loop (above) allow rapid response to perturbations and brings control to uncertain, dangerous, and risky situations. OODA Loop decision making also allows decisions to rapidly migrate up and down the hierarchy to the individual with the most expertise with the situation. Expertise does not equate with experience or rank. Formal teaching of decision making enables individuals to consistently make more complex decisions and allow them to migrate. Redundancy. Through his reliance on team training the medical director built redundancy into the system. Routine clinical discussions and rounds

278

DAVED W. VAN STRALEN ET AL.

are also ways to build redundancy into situations. Effortsa to deﬁne care can also contribute to redundancy because discussions occur in such efforts that illuminate the activities various individuals can substitute for those of other individuals. Senior managers with the big picture. Throughout the previous discussion we see the medical director as a guiding hand. He tries to implement a hands off posture in order to give his staff sufﬁcient ﬂexibility to learn and do their jobs. These jobs are integrated through such processes as round and staff discussions. Opening up information ﬂow adds to this. The Charge Nurse also takes a big picture view by letting his/her staff engage in operations which she manages those operations. Formal rules and procedures. Protocols and preplanned treatments are formal rules and procedures. These are used in the deterministic situations often found in any health care unit. They are supplemented by the other command and control processes in more ﬂuid and changing situations.

Additional Emergent HRO Processes In addition to the HRO processes identiﬁed by Libuser three other processes emerged from our analysis. They are: trust, values, and beliefs. High trust. Focus on the quality of a person’s descriptions helped caregivers develop trust in other individual’s observations. The leader’s development, education, and transformation of individuals also helped grow the person’s trust in his/her judgments. Improvement in observations and judgments occurred when a caregiver saw that decisions and actions resulted from trustworthy information the caregiver provided. Medicine typically engages in low trust during highly uncertain situations. Low trust, when used as a redundancy function can strengthen a system. Low trust used to create self-doubt in a person will weaken it. Values. Values guide the selection of behavior and the evaluation of events, people, and self (Schwartz, 1992). They are concepts or beliefs that help interpret an individual’s goals, concerns, or motivations and transcend speciﬁc situations to guide the selection or evaluation of behaviors and events. Dynamic relations exist between value types with some serving an individual’s interests (e.g. self-direction) while some serve the group’s interests (e.g. conformity). The medical director and RCP manager fostered a system of values dependent on the situation with the expectation that values might shift from obedience in low-tempo times to creativity and leadership in high-tempo

Changing a Pediatric Sub-Acute Facility

279

times. During low-tempo times with predominantly deterministic problems, values included conformity (obedience, self-discipline, and politeness) and security (reciprocation of favors, sense of belonging, and social order) as described by Schwartz (1992). When an indeterminant problem arises or the tempo becomes high, staff members change to self-direction (independence, choosing own goals, curious, and self-respect) and stimulation (varied life and daring). Security is a value necessary for work in high-risk environments. The means to reach security can differ from reliance on obedience to reliance on team formation. The sub-acute facility, in this case, used team formation. Beliefs. Beliefs are the concepts a group shares. The medical director and RCP manager speciﬁcally focused on self-efﬁcacy (e.g. Bandura, 1997) and taught that resilience supersedes rigidity in an emergency or uncertain state. Self-efﬁcacy is a process of self-persuasion that involves motivational and selection processes which give the resilient self-belief that one can inﬂuence an outcome. Self-efﬁcacy decreases perceived vulnerability, reduces incidence of intrusive negative thinking and anxiety, and enhances risk discernment. People develop self-efﬁcacy through progressive mastery of decision making through use of bifurcation. The staff member makes decisions under a manager’s guidance. If a wrong decision is made, more information is given with the goal of ﬁnding the correct response. This identiﬁes how much information a staff member uses in decision making (too much or too early). The staff member always makes the ﬁnal decision. When faced with uncertainty many people ﬁnd comfort in structure and rigidity. The management team encouraged ﬁnding comfort in the team and self-efﬁcacy and that, together, the team can solve the problem. This requires resilience in decision processes and cooperation which are destroyed by overly reliant, rigid, thinking. This also allows for a ﬂattened hierarchy during dynamic states.

THE OUTCOMES During the ﬁve years of the implementation process there were no changes in state reimbursement, patient/nurse ratio, or educational level of caregivers. During the same period the relative number of emergency service calls (911 calls) decreased, referrals from pediatric intensive units (PICUs) increased, PICU ventilators began to be used, and the number of intensive care unit ventilator-dependent residents increased. Facility staff members became adept at initiating ventilation for tracheostomy-dependent

280

DAVED W. VAN STRALEN ET AL.

children in acute respiratory failure without the assistance of an on-site physician or the aid of blood gas analysis or sedative medications. The focus of caregivers shifted from maintaining life to enhancing life through use of ventilator management. The goal of care for these children evolved from life maintenance to children who smile and laugh. The facility’s market expansion came because they entered new markets, accepting patients from different PUCUs than previously. Their reputation improved as evidenced by the fact that state authorities asked physicians thinking about developing new sub-acute facilities to visit and talk with members of this facility. In addition, at least one state inspector used the facility as a training device to show new inspectors how treatment should be given. The level of care given increased in that this facility began to use ICU type ventilators, which are more complex than home ventilators and require training not everyone can absorb. Market share increased with the utilization of this technology because the facility could handle more complex cases than previously.

CONCLUDING STATEMENTS AND IMPLICATIONS HRO is a codiﬁcation type of safety culture found naturalistically in risky, high-tempo environments that operate well, such as naval aviation (Roberts, 1990) and the ﬁre service (Bigley & Roberts, 2001). Resources expended to maintain safety are considered wise investments and a part of business. Greater demands and risks within a system, particularly ﬂuctuating or uncertain demands and risks, require greater resources to ensure safety. In low tempo, less-risky environments, use of such resources may not be available for productivity and could reduce efﬁciency. Under these circumstances increasing resources to decrease risk could decrease productivity and efﬁciency and lead to economic failure of the organization. Without the differential use of resources to separate the capabilities of the system from ﬂuctuating demands, catastrophic failure can result. In the experience of this facility the use of HRO principles increased both safety and efﬁciency. This allowed expansion into a new market of PICU referrals and decreased regulatory sanctions. Over all cost avoidance to the state Medicaid program was great as patients were transferred from the PICU to the sub-acute facility. At the same time, the facility developed a new model which beneﬁted these children. This model of HRO use in a nursing home, which allowed it to become a pediatric sub-acute facility, will beneﬁt medical caregivers in other austere

Changing a Pediatric Sub-Acute Facility

281

medical environments such as under-developed countries or disaster situations. Governments are concerned that a ﬂu epidemic or terrorist event will lead to a surge in hospital ICU admissions when there is little capacity to accommodate this increased load. Use of an HRO model of subacute facility care could allow medical facilities to accept more patients during times of crisis. Further use in medicine can include ventilator management in the ICU after muscle relaxant medications are no longer used. Adjusting the ventilator for patient comfort can decrease use of sedative drugs and improve patient comfort. Safety and HRO process implementation are sometimes thought of as an increased cost. However, rational behavioral models predict that organizations that could prevent crises would prevent them, thus reducing costs. This case demonstrates that HRO can reduce costs, improve safety, and aid in developing new health care markets. Other experiences in implementing high reliability processes in organizations show that each organization tailor must make processes to its own situation (e.g. BP, U.S. Chemical Safety and Hazards Board, Federal Aviation Administration, U.S. Navy Aviation Program, and Kaiser Permanente Health Care System). Just as in the ﬂexibility called for in organizing for high reliability operations, ﬂexibility is called for in deciding which HRO processes work in which situations.

REFERENCES Bandura, A. (1997). Self efﬁcacy: The exercise of control. New York: Freeman. Bigley, G. A., & Roberts, K. H. (2001). Structuring temporary systems for high reliability. Academy of Management Journal, 44, 1281–1300. Briassoulis, G., Filippou, O., Natsi, L., Mavrikiou, M., & Hatzis, T. (2004). Acute and chronic pediatric intensive care patients: Current trends and perspectives on resource utilization. Quarterly Journal of Medicine, 97, 507–518. Coram, R. (2002). Boyd: The ﬁghter pilot who changed the art of war (320pp). New York: Little Brown and Company. Eisenhardt, K. (1993). High reliability organizations meet high velocity environments: Common dilemmas in nuclear power plants. In: K. H. Roberts (Ed.), New challenges to understanding organizations. New York: Macmillan. Endsley, M. (1995). Toward a theory of situation awareness in dynamic systems. Human Relations, 37, 32–64. Gaba, D. M., Singer, S. J., Sinaiko, A. D., Bowen, J. D., & Ciavarelli, A. (2003). Differences in safety climate between hospital personnel and Navy aviators. Human Factors, 45, 173–185. Graham, R. J., Fleegler, E. W., & Robinson, W. M. (2007). Chronic ventilator need in the community: A 2005 pediatric census of Massachusetts. Pediatrics, 119(6), e1280–e1287.

282

DAVED W. VAN STRALEN ET AL.

Hammond, G. T. (2001). The mind of war: John Boyd and American security. Washington, DC: Smithsonian Books. Kalin, N. H. (1993). The neurobiology of fear. Scientiﬁc American Magazine, May. Kerr, S. (1975). On the folly of rewarding A while hoping for B. Academy of Management Journal, 18, 769–783. La Porte, T. R., & Consolini, P. (1991). Working in practice but not theory: Theoretical challenges of high reliability organizations. Journal of Public Administration Research and Theory, 1, 19–47. Libuser, C. (1994). Organization structure and risk mitigation. A dissertation submitted in partial satisfaction of the requirements for the Doctor of Philosophy in Management. Lindsay, M. E., Bijwada, J. S., Schauer, W. W., & Rizich, J. D. (2004). Shifting care of chronic ventilator-dependent patients from the intensive care unit to the nursing home. Joint Commission Journal of Quality and Safety, 30(5), 257–265. Roberts, K. H. (1990). Some characteristics of one type of high reliability organization. Organization Science, 1, 160–176. Rochlin, G. I., La Porte, T. R., & Roberts, K. H. (1987). The self-designing high-reliability organization: Aircraft carrier ﬂight operations at sea. Naval War College Review, 40, 76–90. Schwartz, S. H. (1992). Universals in the content and structure of values: Theoretical advances and empirical tests in 20 countries. Advances in Experimental Social Psychology, 25, 1–65. Weick, K. E. (1987). Organizational culture as a source of high reliability. California Management Review, 29, 116–136. Weick, K. E., & Roberts, K. H. (1993). Collective mind in organizations: Heedful interrelating on ﬂight decks. Administrative Science Quarterly, 38, 357–381.