Manual Therapy Journal - Volume 14, Issue 4, Pages 353-460 (August 2009)

VOLUME 14 NUMBER 4 PAGES 353–460 August 2009

Editors

International Advisory Board

Ann Moore PhD, GradDipPhys, FCSP, CertEd, FMACP Clinical Research Centre for Health Professions University of Brighton Aldro Building, 49 Darley Road Eastbourne BN20 7UR, UK Gwendolen Jull PhD, MPhty, Grad Dip ManTher, FACP Department of Physiotherapy University of Queensland Brisbane QLD 4072, Australia

K. Bennell (Melbourne, Australia) K. Burton (Huddersfield, UK) B. Carstensen (Frederiksberg, Denmark) J. Cleland (Concord, NH, USA) M. Coppieters (Brisbane, Australia) E. Cruz (Setubal, Portugal) L. Danneels (Maríakerke, Belgium) I. Diener (Stellenbosch, South Africa) S. Durrell (London, UK) S. Edmondston (Perth, Australia) L. Exelby (Biggleswade, UK) J. Greening (London, UK) A. Gross (Hamilton, Canada) T. Hall (Perth, Australia) W. Hing (Auckland, New Zealand) M. Jones (Adelaide, Australia) B.W. Koes (Amsterdam, The Netherlands) J. Langendoen (Kempten, Germany) D. Lawrence (Davenport, IA, USA) D. Lee (Delta, Canada) R. Lee (London, UK) C. Liebenson (Los Angeles, CA, USA) L. Maffey-Ward (Calgary, Canada) E. Maheu (Quebec, Canada) C. McCarthy (Coventry, UK) J. McConnell (Northbridge, Australia) S. Mercer (Brisbane, Australia) P. Michaelson (Luleå, Sweden) D. Newham (London, UK) J. Ng (Hung Hom, Hong Kong) S. O’Leary (Brisbane, Australia) N. Osbourne (Bournemouth, UK) M. Paatelma (Jyvaskyla, Finland) N. Petty (Eastbourne, UK) A. Pool-Goudzwaard (The Netherlands) M. Pope (Aberdeen, UK) G. Rankin (London, UK) E. Rasmussen Barr (Stockholm, Sweden) D. Reid (Auckland, New Zealand) A. Rushton (Birmingham, UK) M. A. Schmitt (Amersfoort, The Netherlands) M. Shacklock (Adelaide, Australia) D. Shirley (Sydney, Australia) C. Snijders (Rotterdam, The Netherlands) P. Spencer (Barnstaple, UK) M. Sterling (Brisbane, Australia) M. Stokes (Southampton, UK) P. Tehan (Melbourne, Australia) M. Testa (Alassio, Italy) P. van der Wurff (Doorn, The Netherlands) P. van Roy (Brussels, Belgium) O. Vasseljen (Trondheim, Norway) B.Vicenzino (Brisbane, Australia) M. Wessely (Paris, France) A. Wright (Perth, Australia) M. Zusman (Perth, Australia)

Associate Editor’s Darren A. Rivett PhD, MAppSc, (ManipPhty) GradDipManTher, BAppSc (Phty) Discipline of Physiotherapy Faculty of Health The University of Newcastle Callaghan, NSW 2308, Australia E-mail: [email protected] Deborah Falla PhD, BPhty(Hons) Department of Health Science and Technology Aalborg University, Fredrik BajersVej 7, D-3, DK-9220 Aalborg Denmark Email:[email protected] Tim McClune D.O. Spinal Research Unit. University of Huddersfield 30 Queen Street Huddersfield HD12SP, UK E-mail: [email protected]

Editorial Committee Timothy W Flynn PhD, PT, OCS, FAAOMPT RHSHP-Department of Physical Therapy Regis University Denver, CO 80221-1099 USA Email: [email protected] Masterclass Editor Karen Beeton PhD, MPhty, BSc(Hons), MCSP MACP ex officio member Associate Head of School (Professional Development) School of Health and Emergency Professions University of Hertfordshire College Lane Hatfield AL10 9AB, UK E-mail: [email protected] Case Reports & Professional Issues Editor Jeffrey D. Boyling MSc, BPhty, GradDipAdvManTher, MCSP, MErgS Jeffrey Boyling Associates Broadway Chambers Hammersmith Broadway London W6 7AF, UK E-mail:[email protected] Book Review Editor Raymond Swinkels PhD, PT, MT Ulenpas 80 5655 JD Eindoven The Netherlands E-mail: [email protected]

Visit the journal website at http://www.elsevier.com/math doi:10.1016/S1356-689X(09)00071-X

Manual Therapy 14 (2009) 353–354

Contents lists available at ScienceDirect

Manual Therapy journal homepage: www.elsevier.com/math

Editorial

The primacy of clinical reasoning and clinical practical skills

Intellectual enquiry and the desire for delivery of practice which is evidence based have seen a surge in research and an explosion of knowledge in the musculoskeletal field. There is a constantly increasing volume of research related to musculoskeletal practice. It ranges from studies designed to unravel the mechanisms underpinning pain dysfunction and functional loss, to studies evaluating clinical assessment methods, the physiological effects of interventions, the evidence for the efficacy and cost effectiveness of interventions as well as studies of prognostic indicators. In the clinical arena, clinicians in line with their responsibilities for continuing professional development are attending conferences and courses where they are exposed to information and instruction on a variety of management methods, all providing some evidence of, or claiming effectiveness. All activity aims to enhance clinical practice and the outcomes for patients. However the outcomes of this research activity to date and its synthesis for the clinical setting emphasise the critical element of practitioners’ clinical reasoning and practical skills to ensure good clinical outcomes for patients. To encourage evidence based practice, clinical practice guidelines have been developed from the research evidence to guide patient management decisions. Guidelines provide direction at a ‘first base’ or ‘in principal’ level. For example, we know that encouraging a patient to stay active after an episode of back or neck pain is better than prescribing prolonged bed rest or immobilisation in a collar. However, there is widespread recognition of the limitations of clinical practice guidelines for neck and back pain because of their generic nature in the face of the heterogeneity in patient presentation encountered in clinical practice. Evidence from the results of clinical trials which inform the clinical practice guidelines clearly demonstrate the dilemma. It is readily obvious from the numbers needed to treat or treatment effect size data that a certain treatment modality or method can be highly effective for some patients but ineffective for others, affirming that a ‘one size fits all’ approach to management of patients with neck and back pain is neither possible nor best practice. In the clinical situation, the evidence based approach of the guidelines must be applied in the context of the patient and a thorough understanding of their presenting syndrome which relies on clinician’s high level clinical reasoning, evaluation and therapeutic skills. Recognition of the heterogeneity in patient presentation, causative mechanisms and moderating factors has spurred a growing body of clinical and research work which is aimed at identifying groups of patients who have commonalities in presentation. Such subgrouping is reasoned to better direct management approaches. We are now seeing the development

1356-689X/$ – see front matter Ó 2008 Published by Elsevier Ltd. doi:10.1016/j.math.2009.05.002

of classification systems which have emanated from various perspectives, for example, from movement and motor control characteristics (Dankaerts et al., 2006; van Dillen et al., 2009), pain and movement response characteristics (Clare et al., 2005), or from pathophysiological features using a biopsychosocial framework (Sterling, 2004). An alternate approach to classification is the development of clinical prediction rules which define the primary clinical features of patients who are likely respond to a certain intervention (Tsenga et al., 2006; Cleland et al., 2007; Vicenzino et al., 2008). These classification systems and clinical prediction rules require further validation but have the potential to improve guidelines for the management for the subgroups identified. However on the downside, these classification systems will be relevant for a certain percentage of patients and they do not inform on management of patients who fall outside the classification or clinical rule. Successful outcomes for these patients rely even more heavily on the practitioner’s clinical reasoning and practical skills. A plethora of treatment approaches for musculoskeletal disorders such as neck and back pain are still undergoing scientific scrutiny and this reflects the absence of conclusive evidence for one approach. It confirms the heterogeneity in patient presentation and clinical expectations that some approaches or techniques, even in the spirit of evidence informed practice, stand to be efficacious for some but not all patients. It is the practitioner’s clinical reasoning, assessment and clinical practical skills that are a crucial nexus between the patient, the research evidence and successful clinical outcomes. It might not be too incorrect to assert that the evidence will lose its impact for enhanced patient care in the absence of high level clinical reasoning and practical skills in the practitioner. We are strong advocates for research informed practice, but assert that the fundamental clinical skills of the practitioner should never be undervalued and must attract equal attention in education at professional and post professional levels. References Dankaerts W, O’Sullivan P, Straker L, Burnett A, Skouen J. The inter-examiner reliability of a classification method for non-specific chronic low back pain patients with motor control impairment. Manual Therapy 2006;11:28–39. Clare H, Adams R, Maher C. Reliability of McKenzie classification of patients with cervical or lumbar pain. Journal of Manipulative and Physiological Therapeutics 2005;28:122–7. Cleland J, Childs J, Fritz J, Whitman J, Eberhart S. Development of a clinical prediction rule for guiding treatment of a subgroup of patients with neck pain: use of thoracic spine manipulation, exercise, and patient education. Physical Therapy 2007;87:9–23.

354

Editorial / Manual Therapy 14 (2009) 353–354

Sterling M. A proposed new classification system for whiplash associated disorders. Manual Therapy 2004;9:60–70. Tsenga Y-L, Wanga W, Chena W-Y, Houb T-J, Chenc T-C, Lieuc F-K. Predictors for the immediate responders to cervical manipulation in patients with neck pain. Manual Therapy 2006;11:306–15. van Dillen L, Maluf K, Sahrmann S. Further examination of modifying patientpreferred movement and alignment strategies in patients with low back pain during symptomatic tests. Manual Therapy 2009;14:52–60. Vicenzino B, Smith D, Cleland J, Bisset L. Development of a clinical prediction rule to identify initial responders to mobilisation with movement and exercise for lateral epicondylalgia. Manual Therapy 2008. doi:10.1016/j.math.2008. 08.004.

Gwendolen Jull* Ann Moore NHMRC Centre for Clinical Research Excellence -Spine, School of Health and Rehabilitation Sciences, The University of Queensland, Queensland 4072, Australia  Corresponding author. Tel.: þ61 7 3365 1114; fax: þ61 7 3365 1622. E-mail address: [email protected]

Manual Therapy 14 (2009) 355–362

Contents lists available at ScienceDirect

Manual Therapy journal homepage: www.elsevier.com/math

Systematic Review

The reliability and validity of assessing medio-lateral patellar position: a systematic review Toby O. Smith a, *, Leigh Davies a, Simon T. Donell b a

Orthopaedic Physiotherapy Research Unit, Physiotherapy Department – Out-Patients East, Norfolk and Norwich University Hospital, Colney Lane, Norwich, NR4 7UY, UK b Faculty of Health, University of East Anglia, Norwich, NR4 7TJ, UK

a r t i c l e i n f o

a b s t r a c t

Article history: Received 14 May 2008 Received in revised form 18 July 2008 Accepted 2 August 2008

Medio-lateral patellar position is regarded as a sign of patellofemoral pain syndrome and patellar instability. Its assessment is important in accurately performing patellofemoral therapeutic taping techniques. The purpose of this paper is to systematically review the literature to determine the reliability and validity of evaluating medio-lateral patellar position. An electronic database search was performed accessing AMED, British Nursing Index, CINAHL, the Cochrane database, EMBASE, Ovid Medline, Physiotherapy Evidence Database (PEDro), PubMed and Zetoc to July 2008. Conference proceedings and grey literature were also scrutinised for future publications. All human subject, clinical trials, assessing the inter- or intra-tester reliability, or the criterion validity, were included. A CASP tool was employed to evaluate methodological quality. Nine papers including 237 patients (306 knees) were reviewed. The findings of this review suggest that the intra-tester reliability of assessing medio-lateral patellar position is good, but that inter-tester reliability is variable. The criterion validity of this test is at worse moderate. These are based on a limited evidence-base. Further study is recommended to compare the McConnell (1986) [McConnell J. The management of chondromalacia patellae: a long term solution. Australian Journal of Physiotherapy 1986;32(4):215–23] and Herrington (2002) [Herrington LC. The inter-tester reliability of a clinical measurement used to determine the medial/lateral orientation of the patella. Manual Therapy 2002;7(3):163–7] methods of assessing medio-lateral patellar position in patients with well-defined patellofemoral disorders. Ó 2008 Elsevier Ltd. All rights reserved.

Keywords: Assessment Patellar position Reliability Validity

1. Introduction The aetiology of patellofemoral pain syndrome (PFPS) and patellar instability are multi-factorial (Sutlive et al., 2004). One factor indicated in both patellofemoral disorders is abnormal patellar tracking. The patella in said to be frequently lateralised in both disorders (Mizuno et al., 2001). It is hypothesised that this can cause an increase in retropatellar pressure over the articular surfaces, contributing to articular cartilage degeneration and subsequent pain (Powers et al., 1999; Ota et al., 2006; Fulkerson and Shea, 1990; Hughston, 1968; Insall, 1979). Similarly, patellar maltracking within the femoral sulcus can cause instability symptoms, and predispose the patella to dislocation (Arendt et al., 2002). Taping is one physiotherapeutic strategy aimed at correcting patellar mal-tracking (Warden et al., 2008). This has gained widespread acceptance as a viable treatment option for patients with

* Corresponding author. Tel.: þ44 1603 286990; fax: þ44 1603 287369. E-mail address: [email protected] (T.O. Smith). 1356-689X/$ – see front matter Ó 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.math.2008.08.001

PFPS and patellar instability (Powers et al., 1999; McConnell, 1986, 2007). The patella is taped specifically to address the individual’s abnormal glide, rotation and tilt, and to maintain the patellar correctly within the femoral trochlea throughout range (Warden et al., 2008; Crossley et al., 2001). Since the medial and lateral translation have been associated with PFPS and patellar instability, it is important that the extent and direction of such translation can be accurately assessed. The medial and lateral displacement of the patellar can be measured by two means (Figs. 1 and 2). McConnell (1986) first described assessing this through palpation and visual estimation. She suggested that both the medial and lateral femoral epicondyles should be palpated and identified with both index fingers. The midpatellar point should then be recognised using both thumbs. In normal cases, the distance between index fingers and thumbs will be approximately equal. If the patella is laterally displaced, the distance between the index finger palpating the lateral epicondyle to thumb, will be less than the other fingers measuring the medial patellar position. This is reversed for medial displacement (McConnell, 1986). More recently, Herrington (2002) has assessed

356

T.O. Smith et al. / Manual Therapy 14 (2009) 355–362

a review of the literature suggested that such an evaluation had yet to be performed.

2. Methodology 2.1. Study eligibility criteria

Fig. 1. The McConnell (1986) approach for the assessment of medio-lateral patellar position.

this distance by marking the epicondyles and mid-patellar position on zinc tape and measuring the medial and lateral distance with a tape measure. The validity of a test is the extent to which a test measures what it is intended to measure (Edwards and Talbot, 1994; Polgar and Thomas, 2000). One aspect of validity is criterion validity. This assesses how the test under investigation compares against an established or gold-standard measure (Evans et al., 2004). In the case of patellar position, such gold-standard tests would include magnetic resonance imagery (MRI), computed tomography (CT) or plain radiographs (Grelsamer et al., 1998; Herrington, 2006; Tolouei et al., 2005). Reliability is the extent to which a test is reproducible (Polgar and Thomas, 2000; Edwards and Talbot, 1994). This is subcategorised into two types. Inter-tester reliability assesses the degree to which different examiners give consistent estimates of the same test (Portney and Watkins, 2000). Intra-tester reliability assesses the consistency of a measure on two different occasions (Polgar and Thomas, 2000; Portney and Watkins, 2000). The objective of this study is to systematically review the evidence-base to determine the inter- and intra-tester reliability and criterion validity of medio-lateral patellar position. This has considerable clinical importance given that this assessment forms the basis of therapeutic taping of the patellofemoral joint, a widely used and acceptable intervention in clinical practice for patients with patellofemoral disorders. This is further justified since

Fig. 2. The Herrington (2002) approach for the assessment of medio-lateral patellar position.

The inclusion criteria included all full text papers assessing medio-lateral patellar position by two or more examiners, at one or more time points (inter- or intra-tester reliability). Papers comparing the clinical assessment of medio-lateral patellar position to a radiological assessment using MRI, CT or plain radiograph (criterion validity) were also included. Papers of any language were included, as well as unpublished material including university theses and dissertations and conference proceedings, in an attempt to limit publication bias from impacting on this systematic review’s findings. Papers were excluded if they presented insufficient data on their method of assessing medio-lateral patellar position. Single-subject case reports, comments, letters, editorials, protocols, guidelines, or review papers were excluded. The reference lists of review papers were scrutinised for any clinical papers deemed relevant to the research question. No exclusion was made to subject age or gender. Animal and cadaver studies were excluded.

2.2. Search strategy The primary search was a search of the electronic databases AMED, British Nursing Index, CINAHL, the Cochrane database, EMBASE, Ovid Medline, Physiotherapy Evidence Database (PEDro), PubMed and Zetoc from their inception to July 2008. Key terms and Boolean operators adopted included: patella AND position; orientation. A secondary search of the following specialist journals was undertaken: Knee Surgery Sports Traumatology Arthroscopy (1993– July 2008), The Knee (1994–July 2008), the British and American editions of the Journal of Bone and Joint Surgery (1988–July 2008), American Journal of Sports Science (1988–July 2008), and Journal of Orthopaedic Sports and Physical Therapy (1991–July 2008). Unpublished or grey literature was assessed using the databases SIGLE (System for Information on Grey Literature in Europe), the National Research Register (UK), the National Technical Information Service, the British Library’s Integrated Catalogue, and Current Controlled Trials database for recently completed trials. Conference proceedings from the British Orthopaedic Association Annual Congress and British Association for Surgery of the Knee were searched from 2002 to 2008, for additional studies pertaining to this research question. Using the predefined eligibility criteria, two investigators (TS, LD) independently assessed all identified titles and abstracts. Full manuscripts of citations adhering to the criteria were ordered. Full manuscripts were ordered of those citations the reviewers were uncertain about after reading the abstracts. Reference lists from each full manuscript were scrutinised to identify any publications not initially identified. Each full text was then screened against the eligibility criteria by the same two reviewers. In cases of disagree, a census was reached through discussion. No paper was excluded on poor methodological quality. The two investigators were not blinded to the source or authors of the papers reviewed. The corresponding author of each paper included in the review was then contacted. They were asked whether they knew of any additional papers which had not been identified by the search strategy, to ensure that every paper potentially answering this research question, had been considered in this systematic review.

T.O. Smith et al. / Manual Therapy 14 (2009) 355–362

357

Articles recovered from the search strategy. (n= 1425)

Title or abstract not pertaining to the research question. (n= 1328) Abstracts which appeared relevant to the research question (n=97) Articles deemed not related to the research question after consulting the full abstract. (n= 62) Appropriate studies related to the research question, permitting full manuscripts to be ordered for further scrutiny. (n= 35)

Articles excluded as not adhering to the eligibility criteria. (n= 26) Appropriate studies related to the research question, and adhering to the eligibility criteria. (n =9) Articles excluded due to replication of] data presented. (n = 0) Final included articles. (n =9) Fig. 3. A QUORUM flow-chart.

2.3. Data extraction Data from all studies fully satisfying the eligibility criteria was entered into a spreadsheet by a single investigator (TS), and verified by a second investigator (LD). This spreadsheet tabulated:      

   

Author names and publication date Study design Sample size Population characteristics including diagnosis, subject age and gender Method of assessing Tester details including number of tester, frequency of testing, experience of tester, teaching of tester to the measurement procedure Method of reference test for criterion validity Statistical analysis Results Any relevant methodological limitations

2.4. Critical appraisal All included papers were evaluated against an appraisal based on the Critical Appraisal Skills Programme (CASP, 2007) appraisal tool for diagnostic test studies. This appraisal tool comprises of three sections: an assessment of study validity; an evaluation of methodological quality and presentation of results; an assessment of external validity. Each paper was assessed independently by two reviewers (TS, LD). Any differences in appraisal results were settled

through discussion. There was a difference between the reviews over six items, in two papers (Tomsich et al., 1996; Fitzgerald and McClure, 1995), which was resolved by discussion. 3. Results 3.1. Search results Fig. 3 outlines the results of the search strategy. The search yielded 1425 articles whose titles and abstracts were read. Of these, 9 studies met the eligibility criteria, these are summarised in Tables 1 and 2. The papers have been subdivided by their study aims and discussed below. A total of 213 patients and 282 knees were reviewed. This included 29 patients (37 knees) diagnosed as PFPS, and 104 patients (164 knees) asymptomatic control subjects. One paper did not specify the pathology of its sample, whilst 76 patients (77 knees) were collectively assessed as general knee pathologies. Eighty-four males, and 137 females were reviewed, whilst three studies did not specify the gender of their cohorts. In those with patellofemoral disorders, the age of subjects ranged from 18 to 41 years, with a mean of 30.6 years. This differed from the asymptomatic samples where age ranged from 18 to 28 years, with a mean of 24.2 years. As expected, the method for assessing medio-lateral patellar position was made using two techniques. The majority of studies assessed patellar position following McConnell’s (1986) method. Only Tomsich et al. (1996) assessed patellar orientation differently using visual estimation or pluri-cal callipers. There was some variability in assessment position using this method. McEwan et al. (2007), Herrington (2008), Herrington (2002), Herrington et al.

358

T.O. Smith et al. / Manual Therapy 14 (2009) 355–362

Table 1 Summary of the papers included in this systematic review. Study Design Sample size Population

Medial/lateral position test

Reliability assessment Validity assessment Tester details

Statistical analysis Study Design Sample size Population

Medial/lateral position test

Reliability assessment Validity assessment Tester details Statistical analysis Study Design Sample size Population Medial/lateral position test

Reliability assessment Validity assessment

Tester details

Statistical analysis Study Design Sample size Population

Fitzgerald and McClure (1995) Observational 66 (66 knees) 66 symptomatic subjects; 31 males, 35 females; mean age 29.7  13.1 years (range 14–74); mean weight 73.4  19.6 kg; mean height 171.2  10.2 cm 40 diagnosed with patellar pain syndrome, anterior knee pain, chondromalacia patellae, subluxing patellar, patellar tendonitis or patellar fracture 26 diagnosed with meniscal pathology, ligamentus pathology, femoral or tibia fracture Subjects excluded if they had received a surgical procedure specifically to realign the patella (e.g. lateral retinacular release) McConnell assessment: subject supine, tibiofemoral joint in full extension, quadriceps contraction or lower limb rotation not documented. Palpation of the medial and lateral femoral epicondyles with the index finger and simultaneously palpating the mid-patellar with the thumbs. Normally. Distance between the fingers and thumb should be equal. If a lateral displacement is present, the distance between the index finger palpating the lateral epicondyle to the thumbs will be less than the distance from the fingers palpating the medial epicondyle to the thumbs. If a medial displacement is evident, the distance between the medial epicondyle to the thumbs will be less than the distance from the lateral eipcondyle to the thumbs Subjects were independently assessed once, by 2 different examiners, most often during the same clinic session Not assessed 12 physical therapists from 4 clinics who frequently treat patients with knee or patellofemoral disorders. All testers were familiar with the patellar orientation test prior to participation in the study. One tester had learnt the patellar orientation test from McConnell’s course. The other tester had learnt the test from colleges or from reading texts on the method of assessment. For this study, each tester received a written and photographic description of how to perform the tests based on McConnell (1986). This was provided approximately 2 weeks prior to testing Inter-tester reliability, kappa Herrington (2008) Observational, matched pairs 24 (24 knee) 12 asymptomatic subjects; 12 females; mean age 21.6  2.8 years (range 18–25); mean body mass 62.3  8.4 kg 12 subjects diagnosed with patellofemoral pain syndrome for at least 1 month; 12 females; mean age 21.6  2.6 years (range 18–25); mean body mass 64.5  9.3 kg Subjects were excluded if they had reported previous knee surgery or arthritis, history of patellar dislocation, subluxation or ligament laxity, patellar tendonopathy, chondral damage, spinal referred pain, lower limb abnormalities such as leg length discrepancy (>2 cm), were taking medication as part of their knee treatment, or had received previous knee physiotherapy and acupuncture treatment within the previous 30 days McConnell (1986) and Herrington and Nester (2004) assessment: subject general position, quadriceps contraction or lower limb rotation not documented, knee in 20 degrees flexion. Centre of patella, medial and lateral femoral epicondyle marked on a piece of folded zinc oxide tape placed on subject’s knee. Distance between the medial epidcondyle to mid-point of patella, and lateral epicondyle to mid-point of patella measured .. Not stated what used to measure distances. An assessor marked these distances on the zinc oxide tape, and a second assessor blinded to diagnosis, measured the distance between the markings. This whole procedure was repeated 3 times, with the average of the 3 measurement recorded This procedure was then repeated to assess the 12 matched control subjects on a separate occasion 1–2 days after the original measurement Not assessed The physiotherapist identifying the relevant bony landmarks and marking the zinc oxide tape was an experienced physical therapist. Details for the independent assessor blinded to diagnosis, who measured the markings, was not documented Intra-tester reliability, ICC AND SEM Herrington (2002) Observational 1 (1 knee) 1 subjects; pathology, gender, age not specified McConnell (1986) assessment: subject general position, quadriceps contraction or lower limb rotation not documented, knee in 20 degrees flexion. Centre of patella, medial and lateral femoral epicondyle marked on a piece of folded zinc oxide tape placed on subject’s knee. Distance between the medial epidcondyle to mid-point of patella, and lateral epicondyle to mid-point of patella measured .. Not stated what used to measure distances. Each assessor repeated this procedure 3 times each, re-palpating and applying tape on each occasion. Average of the 3 measurement recorded This procedure was then repeated to assess the subject on 2 separate occasions. Not specified how long duration was between assessments MRI assessment of medial/lateral patellar position with patient supine, in 20 degrees knee flexion, no details on limb rotation. Medio-lateral position determined by assessing lateral patellar displacement in relation to femoral condyles. LPD measured 3 times, with average taken. Not specified which MRI slice used to assess LPD Clinical test: medio-lateral patella orientation measured by 20 chartered physiotherapists, experience musculoskeletal physiotherapists at MACP examination approved level, and a minimum of 5 years specialising in musculoskeletal physiotherapy. Details for this assessor on experience of this test or of teaching of this technique were not detailed. MRI test: all made from one investigator blinded to the clinical examination findings Inter-tester reliability and criterion validity, means of the ICC

Reliability assessment Validity assessment Tester details Statistical analysis

Herrington and Nester (2004) Observational 10 (20 knees) 10 asymptomatic, gender, age not specified Asymptomatic described as physically active asymptomatic individuals with no history of lower limb, spinal or neurological injury Herrington (2002) assessment: subject general position, quadriceps contraction or lower limb rotation not documented, knee in 20 degrees flexion. Centre of patella, medial and lateral femoral epicondyle marked on a piece of folded zinc oxide tape placed on subject’s knee. Distance between the medial epidcondyle to mid-point of patella, and lateral epicondyle to mid-point of patella measured .. Not stated what used to measure distances. Each assessor repeated this procedure 3 times each, re-palpating and applying tape on each occasion. Average of the 3 measurement recorded This procedure was then repeated to assess the subjects on 2 separate occasions. Not specified how long duration was between assessments Not assessed Not documented Intra-tester reliability, ICC and SEM

Study Design

Herrington et al. (2006) Observational

Medial/lateral position test

T.O. Smith et al. / Manual Therapy 14 (2009) 355–362

Sample size Population Medial/lateral position test

Reliability assessment Validity assessment Tester details Statistical analysis Study Design Sample size Population Medial/lateral position test

Reliability assessment Validity assessment

Tester details

Statistical analysis Study Design Sample size Population

Medial/lateral position test

Reliability assessment Validity assessment Tester details Statistical analysis Study Design Sample size Population Medial/lateral position test

Reliability assessment Validity assessment Tester details Statistical analysis Study Design Sample size Population

359

5 (5 knees) 5 asymptomatic subjects; males/females not specified; mean age not specified Asymptomatic described as physically active asymptomatic individuals McConnell (1986) and Herrington (2002) assessment: subject general position, quadriceps contraction or lower limb rotation not documented, knee in 20 degrees flexion. Centre of patella, medial and lateral femoral epicondyle marked on a piece of folded zinc oxide tape placed on subject’s knee. Distance between the medial epidcondyle to mid-point of patella, and lateral epicondyle to mid-point of patella measured .. Not stated what used to measure distances. 1 assessor repeated this procedure 3 times each, re-palpating and applying tape on each occasion. Average of the 3 measurement recorded This procedure was then repeated to assess the subjects on 2 separate occasions. Not specified how long duration was between assessments Not assessed Clinical test: all measures taken by 1 assessor, Details for this assessor on experience of this test or of teaching of this technique was not detailed Intra-tester reliability, means of the ICC and SEM McEwan et al. (2007) Observational 24 (24 knees) 24 asymptomatic subjects; 16 males, 8 females; mean age 24.5  7.9 years, range 18–42 years Asymptomatic consist of no current or previous history of knee or lower extremity injury McConnell (1986) and Herrington and Nester (2004) assessment: subject general position, quadriceps contraction or lower limb rotation not documented, knee in 20 degrees flexion. Centre of patella, medial and lateral femoral epicondyle marked on a piece of folded zinc oxide tape placed on subject’s knee. Distance between the medial epidcondyle to mid-point of patella, and lateral epicondyle to mid-point of patella measured Not stated with what measured. 2 assessors repeated this procedure 3 times each, re-palpating and applying tape on each occasion. Average of the 3 measurement recorded This procedure was then repeated 1 day later to assess the subjects on 2 separate occasions MRI assessment of medial/lateral patellar position with patient supine, in 20 degrees knee flexion, no details on limb rotation. Medio-lateral position determined by assessing lateral patellar displacement in relation to femoral condyles. LPD measured 3 times, with average taken. Not specified which MRI slice used to assess LPD Clinical test: all measures taken by 2 independent assessor. Details for 1 assessor provided as a musculoskeletal physiotherapist with 15 years experience, not specified how long undertaken this testing procedure, or details of teaching of this technique to the assessor. MRI test: all made by one investigator blinded to the clinical examination findings Intra-tester reliability, means of the ICC criterion validity, Pearson’s product moment Powers et al. (1999) Observational 24 (38 knees) Intra-tester assessment: 10 subjects (20 knees) asymptomatic; 4 males, 6 females; mean age 26  2 years. Asymptomatic consisted of pain-free status, not specified of what Criterion validity: 4 subjects (7 knees) asymptomatic – unspecified criteria. 10 subjects (11 knees) symptomatic – described as either tibiofemoral joint osteoarthritis, anterior knee pain, meniscal injury – not specified how these diagnoses were determined. In total: 10 females, 4 males; mean age 41  16 years McConnell (1986) assessment: subject supine, knee extended, quadriceps relaxed, limb rotation not documented. Centre of patellar determined and marked, with a soft tape measure, and bisecting the distance between the most medial and lateral borders of the patella. Distance from centre of patellar to medial femoral epicondyle, and from centre of patella to lateral femoral epicondyle was then assessed using the tape measure 4 measurements of medial/lateral position median and averaged, 2 for medial and 2 for lateral femoral epicondyle distance. Measurements made on 2 separate occasions at least 2 weeks apart MRI assessment of medial/lateral patellar position with patient supine, full extension with quadriceps relaxed, in natural limb rotation for each patient (10–15 degrees external rotation). The image containing the largest patellar cross-section (mid-patellar slice) was used for analysis Clinical test: all measures taken by 1 assessor, who had less than 1 year’s experience of this technique. MRI test: all made by one investigator and measured according to procedure Intra-tester reliability, ICC criterion validity, ANOVA Tomsich et al. (1996) Observational 27 (27 knees) 27 asymptomatic subjects; 7 males, 20 females; mean age 21  5.5 years Asymptomatic defined as no history of knee pathology Assessed by visual estimation and pluri-cal calliper, each subject in supine position, in 0 degrees tibiofemoral flexion, quadriceps relaxed, foot position to maintain leg in neutral rotation using a KT-100 foot stabilizer. (1) Visual estimation of medio-lateral position assessed by positioning the index fingers and thumbs on the sides of the femoral epicondyles and over the patellar midpoint. The distance between the index finger and thumb medially and laterally is an estimate. (2) Medio-lateral position assessed by callipers by positioning the callipers along both femoral epicondyles, the midpoint marker on the ruler of the pluri-cal callipers was placed over the patellar midpoint. The distance between the epicondyle to the patellar midpoint, medially and laterally, was then recorded Subjects assessed 3 times each, by the three different examiners. Assessors were blinded to subject’s identification Not assessed 3 physical therapists who spent a total of 2 h practicing the measurement procedure before the study. Therapist’s ages ranged from 25 to 27 with 2.5 to 5.5 years experience of orthopaedic practice, graduating from 3 different physiotherapy schools Intra- and inter-tester reliability, kappa, ICC and SEM Watson et al. (1999) Observational 56 (101 knees) 56 subjects; 22 males, 34 females; mean age 29  8.0, range 22–34 years 39 (76 knees) asymptomatic subjects mean age 28  6.2 17(25 knees) symptomatic subjects mean age 30  11.4 Asymptomatic defined as no knee pain or pathology Symptomatic defined as patellofemoral pain classified as patellofemoral joint pain reproducible following at least 2 of the following activities in the last month: ascending or descending stairs, prolonged sitting, squatting Subjects excluded if they had a history of knee surgery or patellar dislocation, or if there was evidence on clinical examination of knee ligamentus injury, meniscal injuries, patellar tendonopathy, major joint effusion or plica syndrome (continued on next page)

360

T.O. Smith et al. / Manual Therapy 14 (2009) 355–362

Table 1 (continued) Medial/lateral position test

Reliability assessment Validity assessment Tester details Statistical analysis

McConnell assessment: subject supine, tibiofemoral joint in full extension and femur parallel to the plinth, quadriceps relaxed through lower limb rotation not documented. Distance measured using a tape measure from the mid-patella to the medial femoral epicondyle and to the lateral femoral epicondyle. The midpoint was determined on visual estimation, and marked with a grease pencil. A score of 0 was allocated if the distance from the medial epicondyle to the mid-patella point was equal to the distance from the lateral epicondyle to mid-patella; 1 was allocated if the distance from the medial epicondyle to the mid-patella point was >0.5 cm from the lateral epicondyle to mid-patella Subjects assessed once, by 2 different examiners, on 2 separate occasions, 3–7 days after the initial measure. Assessors were blinded to whether subjects had knee pain Not assessed 2 senior physical therapy students who had received approximately 2 h of didactic instructions and 2 h of practice on the McConnell patellofemoral classification system of medio-lateral position Intra- and inter-tester reliability, kappa

ANOVA, analysis of variance; ICC, intra-class correlation coefficient; LPD, lateral patellar displacement; MACP, Manipulation Association of Chartered Physiotherapists; MRI, magnetic resonance imaging; SEM, standard error of the measurements.

(2006) and Herrington and Nester (2004) assessed the knee in 20 degrees flexion, whereas Powers et al. (1999), Tomsich et al. (1996), Fitzgerald and McClure (1995) and Watson et al. (1999) assessed the knee in full extension. Lower limb rotation was only documented as controlled in Tomsich et al.’s (1996) study in neutral. Only Watson et al. (1999), Tomsich et al. (1996), and Powers et al. (1999) acknowledged that the quadriceps muscles were relaxed during testing.

3.2. Reliability Intra-tester reliability of medio-lateral patellar position as assessed in seven studies. Six studies reported either substantial or near perfect agreement between assessment periods, with intraclass coefficient (ICC)/kappa results ranging from 0.70 to 0.99. Four studies all reported almost perfect agreement between test procedures in McEwen et al. (2007), Powers et al. (1999), Herrington et al. (2006), and Herrington and Nester’s (2004) results. Only Watson et al.’s (1999) study reported poor to fair agreement with 0.11–0.35 kappa results. Four studies assessed inter-tester reliability. These studies reported differing results. Two studies reported near perfect agreement in Herrington (2002) results with 0.91 and 0.94. In contrast, Tomsich et al. (1996), Fitzgerald and McClure (1995) and Watson et al. (1999) reporting poor agreement with results of 0.14, 0.10 and 0.02 respectively.

3.3. Criterion validity The criterion validity of assessment of medio-lateral patellar orientation was evaluated in studies by Herrington (2002), McEwan et al. (2007) and Powers et al. (1999). These studies reported variable agreement between clinical medio-lateral patellar position assessment and MRI evaluation. Herrington (2002) reported near perfect agreement between the measures with an ICC of 0.9. McEwan et al. (2007) reported substantial agreement with an ICC of 0.61, whilst Powers et al. (1999) reported moderate agreement with an ICC of 0.44. 3.4. Critical appraisal results The findings of the CASP appraisal are presented in Table 3. These suggest that the methodological quality of the papers was limited in a number of areas. The CASP review highlighted that all studies stated appropriate research questions and applied suitable study designs to answer their research questions. As Table 1 outlines, three studies used a references test (MRI) to assess the criterion validity of the medio-lateral patella position. Only McEwan et al. (2007) stated that assessors were blinded to the results of this test, whilst Herrington (2002) indicated that different assessors were used for their MRI and clinical findings. Population characteristics such as patient’s knee history and pathology, gender, age or weight and height were poorly described in five papers. Whilst all papers identified the basic method of assessing medio-lateral patellar

Table 2 Summary of results from studies included in this systematic review. Author (date)

Mean clinically assessed medio-lateral position in mm (SD)

Mean reference test medio-lateral position (SD)

Inter-tester reliability (ICC with SEM)

Intra-tester reliability (ICC with SEM)

Criterion validity ICC (p value)

Fitzgerald and McClure (1995) Herrington (2008)

N/D

N/A

0.10 (44%)a

N/A

N/A

PFPS group: lateral 7.5 (2.6); asymptomatic group: lateral 3.8 (2.4) Medial 8.98 (0.51); lateral 8.35 (0.66)

N/A

N/A

0.86, SEM 0.2 mm

N/A

LPD 5.0(2.8)

N/A

0.9

3 mm (6 mm) lateralised from central Medial females 2.5 (1.8)/males 2.3 (1.8); lateral females 6.6 (4.5)/males 5.9 (4.5) Medial 8.9 (0.1); lateral 8.3 (0.1) 6.8  9.6% lateral displacement as percentage of patella width

N/A

Medial measure 0.91; lateral measure 0.94 N/A

0.99 (p < 0.01), SEM 6 mm

N/A

N/A

N/A

0.99 (p < 0.015), SEM 0.1 mm

N/A

LPD 8.1 (2.8) 16.1  12.3% lateral displacement as percentage of patella width N/D N/A

N/A N/A

0.86 (0.1, CI 8.1–8.5) 0.91

0.61 (p ¼ 0.002) 0.44

0.14 (0.55) 0.02 (70%)a

0.70 (0.28) 0.11 (8.4) to 0.35 (0.74)a

N/A N/A

Herrington (2002)

Herrington et al. (2006) Herrington and Nester (2004) McEwan et al. (2007) Powers et al. (1999)

Tomsich et al. (1996) Watson et al. (1999)

N/D Frequency between testers of: score 0, 135–149; score 1, 47–53

ICC, intra-class correlation co-efficient; N/D, not documented; mm, millimetres; SD, standard deviation; N/A, not assessed; SEM, standard error of the measurements. a Kappa coefficient (percentage of agreement).

T.O. Smith et al. / Manual Therapy 14 (2009) 355–362

361

Table 3 A summary of the CASP results. CASP factors

Fitzgerald and McClure (1995)

Herrington Herrington Herrington (2008) (2002) et al. (2006)

Herrington and Nester (2004)

McEwan Powers Tomsich Watson et al. (2007) et al. (1999) et al. (1996) et al. (1999)

Clearly focused question stated Appropriate design Appropriate reference test available Did all receive reference and diagnostic test Could reference test findings influence diagnostic test result Population characteristics clearly defined Diagnostic test clearly defined Appropriate results analysis Precise statistical results presented Appropriate interpretation Ability to generalise results Were the results applicable to clinical practice

Y Y N

Y Y N

Y Y Y

Y Y N

Y Y N

Y Y Y

Y Y Y

Y Y N

Y Y N

N

N

Y

N

N

Y

Y

N

N

N/A

N/A

N

N/A

N/A

N

Y

N/A

N/A

N

Y

N

N

N

Y

Y

Y

Y

N Y N

N Y N

N Y N

N Y N

N Y N

N Y Y

N Y N

Y Y N

N Y N

Y Y Y

Y Y Y

Y N N

N N N

N N N

Y N N

Y Y Y

N N N

Y Y Y

Y, yes; N, no; N/A, not applicable.

position, all studies except Tomsich et al. (1996) did not clearly state either the position of the knee or lower limb, or whether the quadriceps were relaxed or contracted, confounding variables in assessing medio-lateral patellar position. The evidence-base used appropriate statistical tests with ICC and kappa analysis, but only McEwan et al. (2007) documented confidence intervals to assess the precision of their statistical findings. Since only Fitzgerald and McClure (1995), Powers et al. (1999), Watson et al. (1999) and Herrington (2008) recruited patients with patellofemoral pathology, only these studies were regarded as having any clinical significance to be generalisable to the clinic setting. 4. Discussion The findings of this review suggest that the intra-tester reliability of medio-lateral position tests is good, but that inter-tester reliability is variable. The criterion validity of these tests is at worst moderate. However, such conclusions should be interpreted with caution since the evidence-base presently exhibits a number of limitations. The two most important weaknesses identified were the poor documentation of the actual medio-lateral patella positioning test methods and the limited description of subject characteristics. Two distinct methods of assessing medio-lateral patellar position were recognised (Figs. 1 and 2). This review would suggest that Herrington’s (2002) methods appears to have better inter-tester reliability and criterion validity than McConnell’s (1986). This difference may account for the substantial difference in intra-tester results between the findings of Watson et al. (1999) and Herrington et al. (2006) and Herrington and Nester (2004), or inter-tester findings between Tomsich et al. (1996) and Herrington (2002). However, this cannot be categorically stated given the limited size of evidence presently available. Furthermore, with the exception of 12 subjects in Herrington’s (2008) study, all other studies which assessed Herrington’s (2002) method were undertaken on asymptomatic populations. As a result, it is not possible to generalise with confidence, these results to patients with patellofemoral disorders. Accordingly, a direct comparison of these two methods of assessing medio-lateral position is warranted to determine the optimal method of assessing patients with different patellofemoral disorders. A considerable weakness in the evidence-base is the poor description of the medio-lateral patellar position test. Studies did not demonstrate whether they controlled confounding factors such

as lower limb rotation, quadriceps contraction or knee flexion (McEwan et al., 2007; Herrington et al., 2006; Herrington, 2002, 2008; Herrington and Nester, 2004; Fitzgerald and McClure, 1995). These variables can influence the patellar position within the trochlear groove (Herrington and Pearson, 2008; Muhle et al., 1999). Variations in these factors between study methodologies may account for the differences in results between studies for intertester reliability. This should be considered when presenting the findings of similar studies in the future. The degree of knee flexion may be an important factor. Five papers (McEwan et al., 2007; Herrington et al., 2006; Herrington, 2002, 2008; Herrington and Nester 2004) assessed the knee in 20 degrees flexion, compared to full extension. Considering the patella engages with the femoral trochlear at approximately 10–30 degrees of knee flexion (Beasley and Vidal, 2004; Senavongse et al., 2003), it may be hypothesised that the good reliability results of these studies may be related to greater patellar osseous constraint compared to full extension. Further study may therefore be indicated to investigate this assumption assessing medio-lateral patellar position in full extension compared to difference ranges of tibiofemoral flexion. The papers reviewed poorly documented important subject characteristics, particularly in view of weight and height. Papers poorly distinguished between those subjects with patellar instability, patellofemoral pain syndrome or other patellofemoral disorders. Consequently, it was not possible to ascertain the heterogeneity of these papers. In response to this, it was deemed unstable to formally compare these paper’s results using a metaanalysis design. In addition, Egger et al. (2001) suggested that meta-analysis should not be undertaken for observational studies, which was the methodology design used in all the studies reviewed. It appeared that only Fitzgerald and McClure (1995) assessed medio-lateral patellar position in patients with subluxed patellar. The present evidence-base does not state whether the reliability or validity of this test is related to the severity of patellar displacement. Further study is recommended to investigate whether the accuracy of these measurements is related to the degree of patellar displacement in well-defined populations. With the exception of Herrington (2008), Herrington and Nester (2004) and Herrington et al. (2006), the assessors’ experience of the medial-lateral patellar position tests was well described. The assessors in each study appeared to have broadly similar levels of experience and training in each assessment method. However, it remains unclear whether the reliability or validity of the tests to

362

T.O. Smith et al. / Manual Therapy 14 (2009) 355–362

assess medio-lateral position is dependent on the method of teaching these measures. This may inform clinicians as to whether attendance at courses is warranted, or whether text and photographic teaching is sufficient to accurately assess this measurement. The results of this review would suggest that the intra-tester reliability of medio-lateral patellar orientation tests are satisfactory. Accordingly, clinicians can have some confidence that they have consistency in their measures between treatment sessions. It remains unclear whether the results of this measure are reproducible if patients change from one physiotherapist to another with such variable inter-tester reliability. Furthermore, further study is indicated to determine the clinical relevance of these findings. The reliability of physiotherapist’s abilities to translate their mediolateral displacement findings to specific taping placement to address abnormal patellar translation is presently unclear and should be addressed with future study. Although studies frequently described the age and gender of their subjects, only Fitzgerald and McClure (1995) and Herrington (2008) detailed the weight of their subjects to indicate body mass. Body mass may have influenced medio-lateral patellar position measurements. Fitzgerald and McClure (1995) acknowledged that it appeared more difficult to palpate the bony landmarks in those subjects with greater body mass. Accordingly, this may have contributed to variability in the measurement of these patients, and an additional variable which should be considered. 5. Conclusions The findings of this study indicate that the intra-tester reliability of medio-lateral patellar position tests are good, but that intertester reliability is variable. The criterion validity of these tests is at worse moderate. These are based on a limited evidence-base. Further study is recommended to compare the McConnell (1986) and Herrington (2002) methods of assessing medio-lateral patellar orientation. After this rigorous assessment, clinicians will then be better informed on the appropriateness of these tests when assessing patellar orientation in patients with patellofemoral disorders. Acknowledgements We thank the library staff at the Norfolk and Norwich University Hospital’s Sir Benjamin Gooch Library for their assistance in paper retrieval. We also thank Mr Mark Rowlands for his assistance with the photographs used in this paper. References Arendt EA, Fithian DC, Cohen E. Current concepts of lateral patella dislocation. Clinical Sports Medicine 2002;21:499–519. Beasley LS, Vidal AF. Traumatic patellar dislocation in children and adolescents: treatment update and literature review. Current Opinion in Paediatrics 2004;16:29–36. Critical Skills Appraisal Programme (CASP) Homepage on the Internet. Oxford, UK: Learning & Development Public Health Resource Unit; c. 2007. Available from: http://www.phru.nhs.uk/casp/critical_appraisal_tools.htm (accessed 1 May 2007). Crossley K, Bennell K, Green S, McConnell J. A systematic review of physical interventions for patellofemoral pain syndrome. Clinical Journal of Sports Medicine 2001;11(2):103–10.

Edwards A, Talbot R. The hard pressed researcher. A research handbook for the caring professions. London: Longman; 1994. Egger M, Davey Smith G, Schneider M. Systematic reviews of observational studies. In: Egger M, Davey Smith G, Altman DG, editors. Systematic reviews in health care. London: BMJ Books; 2001. p. 211–28. Evans R, Elwyn G, Edwards A. Review of instruments for peer assessment of physicians. BMJ 2004;328:1240. Fitzgerald GK, McClure PW. Reliability of measurements obtained with four tests for patellofemoral alignment. Physical Therapy 1995;75(2):84–92. Fulkerson JP, Shea KP. Disorders of patellofemoral alignment. Journal of Bone and Joint Surgery 1990;72-A:1424–9. Grelsamer RP, Newton PM, Staron RB. The medial-lateral position of the patella on routine magnetic resonance imaging: when is normal not normal? Arthroscopy 1998;14:23–8. Herrington LC. The inter-tester reliability of a clinical measurement used to determine the medial/lateral orientation of the patella. Manual Therapy 2002;7(3):163–7. Herrington L. The effect of corrective taping of the patella on patella position as defined by MRI. Research in Sports Medicine 2006;14:215–23. Herrington LC. The difference in a clinical measure of patella lateral position between individuals with patellofemoral pain and matched controls. Journal of Orthopaedic and Sports Physical Therapy 2008;38:59–62. Herrington L, Nester C. Q-angle undervalued? The relationship between Q-angel and medio-lateral position of the patella. Clinical Biomechanics 2004;19: 1070–3. Herrington L, Pearson S. The applicability of ultrasound imaging in the assessment of dynamic patella tracking: a preliminary investigation. The Knee 2008;15(2):125–7. Herrington L, Rivett N, Munro S. The relationship between patella position and length of the iliotibial band as assessed using Ober’s test. Manual Therapy 2006;11:182–6. Hughston JC. Subluxation of the patella. Journal of Bone and Joint Surgery 1968;50-A:1003–26. Insall J. Chondromalacia patellae: patellar malalignment syndrome. Orthopedic Clinics of North America 1979;10:117–27. McConnell J. The management of chondromalacia patellae: a long term solution. Australian Journal of Physiotherapy 1986;32(4):215–23. McConnell J. Rehabilitation and nonoperative treatment of patellar instability. Sports Medicine and Arthroscopy Reviews 2007;15:95–104. McEwan I, Herrington L, Thom J. The validity of clinical measures of patella position. Manual Therapy 2007;12:226–30. Mizuno Y, Kumagai M, Mattessich SM, Eilas JJ, Ramrattan N, Cosgarea AJ, et al. Q-angle influences tibiofemoral and patellofemoral kinematics. Journal of Orthopaedic Research 2001;19:834–40. Muhle C, Brossmann J, Heller M. Kinematic CT and MR imaging of the patellofemoral joint. European Radiology 1999;9(3):508–18. Ota S, Ward SR, Chen T-J, Tsai Y-J, Powers CM. Concurrent criterion-related validity and reliability of a clinical device used to assess lateral patellar displacement. Journal of Orthopaedic and Sports Physical Therapy 2006;36(9):645–52. Polgar S, Thomas SA. Introduction to research in the health sciences. 4th ed. London: Churchill Livingstone; 2000. Portney LG, Watkins MP. Foundations of clinical research. Applications to practice. 2nd ed. New Jersey: Prentice Hall; 2000. Powers CM, Mortenson S, Nishimoto D, Simon D. Criterion-related validity of a clinical measurement to determine the medial/lateral component of patellar orientation. Journal of Orthopaedic and Sports Physical Therapy 1999;29(7):372–7. Senavongse W, Farahmand F, Jones L, Andersen H, Bull AM, Amis AA. Quantitative measurement of patellofemoral joint stability: force-displacement behaviour of the human patella in vitro. Journal of Orthopaedic Research 2003;21:780–6. Sutlive TG, Mitchell SD, Maxfield SN, McLean CL, Neumann JC, Swiecki CR, et al. Identification of individuals with patellofemoral pain whose symptoms improved after a combined program of foot orthosis use and modified activity: a preliminary investigation. Physical Therapy 2004;84(1):49–61. Tolouei FM, Afshar A, Salarilak S, Sina A. CT patellar cortex tilt angle: a radiological method to measure patellar tilt. Iran Journal of Radiology 2005;3(1):17–21. Tomsich DA, Nitz AJ, Threlkeld AJ, Shapiro R. Patellofemoral alignment: reliability. Journal of Orthopaedic and Sports Physical Therapy 1996;23(3):200–8. Warden SJ, Hinman RS, Watson Jr MA, Avin KG, Bialocerkowski AE, Crossley KM. Patellar taping and bracing for the treatment of chronic knee pain: a systematic review and meta-analysis. Arthritis and Rheumatology 2008;15(1):73–83. Watson CJ, Propps M, Galt W, Redding A, Dobbs D. Reliability of McConnell’s classification of patellar orientation in symptomatic and asymptomatic subjects. Journal of Orthopaedic and Sports Physical Therapy 1999;29(7):378–85.

Manual Therapy 14 (2009) 363–368

Contents lists available at ScienceDirect

Manual Therapy journal homepage: www.elsevier.com/math

Original Article

Clinical measurement of craniovertebral angle by electronic head posture instrument: A test of reliability and validityq Herman Mun Cheung Lau a, Thomas Tai Wing Chiu a, *, Tai-Hing Lam b a b

Department of Rehabilitation Sciences, The Hong Kong Polytechnic University, Hung Hom, Hong Kong Department of Community Medicine, The University of Hong Kong, Hong Kong

a r t i c l e i n f o

a b s t r a c t

Article history: Received 13 August 2007 Received in revised form 6 May 2008 Accepted 19 May 2008

The study was a cross-sectional reliability study with the objective of assessing the reliability and validity of the Electronic Head Posture Instrument (EHPI) in measuring the craniovertebral (CV) angle for subjects with or without neck pain. Twenty-six subjects (mean age ¼ 36.88, SD  9.95) with chronic neck pain and 27 subjects (mean age ¼ 31.85, SD  7.63) without neck pain were recruited. The CV angle was measured by the EHPI which consists of an electronic angle finder, a transparent plastic base and a camera stand. Two therapists were recruited to assess the intra- and inter-rater reliability of the EHPI in two separate sessions of measurement. The difference in CV angle between the two groups was determined. The CV angle of the patient group (mean 43.94, SD  3.61) was significantly smaller (p < 0.001) than that of the normal group (mean 50.58, SD  2.09). Intra-rater (intra-class correlation coefficient (ICC) ranged from 0.86 to 0.94) and inter-rater (ICC ranged from 0.85 to 0.91) reliability of the EHPI in measuring CV angle for both groups of subjects were high. In conclusion the EHPI was found to be reliable and valid in measuring the CV angle for subjects with or without neck pain. Ó 2008 Elsevier Ltd. All rights reserved.

Keywords: Craniovertebral angle Reliability Validity

1. Introduction Neck pain is a common musculoskeletal disorder in the general population. A systematic review of neck pain over the world showed that the one-year prevalence ranged from 16.7% to 75.1% for the entire adult population with a mean of 37.2% (Fejer et al., 2006). In a recent telephone survey done in Hong Kong, Chiu and Leung (2006) reported that the 12 month prevalence was 53.6%. Posture of the head and neck has long been recognized as a factor contributing to the onset and perpetuation of cervical pain and dysfunction (Harrison et al., 2005; Persson et al., 2007). Forward head posture is one of the common poor head postures seen in patients with neck disorders (Hickey et al., 2000). A number of studies suggested that forward head posture predisposes individuals towards pathological conditions such as thoracic outlet syndrome and cervical spondylogenic changes (Rocabado, 1983; Ayub et al., 1984). There are many instruments used for assessing the head posture, such as the Rocabado Posture Gauge

q The various parts of the EHPI are readily available in the commercial market and we had no financial support from the manufacturers. No conflict of financial interest was involved in this measurement tool. * Corresponding author. Tel.: þ852 27666709; fax: þ852 23308656. E-mail address: [email protected] (T.T. Wing Chiu). 1356-689X/$ – see front matter Ó 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.math.2008.05.004

designed by Mariano Rocabado which is a T-shaped instrument. It measures the horizontal distance from the tangent of the most posterior thoracic spinous process to the most anterior cervical spinous process in standing position. The distance measured is the amount of the forward head posture of the subject (Willford et al., 1996). Moreover head posture can be measured by the plumb line and photographic imaging (Wilmarth and Hilliard, 2003). The use of plumb line is, however, limited to the subjective nature of determining the degree of forward head posture and discrepancies can be caused by the viewing angulation of the examiner in relation to the patient. Problems of photographic imaging are related to the time expenditure required for accurate assessment and it does not allow for immediate results (Wilmarth and Hilliard, 2003). One of the objective methods of assessing head posture is through measuring the craniovertebral (CV) angle. This is the angle between a horizontal line through the spinous process of C7 and a line from spinous process of C7 through the tragus of the ear (Raines and Twomey, 1994; Joe et al., 2003). Joe et al. (2003) studied the reliability of measuring the CV angle with Coutts overlay sheet (a transparent grid sheet for mapping) and a protractor in a lateral photograph in 29 female students, and the intra-rater reliability was very high. They also suggested that smaller CV angles indicate greater protraction of the head and larger angles are more representative of ‘ideal’ sagittal plane head/neck alignment. However, the result was limited to young female students only. In many

364

H.M. Cheung Lau et al. / Manual Therapy 14 (2009) 363–368

previous studies, investigators suggested that there were associations between the forward head posture and neck pain and disability (Griegel-Morris et al., 1992; Szeto et al., 2002). They found that those subjects with head, neck and shoulder discomfort or pain are more likely to have a smaller CV angle. In a recent study Yip et al. (2008) concluded that patients with small CV angle had a greater forward head posture and the greater the forward head posture, the greater the disability. Yip et al. (2008) also recommended that the CV angle could provide clinicians with further objective information on the disability and severity of patients with neck pain. An instrument, the Head Posture Spinal Curvature Instrument (HPSCI) was developed to measure both the CV angle and cervical curvature (Willford et al., 1996; Wilmarth and Hilliard, 2003). It was designed to provide more efficient assessment tool with immediate feedback. The HPSCI is a non-invasive, inexpensive measurement method which has been demonstrated to produce consistent and stable intra-rater results (intra-class correlation coefficient, ICC ¼ 0.9) across days and trials in 27 healthy subjects (Willford et al., 1996). However, its measuring scale was accurate to the whole number only; there is potential error in attempting to choose which way to read the scale when the indicator falls between two whole numbers. Despite the good intra-rater reliability, Wilmarth and Hilliard (2003) did not investigate the validity of the instrument which is essential for a clinical measuring tool, because strong reliability does not suggest strong validity (Portney and Watkins, 2000). Moreover, the inter-rater reliability in assessment of patients with neck pain has not been elucidated. Manual therapists utilize different methods to achieve correction of forward head posture in patients with neck pain (Harrison et al., 1996). Assessment of posture is an important component of evaluation and affects the design of a treatment regimen. However, clinical evaluation of head posture is generally based on the clinician’s subjective visual impression. Besides, it is difficult to compare patients with each other and to quantify the improvements. Clinicians are continually looking for easier, safer, more objective and reliable devices to measure the head posture which is not yet available in the literature. Hence, the Electronic Head Posture Instrument (EHPI) was developed. The measuring scale of this instrument is accurate to one decimal place and the reading can be taken automatically by the electronic sensor. The objective of this study was to examine the reliability and validity of the EHPI in subjects with and without neck pain.

months and without any treatment were recruited in the patient group. Any person who had history of fracture injury at the cervical and shoulder region or vertebral column, scoliosis, severe thoracic kyphosis, spasmodic torticollis, rheumatic disease, temporomandibular joint dysfunction, neurological motion disorder or back pain, or loss of standing balance were excluded. Explanation and informed consent were obtained from each subject. The project was approved by the University’s Review Board for Health Science Research involving Human Subjects. 2.2. Apparatus The CV angle was measured by the EHPI which consists of an electronic angle finder, a transparent plastic base and a camera stand. The electronic angle finder ‘SmartTool Angle Finder’ made by M-D Building Products (United States) was fixed on a transparent plastic base. The combined Angle Finder and the plastic base (now named as Angle Finder) were mounted on a tripod camera or video camera stand – HAMA ‘Gamma 74’ which was made in Germany (Fig. 1). The Angle Finder could be used to identify and digitally display degrees/percent slope quickly and pitch to 1/10 degree accuracy. Two parallel lines were marked on the opposite sides of the transparent base. The targeted markers were aligned with the two parallel lines of the plastic base simultaneously, in order to measure the CV angle (Fig. 1). The camera stand was used to adjust the height and the tilting angle of the Angle Finder to measure the CV angle. Measurement validity defines the extent to which an instrument measures what it is intended to measure (Portney and Watkins, 2000). It is therefore necessary to establish the validity of the EPHI to detect different angles of rotation from the horizontal. An ‘index table’, validated in angle measurement was used as a criterion to test the measurement validity of the Angle Finder and the steps were as follows. The Angle Finder was reset to ‘0 ’ and mounted to the index table as shown in Fig. 2. The index table was rotated counter-clockwise and then clockwise again such that the readout from Angle Finder was 0.0 . The pointer of the index table was set to 0 . After that, the index table was rotated clockwise manually by hand to 90 with stops over each 5 rotation. The readings from the index table and the Angle Finder were recorded at each step. The index table was then rotated counter-clockwise back to 0 with stops over on each 15 rotation with the reading recorded. 2.3. Procedure of measuring the CV angle

2. Materials and methods 2.1. Subjects This study was a cross-sectional reliability study with a convenient sampling from the out-patient clinic of the Physiotherapy Department in the Prince of Wales Hospital, Shatin, Hong Kong. In order to carry out a contrast-group comparison to investigate the validity of the EHPI, the present authors adopted the following approach for calculation of the sample size. Using a software package from Number Cruncher Statistical System (NCSS) – power analysis and sample size for Windows (1996), it was estimated that a total of 52 subjects, i.e. 26 subjects in each of the patient group and normal group, would be required (Hintze, 1996). The following parameters were used as inputs to the program: (1) 0.05 alpha; (2) 90% power; (3) two-sided alternative; (4) an effect size of 0.5 (this was chosen because an effect size of 0.5–0.7 was considered moderate by Cohen (1977)). Twenty-seven volunteers aged 19–53 years (mean 31.85 years, SD  7.63) and without neck pain during the past 6 months were recruited in the normal group (Ylinen et al., 2004). Twenty-six subjects aged 20–55 years (mean 36.88 years, SD  9.95) with diagnosis of more than one episode of neck pain during the past 3

The EPHI was put on the standardized marking on the floor and the HAMA stand was adjusted into the position until the bubble of the horizontal indicator and the central marking overlapped. The distance from the subject to the centre of the HAMA stand was standardized to 0.3 m while the distance between the operator and the HAMA stand was 0.5 m because this was the longest distance that the testers can reach. The subjects were asked to put on sportswear in order to expose their neck and the upper thoracic spine. They were also required to remove their socks and shoes. The seventh cervical (C7) spinous process was palpated and identified and an adhesive pin marker (Fig. 1) was attached over its midpoint of the most prominent part. The subject was then asked to stand with his/her left shoulder in front of the EPHI. Another pin marker was fixed at tragus of his/her left ear. The subject was instructed to stand comfortably with their weight distribution evenly on both feet and to keep their eyes looking straight ahead. He/she was then instructed to flex and extend the head for three times and then rest it in a comfortable position. A virtual line was drawn between the two pin makers from midpoints of the tragus to C7. The therapist adjusted the EHPI until the two indicator lines were aligned with the markers. The reading from the Angle Finder represented the CV angle, as seen in Fig. 1.

H.M. Cheung Lau et al. / Manual Therapy 14 (2009) 363–368

365

Fig. 1. Measuring CV angle by the EHPI. When two indicator lines were aligned with the markers (C7 spinous process and tragus), the reading from the Angle Finder would represent the CV angle.

Two physiotherapists with five years of clinical experience in treating patients with neck pain were involved in this experiment to test the intra- and inter-rater reliability of the EHPI. They were trained by one of the authors (HL) to operate the EHPI for measuring CV angles. Two sessions were arranged to measure the CV angle. The reading of each session was taken by two therapists, respectively. In each session, the subject was asked to stand up for the first therapist to take the measurement. He/she was then required to walk around the room and then resume standing for the second therapist, who was blinded to the results of the first measurement, to repeat the measuring procedure. The second session was arranged 7 days later, in order to minimize the memory effect of the therapists. This amounted to four trials in two sessions [i.e. 2(1 þ1)]. Each subject was informed to avoid any unusual activities within the 7 days between the two sessions in order to minimize the changes in his/her neck pain conditions. They were also required to report any change of the neck condition before the second session.

2.4. Statistical analysis 2.4.1. Validity Paired t-test and Pearson’s correlation test were used to examine if there was any significant difference and the correlation between the readings of the index table and the Angle Finder as a test of validity. 2.4.2. Inter/intra-tester reliability ICC Model 1, Form 1 was used to determine the intra-rater and Model 2, Form 1 for the inter-rater reliability of using the EHPI. It has been suggested that ICCs below 0.50 represent poor reliability, ICCs from 0.50 to 0.75 represent moderate reliability and ICCs above 0.75 indicate good reliability (Portney and Watkins, 2000).

Fig. 2. Validity test of smart tool with the index table. The Angle Finder was put on top of the arm of the index table. In the lower figure, the Angle Finder was moved along with the arm of the index table to 30.0 in order to check the measurement validity of the Angle Finder.

2.4.3. Descriptive analysis on normal group and patient group Independent sample t-tests were used to determine if there was any difference in the demographic characteristics and the CV angle between the normal group and the patient group. The minimal level of detectable change (MDC) was calculated according to the formula: standard error of measurement (SEM)  z-score at the

366

H.M. Cheung Lau et al. / Manual Therapy 14 (2009) 363–368

showed that the intra-rater (ICC ranged from 0.86 to 0.94) and inter-rater (ICC ranged from 0.85 to 0.91) reliability of the EHPI in measuring the CV angle for both groups of subjects were high. The SEM was 1.193 and the MDC was 3.31.

Table 1 Demographic characteristics of normal group and patient group. Study part

Normal group (n ¼ 27)

Patient group (n ¼ 26)

Age Mean (standard deviation) (years) Range (years)

31.9 (7.6) 19–53

36.5 (9.7) 20–55

3.4. CV angle of normal group vs patient group

Gender Male Female

n 12 15

n 11 15

The CV angle of the patient group (group mean 43.9, SD  3.6) was significantly smaller (p < 0.001) than that of the patient group (group mean 50.6, SD  2.1) (Table 3)

% 22.6 28.3

% 20.8 28.3

two-sided 95% confidence intervals (z ¼ 1.96)  O2, where SEM ¼ SD  O1  ICC (Beaton, 2000). The level of significance was set to be 0.05. SPSS version 14.0 program was used for statistical analysis.

4. Discussion 4.1. Validity The validity of the EPHI to detect different angles of rotation from the horizontal was demonstrated by the non-significant difference and the high correlation between the readings from the index table which acted as a standard for measurement. Also, the validity of a clinical measuring tool depends on its ability to differentiate the symptomatic subjects from non-symptomatic ones. The EHPI is also valid in this aspect which can be demonstrated by the contrast-group comparison as discussed in the later section.

3. Results 3.1. Descriptive analysis A total of 26 subjects (11 males, 15 females) were recruited in the patient group and 27 subjects (12 males, 15 females) in the normal group. The demographic characteristics of the subjects are shown in Table 1. It was found that there was no significant difference in the age (p ¼ 0.061) or gender (p ¼ 0.88) between the two groups.

4.2. Reliability 3.2. Validity Result of the paired t-test demonstrated that there was no significant difference (p ¼ 1.000) between the readings of the index table and the Angle Finder at different angles of rotation. They are highly correlated with each other as indicated by the result of Pearson’s correlation test (p ¼ 0.000, r ¼ 1.000) and the X–Y plot as shown in Fig. 3.

It was found that the intra- and inter-rater reliability of using the EHPI to measure the CV angle were high, which demonstrated that this CV angle measurement method was highly repeatable for both the normal group and the patient group. The inter-tester reliability results for trial 2 were better than that of trial 1; this may be due to the practice effect which can help to decrease the random errors of measurement (Portney and Watkins, 2000).

3.3. Reliability

4.3. Normal group vs patient group

Reliability of the EHPI was analyzed separately in the normal group and the patient group as shown in Table 2. The results

The contrast-group comparison of the present study confirmed that the CV angle of the non-symptomatic subjects as measured by

100.00

Smart Tool

80.00

60.00

40.00

20.00 R Sq Linear =1

0.00 0.00

20.00

40.00

60.00

80.00

100.00

Index Table Fig. 3. X–Y plot of index table and Angle Finder. Validity of the Angle Finder in correlation to the index table is shown where the linear line represented that these two tools were highly correlated with each other.

H.M. Cheung Lau et al. / Manual Therapy 14 (2009) 363–368

367

Table 2 Intra-rater and inter-rater reliability of therapists A and B in measuring CV angle for subjects of patient group and normal group. Patient group ICC (95% CI) n ¼ 26

Normal group ICC (95% CI) n ¼ 27

Intra-rater reliability

Therapist A Therapist B

0.87 (0.73–0.94) 0.94 (0.86–0.97)

0.87 (0.74–0.94) 0.86 (0.72–0.93)

Inter-rater reliability

First trial Second trial

0.86 (0.71–0.94) 0.91 (0.83–0.96)

0.85 (0.71–0.93) 0.88 (0.76–0.95)

the EHPI was significantly smaller than that of the patients with neck pain. This concurred with the previous findings that patients with neck pain have a forward head posture [bad posture] (Edmondston et al., 2007; Cho, 2008; Yip et al., 2008) as compared to non-symptomatic subjects (Haughie et al., 1995; Hickey et al., 2000). It is well accepted that forward head posture predisposed individual towards pathological condition over the cervical and cranial region. Johnson (1998) suggested that persisted forward head posture might increase loading of the cervical joints and abnormal stress on the posterior cervical structures and cause myofascial pain. However, the exact mechanism is still unknown, further study in this aspect is warranted. As there was no statistically significant difference in terms of the age and gender between the two groups, these two factors should not contribute to the difference in the CV angle between the two groups. Raines and Twomey (1994) reported that, regardless of the difference in population samples, the CV angle is a reliable indicator of variation in the head and neck posture. To be clinically useful and valid an instrument should be able to distinguish normal subjects from the patients. The present study has demonstrated significant difference in the CV angle between the patient group and the normal group. This helps strengthen the validity of the EHPI in clinical measurement. The MDC for the measurement of CV angle by the EHPI was 3.31. This implies that when the CV angle measured by the EHPI demonstrated a difference of 3.31 or more, the therapist can be confident that there is a real change of the CV angle and not just a measurement error. In comparison with previous studies (Joe et al., 2003; Wilmarth and Hilliard, 2003; Yip et al., 2008) the present study achieved the following improvements: the automatic display of all the measurement data in the Angle Finder helped reduce the possibility of transcription and calculation errors. In addition, all the measurement data were digitalized to one decimal place. This helped minimize the inaccuracies caused when reading an analogue scale where the indicator falls between two whole numbers. Moreover, the Angle Finder can quickly identify any angle with good accuracy within 0.1. 4.4. Clinical implications Clinicians always try to correct their patients’ forward head postures by various treatment approaches. In order to assess the

effectiveness of these approaches, it is vital to develop an objective method to measure the posture of the patients. The device involved should be practical, user-friendly, reliable and objective in its quantification. The EHPI serves the above purpose as it is an inexpensive, portable and convenient apparatus that allows clinicians to collect an accurate, reliable and objective reading. The instrument provides clinicians with additional useful and reliable information to monitor patients’ condition and progression. 4.5. Limitations of the study The sagittal plane position of the cervical, thoracic or lumbar spine was not measured. This was one of the limitations of this study because the CV angle depends on the relative position of the entire spine. It must be borne in mind that the CV angle is an index reflective of only one part of the total picture of the cervical and head posture. Accurate assessment of complete head and neck posture requires a cephalometric radiographic analysis which was not available in this study. Moreover, further study may be required to elucidate if there are any differences when the subject is in a standing vs a sitting position for analysis of head posture. Due to the limited number of subjects who participated in the study, and the possibility of selection bias in the recruitment of normal subjects, the CV angle measurement results cannot be regarded as a baseline reference for the particular group. In addition, the age group of this study was limited to 19–55 years; therefore further investigation with larger sample size involving different age groups is indicated. Also, responsiveness of the CV angle as measured by the EHPI should be tested in a longitudinal study for subjects with neck pain. 5. Conclusion We have demonstrated that the EHPI was valid in measuring the CV angle. There was a high degree of test–retest reliability in measuring the CV angle for both the normal subjects and those with neck pain. The CV angle of subjects with neck pain was significantly smaller than that of the normal subjects. With increasing demand being placed on evidence-based practice, a valid and reliable outcome measuring tool should contribute to better clinical service and evaluation.

Table 3 CV angle of subjects in normal group and patient group measured by therapists A and B. Minimum CV angle

Maximum CV angle

Individual mean  SD

Group mean  SDa

Normal (n ¼ 27)

First session therapist A Second session therapist A First session therapist B Second session therapist B

47.4 46.3 47.1 46.5

56.7 57.0 55.4 55.5

51.02.4 51.22.3 50.12.1 50.02.1

50.62.1

Patient (n ¼ 26)

First session therapist A Second session therapist A First session therapist B Second session therapist B

33.7 35.3 35.6 33.7

49.5 48.2 49.5 50.1

43.83.4 44.13.6 44.03.7 43.93.8

43.93.6

Normal/patient

a

Group mean is the mean CV angle measured by both therapists in two sessions of each group.

368

H.M. Cheung Lau et al. / Manual Therapy 14 (2009) 363–368

References Ayub E, Glasheen-Wray M, Kraus S. Head posture: a case study of the effects on the rest position of the mandible. Journal of Orthopaedic and Sports Physical Therapy 1984;5:179–83. Beaton DE. Understanding the relevance of measured change through studies of responsiveness. Spine 2000;25:3192–9. Chiu TW, Leung SL. Neck pain in Hong Kong: a telephone survey on prevalence, consequences and risk groups. Spine 2006;31(16):E540–4. Cho CY. Survey of faulty postures and associated factors among Chinese adolescents. Journal of Manipulative and Physiological Therapeutics 2008;31(3):224–9. Cohen J. Statistical power analysis for the behavioural sciences. New York: Academic Press; 1977. Edmondston SJ, Chan HY, Chi Wing Ngai G, Warren MLR, Williams JM, Glennon S, et al. Postural neck pain: an investigation of habitual sitting posture, perception of good posture and cervicothoracic kinaesthesia. Manual Therapy 2007;12(4):363–71. Fejer R, Kyvik K, Hartvigsen J. The prevalence of neck pain in the world population: a systemic critical review of the literature. European Spine Journal 2006;15: 834–48. Griegel-Morris P, Larson K, Mueller-Klaus K, Oatis CA. Incidence of common postural abnormalities in the cervical, shoulder, and thoracic region and their association with pain in two age groups of healthy subjects. Physical Therapy 1992;72:425–31. Harrison AL, Barry-Greb T, Wojtowicz GW. Clinical measurement of head and shoulder posture variables. The Journal of Orthopaedic and Sports Physical Therapy 1996;23(6):353–61. Harrison DE, Haas JW, Cailliet R, Harrison DD, Holland B, Janik TJ. Concurrent validity of flexicurve instrument measurements: sagittal skin contour of the cervical spine compared with lateral cervical radiographic measurements. Journal of Manipulative and Physiological Therapeutics 2005;28: 597–603. Haughie L, Fiebert IM, Roach KE. Relationship of forward head posture and cervical backward bending to neck pain. Journal of Manual and Manipulative Therapy 1995;3:91–7.

Hickey ER, Rondeau MJ, Corrente JR, Abysalh CJ. Reliability of the cervical range of motion (CROM) device and plumb line techniques in measuring resting head posture (RHP). Journal of Manual and Manipulative Therapy 2000;8:10–7. Hintze JL. Power analysis and sample size for Windows. PASS. Kaysville, Utah, USA: NCSS; 1996. Joe B, Elizabeth B, Aoife NM. Reliability of measuring natural head posture using the craniovertebral angle. Irish Ergonomics Review 2003:37–41. Johnson GM. The correlation between surface measurement of head and neck posture and the anatomic position of the upper cervical vertebrae. Spine 1998; 23:921–7. Persson P, Hirschfeld H, Nilsson-Wikmar L. Associated sagittal spinal movements in performance of head pro- and retraction in healthy women: a kinematic analysis. Manual Therapy 2007;12:119–25. Portney LG, Watkins MP. In: Foundations of clinical research applications to practice. 2nd ed. Upper Saddle River, New Jersey: Prentice Hall Health; 2000. p. 61–77. Raines S, Twomey L. Posture of the head, shoulders and thoracic spine in comfortable erect standing. Australian Journal of Physiotherapy 1994;40:25–32. Rocabado M. Biomechanical relationship of the cranial, cervical and hyoid regions. Journal of Cranio-mandibular Practice 1983;1:1–6. Szeto G, Straker L, Raines S. A field comparison of neck and shoulder posture in symptomatic and asymptomatic office workers. Applied Ergonomics 2002; 33(1):75–84. Willford G, Kisner C, Glenn TM, Sachs L. The interaction of wearing multifocal lenses with head posture and pain. The Journal of Orthopaedic and Sports Physical Therapy 1996;23(3):194–9. Wilmarth MA, Hilliard TS. Measuring head posture via the craniovertebral angle. Orthopaedic Physical Therapy Practice 2003;14:13–5. Yip CHT, Chiu TTW, Poon ATK. The relationship between head posture and severity and disability of patients with neck pain. Manual Therapy 2008;13(2):148–54. Ylinen J, Salo P, Nykanen M, Kautiainen H, Hakkinen A. Decreased isometric neck strength in women with chronic neck pain and the repeatability of neck strength measurements. Archives of Physical Medicine and Rehabilitation 2004;85:1303–8.

Manual Therapy 14 (2009) 369–374

Contents lists available at ScienceDirect

Manual Therapy journal homepage: www.elsevier.com/math

Original Article

Mobilizations of the asymptomatic cervical spine can reduce signs of shoulder dysfunction in adults Lynda McClatchie a, Judi Laprade b, Shelley Martin c, Susan B. Jaglal a, b, Denyse Richardson a, Anne Agur d, * a

Graduate Department of Rehabilitation Sciences, University of Toronto, 500 University Avenue, Toronto, Ontario, Canada M5G 1V7 Department of Physical Therapy, University of Toronto, 160-500 University Avenue, Toronto, Ontario, Canada M5G 1V7 Spine and Sport Physiotherapy Centre, 123 Edward Street, Suite 500, Toronto, Ontario, Canada M5G 1E2 d Division of Anatomy, Department of Surgery, University of Toronto, Medical Sciences Building, 1 King’s College Circle, Toronto, Ontario, Canada M5S 1A8 b c

a r t i c l e i n f o

a b s t r a c t

Article history: Received 12 July 2007 Received in revised form 18 February 2008 Accepted 19 May 2008

Generalized shoulder pain is a common problem that is difficult to treat and frequently recurrent. The asymptomatic cervical spine must be ruled out as a cause of any shoulder pain, as it can have a similar presentation to an isolated shoulder disorder. Previous studies have shown that lateral cervical glide mobilizations to the asymptomatic cervical spine at C5/6 can affect peripheral pain, but none have examined shoulder pain. A randomized, blinded, placebo-controlled, cross-over trial was used to examine the immediate effects of cervical lateral glide mobilizations on pain intensity and shoulder abduction painful arc in subjects with shoulder pain. Twenty-one subjects received interventions of both cervical mobilization and placebo over two sessions. Pain intensity using a visual analog scale (VAS) and painful arc were assessed prior to and following application of cervical mobilization or placebo intervention. Evaluation of cervical mobilization revealed the shoulder abduction painful arc (12.5  15.6 , p ¼ 0.002) and shoulder pain intensity (1.3  1.1 cm, p < 0.001) were significantly decreased. The results of this study suggest that any immediate change in shoulder pain or active shoulder range of motion following cervical mobilizations indicate that treatment directed toward the asymptomatic cervical spine may expedite recovery. Ó 2008 Elsevier Ltd. All rights reserved.

Keywords: Manual therapy Mobilization Radiculopathy Painful arc

1. Introduction The prevalence of shoulder pain as reported in the literature ranges from 7 to 34% in the adult population (Chard et al., 1991; Badley and Tennant, 1992; Van der Windt et al., 1996). Despite its prevalence, the pathophysiology of shoulder pain has not been well defined (de Winter et al., 1999; Groenier et al., 2003), and the diagnosis is complicated by the large number of structures in the shoulder region (Bamji et al., 1996; Pope et al., 1997). Differentiation between various shoulder disorders is important to implementing effective treatment (Green et al., 1998), but the differential diagnosis is often complex as a broad spectrum of intrinsic and extrinsic conditions may produce shoulder pain (Neviaser, 1983). The tendons of the rotator cuff or long head of biceps (Lyons and Orwin, 1998; Van der Heijden, 1999), calcium deposits within tendons (Turner-Stokes, 1996), degenerative changes in the glenohumeral or acromioclavicular joint (Prescher, 2000), or inflammation in the bursae surrounding the shoulder (Koester et al.,

* Corresponding author. Tel.: þ1 416 978 8855; fax: þ1 416 978 3844. E-mail address: [email protected] (A. Agur). 1356-689X/$ – see front matter Ó 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.math.2008.05.006

2005) are some of the structures which cause pain around the shoulder. Generalized shoulder pain is frequently recurrent (Winters et al., 1999) and can remain present for over three years after onset (Chakravarty and Webley, 1993; Croft et al., 1996; Badcock et al., 2002), with one study indicating that 50% of subjects with shoulder pain had persistent problems three years later (MacFarlane et al., 1998). Research has shown that long-term shoulder pain can lead to a considerable restriction of work and leisure activities (Wells, 1982). Persistence of the problem can lead to lengthy treatment in physiotherapy and frequent general practitioner appointments (Wells, 1982). Shoulder pain can arise from a cervical spine disorder, and therefore the neck must be ruled out as a potential cause of pain (Manifold and McCann, 1999). Radiculopathy arising from the cervical spine is difficult to differentiate from localized shoulder pathology because the sensory distribution extends from the base of the neck to the outer edge of the shoulder (Wilson, 2005). Sobel et al. (1997) found that restricted mobility in the cervicothoracic spine in patients with shoulder pain did not seem to recover significantly after 26 weeks. These investigators suggested that if intervention at the cervicothoracic spine was included in

370

L. McClatchie et al. / Manual Therapy 14 (2009) 369–374

management of patients with shoulder pain, the tendency for shoulder disorders to recur could be decreased. Cervical spine mobilization techniques can be used during therapy to affect more peripheral symptoms (Vicenzino et al., 1996; Sterling et al., 2001). Previous studies have shown that chronic elbow pain and temporomandibular joint (TMJ) pain can be reduced by a mobilization intervention to the asymptomatic neck (Vicenzino et al., 1996; Stiesch-Scholz et al., 2003). However, utilizing this technique for the treatment of shoulder symptoms has rarely been studied (Bergman et al., 2004). Bergman et al. (2004) examined the effectiveness of manual therapy directed toward the cervicothoracic spine and adjacent ribs for patients with shoulder pain. Participants were randomized into groups receiving manual therapy in addition to usual medical care (advice, analgesics, nonsteroidal anti-inflammatory drugs), or usual medical care alone. Pain radiating to the neck region was not criterion for exclusion, and the manipulative therapy included specific mobilizations or manipulations at the discretion of the therapist. The primary outcome measure was patient-perceived recovery, which was recorded on a seven point scale. Other outcomes included the severity of shoulder pain and functional disability. The results showed that at 12, 26 and 52 weeks follow-up, the manual therapy group demonstrated significant improvement for all outcome measures. A link between the asymptomatic cervical spine and the painful shoulder has not been clearly established in the literature. The objective of this study was to determine whether the utilization of cervical lateral glide mobilizations of the asymptomatic cervical spine at C5, C6, and C7 could immediately reduce the intensity and/ or range of the painful arc in subjects with generalized shoulder pain who were previously unresponsive to traditional shoulder treatment.

2. Methods Approval for this study was granted by the University of Toronto Research Ethics Board (Protocol reference #14011).

2.1. Subjects Subjects were recruited from a private orthopaedic physical therapy practice in Toronto, Canada, from August 2005 through May 2006. Each subject was currently being treated for generalized unilateral shoulder pain. Both male and female subjects were included in this study if they were age 18 or older, had an insidious onset of unilateral shoulder pain of at least six weeks duration, demonstrated a painful arc with shoulder abduction, and had no current or previous complaints of neck pain within the past year. Patients with shoulder pain were excluded if they had symptoms of paresthesia or neurological deficits, previous surgery or dislocation of the affected shoulder, clinically definitive arthritis of the shoulder on X-ray or had a cortisone injection for the current episode of shoulder pain. Prior to entering the study, subjects must have been unresponsive to 2–4 recent physiotherapy sessions addressing shoulder pain through ‘‘traditional’’ methods of movement patterns, strengthening and modalities such as ultrasound and cryotherapy. The short duration of shoulder treatment to patients prior to admittance into this study was purposeful, as a change in treatment direction would likely be warranted if the patient’s shoulder signs and symptoms had not changed following 2–3 weeks of treatment and home exercises. A sample size calculation of 21 subjects was determined using data from the first 12 subjects to detect a statistically significant difference in shoulder pain with 80% power.

2.2. Study design This was a randomized placebo-controlled cross-over trial. It was an exploratory pilot study to determine if a subsequent RCT would be appropriate. Subjects were randomized by a coin toss to receive either the cervical lateral glide mobilization or the placebo intervention during the first session. The subject returned for the second session within four days and underwent the same measurement protocol, but the second examiner performed the intervention that was not received during the first session. All outcome measures were assessed both before and after the intervention and conducted by the first investigator, who was blinded to which treatment intervention was received. The second examiner performed the predetermined cervical lateral glide mobilization or placebo treatment condition. Informed consent was obtained as per protocol, and all participants were required to attend two sessions of approximately 40 min duration (Fig. 1). 2.3. Intervention 2.3.1. Technique of lateral cervical glide mobilization A lateral cervical glide mobilization was the technique chosen for this study, with the subject seated and the thoracic spine resting against the back of the chair, head in a neutral position, feet resting flat on the floor, and arms relaxed with hands in their lap. The lateral aspect of the spinous processes of C5, C6, and C7 was landmarked on the ipsilateral side of the subject’s painful shoulder. The examiner’s thumb remained on the lateral aspect of the spinous process of C5, with the opposite hand placed on the subject’s non-affected shoulder or head for counterbalance as a lateral movement toward the non-painful side was applied with the mobilizing hand (Mulligan, 1995) (Fig. 3). Mobilizations were conducted for 2 min each at C5, C6 and C7, with small amplitude end range movements (Grade IVþ). The placebo treatment condition involved the examiner resting her hands in the same positions as the mobilization technique, but without the application of external force. 2.3.2. Outcomes measured A 10 cm visual analog scale (VAS) for pain measurement was completed by the subject following the shoulder abduction trials. The VAS has been shown to be a valid, reliable and responsive measure of a subject’s perceived level of pain (Price et al., 1994; Guerra de Hoyos et al., 2004). Active cervical spine range of motion in all planes (flexion, extension, bilateral side-bending) was measured using a cervical range of motion goniometer (CROM), which has been shown to be a valid and reliable tool for the measurement of cervical range of motion in the sagittal and frontal planes (Ordway et al., 1997; Tousignant et al., 2000). A Myrin goniometer was used to measure bilateral cervical rotation, and has been shown to be a reliable tool for measurement in the transverse plane (Balogun et al., 1989; Malstrom et al., 2003). Manual muscle testing of shoulder abduction at 90 was tested on the unaffected side followed by the affected side using an electrodynamometer. The resistance was applied by the examiner proximal to the subject’s elbow joint. Active shoulder range of motion into abduction was performed to determine the presence and extent of a painful arc. Reflective adhesive stickers were placed in five locations on the subject: sternal notch, anterior tip of the right and left shoulder at the acromion process, affected side elbow crease and proximal wrist crease. These points were used as landmarks from which the angle of the painful arc within the shoulder abduction range of motion could be accurately measured. Shoulder abduction was videotaped with the subject facing the video camera so that the reflective markers were clearly seen. The

L. McClatchie et al. / Manual Therapy 14 (2009) 369–374

371

Subject 1

Day 1

Objective Measurements

Randomization

Day 2 Cervical AROM Abduction MMT FHP Arc of Pain

Objective Measurements

Placebo Condition

Mobilization Condition

Re-measure

Re-measure

Fig. 1. Summary of methodology.

starting position for this test was with the palm of the affected arm facing outward (external rotation of the shoulder) and the arm by their side. While performing active shoulder abduction, the subject kept the thumb on the uninvolved side down until the start of their shoulder pain, then kept the thumb raised until the end of the painful arc (Fig. 2A and B). The digital video camera (Samsung SCD353, China) was mounted on a standard adjustable tripod at a standard position 220 cm from the subject. The beginning and end of the arc of pain with shoulder abduction was later quantified using software (Virtual Dub freeware) to identify the precise frames in which the subject began to raise and lower their thumb on the uninvolved side to indicate the start and end of their shoulder pain,

respectively. The abduction angle was determined on those chosen frames by using the measuring tool in Adobe PhotoshopÔ CS2. Cervical AROM was performed first, followed by shoulder abduction manual muscle testing. The subject was then positioned in front of the camera with adhesive markers as outlined previously to perform shoulder abduction. Three trials of shoulder abduction were recorded, with the subject indicating the start and end of the arc of pain each time. The measurements from the painful arc from each of the trials were averaged together to represent the true arc of pain. Subjects indicated the intensity of their pain during shoulder abduction on the VAS. Following these measures, the predetermined intervention (i.e. mobilization or placebo) was then performed by the second examiner as outlined previously. Subsequent to this intervention, the first examiner returned for re-measurement of all pre-intervention outcome measures. A second VAS was then completed by each subject to indicate the intensity of post-intervention shoulder pain. After the completion of the second session, each subject was

Fig. 2. Measurement of painful arc during shoulder abduction.

Fig. 3. Positioning for cervical lateral glide mobilization.

372

L. McClatchie et al. / Manual Therapy 14 (2009) 369–374

asked by the examiner to describe the perceived difference between the two treatment techniques. Answers were recorded on subject data sheets. 2.4. Data analysis The Statistical Package for the Social Sciences (SPSS) version 14.0 was used for data analysis. Paired t-tests were used to evaluate prepost mobilization as well as pre-post placebo for the following outcomes: cervical spine range of motion, shoulder abduction manual muscle testing, shoulder pain intensity and shoulder range of motion. A paired t-test was used to determine if there was a significant difference (p < .05) between the mean averages of the placebo and cervical mobilization conditions. A Pearson correlation test was used to determine if a relationship existed between shoulder abduction arc of pain and shoulder pain intensity. 3. Results 3.1. Subjects Twenty-one subjects (14 females, seven males) with an average age of 49.8 (9.8) years volunteered and consented to participate in the study. Forty-three percent of subjects reported a unilateral shoulder problem, and 57% reported a previous resolved neck problem more than one year prior to commencing participation in the study. Seven subjects were randomized to the mobilization condition during the first session, while 14 subjects received the placebo condition (Fig. 1). Each subject recognized that the mobilization and placebo interventions were different from each other, however, no subject realized that the placebo intervention was not therapeutic. 3.2. Pain measures There was a significant difference (p < 0.001) in the intensity of shoulder pain as measured by VAS before and after the mobilization treatment condition (Table 1). A 1.3  1.1 cm decrease in the mean VAS score post-mobilization was recorded when compared with the pre-mobilization pain score. Eighteen subjects (86%) showed an average decrease of 1.5 cm of shoulder pain postmobilization condition. Following the placebo condition, the 0.2  0.6 cm decrease in the mean VAS score was not statistically significant (p ¼ 0.078) (Table 1). Using a paired t-test, the mobilization and placebo conditions were found to be significantly different (p ¼ 0.0002), with a treatment effect of 1.038. The radius of the arc of pain with shoulder abduction diminished significantly after both the mobilization (12.5  15.6 , p ¼ 0.002) and the placebo (8.8  12.7, p ¼ 0.005) conditions (Table 1). In comparing pre- and post-mobilization shoulder arc of pain to pre- and post-mobilization VAS scores, 14 subjects (66.7%)

Table 1 Pre-post condition (placebo and mobilization) scores. Pre-condition mean  SD VAS – placebo (cm) 3.5  2.3 VAS – mobilization 3.7  2.0 (cm) Arc of pain – placebo 31.4  22.3 ( ) Arc of pain – 33.0  21.6 mobilization ( )

Post-condition mean  SD

Mean difference  SD

P* [twotailed]

3.2  2.5 2.4  2.1

0.2  0.6 1.3  1.1

0.078 <0.001

22.6  18.0

8.8  12.7

0.005

20.5  17.6

12.5  15.6

0.002

Abbreviation: SD, standard deviation; * P-value of two-tailed t-test. Paired samples significance.

showed a decrease in the arc of pain with shoulder abduction of 21.2  12.49 with a concurrent decrease in VAS score averaging 1.4  0.9 cm. The post-mobilization painful arc with shoulder abduction and post-mobilization decrease of shoulder pain intensity showed a moderate correlation (r ¼ 0.595). 3.3. Physical measures There was no significant difference in pre-post placebo cervical flexion (1.4  5.3 , p ¼ 0.242), extension (0.5  5.5 , p ¼ 0.693), right side-bending (0.1  5.3 , p ¼ 0.595), left sidebending (0.3  4.4 , p ¼ 0.769), right rotation (0.4  5.9 , p ¼ 0.769), or left rotation (0.3  4.9 , p ¼ 0.793). There was no significant difference in the pre-post mobilization cervical flexion (1.2  6.5 , p ¼ 0.410), extension (0.8  5.5 , p ¼ 0.534), right p ¼ 0.533), left side-bending side-bending (0.7  5.2 , (0.4  4.1, p ¼ 0.636), right rotation (1.1  4.4 , p ¼ 0.251), or left rotation (1.3  6.5 , p ¼ 0.359). Shoulder abduction strength was not significantly different following the placebo condition (0.4  0.9 kgf, p ¼ 0.057), nor following the mobilization condition (0.01 1.1 kgf, p ¼ 0.984). 4. Discussion This is the first study to show that cervical lateral glide mobilizations can immediately decrease the intensity of shoulder pain beyond a placebo effect, with movement into shoulder abduction in the likely absence of shoulder pathology. Eight subjects (38.1%) showed a decrease in the intensity of shoulder pain of more than 1.3 cm following the mobilization condition, compared to one subject (4.8%) following the placebo intervention. The 1.3 cm difference on VAS score following the cervical lateral glide mobilization is statistically significant, and also indicates a clinically relevant change, as the minimal clinically important difference has been shown to be 1.3 cm (Bird and Dickson, 2001; Gallagher et al., 2001). No studies have been found to date which examine the effect of cervical lateral glide mobilizations on shoulder pain, but there are similar studies involving the elbow. A study by Stiesch-Scholz et al. (2003) demonstrated an immediate positive effect on TMJ pain following cervical mobilizations. However, they did not examine the cervical spine for concurrent symptoms. Vicenzino et al. (1996) examined 15 subjects with chronic elbow pain that were previously unresponsive to conventional elbow treatment, none of whom reported neck pain. A significant decrease in elbow pain and increase in grip strength was noted immediately following the application of cervical lateral glide mobilizations at the C5/6 level with the subject supine. The cervical lateral glide mobilization technique in the study by Vicenzino et al. (1996) was performed in a supine position, while the mobilizations in the current study were performed in a weight-bearing position. This choice of subject positioning in the current study was supported by Mulligan (1995), who found that any gains achieved by mobilizations in supine may be diminished or lost upon returning to an upright position. It is likely that in a clinical setting, patients would present with various types of shoulder area pain, so the inclusion criteria in the current study were broad to allow for generalizability of results. The subjects included were non-symptomatic in their cervical spine with no limitation in cervical range of motion during the study, but a treatment effect with shoulder pain beyond placebo was demonstrated following a cervical lateral glide mobilization at C5, C6 and C7. In assessing joints such as the shoulder, the common practice is to examine the joints above and below the localized pain during an assessment to rule out involvement from those areas (Magee, 1987; Richardson and Iglarsh, 1994). It may not be appropriate to only use tissue-based reasoning, since that alone does not

L. McClatchie et al. / Manual Therapy 14 (2009) 369–374

always accurately indicate the true source of the problem and justify the course of management (Aina and May, 2005). The results of the current study indicated that a statistically and clinically significant decrease in the intensity of shoulder pain was observed post-mobilization condition, which lends credence to the argument that the cervical spine might still be involved in shoulder pain in the absence of any objective cervical limitations or symptom reproduction. If the asymptomatic cervical spine was examined thoroughly through manual therapy techniques during the assessment of a painful shoulder, the patient might be able to determine any immediate change in the intensity of shoulder pain or improvement in active shoulder range of motion. The patient should take note of any existing pain while initially performing shoulder abduction, and notice the intensity of shoulder pain during abduction again after a mobilization technique has been performed. This would allow for any immediate changes in shoulder pain to be observed. If any increases or decreases of pain intensity or shoulder range of motion is noted, the clinician can be more confident that a cervical component to that patient’s pain likely exists. Treatment could be planned accordingly to involve the cervical spine. There was no change in cervical range of motion before and after either the placebo condition or the mobilization condition. Since cervical movement was not limited at the onset, no significant increase or decrease was expected. If the cervical spine did exhibit any initial limitations in range of motion, it would likely have been addressed immediately as part of therapy for the subject’s shoulder pain. Similarly, no change in shoulder strength was noted following the placebo or mobilization interventions. This might suggest that neither the shoulder abductor muscles nor the C5 myotome was affected. The significant decrease in perceived intensity of shoulder pain immediately following the cervical lateral glide mobilization may be indicative of cervical spine involvement without concurrent clinically significant shoulder abduction weakness. The outcome measures of cervical range of motion and shoulder abduction strength were included in this study to demonstrate that those objective measurements may not show limitations during a clinical shoulder assessment, but mobilizations to the asymptomatic cervical spine can still alter shoulder symptoms. A significant decrease in the radius of the arc of pain with shoulder abduction was noted both post-placebo and post-mobilization condition. It could be inferred into clinical practice that there may be a decrease in the intensity of shoulder pain with a concurrent decrease in the radius of the painful arc with shoulder abduction. In those patients who present with an arc of pain with shoulder abduction, cervical mobilization techniques to the asymptomatic cervical spine may serve to decrease both the intensity of shoulder pain and the radius of the abduction arc of pain. There have been no other studies found which have examined the change in a painful arc in shoulder abduction following a cervical intervention, but there is existing research examining improvement in shoulder external rotation range of motion following cervical mobilizations. Schneider (1989) reported significantly increased shoulder external rotation range of motion following cervical mobilizations in patients with suspected capsular contractures of the glenohumeral joint. It was proposed by Schneider (1989) that the restriction in shoulder movement was likely not capsular but perhaps due to cervical somatic structures referring pain to the shoulder region and initiating spasm in shoulder musculature. It was also suggested that the improvement in shoulder movements following cervical mobilization may have had a neurological basis by positively affecting a nerve root impingement. A placebo intervention was included in this study to strengthen the internal validity (Tundle, 2006) and provide a control group. Placebo has been defined as an intervention used in a clinical trial

373

that is administered with the intention of mimicking some other intervention so that a comparison can be made (Vickers and de Craen, 2000). Placebos are common in drug studies, and have been used previously for physical interventions such as ultrasound, spinal manipulation and surgery (Vickers and de Craen, 2000). The placebo response is complex, and is thought to be influenced by patient expectation and the enthusiasm and belief of the therapist (Vicenzino et al., 1996). In this study, all subjects were asked to differentiate between the two interventions after both treatment sessions had been completed. Interestingly, each subject recognized that the mobilization intervention and the placebo condition were different from each other, but no subject realized that the placebo condition was not therapeutic. Subjects described the mobilization condition as more aggressive or less comfortable than the placebo condition. This exploratory pilot study has demonstrated that a cervical lateral glide mobilization can provide a clinically significant improvement in shoulder beyond that of placebo. This suggests that despite the lack of objective findings at the cervical spine upon assessment, it can be a source of shoulder pain. Clinicians choose appropriate treatment techniques to improve pain or movement in an affected area, so a clinician should not rule out the cervical spine as a source of non-specific shoulder pain, especially when the patient has been non-responsive to shoulder treatment. The limitations to this study include the small sample size and the use of a standardized treatment intervention. The number of subjects recruited may not be representative of the population that might present with generalized shoulder pain. As well, in order to maintain consistency among subjects, one type of cervical mobilization was utilized. Clinically, several different mobilizations would likely be attempted to determine which was most effective to decrease pain or increase range of motion in the affected area. Perhaps if the mobilization techniques were more individualized for each subject, a larger change in shoulder pain might have been seen following the mobilization condition. 5. Conclusion This study has shown that cervical lateral glide mobilizations can decrease the intensity of shoulder pain with movement into shoulder abduction, and this treatment effect is beyond placebo. It should be emphasized that the design of this study was an exploratory pilot, and not to provide ongoing treatment to address each subject’s generalized shoulder pain. The intent was to apply a specific cervical lateral glide mobilization to determine if that subject’s shoulder pain or shoulder abduction range of motion could be immediately altered. No previous studies have been found that examine the effect of cervical lateral glide mobilizations on shoulder pain, but there are similar studies involving the elbow. These findings bolster the argument that addressing the cervical spine in patients with more peripheral symptoms, even in the absence of cervical signs and symptoms, may expedite the rate of recovery. If during an assessment for shoulder pain any initial change in shoulder pain is noticed through cervical manual therapy techniques, continued therapy involving the cervical spine might provide more efficient management of shoulder pain and functional limitations, increasing patient satisfaction and decreasing treatment costs. Acknowledgements The author would like to acknowledge the financial support of the Graduate Department of Rehabilitation Science, University of Toronto, Toronto, Canada. Dr. Susan B. Jaglal is the Toronto Rehabilitation Institute Chair at the University of Toronto. Thank you to

374

L. McClatchie et al. / Manual Therapy 14 (2009) 369–374

Kajeandra Ravichandrian for technical support. Sincere thanks to all participants who volunteered for this study. References Aina A, May S. Case report: a shoulder derangement. Manual Therapy 2005;10: 159–63. Badcock LJ, Lewis M, Hay EM, McCarney R, Croft P. Chronic shoulder pain in the community: a syndrome of disability or distress? Annals of the Rheumatic Diseases 2002;61:128–31. Badley E, Tennant A. Changing profile of joint disorders with age: findings from a postal survey of the population of Calderdale, West Yorkshire, United Kingdom. Annals of the Rheumatic Diseases 1992;51:366–71. Balogun JA, Oladehinde KA, Olaogun MO, Obajuluwa VA. Inter- and intratester reliability of measuring neck motions with tape measure and Myrin gravityreference goniometer. Journal of Orthopedic and Sport Physical Therapy 1989;10:248–53. Bamji AN, Erhardt CC, Price TR. The painful shoulder: can consultants agree? British Journal of Rheumatology 1996;35:1172–4. Bergman GJD, Winters JC, Groenier KH, Pool JJM, Meyboom de-Jong B, Postema K, et al. Manipulative therapy in addition to usual medical care for patients with shoulder dysfunction and pain: a randomized, controlled trial. Annals of Internal Medicine 2004;141:432–9. Bird SB, Dickson EW. Clinically significant changes in pain along the visual analog scale. Annals of Emergency Medicine 2001;38:639–43. Chakravarty K, Webley M. Shoulder joint movement and its relationship to disability in the elderly. The Journal of Rheumatology 1993;20:1359–61. Chard MD, Hazleman BL, Hazleman R, King RH, Reiss BB. Shoulder disorders in the elderly: a community survey. Arthritis and Rheumatism 1991;34:766–9. Croft P, Pope D, Silman A. The clinical course of shoulder pain: prospective cohort study in primary care. British Medical Journal 1996;313:601–3. Gallagher EJ, Liebman M, Bijur PE. Prospective validation of clinically important changes in pain severity measured on a visual analog scale. Annals of Emergency Medicine 2001;38:633–8. Green S, Buchbinder R, Glazier R, Forbes A. Systematic review of randomised controlled trials of interventions for painful shoulder: selection criteria, outcome assessment, and efficacy. British Medical Journal 1998;316:354–60. Groenier KH, Winters JC, Meyboom de-Jong B. Classification of shoulder complaints in general practice by means of nonmetric multidimensional scaling. Archives of Physical Medicine and Rehabilitation 2003;84:812–7. Guerra de Hoyos JA, Martin MCA, de Leon EBB. Randomized trial of long term effect of acupuncture for shoulder pain. Pain 2004;112:289–98. Koester MC, George MS, Kuhn JE. Shoulder impingement syndrome. The American Journal of Medicine 2005;118:452–5. Lyons PM, Orwin JF. Rotator cuff tendinopathy and subacromial impingement syndrome. Medicine and Science in Sports and Exercise 1998;30:12–7. MacFarlane JG, Hunt IM, Silman AJ. Predictors of chronic shoulder pain: a population-based prospective study. Journal of Rheumatology 1998;25:1612–5. Magee D. Orthopedic physical assessment. Philadelphia: WB Saunders Company; 1987. Malstrom EM, Karlberg M, Melander A, Magnusson M. Zebris vs. Myrin: a comparative study between a three-dimensional ultrasound movement analysis and an inclinometer/compass method. Spine 2003;28:E433–40. Manifold SG, McCann PD. Cervical radiculitis and shoulder disorders. Clinical Orthopedics and Related Research 1999;368:105–13. Mulligan B. Manual therapy: NAGS, SNAGS and MWMS. Wellington: Plane View Services, Ltd; 1995.

Neviaser RJ. Painful conditions affecting the shoulder. Clinical Orthopedics and Related Research 1983;173:63–9. Ordway NR, Seymour R, Donelson RG, Hojnowski L, Lee E, Edwards WT. Cervical sagittal range of motion analysis using three methods. Spine 1997;22:501–8. Pope D, Croft P, Pritchard CM, Silman A. Prevalence of shoulder pain in the community: the influence of case definition. Annals of the Rheumatic Diseases 1997;56:308–12. Prescher A. Anatomical basics, variations, and degenerative changes of the shoulder joint and shoulder girdle. European Journal of Radiology 2000;35:88–102. Price DD, Bush FM, Long S, Harkins SW. A comparison of pain measurement characteristics of mechanical visual analogue and simple numerical rating scales. Pain 1994;56:217–26. Richardson JK, Iglarsh ZA. Clinical orthopaedic physical therapy. Philadelphia: W.B. Saunders Company; 1994. Schneider G. Restricted shoulder movement: capsular contracture or cervical referral– a clinical study. Australian Journal of Physiotherapy 1989;35: 97–100. Sobel JS, Winters JC, Groenier KH, Arendzen JH, Meyboom de-Jong B. Physical examination of the cervical spine and shoulder girdle in patients with shoulder complaints. Journal of Manipulative and Physiological Therapeutics 1997;20: 257–62. Sterling M, Jull G, Wright A. Cervical mobilisation: concurrent effects on pain, sympathetic nervous system activity and motor activity. Manual Therapy 2001;6:72–81. Stiesch-Scholz M, Fink M, Tschernitschek H. Comorbidity of internal derangement of the temporomandibular joint and silent dysfunction of the cervical spine. Journal of Oral Rehabilitation 2003;30:386–91. Tousignant M, de Bellefeuille L, O’Donoughue S, Grahovac S. Criterion validity of the cervical range of motion (CROM) goniometer for cervical flexion and extension. Spine 2000;25:324–30. Tundle AN. Redefining when to use placebo in clinical trials in Canada: National placebo working committee report. Orthopaedic Division Review May/June 2006:31–2. Turner-Stokes L. Clinical differential diagnosis of shoulder pain. British Journal of Hospital Medicine 1996;56:73–7. Van der Heijden GJMG. Shoulder disorders: a state-of-the-art review. Bailliere’s Best Practice and Research in Clinical Rheumatology 1999;13:287–309. Van der Windt D, Koes BW, Boeke AJP, deJong BA, Bouter LM. Shoulder disorders in general practice: prognostic indicators of outcome. British Journal of General Practice 1996;46:519–23. Vicenzino B, Collins D, Wright A. The initial effects of a cervical spine manipulative physiotherapy treatment on the pain and dysfunction of lateral epicondylalgia. Pain 1996;68:69–74. Vickers AJ, de Craen AJM. Why use placebos in clinical trials? A narrative review of the methodological literature. Journal of Clinical Epidemiology 2000;53: 157–61. Wells P. Cervical dysfunction and shoulder problems. Physiotherapy 1982;68: 66–74. Wilson C. Rotator cuff versus the cervical spine: making the diagnosis. The Nurse Practitioner 2005;30:45–50. de Winter AF, Jans MP, Scholten RJPM, Deville W, van Schaardenburg D, Bouter LM. Diagnostic classification of shoulder disorders: interobserver agreement and determinants of disagreement. Annals of the Rheumatic Diseases 1999;58: 272–7. Winters JC, Jorritsma W, Groenier KH, Sobel JS, deJong M. Treatment of shoulder complaints in general practice: long term results of a randomised, single-blind study comparing physiotherapy, manipulation and corticosteroid injection. British Medical Journal 1999;318:1395–6.

Manual Therapy 14 (2009) 375–380

Contents lists available at ScienceDirect

Manual Therapy journal homepage: www.elsevier.com/math

Original Article

The short-term effects of thoracic spine thrust manipulation on patients with shoulder impingement syndromeq Robert E. Boyles a, *, Bradley M. Ritland b, Brian M. Miracle b, Daniel M. Barclay b, Mary S. Faul b, Josef H. Moore b, Shane L. Koppenhaver b, Robert S. Wainner c, d a

School of Physical Therapy, University of Puget Sound, 1500N Warner St, Tacoma, WA 98416, United States US Army-Baylor University Doctoral Program in Physical Therapy, Ft. Sam Houston, TX, United States Texas State University, Physical Therapy Department, San Marcos, TX, United States d Texas Physical Therapy Specialists, PC, United States b c

a r t i c l e i n f o

a b s t r a c t

Article history: Received 17 May 2007 Received in revised form 5 May 2008 Accepted 19 May 2008

The study was an exploratory, one group pretest/post-test study, with the objective of investigating the short-term effects of thoracic spine thrust manipulations (TSTMs) on patients with shoulder impingement syndrome (SIS). There is evidence that manual physical therapy that includes TSTM and non-thrust manipulation and exercise is effective for the treatment of patients with SIS. However, the relative contributions of specific manual therapy interventions are not known. To date, no published studies address the short-term effects of TSTM in the treatment of SIS. Fifty-six patients (40 males, 16 females; mean age 31.2  8.9) with SIS underwent a standardized shoulder examination, immediately followed by TSTM techniques. Outcomes measured were the Numeric Pain and Rating Scale (NPRS) and the Shoulder Pain and Disability Index (SPADI), all collected at baseline and at a 48-h follow-up period. Additionally, the Global Rating of Change Scale (GRCS) was collected at 48-h follow-up to measure patient perceived change. At 48-h follow-up, the NPRS change scores for Neer impingement sign, Hawkins impingement sign, resisted empty can, resisted external rotation, resisted internal rotation, and active abduction were all statistically significant (p < 0.01). The reduction in the SPADI score was also statistically significant (p < 0.001) and the mean GRCS score ¼ 1.4  2.5. In conclusion, TSTM provided a statistically significant decrease in self reported pain measures and disability in patients with SIS at 48-h follow-up. Ó 2008 Elsevier Ltd. All rights reserved.

Keywords: Shoulder impingement Shoulder pain Thoracic spine thrust manipulation Manual physical therapy

1. Introduction Shoulder impingement syndrome (SIS) is a common shoulder diagnosis that accounts for 44–65% of all shoulder related medical visits to general practitioners (Price et al., 1994; Michener and Leggin, 2001). Neer (1972) described SIS as impingement of the rotator cuff beneath the coracoacromial arch. Fu et al. (1991) summarized SIS as a poorly defined term for a variety of shoulder disorders. While the etiology and location of SIS have been disputed, the emerging consensus has been that it is a multifactorial mechanical syndrome characterized by a compromise of the subacromial soft tissues. Common etiological factors of SIS are

q This study was approved by the Institutional Review Boards of Brooke Army Medical Center, Wilford Hall Medical Center, and Keller Army Community Hospital. * Corresponding author. þ1 253 879 3633; fax: þ1 253 879 2933. E-mail address: [email protected] (R.E. Boyles). 1356-689X/$ – see front matter Ó 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.math.2008.05.005

weakness, decreased activity of the rotator cuff muscles, anatomic variation of the acromion, spurring, trauma, or other underlying conditions (Wilk and Arrigo, 1993; Bak and Faunl, 1997; Morrison et al., 1997; Prato et al., 1998; Bang and Deyle, 2000; Paley et al., 2000;Yanai et al., 2000; Almekinders, 2001; Armfield et al., 2003; Kim and McFarland, 2004). Frequent complaints with SIS are pain, weakness, crepitus, and stiffness which may result in loss of activity and sleep disturbances (Lo et al., 1990; van der Windt et al., 1996). Reproduction of pain occurs with movements that compress the subacromial bursa and the supraspinatus muscle between the acromion and the humeral head (Neer and Welsh, 1977; Hawkins and Kennedy, 1980). Pain is common when the arm is elevated above the height of the shoulder while being internally rotated (Warner et al., 1990; Fu et al., 1991; Kamkar et al., 1993; Wilk and Arrigo, 1993; van der Windt et al., 1996; Yanai et al., 2000). Although earlier systematic reviews of randomized controlled trials (RCTs) found little support for physical therapy intervention

376

R.E. Boyles et al. / Manual Therapy 14 (2009) 375–380

in patients with SIS, more recent high quality RCTs have shown that manual physical therapy and exercise are effective for reducing pain and disability in patients with disorders of the rotator cuff when compared to competing interventions (Winters et al., 1999; Bang and Deyle, 2000; Bergman et al., 2004). In several studies, patients who received manual physical therapy and exercise directed at the cervicothoracic spine and ribs in addition to Primary Care Medicine (PCM) had improved success rates when compared to patients receiving PCM care alone or exercise alone intervention (Winters et al., 1999; Bang and Deyle, 2000; Bergman et al., 2004). In addition, these effects were still maintained at 1 year (Winters et al., 1999; Bergman et al., 2004). While manual therapy to the thoracic spine and ribs appears to be an important treatment component in patients with SIS, the individual contribution of manipulative intervention or which patients may benefit is not known. Given the positive results of these trials in which thrust manipulation was a treatment component, further studies investigating the specific contributions of thrust manipulation are warranted. While manual physical therapy and local corticosteroids have been shown to be of similar effectiveness for treating unilateral shoulder pain in primary care, patients receiving manual physical therapy had fewer co-interventions at 6 months (Bergman et al., 2004). Traditional passive interventions and modalities have not been shown to be effective for significantly reducing shoulder symptoms and disability associated with SIS (Leduc et al., 2003) and the use of non-steroidal anti-inflammatory drugs (NSAIDs) have not been demonstrated to provide clinically meaningful relief (Green et al., 2003; Petri et al., 2004). Patients with SIS may develop functional impairments with chronic symptoms. Therefore, it is essential to explore treatment options for SIS that extend beyond the glenohumeral joint and subacromial space. The clinical rationale for the use thoracic spinal thrust manipulation on SIS patients is based upon regional interdependence (Wainner et al., 2001), or the theory that dysfunction of one body part imparts dysfunction upon another. The concept of regional interdependence is supported by evidence demonstrating the effectiveness of spinal manipulation in the treatment of upper and lower extremity disorders (Williams et al., 1995; Bergman et al., 2004). More specifically, mobilization and manipulation of the thoracic spine and upper quarter structures were components of combined intervention protocols demonstrated to be effective for the treatment of patients with SIS (Bang and Deyle, 2000; Bergman et al., 2004). However, the studies did not standardize the manual therapy techniques that each patient received and the relative contribution of manual therapy intervention of the thoracic spine is unknown. To date, there are no published studies that address the shortterm benefits of treating SIS patients with only thoracic thrust manipulation. Therefore, the purpose of our study was to investigate the short-term effects of thoracic spine thrust manipulations (TSTMs) on patients with SIS symptoms.

2.1. Subjects Fifty-six meeting the inclusion criteria participated in the study (age 31.2 years, 8.9; 40 males, 71.4%). Subjects were recruited by referral from physical therapists, advertising, group announcements, or word-of-mouth contact. Subjects were screened by licensed physical therapists for the diagnosis of impingement syndrome according to the established inclusion criteria. All subjects were informed of the purpose of the study and signed an informed consent and Health Insurance and Portability and Accountability Act (HIPAA) forms approved by the Institutional Review Boards at Brooke Army Medical Center, Keller Army Community Hospital, and Wilford Hall Air Force Medical Center. The inclusion criteria used in this study were modified from those used by Bang and Deyle (Bang and Deyle, 2000; Deyle et al., 2000). Subjects were between the ages of 18 and 50 years old, reported a two or greater pain rating on a 10-point Numeric Pain Rating Scale (NPRS) with either test in Category 1 and score an NPRS of two or greater in either Category 2 or on any resisted test in Category 3 (Table 1). In addition, all patients were required to be TRICARE beneficiaries (military health care insurance affiliate). Patients were excluded from the study if they had a positive result on a cervical distraction, or Spurling’s test as described by Wainner et al. (2003), reported primary complaint of neck or thoracic pain, received previous treatment of shoulder mobilization or thoracic manipulation since the onset of current shoulder pain, received a cortisone or other fluid injection into the shoulder joint within the last 30 days, a history of osteoporosis or fracture of shoulder girdle bones, a neurologic deficit, received any hormone replacement therapy, had contraindications to thrust manipulations, unwillingness to participate in the study, or inability to attend a 48-h follow-up. Refer to Fig. 1 for subject flow diagram. 2.2. Outcome measures Intensity of shoulder pain was measured utilizing the NPRS. The NPRS is an 11-point pain rating scale ranging from 0 (no pain) to 10 (worst imaginable pain) used to assess pain intensity in the shoulder (Jensen et al., 1994). This scale has been demonstrated to be reliable, generalizable, and an internally consistent measure of clinical and experimental pain sensation intensity (Price et al., 1994). A two-point change on the NPRS has been identified as the minimally clinically important change needed to be confident that a change has actually occurred (Farrar et al., 2001; Childs et al., 2005). The NPRS was explained verbally and recorded prior to the shoulder evaluation and treatment and at the 48-h follow-up. The patient rated their pain during the following tests: Neer impingement test, Hawkins impingement test, active shoulder abduction, resisted external rotation, resisted internal rotation, and empty can. Table 1 Inclusion criteria. TRICARE beneficiary

2. Methods

Age 18–50 Plusa

Three faculty members and four doctoral students in the US Army-Baylor University Doctoral Program in Physical Therapy were responsible for all study procedures. In addition to regular classroom instruction, faculty members trained the students in all examination and intervention procedures used in this study during four separate training sessions. Subject enrollment commenced once faculty judged that students were proficient with all study procedures. Students performed the shoulder examination and intervention after initial screening by the primary investigator for inclusion/exclusion criteria.

Category I: impingement signs Neer’s sign Hawkins sign Category II: active shoulder abduction Category III: resisted break tests Internal rotation External rotation Empty can a Required to have NPRS of 2 with one of the two tests. In Category I and NPRS of 2 with one test from either Category II or III.

R.E. Boyles et al. / Manual Therapy 14 (2009) 375–380

101 Patients with Shoulder pain Screened for Eligibility

Not Eligible (n=42)

Did not want to participate (n=2)

Eligible (n=59)

Agreed to Participate, Signed Informed Consent/HIPPA, and received treatment (n=56)

377

Primary Complaint of Neck/Thoracic Pain (n=4) Cortisone/fluid injection in last 30 days (n=1) Positive Spurlings (n=1) Did not meet inclusion criteria (n=36)

Would not be able to return for follow-up (n=1)

Successful Outcome on 48 hour Follow-up Fig. 1. Flow diagram showing subject recruitment and retention.

Pain and disability associated with the patient’s shoulder were measured using the Shoulder Pain and Disability Index (SPADI). The SPADI is a 100-point, 13-item self-administered questionnaire which is divided into two subscales: a five-item pain subscale and an eight-item disability subscale. Williams et al. (1995) have shown that the SPADI is responsive to change and accurately discriminates between patients who are improving or worsening. Michener and Leggin (2001) also reported a high test–retest reliability and internal consistency for this instrument. A 10-point change on the SPADI has been identified as the minimally clinically important change needed to be confident that a change has actually occurred (Heald et al., 1997). The SPADI was administered prior to the shoulder evaluation and treatment and at the 48-h follow-up. Change in the perception in the quality of life after the treatment was measured by a Global Rating of Change Scale (GRCS) (Jaeschke et al., 1989). The GRCS is a 15-point global rating scale ranging from 7 (‘‘a very great deal worse’’) to 0 (‘‘about the same’’) to þ7 (‘‘a very great deal better’’). The use of a GRCS is a common, feasible, and useful method for assessing outcome (Fritz and Irrgang, 2001) and has been shown to be a valid measurement of change in patient status in other pain populations. It has been reported that scores of þ4 and þ5 are indicative of moderate changes in patient status and scores of þ6 and þ7 indicate large changes in patient status (Jaeschke et al., 1989). The GRCS was administered at the 48-h follow-up. The patient selected a number from 0 to 7 which corresponded to their perceived change in quality of life.

2.3. Examination procedure Patients completed a form listing the inclusion and exclusion criteria of the study and were questioned with regard to the duration, mode of onset, distribution of symptoms, nature of symptoms, aggravating/easing factors, and any prior shoulder treatments. The physical examination measures consisted of shoulder range of motion (ROM), cervical ROM, Spurling’s test, distraction test, cervical and thoracic posterior to anterior (PA) and unilateral mobility testing, and special tests for the shoulder. After a standardized shoulder examination was conducted, the examiner performed the standardized treatment.

2.4. Treatment Upon completion of the physical exam, all patients received a high velocity low amplitude TSTM focusing on the mid-thoracic spine (Fig. 2A) and cervicothoracic junction (Fig. 2B). Care is taken to protect the shoulders and minimize force through the shoulders. No patients complained of shoulder pain during the set-up or actual manipulation. Only subjects with rib angle tenderness on the exam received a rib manipulation (Fig. 2C) at the level of tenderness. Manipulations were performed in the following order for all patients: mid-thoracic, cervicothoracic junction, and rib manipulation (if required). A description of all three techniques can be reviewed in Appendix A. If a cavitation, or ‘‘pop’’, was experienced, the researcher proceeded to the next technique. If no cavitation was heard by the examiner or felt by the patient, on the first attempt, the patient was repositioned, and a second manipulation was attempted. A maximum of two attempts per technique were administered. At the end of the examination and treatment session, patients were instructed to maintain normal daily living activities within their pain tolerance, to avoid activities that exacerbate symptoms, and instructed to perform an active ROM exercise for the thoracic spine two or three times daily. 2.5. Follow-up Patients followed up with the same provider 48 h later. In order to determine the short-term effects of the treatment, follow-up SPADI, NPRS, and GRCS outcome measurements were conducted. No treatment/examination was conducted at the follow-up and the patient was considered to have completed the study at this time. 2.6. Data analysis Statistical analysis was performed utilizing SPSS version 12.0. Data were analyzed using paired t-tests with an alpha level set at 0.01. The alpha level adjusted taking into account the overly conservative nature of a Bonferroni correction and concern for a Type II error. The SPADI scores as well as the NPRS values were analyzed in the same manner.

378

R.E. Boyles et al. / Manual Therapy 14 (2009) 375–380

3. Results Between August 2005 and January 2006, 56 patients were recruited for the study. The total number of patients screened and reasons for exclusion can be seen in Fig. 1. Initial and follow-up NPRS and SPADI scores results can be found in Table 2. Significant differences between baseline and 48-h follow-up NPRS scores were found for all provocative shoulder tests, resisted tests, and SPADI scores, Table 2. Mean scores for all are depicted in Table 2. The mean SPADI change score was 6.85, Table 2. The mean GRCS value was 1.4 þ (2.5), frequency of GRCS can be found in Fig. 3. 4. Discussion Due to the limitations of this study, we want to make it clear to readers that this is an exploratory study from which no cause and effect relationship can be inferred. There was no control group, nor randomization of subjects. Other significant limitations include a short-term follow-up, limited sample size and lack of outcomes assessor blinding. This study was based on previous research (Bang and Deyle, 2000; Flynn et al., 2007) from the literature and clinical observations and we endeavored to lay ground work for a future line of inquiry and high quality research regarding the influence of thoracic thrust manipulation on SIS. Based on the results of this study, the use of TSTM in patients may have an impact on short-term pain and disability resulting in statistically significant changes. However, the changes observed did not reach the level of clinically meaningful significance. According to Heald et al. (1997), the SPADI requires a reduction of at least 10 points to be considered a clinically meaningful change. Childs et al. (2005) reported that a two-point reduction is needed for the NPRS. One possible reason why clinically significant change was not realized was that some patients in this study with shoulder complaints were not actively seeking care for their symptoms. This may explain the relatively lower NPRS and SPADI scores (Bang and Deyle, 2000; Bergman et al., 2004) which could have resulted in a floor effect and two different patient type populations represented. The mean GRCS was not large. Because the GRCS assesses a more global construct than pain and disability, longer-term follow-up may be required to adequately assess this variable. We have included the mean differences between the outcome scores in Table 2. While the mean differences we observed were all statistically significant, none meet the minimal clinically important difference MCID for NPRS (Childs et al., 2005) or SPADI scores (Williams et al., 1995). However, all score differences were not only statistically significant, but they were consistently lower and were measured 48 h after a single intervention. Therefore, we believe the meaning of these results may be clinically important and could be further clarified in a future study of stronger design that includes repeated application of the intervention. Another possible reason for the lack of clinical significance is that any effect TSTM contributed to the change observed was Table 2 Results. Initial NPRS Neer Hawkins EC, resisted IR, resisted ER, resisted Active ABD Fig. 2. Thoracic and rib manipulations used in this study. A: Seated mid-TSTM. B: Seated cervicothoracic spine thrust manipulation. C: Supine rib opening manipulation used in this study. Please see Appendix A for complete description of each.

SPADI

48-h FU

Mean difference

p-Value

4.0  2.5 4.5  2.2 3.3  2.5 2.4  2.4 2.9  2.6, 3.1  2.5

2.9  2.5 3.3  2.6 2.5  2.3 1.8  2.4 1.9  2.5 2.3  2.6

1.1 1.2 0.80 0.63 1.0 0.8

0.001 <0.001 0.007 0.008 <0.001 0.001

34.7  17.4

27.9  21.4

6.8

<0.001

Initial and 48-h follow-up NPRS and SPADI (mean values þ SD). FU ¼ follow-up, EC ¼ empty can, IR ¼ internal rotation, ER ¼ external rotation, ABD ¼ abduction.

R.E. Boyles et al. / Manual Therapy 14 (2009) 375–380

GRCS Frequency Frequency

16

12

12

8

8 4

12

2

2

3

2

5

6 2

2

M

od

er a So tely w m or ew se ha t Li w ttl or e se bi tw Ti o ny rs e bi tw Ab ou or se tt he S Ti am ny e bi t Li be ttl t e bi ter So tb m et ew te r ha M od tb er e tte at el r Q y ui be te t t er a bi G tb re et at te de r al be tte r

0

Fig. 3. GRCS frequency.

assessed in isolation from other interventions such as exercise and manual therapy of the shoulder girdle complex. These other interventions are often combined with TSTM and included in overall plan of care for patients with TSTM (Winters et al., 1999; Bang and Deyle, 2000; Bergman et al., 2004). While the relative contribution of specific manual therapy procedures is of interest, it has been demonstrated in patients with cervical spine disorders that the combined intervention of manual therapy and exercise is superior to either intervention alone (Bronfort et al., 2001; Jull et al., 2002). This would imply there is an interaction effect between manual therapy procedures and exercise. The individual responses of subjects to TSTM varied. Of the 56 subjects in this study, one-third of them had a reduction of at least 10 points on their SPADI scores and approximately one-third had a clinically meaningful reduction in their NPRS scores for various provocative tests and movements. Thirteen subjects had a GRCS of four or greater. It is possible that if TSTM had an effect that patients may respond differentially to the intervention. A clinical prediction rule has been established for determining which patients with back pain and neck pain respond best to manipulation (Childs et al., 2004; Cleland et al., 2006). Similarly, the development of a clinical prediction rule for determining which patients with SIS respond best to TSTM may be possible. Because of the design of this study, we were unable to demonstrate a cluster of either subjective or objective signs, symptoms, or clinical findings that might lead the physical therapist to a decision to manipulate. Again, this is where the development of a clinical prediction rule would be beneficial. As stated earlier, the clinical rationale for the use thoracic spinal thrust manipulation on SIS patients is based upon regional interdependence (Wainner et al., 2001), or the theory that dysfunction of one body part imparts dysfunction upon another. This rationale may be acceptable at this point, however, to truly understand the relationship between manipulative interventions effects on adjacent areas, rigorous mechanistic studies need to be conducted. Another possible explanation for the changes that occurred are biomechanical in nature. Two studies (Bullock et al., 2005; Lewis et al., 2005) have been able to show the effects of thoracic posture on shoulder pain. Specifically the relationship of postural corrections in the thoracic spine and its effects of decreasing pain and increasing shoulder motion. There has also been research that has shown a relationship between scapular positional dysfunction and shoulder pathology (Kibler, 1998; Lukasiewicz et al., 1999; Ludewig and Cook, 2000; Laudner et al., 2006). It is possible that the TSMT administered in this study had effected the scapular position on the thoracic spine, thus bringing about the changes reported. Again, this is speculation as the design of this study does not allow for cause and effect. The exploratory nature and design of this study do not permit us to establish a cause and effect relationship between TSTM and the findings observed in this study, or whether an interaction effect exists between TSTM and exercise in the management of patients

379

with SIS. Given the relatively short follow-up time, it is not likely that maturation or history effects were solely responsible for the changes observed in this study. Although TSTM may be of benefit in the management of patients with SIS, future work should consider a multifactorial, RCT of patients with moderate to severe finding of SIS that includes a standardized, evidence-based exercise program (Bang and Deyle, 2000) as well as manual therapy procedures to the shoulder girdle complex and thoracic spine. This would allow the questions of efficacy and interaction with regard to TSTM to be addressed. As this was an exploratory study, we wanted to emulate the clinical experience as close as possible. Therefore, each subject had the same provider perform the initial evaluation, the treatment, and the follow-up. Similarly, we chose outcome measures that are commonly used to assess a patient’s response to treatment or progression of pain/symptoms. One could argue that by having the same provider performing both the evaluation and the treatment, the results may have been impacted one way or another. Arguments could be made either way, but we believed the effects of having one provider, as opposed to two providers; one performing the treatment and another performing the evaluation, were negligible in this specific exploratory study. While statistically significant, the results did not represent clinically meaningful change. The type of thrust techniques used was based on region rather than specific joint(s) treated. Other than rib palpation, no attempt was made to attempt to clinically determine the type or location of manipulation. Although this study has significant threats to internal validity which limits interpretation of our results and any clinical inference made from them, it does provide an initial look at a research frame work and clinical model of manual therapy regional interventions for patients with SIS. We look forward to seeing future studies of manual therapy intervention in patients with SIS which include a control group and blinded collection of outcome measures. 5. Conclusion Subjects with SIS who received TSTM demonstrated statistically significant changes in pain and disability scores at 48 h. The efficacy of TSTM in isolation or combined with exercise and other manual therapy interventions is not known, nor can we identify which patients may benefit most. Further research with a more complex design is needed to determine questions of efficacy, relative contribution and predicted outcome. The opinions and assertions contained herein are the private views of the authors and are not to be construed as official or as reflecting the views of the Departments of the Army, Air Force, or Defense. Acknowledgments CAPT Edward Kane, PT, PhD, ECS, SCS, ATC, for his valuable input during conception and initial write-up of the original research proposal. Appendix A Mid thoracic spine thrust manipulation With patient sitting, researcher stands behind the patient and tells patient to slide all the way to the edge of the table. Researcher places upper right or left pectoral region on the area of the spine to be thrust. Researcher reaches around the patient and grasps subject’s elbows, with knees slightly flexed. Researcher then tells the patient to relax and take a deep breath. When patient is on a natural breathing relaxation after exhaling, the researcher will then

380

R.E. Boyles et al. / Manual Therapy 14 (2009) 375–380

compress the patient’s upper body through patient’s arms. Simultaneously, researcher extends knees to lift subject’s body slightly up and over the fulcrum established by chest making a J-stoke with the patient’s arms (Fig. 2a). Cervicothoracic junction thrust manipulation With patient sitting, have him/her inter-lock fingers behind lower cervical spine. Researcher stands behind patient with shoulders at same height as patients shoulders. Researcher then threads arms through patients so that hands are on top and inferior to the patient’s hands at the CT Junction. Patient is told to move hands as low on the neck as possible and to relax his/her arms. Care is taken not to hyperextend the patients shoulder, but rather use the researchers forearms in a compressive manner anterior to the shoulder. Recline patient slightly so that C-Spine is oriented perpendicular to floor. When the patient is on a natural breathing relaxation after exhaling the researcher uses legs and lumbar extension to apply a high velocity, short amplitude force against gravity to distract (Fig. 2b). Rib opening manipulation Patient is supine with arms folded across chest, positioned as close to edge of plinth as possible. Examiner stands on the side of the patient ipsilateral to the pain and rolls patient over just far enough to get his/her hand under patient to get a skin lock on the affected rib angle. Examiner rolls patient supine so examiner’s hand is still underneath patient and skin is still locked. Examiner moves patient’s body into trunk flexion, ipsilateral side bend and rotation. Patient told to breathe deeply. At the natural relaxation after the exhale, examiner applies thrust with own body over the hand that’s under the patient. If there is no cavitation, reposition patient and repeat the manipulation once (Fig. 2c). References Almekinders LC. Impingement syndrome. Clinics in Sports Medicine 2001; 20:491–504. Armfield DR, Stickle RL, Robertson DD, Towers JD, Debski RE. Biomechanical basis of common shoulder problems. Seminars in Musculoskeletal Radiology 2003;7:5–18. Bak K, Faunl P. Clinical findings in competitive swimmers with shoulder pain. American Journal of Sports Medicine 1997;25:254–60. Bang MD, Deyle GD. Comparison of supervised exercise with and without manual physical therapy for patients with shoulder impingement syndrome. Journal of Orthopaedic and Sports Physical Therapy 2000;30:126–37. Bergman GJ, Winters JC, Groenier KH, Pool JJ, Meyboom-de Jong B, Postema K, et al. Manipulative therapy in addition to usual medical care for patients with shoulder dysfunction and pain: a randomized, controlled trial. Annals of Internal Medicine 2004;141:432–9. Bullock MP, Foster NE, Wright CC. Shoulder impingement: the effect of sitting posture on shoulder pain and range of motion. Manual Therapy 2005;10:28–37. Bronfort G, Evans R, Nelson B, Aker PD, Goldsmith CH, Vernon H. A randomized clinical trial of exercise and spinal manipulation for patients with chronic neck pain. Spine 2001;26:788–97 [Discussion 798–789]. Childs JD, Fritz JM, Flynn TW, Irrgang JJ, Johnson KK, Majkowski GR, et al. A clinical prediction rule to identify patients with low back pain most likely to benefit from spinal manipulation: a validation study. Annals of Internal Medicine 2004;141:920–8. Childs JD, Piva SR, Fritz JM. Responsiveness of the numeric pain rating scale in patients with low back pain. Spine 2005;30:1331–4. Cleland JA, Fritz JM, Childs JD, Kulig K. Comparison of the effectiveness of three manual physical therapy techniques in a subgroup of patients with low back pain who satisfy a clinical prediction rule: study protocol of a randomized clinical trial [NCT00257998]. BMC Musculoskeletal Disorders 2006;7:11. Deyle GD, Henderson NE, Matekel RL, Ryder MG, Garber MB, Allison SC. Effectiveness of manual physical therapy and exercise in osteoarthritis of the knee. A randomized, controlled trial. Annals of Internal Medicine 2000;132:173–81. Farrar JT, Young Jr JP, LaMoreaux L, Werth JL, Poole RM. Clinical importance of changes in chronic pain intensity measured on an 11-point numerical pain rating scale. Pain 2001;94:149–58. Flynn TW, Wainner RS, Whitman JM, Childs JD. The immediate effects of thoracic spine manipulation on cervical range of motion and pain in patients with a primary complaint of neck pain – a technical note. Orthopaedic Division Review 2007 March/April:31–6.

Fritz JM, Irrgang JJ. A comparison of a modified Oswestry Low Back Pain Disability Questionnaire and the Quebec Back Pain Disability Scale. Physical Therapy 2001;81:776–88. Fu FH, Harner CD, Klein AH. Shoulder impingement syndrome. A critical review. Clinical Orthopaedics and Related Research 1991:162–73. Green S, Buchbinder R, Hetrick S. Physiotherapy interventions for shoulder pain. Cochrane Database of Systematic Reviews 2003:CD004258. Hawkins RJ, Kennedy JC. Impingement syndrome in athletes. American Journal of Sports Medicine 1980;8:151–8. Heald SL, Riddle DL, Lamb RL. The shoulder pain and disability index: the construct validity and responsiveness of a region-specific disability measure. Physical Therapy 1997;77:1079–89. Jaeschke R, Singer J, Guyatt GH. Measurement of health status. Ascertaining the minimal clinically important difference. Controlled Clinical Trials 1989;10:407–15. Jensen MP, Turner JA, Romano JM. What is the maximum number of levels needed in pain intensity measurement? Pain 1994;58:387–92. Jull G, Trott P, Potter H, Zito G, Niere K, Shirley D, et al. A randomized controlled trial of exercise and manipulative therapy for cervicogenic headache. Spine 2002;27:1835–43 [Discussion 1843]. Kamkar A, Irrgang JJ, Whitney SL. Nonoperative management of secondary shoulder impingement syndrome. Journal of Orthopaedic and Sports Physical Therapy 1993;17:212–24. Kibler WB. The role of the scapula in athletic shoulder function. American Journal of Sports Medicine 1998;26:325–37. Kim TK, McFarland EG. Internal impingement of the shoulder in flexion. Clinical Orthopaedics and Related Research 2004:112–9. Laudner KG, Myers JB, Pasquale MR, Bradley JP, Lephart SM. Scapular dysfunction in thowers with pathologic internal impingement. Journal of Orthopaedic and Sports Physical Therapy 2006;36:485–94. Leduc BE, Caya J, Tremblay S, Bureau NJ, Dumont M. Treatment of calcifying tendinitis of the shoulder by acetic acid iontophoresis: a double-blind randomized controlled trial. Archives of Physical Medicine and Rehabilitation 2003;84:1523–7. Lewis JS, Wright C, Green A. Subacromial impingement syndrome: the effect of changing posture on shoulder range of motion. Journal of Orthopaedic and Sports Physical Therapy 2005;35:72–87. Lo YP, Hsu YC, Chan KM. Epidemiology of shoulder impingement in upper arm sports events. British Journal of Sports Medicine 1990;24:173–7. Ludewig PM, Cook TM. Alterations in shoulder kinematicsand associated muscle activity in people with symptoms of shoulder impingement. Physical Therapy 2000;80:276–91. Lukasiewicz AC, McClure P, Michner L, Pratt N, Sennett B. Comparison of 3-dimensional scapular position and orientation between subjects with and without shoulder impingement. Journal of Orthopaedic and Sports Physical Therapy 1999;29:574–83. Michener LA, Leggin BG. A review of self-report scales for the assessment of functional limitation and disability of the shoulder. Journal of Hand Therapy 2001;14:68–76. Morrison DS, Frogameni AD, Woodworth P. Non-operative treatment of subacromial impingement syndrome. Journal of Bone and Joint Surgery American Volume 1997;79:732–7. Neer 2nd CS. Anterior acromioplasty for the chronic impingement syndrome in the shoulder: a preliminary report. Journal of Bone and Joint Surgery American Volume 1972;54:41–50. Neer 2nd CS, Welsh RP. The shoulder in sports. Orthopedic Clinics of North America 1977;8:583–91. Paley KJ, Jobe FW, Pink MM, Kvitne RS, El Attrache NS. Arthroscopic findings in the overhand throwing athlete: evidence for posterior internal impingement of the rotator cuff. Arthroscopy 2000;16:35–40. Petri M, Hufman SL, Waser G, Cui H, Snabes MC, Verburg KM. Celecoxib effectively treats patients with acute shoulder tendinitis/bursitis. The Journal of Rheumatology 2004;31:1614–20. Prato N, Peloso D, Franconeri A, Tegaldo G, Ravera GB, Silvestri E, et al. The anterior tilt of the acromion: radiographic evaluation and correlation with shoulder diseases. European Radiology 1998;8:1639–46. Price DD, Bush FM, Long S, Harkins SW. A comparison of pain measurement characteristics of mechanical visual analogue and simple numerical rating scales. Pain 1994;56:217–26. van der Windt DA, Koes BW, Boeke AJ, Deville W, De Jong BA, Bouter LM. Shoulder disorders in general practice: prognostic indicators of outcome. The British Journal of General Practice 1996;46:519–23. Wainner RS, Flynn TW, Whitman J. Spinal and extremity manipulation: the basic skill set for physical therapists. ed. Manipulations, I. Manipulations, Inc; 2001. Wainner RS, Fritz JM, Irrgang JJ, Boninger ML, Delitto A, Allison S. Reliability and diagnostic accuracy of the clinical examination and patient self-report measures for cervical radiculopathy. Spine 2003;28:52–62. Warner JJ, Micheli LJ, Arslanian LE, Kennedy J, Kennedy R. Patterns of flexibility, laxity, and strength in normal shoulders and shoulders with instability and impingement. American Journal of Sports Medicine 1990;18:366–75. Wilk KE, Arrigo C. Current concepts in the rehabilitation of the athletic shoulder. Journal of Orthopaedic and Sports Physical Therapy 1993;18:365–78. Williams Jr JW, Holleman Jr DR, Simel DL. Measuring shoulder function with the Shoulder Pain and Disability Index. Journal of Rheumatology 1995;22:727–32. Winters JC, Jorritsma W, Groenier KH, Sobel JS, Meyboom-de Jong B, Arendzen HJ. Treatment of shoulder complaints in general practice: long term results of a randomised, single blind study comparing physiotherapy, manipulation, and corticosteroid injection. British Medical Journal 1999;318:1395–6. Yanai T, Hay JG, Miller GF. Shoulder impingement in front-crawl swimming: I. a method to identify impingement. Medicine and Science in Sports and Exercise 2000;32:21–9.

Manual Therapy 14 (2009) 381–386

Contents lists available at ScienceDirect

Manual Therapy journal homepage: www.elsevier.com/math

Original Article

Biomechanical analysis of axial distraction mobilization of the glenohumeral joint – A cadaver study Ar-Tyan Hsu a, b, *, Jing-Fang Chiu c, Jia Hao Chang d a

Institute of Allied Health Sciences, College of Medicine, National Cheng Kung University, 1 University Road, Tainan 701, Taiwan Department of Physical Therapy, College of Medicine, National Cheng Kung University, Tainan, Taiwan c Department of Rehabilitation Medicine, Wan Fang Hospital, Taipei, Taiwan d Department of Physical Education, National Taiwan Normal University, Taipei, Taiwan b

a r t i c l e i n f o

a b s t r a c t

Article history: Received 2 February 2008 Received in revised form 28 May 2008 Accepted 29 June 2008

The axial distraction mobilization techniques are frequently employed for treating patients with joint hypomobility. However, there is a lack of basic biomechanical studies and description of this procedure. The purpose of this study was to analyze humeral head displacement while performing an axial distraction mobilization of the glenohumeral joint. Twelve experienced orthopedic physical therapists participated. Distraction mobilization techniques were performed in three different positions of glenohumeral abduction on a fresh cadaveric specimen. Outcome measures were displacements of the humeral head center during distraction mobilization. Result indicated that displacement of the humeral head was largest in the resting position (27.38 mm) followed by the neutral (22.01 mm) and the end range position (9.34 mm). There were significant differences for both the displacement of the humeral head (p < 0.002) and the distraction forces used (p < 0.015) among the three joint positions. Greater gain in mobility was obtained in distraction at the end range position. In conclusion, during distraction mobilization, the force applied by the therapist and displacement of the humeral head depends on the joint position tested. Our results also provide rationales for choosing end range distraction mobilization for improving joint mobility. Ó 2008 Published by Elsevier Ltd.

Keywords: Distraction Mobilization Shoulder Mobility

1. Introduction In manual therapy practice, distraction mobilization techniques are frequently performed on joints of extremities, spine and jaw as a diagnostic procedure to assess joint play, or as treatment to relieve pain, maintain or improve joint mobility (Kaltenborn, 1989; Conroy and Hayes, 1998; Deodato et al., 2003; Magee, 2006). Distraction mobilization usually involves the exertion of a force along the shaft of the involved segment or in a direction perpendicular to the joint surface by the therapist (Kaltenborn, 1989; Threlkeld, 1992) or postoperatively through a fixator (Gausepohl et al., 2006). In certain shoulder dysfunctions, such as adhesive capsulitis and shoulder impingement syndrome, distraction mobilization techniques are often included as an integral part of the rehabilitation program (Kaltenborn, 1989; Hertling and Kessler, 1996; Culham and Peat, 1999). However, the kinematic and kinetic characteristics of the distraction mobilization technique have not yet been reported. * Corresponding author. Institute of Allied Health Sciences, College of Medicine, National Cheng Kung University, 1 University Road, Tainan 701, Taiwan. Tel.: þ886 6 235 3535x5931; fax: þ886 6 237 0411. E-mail address: [email protected] (A.-T. Hsu). 1356-689X/$ – see front matter Ó 2008 Published by Elsevier Ltd. doi:10.1016/j.math.2008.06.003

In vitro studies of glenohumeral stability and laxity tests following selective cutting of capsuloligamentous structures have shed light on the roles of specific glenohumeral capsular ligaments on joint stability (Harryman et al., 1992; Warner et al., 1992; Warner, 1993; Bigliani et al., 1996; Malicky et al., 1996; Blasier et al., 1997; Wilk et al., 1997; Debski et al., 1999a; Brenneke et al., 2000). Results of these studies provided rationales for glenohumeral joint mobilization at different joint positions and directions. In an in vivo study, Gokeler et al. (2003) investigated changes in distance between the humeral head and the glenoid fossa during traction in the loose-packed and the closed-packed positions. The authors, however, did not find differences in mean distance between the humeral head and the glenoid fossa during traction in these two positions. Distraction with loading applied along the longitudinal axis of the humerus, however, has not yet been studied, except for the inferior translation at the neutral position where the direction of axial distraction and inferior translation coincide. The purpose of this study was to analyze the humeral head displacements pattern during distraction mobilization procedures in the neutral position, the resting position, and the end range position of the glenohumeral abduction.

382

A.-T. Hsu et al. / Manual Therapy 14 (2009) 381–386

2. Materials and methods 2.1. Subjects Twelve physical therapists (five men, seven women, mean age 30.92  7.54 years) with an average of 6.58 years of clinic experience in physical therapy and at least 3 years in orthopedic or manual therapy participated in this study. They completed questionnaires regarding their past experiences in joint mobilization and signed an informed consent form. A written instruction regarding the distraction mobilization technique was provided to all participants two weeks prior to the experiment. In this written instruction, guidance regarding the specific glenohumeral joint distraction mobilization technique under study including descriptions of the joint position, long axis traction force, and the grade of movement (Kaltenborn grades III) were included (Kaltenborn, 1989). The three positions of distraction employed in the current study were defined as follows. The neutral position was the position when the arm was placed at the side of the body in neutral rotation with its shaft paralleled the medial border of the scapula. The resting position was defined as the position where the therapists detected the greatest amount of joint laxity. The end range position was the end range of abduction in neutral rotation. The protocol of this study was approved by the Institutional Review Board of the National Cheng Kung University Hospital. 2.2. Specimen preparation The glenohumeral specimen was prepared according to the procedures described by Hsu et al. (2002a,b) except that a tensile force of 22.2 N was equally distributed to the tendons of supraspinatus, subscapularis, and the infraspinatus–teres minor complex. The lines of action for each of these muscles in relation to the spine of the scapula were 51, 58 , and 8 for the infraspinatus– teres minor complex, subscapularis, and supraspinatus, respectively (Apreleva et al., 1998). 2.3. Instrumentation A six-axis load cell (AMTI MC3A-6-250, Advanced Mechanical Technology, Inc. Massachusetts, USA) and an instruNet data acquisition system (GW Instrument Inc., Massachusetts, USA) were used to monitor and record forces and moments applied to the scapulohumeral complex. A six-camera VICON 370 motion analysis system (VICON Motion Systems Limited, Oxford, UK) was used to track trajectories of retroreflective markers and movements of the humerus and the scapula. The load cell was mounted on a 3.83 cm thick, L-shaped aluminum plate clamped to the edge of a sturdy table. Prior to mounting the specimen, three retroreflective markers were placed on the top plate of the load cell and recorded by the VICON motion analysis system for referencing the load cell coordinate system to that of the global. The scapular block was rigidly fixed on the top plate of the load cell such that the plane of scapula was horizontal and paralleled the top plate with the anterior aspect of scapula facing superiorly. By aligning the medial border of the scapula parallel to the x axis of the load cell, we defined the scapular coordinate system identical to the load cell coordinate with the x, y, and z axes directed inferiorly, medially and anteriorly, respectively. The humerus segment was defined by a triplet of retroreflective markers drilled into the distal end of the humerus (Fig. 1) with one of its arms paralleled the shaft of the humerus. As the three vectors formed by each of the three markers and the origin of the triplet were mutually perpendicular to one another, once the coordinates of the three markers were known the position of the origin of the triplet could be solved and the humeral coordinate system was

Fig. 1. Set up for the axial distraction test with the specimen in neutral position. (A) Scapular clamp, (B) triplet markers, (C) pulley systems for applying force to rotator cuff muscles, (D) specimen.

defined. The neutral position of the humerus was designated as the position when the humerus was placed in neutral rotation on the plane of scapula with its shaft paralleled the medial border of scapula. Once the scapular and the humeral coordinates systems had been established, we employed a method incorporating Gamage and Lasenby (2002) least square method to compute the center of rotation of the glenohumeral joint (COR) and the relative displacement of the center of the head of the humerus (CHH) and the center of the glenoid fossa (CGF). A detailed description of the method was presented in a recent report by Lin et al. (2007). It was assumed that COR did not change during small-arc movements of the arm and that CGF and CHH coincided at this time. As the CGF was rigidly linked to the scapula its position within the scapular coordinate system would not change during the axial distraction procedures. Therefore, the amount of translation of the CHH could be derived by computing the difference between positional vectors of the CGF and the CHH (calculated from the coordinates of the triplet markers) at the beginning and at the end of the distraction procedure. This method has been validated by Lin et al. (2006) with an acrylic model of a ball-and-socket joint. The root-mean-square errors for the estimation of the COR were 0.04, 0.19, and 1.31 mm in the anteroposterior, mediolateral, and superoinferior directions, respectively, and those for the relative translation between the head and the scapular components ranged from 0.21 to 0.51 mm.

2.4. Experimental procedures After COR had been derived, the following experimental procedures were conducted in the following sequences (Fig. 2): 1. five repetitions of grade III distraction mobilization at the neutral position, 2. five repetitions of grade III distraction mobilization at the resting position,

A.-T. Hsu et al. / Manual Therapy 14 (2009) 381–386

383

Fig. 2. Superior views of the experimental procedure – axial distraction in the neutral (A), the resting (B), and the end range (C) positions, respectively.

3. five repetitions of grade III distraction mobilization at the end range position. Subjects were instructed to hold with the same amount of force applied at the end of the distraction for about 5 s (a count of 1–5). The rate of loading, however, was not controlled. Subjects were simply instructed to perform as they always do in the clinic. These procedures were repeated half an hour later to test the intersession test–retest reliability. 2.5. Data analysis Both the displacement of the CHH and the distraction force were calculated along the direction of the longitudinal axis of the humerus. As the translational mobility of a joint is position dependent (Hsu et al., 2000), a better measure of gain in translational mobility across different positions is strain and not the magnitude of displacement. For this reason we also expressed the gain (difference in translation between the fifth and the first repetition) in translational mobility as a percentage of first repetition. Because the sample size was relatively small, nonparametric statistical methods were used. The Friedman test was employed to evaluate the differences of the peak magnitude of the humeral head displacements between the first and the fifth repetitions of the distraction procedure among the three different joint positions both as the absolute magnitude of displacement and as a percentage of the first repetition. The Wilcoxon signed rank test was used for post hoc cross comparison of the differences among different positions with an adjusted a value of 0.017. The mean peak magnitude of distraction forces acquired from the load cell during testing was also calculated with the Friedman test and the Wilcoxon signed rank tests. Intraclass correlation coefficients ICC2, 1 were used to compute the intra- and intersession test–retest reliability. The p < 0.05 was regarded as statistically significant. We used SPSS 11.0 (SPSS for Windows, Chicago) for all statistical analyses.

37.97  3.01 and 86.18  4.18 , respectively. The mean values of distraction forces applied by the participating therapists in the three joint positions tested are listed in Table 1, and values for the coefficients of variation (CV) ranged from 30.99% to 40.20%. The mean peak forces registered during the distraction mobilization technique in the neutral, the resting, and the end range position were 57.61  23.16 N, 57.24  22.32 N, and 72.28  22.40 N, respectively. Results of the Friedman test showed a significant difference (F ¼ 10.17, p ¼ 0.006) in the magnitudes of distraction force among three different glenohumeral joint positions. There were significant differences in the distraction force (a ¼ 0.017, after Bonferroni adjustment) between the end range position and the neutral position (z ¼ 2.43, p ¼ 0.015), and between the end range position and the resting position (z ¼ 2.98, p ¼ 0.003). The values of ICC2, 1 for the intra- and the intersession test– retest reliability are listed in Table 2. The intrasession test–retest reliability was excellent (0.97–0.98), and that of intersession test– retest reliability was good to excellent (0.88–0.97) for each of the positions tested. 3.2. Effect of joint position on displacement of the head of the humerus The mean and SD of the peak total displacements of the humeral head of the five repetitions in the three positions tested is presented in Table 3. Results of the Friedman test showed significant differences among the three joint positions (F values all equal 22.2, p ¼ 0.000) and the five repetitions (F values ranged from 42.1 to 45.9, p ¼ 0.000). Results of the post hoc analyses are presented in Table 3 and showed significant differences among five distraction repetitions in the neutral position (Z values ranged from 3.06 to 3.07, p values ranged from 0.003 to 0.002), in the resting position (Z values ranged from 3.06 to

Table 1 Mean, SD, and CV (coefficients of variation) of the distraction force in the axial direction of the humerus at three joint positions tested.

3. Results 3.1. Forces and joint angle The mean glenohumeral joint abduction angles actually achieved during the distraction mobilization procedures in the neutral, the resting and the end range positions were 2.72  0.67,

Neutral position Resting position End range position

N

Mean (N)

SD (N)

CV (%)

12 12 12

57.61 57.24 72.28

23.16 22.32 22.40

40.20 38.99 30.99

Note: Mean and SD values are reported as Newton. Wilcoxon signed rank tests: the end range position vs the neutral position, z ¼ 2.43, p ¼ 0.015; the end range position vs the resting position, z ¼ 2.98, p ¼ 0.003; a ¼ 0.017.

384

A.-T. Hsu et al. / Manual Therapy 14 (2009) 381–386

Table 2 Intraclass correlation coefficients (ICC) of intrasession and intersession reliability for the distraction mobilization procedures performed at the neutral, resting, and the end range position.

Neutral position Resting position End range position

N

Intrasession ICC2,1

Intersession ICC2,1

12 12 12

0.97 0.98 0.97

0.88 0.94 0.97

Table 4 Means and SDs of gains of displacement of the humeral head in the axial direction of the humerus expressed in millimeters and as a percentage of the first repetition (N ¼ 12).

Neutral Resting End range

Gaina in mm (SD)

Gainb in % (SD)

1.7(0.8) 1.4(0.5) 1.7(0.6)

8.6(4.4) 5.5(2.0) 22.4c(6.9)

a

No position effect (Chi-Square ¼ 4.044, p ¼ 0.132). Significant position effect (Chi-Square ¼ 16.167, p ¼ 0.000). Post hoc Wilcoxon signed ranks test (a ¼ 0.017) showed gains in the end range position > resting position (Z ¼ 3.059, p ¼ 0.002), the end range position > the neutral position (Z ¼ 2.903, p ¼ 0.004). No difference between the neutral position and the resting position (Z ¼ 2.197, p ¼ 0.028). b

2.3, p values ranged from 0.021 to 0.002) as well as the end range position (Z values ranged from 3.07 to 2.81, p values ranged from 0.005 to 0.002). Gains in the humeral head displacement between the first and the fifth trials during the distraction procedures were largest in the end range position (1.7  0.59 mm or 22.4  6.9%) followed by the neutral position (1.7  0.84 mm or 8.6  4.4%) and the resting position (1.4  0.54 mm or 5.5  1.9%, Table 4). There was a significant difference in gains in the humeral head displacements when expressed as percentages of the magnitude of the first trial among the three positions (F ¼ 16.2, p ¼ 0.000), but not when expressed in millimeters. There were significant difference in the percentage gain in displacement (a ¼ 0.017, Wilcoxon Signed Ranks test) between the end range position and the neutral position (z ¼ 2.9, p ¼ 0.004), and between the end range and the resting position (z ¼ 3.1, p ¼ 0.002). A trend was noted between the neutral position and the resting position (z ¼ 2.2, p ¼ 0.03). 4. Discussion 4.1. Intrasession and intersession test–retest reliability Most clinical manual procedures are characterized by good intrasession, poor intersession and poor inter-tester reliability. The mobilization maneuvers are no exception. Simmonds et al. (1995) reported that high inter-therapist variability and poor reliability in 10 physical therapists while performing ventrally directed translational mobilization on a mechanical spinal model under different conditions of stiffness (Simmonds et al., 1995). The authors also noted an interesting systematic bias in underestimating the magnitude of the applied force and overestimating the amount of motion. The variability in force application and the general overestimation of motion detection may explain the poor reliability of measurements obtained in clinical tests based upon motion palpation. As for the glenohumeral joint, Hsu et al. (2002b) tested 12 experienced therapists for grade II and grade III dorsal glide mobilization forces in resting and end range position. The authors reported that the intrasession test–retest reliability was excellent

Table 3 Mean and SD of the peak displacements of the humeral head in the direction of the humeral axis of the five repetitions in three different joint positions.

N Neutral Resting End range

REP1 (SD)

REP2 (SD)

REP3 (SD)

REP4 (SD)

REP5 (SD)

12 20.4(3.8) 25.9(4.0) 7.7(1.6)

12 20.8(3.8) 26.3(4.2) 8.1(1.6)

12 21.5(3.9) 27.1(4.5) 8.8(1.9)

12 21.8(3.9) 26.9(4.3) 8.9(1.8)

12 22.0(3.8) 27.4(4.2) 9.3(1.9)

Note: REP1–REP5: displacement values of the first to fifth repetitions of distraction. Unit: mm. Results of the Friedman test in five repetitions among three joint positions showed a significant difference (F ¼ 22.2, p ¼ 0.000); between five repetitions in neutral position (F ¼ 45.9, p ¼ 0.000); in resting position (F ¼ 42.1, p ¼ 0.000); in end range position (F ¼ 44.8, p ¼ 0.000). Wilcoxon signed rank tests: Resting position vs Neutral position, z ¼ 2.9, p ¼ 0.003; End range position vs Neutral position, z ¼ 3.06, p ¼ 0.002; End range position vs Resting position, z ¼ 3.06, p ¼ 0.002. Between five distraction repetitions in neutral position (Z ¼ 3.06 to 3.07, p ¼ 0.003 to 0.002); in resting position (Z ¼ 3.06 to 2.3, p ¼ 0.021 to 0.002); in end range position (Z ¼ 3.07 to 2.81, p ¼ 0.005 to 0.002). a ¼ 0.017.

c

(0.90–0.94) and that of intersession test–retest reliability, however, was poor (0.1–0.54). In the present study, we also noted large values of coefficients of variation (ranged from 30.99% to 40.20%) indicating a high inter-therapist variability for the mean peak force exerted during distraction mobilization in different joint positions. However, both the intrasession and intersession test–retest reliability of the mean peak forces measured were excellent. This departure from the results of prior reports might have been the consequence of a short inter-test period (30 min) used in the present study. 4.2. Distraction forces and joint position Results of the present study indicate that the magnitude of force exerted by therapists during distraction mobilization appears to be influenced by the joint position. At positions where the joint capsule is lax (the neutral position and the resting position) therapists exert less force than the position where the joint capsule is taut (the end range of abduction position). Hsu et al. (2002b) reported a similar finding in a dorsally directed translational mobilization regime. The authors suggested that therapists’ perception of the resistance encountered in the loose-packed position provided higher contrast, as the mobilization force was applied the target tissue went through a period of low resistance in the neutral zone and the toe region before standing out against a rapid rise in stiffness at the linear elastic region of the load-displacement curve. In contrast, at the end range position the slack on the capsular tissue has been taken up by prior positioning of the joint towards the end range making it more difficult for a therapist to accurately identify the resistance encountered (Hsu et al., 2002b). A similar phenomenon might have occurred during the axial distraction mobilization procedure employed in the present study and led to a greater force exerted by therapists at the end range of distraction. In the present study the amount of force exerted by the therapists during axial distraction mobilization (57.24 N and 72.28 N for the resting and the end range position, respectively) was about twice the magnitude obtained by Hsu et al. (2002b) in a dorsal translational mobilization (28.1 N and 38.8 N for the corresponding positions). Such discrepancy in the magnitude of the force exerted by the therapists might be explained by the fact that the primary restraining structures to a dorsally directed translational mobilization, according to the circle stability concept, is the posterior capsule for the resting position and posterior band and the axillary pouch of the inferior glenohumeral ligament for the end range position. During axial distraction mobilization, however, capsular ligaments located at both the anterior and the posterior aspects of the joint are equally stressed if performed at the resting position, and the whole inferior glenohumeral ligament will be added to the list of primary restraints if performed at the end range position. This finding also suggests that the magnitude of force may not be the sole criterion therapists consider when

A.-T. Hsu et al. / Manual Therapy 14 (2009) 381–386

determining the ‘final stop’, some measure of displacement may be involved as well. The magnitude of axial distraction force obtained in the present study was smaller comparing to a recent in vivo measurement by Gokeler et al. (2003) who reported a mean traction force of 12 kg. The discrepancy might have been due to differences in the experimental design especially regarding the fixation of the scapula and muscle guarding. In the present cadaver study, no muscle guarding was possible while rigid fixation of the scapula had rendered a more effective distraction of the humerus relative to the glenoid fossa and, thus, less force was required. 4.3. Joint position and displacement of the humeral head In the present study, the mean peak total displacement of the CHH during distraction procedure was largest in the resting position and followed by the neutral and the end range position. Our results indicate that the variation of the humeral head displacement during the distraction procedures is influenced by joint positions. Similar results were reported by Black et al. (1999), Debski et al. (1999a,b) and Hsu et al. (2002a) on the displacement of the humeral head in response to ventrally and/or dorsally directed forces. These studies demonstrated that more humeral head displacement occurred in the mid-range position than that of the end range position in cadaver studies. Such findings could be attributable to the cradling of the tightened inferior glenohumeral ligament around the humeral head at the end range of glenohumeral abduction position (O’Brien et al., 1995). In contrast, the resting position where the joint capsule is lax allows greater humeral head displacements. However, in their in vivo study Gokeler et al. (2003) did not find differences in mean distance between the humeral head and the glenoid fossa during traction in the maximally loose-packed position compared to that of the maximally close-packed position even when the subjects were reported to be relaxed. The authors suggested that muscle guarding might have been a major source of error in their study. No such concern was needed for our in vitro cadaver study. In the present study, during the distraction procedures gains in the humeral head displacement when expressed in millimeters did not differ among positions tested, but when expressed as percentages of the magnitude of the first trial the gain in the end range position was greater than those of the neutral and the resting positions. Such results appear to suggest that the end range distraction may be more effective in gaining mobility of the glenohumeral joint. A similar conclusion was reported by Hsu et al. (2000, 2002a) in a dorsal and ventral mobilization paradigm and Yang et al. (2007) in a frozen shoulder population. 4.4. Study limitations A single cadaveric shoulder specimen was used throughout the testing period in our study. While the use of a single specimen may subject the specimen to changes in the properties of the joint capsule after repetitive distraction mobilization, different specimens might differ greatly in the mechanical properties of the joint capsules and might have introduced another source of variability into the study. Therefore, the authors opted to adopt a single specimen for the purpose of this study. Forces applied in our study were exerted by experienced physical therapists performing axial distraction mobilization procedures on the cadaveric specimen shoulder model. Compared with live subjects, the cadaver model lacks active tension from muscles crossing the glenohumeral joint. In actual clinical practice, the loading force used and the mechanical responses obtained might

385

differ from those used in the present study. Accordingly, generalization to living patients should proceed with caution. In the present study, the sequence of the distraction positions was not randomized. However, as was discussed previously, axial distraction at different positions stresses predominately different portion of the glenohumeral joint capsule. In addition, gain in displacement due to the fifth distraction bout was expressed as a percentage of the first bout in each position. We felt that the potential biases due to the fixed order of distraction position were most likely avoided in the present study. Like all viscoelastic materials, joint tissues exhibit time- and history-dependent properties and are influenced by the magnitude and repetition of the applied force, and the mechanical constraints that might have affected our study (Nigg and Herzog, 1994; Woo et al., 1994). Tissue deformation and the mechanical properties of the glenohumeral joint might have changed with repeated distraction mobilization, which might have affected the results in our study. Comparing with the in vivo results reported by Gokeler et al. (2003), the magnitude of displacement of the CHH obtained in the present study, while equivalent to those reported in other cadaver studies, was relatively large. In view of the reported size of the glenoid fossa (von Shroeder et al., 2001) it was likely that dislocation of the shoulder might have occurred during axial distraction in the neutral position and probably in the resting position as well, although distraction at the resting position might not place the humeral head in an alignment we would call dislocation. 5. Conclusion The present study investigated the kinematic and kinetic characteristics obtained from the axial distraction mobilization of the glenohumeral joint in different positions. Our results suggest that both the forces applied and the displacements of the head of the humerus during distraction mobilization are position dependent. The magnitude of the force applied during the distraction mobilization was about twice that of the dorsally directed translational mobilization. The results of the present study also suggest that axial distraction mobilization at the end range is more effective in gaining joint mobility when compares with those of the neutral and the resting positions. The findings of the present study provide the rationales for choosing distraction mobilization technique at appropriate joint position. Further in vivo studies are necessary for a definite conclusion on the efficacy of axial distraction on normal subjects and patients with glenohumeral mobility disorders. References Apreleva M, Hasselman CT, Debski RE, Fu FH, Woo SL, Warner JJ. A dynamic analysis of glenohumeral motion after simulated capsulolabral injury. A cadaver model. Journal of Bone and Joint Surgery 1998;80-A(4):474–80. Bigliani LU, Kelkar R, Flatow EL, Pollock RG, Mow VC. Glenohumeral stability. Clinical Orthopaedics and Related Research 1996;330:13–30. Blasier RB, Soslowsky LJ, Malicky DM, Palmer ML. Posterior glenohumeral subluxation: active and passive stabilization in a biomechanical model. Journal of Bone and Joint Surgery 1997;79-A(3):433–40. Black KP, Schneider DJ, Yu JR, Jacobs CR. Biomechanics of the Bankart repair: the relationship between glenohumeral translation and labral fixation site. American Journal of Sports Medicine 1999;27(3):339–44. Brenneke SL, Reid J, Ching RP, Wheeler DL. Glenohumeral kinematics and capsuleligamentous strain resultind from laxity exams. Clinical Biomechanics 2000;15(10):735–42. Conroy DE, Hayes KW. The effect of joint mobilization as a component of comprehensive treatment for primary shoulder impingement syndrome. Journal of Orthopedics and Sports Physical Therapy 1998;28(1):3–13. Culham E, Peat M. Functional anatomy of the shoulder complex. Journal of Orthopedics and Sports Physical Therapy 1999;18(1):342–50. Debski RE, Wong EK, Woo S, Sakane M, Fu FH, Warner JJ. In situ force distribution in the glenohumeral joint capsule during anterior-posterior loading. Journal of Bone and Joint Surgery 1999a;17-A(5):769–75.

386

A.-T. Hsu et al. / Manual Therapy 14 (2009) 381–386

Debski RE, Sakone M, Woo SL, Wong EK, Fu FH, Warner JJ. Contribution of the passive properties of the rotator cuff to glenohumeral stability during anteriorposterior loading. Journal of Shoulder and Elbow Surgery 1999b;8(4):324–9. Deodato F, Cristiano S, Trusendi R, Giorgetti R. A functional approach to the TMJ disorders. Progress in Orthodontics 2003;4(2):20–37. Gamage HU, Lasenby J. New least squares solutions for estimating the average centre of rotation and the axis of rotation. Journal of Biomechanics 2002;35(1):87–93. Gausepohl T, Mader K, Pennig D. Mechanical distraction for the treatment of posttraumatic stiffness of the elbow in children and adolescents. Journal of Bone and Joint Surgery 2006;88-A(5):1011–21. Gokeler A, van Paridon-Edauw GH, DeClercq S, Matthijs O, Dijkstra PU. Quantitative analysis of traction in the glenohumeral joint. In vivo radiographic measurements. Manual Therapy 2003;8(2):97–102. Harryman DT, Sidles JA, Harris SL, Matsen FA. The role of the rotator interval capsule in passive motion and stability of the shoulder. Journal of Bone and Joint Surgery 1992;74-A(1):53–66. Hertling D, Kessler RM. Management of common musculoskeletal disorders: physical therapy principles and methods. 3rd ed. New York: Lippincott; 1996. Hsu AT, Ho L, Ho S, Hedman T. Joint position during anterior-posterior glide mobilization: its effect on glenohumeral abduction range of motion. Archives of Physical Medicine and Rehabilitation 2000;81(2):210–4. Hsu AT, Hedman T, Chang JH, Vo C, Ho L, Ho S, Chang GL. Changes in abduction and rotation range of motion in response to simulated dorsal and ventral translational mobilization of the glenohumeral joint. Physical Therapy 2002a;82(6):544–55. Hsu AT, Ho L, Ho S, Hedman T. Characterization of tissue resistance during a dorsally directed translational mobilization of the glenohumeral joint. Archives of Physical Medicine and Rehabilitation 2002b;83(3):360–6. Kaltenborn F. Manual mobilization of the extremity joints. Oslo: Olaf Norlis Bokhandel; 1989. Lin HT, Hsu AT, An KN, Chang GL. Glenohumeral joint laxity measurement in normal subjects and its validation. In: Proceedings of the 15th International Conference on Mechanics in Medicine and Biology, Singapore, December 6; 2006.

Lin HT, Hsu AT, Chang GL, Chang Chien JR, An KN, Su FC. Determining the resting position of the glenohumeral joint in subjects who are healthy. Physical Therapy 2007;87(12):1669–82. Magee DJ. Orthopedic physical assessment. 4th ed. Philadelphia: WB Saunders; 2006. Malicky DM, Soslowsky LJ, Blasier RB, Shyr Y. Anterior glenohumeral stabilization factors: progressive effects in a biomechanical model. Journal of Orthopedic Research 1996;14(2):282–8. Nigg BM, Herzog W. Biomechanics of the musculo skeletal system. New York: Wiley; 1994. O’Brien SJ, Schwartz RS, Warren RF, Torzilli PA. Capsular restraints to anteriorposterior motion of the abducted shoulder: a biomechanical study. Journal of Shoulder and Elbow Surgery 1995;4(4):298–308. Simmonds MJ, Kumar S, Lechelt E. Use of a spinal model to quantify the forces and motion that occur during therapists’ tests of spinal motion. Physical Therapy 1995;75(3):212–22. von Shroeder HP, Kuiper SD, Botte MJ. Osseous anatomy of the scapula. Clinical Orthopaedics 2001;383:131–9. Threlkeld AJ. The effects of manual therapy on connective tissue. Physical Therapy 1992;72(12):893–902. Warner JP. The gross anatomy of the joint surfaces, ligaments, labrum, and capsule. In: Matsen FA, Fu FH, Hawkins RJ, editors. The shoulder: a balance of mobility and function. Rosemont, IL: AAOS; 1993. p. 7–27. Warner JJ, Deng XH, Warren RF, Torzilli PA. Static capsuloligamentous restraints to superior-inferior translation of the glenohumeral joint. American Journal of Sports Medicine 1992;20(6):675–85. Wilk K, Arrigo C, Andrews J. Current concepts: the stabilizing structures of the glenohumeral joint. Journal of Orthopaedics and Sports Physical Therapy 1997;25(6):364–79. Woo SLY, An KN, Arnoczky SP, Wayne JS, Fithian DC, Myer BS. Anatomy, biology, and biomechanics of tendon, ligament, and meniscus. In: Simon SR, editor. Orthopedic basic sciences. Rosemont, IL: American Academy of Orthopedic Surgeons; 1994. Yang JL, Chang CW, Chen SY, Wang SF, Lin JJ. Mobilization techniques in subjects with frozen shoulder syndrome: randomized multiple-treatment trial. Physical Therapy 2007;87(10):1307–15.

Manual Therapy 14 (2009) 387–396

Contents lists available at ScienceDirect

Manual Therapy journal homepage: www.elsevier.com/math

Original Article

Searching for a conceptual framework for nonspecific low back pain Peter M. Kent a, b, c, *, Jennifer L. Keating c, Rachelle Buchbinder a, b a

Monash Department of Clinical Epidemiology at Cabrini Hospital, 183 Wattletree Road, Malvern, Victoria 3199, Australia Department of Epidemiology and Preventive Medicine, Monash University, Melbourne, Australia c Department of Physiotherapy, Monash University, Melbourne, Australia b

a r t i c l e i n f o

a b s t r a c t

Article history: Received 20 December 2007 Received in revised form 30 June 2008 Accepted 7 July 2008

Diverse views exist regarding the underlying nature of nonspecific low back pain (NSLBP). This study aimed to (i) develop a conceptual framework of NSLBP based on the expressed beliefs of those who treat and/or research NSLBP and (ii) determine whether these beliefs are discretely clustered and whether they are associated with participant characteristics. Surveys were completed by participants (n ¼ 162) of the 2006 Amsterdam International Low Back Pain Forum and a low back pain meeting (n ¼ 488) in Melbourne. Respondents reported beliefs regarding the nature of NSLBP. Probabilistic data-mining was used to detect ‘clusters of belief’ and between group differences were tested using Mann–Whitney U tests. Overall, there was an 84% response rate. Diverse beliefs were reported but multiple ‘clusters of belief’ to explain this diversity were not apparent. Whether predominantly engaged in clinical or research activity, people expressed similar beliefs, except that clinicians placed greater value on measures of physical impairment. There were conflicting views within the clinical and research community regarding the underlying nature of NSLBP. Within the constructs sampled, no unifying framework could explain the diversity of current beliefs. This is likely to reflect pervasive uncertainty about the etiology, and therefore best practice assessment, of NSLBP. Ó 2008 Elsevier Ltd. All rights reserved.

Keywords: Low back pain Conceptual framework Health beliefs

1. Introduction Less than 15% of people who seek care for low back pain (LBP) in primary care have specific LBP that is either associated with serious pathology (infection, fracture, and cancer) or neurocompression (most commonly due to nerve root compression) (Deyo et al., 1992). Most low back pain in primary care remains a diagnostic enigma and is appropriately classified as nonspecific low back pain (NSLBP). There are diverse views regarding the underlying nature of NSLBP and the appropriate management of this prevalent condition. For example, some have argued that NSLBP is a number of conditions (subgroups) for which targeted treatment can improve patient outcomes (Leboeuf-Yde, 2001; Childs et al., 2004; Long et al., 2004; Brennan et al., 2006), while others have argued that conceptualising and treating NSLBP as a single condition is more ideal (Waddell, 1998; Zusman, 1998). Even within those who subscribe to the notion of NSLBP subgrouping, there are different views regarding whether clinically important subgroups are likely to be recognized on the basis of pathoanatomical diagnoses (Petersen et al., 1999, 2003) or on the basis of patterns of symptoms and signs where the

* Corresponding author. Monash Department of Clinical Epidemiology at Cabrini Hospital, 183 Wattletree Road, Malvern, Victoria 3199, Australia. Tel.: þ61 394 898 729; fax: þ61 394 892 819. E-mail address: [email protected] (P.M. Kent). 1356-689X/$ – see front matter Ó 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.math.2008.07.003

pathoanatomy remains indistinct (Delitto et al., 1995; Fritz et al., 2003). In a similar way, views vary regarding the contribution of physiological and psychological factors to the experience of NSLBP and associated disability (activity limitation and participation restriction). Using a biopsychosocial model, some have argued that physiological factors play a minor role in NSLBP (Waddell, 1998; Zusman, 1998), especially chronic NSLBP (>12 weeks duration), while others have argued that physiological factors remain important (Murphy and Hurwitz, 2007; O’Sullivan and Beales, 2007a,b). Another area of contention in NSLBP is the relative usefulness of clinical measures from different domains of health status: pain, physical impairment (range of movement, tenderness, strength, etc.), activity limitation (ability to perform activities of daily living such as sitting, lifting, etc.), participation restriction (ability to perform societal roles such as work) and psychosocial function (depression, anxiety, coping, and fear-avoidance beliefs) (World Health Organisation, 2001). Traditionally, clinicians, such as manual therapists, have focused on measures of pain and physical impairment to guide clinical reasoning (Kent et al., 2009) but some have argued that measures of activity limitation, participation restriction and psychosocial function are more informative (Waddell, 1987; Victorian WorkCover Authority, 2004).

388

P.M. Kent et al. / Manual Therapy 14 (2009) 387–396

It is likely that beliefs regarding these aspects of NSLBP do influence clinical reasoning, treatment choices and the way in which research findings are interpreted. For example, most Australian primary care clinicians (93%) believe NSLBP to be many conditions (subgroups), although more physiotherapists believe it is possible to recognize those subgroups than do their primary care medical colleagues (Kent and Keating, 2004). Similarly, most Australian primary care clinicians (93%) base their NSLBP treatment decisions on patterns of symptoms and signs that they believe differentiate homogenous subgroups with particular treatment needs. However, primary care medical practitioners are less likely to do this than their manual therapy colleagues (chiropractors, osteopaths and physiotherapists) (Kent and Keating, 2004). In recognition of the view that there are clinically important subgroups of NSLBP, some have argued that research conducted using cohorts of people with heterogeneous NSLBP has limited generalisability to clinical practice (Rosner, 2003a,b). 1.1. Clusters of belief We hypothesised that beliefs regarding NSLBP might be discretely clustered and that clusters of belief might be associated with professional discipline, clinical training and whether a person predominately treats NSLBP patients or performs NSLBP research. For example, people who believe that NSLBP is many conditions and mostly physiological might believe that clinically useful NSLBP subgroups will be based on symptoms and signs rather than pathoanatomy, might more commonly be manual therapists, and might rate the clinical usefulness of measures of physical impairment more highly. Similarly, people who believe that NSLBP is one condition and mostly psychological might more commonly be researchers or non-manual therapy clinicians, and might more highly rate the clinical usefulness of measures of activity limitation, participation restriction and psychosocial function. We therefore hypothesised that developing a conceptual framework of NSLBP might help identify whether some groups of people have particular views, and understand more about those views. A pragmatic survey of NSLBP beliefs provided a suitable method to gather data appropriate for modeling such a conceptual framework. Although this conceptual framework would have as many dimensions as the survey questions used to create it, a simple three dimensional model is shown in Fig. 1 to illustrate the concept. 1.2. Current research aims This research aimed to (1) develop a conceptual framework of NSLBP from the views of attendees at two professional meetings where the focus was low back pain; (2) test the hypothesis that diverse beliefs within that framework are discretely ‘clustered’; and (3) test whether detected ‘clusters of belief’ are associated with demographic factors or with views regarding the clinical usefulness of different measures of health status. The relationship between these aims is shown in Fig. 2. 2. Materials and methods 2.1. Questionnaire Questionnaire development included a pilot study involving structured and systematic interviews (Dillman, 1978) of six NSLBP clinicians and researchers from a sample of convenience. The questionnaire, which is shown in Appendix 1, collected the demographic details of participants and their reported beliefs

regarding: (question 1) whether NSLBP is one or many conditions; (question 2) if clinically useful subgroups of NSLBP are to be detected is this likely to be on the basis of pathoanatomy or clusters of symptoms and signs; (question 3 with five subquestions) whether measures of pain, physical impairment, activity limitation, participation restriction and psychosocial function are clinically useful; and (questions 4–7) whether the pain and disability associated with NSLBP are primarily physiological or psychological (in the contexts of both acute and chronic pain). 2.2. Survey population and procedures Surveys were performed of two samples of convenience: 162 attendees at the 2006 Amsterdam International Forum VIII on Primary Care Research on Low Back Pain in the Netherlands and 488 attendees at a multidisciplinary professional meeting on low back pain in Melbourne, Australia (Australian Physiotherapy Association Physiotherapists’ Breakfast 2006) (total sample frame n ¼ 640). All attendees at the 2006 Amsterdam Forum were contacted after the event by email and invited to complete an on-line version of the questionnaire. All attendees at the Melbourne meeting were invited to complete a printed version of the questionnaire at the venue before the meeting started. Both forms of the questionnaire contained identical content and layout. Both meetings were open to a wide range of professional disciplines and it was not possible to anticipate the numbers and professional background of attendees. Therefore, as it was not possible to selectively sample participants to match any predetermined sample size requirements for each professional discipline, we pragmatically collected responses from all attendees who agreed to participate. Participation was voluntary and anonymous and ethics approval was gained from the Cabrini Human Research Ethics Committee. 2.3. Statistical procedures Probabilistic data-mining was used to detect the presence of ‘clusters of belief’ within respondents’ answers to questions 1–2 and 4–7. This data-mining used the ‘Vanilla SNOB’ software (Monash Data Mining Centre, Melbourne). SNOB is a form of cluster analysis/automatic classification using minimum message length methods (Wallace and Boulton, 1968; Wallace and Dowe, 2000; Wallace, 2005). It has been shown to be more accurate at recovering latent data structures than more traditional forms of cluster analysis and other current forms of automatic classification (Upal and Neufeld, 1996). Probabilistic data-mining is a form of multivariable analysis that is purely descriptive. There are no definitive sample size recommendations but Donicar (2002) suggests 5  2k (k ¼ number of variables). In the data-mining in this study, there were six variables and this formula would indicate a power requirement of 320 participants. As data reporting clinician beliefs were not normally distributed, descriptive statistics (medians, modes, ranges and proportions) were used. Between group differences were tested using Mann–Whitney U and Kruskal–Wallis tests (P < 0.05). Differences between groups were reported if they were both statistically significant and clinically important in size. Clinically important was arbitrarily defined as a difference between the most common (modal) scores of two groups that was 2 on a 7-point scale. Comparisons between groups were planned based on four groupings: comparisons between the participants identified by data-mining as belonging to groups with particular clusters of belief; comparisons between participants from different professional disciplines; comparisons between those with

P.M. Kent et al. / Manual Therapy 14 (2009) 387–396

389

Fig. 1. Example of hypothesis setting within a NSLBP conceptual framework (data points are joined with lines to accentuate the three dimensionality of the data).

a special interest in low back pain and those without this special interest; and comparisons between participants who work with NSLBP predominately as clinicians and participants who work with NSLBP predominately as researchers. ‘Predominantly clinicians’ was arbitrarily defined as participants who spend

>20% of their professional time caring for people with NSLBP and no time researching NSLBP. ‘Predominantly researchers’ was arbitrarily defined as participants who spend 20% of time caring for people with NSLBP and 20% of their time researching NSLBP.

Fig. 2. Relationship of aims of this research in developing a conceptual framework for NSLBP.

390

P.M. Kent et al. / Manual Therapy 14 (2009) 387–396

3. Results 3.1. Survey response Overall, there was an 84% response rate (n ¼ 544), consisting of an 89.4% response rate from attendees at the Melbourne meeting and a 66.7% response from attendees at the Amsterdam Forum. Participant demographics are shown in Table 1. Overall, most participants worked predominately as clinicians, most had a special interest in low back pain and the largest professional discipline represented was physiotherapy (70.3%). As the descriptive characteristics of people who declined participation in the survey were not measured, non-responder bias cannot be determined. The number of participants from different professional disciplines ranged from 383 physiotherapists to three psychologists. However, sample size calculations in previous work (Kent and Keating, 2004) indicate that there were too few participants in the current study for the results of comparisons between professional groups to be generalizable to the broader population. Therefore, although a summary of the modal responses of participating professional groups are included as Appendix 2, no analysis of these data using inferential statistics was performed. Of the responders, 273 (50.2%) were classified as working predominately as clinicians and 69 (12.7%) were classified as working predominately as researchers. A further 202 (37.1%) did not fall within either classification, as their clinical and research workload was too mixed to be classified with our arbitrary classification criteria. Regardless of classification, the views of all responders were included in the data used to model a conceptual framework.

3.2. Question responses Participant responses to questions 1, 2, 4–7 are shown in Fig. 3. Their responses to these questions showed central tendencies, with most respondents being prepared to commit to responses that favoured one end of the response scale. Most respondents favoured the beliefs that NSLBP is many conditions (88.4%), that clinically useful subgroups are most likely to be classified on the basis of symptoms and signs rather than

Table 1 Demographic details of responders. Overall (n ¼ 544) Age: mean (SD), range Years of clinical practice: mean (SD), range Physiotherapists General Medical Practitioners Specialist Physicians (e.g. Rheumatologists, etc.) Musculoskeletal Medicine Practitionersa Surgeons Chiropractors Osteopaths Psychologists Other occupation Missing occupation Predominantly clinicians (20% of time is spent caring for people with NSLBPb & no time spent researching NSLBP) Predominantly researchers (20% of time is spent caring for people with NSLBP & >20% time spent researching NSLBP) Special interest in low back pain a

38.5 (11.2), 21–80 13.2 (13.2), 0–44 384 (70.3%) 41 (7.5%) 27 (4.9%) 16 (2.9%) 14 (2.6%) 9 (1.6%) 4 (0.7%) 3 (0.5%) 42 (7.7%) 4 (0.7%) 273 (50.2%) 69 (12.7%) 374 (68.8%)

Musculoskeletal Medicine Practitioners either have a university-based postgraduate qualification in orthopaedic, spinal and musculoskeletal evaluation and treatment, or have a special interest in orthopaedic, spinal and musculoskeletal evaluation and treatment. b NSLBP ¼ nonspecific low back pain.

pathoanatomy (62.5%), that the pain (65.1%) and disability (57.7%) associated with acute NSLBP are more physiological than psychological, and that the pain (68.1%) and disability (72.1%) associated with chronic NSLBP are more psychological than physiological. However, within these proportions of respondents, participants varied in their strength of belief. Similarly, although participants’ responses showed these central tendencies, their beliefs also traverse the full spectrum of response options, indicating a diversity of beliefs. Where present, missing data were less than 1.0% and were therefore ignored. Table 2 shows responses to the question 3 ‘For providing care in nonspecific low back pain, how useful are measures of physical impairment, pain, activity limitation, participation restriction, psychosocial function?’ Most responses indicate that participants believe all these domains of health status to be quite useful, however, responses also ranged across the full spectrum from ‘not useful at all’ to ‘extremely useful’. Where present, missing data were less than 0.6% and were therefore ignored. Overall, for the 11 questions and part-questions about the nature and behaviour of NSLBP, participant responses ranged across nearly all the possible response options for each question. As all these questions about NSLBP had seven possible response options, there were a total of 77 possible responses. Participants collectively used 74 (96.1%) of these response options, illustrating the very diverse views that were expressed. 3.3. Clusters of belief Probabilistic data-mining did not detect the presence of multiple ‘clusters of belief’ within respondents’ answers. The SNOB software reported that the variability in participants’ scores across questions 1–2 and 4–7 was best explained by all participants belonging to one population of respondents and not by any statistical models containing subgroups of participants. As there were 544 subjects in the study, it is likely that this data-mining had adequate statistical power. 3.4. Comparisons between groups Participants who had a special interest in low back pain reported similar beliefs to those without this special interest. In the same way, participants who were predominately clinicians reported similar beliefs to those who were predominantly researchers, however, there was one question where they differed. For the question ‘For providing care in NSLBP how clinically useful are measures of physical impairment?’ the modal response from clinicians was 6 and for researchers was 2 (P ¼ .000) indicating that people working predominantly as clinicians think that physical impairment measures are much more useful than people working predominantly as researchers. 4. Discussion This research aimed to develop a conceptual framework of NSLBP from the views of people who treat and research NSLBP. The first step in this process was to test the hypothesis that beliefs within that framework are diverse but discretely ‘clustered’. Very diverse beliefs were reported but multiple ‘clusters of belief’ were not detected by data-mining. Therefore, it was not possible to undertake the further aims of testing if such clusters were associated with demographic factors or with views regarding the clinical usefulness of different measures of health status. Reinforcing earlier findings (Kent and Keating, 2004), there was variability in the beliefs commonly reported by participants from

38.4%

40% 30.7%

30% 19.3%

20% 10% 2.4% 2.8% 1.1%

0%

Proportion of respondents

50%

1

2

5.3%

3

One condition

4

5

6

7

Proportion of respondents

50% 36.5%

30%

26.7%

26.3%

20% 10%

6.5% 1.9%

0%

1

2.0%

2

3

4

5

21.5%

20% 10% 0%

16.9% 12.2%

8.9%

7.6%

2

3

4.1%

1

4

5

6

Pathoanatomy Unsure which method

0.0%

6

7

50% 40% 31.9%

30%

25.8%

24.5%

20% 12.1%

10%

4.5%

1.3%

0%

All psychological

1

0.0%

2

3

All physiological

4

5

Equally both

6

Question 5: In ACUTE NSLBP the DISABILITY is based in physiological or psychological phenomena?

50% 42.4%

40% 25.1%

23.1%

20% 10% 0%

6.3% 0.4%

1

2.2%

0.6%

2

3

All physiological

4 Equally both

5

6

7

All psychological

Question 6: In CHRONIC NSLBP the PAIN is based in physiological or psychological phenomena?

7

All psychological

Question 4: In ACUTE NSLBP the PAIN is based in physiological or psychological phenomena?’

30%

7

Symptoms & signs

Proportion of respondents

Equally both

28.8%

30%

Proportion of respondents

All physiological

40%

Question 2: Subgrouping NSLBP is likely to be on the basis of pathoanatomic diagnoses or on clusters of symptoms and signs (with indistinct pathoanatomy)?

Question 1: NSLBP is one condition or a number of conditions?

40%

391

50%

Many conditions

Uncertain

Proportion of respondents

Proportion of respondents

P.M. Kent et al. / Manual Therapy 14 (2009) 387–396

50% 40%

33.7%

36.7%

30% 21.1%

20% 10% 0%

5.9% 0.0%

0.9%

1

2

All physiological

1.7%

3

4 Equally both

5

6

7

All psychological

Question 7: In CHRONIC NSLBP the DISABILITY is based in physiological or psychological phenomena?

Fig. 3. Responses to questions regarding nonspecific low back pain (NSLBP).

different professional disciplines (Appendix 2) and these may be useful in hypothesis setting for further studies designed to sample sufficient participants for the testing of significant differences of opinion across disciplines. The finding that most respondents favoured the belief that NSLBP is many conditions verifies findings from our earlier study of clinician beliefs (Kent and Keating, 2004). The current study additionally found that low back pain researchers attending a primary care research conference had similar views. Other new insights were: that many respondents believed that clinically useful

subgroups are most likely to be classified on the basis of symptoms and signs rather than pathoanatomy; that the pain and disability associated with acute NSLBP are mostly physiological; and that the pain and disability associated with chronic NSLBP are mostly psychological. However, these findings are not sufficient in themselves to explain the diversity of beliefs. The beliefs of people with and without a special interest in LBP were similar. The beliefs of people working predominantly as clinicians and researchers only markedly diverged over one question. Those who were predominantly clinicians tended to rate the

392

P.M. Kent et al. / Manual Therapy 14 (2009) 387–396

Table 2 Responses to question: ‘For providing care in nonspecific low back pain, how useful are measures of.’.

Physical impairment Pain Activity limitation Participation restriction Psychosocial function

Median

25th–75th percentile

Range

5 5 6 6 6

4–6 4–6 5–7 5–6 5–7

1–7 1–7 1–7 1–7 1–7

1 ¼ not at all useful and 7 ¼ extremely useful.

clinical usefulness of measures of physical impairment as ‘extremely useful’ while those who were predominantly researchers tended to rate physical impairment measures as ‘not useful at all’. The criteria used to classify people based on their NSLBP workload as predominantly clinicians or researchers were arbitrary and not everyone will agree that these criteria were ideal. However, these criteria were set prior to data collection and the results reinforce earlier findings that clinicians highly value measures of physical impairment in NSLBP (Kent et al., 2009). To our knowledge, these data are the first to quantify that people active in NSLBP research place a much lower value on these measures. The strength of evidence for the clinical utility of measures of physical impairment varies depending on the purpose for which these measures are applied and a brief summary of that evidence follows. Symptoms and signs are usually assessed for four clinical purposes: making a diagnosis, informing treatment decisions, assisting estimation of likely prognosis and monitoring outcome. Despite popular sentiment, the evidence is weak for physical impairment assisting with definitive pathonatomic diagnoses in NSLBP. In mixed LBP that includes neurocompressive LBP and NSLBP, an association has been demonstrated between centralization/peripheralisation and pain reproduction on provocative discography. However, the overall accuracy of this test has been variable and moderate at best: 51.8% (Laslett et al., 2005), 66.7% (Young et al., 2003) and 70.1% (Donelson et al., 1997). Similarly, investigators have found only very weak associations between particular physical impairments and a positive response to anaesthetic blocks of the lumbar zygapophyseal joints (Schwarzer et al., 1995; Laslett et al., 2004; Laslett, 2005). For example, ipsilateral rotation and extension, which historically have been promoted as an indicator of zygapophyseal joint pain, have an overall accuracy of 41.1% (Laslett, 2005). Furthermore, the use of diagnostic injections as reference standards of definitive diagnoses remains controversial, with some arguing that pain reproduction or ablation does not unequivocally isolate the site of primary pain generation (North et al., 1996; Carragee et al., 1999). There is preliminary evidence that some physical impairments can inform the NSLBP treatment decisions of manual therapists. In mixed LBP that includes neurocompressive LBP, centralization/peripheralisation (directional preference) has been shown to be strongly predictive of the type of exercise most likely to benefit people (Long et al., 2004). Physical impairments also are statistically significant components of clinical prediction rules shown to predict response to either manipulation or stabilization exercises (Flynn et al., 2002; Childs et al., 2004; Hicks et al., 2005). There is also preliminary evidence that some physical impairments, such as centralization, limited flexion range of movement (ROM) and greater body mass are as strongly associated with prognosis (when defined as poor outcome from an episode of NSLBP) as factors from other domains of health status (Kent and Keating, 2009). However, due to disparate methods, contradictory

findings, the likely interaction between factors and highly variable study quality, considerable uncertainty still remains regarding prognostic factors. Similarly, there is also evidence (Hahne et al., 2004) supporting the common practice of manual therapists using within-treatment-session change in impairment of range of movement as a predictor of sustained improvement (overall accuracy 74–88%). Change in particular physical impairments associated with NSLBP can be reliably measured (Van Dillen et al., 1998; Hahne et al., 2004). However, change in physical impairment is very weakly associated with change in other measures of health status (such as activity limitation or participation restriction) suggesting that it cannot be used as a proxy for directly measuring change in these other domains of health status (Deyo, 1986; Mellin, 1986; Riddle, 1997; Sullivan et al., 2000). This has led to some disagreement about the utility of measurements of impairment, as third party payers are predominantly interested in reductions of activity limitation and participation restriction. However, as changes in measures of activity and participation do not typically occur within a treatment session they often do not provide information that assists with early decisions regarding the best choice of manual therapy treatment. In contrast, immediate changes in impairments can be observed in response to clinical interventions. These varying needs of outcome measures by different users may account for some observed differences in beliefs. The overall finding of the current research is that despite there being very diverse views of the nature and behaviour of NSLBP expressed by those participating in this survey, there was no apparent unifying framework that explained the diversity of their beliefs. There may be a number of reasons for this. We asked simple closed questions that sought to elicit beliefs about contentious concepts in NSLBP. It may have been that these questions were insensitive to the subtleties of participants’ beliefs and that research using qualitative methods might better elucidate a framework that explains current beliefs. It is also possible that providing a variety of specific clinical scenarios may have provided a richer data set to model a NSLBP conceptual framework. Alternatively, it may be that there is no underlying framework that explains the diversity of current beliefs. The observed diversity of beliefs may also reflect historical views coupled with limited knowledge about the etiology of NSLBP.

5. Conclusion Conflicting views exist amongst those that participated in this survey regarding the underlying nature of NSLBP. However, within the constructs that were sampled in these surveys, there does not appear to be any unifying framework that explains the diversity of current beliefs. This is likely to reflect pervasive uncertainty about the etiology and best practice assessment of NSLBP.

Acknowledgement Peter Kent is supported by a NHMRC Health Professional Fellowship (384366) and Rachelle Buchbinder by a NHMRC Practitioner Fellowship (334010). No benefits in any form have been, or will be, received from a commercial party related directly or indirectly to the subject of this manuscript. Ethics approval provided by the Cabrini Human Research Ethics Committee (05111206).

P.M. Kent et al. / Manual Therapy 14 (2009) 387–396

Appendix 1. Questionnaire

393

394

P.M. Kent et al. / Manual Therapy 14 (2009) 387–396

P.M. Kent et al. / Manual Therapy 14 (2009) 387–396

395

Appendix 2. Modal responses (most common answer) by participants from professional groups

Question

Response options

Physio (n ¼ 383)

GP (n ¼ 41)

Physic (n ¼ 27)

Musc Med (n ¼ 16)

Surg (n ¼ 14)

Chiro (n ¼ 9)

Osteo (n ¼ 4)

Psych (n ¼ 3)

Is NSLBP one condition or a number of conditions?

1¼one condition 4¼uncertain 7¼many conditions

7

7

7

7

7

7

4

4

If NSLBP were able to be classified into clinically useful subgroups, is this likely to be on the basis of pathoanatomy or symptoms & signs?

1¼pathoanatomy 4¼unsure which method 7¼symptoms & signs

6

6

2

6

6

7

1

4

For providing care for NSLBP, measures of Physical Impairment are. For providing care for NSLBP, measures of Pain are . For providing care for NSLBP, measures of Activity Limitation are . For providing care for NSLBP, measures of Participation Restriction are . For providing care for NSLBP, measures of Psychosocial Function are.

1¼not at all useful 7¼extremely useful

6

6

6

5

4

5

2

3

5

5

4

5

4

6

7

2

6

6

6

7

4

5

4

7

6

6

7

5

5

5

5

7

6

6

6

7

5

5

6

7

The pain in acute NSLBP is . The pain in chronic NSLBP is . The disability in acute NSLBP is . The disability in chronic NSLBP is .

1¼all physiological 4¼equally both 7¼all psychological

3 5 3

3 5 3

4 5 3

2 5 4

3 4 2

4 4 3

3 4 4

2 3 2

6

5

6

6

6

4

6

6

References Brennan GP, Fritz JM, Hunter SJ, Thackeray A, Delitto A, Erhard RE. Identifying subgroups of patients with acute/subacute ‘‘nonspecific’’ low back pain – results of a randomized clinical trial. Spine 2006;31(6):623–31. Carragee EJ, Tanner CM, Yang B, Brito JL, Truong T. False-positive findings on lumbar discography. Reliability of subjective concordance assessment during provocative disc injection. Spine 1999;24(23):2542–7. Childs JD, Fritz JM, Flynn TW, Irrgang JJ, Johnson KK, Majkowski GR, Delitto A. A clinical prediction rule to identify patients with low back pain most likely to benefit from spinal manipulation: a validation study. Annals of Internal Medicine 2004;141: 920–8. Delitto A, Erhard RE, Bowling RW. A treatment-based classification approach to low back syndrome: identifying and staging patients for conservative treatment. Physical Therapy 1995;75(6):470–89. Deyo R, Rainville J, Kent D. What can the history and physical examination tell us about low back pain. Journal of the American Medical Association 1992;268(6):760–5. Deyo RA. Comparative validity of the sickness impact profile and shorter scales for functional assessment in low-back pain. Spine 1986;11(9):951–4. Dillman D. Mail and telephone surveys, the total design method. New York: John Wiley and Sons; 1978. Donelson R, Aprill C, Medcalf R, Grant W. A prospective study of centralization of lumbar and referred pain: a predictor of symptomatic discs and annular competence. Spine 1997;22(10):1115–22. Donicar S. A review of unquestioned standards in using cluster analysis for datadriven market segmentation. Australia: Faculty of Commerce, University of Wollongong, http://ro.uow.edu.au/commpapers/273; 2002 [accessed 24.06.08]. Flynn T, Fritz JW, Whitman M, Wainner RS, Magel J, Rendeiro D, Butler B, Garber M, Allison S. A clinical prediction rule for classifying patients with low back pain who demonstrate short-term improvement with spinal manipulation. Spine 2002;27(24):2835–43. Fritz JM, Delitto A, Erhard RE. Comparison of classification-based physical therapy with therapy based on clinical practice guidelines for patients with acute low back pain a randomized clinical trial. Spine 2003;28(13):1363–72. Hahne AJ, Keating JL, Wilson SC. Do within-session changes in pain intensity and range of motion predict between-session changes in patients with low back pain? Australian Journal of Physiotherapy 2004;50(1):17–23. Hicks GE, Fritz JM, Delitto A, McGill SM. Preliminary development of a clinical prediction rule for determining which patients with low back pain will respond to a stabilization exercise program. Archives of Physical Medicine & Rehabilitation 2005;86(9):1753–62.

Kent PM, Keating J. Do primary-care clinicians think that non-specific low back pain is one condition? Spine 2004;29:1022–31. Kent PM, Keating JL. Can we predict poor recovery from recent-onset nonspecific low back pain? A systematic review. Manual Therapy 2008;13:12–28. Kent PM, Keating JL, Taylor NF. Methods used by primary care clinicians to assess recent-onset non-specific low back pain. Manual Therapy 2009;14(1):88–100. Laslett M. Diagnostic accuracy of the clinical examination compared to available reference standards in chronic low back pain patients. Division of Physiotherapy, Linkoping, Sweden: Linkopings Universitet; 2005. p. 113. Laslett M, Oberg B, Aprill CN, McDonald B. Zygapophysial joint blocks in chronic low back pain: a test of Revel’s model as a screening test. BMC Journal of Musculoskeletal Disorders 2004;5(1):43. Laslett M, Oberg B, Aprill CN, McDonald B. Centralization as a predictor of provocation discography results in chronic back pain, and the influence of disability and distress on diagnostic power. The Spine Journal 2005;5:370–80. Leboeuf-Yde C. Low back pain: time to get off the treadmill. Journal of Manipulative and Physiological Therapeutics 2001;24(1):63–6. Long A, Donelson R, Fung T. Does it matter which exercise? A randomized control trial of exercise for low back pain. Spine 2004;29:2593–602. Mellin G. Chronic low back pain in men 54–63 years of age. Correlations of physical measurements with the degree of trouble and progress after treatment. Spine 1986;11(5):421–6. Monash Data Mining Centre, Monash University, Clayton, VIC, Australia. SNOB datamining software. Murphy DR, Hurwitz EL. A theoretical model for the development of a diagnosisbased clinical decision rule for the management of patients with spinal pain. BMC Musculoskeletal Disorders 2007;8:75. doi:10.1186/1471-2474-8-75. North RB, Kidd DH, Zahurak M, Piantadosi S. Specificity of diagnostic nerve blocks: a prospective, randomized study of sciatica due to lumbosacral spine disease. Pain 1996;65(1):77–85. O’Sullivan PB, Beales DJ. Diagnosis and classification of pelvic girdle pain disorders – Part 1: A mechanism based approach within a biopsychosocial framework. Manual Therapy 2007a;12:86–97. O’Sullivan PB, Beales DJ. Diagnosis and classification of pelvic girdle pain disorders, Part 2: Illustration of the utility of a classification system via case studies. Manual Therapy 2007b;12:e1–12. Petersen T, Laslett M, Thorsen H, Manniche C, Ekdahl C, Jacobsen S. Diagnostic classification of non-specific low back pain. A new system integrating patho-anatomic and clinical categories. Physiotherapy Theory & Practice 2003;19(4):213–37. Petersen T, Thorsen H, Manniche C. Classification of non-specific low back pain: a review of the literature on classifications systems relevant to physiotherapy. Physical Therapy Reviews 1999;4(4):265–81.

396

P.M. Kent et al. / Manual Therapy 14 (2009) 387–396

Riddle D. Relationships between disability and physical impairment in patients with low back pain. Virginia Commonwealth University; 1997. Rosner A. Fables or foibles: inherent problems with RCTs. Journal of Manipulative & Physiological Therapeutics 2003a;26(7):460–7. Rosner A. Response to meta-analysis published by Assendelft et al., in the Annals of Internal Medicine. Dynamic Chiropractic 2003b;21(19):34–6. Schwarzer AC, Wang SC, Bogduk N, McNaught PJ, Laurent R. Prevalence and clinical features of lumbar zygapophysial joint pain: a study in an Australian population with chronic low back pain. Annals of the Rheumatic Diseases 1995;54(2):100–6. Sullivan MS, Shoaf LD, Riddle DL. The relationship of lumbar flexion to disability in patients with low back pain. Physical Therapy 2000;80(3):240–50. Upal MA, Neufeld EM. Comparison of unsupervised classifiers. In: Proceedings of the first international conference on information, statistics and induction in science (ISIS). River Edge, NJ, USA:: World Scientific; 1996. p. 342–53. Van Dillen LR, Sahrmann SA, Norton BJ, Caldwell CA, Fleming DA, nnell MK, Woolsey NB. Reliability of physical examination items used for classification of patients with low back pain. Physical Therapy 1998;78(9):979–88.

Victorian WorkCover Authority. Clinical framework for the delivery of physiotherapy services to injured workers. Melbourne, Australia: Victorian WorkCover Authority; 2004. p. 1–15. Waddell G. A new clinical model for the treatment of low-back pain. Spine 1987;12(7):632–44. Waddell G. The back pain revolution. Edinburgh: Churchill Livingstone; 1998. p. 438. Wallace CS. Statistical and inductive inference by minimum message length. New York, USA: Springer; 2005. p. 429. Wallace CS, Boulton DM. An information measure for classification. Computer Journal 1968;11(2):185–94. Wallace CS, Dowe DL. MML clustering of multi-state, Poisson, von Mises circular and Gaussian distributions. Statistics and Computing 2000;10(1):73–83. World Health Organisation. International classification of functioning, disability and health. Geneva, Switzerland: World Health Organisation; 2001. Young S, Aprill C, Laslett M. Correlation of clinical examination characteristics with three sources of chronic low back pain. The Spine Journal 2003;3:460–5. Zusman M. Structure-oriented beliefs and disability due to back pain. Australian Journal of Physiotherapy 1998;44(1):13–20.

Manual Therapy 14 (2009) 397–403

Contents lists available at ScienceDirect

Manual Therapy journal homepage: www.elsevier.com/math

Original Article

The validity and intra-tester reliability of a clinical measure of humeral head position Leanda McKenna*, Leon Straker, Anne Smith Curtin University of Technology, School of Physiotherapy, G.P.O. Box U1987, Perth, WA 6845, Australia

a r t i c l e i n f o

a b s t r a c t

Article history: Received 18 December 2007 Received in revised form 6 May 2008 Accepted 29 June 2008

The purpose of this study was to determine the degree of criterion validity and intra-tester reliability of humeral head palpation in subjects with shoulder pathology. The study also sought to determine whether there was any effect of arm position on humeral head position in subjects with shoulder pathology. In a same day repeated measures design, 27 subjects had the distance between the most anterior portion of the humeral head and the anterior edge of the acromion measured by a radiologist using MRI (supine), and by a physiotherapist using palpation and photography (supine, sit with arm in neutral and in abduction). The Standard Error of Measurement (SEM) for the difference between MRI and palpation ranged from 3.4 to 4.4 mm and correlated significantly with palpation measures in sit (r ¼ 0.57–0.64, p  0.002). The Intraclass Correlation Coefficients (ICCs) and SEMs for intra-tester reliability were 0.85 and 2.6 mm for supine, 0.86 and 2.2 mm for sit (glenohumeral neutral), and 0.91 and 3.0 mm for sit (glenohumeral abduction). Significant differences between the positions of sit neutral and sit with abduction were found (p < 0.001). Humeral head palpation in sit abduction demonstrates sufficient validity and reliability for clinical use. Ó 2008 Elsevier Ltd. All rights reserved.

Keywords: Shoulder Palpation Validity Reliability

1. Introduction Subacromial impingement is a well recognized, painful and potentially limiting condition. Theoretically, anterior position of the humeral head in relation to the acromion may compromise the subacromial space (Michener et al., 2003), because the humerus is in greater proximity to anterior structures in elevation (Ludewig and Cook, 2002; Werner et al., 2004; Bach and Goldberg, 2006). Measuring the habitual anterior position of the humeral head in relation to the acromion may therefore be important when assessing patients with impingement syndromes. Measurement of in vivo humeral head position usually quantifies the minimal acromiohumeral or coracohumeral interval by imaging or complicated methods. These methods are not practical for repeated clinical use and a simple, palpatory technique may be more appropriate. Published methods have concentrated on evaluating the superior position of the humeral head in relation to the acromion. However, the superior aspect of the humeral head under the acromion is difficult to palpate and assessment of the superior position of the humeral head in relation to the acromion may not be possible in the clinic. Given

* Corresponding author. Tel.: þ61 8 9266 3660; fax: þ61 8 9266 3699. E-mail address: [email protected] (L. McKenna). 1356-689X/$ – see front matter Ó 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.math.2008.06.004

that anterior and superior migrations of the humeral head (and resultant proximity to subacromial structures) during flexion appear to be closely linked (Harryman et al., 1990a,b; Werner et al., 2004), anterior assessment may also provide insight into acromiohumeral proximity. The anterior portion of the humeral head and acromion are easily located by palpation may provide a useful technique to enhance the therapeutic decision-making process and has been used previously as an assessment tool (Mckenna et al., 2001). 1.1. Validity and reliability of a palpatory method Humeral head palpation has been examined for reliability (Bryde et al., 2005) in a healthy population, but not in a population with pathology. Additionally, there does not appear to be any validity study examining anterior humeral head position assessment that is applicable to impingement. 1.2. Criterion standard X-rays have previously been used as the criterion of gold standard against which palpatory techniques were judged in live subjects, but has possible errors of magnification, projection, patient positioning and identification of landmarks (Vaatainen et al., 1991; Sutherland and Bresina, 1992; Kladny et al., 1996; Duralde and Gauntt, 1999; Graichen et al., 1999b; Burckhardt et al.,

398

L. McKenna et al. / Manual Therapy 14 (2009) 397–403

2000; Schulze and d’Hoedt, 2001). Ultrasound accuracy is highly operator dependent (Tirman et al., 1997; Torriani and Kattapuram, 2003; Roemer et al., 2005) and CT involves high levels of radiation exposure. As a criterion standard, MRI appears to have the advantages of high tissue contrast (Kladny et al., 1996; Eckstein et al., 2001) with high resolution (Torriani and Kattapuram, 2003) and therefore may be theoretically superior in measuring distances.

1.3. Influence of testing position One disadvantage of using MRI as a validation tool is that closed MRI machines do not allow shoulder joint elevation and images are generally taken in a neutral position that can be sustained by the patient. Closed MRI is often the standard choice for patients with impingement type symptoms who require imaging. Thus any validity study in patients with impingement that uses closed MRI as the reference comparison should use the same glenohumeral joint position as that utilized within the MRI machine. However, neutral positions may not be provocative enough to elicit a pathological habitual humeral head position. Elevation is recognized as a provocative position for glenohumeral impingement because the subacromial space narrows in elevation (Warner et al., 1994; Allmann et al., 1997; Moffet et al., 1998; Graichen et al., 1999a,b; Hinterwimmer et al., 2003) and may be reduced to 1.0 mm in patients with impingement (Allmann et al., 1997; Graichen et al., 1999b). Measurement of humeral head position in elevation is therefore likely to be used where clinicians wish to elicit symptoms that are difficult to provoke. Validity therefore needs to include both the neutral and the elevation positions to address the dual issues of appropriate validation and clinical need.

1.4. Purpose of this study Firstly, to determine the degree of criterion validity of a palpatory method of humeral head position measurement in subjects who have shoulder pathology. Secondly, to determine the degree of intra-tester reliability of a palpatory method of humeral head position measurement in subjects who have shoulder pathology. Thirdly, to determine the effect of arm position on humeral head position (measured by a palpatory method) in subjects who have shoulder pathology. 2. Methods 2.1. Study design This was a double blind validity and intra-tester reliability study utilizing a qualified physiotherapist (LMcK) and consultant radiologist. Curtin University of Technology Human Research Ethics Committee approved this study and all rights of the individual were protected. 2.2. Subjects Subject recruitment was conducted over one year as shown in Fig. 1. Participating surgeons invited patients undergoing shoulder MRI into the study if they were aged above 18, and had a provisional diagnosis of impingement or rotator cuff disease, as determined by an orthopaedic surgeon. Subjects were excluded from the

Patients referred to Orthopaedic Specialist by General Practitioner

Some patients referred for MRI of the shoulder by Orthopaedic Specialist Patients that met the inclusion no criteria invited to join study

yes

1 patient was unable to allow extra time for testing

32 patients indicated they would like to join the study.

no

3 patients could not be contacted 1 patient could not be tested as the testing room was double booked

yes 25 patients tested immediately yes prior to MRI 27 subjects in study yes

2 patients tested immediately post MRI

Patients underwent MRI Fig. 1. Flowchart of study participants.

L. McKenna et al. / Manual Therapy 14 (2009) 397–403

study if they had a suspected shoulder fracture or grossly unstable glenohumeral joint. Potentially interested patients received written information from the surgeon’s secretary to read in the waiting room. Those interested in participating in the study indicated to the secretary that their phone number could be given to the chief investigator. Inclusion to the study was dependent upon the volunteers’ reading and signing an informed consent immediately prior to measurement at the MRI centre. A priori analysis suggested that a minimum of 24 subjects was needed to achieve 80% power to detect a systematic difference between MRI and palpatory measures of 3.0 mm (5.0 mm) using a 2-tailed t-test (a ¼ 0.05). A sample of 26 subjects has 80% power (a ¼ 0.05) to detect a correlation of 0.52.

399

weighted sagittal and proton density weighted fat saturated axial sequences in patients requiring gadolinium injection (see Fig. 3). Sagittal image resolution matrix was 256  192 pixels (repetition time 500–3250 ms, echo time 14.0–35.4 ms) and 384  224 for the axial views (repetition time 640–3000 ms, echo time 13.5– 28.5 ms). Slice thickness was 3.0 mm (1.0 mm inter-slice gap) and field of view 14.0 cm  14.0 cm. Using digital MRI images, the radiologist identified the anterior margin of the acromion and the anterior margin of the Deltoid where it overlaid the most anterior portion of the humeral head (see Fig. 3). Values drawn from images were made available to the chief investigator at the conclusion of the study. 2.4. Data processing

2.3. Data collection 2.3.1. Palpatory method Subjects completed a questionnaire, immediately prior to palpation testing at the MRI centre, which enquired about date of birth, hand dominance and shoulder pathology. Height and weight were measured. The shoulder was prepared with the application of a patch of clear adhesive plastic dressing (Tegaderm, 3M Health Care, St. Paul, Minnesota, USA). Subjects were tested using a random order for position: (1) supine with arm by side and slight external rotation, (2) sitting, neutral glenohumeral joint, and (3) sitting, arm supported in abduction. The arm was supported in abduction using a removable wooden platform set at axillary height and attached to the camera tripod. The anterior acromion was palpated and marked with a line. The examiner placed a mark on their index finger that corresponded with the point of maximal palpatory pressure. The most anterior aspect of the subject’s humeral head was palpated and the examiner placed their marked finger on this point. A photograph (see Fig. 2) was taken using a digital camera (Kodak DC4800, Eastman Kodak Company, NY) mounted 20.0 cm superiorly above the shoulder. All subject marks were removed and the procedure was repeated twice for each position.

Digital photographic images were corrected for lens pincushion distortion using a 32% correction from Colour Science Image factory Home Edition program (http://www.colour-science.com/if/ registration.htm). A custom program (High-Tech Laboratories, Perth, Western Australia) was used to calculate the humeral head distance following correction for scaling and parallax error using the formulas shown in Appendix 1. The scaling factor was taken from a rule included in the field of view. Calculation of parallax error required the vertical distance (dv in Appendix 1) between the acromion and most anterior portion of the anterior humeral head as measured from MRI by the chief investigator. Using the computer software, the investigator placed digital points on either end of the marked anterior margin of the acromion and at the marked investigator’s finger (placed against the anterior point of the humeral head). The software calculated the line of best fit along the acromion and the perpendicular distance from the line to the most anterior point of the humeral head (see Appendix 1 and Fig. 2). 2.5. Statistical analysis Statistical processing was performed using SPSS (Statistical Packages for the Social Sciences) version 13 for Macintosh. A trial

2.3.2. MRI The subject underwent MRI in supine with arm by side and in slight external rotation. A closed MRI system (General Electric Sigma 1.5T, Milwaukee, USA) with dedicated phased array shoulder coil was used. Two image sequences were used – either proton density weighted sagittal and fat saturated axial sequences or T1

Fig. 2. Photograph showing palpation technique for measurement of anterior humeral head to acromion distance (viewed from the superior aspect of the shoulder). Dashed line indicates the extrapolated acromion line. The continuous line (points 1–2) indicates the perpendicular distance between the examiner’s finger overlaying the anterior humeral head and the dashed acromial line.

Fig. 3. MRI showing measurement of anterior humeral head to acromion distance. Dashed line extrapolates the most anterior portion of the humeral head to level with the acromion. Continuous line (points 1–2) indicates the distance between anterior acromion and the dashed line.

400

L. McKenna et al. / Manual Therapy 14 (2009) 397–403

order effect was examined with a Repeated measures Analysis of Variance (RANOVA), with posthoc examination of individual trial comparisons. Validity was examined for agreement between the two methods of measuring humeral head position by calculating a mean difference between methods (Bland and Altman, 1986), Pearson’s correlation coefficient (r) and Standard Error of Measurement (SEM). SEM ¼ OMSE, where MSE is the square root of the RANOVA Mean Square Error (Stratford and Goldsmith, 1997). Intra-tester reliability of the palpatory method was examined using Intraclass Correlation Coefficient (ICC) for consistency and for absolute agreement (McGraw and Wong, 1996), and SEM. Comparison between positions was analysed using a one factor RANOVA. Posthoc analysis of contrasts between individual positions was performed to identify significant differences in position. Bivariate correlations between arm positions were also performed.

Table 2 Anterior humeral head to acromion distances (mm) as measured by palpatory and MRI methods (mean (sd)). Palpation

Supine Sit neutral Sit abduction

MRI

Trial 1

Trial 2

Trial 3

Mean of 3 trials

18.7 (6.5) 17.0 (6.4) 22.9 (7.8)

18.3 (6.4) 17.0 (6.6) 21.5 (7.8)

17.8 (5.5) 16.0 (6.1) 20.7 (8.0)

18.2 (5.8) 16.7 (6.1) 21.7 (7.5)

14.0 (5.3) NA NA

Larger values indicate a more anterior humeral head, in comparison to the acromion.

(F ¼ 49.43, p < 0.001). These two positions were also highly correlated to each other (r ¼ 0.89, p < 0.001), whereas there appeared to be no correlation with supine measures (sit neutral–supine r ¼ 0.26, p ¼ 0.200; sit abduction–supine r ¼ 0.12, p ¼ 0.555).

4. Discussion

3. Results One subject’s data were excluded from analysis as the MRI showed evidence of previous anterior dislocation (marrow oedema was evident within the posterosuperior aspect of the humeral head, with slight flattening) and therefore possible instability, despite fulfilling initial inclusion criteria. Patient demographic data are presented in Table 1. There were 21 males and 5 females. 54% of subjects had dominant arm symptoms. Pathology found at MRI included calcification, atrophy, thickening, fraying and tears of the rotator cuff tendons, bursal thickening and effusion, mild subluxation, tearing and rupture of the biceps tendon, acromioclavicular joint degeneration, glenohumeral joint synovitis, SLAP lesions, and paralabral cysts. 7 Subjects had been injected with gadolinium to enhance diagnostic findings. Mean anterior humeral head distance from the anterior acromion measured between 14.0 and 21.7 mm, depending on position and method of assessment (see Table 2). There was a trial effect within the palpatory method for sit abduction (F ¼ 3.49, p ¼ 0.039) due to a difference between trial 1 and trial 3 (F ¼ 5.31, p ¼ 0.030). Other positions did not demonstrate a significant trial effect. The mean of three trials was used for further analysis. Soft tissue overlay measured between the anterior humeral head and skin on MRI was on average 17.7 mm thick (Standard deviation (sd) ¼ 0.5 mm, range ¼ 8.0–32.0 mm). MRI measured the average anterior difference between the humeral head and acromion as 4.2, 2.7 (see Fig. 5) and 7.7 mm smaller than the palpatory measurement of the humeral head in supine, sit neutral and sit abduction, respectively (see Table 3). A pattern of higher validity for palpation in sit positions compared to supine positions was demonstrated by lower SEMs and greater correlations between palpation in sit and MRI (see Table 3 and Fig. 4). Intra-tester reliability revealed ICCs (repeated measures in each position) that were all above 0.85. The SEMs were all below 3.0 mm (see Table 4). There was a significant difference between positions (F ¼ 7.07, p ¼ 0.009) with sit abduction demonstrating the most anterior humeral head position (see Table 2). This overall difference was due to the significant difference between sit abduction and sit neutral

Table 1 Subject attributes. Attribute (unit)

Mean (standard deviation)

Range

Age (years) Weight (kg) Height (cm) Chronicity of pain (months)

52.0 87.2 175.3 30.5

22.4–76.3 60.0–146.0 157.0–188.5 2.0–120.0

(13.6) (15.8) (8.6) (32.3)

4.1. Validity Previous criterion validity studies of palpation assessment techniques of the humeral head were confined to assessment of inferior subluxation. These studies reported slightly higher correlation coefficients (rho ¼ 0.70–0.76) than that found in this study (r ¼ 0.64), which may be a result of the differences in recording (fingerbreadths versus mm) and the type of correlation coefficients used (Spearman versus Pearson). Despite finding similar correlations to each other, these authors have concluded the technique to be both valid (Prevost et al., 1987; Boyd et al., 1993) and invalid (Hall et al., 1995). These conclusions were based only on correlations and demonstrate the subjective nature of using an association to decide whether to use a technique or not. This demonstrates the need for reporting additional statistical analyses such as SEM and mean difference to allow adequate assessment of validity. Comparison of MRI to anterior palpation in the present study demonstrated that MRI values were consistently smaller than palpation values, which is to be expected as palpation included overlying soft tissue. This systematic mean difference was greatest in sit abduction (see Table 3), possibly due to the superior migration of deltoid during abduction. The systematic differences between MRI and palpation in upright positions were also expected, given the changing orientation of structures in the different body and glenohumeral positions. Systematic differences between MRI and palpation in supine measures may also have been caused by the longer measurement period for MRI, which may have allowed greater soft tissue creep. Random error factors that may have affected validity in all positions include soft tissue creep, the use of gadolinium contrast injection in 7 patients, phasic muscle firing, examiner line of vision difficulties and MRI accuracy. Patients may have different rates of creep (e.g. patients with occult glenohumeral laxity versus those with rotator cuff thickening) and this may have contributed to random error. Contrast injection was rejected as a contribution to

Table 3 Differences and relationships of palpatory and MRI measures for anterior humeral head to acromion distance. Position

Mean difference (mm) (LOA)

SEM (mm)

r (p)

MRI – supine MRI – sit neutral MRI – sit abduction

4.2 (7.3 to 1.2) 2.7 (4.6 to 0.7) 7.7 (10.2 to 5.2)

5.3 3.4 4.4

0.10 (0.630) 0.64 (<0.001) 0.57 (0.002)

Mean difference ¼ mean difference between MRI and palpation (mean of 3 trials); SEM ¼ Standard Error of Measurement; r ¼ Pearson’s correlation coefficient. LOA ¼ Limits of Agreement.

L. McKenna et al. / Manual Therapy 14 (2009) 397–403

Average difference between MRI + Palpation in Sit neutral (mm)

Palpation measures (mm)

35 30 25 20 15 10 5 0

0

5

10

15

20

25

30

35

MRI (mm)

401

15 10

7.1

5 0

0

5

10

15

20

25

30

-2.6

-5 -10 -12.4 -15

Average by MRI + Palpation in Sit neutral (mm)

Fig. 5. Bland and Altman plot of the differences between MRI and palpation in sit neutral position measures of anterior humeral head position.

Fig. 4. Scattergram of anterior humeral head to acromion distance measured by MRI versus palpation in supine, sit neutral and sit abduction (mean of 3 trials). Dotted line ¼ MRI – supine trendline. MRI – supine data point. Dashed line ¼ MRI – sit neutral trendline. ¼ MRI – sit neutral data point. Solid line ¼ MRI – sit abduction trendline. ¼ MRI – sit abduction data point.

random error after posthoc examination showed no difference between those injected and not injected. Intermittent inhibition or phasic muscle guarding by the rotator cuff may have intermittently affected humeral head position during the test procedure. Supine palpation appeared to have greater random error, as palpation in this position correlated poorly with MRI and had the largest SEM (see Table 3) compared to other positions. The line of vision for estimating the position of landmarks varied between supine and the sit positions. In supine, posterior palpation of the acromion and humeral head was more difficult, diminishing 3D appreciation of anatomical geometry. Accuracy of MRI measures may be dependent on a number of issues including the anatomical structure being imaged, the equipment used (Eckstein et al., 2001) and the specific MRI sequence (Sittek et al., 1996). Linear error for MRI can be as high as 3.0 mm (Chandnani et al., 1991), therefore some of the random ‘error’ may have been error in the ‘gold standard’ rather than the palpatory method. In summary, systematic and random errors found in this study favour palpation in sit positions rather than the supine for measurement of the habitual humeral head position. 4.2. Reliability Previous research (Bryde et al., 2005) on anterior humeral head palpation reliability in subjects without shoulder pathology demonstrated similar results (intra-tester SEM 2.3 mm and ICC 0.86 for the arm at side position) to the present study. The arm at side position described by these authors (Bryde et al., 2005) was directly comparable to the sit neutral position used in the present study. Little difference between the ICCs for consistency and agreement was evident in the present study, demonstrating a minimal effect of systematic error for intra-tester reliability. There was

Table 4 Intra-tester reliability for palpatory measures of anterior humeral head to acromion distance. Position

ICCa (CI)

ICCc (CI)

SEM (mm)

Supine Sit neutral Sit abduction

0.85 (0.74–0.92) 0.85 (0.73–0.92) 0.91 (0.82–0.95)

0.85 (0.73–0.92) 0.86 (0.75–0.93) 0.91 (0.83–0.96)

2.6 2.2 3.0

ICCa (CI) ¼ Intraclass Correlation Coefficient for absolute agreement (confidence interval); ICCc (CI) ¼ Intraclass Correlation Coefficient for consistency (confidence interval); SEM ¼ Standard Error of Measurement.

a slight systematic trend across all arm positions for a reduction in values over time that may be a slight posterior ‘‘settling’’ of the humeral head. This was most evident in the most provocative position of sit abduction. This may be due to anterior relaxation of muscle after active abduction or guarding, or increasing posterior muscle spasm due to growing discomfort during the provocative abduction position.

4.3. Arm position It appears that glenohumeral abduction has a significant effect on humeral head position, causing an anterior shift of 5.0 mm compared to the acromion in patients with shoulder pain. This may demonstrate the provocative nature of glenohumeral abduction or alternatively the influence of a larger bulk of deltoid in the abduction position. Previous literature demonstrates that the humeral head glides an average 1.0 mm anteriorly (range 0.5– 1.5 mm) with abduction in healthy live shoulders (Graichen et al., 2000; von Eisenhart-Rothe et al., 2002; von Eisenhart-Rothe et al., 2005; Hallstrom and Karrholm, 2006; Ogston and Ludewig, 2007). One study demonstrated that shoulders with impingement lose this usual anterior glide during 10–60 abduction (Hallstrom and Karrholm, 2006). It is possible that shoulders with pathology have early or late anterior translation of the humeral head, which may explain the discrepancy between the findings from those authors and this research. Other factors such as an increase in deltoid thickness with abduction may also account for differences between results.

4.4. Clinical implications A clinical measurement of anterior/posterior humeral head position may be required to accurately and reliably detect differences of at least 5 mm, as this appears to be a possible pathological threshold, calculated from anatomical distances near glenohumeral neutral in the sagittal plane given in the literature (see Appendix 2). The literature did not provide a single linear measurement of anterior humeral head position in relation to the anterior acromion in normal or pathological shoulders and the proposed threshold is based on the calculations outlined in Appendix 2 and Fig. 6. Comparison of the pathological threshold value of 5.0 mm to SEM found in the present study demonstrates acceptable validity. The SEMs in sit (3.4 and 4.4 mm) were lower than the pathological value. However, palpation in supine should be avoided as the error (5.3 mm) was higher than the pathological threshold and neutral positions in supine do not normally elicit symptoms.

402

L. McKenna et al. / Manual Therapy 14 (2009) 397–403

5. Conclusion The findings of this study indicate that palpatory measurement of anterior humeral head position has sufficient validity and reliability for clinical use in a sit position. The errors associated with the measurement performed in sit are lower than the pathological threshold (average difference between pathology and normal). The pathological threshold was derived theoretically and the true value needs to be determined in further research. In glenohumeral abduction the humeral head occupied a position 5.0 mm more anterior than when in the neutral position, in shoulders with impingement type pathology, and clinicians may need to expect this change. Acknowledgements We thank Klaus Sussenbach for technical assistance; Dr. Ritu Gupta for statistical advice, Perth Radiological Centre and surgeon Mr. Kon Kozak for examining rooms, orthopaedic surgeons Mr. Greg Janes, and Mr. Hari Goonatillake for subject recruitment. In particular, the help of radiologist Dr. Bill Breidahl made this study possible. Appendix. Supplementary material

Fig. 6. Theoretical pathological threshold for anterior humeral head distance. D (theoretical horizontal distance from acromion to anterior humeral head) ¼ A  B þ C where A ¼ anterior margin of the acromion to the centre of the glenoid fossa distance; B ¼ the centre of the glenoid fossa to the centre of the humeral head distance; C ¼ the humeral head radius; D ¼ the anterior humeral head to anterior acromion distance. Small circle ¼ centre of humeral head. Star ¼ centre of glenoid fossa. Large circle ¼ humeral head.

Intra-tester reliability demonstrated errors (2.2–3.0 mm) lower than the critical pathological threshold of 5.0 mm in all positions. Together with good ICCs, this indicates that the method has good intra-tester reliability for clinical use. Clinicians can be 68% confident that changes of greater than 3 mm in sit positions are a result of any treatment and not measurement error. As both sit positions demonstrated adequate validity and intratester reliability, clinicians may wish to use the sit neutral position for patients who have severe or irritable pain and use the sit abduction position in those whose pain is more difficult to provoke.

4.5. Limitations The samples of subjects used in this study were not necessarily representative of the broader community that develop arm pain and impingement as these were all cases that had severe enough symptoms to warrant referral to an orthopaedic surgeon. A variety of impingement pathologies were included in this study, which prevents focus on one particular discrete syndrome. Pathologies such as traumatic acute dislocations or atraumatic multidirectional instability were not included in the study and may be subject to greater errors in reliability and validity due to the influence of greater, inconsistent and changeable rate of tissue creep. This study was limited to examining intra-tester reliability in subjects with pathology. Intra-tester reliability is an important first step in establishing reliability. The reliability results are therefore limited to one clinician reexamining a patient. Further research is required to examine inter-tester reliability and has been completed in a separate study.

Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.math.2008.06.004. References Allmann KH, Uhl M, Gufler H, Biebow N, Hauer MP, Kotter E, et al. Cine-MR imaging of the shoulder. Acta Radiologica 1997;38(6):1043–6. Bach HG, Goldberg BA. Posterior capsular contracture of the shoulder. The Journal of the American Academy of Orthopaedic Surgeons 2006;14(5):265–77. Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. The Lancet 1986;(February)::8307–10. Boyd EA, Goudreau L, O’Riain MD, Grinnell DM, Torrance GM, Gaylard A. A radiological measure of shoulder subluxation in hemiplegia: it’s reliability and validity. Archives of Physical Medicine & Rehabilitation 1993;74(2):188–93. Bryde D, Jane Freure B, Jones L, Werstine M, Kathryn Briffa N. Reliability of palpation of humeral head position in asymptomatic shoulders. Manual Therapy 2005;10(3):191–7. Burckhardt K, Gerber CH, Hodler J, Notzli H, Szekely G. Precision of distance determination using 3D to 2D projections: the error of migration measurement using X-ray images. Medical Image Analysis 2000;4(4):375–88. Chandnani VP, Ho C, Chu P, Trudell D, Resnick D. Knee hyaline cartilage evaluated with MR imaging: a cadaveric study involving multiple imaging sequences and intraarticular injection of gadolinium and saline solution. Radiology 1991;178(2):557–61. Duralde XA, Gauntt SJ. Troubleshooting the supraspinatus outlet view. Journal of Shoulder & Elbow Surgery 1999;8(4):314–9. Eckstein F, Reiser M, Englmeier KH, Putz R. In vivo morphometry and functional analysis of human articular cartilage with quantitative magnetic resonance imaging-from image to data, from data to theory. Anatomy and Embryology 2001;203(3):147–73. Graichen H, Bonel H, Stammberger T, Englmeier KH, Reiser M, Eckstein F. Subacromial space width changes during abduction and rotation-a 3-D MR imaging study. Surgical & Radiologic Anatomy 1999a;21(1):59–64. Graichen H, Bonel H, Stammberger T, Haubner M, Rohrer H, Englmeier KH, et al. Three-dimensional analysis of the width of the subacromial space in healthy subjects and patients with impingement syndrome. AJR. American Journal of Roentgenology 1999b;172(4):1081–6. Graichen H, Stammberger T, Bonel H, Karl-Hans E, Reiser M, Eckstein F. Glenohumeral translation during active and passive elevation of the shoulder – a 3D open-MRI study. Journal of Biomechanics 2000;33(5):609–13. Hall J, Dudgeon B, Guthrie M. Validity of clinical measures of shoulder subluxation in adults with poststroke hemiplegia. The American Journal of Occupational Therapy 1995;49(6):526–33. Hallstrom E, Karrholm J. Shoulder kinematics in 25 patients with impingement and 12 controls. Clinical Orthopaedics & Related Research 2006;448(July):22–7. Harryman D, Sidles JA, Matsen F. The humeral head translates on the glenoid with passive motion. In: Hawkins R, Morrey B, Post M, editors. Surgery of the shoulder. St Louis: Mosby; 1990a. p. 186–9. Harryman DT, Sidles JA, Clark JM, McQuade KJ, Gibb TD, Matsen FA. Translation of the humeral head on the glenoid with passive glenohumeral motion. The Journal of Bone & Joint Surgery – British Volume 1990b;72A(9):1334–43.

L. McKenna et al. / Manual Therapy 14 (2009) 397–403 Hinterwimmer S, Von Eisenhart-Rothe R, Siebert M, Putz R, Eckstein F, Vogl T, et al. Influence of adducting and abducting muscle forces on the subacromial space width. Medicine & Science in Sports & Exercise 2003;35(12):2055–9. Kladny B, Bail H, Swoboda B, Schiwy-Bochat H, Beyer WF, Weseloh G. Cartilage thickness measurement in magnetic resonance imaging. Osteoarthritis & Cartilage 1996;4(3):181–6. Ludewig P, Cook T. Translations of the humerus in persons with shoulder impingement symptoms. The Journal of Orthopaedic and Sports Physical Therapy 2002;32(6):248–59. McGraw KO, Wong SP. Forming inferences about some intraclass correlation coefficients. Psychological Methods 1996;1(1):30–46. Mckenna LJ, Cunningham J, Straker L. Differences in scapular and humeral head position between junior competitive swimmers and non-competitive swimmers. Presented at Inaugural International Physiotherapy Congress of the World Confederation for Physical Therapy Asia Western Pacific region, Singapore; 2001. Michener LA, McClure PW, Karduna AR. Anatomical and biomechanical mechanisms of subacromial impingement syndrome. Clinical Biomechanics 2003;18(5):369–79. Moffet H, Hebert LJ, Dufour M, Tardif J. Variation in the sub-acromial distance measured by magnetic resonance imaging during shoulder flexion and abduction movements. Canadian Journal of Rehabilitation 1998;11(4):265–7. Ogston JB, Ludewig PM. Differences in 3-dimensional shoulder kinematics between persons with multidirectional instability and asymptomatic controls. The American Journal of Sports Medicine 2007;35(8):1361–70. doi:10.1177/ 0363546507300820. Prevost R, Arsenault AB, Dutil E, Drouin G. Shoulder subluxation in hemiplegia: a radiologic correlational study. Archives of Physical Medicine & Rehabilitation 1987;68(11):782–5. Roemer FW, van Holsbeeck M, Genant HK. Musculoskeletal ultrasound in rheumatology: a radiologic perspective. Arthritis & Rheumatism 2005;53(4):491–3.

403

Schulze RK, d’Hoedt B. Mathematical analysis of projection errors in ‘‘paralleling technique’’ with respect to implant geometry. Clinical Oral Implants Research 2001;12(4):364–71. Sittek H, Eckstein F, Gavazzeni A, Milz S, Kiefer B, Schulte E, et al. Assessment of normal patellar cartilage volume and thickness using MRI: an analysis of currently available pulse sequences. Skeletal Radiology 1996;25(1):55–62. Stratford PW, Goldsmith CH. Use of the standard error as a reliability index of interest: an applied example using elbow flexor strength data. Physical Therapy 1997;77(7):745–50. Sutherland CJ, Bresina SJ. Measurement of acetabular component migration using two-dimensional radiography. Journal of Arthroplasty 1992;7(Suppl.):377–9. Tirman PF, Steinbach LS, Belzer JP, Bost FW. A practical approach to imaging of the shoulder with emphasis on MR imaging. The Orthopedic Clinics of North America 1997;28(4):483–515. Torriani M, Kattapuram SV. Musculoskeletal ultrasound: an alternative imaging modality for sports-related injuries. Topics in Magnetic Resonance Imaging 2003;14(1):103–11. Vaatainen U, Pirinen A, Makela A. Radiological evaluation of the acromioclavicular joint. Skeletal Radiology 1991;20(2):115–6. von Eisenhart-Rothe R, Matsen 3rd FA, Eckstein F, Vogl T, Graichen H. Pathomechanics in atraumatic shoulder instability: scapular positioning correlates with humeral head centering. Clinical Orthopaedics & Related Research 2005;433(April):82–9. von Eisenhart-Rothe R, Jager A, Englmeier K-H, Vogl TJ, Graichen H. Relevance of arm position and muscle activity on three-dimensional glenohumeral translation in patients with traumatic and atraumatic shoulder instability. The American Journal of Sports Medicine 2002;30(4):514–22. Warner JJ, Kann S, Maddox LM. The ‘‘arthroscopic impingement test. Arthroscopy 1994;10(2):224–30. Werner CM, Nyffeler RW, Jacob HA, Gerber C. The effect of capsular tightening on humeral head translations. Journal of Orthopaedic Research 2004;22(1):194–201.

Manual Therapy 14 (2009) 404–408

Contents lists available at ScienceDirect

Manual Therapy journal homepage: www.elsevier.com/math

Original Article

Is ‘ideal’ sitting posture real?: Measurement of spinal curves in four sitting postures Andrew P. Claus a, *, Julie A. Hides a, G. Lorimer Moseley b, Paul W. Hodges a a

NHMRC Centre of Clinical Research Excellence in Spinal Pain, Injury and Health, School of Health and Rehabilitation Sciences, The University of Queensland, Brisbane, St. Lucia, QLD 4072, Australia b Department of Physiology, Anatomy and Genetics, University of Oxford, Oxford, UK

a r t i c l e i n f o

a b s t r a c t

Article history: Received 22 October 2007 Received in revised form 11 March 2008 Accepted 28 June 2008

There is a lack of quantitative evidence for spinal postures that are advocated as ‘ideal’ in clinical ergonomics for sitting. This study quantified surface spinal curves and examined whether subjects could imitate clinically ‘ideal’ directions of spinal curve at thoraco-lumbar and lumbar regions: (i) flat – at both regions (ii) long lordosis – lordotic at both regions (iii) short lordosis – thoracic kyphosis and lumbar lordosis. Ten healthy male subjects had 3-D motion sensors adhered to the skin so that sagittal spinal curves were represented by angles at thoracic (lines between T1–T5 and T5–T10), thoraco-lumbar (T5– T10 and T10–L3) and lumbar regions (T10–L3 and L3–S2). Subjects attempted to imitate pictures of spinal curves for the flat, long lordosis, short lordosis and a slumped posture, and were then given feedback/ manual facilitation to achieve the postures. Repeated measures analysis of variance was used to compare spinal angles between posture and facilitation conditions. Results show that although subjects imitated postures with the same curve direction at thoraco-lumbar and lumbar regions (slumped, flat or long lordosis), they required feedback/manual facilitation to differentiate the regional curves for the short lordosis posture. Further study is needed to determine whether the clinically proposed ‘ideal’ postures provide clinical advantages. Ó 2008 Elsevier Ltd. All rights reserved.

Keywords: Lumbosacral region Thoracic vertebrae Sitting Posture

1. Introduction Spinal posture refers to the position of spinal segments with respect to each other and with respect to gravity. A ‘good’ posture for a specific task represents a complex interplay between biomechanics and neuromuscular function. ‘Good’ posture may be influenced by demands to prevent movement, coordinate movement, safely load spinal segments or conserve energy. Before we can examine the efficiency and safety of dynamic spinal control it is necessary to examine common, low load, and stationary postures such as standing and sitting. There are widely accepted clinical beliefs concerning ‘good’ or ‘bad’ postures, but there is little quantitative basis to define these postures. Postures have been qualitatively described according to spinal curves at the skin surface. For the standing posture, clinical literature has described ‘ideal’ spinal posture as a slight lordosis at lumbar and slight kyphosis at the thoracic regions (Kendall et al., 1983; p. 280). This ‘ideal’ sought to ‘‘involve a minimal amount of stress and strain and which is conducive to maximal efficiency in the use of the body’’ (Kendall et al., 1952; p. 5). It is normal for the

* Corresponding author. Fax: þ61 7 3365 1622. E-mail address: [email protected] (A.P. Claus). 1356-689X/$ – see front matter Ó 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.math.2008.06.001

lumbar spine to have a lordotic curve at rest (standing, supine or prone lying) (Bogduk, 2005; p. 53), but with only qualitative description of ‘ideal’ or ‘acceptable’ lumbar lordosis when observed at the skin surface, it is difficult for researchers, health practitioners and patients to know if they are talking about the same spinal curves. Current textbooks on musculoskeletal assessment (published since 2000) are used as a basis for ergonomic advice, but they lack consensus about optimal spinal curves in sitting. Qualitative descriptions and figures from textbooks appear to advocate three different spinal curve combinations as ‘ideal’ sitting posture. Firstly, a flat lower thoracic and lumbar posture have been advocated (Magee, 2006), or a flat lumbar posture with backrest support, arguing that lordosed sitting postures demand too much muscle activity (Kendall, 2005). Secondly, lordosis at both lower thoracic and lumbar regions has been advocated (Sprague, 2001). Thirdly, spinal curves similar to the standing ‘ideal’ – thoracic kyphosis with lumbar lordosis – have been advocated by some authors (Lee, 2003; O’Sullivan, 2004), and others suggest a lumbar lordosis without detail of lower thoracic curve (Bullock and Bullock-Saxton, 2000; Sahrmann, 2002). As yet, the spinal curves in these three upright postures have not been quantitatively defined. Skin surface tracking (markers/sensors adhered to skin overlying spinous processes) is an appropriate tool to quantify spinal

A.P. Claus et al. / Manual Therapy 14 (2009) 404–408

curves because it has been validated, against radiography (Gracovetsky et al., 1990) and MRI (Morl and Blickhan, 2006), to quantify the change in lumbar spinal curve between positions from flexion to extension. Furthermore, skin surface measures are relevant because clinical evaluation of posture is based on surface observation. Several studies have used skin surface measures of lumbar posture, and taught subjects to sit in the clinical ‘ideal’ slight lordosis at the lumbar spine as described for standing. In one study, subjects were given a 12-week exercise programme to improve physical range of motion and neuromuscular control to sit with a lordotic lumbar curve similar to their standing posture (Scannell and McGill, 2003). Despite the advice and intervention, all subjects sat in a more flexed lumbar position than they had during standing. Subjects had little or no ability to maintain a lumbar lordosis when they sat for 1 h (the mean lumbar angle was calculated from surface angles at L1 and S1 spinal levels). In another study, subjects were given prior training to adopt two different lumbar lordosed sitting postures, with surface angles measured at lower thoracic (angles at T6 and T12) and lumbar regions (angles at T12 and S2) (O’Sullivan et al., 2006b). One posture was lordotic at both the lower thoracic and lumbar regions. The other posture was similar to the standing ‘ideal’ with a kyphotic lower thoracic angle, and a lordotic lumbar angle. These studies provide evidence that trained subjects could sit with lordotic lumbar posture during 5 s trials (O’Sullivan et al., 2006b), but could not maintain a lordotic sitting posture for extended periods (Scannell and McGill, 2003). If therapist intervention was needed in both studies, this raises the question: can untrained subjects adopt the supposedly ‘ideal’ upright sitting postures (flat or lordotic lumbar curves) without therapist facilitation? Or put another way, are ‘ideal’ sitting postures realistically achievable? The objective of the present study was to compare the surface spinal curves of subjects when they attempted to sit in upright postures (flat and lordotic lumbar postures as clinically advocated) in two conditions (i) imitating pictures and descriptions of the postures, and (ii) with manual facilitation and feedback similar to that used in clinical rehabilitation. It was hypothesised that although the three upright postures would be physically achievable, they may not be very intuitive and may require manual facilitation and feedback. 2. Methods

405

palpation and ultrasound imaging were used to identify spinal processes, and the overlying skin was marked with ink. Sensor positions are shown in Fig. 1. For attachment of the sensors, subjects were positioned in prone lying with pillows placed under the abdomen so that the skin surface was flat from the mid thorax to sacrum. To prevent paraspinal muscle bulk from affecting sensor positions (especially during spinal extension), mounting blocks (12  25  16 mm) were used as spacers between skin and sensors. Mounting blocks were adhered to the skin with double-sided tape, and a flexible dressing tape (Fixomull stretch, BSN medical, Hamburg, Germany) was fixed over the block and surrounding skin. Spinal sensors were adhered to mounting blocks with double-sided tape, and sports tape (Leuko, BSN medical, Hamburg, Germany) was used to fix the sensor to the flexible dressing tape. Sensor cables were secured laterally with sports tape, to avoid distortion of sensors during spinal movement. The tracking device recorded 3-D position of each sensor relative to the electromagnetic source, from which relative sagittalplane positions of the sensors were derived. Spinal curves at three regions of the spine were represented by sagittal angles. These angles were derived from the line segments between 3-D sensors (Hodges et al., 1999) Fig. 1. (1) T1–T5 and T5–T10: thoracic angle. (2) T5–T10 and T10–L3: thoraco-lumbar angle. (3) T10–L3 and L3–S2: lumbar angle. T10 was chosen as the boundary between thoracic and lumbar curves, based on variation in facet joint orientation (Singer et al., 1994) and radiographs of standing posture (Roussouly et al., 2005). To avoid interference of metal with the electromagnetic source for the 3-D tracking system, a wooden stool was constructed for the sitting posture trials, and positioned within 0.7 m of the electromagnetic source. For subject comfort, the flat surface of the stool was covered with closed cell foam (8 mm), and stool height was adjusted to the level of the posterior knee crease (popliteal height). Subjects were seated towards the front of the stool (w15 cm supported from ischial tuberosities to upper thighs), knees were flexed below the height of their hips and heels were under the stool, so that the subjects could adequately tilt their pelvis in order to achieve the lordosed postures (Keegan, 1953; Mandal, 1984). Hands rested lightly on their thighs.

2.1. Subjects Ten males with a mean (SD) age of 23 (9) years, height of 178 (10) cm, and weight of 75 (9) kg participated in the experiment. Subjects were excluded if they had a history of respiratory conditions, neurological conditions, or if they had ever experienced thoracic or lumbar spinal pain that required treatment or rest from normal activities for more than 2 days. An experienced musculoskeletal physiotherapist undertook a physical examination to ensure that participants had no abnormal restriction of hip mobility, spinal mobility or scoliosis that would limit symmetrical performance of sitting postures. Written informed consent was obtained from each subject, and all procedures were approved by the institutional research ethics committee.

T1 Thoracic angle T5

T10

Thoracolumbar angle

L3 S2

Lumbar angle

2.2. Spinal curve analysis A 3-D electromagnetic tracking system (Ascension, USA, with Motion Monitor software by Innovative Sports Training) with sensor static position accuracy specified to be within 1.8 mm was used to record position data from 5 sensors adhered to the skin surface over the spinous processes at T1, T5, T10, L3 and S2. Manual

Fig. 1. The left image shows positions of the 3-D sensors attached at the skin overlying five spinal levels. The right image shows angles derived from sensor positions, to measure thoracic, thoraco-lumbar and lumbar spinal angles.

406

A.P. Claus et al. / Manual Therapy 14 (2009) 404–408

2.3. Procedure Measures of spinal alignment were made in four postures (Fig. 2), distinguished by the direction of curve at thoraco-lumbar and lumbar angles. (1) Slump: thoraco-lumbar and lumbar angles kyphosed approaching end of range. (2) Flat: thoraco-lumbar and lumbar angles vertically aligned. (3) Long lordosis: thoraco-lumbar and lumbar angles lordosed. (4) Short lordosis: thoraco-lumbar angle kyphosed or flat, lumbar angle lordosed, i.e. the clinically proposed ‘ideal’ spinal curves in standing. Short lordosis uniquely involves dissociation of curve directions between thoraco-lumbar and lumbar regions. Two intervention conditions were used (imitated and facilitated) for each of the four postures. For the imitated condition, subjects were shown pictures of each posture (Fig. 2) and the specific thoraco-lumbar and lumbar curve features were verbally described. Subjects were advised to imitate the shape of spinal curves at thoraco-lumbar and lumbar regions for each posture. For the facilitated condition, manual facilitation at spinal curves and verbal feedback of performance were provided for the three upright postures (flat, long lordosis and short lordosis shown in Fig. 2) in addition to pictures and demonstration of the spinal curves. Adoption of the slump posture did not require facilitation. For the flat posture, subjects sat towards the rear of their ischia with their coccyx almost resting on the seat. The spine was manually guided so that the skin surface was in a vertical line from lower thoracic levels to lumbar spine and sacrum. For the long and short lordosis postures, participants were taught to tilt the upper aspect of the sacrum forwards, sitting towards the front of their ischia and perineum, similar to the pelvis position on a bicycle saddle. For the long lordosis posture, the T10–L1 spinal region was manually guided anteriorly relative to the mid thoracic levels and sacrum (T10–L1 – vertically aligned with L3). In contrast, for the short lordosis posture, the T10–L1 segments were guided posteriorly relative to L3 (T10–L1 – vertically aligned over the sacrum) Fig. 2. Sitting postures were performed in random order for three trials of w45 s each. Spinal position data were collected at 100 Hz during the middle 15 s of each trial. Subjects were advised to breath naturally, avoid talking, and face forwards during data collection trials. Between trials, subjects stood up briefly, to minimise the effects of fatigue or task sequence. 2.4. Data analysis Three-dimensional spinal position data from the five sensors were exported for analysis (Matlab, The Mathworks, USA). Data

Slump

Flat

from one full respiratory cycle were extracted (w4 s, determined from antero-posterior movement of chest sensors) from each 15 s trial, to account for spinal movement due to respiration. Data were then averaged over three trials for each posture and condition. Kyphotic curves were represented as positive angles, and lordotic curves were represented as negative angles. Thoracic, thoracolumbar and lumbar angles are reported for each posture with angle means (95% CI). Although results from slump are included, the emphasis is upon comparison of the flat, long lordosis and short lordosis postures in the imitated and facilitated conditions. 2.5. Statistical analysis The three spinal angles (thoracic, thoraco-lumbar and lumbar) were compared between the seven test conditions (four postures imitated and three postures facilitated) with a repeated measures’ analysis of variance using one repeated measure (posture). The alpha level was set at p < 0.05. Where significant differences were found, post-hoc analysis was performed with Duncan’s multiple range test. 3. Results Angles were different between the three spinal regions (main effect angle – p < 0.001) and there was a significant interaction between angle and posture (main effect posture  angle  p < 0.001). Therefore the difference in angle between postures was described separately for each of the three spinal angles below. Thoracic angle: At the thoracic angle, all postures were kyphotic (Fig. 3). Post-hoc analysis showed that within postures, thoracic angles did not differ between the imitated and facilitated conditions (p > 0.10). Mean angles (95 % CI) for imitated and facilitated conditions were: flat 19.0 (15.0–23.0) and 18.0 (13.5–22.5) deg, long lordosis 16.2 (12.2–20.2) and17.6 (13.4–21.8) deg, short lordosis 18.9 (14.3–24.5) and 22.0 (17.1–26.9) deg. Thoraco-lumbar angle: At the thoraco-lumbar angle (Fig. 3), slump was kyphotic (19.2 (16.2–22.2) deg), flat showed a small degree of kyphosis in most subjects (imitated: 4.6 (0.0–9.2) deg, facilitated: 3.4 (0.5–6.3) deg conditions), long lordosis showed a small degree of lordosis (imitated: 2.6 (7.0 to 0.7) deg, facilitated: 3.0 (6.8 to 0.8) deg), and short lordosis showed a small degree of kyphosis at thoraco-lumbar angle (imitated: 3.1 (0.6 to 6.8) deg, facilitated: 3.8 (0.3–7.3) deg). Analysis of variance showed that the thoraco-lumbar angle was significantly more kyphosed in slump than the other postures (p < 0.001); similar between flat and short lordosis (p > 0.40); and significantly more lordosed in long lordosis than in the other postures (p < 0.001). Subjects were able to

Long lordosis

Short lordosis

Fig. 2. The four postures examined in this study. Postures were defined by the curve directions at thoraco-lumbar and lumbar regions. Angles were measured at thoraco-lumbar and lumbar regions as indicated by the arcs.

A.P. Claus et al. / Manual Therapy 14 (2009) 404–408

Lordosis

Kyphosis

Slump Imitated Facilitated

Flat

Long lordosis Short lordosis

407

the imitated 4.1 (7.6 to 0.6) deg, but a larger degree of lordosis in the facilitated condition 15.0 (18.3 to 11.7) deg (p < 0.001). Analysis of variance showed that the lumbar angle was significantly more kyphosed in slump than the other postures (p < 0.001); imitated flat and imitated short lordosis angles were similar (p ¼ 0.14, NS in Fig. 3) and close to 0 ; but the facilitated short lordosis and facilitated long lordosis angles were similar (p ¼ 0.32, NS in Fig. 3) and more lordosed than the other postures (p < 0.001). At the lumbar angle, results show that subjects achieved a lordotic curve in the imitated and facilitated long lordosis, but for the short lordosis most subjects required facilitation to achieve an angle that was beyond flat. 4. Discussion

-15

-5

0

15

5

25

35

15

25

Thoracic Angle (deg)

Slump

Flat

Long lordosis

NS

Short lordosis -25

-15

-5

0

5

Thoraco-lumbar Angle (deg)

Slump

Flat

NS

Long lordosis

* NS

Short lordosis

* -25

-15

-5

0

5

15

25

Lumbar Angle (deg) Fig. 3. Thoracic, thoraco-lumbar and lumbar angles are shown with mean and 95% confidence intervals for each of the postures, in the imitated and facilitated intervention conditions. *p < 0.01 – comparison between imitate and facilitate conditions that was statistically different; imitated and facilitated were similar within posture for all other comparisons. Grey bars with NS – comparison between postures where there was no difference; thoraco-lumbar and lumbar angles were different between postures for all other comparisons.

imitate thoraco-lumbar angles similar to those in the facilitated condition for each posture. Lumbar angle: At the lumbar angle (Fig. 3), slump was kyphotic 10.8 (5.8–15.8) deg, flat was close to zero (imitated: 1.5 (5.2 to 2.2) deg, facilitated: 0.1 (1.8 to 2.0) deg), long lordosis showed a lordotic curve (imitated: 9.2 (12.4 to 6.0) deg, facilitated: 13.4 (17.3 to 9.5) deg), and short lordosis showed a small degree of lordosis for

The results show that most subjects could not attain the short lordosis spinal curves (kyphotic/flat thoraco-lumbar, and lordotic lumbar region) with visual and verbal description alone. Facilitation and feedback were needed in the short lordosis to achieve a lumbar angle that was more lordotic than the flat posture. Yet most subjects were able to imitate the slump, flat and long lordosis postures (similar curve directions at thoraco-lumbar and lumbar regions) without manual facilitation. The short lordosis posture was unique in demanding different directions of spinal curve at thoraco-lumbar and lumbar regions. Results from the thoraco-lumbar angle were closely matched between the imitated and facilitated condition for all postures, and achieved the appropriate directions of spinal curve. It is interesting to note that the failure to intuitively imitate a spinal curve mostly occurred at the lumbar angle (short lordosis required facilitation to lordose at the lumbar angle). This raises the question, if a short lordosis posture is commonly adopted in standing (Berthonnaud et al., 2005) why wouldn’t it be easily achievable in sitting? The difference in hip positions between standing and sitting could be a reason for lordosis to be commonly achieved in standing but not in sitting. In a radiographic study from 1953, it was observed that hip flexion to 90 in side-lying (the hip angle commonly advocated in sitting) caused the subjects to adopt a kyphotic lumbar curve, and hip extension caused a lordotic lumbar curve (Keegan, 1953). Although subjects in the current study sat with their knees below the height of their hips, the observation that hip flexion encourages a kyphotic lumbar curve might explain why subjects in the current study had difficulty in achieving a lumbar lordosis in sitting, and subjects in another study (Scannell and McGill, 2003) had difficulty in maintaining a lumbar lordosis in sitting. These results give reason to reconsider what is the natural posture for the lumbar spinal joints. Although the anatomical position is often assumed to be a natural or ‘ideal’ lordosed posture for the lumbar spine, it is derived from postures with the hip in an extended position (standing, supine or prone). The anatomical position is not necessarily a mid-range or natural resting position for joints (e.g. glenohumeral, tibiofemoral, hip or talocrural). Despite the wedge-shape of vertebral bodies and intervertebral discs, the assumed natural or ‘ideal’ lordosis could be at least partly due to anterior tilt of the sacral base and pelvis, as a result of hip extension. The relative merits of various spinal curves need to be understood. What then, is a ‘good’ spinal posture to adopt in sitting? Kyphosed lumbar postures require less muscle activity than upright postures (Floyd and Silver, 1951; O’Sullivan et al., 2006a), but may cause greater stress to articular and ligamentous structures (Gracovetsky et al., 1990). Upright lumbar postures such as flat, long lordosis and short lordosis examined in this study are likely to approach mid-range for the lumbar joints. Although mid-range postures avoid end-range stress to ligaments, they are prone to bend, twist and shear (buckling) (Crisco and Panjabi, 1992; Adams,

408

A.P. Claus et al. / Manual Therapy 14 (2009) 404–408

1995). Computational modeling of spinal stability with the objective to prevent buckling of spinal segments has shown that midrange low-load postures place more demands upon neuromuscular control than end-range postures or high-load tasks do (Cholewicki and McGill, 1996). Neuromuscular control of spinal segments in mid-range was hypothesised to be an important determinant for safety of loading/movement of the spine (Panjabi, 1992a,b). In support of this hypothesis, a recent study with subjects in an upright semi-seated posture showed that a deficit in one aspect of neuromuscular control was a predictor for consequent development of low back pain (Cholewicki et al., 2005). The present study showed that sitting in the flat and long lordosis postures was intuitive, but further facilitation was needed for the neuromuscular control to adopt the short lordosis posture, as advocated in clinical rehabilitation texts (Lee, 2003; O’Sullivan, 2004). This suggests that the short lordosis posture is attainable, but we cannot recommend whether the posture is a realistic goal of intervention or useful training for neuromuscular control of the spine. Further research is needed to determine whether the sitting postures quantified in this study present advantages for safety or efficiency, to prevent and/or coordinate spinal movement. There are several important considerations for interpretation of the results of this study. First, this study examined only males. Replication with female subjects would be required to determine whether results are applicable to females. Second, the laboratory environment and adhesive tape used to attach electrodes to subjects’ spine, chest and pelvis may have influenced performance, although one would not expect this to compromise or enhance any specific postures. Third, although skin measures accurately represent changes in lumbar flexion/extension, skin surface measures may show smaller angles of lordosis than measures made with spinal imaging (Gracovetsky et al., 1990). Lastly, postures in this study focused only on sagittal spinal curves without back support. Other functional postures (including asymmetrical, backrest or armrest supported postures) remain to be investigated. 5. Conclusions Sitting in the short lordosis posture (thoraco-lumbar region kyphotic/flat, and lumbar region lordotic) required facilitation and feedback, but subjects imitated the slump, flat and long lordosis postures (consistent spinal curve directions at thoraco-lumbar and lumbar regions) without facilitation. Postures quantified in this study provide a foundation to examine whether particular spinal curves are advantageous in sitting. Acknowledgements (1) Paul Hodges is supported by the National Health and Medical Research Council of Australia. (2) G. Lorimer Moseley is supported by the Nuffield Medical Research Fellowship from the University of Oxford. (3) The Dorothy Hopkins award is acknowledged for financial assistance with reimbursing research subjects.

References Adams MA. Mechanical testing of the spine. An appraisal of methodology. Spine 1995;20(19):2151–6. Berthonnaud E, Dimnet J, Roussouly P, Labelle H. Analysis of the sagittal balance of the spine and pelvis using shape and orientation parameters. Journal of Spinal Disorders and Techniques 2005;18(1):40–7. Bogduk N. Clinical anatomy of the lumbar spine and sacrum. New York: Elsevier/ Churchill Livingstone; 2005. p. 53. Bullock MI, Bullock-Saxton JE. Control of low back in the workplace using an ergonomic approach. In: Twomey LT, Taylor JR, editors. Physical therapy of the low back. 3rd ed. New York: Churchill Livingstone; 2000. p. 297–326. Cholewicki J, McGill SM. Mechanical stability of the in vivo lumbar spine: implications for injury and chronic low back pain. Clinical Biomechanics (Bristol, Avon) 1996;11(1):1–15. Cholewicki J, Silfies SP, Shah RA, Greene HS, Reeves NP, Alvi K, Goldberg B. Delayed trunk muscle reflex responses increase the risk of low back injuries. Spine 2005;30(23):2614–20. Crisco JJ, Panjabi MM. Euler stability of the human ligamentous lumbar spine. Part II: experiment. Clinical Biomechanics (Bristol, Avon) 1992;7(1):27–32. Floyd WF, Silver PH. Function of erectores spinae in flexion of the trunk. Lancet 1951;1(3):133–4. Gracovetsky S, Kary M, Levy S, Ben Said R, Pitchen I, Helie J. Analysis of spinal and muscular activity during flexion/extension and free lifts. Spine 1990;15(12):1333–9. Hodges P, Cresswell A, Thorstensson A. Preparatory trunk muscle motion accompanies rapid upper limb movement. Experimental Brain Research 1999;124(1):69–79. Keegan JJ. Alterations of the lumbar curve related to posture and seating. The Journal of Bone and Joint Surgery (Am) 1953;35-A(3):589–603. Kendall FP. Posture. In: Muscles: testing and function with posture and pain. 5th ed. Baltimore: Lippincott Williams & Wilkins; 2005. p. 49–117. Kendall FP, McCreary EK, Kendall HO. Muscles, testing and function. Baltimore: Williams & Wilkins; 1983. p. 280. Kendall HO, Kendall FP, Boynton DA. Posture and pain. Baltimore: Williams & Wilkins; 1952. p. 5. Lee L. Ch 7: restoring force closure/motor control of the thorax. In: Lee D, editor. The thorax: an integrated approach. 2nd ed. Minneapolis: OPTP; 2003. p. 103–35. Magee DJ. Thoracic (dorsal) spine. In: Orthopedic physical assessment. 4th ed. Philadelphia: Saunders Elsevier; 2006. p. 425–65. Mandal AC. The correct height of school furniture. Physiotherapy 1984;70(2):48–53. Morl F, Blickhan R. Three-dimensional relation of skin markers to lumbar vertebrae of healthy subjects in different postures measured by open MRI. European Spine Journal 2006;15(6):742–51. O’Sullivan PB. ‘Clinical instability’ of the lumbar spine: its pathological basis, diagnosis and conservative management. In: Boyling JD, Jull GA, editors. Grieve’s modern manual therapy: the vertebral column. 3rd ed. Edinburgh: Churchill Livingstone; 2004. p. 311–31. O’Sullivan PB, Dankaerts W, Burnett A, Chen D, Booth R, Carlsen C, Schultz A. Evaluation of the flexion relaxation phenomenon of the trunk muscles in sitting. Spine 2006a;31(17):2009–16. O’Sullivan PB, Dankaerts W, Burnett AF, Farrell GT, Jefford E, Naylor CS, O’Sullivan KJ. Effect of different upright sitting postures on spinal–pelvic curvature and trunk muscle activation in a pain-free population. Spine 2006b;31(19):E707–12. Panjabi MM. The stabilizing system of the spine. Part I. Function, dysfunction, adaptation, and enhancement. Journal of Spinal Disorders 1992a;5(4):383–9. Panjabi MM. The stabilizing system of the spine. Part II. Neutral zone and instability hypothesis. Journal of Spinal Disorders 1992b;5(4):390–6. Roussouly P, Gollogly S, Berthonnaud E, Dimnet J. Classification of the normal variation in the sagittal alignment of the human lumbar spine and pelvis in the standing position. Spine 2005;30(3):346–53. Sahrmann S. Movement impairment syndromes of the lumbar spine. In: Diagnosis and treatment of movement impairment syndromes. St Louis: Mosby; 2002. p. 51–119. Scannell JP, McGill SM. Lumbar posture: should it and can it be modified? A study of passive tissue stiffness and lumbar position during activities of daily living. Physical Therapy 2003;83(10):907–17. Singer KP, Edmondston SJ, Day RE, Breidahl WH. Computer-assisted curvature assessment and Cobb angle determination of the thoracic kyphosis. An in vivo and in vitro comparison. Spine 1994;19(12):1381–4. Sprague RB. Differential assessment and mobilisation of the cervical and thoracic spine. In: Donatelli R, Wooden MJ, editors. Orthopaedic physical therapy. 3rd ed. New York: Churchill Livingstone; 2001. p. 108–43.

Manual Therapy 14 (2009) 409–414

Contents lists available at ScienceDirect

Manual Therapy journal homepage: www.elsevier.com/math

Original Article

Reliability and validity of a palpation technique for identifying the spinous processes of C7 and L5 Roar Robinson a, b, *, Hilde Stendal Robinson c, Gustav Bjørke b, Alice Kvale a a

University of Bergen, Department of Public Health and Primary Health Care, Section for Physiotherapy Science, Bergen, Norway Hans and Olaf Physiotherapy Clinic, Oslo, Norway c University of Oslo, Faculty of medicine, Institute for nursing and health sciences, Oslo, Norway b

a r t i c l e i n f o

a b s t r a c t

Article history: Received 16 May 2007 Received in revised form 5 June 2008 Accepted 29 June 2008

The objective was to examine inter-tester reliability and validity of two therapists identifying the spinous processes (SP) of C7 and L5, using one predefined surface palpation procedure for each level. One identification method made it possible to examine the reliability and the validity of the procedure itself. Two manual therapists examined 49 patients (29 women). Aged between 26 and 79 years, 18 were cervical and 31 lumbar patients. An invisible marking pen and ultraviolet light were used, and the findings were compared. X-rays were taken as an objective measure of the correct spinal level. Percentage agreement and kappa statistics were used to evaluate reliability and validity. The best inter-therapist agreement was found for the skin marks. Percentage agreement within 10 mm and 20 mm was 67% and 85%, respectively. The inter-tester reliability for identifying a radiological nominated SP by palpation was found to be poor for C7 and moderate for L5, with kappa of 0.18 and 0.48, respectively. The results indicated acceptable inter-therapist surface palpation agreement, but the chosen procedures did not identify the correct SP. This indicates that the procedures are not precise enough. Future reliability studies should test other non-invasive palpation procedures, both individually and in combination, and compare these with radiological investigation. Ó 2008 Elsevier Ltd. All rights reserved.

Keywords: Palpation procedure Reliability Validity Spinous process

1. Introduction The ability to palpate a spinous process (SP) is considered to be a basic skill and a prerequisite for other manual therapy techniques (Jull, 1986; Downey et al., 1999, 2003). If physiotherapists are unable to locate the same SP by palpation, it would be unreasonable to assume that other spinal manual therapy techniques have better reproducibility (Billis et al., 2003). For a variety of reasons, colleagues often examine the same patient, yet often their findings differ or are conflicting. Several studies have assessed the reliability of palpation tests for locating bony landmarks in the lumbar and sacral spine (Burton et al., 1990; Keating et al., 1990; Byfield et al., 1992; Simmonds and Kumar, 1993; McKenzie and Taylor, 1997; O’Haire and Gibbons, 2000; Harlick et al., 2007). Studies of static palpation of the L5 SP have shown acceptable intra-tester reliability, but generally poor inter-tester reliability (Burton et al.,1990; Breen,1992; Russell,1993;

* Corresponding author. Hans og Olaf fysioterapi A/S, Torggt 16, N – 0181 Oslo, Norway. Tel.: þ47 22993177; fax: þ47 22203019. E-mail address: [email protected] (R. Robinson). 1356-689X/$ – see front matter Ó 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.math.2008.06.002

Simmonds and Kumar, 1993). Some studies have demonstrated a trend for better agreement when using highly experienced physiotherapists specialised in manual therapy (Byfield et al., 1992; Jull et al., 1994; Downey et al., 1999; Billis et al., 2003). However, a systematic review of the reliability of spinal palpation for diagnosing back and neck pain concluded that neither the examiners’ education nor their experience improved reliability (Seffinger et al., 2004). In contrast to this, a recently published study by Harlick et al. (2007) concluded that inter-therapist variability had a greater effect on accuracy than any patient-defined factor. Due to common anatomical variations of the spine, advised palpation procedures may fail (Lewit, 1985; Grieve, 1994). The assumption that SPs are points rather than having a surface area (Burton et al., 1990; Simmonds and Kumar, 1993) might influence palpation results. McKenzie and Taylor (1997) attempted to correct this source of error and included an average surface area for each of the SPs of L1–L5. They examined 13 cadavers and found that the height of the SPs ranged from 16.4 to 20.4 mm and the width from 7.4 to 9.4 mm. Harlick et al. (2007) measured the height of the SPs on L1–L5 by means of X-ray and reported the mean height of L5 to be 14.1 mm. They suggested a mean SP height of 18.3 mm as level of acceptance for agreement.

410

R. Robinson et al. / Manual Therapy 14 (2009) 409–414

Binkley et al. (1995) had six orthopaedic physiotherapists to examine the mobility of the lumbar spine of 18 patients. The examination included identification of an arbitrary marked SP, and agreement on numbering this SP was found to be low. With no predefined method, they demonstrated inter-rater agreement only within 1.4 segments. Downey et al. (1999) reported a kappa of 0.69 and a weighted kappa of 0.92 for three pairs of therapists palpating the appropriate levels in the lumbar spine. If physiotherapists are unable to identify the same level, treatment and evaluation may be applied to the wrong level. Hence, there is no need to use weighted kappa. McKenzie and Taylor (1997) reported a lower overall kappa, even though they used the same palpation method as Downey et al. (1999). Christensen et al. (2002) studied the thoracic spine using the term expanded agreement; which also included the neighbouring segment. They argued that pain does not come from one segment, but from an area. However, we find this problematic, particularly when therapists use palpation to identify and treat symptomatic spinal levels, and also when evaluating patient progress. The literature on the topic is sparse, and with conflicting levels of intra- and inter-tester reliability it is difficult to argue for palpation as a reliable assessment tool. Najm et al. (2003) claimed that most research results are not comparable, due to variability in the tests, terminology, design, methodology, and statistical analyses utilised. Such inconsistencies make it difficult to rate the value of reliability studies and the results of effect studies. Furthermore, to our knowledge, no former studies have reported inter-tester reliability results for C7 identification. Although a wide variety of studies have related to radiological examination of the spine, very few have investigated either the reliability or the validity of locating SPs by palpation in relation to X-ray (Harlick et al., 2007). Hence, earlier studies might have shown satisfactory reliability, but the wrong spinal level may nevertheless have been palpated. Consequently; validity cannot be examined without confirmation. Postural assessment instruments and radiographic measurement are suggested as valid and reliable objective tools for identifying spinal levels (Haldeman et al., 1993; Troyanovich and Harrison, 1999). The SPs of C7 and L5 are easily identified on X-rays and are considered accurate if standard procedures are followed (A. Høiseth radiologist, personal communication 2004). Accordingly, we decided to use X-ray as the gold standard for identifying SPs, as used in a recently published study (Harlick et al., 2007). The main purpose of this study was to examine the inter-tester reliability of experienced manual therapists (MTs) using a predefined surface palpation technique to identify the SPs of C7 and L5, and to verify whether the markings were at the correct SP by means of X-ray (concurrent validity). We wanted to test the actual procedures described in textbooks and taught in physiotherapy schools. Hence, the focus was not to optimise the MTs’ possibilities of identifying the correct SP, but to evaluate the procedure itself. Allowing more than one procedure for each segment would have made this impossible. The amount of subcutaneous fat varies among patients and might influence palpation results (Harlick et al., 2007). Accordingly, we also decided to investigate whether patient’s body mass index (BMI) and gender influenced the findings. 2. Methods In clinical practice, the SPs of C7 and L5 are commonly used key points for identifying cervical and lumbar levels, respectively, before starting motion assessment (Magee, 2002). Guidelines for clinical identification of segmental levels are based on published descriptions (Hoppenfeld, 1976; Lewit, 1985; Grieve, 1994; Magee,

2002). In this study, C7 SP was identified through an assisted movement of the cervical spine into extension, where the C6 SP appears to move anterior (or ‘‘disappear’’) and C7 is thus the first cervical SP remaining stationary during the movement (Lewit, 1985; Magee, 2002). L5 was identified as the first SP under an imaginary line connecting the two iliac crests (Hoppenfeld, 1976; Lewit, 1985; Magee, 2002). The MTs used these procedures, among others, in their clinical practice. They attended three training sessions with the researcher to test the procedure details, including standard positioning of the participant, palpation, use of marking pen, blinding, and time consumption. Twelve healthy volunteers were tested, but no X-rays taken. The Regional Committee for Medical Research Ethics, Western Norway, provided formal ethical approval for this study. 2.1. Testers The participating MTs had between16 and 18 years in practice since completing the Norwegian postgraduate qualification in manual therapy (IFOMT standards). One radiologist with 30 years experience examined all the X-rays. 2.2. Subjects Patients referred to a particular radiology institute in Oslo, Norway, for X-ray examination of either their cervical or lumbar spine, were invited to participate in the study. Patients who had undergone surgery in the area of interest were not invited. If the MTs could neither identify the iliac crests properly nor palpate the SP in question, the patients were excluded. Cervical patients who could not maximally extend the lower cervical spine were also excluded. Information regarding diagnosis, age, gender, height and weight was collected. 2.3. Testing procedure Prior to radiographs, each participant was examined separately and in random order by both of the two MTs (GB and HSR). For practical reasons, examination was conducted in a different room to the X-ray table. Lumbar patients were examined in side lying (the same position in which the X-rays were subsequently taken). Cervical patients had their X-rays taken in the standing position, but the palpation procedures were performed with the participant seated. Reliability: Each MT palpated the SP in question and marked the position on the skin with a pen containing ink visible only under ultraviolet light, thus making it possible to blind the result for the next tester. The method has been utilised in previous studies (Burton et al., 1990; Simmonds and Kumar, 1993; McKenzie and Taylor, 1997; Downey et al., 1999; Billis et al., 2003). The researcher (RR) measured the distance between the marks using a hand held calliper and an UV lamp. Validity: To visualize the marks on X-rays, one small magnet was taped on each skin mark (Accu-Band, Magnetic Plaster, Gauss, Ito co., Ltd). The X-rays were taken and examined according to standard procedures with one anterior–posterior picture and one lateral projection; no extra X-rays were taken for the study. The outlines of the vertebrae were inspected on the X-rays, and lines were drawn perpendicular to the skin when defining the sector for C7 (L5). The area between two lines defined the sector (Figs. 1 and 2); line one midway from the lower demarcation of the C6 (L4) and the upper demarcation of C7 (L5) to the skin and line two from the upper demarcation of Th1 (S1) and the lower demarcation of C7 (L5) to the skin. The markers (magnets) were inspected and noted as being within or outside the sector. The

R. Robinson et al. / Manual Therapy 14 (2009) 409–414

411

results were reported as positive or negative identification of the actual SP marked by each of the therapists. 2.4. Statistical analysis SPSS version 12.0 was used. The percentage agreement was calculated as a measure of reliability between the two therapists, firstly based on skin markings alone and secondly on X-ray identification of the magnets and the SPs. To evaluate palpation precision, we defined and used three levels of agreement: marking at the same point (0 mm), within 10 mm and within 20 mm, respectively. These distances are within formerly reported sizes of the SPs (McKenzie and Taylor, 1997; Harlick et al., 2007). Kappa statistics were not used, since no contingency (2  2) table could be made with this data format. However, when the palpation marks were compared with the gold standard (X-ray), Cohen’s kappa (k) was used to calculate agreement between the two testers, the marks being either within or outside the sector. k effectively discounts the proportion of agreement expected by chance, and ranges in values from 1 to þ1 (Landis and Koch, 1977). Both inter-tester reliability and concurrent validity could be examined using these calculations. We also split the participants into three groups according to their BMI and used the Kruskal–Wallis test to compare the palpation results for the groups to see if BMI influenced the results. 3. Results Fig. 1. The identification sector of C7 on the X-ray was defined as: the area between one line drawn from a midpoint between the upper demarcation of C7 SP and the lower demarcation of C6 SP, and another line drawn from the midpoint between the lower demarcation of C7 SP and upper demarcation of Th1 SP, the lines being perpendicular to the skin.

The 52 patients (aged 26–79 years) who were asked to participate included 20 cervical and 32 lumbar patients. All participants were examined twice prior to radiographs, one patient was excluded because one magnet had fallen off and two patients did not meet the inclusion criteria. The study sample thus contained 49 patients; 18 cervical patients (8 females), and 31 lumbar patients (21 females) (Table 1). 3.1. Reliability: therapist agreement on skin marks The therapists had marked the same point in 18 out of 49 patients (37%). The agreement increased to 33 (67%) and 40 (82%) of 49 patients for markings within 10 and 20 mm, respectively. When the results for C7 and L5 were analysed separately, we found corresponding results (Tables 2 and 4). 3.2. Validity: therapist agreement examined on X-rays The magnets’ positions were examined using two X-ray projections; the anterior–posterior projection was used to check that both magnets were in place, and the lateral projection to evaluate the position of the magnets in relation to the defined sector. The two MTs identified the correct SP in a total of 25 (51%) and 24 (49%) of 49 participants, respectively. C7 was correctly identified in 10 (55%) and 13 (72%) participants, respectively. There was agreement between the two MTs in 8 (44%) out of 18 cervical participants. For 3 (17%) of the cervical participants, both MTs

Table 1 Demographic data. Demographic data

Fig. 2. The identification sector of L5 on the X-ray was defined as: the area between one line drawn from a midpoint between the upper demarcation of L5 SP and the lower demarcation of L4 SP, and another line drawn from the midpoint between the lower demarcation of L5 SP and the upper demarcation of S1 SP, both lines being perpendicular to the skin.

Age, range (mean) Women Normal weight (BMI 18.5–24.9) Overweight (BMI 25–29.9) Obese (BMI >30)

Cervical

Lumbar

n ¼ 18

n ¼ 31

Total n ¼ 49

26–79 (53.6) 8 7 7 4

26–79 (49) 21 14 12 5

26–79 (50.7) 29 21 19 9

412

R. Robinson et al. / Manual Therapy 14 (2009) 409–414

Table 2 Agreement between the two therapists, according to skin marks only.

C7 (n ¼ 18) L5 (n ¼ 31) Total (n ¼ 49)

Total agreement (%)

Agreement within:

0 mm

10 mm

20 mm

7 (39) 11 (35) 18 (37)

11 (61) 22 (71) 33 (67)

14 (78) 27 (87) 40 (82)

The results are presented within three levels of agreement, 0 mm, 10 mm, and 20 mm, to permit determination of the degree of palpation precision.

marked outside the sector (Table 3). Kappa was 0.18 (95% CI; 0.25 to 0.62) for identifying C7 SP. L5 SP was correctly identified by the two MTs for 15 (48%) and 11 (36%) participants, respectively. There was agreement between the MTs in 9 (29%) out of 31 participants. For 14 (45%) participants, both MTs had marked outside the sector (Table 3). Kappa was 0.48 (95% CI: 0.18–0.78) for identifying L5 SP. For 15 participants, one MT marked within and the other outside the sector (Table 3). For 7 of these 15, the distance between the marks was <20 mm. We also compared the palpation results for the participants when categorised according to international BMI classification (WHO, 1995; Seidell and Flegal, 1997) (Table 1). The Kruskal–Wallis test indicated no significant difference in distance between the MTs skin marks in the three different BMI groups (p ¼ 0.2). Neither height or weight nor gender influenced the results. 4. Discussion In this study, the best agreement between two experienced MTs was found when skin markings were compared. Agreement was 67% and 85% for agreement levels within 10 mm and 20 mm, respectively. Inter-tester reliability as determined by X-ray sector definition was found to be poor for identifying C7 SP, yet moderate for identifying L5 SP. Previous studies have used small sample sizes, and included both healthy volunteers and patients (Downey et al., 1999; Billis et al., 2003). In the present study, two therapists examined 18 cervical and 31 lumbar patients, all recruited at the same X-ray institute. Although somewhat small, our sample is heterogeneous, representing the patient population in a clinical setting, similar to that of Harlick et al. (2007) who recruited 75 patients and distributed them among five physiotherapists. A recent review on manual examination of the spine found poor reproducibility and questioned the utility of manual examination procedures in spinal diagnosis altogether (Stochkendahl et al., 2006). The present study was performed in accordance with their recommendations concerning blinding and randomisation procedures, and included standardisation and training sessions to enhance internal validity. Acceptable agreement also depends on the circumstances. In the present study, the testers were experienced MTs. They worked in the same setting and attempted to optimise agreement by using the required procedures for testing and marking. Thus, poorer agreement might be expected in an ordinary clinical setting or between different medical specialists. In clinical practice the identification procedure for SPs might be based upon the use of several identification procedures. However, in this study the purpose was not to optimise the identification of the SPs, but to evaluate one palpation procedure at each level. In our view, more studies should focus on evaluating the reliability and validity of individual procedures used in clinical practice. The next step would be to examine a cluster of the best procedures to find the best combination for identifying the SP in question. Whether inadequate reliability will result in imprecise treatment cannot be inferred from our study.

Table 3 Agreement between the two testers with X-ray results for 49 participants, split for C7 and L5. C7

N ¼ 18

MT 2 Within sector (%)

Outside sector (%)

Total

MT 1

Within sector Outside sector Total

9 (29) 2 (6) 11 (35)

6 (19) 14 (45) 20 (65)

15 (48) 16 (52) 31 (100)

L5

N ¼ 31

MT 2

MT 1

Within sector Outside sector Total

Within sector (%)

Outside sector (%)

Total

8 (44) 5 (28) 13 (72)

2 (11) 3 (17) 5 (28)

15 (48) 8 (44) 18 (100)

The table shows that for 17 participants both MTs marked inside sector and in 17 both marked outside sector. In 15 participants one MT marked outside and the other inside sector. Sector refers to the demarcation of the SP (see Figs. 1 and 2).

4.1. Identifying C7 The results suggest that there is agreement between MTs, but they fail to identify the C7 SP. This indicates poor validity for the used identification method. The therapists agreed on 12 participants out of the 18 examined, but the C7 SP was correctly identified in only 8 (44%). This could be attributable to the MTs, the participants or the procedure itself. The procedure for identifying C7 SP requires that the patient extend the cervical spine. As some of the participants were likely to be apprehensive about extending their cervical spine due to pain, stiffness, or both, this might have influenced the results. A combination of other palpation techniques, including counting the cervical SPs from occiput to C7, might have improved the results. However, this would have disqualified us from the possibility of evaluating this particular procedure. The description of the procedure in textbooks says nothing about precision (Hoppenfeld, 1976; Lewit, 1985; Grieve, 1994; Magee, 2002). The study was conducted in the X-ray institute during working hours, and we could not occupy the X-ray lab for our examinations. Hence, the procedure was carried out in an adjoining room. Repositioning in the X-ray lab might have interfered with posture and neck position, and hence influenced the palpation results. However, we concluded that this would be of little importance at the C7 level. We had some unexpected problems in demarcating the C7 SP on the X-ray and in defining the sector. C7 SP is longer than the neighbouring SPs and points both dorsally and caudally. Both the dorsal part and the distal tip of C7 then define the palpation zone. The dorsal part was often 2–3 times longer than the tip of the SP as visualized on the X-rays (Fig. 1). In some participants, the tip of the C6 SP almost covered the dorsal, proximal part of the C7 SP in a manner similar to a roof tile. This could have confused the MTs, since the best bony contact for palpation might not have been the tip of the C7 SP. Using X-rays to confirm identification of the C7 SP might then be less precise than expected. More research is needed to further investigate this phenomenon. 4.2. Identifying L5 L5 was identified as the first SP lying under an imaginary line between the iliac crests. There might be uncertainty concerning the level at which this line crosses the spine. The amount of subcutaneous tissue and difficulties in compressing the skin overlying the iliac crest could displace the line. Differences in the size of the iliac crests, both within and between individuals, have also been reported (Grieve, 1994). One earlier study has concluded that comparing iliac crest heights using observation and palpation was

R. Robinson et al. / Manual Therapy 14 (2009) 409–414

413

Table 4 Distribution of participants at each mm distance between skin marks for the two testers. Distance between skin marks, in mm

Level

C7 (n ¼ 18) L5 (n ¼ 31)

Mean (SD)

0

2

3

4

5

6

7

8

9

10

11

13

14

15

17

18

19

20

21

24

25

28

32

34

9 (10) 8 (10)

7 11

1 1

0 2

0 2

0 1

1 1

1 1

0 2

0 1

1 0

0 1

1 0

1 0

0 1

0 1

0 1

0 1

1 0

1 0

1 1

2 0

0 1

0 1

0 1

Mean distance between skin marks was 9 mm (SD 10) and 8 mm (SD 10) for C7 SP and L5 SP, respectively.

unreliable (Mann et al., 1984). According to Lewit (1985) and Grieve (1994), the horizontal or vertical position of the sacrum between the two iliac bones also varies: this is called ‘‘high’’ or ‘‘low’’ pelvis, respectively, relative to the spinal column (Lewit, 1985; Grieve, 1994). Two of our participants were on X-ray classified by the examining radiologist as having a ‘‘horizontal sacrum’’ and neither of the two MTs had palpated the correct SP. Owing to methodological differences, it is difficult to compare our results with earlier studies. Levels of the spine, positioning, inclusion criteria as well as sample sizes vary from study to study. We wanted to carry out examinations in the same positions as the X-rays were taken, and used side-lying position in the lumbar identification. This was done to minimize errors due to postural changes. The position was, however, the same for both testers and should not interfere with inter-tester results. Harlick et al. (2007) allowed the physiotherapists to use their preferred clinical assessment technique for palpating and identifying the lumbar SPs, while other studies used prone position (Simmonds and Kumar, 1993; Binkley et al., 1995). Different positions may affect the spine and the inter-vertebral position; this in turn might influence palpation accuracy. The analyses of the X-rays in this study showed variation in the size of L5 SP, consistent with the results of McKenzie and Taylor (1997) and Harlick et al. (2007). The location of L5 SP has been described in the literature as being very deep compared with L4 SP (Ebraheim et al., 1999), and L4 SP can be very prominent compared to L5 SP (Lewit, 1985; Grieve, 1994; Billis et al., 2003). Interestingly, the radiologist described L5 SP as lying very deep in four of our participants, but no measurements were made to confirm this. Both MTs marked the same skin spot in these participants; however, they missed the correct spinal level. 4.3. Reading the X-rays The same experienced radiologist, together with the researcher (RR), examined all X-rays on the testing day. The SPs were identified on the X-rays according to standard radiological definitions. There was no validation procedure between the palpation and the X-ray prior to the study, which might have improved the results. Only one former study where X-ray was used as the gold standard has been available for comparing results (Harlick et al., 2007). Their methodology was different from ours, in that the therapists were allowed to choose palpation procedures and that accuracy was evaluated differently. However, no correlation was observed between the palpation procedure used and accuracy, and they concluded that therapist variability had the greatest effect on accuracy in connection with palpation of the L1, L3 and L5 SPs. We did not find that BMI influenced the results. The study population contained few obese participants (18%) (Table 1), and the BMI distribution is representative of the distribution in the general Norwegian population (Directorate for Health and Social Affairs, 2004). 5. Conclusion The best inter-tester agreement was found when distances between skin marks were compared independently of the X-rays.

Inter-tester reliability relative to radiological identification was found to be poor for identifying the C7 SP, and moderate for identifying L5 SP. The MTs missed the correct segment in several participants, but agreed on the skin markings. This indicated good reliability, but low validity, as the used palpation methods failed to identify the correct spinal level determined through radiology. The results of this study are consistent with earlier studies. Future reliability studies should test other non-invasive palpation procedures as well, individually and in combination, and compare with radiological investigation. In the clinical setting, more than one palpation method should also be used. Acknowledgements The authors would like to thank the staff at Centrum Radiology Institute, Oslo, Norway for practical assistance during the project, particularly Arne Høiseth, MD, specialised in radiology, for valuable advice and for examining all of the X-rays in this study. References Billis EV, Foster NE, Wright CC. Reproducibility and repeatability: errors of three groups of physiotherapists in locating spinal levels by palpation. Manual Therapy 2003;8(4):223–32. Binkley JE, Stratford PW, Gill C. Interrater reliability of lumbar accessory motion mobility testing. Physical Therapy 1995;75(9):786–95. Breen A. The reliability of palpation and other diagnostic methods. Journal of Manipulative & Physiological Therapeutics 1992;15(1):54–6. Burton AK, Edwards VA, Sykes DA. Invisible skin marking for testing palpation reliability. Journal of Manual Medicine 1990;5:27–9. Byfield DC, Mathiasen J, Sangren C. The reliability of osseous landmark palpation in the lumbar spine and pelvis. European Journal of Chiropractic 1992;40: 83–8. Christensen HW, Vach W, Vach K, Manniche C, Haghfelt T, Hartvigsen L, et al. Palpation of the upper thoracic spine: an observer reliability study. Journal of Manipulative and Physiological Therapeutics 2002;25(5):285–92. Downey BJ, Taylor NF, Niere KR. Manipulative physiotherapists can reliably palpate nominated lumbar spinal levels. Manual Therapy 1999;4(3):151–6. Downey BJ, Taylor N, Niere K. Can manipulative physiotherapists agree on which lumbar level to treat based on palpation? Physiotherapy 2003;89(2):74–81. Directorate for Health and Social Affairs. Sosial- og helsedirektoratet, Forebygging og behandling av overvekt/fedme i helsetjenesten. [Prevention and treatment of overweight/obese in the national health service];, ISBN 82-8081-036-6; 2004 [In Norwegian]. Ebraheim NA, Inzerillo C, Xu R. Are anatomic landmarks reliable in determination of fusion level in posterolateral lumbar fusion? Spine 1999;24(10):973–4. Grieve G. Bony and soft-tissue anomalies of the vertebral column. In: Boyling JD, Palastanga N, editors. Grieve‘s modern manual therapy. 2nd ed. Edinburg: Churchill Livingston; 1994. p. 227–50 [chapter 17]. Haldeman S, Chapman-Smith D, Peterson DJ. Guidelines for chiropractic quality assurance and practice parameters. Burlingame, CA: Aspen Publishers; 1993. Harlick JC, Milosavljevic S, Milburn PD. Palpation identification of spinous processes in the lumbar spine. Manual Therapy 2007;12(1):56–62. Hoppenfeld S. Physical examination of the spine and extremities. AppeltonCentury-Crofts; 1976. Jull GA. Examination of the lumbar spine. In: Grieve GP, editor. Modern manual therapy of the vertebral column. Churchill Livingstone; 1986. p. 547–60. Jull G, Treleaven J, Versace G. Manual examination: is pain provocation a major cue for spinal dysfunction? Australian Physiotherapy 1994;40(3):159–65. Keating JC, Bergmann TF, Jacobs GE, Finer BA, Larson K. Interexaminer reliability of eight evaluative dimensions of lumbar segmental abnormality. Journal of Manipulative Physiological Therapeutics 1990;13(8):463–70. Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics 1977. Lewit K. Manipulative therapy in rehabilitation of the motor system. Buttersworths; 1985. Magee DJ. Orthopedic physical assessment. 4th ed. Philadelphia: WB Saunders Company; 2002.

414

R. Robinson et al. / Manual Therapy 14 (2009) 409–414

Mann M, Glasheen-Wray M, Nyberg R. Therapist agreement for palpation and observation of iliac crest heights. Physical Therapy 1984;64(3):34–8. McKenzie AM, Taylor NF. Can physiotherapists locate lumbar spinal levels by palpation? Physiotherapy 1997;83(5):235–9. Najm WI, Seffinger MA, Mishra SI, Dickerson VM, Adams A, Reinsch S, et al. Content validity of manual spinal palpatory exams – a systematic review. BMC Complement Alternative Medicine 2003;3(1):1–14. O’Haire C, Gibbons P. Inter-examiner and intra-examiner agreement for assessing sacroiliac anatomical landmarks using palpation and observation: pilot study. Manual Therapy 2000;5(1):13–20. Russell R. Diagnostic palpation of the spine: a review of procedures and assessment of their reliability. Journal of Manipulative and Physiological Therapeutics 1993;6(4):181–3. Seidell JC, Flegal KM. Assessing obesity: classification and epidemiology. British Medical Bulletin 1997;53(2):238–52.

Seffinger MA, Najm WI, Mishra SI, Adams A, Dickerson VM, Murphy LS, et al. Reliability of spinal palpation for diagnosis of back and neck pain: a systematic review of the literature. Spine 2004;29(19):E413–25. Simmonds MJ, Kumar S. Health care ergonomics. Part 2: location of body structures by palpation – a reliability study. International Journal of Industrial Economics 1993;11:145–51. Stochkendahl MJ, Christensen HW, Hartvigsen J, Vach W, Haas M, Hestbaek L, et al. Manual examination of the spine: a systematic critical literature review of reproducibility. Journal of Manipulative and Physiological Therapeutics 2006;29(6):475–85. Troyanovich SJ, Harrison DD. In reply to letter to the editor. Journal of Manipulative and Physiological Therapeutics 1999;22:182–3. WHO Expert Committee. Physical status: the use and interpretation of anthropometry. In: WHO technical report series, no. 854. Geneva: WHO; 1995.

Manual Therapy 14 (2009) 415–420

Contents lists available at ScienceDirect

Manual Therapy journal homepage: www.elsevier.com/math

Original Article

Effects of lumbopelvic joint manipulation on quadriceps activation and strength in healthy individuals Terry L. Grindstaff a, *, Jay Hertel a, James R. Beazell b, Eric M. Magrum b, Christopher D. Ingersoll a a b

University of Virginia, Charlottesville, VA, USA University of Virginia-HEALTHSOUTH, Charlottesville, VA, USA

a r t i c l e i n f o

a b s t r a c t

Article history: Received 4 June 2007 Received in revised form 24 May 2008 Accepted 28 June 2008

Lumbopelvic joint manipulation has been shown to increase quadriceps force output and activation, but the duration of effect is unknown. It is also unknown whether lower grade joint mobilisations may have a similar effect. Forty-two healthy volunteers (x  SD; age ¼ 28.37.3 yr; ht ¼ 172.8  9.8 cm; mass ¼ 76.6  21.7 kg) were randomly assigned to one of three groups (lumbopelvic joint manipulation, 1 min lumbar passive range of motion (PROM), or prone extension on elbows for 3 min). Quadriceps force and activation were measured using the burst-superimposition technique during a seated isometric knee extension task before and at 0, 20, 40, and 60 min following intervention. Collectively, all groups demonstrated a significant decrease (p < 0.001) in quadriceps force output without changes in activation (p > 0.05) at all time intervals following intervention. The group that received a lumbopelvic joint manipulation demonstrated a significant increase in quadriceps force (3%) and activation (5%) (p < 0.05) immediately following intervention, but this effect was not present after the 20 min interval. Since participants in this study were free of knee joint pathology, it is possible that they did not have the capacity to allow for large changes in quadriceps muscle activation to occur. Ó 2008 Elsevier Ltd. All rights reserved.

Keywords: Force output Manual therapy Muscle activation Sacroiliac

1. Introduction Manual therapeutic interventions such as joint mobilisation or manipulation have been shown to alter muscle force output and activation. Specific to the lower extremity, changes in muscle force output and activation have been demonstrated in the hip extensors (Yerys et al., 2002), hamstrings (Cibulka et al., 1986), quadriceps (Suter et al., 1999, 2000; Hillermann et al., 2006), soleus (Murphy et al., 1995), and gastrocnemius (Dishman and Bulbulian, 2000, 2001; Dishman et al., 2002b, 2005; Dishman and Burke, 2003). Joint mobilisation or manipulation is thought to stimulate sensory receptors in and around the joint and affects the central nervous system at the spinal segmental level (Suter et al., 1994; Murphy et al., 1995; Herzog et al., 1999; Pickar, 2002; Colloca et al., 2003, 2004; Sung et al., 2004) as well as the cortical level (Dishman et al., 2002a). The associated neurophysiological effect may be dependent on the forces (high vs. low grade joint mobilisations) applied during the manual intervention (Dishman et al., 2002a, 2005). Changes in muscle activation have been demonstrated in symptomatic (Suter et al., 1999, 2000) and healthy individuals * Corresponding author. University of Virginia, 290 Massie Road, McCue Center, PO Box 400834, Charlottesville, VA 22903, USA. Tel.: þ1 434 823 5031/þ1 434 243 2419; fax: þ1 434 243 2430. E-mail address: [email protected] (T.L. Grindstaff). 1356-689X/$ – see front matter Ó 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.math.2008.06.005

(Murphy et al., 1995; Dishman and Bulbulian, 2000, 2001; Dishman et al., 2002b; Dishman and Burke, 2003). A single lumbopelvic joint manipulation has been shown to acutely increase quadriceps force output (Suter et al., 1999, 2000; Hillermann et al., 2006) and quadriceps activation (Suter et al., 1999, 2000) in individuals with anterior knee pain. Unfortunately these studies are limited by not examining the underlying physiological mechanisms for changes in strength and function (Hillermann et al., 2006; Iverson et al., 2008) or the duration of effects (Suter et al., 1999, 2000). It is also unknown whether lower grade joint mobilisations would have a similar effect. Studying the effects of joint mobilisation and manipulation on asymptomatic individuals may provide additional insight into the neurophysiological muscle response of the intervention without the confounding, uncontrolled effects of altered muscle activation related to injury. Further understanding the neurophysiological response and duration of altered muscle activation of the quadriceps following manual intervention will help guide future studies and begin to provide scientific rational for treatment selections. Since previous studies (Suter et al., 1999, 2000) have only examined immediate changes in quadriceps force output and activation further investigation is necessary to determine if changes would be maintained over a 60 min period of time. Therefore, the purpose of this study was to determine the amount and duration of altered quadriceps force output and activation following a single high or

416

T.L. Grindstaff et al. / Manual Therapy 14 (2009) 415–420

low grade joint mobilisation/manipulation applied at the lumbopelvic region in healthy individuals over the course of 1 h. 2. Methods 2.1. Design A randomised controlled trial with one between factor, treatment group (lumbopelvic joint manipulation, passive lumbar range of motion, and prone extension) and one within factor, time (pre/ post 0, 20, 40, 60 min) was used to examine the effects of high and low grade joint mobilisation/manipulation on quadriceps force output and activation in this clinic-based study. Main outcome variables included quadriceps force output and percentage quadriceps activation. 2.2. Participants Forty-two healthy subjects volunteered for this study (Table 1). Subjects self-reported they had pain free lumbar spine and lower extremities for the past six months. Exclusion criteria included signs/symptoms indicating nerve root compression, previous spine or lower extremity surgery, osteoporosis, pregnancy, and spinal or neurological disorders. A brief health history form was completed by each subject and a standard musculoskeletal evaluation was performed and included assessment of the lumbar spine, sacroiliac, and knee joints to screen for exclusionary criteria. All subjects signed a consent form prior to participation and the study was approved by our Institutional Review Board. 2.3. Instrumentation 2.3.1. Quadriceps force output Isometric quadriceps force was measured using a load cell (Model 41, Range 1–1000 lbs; Sensotec, Columbus, OH) interfaced with a data acquisition system (MP150; Biopac Systems, Inc., Goleta, CA) and amplifier (DA100B; Biopac Systems Inc.), and sampled at 125 Hz. Subjects were seated in a custom-made chair with their hips flexed at 85 , knees flexed at 90 , and arms folded across their chest The pelvis was secured to the chair using Velcro straps, while a padded ankle strap was placed 3 cm proximal to the lateral malleolus and connected to the load cell via an ‘‘S’’ hook. 2.3.2. Burst-superimposition technique Quadriceps activation was estimated by utilising the burstsuperimposition technique on a maximum voluntary isometric contraction (MVIC). The burst-superimposition technique provides the muscle with a percutaneous supramaximal stimulus to recruit any remaining muscle fibres which have not been stimulated (Rutherford et al., 1986; Snyder-Mackler et al., 1994; Stevens et al., 2001). A superimposed burst (100 pulses/s, 600 ms pulse duration, 10 pulse tetanic train, 125 V, 100 ms duration) was manually applied to the quadriceps approximately 2 s after the beginning of the MVIC when the experimenter determined a plateau in force had occurred. Amount of muscle activation was quantified using the

Table 1 Subject demographics.

Age Height (cm) Mass (kg) Force (N) CAR (%)

Manipulation (n ¼ 15)

PROM (n ¼ 13)

Prone extension (n ¼ 13)

24.6 168.4 69.1 495.1 83.4

28.6 (8.2) 170.2 (7.0) 68.7 (8.1) 431.5 (105.7) 76.2 (12.3)

27.0 168.0 70.7 450.9 75.6

(6.2) (8.4) (16.1) (122.1) (9.9)

Values are mean (SD).

(5.9) (10.4) (14.9) (113.9) (11.9)

central activation ratio (CAR) and calculated by dividing the volitional MVIC force by total force (combined effect of the electrical superimposed burst stimulation upon the MVIC, Eq. (1)) (KentBraun and Le Blanc, 1996). A CAR of 1.00 represents complete quadriceps activation (Stackhouse et al., 2000, 2001; Mizner et al., 2003; Stevens et al., 2003; Fitzgerald et al., 2004; Lewek et al., 2004).

CAR ¼

Fvolitional Fvolitionalþelectrical

(1)

An S88 Grass Stimulator (Astro-Med, West Warwick, RI) was used with the SIU8T isolation unit (125 V stimulus) and two rubber–carbon electrodes (8  14 cm) to deliver the electrical stimuli over quadriceps. Electrode surfaces were covered with conductive gel and secured with an elastic bandage over the proximal lateral aspect and the distal medial aspect of the quadriceps muscle. The burst-superimposition technique has been shown to be highly reliable with repeated testing of healthy subjects (ICC ¼ 0.98) (Snyder-Mackler et al., 1993). 2.4. Study protocol After initial evaluation, all participants had baseline testing of quadriceps strength and quadriceps activation. Test leg was randomly determined by coin toss and all interventions and tests were performed on the same side. Participants performed a standardised warm-up consisting of four submaximal isometric contractions (50–75% MVIC) with submaximal electrical stimulation of the quadriceps and one MVIC with submaximal electrical stimulation to orient them to the test procedures. Participants were instructed to slowly build up force and hold an MVIC for 3–5 s. Verbal encouragement and visual feedback of real time force output were given. A superimposed burst (100 pulses/s, 600 ms pulse duration, 10 pulse tetanic train, 125 V, 100 ms duration) was manually applied to the quadriceps approximately 2 s after the beginning of the MVIC when the experimenter determined a plateau in force had occurred. If force did not plateau, a stimulus was not applied. A 90 s rest period was given between trials. Participants performed three trials with superimposed burst, with the average MVIC and CAR values used for data analysis. Following baseline testing, participants were randomised to one of three treatment interventions: lumbopelvic joint manipulation, side-lying lumbar mid-range flexion/extension PROM for 1 min, or lying prone (Prone Ext) on elbows for 3 min. The total duration to perform each of the three interventions was estimated at 3 min and accounted for subject positioning and intervention. Lumbopelvic joint manipulation was selected as a high grade mobilisation, while lumbar PROM was selected as a lower grade joint mobilisation. The prone on elbows intervention was selected as a sham treatment to reduce potential participant bias. Quadriceps strength and quadriceps activation were tested immediately following intervention (post 0), and at 20, 40, and 60-min post-intervention time intervals, using the same methods described above. Sixty minutes was chosen to coincide with a 60 min rehabilitation session based on common clinical practice. During rest periods between testing intervals, participants were asked to remain seated. Testing concluded after 60-min post-intervention data were collected. 2.4.1. Lumbopelvic joint manipulation The lumbopelvic joint manipulation (Flynn et al., 2006) was performed on the ipsilateral side of the test limb (Fig. 1). The term lumbopelvic was used to describe the targeted region since this manipulation technique is not exclusively specific to the lumbar, sacroiliac, or pelvic regions (Flynn et al., 2006). The manipulation procedure utilised in this study was consistent with previously

T.L. Grindstaff et al. / Manual Therapy 14 (2009) 415–420

417

Fig. 2. Passive range of motion.

2.5. Statistical analyses

Fig. 1. Lumbopelvic joint manipulation in supine with side bending (a) and rotation (b).

used methods (Flynn et al., 2002; Fritz et al., 2004; Iverson et al., 2008) and was performed by one of two physical therapists (initials, initials) with advanced manual therapy training. Subjects were positioned supine on a treatment table, while the experimenter stood on the opposite side to be manipulated. The participant was passively side-bent towards and rotated away from the selected lumbopelvic joint which was followed by the delivery of a posterior/inferior force through the opposite anterior superior iliac spine (ASIS). If a cavitation was not heard or felt by the subject or examiner, the technique was repeated. If the second attempt was not successful the procedure was repeated on the contralateral side using similar methods (Flynn et al., 2002; Fritz et al., 2004; Iverson et al., 2008). If cavitation was not heard or felt by the participant or examiner following the fourth attempt, the participant proceeded with the assessment of quadriceps activation as usual.

Subject demographics and baseline values for MVIC and CAR were compared using a one-way ANOVA. Two separate single factor repeated measures ANOVAs were performed to compare MVIC and CAR percent change scores from baseline between groups (manipulation, PROM, and Prone Ext) across each time period. A secondary analysis consisting of two post-hoc one-way ANOVAs was performed to analyse the percent change from baseline for quadriceps MVIC force and CAR values immediately following intervention. This analysis was performed to assess immediate effects of the intervention, allowing direct comparisons to similar studies. The level of statistical significance was set a priori at p < 0.05. Statistical analyses were performed with SPSS Version 14.0 (SPSS Inc., Chicago, IL).

3. Results There were no significant differences (p > 0.05) between any of the subject group demographics or baseline MVIC or CAR values (Table 1). Lumbopelvic joint cavitation was achieved in 86.7% of the individuals (54% with one attempt, 46% requiring 2–3 attempts). Only two of the subjects in the manipulation group were unable to achieve joint cavitation after four attempts (two per side), but were retained in the statistical analysis, since cavitation may not be

2.4.2. PROM Subjects were positioned side-lying on the opposite side of the test limb (Fig. 2). The experimenter held both knees with one arm while placing their opposite hand on the participant’s lumbar spine. The experimenter performed 1 min of flexion and extension PROM without reaching physiological end range in either direction of movement. 2.4.3. Prone extension on elbows Subjects were positioned prone with lumbar spine extension (Fig. 3) while using their elbows for support to maintain the position for 3 min.

Fig. 3. Prone extension on elbows.

T.L. Grindstaff et al. / Manual Therapy 14 (2009) 415–420

necessary to achieve clinically relevant changes (Flynn et al., 2003, 2006). For quadriceps MVIC force (Fig. 4) there was not a significant time by group interaction (F8,152 ¼ 1.41, p ¼ 0.20) or a significant difference between groups (F2,38 ¼ 2.55, p ¼ 0.09). When all subjects were examined independent of group assignment there was a significant difference across time intervals (F4,152 ¼ 18.45, p < 0.001) with a decrease (p < 0.01) in MVIC force from baseline at all time (0, 20, 40, 60 min) intervals following intervention. For quadriceps CAR (Fig. 5) there was not a significant time by group interaction (F8,152 ¼ 1.02, p ¼ 0.43) or significant differences in time (F4,152 ¼ 1.08, p ¼ 0.37) or between groups (F2,38 ¼ 1.12, p ¼ 0.34). To allow for comparison with previous studies (Suter et al., 1999, 2000; Hillermann et al., 2006) two post-hoc one-way ANOVAs were performed to analyse the percent change from baseline for quadriceps MVIC force and CAR values immediately following intervention. There was a significant difference between groups for MVIC (F2,38 ¼ 6.93, p ¼ 0.003) and CAR (F2,38 ¼ 3.98, p ¼ 0.03). The manipulation group demonstrated a significant increase in quadriceps force output (3.1%) compared to the PROM group (p ¼ 0.001, 95% CI ¼ 5.52, 19.44) and the Prone Ext group (p ¼ 0.02, 95% CI ¼ 1.36, 15.28). The manipulation group also significantly increased quadriceps activation (4.7%) compared to the PROM group (p ¼ 0.04, 95% CI ¼ 0.37, 9.92) and Prone Ext group (p ¼ 0.01, 95% CI ¼ 1.34, 10.90). Effect size for immediate changes in MVIC force (d ¼ 0.12) and CAR (d ¼ 0.38) following joint manipulation was also calculated.

4. Discussion The results of this study indicate that changes in quadriceps force output and activation are not present over the course of 1 h following high or low grade joint mobilisation/manipulation directed at the lumbopelvic region. The original data analysis suggested an immediate change in quadriceps force output and activation was not present following lumbopelvic joint manipulation and contrasted findings of similar studies (Suter et al., 1999, 2000; Hillermann et al., 2006). The secondary analysis was conducted to only examine immediate changes following intervention and allowed for direct comparisons with previous findings. This analysis indicated following lumbopelvic joint manipulation an acute increase in quadriceps force output (3.1%) and quadriceps activation (4.7%) was present. Although changes in quadriceps force output were less than previously reported values (11–17%) (Suter et al., 1999, 2000; Hillermann et al., 2006), changes in quadriceps activation (5–7.5%) were consistent with previously reported values (Suter et al., 1999, 2000). It appears with the

10

Percent Change

Manipulation PROM Prone Ext

*

5 0 -5 -10 -15 -20 -25 Baseline

Post 0

Post 20

Post 40

Post 60

Fig. 4. Quadriceps MVIC force. Values are expressed as percent change from baseline and standard error of the mean. *Significant from baseline p  0.05.

10 Manipulation PROM Prone Ext

* Percent Change

418

0

-5

-10 Baseline

Post 0

Post 20

Post 40

Post 60

Fig. 5. Quadriceps activation CAR. Values are expressed as percent change from baseline and standard error of the mean. *Significant from baseline p  0.05.

multifactorial design originally used in this study that an immediate increase in quadriceps force output and activation may have gone unrecognised. Thus we cannot discount the immediate findings which are in agreement with previous studies. Examination of effect sizes for the manipulation group demonstrates a small effect size (d ¼ 0.12) for immediate changes in quadriceps force output and a small, but approaching moderate (d ¼ 0.38) effect size for quadriceps activation. Due to the extremely short term effect following lumbopelvic joint manipulation the clinical relevance of these findings is questionable and interpretation should be left to the reader. The immediate increase in quadriceps force output and activation following lumbopelvic joint manipulation could be attributed to a facilitation of the motoneuron pool mediated at the spinal or cortical level. We hypothesise the underlying physiological mechanism for the distant response associated with lumbopelvic joint manipulation may be due to common sensory and motor nerve root levels with the same interneurons. Joint manipulation is thought to affect the central nervous system at the segmental level by activating structures in and around the manipulated joint (mechanoreceptors, proprioceptors, and free nerve endings) (Suter et al., 1994; Murphy et al., 1995; Herzog et al., 1999; Pickar, 2002; Colloca et al., 2003, 2004; Sung et al., 2004). Cortical changes have also been demonstrated following spinal manipulation (Dishman et al., 2002a) and may also subsequently affect motoneuron pool excitability. Since the sacroiliac joint (L2–S3), quadriceps (L2–4) and knee joints (L2–S2) share common nerve root levels (Moore and Dalley, 1999) it is possible that afferent information from one structure may alter efferent signals to all structures innervated by a similar nerve root level. Lumbopelvic joint manipulation has been shown to briefly decrease H-reflexes of the soleus (Murphy et al., 1995) and gastrocnemius (Dishman and Bulbulian, 2000, 2001; Dishman et al., 2002b, 2005; Dishman and Burke, 2003) muscles and increase quadriceps force output (Suter et al., 1999, 2000; Hillermann et al., 2006) and activation (Suter et al., 1999, 2000). This relationship has been demonstrated in reverse using a knee joint effusion model, where quadriceps inhibition and soleus facilitation were demonstrated using H-reflex measures (Hopkins et al., 2001). It is proposed that the increase in afferent information due to joint manipulation is mediated at the interneuron and can affect efferent motor output to the surrounding musculature. The clinical implication of this study is in agreement that lumbopelvic joint manipulation has the ability to immediately increase quadriceps force output and activation, but the effects in a healthy population

T.L. Grindstaff et al. / Manual Therapy 14 (2009) 415–420

are of limited duration and small, but approaching a moderate effect size (d ¼ 0.38) for changes in quadriceps activation. A limitation of this study is only healthy individuals without joint pathology were studied. It is possible since participants were free of joint pathology that their motoneuron pool availability could not be altered beyond the acute response observed in this study. An explanation for the differences in magnitude of quadriceps activation between this study (4.7%) and previous studies (Suter et al., 1999, 2000) demonstrating that a 5–7.5% increase in quadriceps activation occurs may be due to the fact that facilitation of the quadriceps may not occur as easily as disinhibition. Individuals with knee joint pathology typically have some level of quadriceps inhibition (Stratford, 1982; SnyderMackler et al., 1994; Maitland et al., 1999; Suter et al., 1999, 2000; Urbach et al., 1999, 2001; Hopkins et al., 2001; Palmieri et al., 2003, 2004; Williams et al., 2003; Chmielewski et al., 2004) which may allow for greater changes in quadriceps activation to occur following intervention. Participants in this study did have quadriceps activation levels which were below values previously published literature on healthy individuals (Stackhouse et al., 2001; Lewek et al., 2004; Hart et al., 2006a,b). CAR values greater than or equal to 0.95 have been used to describe a situation where all motoneurons have been activated during an MVIC (Stackhouse et al., 2001; Lewek et al., 2004). Approximately 80% of the individuals in this study could not achieve quadriceps activation levels above 90%. Previous studies (Stackhouse et al., 2001; Lewek et al., 2004) have estimated that 25% of younger to middle aged adults are not able to achieve full activation. Procedures used in this study including warm-up, verbal encouragement, visual feedback, and rest periods between trials were similar to other studies (Manal and Snyder-Mackler, 2000; Stackhouse et al., 2001; Lewek et al., 2002, 2004; Mizner et al., 2003, 2005; Stevens et al., 2003; Chmielewski et al., 2004; Williams et al., 2005; Hart et al., 2006a,b,c). Although the CAR values in this study were lower, measurements in the sham group were extremely stable (ICC3,k ¼ 0.97, 95% CI: 0.93–0.99; SEM ¼ 2.06%) over the course of an hour indicating stability in our measurement technique. Even if the sham intervention had a therapeutic effect, at best, this value would be underestimated and likely represents normal variability within each subject over the 60 min testing period. Results also indicated decreased quadriceps force output occurred over the course of 1 h for all groups. Obtaining valid and reliable CAR values is dependent on participants giving a maximal effort (Behm et al., 1996). Subjects were verbally encouraged during all trials to give 100% effort, but deceased force output occurred over the course of 1 h. This may have been due local muscle fatigue due to performing 15 MVICs augmented with a supramaximal electrical stimulus. The decrease in quadriceps force output could not be attributed to changes in CAR, but were more likely due to local muscle fatigue. Future studies should utilise electromyographic analysis of muscle median frequency to quantify local muscle fatigue associated with repeated MVICs (Hart et al., 2006a,b).

5. Conclusion The findings of this study indicate that there was an acute increase in quadriceps force output and quadriceps activation following lumbopelvic joint manipulation in healthy individuals, but these effects diminished within 20 min of intervention. All groups demonstrated a progressive loss of quadriceps force output over the course of testing, which was thought to be due to muscle fatigue. Additional research is necessary to accurately determine duration of changes in quadriceps force output and activation

419

following lumbopelvic joint manipulation in healthy and symptomatic individuals. References Behm DG, St-Pierre DM, Perez D. Muscle inactivation: assessment of interpolated twitch technique. Journal of Applied Physiology 1996;81:2267–73. Chmielewski TL, Stackhouse S, Axe MJ, Snyder-Mackler L. A prospective analysis of incidence and severity of quadriceps inhibition in a consecutive sample of 100 patients with complete acute anterior cruciate ligament rupture. Journal of Orthopaedic Research 2004;22:925–30. Cibulka MT, Rose SJ, Delitto A, Sinacore DR. Hamstring muscle strain treated by mobilizing the sacroiliac joint. Physical Therapy 1986;66:1220–3. Colloca CJ, Keller TS, Gunzburg R. Neuromechanical characterization of in vivo lumbar spinal manipulation. Part II. Neurophysiological response. Journal of Manipulative and Physiological Therapeutics 2003;26:579–91. Colloca CJ, Keller TS, Gunzburg R. Biomechanical and neurophysiological responses to spinal manipulation in patients with lumbar radiculopathy. Journal of Manipulative and Physiological Therapeutics 2004;27:1–15. Dishman JD, Ball KA, Burke J. Central motor excitability changes after spinal manipulation: a transcranial magnetic stimulation study. Journal of Manipulative and Physiological Therapeutics 2002a;25:1–9. Dishman JD, Bulbulian R. Spinal reflex attenuation associated with spinal manipulation. Spine 2000;25:2519–24. Dishman JD, Bulbulian R. Comparison of effects of spinal manipulation and massage on motoneuron excitability. Electromyography and Clinical Neurophysiology 2001;41:97–106. Dishman JD, Burke J. Spinal reflex excitability changes after cervical and lumbar spinal manipulation: a comparative study. The Spine Journal 2003;3:204–12. Dishman JD, Cunningham BM, Burke J. Comparison of tibial nerve H-reflex excitability after cervical and lumbar spine manipulation. Journal of Manipulative and Physiological Therapeutics 2002b;25:318–25. Dishman JD, Dougherty PE, Burke JR. Evaluation of the effect of postural perturbation on motoneuronal activity following various methods of lumbar spinal manipulation. The Spine Journal 2005;5:650–9. Fitzgerald GK, Piva SR, Irrgang JJ, Bouzubar F, Starz TW. Quadriceps activation failure as a moderator of the relationship between quadriceps strength and physical function in individuals with knee osteoarthritis. Arthritis & Rheumatism 2004;51:40–8. Flynn T, Fritz J, Whitman J, Wainner R, Magel J, Rendeiro D, et al. A clinical prediction rule for classifying patients with low back pain who demonstrate short-term improvement with spinal manipulation. Spine 2002;27:2835–43. Flynn TW, Childs JD, Fritz JM. The audible pop from high-velocity thrust manipulation and outcome in individuals with low back pain. Journal of Manipulative and Physiological Therapeutics 2006;29:40–5. Flynn TW, Fritz JM, Wainner RS, Whitman JM. The audible pop is not necessary for successful spinal high-velocity thrust manipulation in individuals with low back pain. Archives of Physical Medicine and Rehabilitation 2003;84:1057–60. Fritz JM, Whitman JM, Flynn TW, Wainner RS, Childs JD. Factors related to the inability of individuals with low back pain to improve with a spinal manipulation. Physical Therapy 2004;84:173–90. Hart JM, Fritz JM, Kerrigan DC, Saliba EN, Gansneder BM, Ingersoll CD. Quadriceps inhibition after repetitive lumbar extension exercise in persons with a history of low back pain. Journal of Athletic Training 2006a;41:264–9. Hart JM, Fritz JM, Kerrigan DC, Saliba EN, Gansneder BM, Ingersoll CD. Reduced quadriceps activation after lumbar paraspinal fatiguing exercise. Journal of Athletic Training 2006b;41:74–8. Hart JM, Kerrigan DC, Fritz JM, Saliba EN, Gansneder B, Ingersoll CD. Contribution of hamstring fatigue to quadriceps inhibition following lumbar extension exercise. Journal of Sports Science & Medicine 2006c;5:70–9. Herzog W, Scheele D, Conway J. Electromyographic responses of back and limb muscles associated with spinal manipulative therapy. Spine 1999;24:146–53. Hillermann B, Gomes AN, Korporaal C, Jackson D. A pilot study comparing the effects of spinal manipulative therapy with those of extra-spinal manipulative therapy on quadriceps muscle strength. Journal of Manipulative and Physiological Therapeutics 2006;29:145–9. Hopkins JT, Ingersoll CD, Krause BA, Edwards JE, Cordova ML. Effect of knee joint effusion on quadriceps and soleus motoneuron pool excitability. Medicine & Science in Sports & Exercise 2001;33:123–6. Iverson CA, Sutlive TG, Crowell MS, Morrell RL, Perkins MW, Garber MB, et al. Lumbopelvic manipulation for the treatment of patients with patellofemoral pain syndrome: Development of a clinical prediction rule. Journal of Orthopaedic and Sports Physical Therapy 2008;33(6):297–309. Kent-Braun JA, Le Blanc R. Quantitation of central activation failure during maximal voluntary contractions in humans. Muscle & Nerve 1996;19:861–9. Lewek M, Rudolph KS, Axe MJ, Snyder-Mackler L. The effect of insufficient quad strength on gait after anterior cruciate ligament reconstruction. Clinical Biomechanics 2002;17:56–63. Lewek MD, Rudolph KS, Snyder-Mackler L. Quadriceps femoris muscle weakness and activation failure in patients with symptomatic knee osteoarthritis. Journal of Orthopaedic Research 2004;22:110–5.

420

T.L. Grindstaff et al. / Manual Therapy 14 (2009) 415–420

Maitland ME, Ajemian SV, Suter E. Quadriceps femoris and hamstring muscle function in a person with an unstable knee. Physical Therapy 1999;79:66–75. Manal TJ, Snyder-Mackler L. Failure of voluntary activation of the quadriceps femoris muscle after patellar contusion. Journal of Orthopaedic and Sports Physical Therapy 2000;30:655–60. Mizner RL, Petterson SC, Stevens JE, Vandenborne K, Snyder-Mackler L. Early quadriceps strength loss after total knee arthroplasty. The Contributions of muscle atrophy and failure of voluntary muscle activation. Journal of Bone and Joint Surgery American 2005;87:1047–53. Mizner RL, Stevens JE, Snyder-Mackler L. Voluntary activation and decreased force production of the quadriceps femoris muscle after total knee arthroplasty. Physical Therapy 2003;83:359–65. Moore KL, Dalley AF. Clinically oriented anatomy. 4th ed. Philadelphia: Lippincott Williams & Wilkins; 1999. Murphy BA, Dawson NJ, Slack JR. Sacroiliac joint manipulation decreases the H-reflex. Electromyography and Clinical Neurophysiology 1995;35:87–94. Palmieri RM, Ingersoll CD, Edwards JE, Hoffman MA, Stone MB, Babington JP, et al. Arthrogenic muscle inhibition is not present in the limb contralateral to a simulated knee joint effusion. American Journal of Physical Medicine & Rehabilitation 2003;82:910–6. Palmieri RM, Tom JA, Edwards JE, Weltman A, Saliba EN, Mistry DJ, et al. Arthrogenic muscle response induced by an experimental knee joint effusion is mediated by pre- and post-synaptic spinal mechanisms. Journal of Electromyography and Kinesiology 2004;14:631–40. Pickar JG. Neurophysiological effects of spinal manipulation. The Spine Journal 2002;2:357–71. Rutherford OM, Jones DA, Newham DJ. Clinical and experimental application of the percutaneous twitch superimposition technique for the study of human muscle activation. Journal of Neurology. Neurosurgery & Psychiatry 1986;49:1288–91. Snyder-Mackler L, Binder-Macleod SA, Williams PR. Fatigability of human quadriceps femoris muscle following anterior cruciate ligament reconstruction. Medicine & Science in Sports & Exercise 1993;25:783–9. Snyder-Mackler L, De Luca PF, Williams PR, Eastlack ME, Bartolozzi ARI. Reflex inhibition of the quadriceps femoris muscle after injury or reconstruction of the anterior cruciate ligament. Journal of Bone and Joint Surgery American 1994;76:555–60. Stackhouse SK, Dean JC, Lee SCK, Binder-MacLeod SA. Measurement of central activation failure of the quadriceps femoris in healthy adults. Muscle & Nerve 2000;23:1706–12.

Stackhouse SK, Stevens JE, Lee SC, Pearce KM, Snyder-Mackler L, Binder-Macleod SA. Maximum voluntary activation in nonfatigued and fatigued muscle of young and elderly individuals. Physical Therapy 2001;81:1102–9. Stevens JE, Binder-Macleod S, Snyder-Mackler L. Characterization of the human quadriceps muscle in active elders. Archives of Physical Medicine and Rehabilitation 2001;82:973–8. Stevens JE, Mizner RL, Snyder-Mackler L. Quadriceps strength and volitional activation before and after total knee arthroplasty for osteoarthritis. Journal of Orthopaedic Research 2003;21:775–9. Stratford P. Electromyography of the quadriceps femoris muscles in subjects with normal knees and acutely effused knees. Physical Therapy 1982;62:279–83. Sung PS, Kang YM, Pickar JG. Effect of spinal manipulation duration on low threshold mechanoreceptors in lumbar paraspinal muscles: a preliminary report. Spine 2004;30:115–22. Suter E, Herzog W, Conway PJ, Zhang YT. Reflex response associated with manipulative treatment of the thoracic spine. Journal of the Neuromusculoskeletal System 1994;2:124–30. Suter E, McMorland G, Herzog W, Bray R. Decrease in quadriceps inhibition after sacroiliac joint manipulation in patients with anterior knee pain. Journal of Manipulative and Physiological Therapeutics 1999;22:149–53. Suter E, McMorland G, Herzog W, Bray R. Conservative lower back treatment reduces inhibition in knee-extensor muscles: a randomized controlled trial. Journal of Manipulative and Physiological Therapeutics 2000;23:76–80. Urbach D, Nebelung W, Becker R, Awiszus F. Effects of reconstruction of the anterior cruciate ligament on voluntary activation of quadriceps femoris a prospective twitch interpolation study. Journal of Bone and Joint Surgery British 2001;83:1104–10. Urbach D, Nebelung W, Weiler HT, Awiszus F. Bilateral deficit of voluntary quadriceps muscle activation after unilateral ACL tear. Medicine & Science in Sports & Exercise 1999;31:1691–6. Williams GN, Barrance PJ, Snyder-Mackler L, Axe MJ, Buchanan TS. Specificity of muscle action after anterior cruciate ligament injury. Journal of Orthopaedic Research 2003;21:1131–7. Williams GN, Buchanan TS, Barrance PJ, Axe MJ, Snyder-Mackler L. Quadriceps weakness, atrophy, and activation failure in predicted noncopers after anterior cruciate ligament injury. American Journal of Sports Medicine 2005;33:402–7. Yerys S, Makofsky H, Byrd C, Pennachio J, Conkay J. Effect of mobilization of the anterior hip capsule on gluteus maximus strength. Journal of Manual & Manipulative Therapy 2002;10:218–24.

Manual Therapy 14 (2009) 421–426

Contents lists available at ScienceDirect

Manual Therapy journal homepage: www.elsevier.com/math

Original Article

Validity and reliability of ultrasonography for the longus colli in asymptomatic subjectsq Barbara Cagnie a, *, Erwin Derese a, Leen Vandamme a, Koenraad Verstraete b, Dirk Cambier a, Lieven Danneels a a b

Ghent University, Department of Rehabilitation Sciences and Physiotherapy, De Pintelaan 185, 3B3, 9000 Ghent, Belgium Ghent University, Department of Radiology, Ghent University Hospital, De Pintelaan 185, 1K12, 9000 Ghent, Belgium

a r t i c l e i n f o

a b s t r a c t

Article history: Received 11 September 2007 Received in revised form 9 July 2008 Accepted 28 July 2008

The purposes of this study were to evaluate the reliability and validity of ultrasound (US) for measuring the cross-sectional area (CSA) of the longus colli (LC) as compared with magnetic resonance imaging (MRI), and to determine the change in CSA of the LC during contraction. 27 healthy volunteers participated in the study. In order to assess the validity of US, the US measurements of the CSA of the LC were compared to those determined with MRI. Two testers established the measurements to ascertain intra- and interrater reliability. The widely spaced limits of agreement (2SD ¼ 0.45) reflect the large variability between the measurements by US and MRI. The ICC for the intra- and interrater reliability for the CSA of the LC was respectively 0.71 (95% CI, 0.57–0.81; SEM, 0.17; SDD, 0.48) and 0.68 (95% CI, 0.48–0.81; SEM, 0.18; SDD, 0.50). The CSA of the LC increased significantly during contraction of the LC (p ¼ 0.006). Results from this study show that the validity and reliability of US to evaluate the CSA of the LC is questionable, which may be due to both anatomical characteristics and methodological limitations. Ó 2008 Elsevier Ltd. All rights reserved.

Keywords: Ultrasonography Longus colli Reliability Validity

1. Introduction In recent years evidence has accumulated of impairment in deep cervical flexor muscle (DCF) function in neck pain sufferers (Falla et al., 2004a; Falla et al., 2004b; Falla, 2004; Jull et al., 2004; O’leary et al., 2007b). Evidence suggests that these muscles are important for the control and support of the cervical lordosis and maintenance of cervical spine postural form (Mayoux-Benhamou et al., 1994; Conley et al., 1995; Vasavada et al., 1998; Boyd-Clark et al., 2001; Boyd-Clark et al., 2002). Furthermore, specific therapeutic retraining of the DCF muscles has demonstrated efficacy in the management of patients with chronic neck pain and cervicogenic headache (Jull et al., 2002; O’leary et al., 2003; Falla et al., 2006a). Unfortunately, because these muscles are deeply situated, traditional methods such as palpation and manual muscle testing are unreliable for assessment of their function. Secondly, it is difficult to reach the DCF with surface electromyography (EMG). Nevertheless, Falla et al. (2003) described a novel surface EMG technique for the detection of DCF muscle activity. However, this technique is not applicable for

q This study was supported by the Research Foundation – Flanders (FWO). * Corresponding author. Tel.: þ32 9 332 52 65; fax: þ32 9 332 38 11. E-mail address: [email protected] (B. Cagnie). 1356-689X/$ – see front matter Ó 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.math.2008.07.007

routine assessment of the DCF in clinical practice (Falla et al., 2006b; O’leary et al., 2007a). Quantitative measurements of paraspinal muscle size can be obtained with both real-time ultrasonography (US) and magnetic resonance imaging (MRI) and there is growing support for their use in investigations of patients with spinal pain (Stokes et al., 2007; Whittaker et al., 2007; Elliott et al., 2008). MRI can be regarded as the gold standard for muscle imaging. Real-time US, however, has the advantages of widespread accessibility and lower cost. The muscles are visualized in real-time and measurements can be obtained in a relaxed state and in different states of contraction as well as during movements (Rezasoltani et al., 2002; Kiesel et al., 2007). The disadvantages of real-time US are its relatively limited field of view and its inability to provide pilot sections for confirmation of vertebral levels when the spine is imaged. This means that strict protocols must be followed to allow accurate measurement of the soft tissues. Today, US is frequently used as a diagnostic tool and as a biofeedback method for the muscles in the lumbo-pelvic region (Van et al., 2006). However, the use of US for the evaluation of the cervical muscles is sparse. Previous researchers have used US to evaluate the involvement of the dorsal neck muscles in chronic pain, while, to the best of our knowledge, this has not been used for the evaluation for the cervical flexor muscles (Rezasoltani et al., 1998; Rezasoltani et al., 1999; Kristjansson, 2004; Rankin et al., 2005).

422

B. Cagnie et al. / Manual Therapy 14 (2009) 421–426

The purposes of this study were 1) to set up a clear standardized protocol for the evaluation of the longus colli (LC) with US, 2) to evaluate the validity of US as compared with MRI (CSA) of the LC during rest 3) to determine the reliability of US for measuring the cross-sectional area and 4) to determine the change in CSA of the LC during contraction. 2. Materials and methods 2.1. Subjects US images were obtained in 27 healthy subjects (14 men and 13 women). Demographic details are shown in Table 1. Subjects were either sedentary or moderately active (0–10 h of sports a week). Exclusion criteria were pronounced neck or back trauma, neurological or inflammatory disorders, dizziness, vestibular symptoms or diabetes. The subjects should not have had neck pain, back pain or headache from cervical origin in 6 weeks before the scanning and at the moment of the scanning. MRI images were obtained in 18 of the 27 volunteers. Exclusion criteria for MRI were a cardiac pacemaker, claustrophobia, implanted metals, unremovable piercings, aneurysm clips, carotid artery vascular clamp, neurostimulator, cochlear or ear implants, and (possible) pregnancy within the first 3 months. The project was approved by the local ethics committees. Written informed consent was obtained from all subjects. 2.2. Study design The CSA of the LC was measured using real-time US and MRI. 2.2.1. Validity study In order to assess the validity of US, one examiner compared the US measurements of the CSA of the LC to those determined with MRI, both at the C5-C6 level. 2.2.2. Reliability study One tester established the measurements on three different occasions to ascertain intrarater reliability of US at rest. The time interval between the tests was one week. To determine the interrater reliability, each subject was imaged and measured by a second tester on the first test day. Each rater independently established anatomical landmarks and subject positioning. One single measurement on each side was taken. In order to determine the interrater reliability of the CSA measurement of the LC on MRI, both testers evaluated independently the same MRI images. In order to interpret the values obtained, the reliability of the CSA of the sternocleidomastoid (SCM) was also determined.

range of motion while trying to keep the SCM relaxed and without lifting their head of the surface (Falla et al., 2004b). A pressure cuff was used (Chattanooga Group Inc.), which was placed suboccipitally behind the subject’s cervical spine and inflated until a stable pressure of 20 mmHg was achieved. Prior to the test day, subjects were instructed in the movement and practiced targeting their maximal pressure without excessive action of the SCM. The pressure level that the subject could achieve and hold in a steady manner for 10  10 s with the SCM relaxed and without lifting the head off the surface was determined. 2.3. Ultrasound imaging technique The US measurements were performed by two experienced manual therapists using a real-time US apparatus (MyLab 30 CV - Esaote) with a 12 MHz linear array transducer (type LA 532). Both raters were trained by an experienced ultrasonographer which included instruction in the measurement technique, visualization of the LC and processing of the data. The measurement protocol using US was designed through several pilot trials and based on a fundamental knowledge of cervical anatomy and US. The subjects were supine in a neutral position, with their knees and hips bent, and the arms lying along the sides of the body. The subjects were first measured at rest, followed by a measurement during contraction. At rest, the cervical lordosis was supported by a folded towel. LC was scanned at C5-C6 level. This level was selected because of its clear visualization. Secondly, at this level, there is no overlap between LC and longus capitis muscle, as this muscle runs more laterally and attaches to the anterior tubercle of C6. Axial images were obtained by placing the middle of the probe perpendicular to the long axis of the anterior neck (Fig. 1) The bottom of the laryngeal prominence of the thyroid cartilage, which corresponds with the C5 level, served as a reference point for the probe, to assure that all measurements took place at the same level (Putz & Pabst, 1994; Moeller & Reif, 2007). The transducer was then moved approximately 1 cm laterally from the bottom of the laryngeal prominence to each side to image the left and right LC. Images were captured, stored and measured afterwards with the device-linked program MyLabDesk. During contraction, participants were asked to accurately maintain the predetermined level of contraction for at least 10 s while US measurements of the LC were made. The probe was positioned at the same place as during rest. The CSA was measured by using on-screen callipers. The LC was localized on an average depth of 2–3 cm. The outlines of the LC were identified by the following landmarks: anterolaterally by the

2.2.3. Rest versus contraction In addition to the measurement at rest, the CSA of the LC was also measured during contraction on the third test day by one examiner. To isolate the contraction, subjects were asked to perform a gentle nodding action to reach full cranio-cervical flexion

Table 1 Anthropometric variables. N ¼ 27 Age (year) Length (cm) Weight (kg) BMI (kg/m2) Sports (h/week) Neck perimeter (cm)

22.4  1.2 174.3  7.7 66.2  8.1 21.7  2.1 4.5  3.1 35.0  2.8

Fig. 1. Position of the transducer for imaging of the longus colli.

B. Cagnie et al. / Manual Therapy 14 (2009) 421–426

423

common carotid artery and internal jugular vein; anteromedially by the thyroid gland and posteriorly by the echogenic vertebra (Fig. 2). 2.4. MRI MRI was performed on a 3 Tesla magnet (Siemens Magnetom ‘Trio a Tim System’ with Syngo MR B13). A flexible surface coil, 20  50 cm, fixed over the anterior aspect of the participant’s neck was combined with the phased-array spine coil as a receiver coil combination. The subjects were placed in a comfortable and relaxed supine position, with their hips flexed to 45 and legs supported by foam wedges. The head was positioned in a neutral position, without rotation, lateral flexion or exaggerated lordosis. Compared to US, no towel was added as this was not possible due to the used coils. T2-weighted images of the cervical spine were obtained from the C0 to the C6 segmental level using the following imaging sequences: 20 slices with a slice thickness of 3 mm; Field of view read: 200 mm; relaxation time: 5130 ms; Echo time: 103 ms; Flip angle: 160 ; Acquisition time: 3:22 s. (Fig. 3). The measurements at the C5-C6 level were used to compare with the measurement results of the US. The location of the C5-C6 level was determined from parallel images in the mid-sagittal plane by identifying the spinous process of C5 and C6 and the C5–C6 intervertebral disc. Axial images were positioned parallel to the C2–C3 intervertebral disk, which may produce a slight measurement error for the CSA measures at the C5–C6 level due to the cervical lordosis. As a result, relative CSAs (rCSA) are reported. Muscle CSA was measured using on-screen callipers. The rCSA of LC was calculated by the number of pixels under each ROI in the x and y axes with DicomWorks. 2.5. Statistical analysis Data were analyzed with SPSS 15.0. All data are presented as mean  standard deviation (SD). Because there were no significant differences between left and right side, the values of both sides were averaged for further analyses. For analysis of validity, differences between the measurements by the two methods for each subject were plotted against their mean (Bland & Altman, 1999). The 95% limits of agreement were

Fig. 3. MRI image on the C6-level. 1:longus colli, 2: sternocleidomastoid, 3: trachea, 4: vertebral body C6, 5: multifidus, 6: trapezius muscle, 7: common carotid artery, 8: internal jugular vein.

calculated, to indicate the extent the two methods are likely to differ due to observer error when used in practice. For analysis of intra- and intertester reliability, intra-class correlation coefficients [ICC1,1 and ICC2,1], standard error of measurement (SEM) and smallest detectable difference (SDD) were used. Defined with respect to a 95% level of confidence, the SDD is equal to 1.96 O2*SEM. The changes in CSA of the LC between rest and contraction were evaluated by a paired t-test with a 95% CI for the mean difference. 3. Results 3.1. Validity The within-subject differences between the MRI and US values plotted against the mean of measurements with 95% limits of agreement are seen in Fig. 4. The widely spaced limits of agreement (2SD ¼ 0.45) reflect the variability between the two methods. The average (SD) CSA of the LC on MRI (1.25 cm2  0.28) was larger than the CSA on US (1.22 cm2  0.37), although not significant (p ¼ 0.309). 3.2. Reliability The ICC for the intra- and interrater reliability for the CSA of the LC was respectively 0.71 (95% CI, 0.57–0.81) and 0.68 (95% CI, 0.48–0.81) (Table 2). The interrater measurements of LC on MRI showed good reliability (ICC ¼ 0.81; 95% CI, 0.64–0.90), whereas the interrater measurements of SCM showed excellent reliability (ICC ¼ 0.96; 95% CI, 0.87–0.98) (Table 3). 3.3. Rest versus contraction

Fig. 2. Ultrasound image of the longus colli at the C5/C6 level. 1: acoustic shadow of the trachea, 2: internal jugular vein, 3: sternocleidomastoid muscle, 4: common carotid artery, 5: longus colli, 6: thyroid gland.

The CSA of the LC during contraction (1.35 cm2  0.32) was significantly higher compared to rest (1.22 cm2  0.37) (p ¼ 0.006), with a 95% CI for the difference ranging from 0.04 to 0.20 (Table 4). This corresponds with an enhancement of 12%.

424

B. Cagnie et al. / Manual Therapy 14 (2009) 421–426 Table 3 Inter rater reliability of the CSA measurement of longus colli (LC) and sternocleidomastoid (SCM) on MRI.

0,60 +2SD

LC SCM

Difference US and MRI (cm2)

0,40

0,20

0,00

-0,20

-0,40 -2SD -0,60 0,00

0,50

1,00

1,50

2,00

2,50

Average US and MRI (cm2) Fig. 4. Agreement between US and MRI recordings plotted as mean values against differences between the two methods. The solid horizontal line illustrates the mean difference value with  2SD (dotted lines).

4. Discussion The present study is the first, to our knowledge, to document the evaluation of CSA of the longus colli with use of US. Results from this study show that the validity and reliability of US to evaluate the CSA of the LC is doubtful, which may be due to both the anatomic structure of the muscle and some methodological limitations. 4.1. Validity Compared with MRI, US seems to have a questionable validity. In previous comparable validity studies, Hides et al. (1995) did not found a difference between US and MRI measurements of the CSA of the lumbar multifidus, whereas Lee et al. found no acceptable validity for measuring the area of the cervical multifidus muscle (Hides et al., 1995; Lee et al., 2007). In this study, the widely spaced limits of agreement (2SD ¼ 44.58) reflects a large variability between the two methods. These differences may be explained by the size and anatomic structure of this muscle. The small CSA of the LC (about 1.20 cm2 compared with 2–7 cm2 for lumbar multifidus depending on the level) may amplify errors, thus influence the variability of measurements (Lee et al., 2007; Stokes et al., 2007). Secondly, the boundaries are not always distinct, which may influence the accuracy of the US measurement. This is especially true for the posteromedial boundary of the LC which is difficult to outline on US, due to the acoustic shadow of the trachea, which may explain the large variability. It should have been useful to watch the muscle contracting before taking Table 2 Inter- and intrarater reliability of the CSA measurement of longus colli (LC) on US.

Intra Inter

ICC (95% CI)

SEM

SDD

0.71 (0.57–0.81) 0.68 (0.48–0.81)

0.17 0.18

0.48 0.50

ICC (95% CI)

SEM

SDD

0.81 (0.64–0.90) 0.96 (0.87–0.98)

0.14 0.27

0.39 0.76

measurements at rest, as this technique has been found helpful for identifying the borders of multifidus more clearly (Stokes et al., 2005). It should also be noted that MRI as being the gold standard can be questioned for smaller muscles. MRI seems a very reliable method for large muscles, but it is less obvious to obtain reliable results for smaller and deeper muscles. This was also demonstrated in our study, in which the reliability of the LC was much lower on MRI than the reliability of the SCM. This fact should be kept in mind when interpreting the results of the validity study. The validity could have been affected by some methodology flaws. The positioning of the patient may have influenced the validity. In both the US and MRI study, patients were placed in a supine position, with their hips and knees bent. The head was positioned in a neutral position, without rotation, lateral flexion or exaggerated lordosis. However, in the US study, the cervical lordosis was supported by a folded towel, whereas this was not the case in the MRI scanner as this was not possible due to the fixed coils. Minimal differences in cervical lordosis probably may influence the results. Although not significant, the average CSA of the LC on MRI was larger than the CSA on US. This is in accordance with the study of Lee et al. who found that the values obtained by MRI are on average 2–3 mm2 larger than those obtained by US (Lee et al., 2007). This can be explained by the different scanning planes of US and MRI. In the cervical region, the lordosis may play a role in the measurement of muscle thickness. The T2-weighted MRI images were taken perpendicular to the C2–C3 vertebral disk, while the US images are supposed to be taken perpendicular to the spine and thus the LC. This gives an overestimation of the CSA on MRI which may explain why MRI shows a higher value for the CSA in comparison with US (Lee et al., 2007). In order to reduce measurement errors and thus increase validity, it would have been more accurate to take an image parallel to the C5–C6 intervertebral disk instead of the C2–C3 level. In addition, as the bottom of the laryngeal prominence of the thyroid cartilage was taken as a reference point for C5–C6, it is possible that the MRI and US levels were not exactly the same, which may also could have influenced the validity. Measurements were taken by two trained manual therapists. Although trained in this technique, it is possible that the fact that they were no experienced radiologists may have affected validity. 4.2. Reliability The reliability of the protocol was moderate (ICC: 0.68–0.71) (Shrout, 1998). Kiesel et al. (2007), Lee et al. (2007) and Pressler et al. (2006) demonstrated quite similar results as our study (Pressler et al., 2006; Kiesel et al., 2007; Lee et al., 2007). In contrast, Rankin et al. (2005), Stokes et al. (2005) and Van et al. (2006), did however found quite high ICC values (varying between 0.97 and 1.00) for the evaluation of the dorsal neck muscles and the lumbar multifidus (Rankin et al., 2005; Stokes

Table 4 CSA of the Longus colli (LC) (cm2) at rest and during contraction. LC rest

LC contraction

Mean difference (95% CI)

p-Value

1.22  0.27

1.35  0.32

0.04–0.20

0.006

B. Cagnie et al. / Manual Therapy 14 (2009) 421–426

et al., 2005; Van et al., 2006). The fact that in these studies mean measurements rather than single measurements were used, should be borne in mind. In addition to the findings of moderate intrarater and interrater reliability, the reported SEM and SDD values are quite high when compared to the resting CSA of the muscle. SEM and SDD values ranged from 0.17 to 018 and 0.48 to 0.50 respectively. The SDD values indicate that the CSA would change by at least 0.48–0.50 cm, to be 95% confident that true change occurred, which would require a 40% change. This is rather high compared to previous studies (Pressler et al., 2006; Van et al., 2006; Wallwork et al., 2007). These high SEM and SDD values will limit the ability to detect a change in muscle CSA with US in longitudinal studies. The fact that only one measurement was taken, could partly explain the moderate reliability. It is likely that a procedure which utilises more than one measurement would increase the SEM and SDD values. In order to reduce measurement error, averaged values of repeated trials should be considered to determine the CSA. Springer et al. demonstrated that by taking an average of three measures of the transversus abodminis muscle with US, the SEM reduced by approximately 50% (Springer et al., 2006). 4.3. Rest versus contraction During contraction, a significant increase in muscle CSA was found. However, the questionable validity and reliability should be kept in mind when interpreting the results. Although significant, the amount of increase is not comparable to previous studies, in which the lumbar and cervical multifidus were investigated during contraction (McMeeken et al., 2004; Kiesel et al., 2007; Lee et al., 2007). In these studies, an increase of up to 60% between rest and contraction was noted, compared to 12% in our study. One of the hypotheses of having a lower increase may be the function of the LC. In order to isolate the contraction, we asked the subjects to perform a cranio-cervical flexion. Previous studies have indicated that this method is specific to the anatomical action of the DCF, which encompasses the LC and longus capitis. In the studies of Conley et al. (1995), Falla et al. (2003) and O’leary et al. (2006), no difference is made between both muscles (Conley et al., 1995; Falla et al., 2003; O’leary et al., 2007b). However, a recent study has indicated the need to differentiate between the LC and longus capitis, since there is a clear difference in activation of both muscles (Cagnie et al., 2008). The longus capitis seems to play a more important role in the performance of the cranio-cervical flexion compared to the LC. This could be due to the fact that the primary anatomical action of Lca is flexion of the cranio-cervical junction, whereas the primary action of LC is flattening of the cervical lordosis (Falla et al., 2006b). It may be possible that greater changes in CSA would occur if 1) a different task was used to contract the parts of longus colli that were imaged or 2) the LC was measured at a higher segmental level. 4.4. Clinical implications As mentioned in the introduction section, a key advantage of US imaging is the ability to visualise a muscle contracting. If further testing suggests that measurement of changes in CSA is not reliable, the use of imaging as a biofeedback tool should still be considered. As it is not possible to palpate the longus colli, neither to capture this muscle with surface EMG, US may provide visual feedback in order to enhance motor learning for contracting this muscle. Evidence that feedback with US enhances motor learning holds promise for the future but studies are needed on subjects with neck pain.

425

5. Conclusion Results from this study show that the validity and reliability of US to evaluate the CSA of the LC is questionable, which may be due to both anatomical characteristics and methodological limitations. The small size of the muscle and difficulties in outlining the boundaries of the muscle may explain the large variability. Taking an average of different measures should be considered in further research. If further testing suggests that measurement of changes in CSA is not reliable, the use of imaging as a biofeedback tool should still be considered. Acknowledgements We would like to thank Dr. H. Vinck for his useful help during the set up of the ultrasound protocol. Reference Bland J, Altman D. Measuring agreement in method comparison studies. Stat Methods Res 1999;8:135–60. Boyd-Clark LC, Briggs CA, Galea MP. Comparative histochemical composition of muscle fibres in a pre- and a postvertebral muscle of the cervical spine. J Anat 2001;199:709–16. Boyd-Clark LC, Briggs CA, Galea MP. Muscle spindle distribution, morphology, and density in longus colli and multifidus muscles of the cervical spine. Spine 2002;27:694–701. Cagnie B, Dickx N, Peeters I, Tuytens J, Achten E, Cambier D, et al. The use of functional MRI to evaluate cervical flexor activity during different cervical flexion exercises. J Appl Physiol 2008;104:230–5. Conley MS, Meyer RA, Bloomberg JJ, Feeback DL, Dudley GA. Noninvasive analysis of human neck muscle function. Spine 1995;20:2505–12. Elliott J, Jull G, Noteboom JT, Galloway G. MRI study of the cross-sectional area for the cervical extensor musculature in patients with persistent whiplash associated disorders (WAD). Man Ther 2008;13:258–65. Falla D. Unravelling the complexity of muscle impairment in chronic neck pain. Man Ther 2004;9:125–33. Falla D, Jull G, Dall’Alba P, Rainoldi A, Merletti R. An electromyographic analysis of the deep cervical flexor muscles in performance of craniocervical flexion. Phys Ther 2003;83:899–906. Falla D, Jull G, Hodges P, Vicenzino B. An endurance-strength training regime is effective in reducing myoelectric manifestations of cervical flexor muscle fatigue in females with chronic neck pain. Clin Neurophysiol 2006a;117: 828–37. Falla D, Jull G, O’leary S, Dall’Alba P. Further evaluation of an EMG technique for assessment of the deep cervical flexor muscles. J Electromyogr Kinesiol 2006b;16:621–8. Falla D, Jull G, Rainoldi A, Merletti R. Neck flexor muscle fatigue is side specific in patients with unilateral neck pain. Eur J Pain 2004a;8:71–7. Falla DL, Jull GA, Hodges PW. Patients with neck pain demonstrate reduced electromyographic activity of the deep cervical flexor muscles during performance of the craniocervical flexion test. Spine 2004b;29:2108–14. Hides JA, Richardson CA, Jull GA. Magnetic resonance imaging and ultrasonography of the lumbar multifidus muscle. Comparison of two different modalities. Spine 1995;20:54–8. Jull G, Kristjansson E, Dall’Alba P. Impairment in the cervical flexors: a comparison of whiplash and insidious onset neck pain patients. Man Ther 2004;9:89–94. Jull G, Trott P, Potter H, Zito G, Niere K, Shirley D, et al. A randomized controlled trial of exercise and manipulative therapy for cervicogenic headache. Spine 2002;27:1835–43. Kiesel KB, Uhl TL, Underwood FB, Rodd DW, Nitz AJ. Measurement of lumbar multifidus muscle contraction with rehabilitative ultrasound imaging. Man Ther 2007;12:161–6. Kristjansson E. Reliability of ultrasonography for the cervical multifidus muscle in asymptomatic and symptomatic subjects. Man Ther 2004;9:83–8. Lee JP, Tseng WY, Shau YW, Wang CL, Wang HK, Wang SF. Measurement of segmental cervical multifidus contraction by ultrasonography in asymptomatic adults. Man Ther 2007;12:286–94. Mayoux-Benhamou MA, Revel M, Vallqe C, Roudier R, Barbet JP, Bargy F. Longus colli has a postural function on cervical curvature. Surg Radiol Anat 1994;16: 367–71. McMeeken JM, Beith ID, Newham DJ, Milligan P, Critchley DJ. The relationship between EMG and change in thickness of transversus abdominis. Clin Biomech (Bristol, Avon) 2004;19:337–42. Moeller T, Reif E. Pocket atlas of sectional anatomy. Georg Thieme Verlag 2007. O’leary S, Falla D, Jull G. Recent advances in therapeutic exercise for the neck: implications for patients with head and neck pain. Aust Endod J 2003;29: 138–42. O’leary S, Falla D, Jull G, Vicenzino B. Muscle specificity in tests of cervical flexor muscle performance. J Electromyogr Kinesiol 2007a;17:35–40.

426

B. Cagnie et al. / Manual Therapy 14 (2009) 421–426

O’leary S, Jull G, Kim M, Vicenzino B. Cranio-cervical flexor muscle impairment at maximal, moderate, and low loads is a feature of neck pain. Man Ther 2007b;12:34–9. Pressler JF, Heiss DG, Buford JA, Chidley JV. Between-day repeatability and symmetry of multifidus cross-sectional area measured using ultrasound imaging. J Orthop Sports Phys Ther 2006;36:10–8. Putz R, Pabst R. Sobotta Atlas van de menselijke anatomie. Houten: Bohn Stafleu Van Loghum; 1994. Rankin G, Stokes M, Newham DJ. Size and shape of the posterior neck muscles measured by ultrasound imaging: normal values in males and females of different ages. Man Ther 2005;10:108–15. Rezasoltani A, Kallinen M, Malkia E, Vihko V. Neck semispinalis capitis muscle size in sitting and prone positions measured by real-time ultrasonography. Clin Rehabil 1998;12:36–44. Rezasoltani A, Malkia E, Vihko V. Neck muscle ultrasonography of male weightlifters, wrestlers and controls. Scand J Med Sci Sports 1999;9:214–8. Rezasoltani A, Ylinen J, Vihko V. Isometric cervical extension force and dimensions of semispinalis capitis muscle. J Rehabil Res Dev 2002;39:423–8. Shrout P. Measurement of reliability and agreement in psychiatry. Statisical methods in medical research. Stat Methods Med Res 1998;7:301–17.

Springer BA, Mielcarek BJ, Nesfield TK, Teyhen DS. Relationships among lateral abdominal muscles, gender, body mass index, and hand dominance. J Orthop Sports Phys Ther 2006;36:289–97. Stokes M, Hides J, Elliott J, Kiesel K, Hodges P. Rehabilitative ultrasound imaging of the posterior paraspinal muscles. J Orthop Sports Phys Ther 2007;37:581–95. Stokes M, Rankin G, Newham DJ. Ultrasound imaging of lumbar multifidus muscle: normal reference ranges for measurements and practical guidance on the technique. Man Ther 2005;10:116–26. Van K, Hides JA, Richardson CA. The use of real-time ultrasound imaging for biofeedback of lumbar multifidus muscle contraction in healthy subjects. J Orthop Sports Phys Ther 2006;36:920–5. Vasavada AN, Li S, Delp SL. Influence of muscle morphometry and moment arms on the moment-generating capacity of human neck muscles. Spine 1998;23: 412–22. Wallwork TL, Hides JA, Stanton WR. Intrarater and interrater reliability of assessment of lumbar multifidus muscle thickness using rehabilitative ultrasound imaging. J Orthop Sports Phys Ther 2007;37:608–12. Whittaker JL, Teyhen DS, Elliott JM, Cook K, Langevin HM, Dahl HH, et al. Rehabilitative ultrasound imaging: understanding the technology and its applications. J Orthop Sports Phys Ther 2007;37:434–49.

Manual Therapy 14 (2009) 427–432

Contents lists available at ScienceDirect

Manual Therapy journal homepage: www.elsevier.com/math

Original Article

Evaluation of stretching position by measurement of strain on the ilio-femoral ligaments: An in vitro simulation using trans-lumbar cadaver specimens Egi Hidaka a, Mitsuhiro Aoki b, *, Takayuki Muraki a, Tomoki Izumi a, Misaki Fujii a, Shigenori Miyamoto b a b

Graduate School of Health Sciences, Sapporo Medical University, Sapporo, Japan Department of Physical Therapy, Sapporo Medical University School of Health Sciences, Sapporo, Japan

a r t i c l e i n f o

a b s t r a c t

Article history: Received 12 March 2008 Received in revised form 2 July 2008 Accepted 28 July 2008

The ilio-femoral ligament is known to cause flexion contracture of the hip joint. Stretching positioning is intended to elongate the ilio-femoral ligaments, however, no quantitative analysis to measure the effect of stretching positions on the ligament has yet been performed. Strains on the superior and inferior iliofemoral ligaments in 8 fresh/frozen trans-lumbar cadaveric hip joints were measured using a displacement sensor, and the range of movement of the hip joints was recorded using a 3Space Magnetic Sensor. Reference length (L0) for each ligament was determined to measure strain on the ligaments. Hip positions at 10 adduction with maximal external rotation, 20 adduction with maximal external rotation, and maximal external rotation showed larger strain for the superior ilio-femoral ligament than the value obtained from L0, and hip positions at 20 external rotation with maximal extension and maximal extension had larger strain for the inferior ilio-femoral ligament than the value obtained from L0 (p < 0.05). Superior and inferior ilio-femoral ligaments exhibited positive strain values with specific stretching positions. Selective stretching for the ilio-femoral ligaments may contribute to achieve lengthening of the ligaments to treat flexion contracture of the hip joint. Ó 2008 Elsevier Ltd. All rights reserved.

Keywords: Fresh cadaver Hip Ligaments Biomechanics

1. Introduction The ilio-femoral ligament is known to be the strongest ligament in human body and provides stability to the hip in the standing position (Hewitt et al., 2001; Neumann, 2002; Werner, 2004). The ilio-femoral ligament, which is located on the anterior aspect of the hip joint limits hyperextension of the hip and is known to be a cause of flexion contracture of the hip joint (O’Malley, 1959; Fuss and Bacher, 1991; French, 2007). For the treatment of flexion contracture associated with adult hip osteoarthritis, soft tissue release surgery, as advocated by O’Malley, has been performed (O’Malley, 1959). To prevent recurrence of flexion contracture of the hip, passive stretching of the released ilio-femoral ligament is performed postoperatively. On the other hand, stretching of the hip has been used for conditioning the limbs and trunk to improve flexibility of the lower extremities prior to strenuous physical activities, such as soccer, football, and handball (Mo¨ller et al., 1985; Zakas et al., 2003; Zakas et al., 2006). Thus, it is important for therapists and trainers to understand effective stretching positions for the ilio-femoral ligament.

* Corresponding author. Department of Physical Therapy, Sapporo Medical University School of Health Sciences, South-3, West-17, Chuo-ku, Sapporo 0608556, Japan. Tel.: þ81 11 611 2111; fax: þ81 11 611 2150. E-mail address: [email protected] (M. Aoki). 1356-689X/$ – see front matter Ó 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.math.2008.07.006

With regard to stretching positions for the ilio-femoral ligaments, Kapandji (1970) reported that the superior ilio-femoral ligament was stretched during hip external rotation and adduction, whereas the inferior ilio-femoral ligament was stretched during hip extension. Werner (2004) reported that the superior ilio-femoral ligament was stretched during hip external rotation, adduction and extension, whereas the inferior ilio-femoral ligament was stretched during hip internal rotation and extension. Lee (2004) reported that the superior ilio-femoral ligament was stretched during the combination of hip external rotation and adduction, with slight extension, whereas the inferior ilio-femoral ligament was stretched during hip extension. These stretching positions were determined from anatomical observations of the origin and insertion of the ligaments or kinesiological assumptions based on motion analysis. On the basis of a cadaveric study, Fuss and Bacher (1991) reported that the superior ilio-femoral ligament was stretched mainly during hip external rotation, whereas the inferior iliofemoral ligament was stretched mainly during hip extension. Although several authors have advocated various stretching positions, no consensus has been reached. Therefore, quantitative analysis to measure the effect of stretching on the ilio-femoral ligament is required to resolve this question. The purpose of this study was to measure strain on the superior and inferior ilio-femoral ligaments of the hip joints during passive stretching by using trans-lumbar cadaver specimens, and to

428

E. Hidaka et al. / Manual Therapy 14 (2009) 427–432

determine joint positions with positive ligament strain for the selective stretching of the ilio-femoral ligaments. 2. Methods 2.1. Preparation of specimens Eight fresh/frozen trans-lumbar specimens (four left hips and four right hips) with no evidence of osteoarthritis or fracture were used in this experiment. The age of the specimens at death was 67– 91 years (average, 80.3 years). Within 24 h after death, the specimens were transferred from their respective hospitals to the Department of Anatomy at the University and were kept in a freezer at 20  C. Frozen specimens were separated into 2 parts through the waist at the L1 level; i.e., into trans-thoracic specimen and trans-lumbar specimens. Thawing of the trans-lumbar specimens at room temperature began 24 h before preparation. After removal of the internal organs in a saline solution containing 0.1% sodium azide, the specimens were washed with clean saline. The skin, fascia, muscles, nerves and vessels were all removed leaving the lumbar spine, pelvic ring, hip joints, and both femurs. Ligaments of the hip joint were clearly exposed. To achieve the full range of passive hip joint motion, each translumbar specimen was secured by Kirschner wires and screws onto a thick wooden pole while maintaining the pelvis in an upright position with 30 of anterior tilt (Saito and Nagasaki, 2002). The anterior torsion angle of the femoral neck was set at 0 , and this position was regarded as the neutral hip position in this experiment. A plastic rod was inserted into the proximal femur in the vertical direction. The distal femur was then amputated (Fig. 1). According to the classification advocated by Kapandji (1970), the ilio-femoral ligament is divided into two parts. The ligament that originates from the antero-superior acetabular wall and inserts into the supero-lateral aspect of the inter-trochanteric line of the femur is designated as the superior ilio-femoral ligament, and the ligament that originates from anterior wall of the acetabulum and inserts into the infero-medial aspect of the inter-trochanteric line of the femur is designated as the inferior ilio-femoral ligament (Fig. 2). 2.2. Measurement devices Strain on the ilio-femoral ligament was measured using a displacement sensor. This device consisted of a coil sensor and

3SPACE Pulse Coder

Fig. 2. The superior and inferior ilio-femoral ligaments. (a) The superior ilio-femoral ligament originates from the antero-superior acetabular wall, extends parallel to the femoral neck axis, and inserts into the supero-lateral aspect of the inter-trochanteric line of the femur. (b) The inferior ilio-femoral ligament originates from the anterior wall of the acetabulum, extends parallel to the femoral shaft axis, and inserts into the infero-medial aspect of the inter-trochanteric line of the femur. The displacement sensor, which had 10-mm long points, was mounted on the center of the superior and inferior ilio-femoral ligament parallel to the ligament fibers.

a brass pipe. Changes in the distance between the points of the coil and the brass pipe enabled the measurement of the strain on the ligament. The range of measurement was 3–15 mm. The sensor, equipped with 10-mm long points, was mounted on the center of the superior and inferior ilio-femoral ligament parallel to the ligament fibers. Fishhook-like barb points prevented the sensors from slipping out of the ligament (Fig. 2). By using 3-power digital images projected on SCION Images, the accuracy of the preliminary calibration of the Pulse Coder, which was attached to the isolated iliofemoral ligament, was found to be 0.01 mm root mean square (RMS). A six-degree-of-freedom electromagnetic tracking device 3Space Tracker System was used for the measurement and monitoring of the hip angles. The rotation angle was defined as the rotation of the femur along the longitudinal axis of the femoral shaft. Within 750-mm range from the source, the positional accuracy was 0.8 mm RMS, and the angular accuracy was 0.5 RMS (Muraki et al., 2006). 2.3. Experimental procedure

Rod

Wooden jig

X

Z Y

Fig. 1. Experimental set-up. A trans-lumbar specimen was fixed on the wooden pole keeping the pelvis in an upright position with 30 of anterior tilt. A six-degree-offreedom electromagnetic tracking device 3Space Tracker System was used to measure hip angles.

On the basis of data obtained from preliminary experiments, flexion, abduction, and internal rotation for the superior iliofemoral ligament, and flexion, adduction, and internal rotation for the inferior ilio-femoral ligament, were excluded from this study, as the ilio-femoral ligaments were loose in those hip positions. For measurement of the superior ilio-femoral ligament, external rotation was included as a primary stretching position as it was shown to stretch the ligament in preliminary observation. The 6 stretching positions for the superior ilio-femoral ligament were determined as follows: (a) maximal external rotation, (b) maximal adduction, (c) maximal extension, and (d) 10 adduction with maximal external rotation, (e) 20 adduction with maximal

E. Hidaka et al. / Manual Therapy 14 (2009) 427–432

external rotation, and (f) 10 extension with maximal external rotation. For measurement of the inferior ilio-femoral ligament, extension was included as a primary stretching position as it was shown to stretch the ligament in preliminary observations. Thus, the 7 stretching positions for the inferior ilio-femoral ligament were determined as follows: (a) maximal extension, (b) maximal abduction, (c) maximal external rotation, (d) 20 external rotation with maximal extension, (e) 40 external rotation with maximal extension, (f) 20 abduction with maximal extension, and (g) 40 abduction with maximal extension. Strain on the ligament at each hip position was measured at the end of range, which was determined by Grade III mobilization after Kaltenborn’s procedure (Kaltenborn, 1999). The extent of unidirectional passive motion, applied by therapists, on the joints was graded from I to III. In that grading, Grade III mobilization comprised the manual application of force at a point which therapist perceived end-feel of the joint and observed no further stretching of the capsule or ligament. As the stretching procedure described in the Manual (Evjenth and Hamberg, 1984) is applied to the joint for 10–12 s after passive motion of the hip joint has reached the end of the range of movement, each position was maintained for more than 10 s until no increase or decrease in strain values was observed. Strain measurement was randomized in order for the superior and inferior ligament. During the mobilizations of the hip joint, the interclass correlation coefficient (1,1) of strain values obtained from 3 independent measurements using a displacement sensor was 0.936 for the ilio-femoral ligament. During the stretching procedures, the 3Space Tracker System during stretching procedures demonstrated the following average maximal ranges of hip motion; 23.9  6.4 for extension, 41.8  11.3 for external rotation, 22.0  11.1 for adduction, and 42.1 19.5 for abduction. The inter-class correlation coefficient (1,1) of measured ranges of motion during mobilization was 0.954. Throughout the experiment, which was performed at room temperature (22  C), the specimens were kept moist by spraying with saline solution every 10 min. To preserve smooth joint motion and reduce hysteresis of the ligaments, passive hip movement to the end-range was performed more than 10 times as preconditioning (Woo et al., 1986). 2.4. Identification of reference length (L0) and data analysis Based on a previous report using cadaver shoulders (Urayama et al., 2001), reference length (L0) was determined for each ligament. L0 was the length at which the angle-deformation curve of the ligament started to indicate a sudden increase in angle (Fig. 3). The L0 was the point between the laxity and linear regions in the angle-deformation curve. The displacement of the ligament (DL) was defined as the change in length from the L0. Since the distance between the points of the displacement sensor attached on the ligaments was not always the same, the strain on each ligament was calculated using the following formula (Muraki et al., 2006).

Strainð%Þ ¼ DLðmmÞ=L0 ðmmÞ  100 Strain values greater than 0% indicated positive stretching of the ligament from the L0. Values less than 0% indicated no stretching, and were presented as 0% strain. 2.5. Statistical analysis Statistical analysis was performed using SPSS for Windows ver.11.5J. One-way analysis of variance and post hoc Dunnett’s

a

429

b a laxity region

c c linear region

angle

b L0 deformation Fig. 3. Identification of reference length (L0). Reference length is the length at which the angle-deformation curve of the ligament starts to indicate a sudden increase in angle.

multiple comparisons test were used to determine the hip joint positions that showed significantly greater strain than that observed at the L0. The significance level was set at 0.05. 3. Results 3.1. The superior ilio-femoral ligament Strain on the superior ilio-femoral ligament with the hip at 10 adduction with maximal external rotation (3.2  3.3%), 20 adduction with maximal external rotation (4.0  4.2%), and maximal external rotation (3.7  3.0%) was significantly larger than the value obtained for L0 (p < 0.05). Few other hip positions demonstrated any positive strain on the superior ilio-femoral ligament (Fig. 4). 3.2. The inferior ilio-femoral ligament Strain on the inferior ilio-femoral ligament with the hip at maximal extension (2.1  2.1%) and 20 external rotation with maximal extension (1.8  2.1%) was significantly larger than the value obtained for L0 (p < 0.05). Few other hip positions demonstrated any positive strain on the inferior ilio-femoral ligament (Fig. 5). 4. Discussion Based on the traditional concepts of stretching to obtain effective lengthening of the ligaments, joint positioning by separating the origin and insertion of the ligament is important. The present study demonstrated that a significant amount of strain was observed on both the superior ilio-femoral ligament and the inferior ilio-femoral ligament in several hip positions.

430

E. Hidaka et al. / Manual Therapy 14 (2009) 427–432

10

*

8

*

*

6

4

2

0

ER

Add

Ext

Ext10°+ER

Add10°+ER

Add20°+ER

Fig. 4. Strain on the superior ilio-femoral ligament. The Y-axis indicates strain (%). Error bars indicate standard deviations. *A statistically significant difference between the obtained strain value and L0 value (p < 0.05).

The superior ilio-femoral ligament originates from the anterosuperior acetabular wall, extends parallel to the femoral neck axis, and inserts into the supero-lateral aspect of the inter-trochanteric line of the femur (Werner, 2004). As for external rotation, Kapandji (1970), Oatis (2004), and Werner (2004) indicated that adduction also stretched the ligament. The limiting factor of hip adduction was not only the superior ilio-femoral ligament but also the ischiofemoral ligament which extends from the ischium to the posterior aspect of the femoral neck (Luttgens and Wells, 1982; Fuss and Bacher, 1991; Oatis, 2004). The inferior ilio-femoral ligament originates from the anterior wall of the acetabulum, extends parallel to the femoral shaft axis, and inserts into the infero-medial aspect of the inter-trochanteric line of the femur (Werner, 2004). As for extension, Werner (2004) indicated that internal rotation also stretched the ligament. The limiting factor of hip internal rotation was reported to be the ischiofemoral ligament (Kapandji, 1970; Luttgens and Wells, 1982; Fuss and Bacher, 1991; Oatis, 2004; Werner, 2004). Although several authors have advocated various stretching positions, no consensus has been reached. While the anatomy of the hip joint capsule and ligaments has been well described, little or no information had been available on the material properties of the ilio-femoral ligaments (Fuss and Bacher, 1991). Recently the tensile properties of the ilio-femoral ligaments were reported by Hewitt et al. (2002). By using isolated ilio-femoral ligaments from cadaver hips, breaking strength and stiffness were measured. According to the report, the breaking strength and stiffness of the superior and inferior ilio-femoral ligaments were 320 N and 351 N, and 98.7 N/mm and 100.7 N/mm, respectively. In addition, it was noted that the ilio-femoral ligaments were the strongest capsular ligaments in the human body.

Further, from the stress–strain curves which were obtained from the cross-sectional area of the ligaments, maximal stress and strain on the ilio-femoral ligaments were found to be 2.7 MPa and 6.2%, respectively (Hewitt et al., 2001). This strain value was small compared with the values reported for the inferior gleno-humeral ligaments of the shoulder, i.e., 10.9% (Bigliani et al., 1992) and 9.3% (Ticker et al., 1996). In order to apply these biomechanical values to mobilization of the hip in patients, it is supposed that passive mobilization of the joint is performed by the application of a degree of force that can be estimated form the cross-sectional area of the ligaments. However, accurate evaluation of the cross-sectional area of the ilio-femoral ligament in vivo for each subject is difficult and the joint torque that should be applied to the individual hip joint cannot be clearly determined. Excessive torque in an inappropriate direction on a painful and contracted hip joint can occasionally cause ligament and cartilage damage, or even fracture (French, 2007; Hewitt et al., 2001; Wainner, 2004; Izumi et al., in press). Therefore, the direct application of these biomechanical values to patients involves considerable risk. For these reasons, this experimental study was designed to reduce the disparity between the biomechanical values and clinical practice. Taking into consideration a suitable safety range for the mobilization force used for the treatment of patients, in this study Grade III uni-directional mobilization after Kaltenborn’s procedure (Kaltenborn, 1999) was applied to the hip to simulate adequate torque to stretch and measure strain on the ilio-femoral ligaments. As a result, the maximal strains on the superior and inferior iliofemoral ligaments obtained by the mobilization were 3.2–4.1% and 1.8–2.1%, respectively, and those values were consistent with the strain values of toe lesions of the ilio-femoral ligaments (2–4%)

10 8 6 4

*

*

2 0

Ext

Abd

ER

ER20° +Ext ER40° +Ext Abd20° +Ext Abd40° +Ext

Fig. 5. Strain on the inferior ilio-femoral ligament. The Y-axis indicates strain (%). Error bars indicate standard deviation. *A statistically significant difference between the obtained strain value and L0 value (p < 0.05).

E. Hidaka et al. / Manual Therapy 14 (2009) 427–432

(Hewitt et al., 2001). This suggests that Grade III mobilization can apply adequate torque to the hip joint so as to provide maximal strain without resulting in injury to the ilio-femoral ligaments (Woo et al., 1994; Tillan and Hertling, 1996). The cause of flexion contracture of the hip joint after hip osteoarthritis or cerebro-vascular disease is thought to be the coexisting contracture of the ilio-femoral ligaments and the iliopsoas, adductor and internal rotator muscles of the hip (Sage, 1987; Lee et al., 1997; French, 2007). If muscle contracture has progressed and the hip is in a flexed position, surgical release of both the psoas muscle and ilio-femoral ligaments is required (O’Malley, 1959; Morais Filho et al., 2006). To prevent recurrence of the contracture, postoperative stretching of the ilio-femoral ligament and ilio-psoas muscle will be indicated using this procedure. Patients may require long-term physical therapy to treat the contracture of the hip. To improve these pathological conditions, a selective stretching procedure for the ilio-femoral ligaments is considered to be very important. Selective stretching applied separately for the superior and inferior ilio-femoral ligaments is thought to be effective in reducing total stress on the contracted ligaments of the hip joint. This may diminish the risk of chronic inflammation generated by micro rupture due to excessive strain (Tillan and Hertling, 1996). Based on the data obtained from this study, selective stretching for the superior ilio-femoral ligament with the hip at 10 or 20 adduction with maximal external rotation, and maximal external rotation, and that for the inferior ilio-femoral ligament with the hip at 20 external rotation with maximal extension and maximal extension should be performed as specific procedures. A series of selective stretches for the ilio-femoral ligaments may contribute to an effective treatment for flexion contracture of the hip joint. There are limitations to this study. First, since the trans-lumbar specimens were harvested from aged cadavers, the range of motion of the specimens might be different from those of specimens from younger people. The range of passive hip motion measured in the current study was 23.9  6.4 for extension, 41.8  11.3 for external rotation, 22.0  11.1 for adduction, and 42.1 19.5 for abduction. These values were similar to the standard values obtained in clinical measurements (Boone and Azen, 1979; Walker et al., 1984; Roach and Miles, 1991). Second, the mechanical properties of the ligaments in aged specimens might be different from those of specimens from younger people. Because tendon strain also decreases with age, the strain on the ilio-femoral ligaments observed in aged cadavers was thought to be less than that on tendons in younger cadavers (Yamada, 1970). Third, the displacement sensor had no capacity to measure strain in all dimensions. Therefore, the reference length (L0) was determined for each fiber direction at the center of the superior and inferior ilio-femoral ligaments. We considered that strain values obtained by the displacement sensor in those location represented strains of the superior and inferior ilio-femoral ligaments. Although we adopted uni-directional Grade III hip mobilization after Kaltenborn’s procedure (Kaltenborn, 1999) to obtain consistent strain on the ilio-femoral ligaments in this experimental study, Grade III or IV mobilization advocated by Maitland with reciprocating motions near the end-range is also considered to be applicable to clinical cases (Maitland, 1991). To resolve contracture of the hip joint with joint mobilization, long-term mechanical stress through an adequate amount of stretching should be applied to the contracted ligaments. This process may induce remodeling of the collagen tissue of the ligament and joint capsule, which will, in turn, help transform the tight, stiff joint into a flexible, mobile joint (Tillan and Hertling, 1996). For this reason, together with fact that the basic data support the beneficial mechanical effect of this procedure, joint mobilization should become a widely accepted clinical practice for the treatment of hip joint contracture.

431

5. Conclusion This study is designed to elucidate selective stretching positions for the superior and inferior ilio-femoral ligaments of the hip by measurement of strain. Based on the data obtained from this study, selective stretching for the superior ilio-femoral ligament with the hip at 10 or 20 adduction with maximal external rotation and maximal external rotation, and that for the inferior ilio-femoral ligament with the hip at 20 external rotation with maximal extension and maximal extension should be performed as specific procedures. A series of selective stretches for the ilio-femoral ligaments may contribute to an effective treatment for flexion contracture of the hip joint. Acknowledgments Authors sincerely thank E. Uchiyama, D. Suzuki, M. Fujii, S. Ohshiro, H. Takasaki, and H. Miyamoto for their useful suggestions and technical support. References Bigliani LU, Pollock RG, Soslowsky LJ, Flatow EL, Pawluk RJ, Mow VC. Tensile properties of the inferior glenohumeral ligament. Journal of Orthopaedic Research 1992;10:187–97. Boone DC, Azen SP. Normal range of motion of joints in male subjects. Journal of Bone and Joint Surgery 1979;61A:756–9. Evjenth O, Hamberg J. The extremities. In: Muscle stretching in manual therapy: a clinical manual, vol. 1. Alfta: Alfta Rehab Forlag; 1984. p. 7–12. French HP. Physiotherapy management of osteoarthritis of the hip: a survey of current practice in acute hospitals and private practice in the Republic of Ireland. Physiotherapy 2007;93:253–60. Fuss FK, Bacher A. New aspects of the morphology and function of the human hip joint ligaments. American Journal of Anatomy 1991;192:1–13. Hewitt J, Guilak F, Glisson R, Vail TP. Regional material properties of the human hip joint capsule ligaments. Journal of Orthopaedic Research 2001;19:359–64. Hewitt JD, Glisson RR, Guilak F, Vail TP. The mechanical properties of the human hip capsule ligaments. Journal of Arthroplasty 2002;17:82–9. Izumi T, Aoki M, Muraki T, Hidaka E, Miyamoto S. Stretching positions for the posterior capsule of the glenohumeral joint: strain measurement using cadaver specimens. American Journal of Sports Medicine, in press. Kaltenborn FM. The extremities. In: Manual mobilization of the joint: the Kaltenborn method of joint examination and treatment. 5th ed., vol. 1. Oslo: Olaf Norlis Bokhandel; 1999. p. 21–8. Kapandji IA. Lower limb. In: The physiology of the joints. 2nd ed., vol. 2. New York: Churchill Livingstone; 1970. p. 34–41. Lee D. The pelvic girdle: approach to be the examination and treatment of the lumbopelvic-hip. 3rd ed. New York: Churchill Livingstone; 2004. p. 105. Lee LW, Kerrigan DC, Della Croce U. Dynamic implications of hip flexion contractures. American Journal of Physical Medicine and Rehabilitation 1997;76:502–8. Luttgens K, Wells KF. Kinesiology: scientific basis of human motion. 7th ed. Philadelphia: Saunders College; 1982. p. 148–9. Maitland GD. Peripheral manipulation. 3rd ed. Oxford: Butterworth-Heinemann; 1991. p. 183–9. Mo¨ller MH, Oberg BE, Gillquist J. Stretching exercise and soccer: effect of stretching on range of motion in the lower extremity in connection with soccer training. International Journal of Sports Medicine 1985;6:50–2. Morais Filho MC, de Godoy W, Santos CA. Effects of intramuscular psoas lengthening on pelvic motion in patients with spastic diparetic cerebral palsy. Journal of Pediatric Orthopaedics 2006;26:260–4. Muraki T, Aoki M, Uchiyama E, Murakami G, Miyamoto S. The effect of arm position on stretching of the supraspinatus, infraspinatus, and posterior portion of deltiod muscles: a cadaveric study. Clinical Biomechanics 2006;21:474–80. Neumann D. Kinesiology of the musculoskeletal system: foundations for physical rehabilitation. London: Mosby Elsevier; 2002. p. 397–400. Oatis CA. Kinesiology: the mechanics and pathomechanics of human movement. Philadelphia: Lippincott; 2004. p. 667–9. O’Malley AG. Correspondence and preliminary communications. Journal of Bone and Joint Surgery 1959;41B:888–9. Roach KE, Miles TP. Normal hip and knee active range of motion: the relationship to age. Physical Therapy 1991;71(9):656–65. Saito H, Nagasaki H. In: Nakamura R, editor. Clinical kinesiology. 3rd ed. Tokyo: Ishiyaku Ltd; 2002. p. 455–6 [in Japanese]. Sage FP. Cerebral palsy. In: Crenshaw AA, editor. Campbell’s operative orthopaedics. 7th ed. St. Louis: Mosby; 1987. p. 2891–5. Ticker JB, Bigliani LU, Soslowsky LJ, Pawluk RJ, Flatow EL, Mow VC. Inferior glenohumeral ligament: goniometric and strain rate dependent properties. Journal of Shoulder and Elbow Surgery 1996;5:269–79. Tillan LJ, Hertling D. Properties of dense connective tissue and wound healing. In: Hertling D, Kessler RM, editors. Management of common musculoskeletal disorders. 3rd ed. Philadelphia: Lippincott; 1996. p. 8–21.

432

E. Hidaka et al. / Manual Therapy 14 (2009) 427–432

Urayama M, Itoi E, Hatakeyama Y, Pradhan RL, Sato K. Function of the 3 portions of the inferior glenohumeral ligament: a cadaveric study. Journal of Shoulder and Elbow Surgery 2001;10:589–94. Wainner RS. Passive versus active stretching of hip flexor muscles in subjects with limited hip extension: a randomized clinical trial. Physical Therapy 2004;84:800–7. Walker JM, Sue D, Mile-Elkousy N, Ford G, Trevelyan H. Active mobility of the extremities in older subjects. Physical Therapy 1984;64(6):919–23. Werner P. Locomotor system. In: Color atlas and textbook of human anatomy. 5th ed., vol. 1. Stuttgart, New York: George Thieme Verlag; 2004. p. 200–1. Woo S-LY, Orlando CA, Camp JF, Akeson WH. Effects of postmortem storage by freezing on ligament tensile behavior. Journal of Biomechanics 1986;10:399–404.

Woo SLY, An KN, Arnoczky SP, Wayne JS, Fithian DC. Anatomy, biology, and biomechanics of tendon, ligaments, and meniscus. In: Simon SR, editor. Orthopaedic basic science. Chicago: American Academy of Orthopaedic Surgeons; 1994. p. 45–74. Yamada H. In: Evans FG, editor. Strength of biological materials. Maryland: Williams and Wilkins; 1970. p. 248–80. Zakas A, Vergou A, Grammatikkopoulou NG, Zakas N, Sentelidis T, Vamvakoudis S. The effect of stretching during worming-up on the flexibility of junior handball players. Journal of Sports Medicine and Physical Fitness 2003;43:145–9. Zakas A, Grammatikkopoulou NG, Zakas N, Zahariadis P, Vamvakoudis E. The effect of active worm-up and stretching on the flexibility of adolescent soccer players. Journal of Sports Medicine and Physical Fitness 2006;46:57–61.

Manual Therapy 14 (2009) 433–438

Contents lists available at ScienceDirect

Manual Therapy journal homepage: www.elsevier.com/math

Original Article

Validity of the Neck Disability Index and Neck Pain and Disability Scale for measuring disability associated with chronic, non-traumatic neck pain Mark Chan Ci En a, Dean A. Clair b, Stephen J. Edmondston c, * a

Physiotherapy Department, Tan Tock Seng Hospital, Singapore Physiotherapy Department, Osborne Park Hospital, WA, Australia c School of Physiotherapy, Curtin University of Technology, GPO Box U1987, Perth, WA 6845, Australia b

a r t i c l e i n f o

a b s t r a c t

Article history: Received 17 January 2008 Received in revised form 13 May 2008 Accepted 27 July 2008

The purpose of this study was to evaluate the construct and content validity of the Neck Disability Index (NDI) and the Neck Pain and Disability Scale (NPAD) in patients with chronic, non-traumatic neck pain. Twenty patients (mean age ¼ 64.5 years) completed a patient-specific questionnaire, the Problem Elicitation Technique (PET), followed by the NDI and NPAD. Content validity was assessed by comparing the items of the NDI and NPAD with problems identified from the PET. Construct validity of the fixed-item questionnaires was examined by establishing the correlation with each other, and with the PET score. Eleven common problems were identified by patients through the PET, of which six were included in the NDI and seven included in the NPAD. The NDI and NPAD scores were strongly correlated (r ¼ 0.86, p < 0.01), while the correlation between the PET and the fixed-item questionnaires was moderate (NDI: r ¼ 0.62, p < 0.01; NPAD: r ¼ 0.71, p < 0.01). Both the NDI and the NPAD include most of the functional problems common to this patient group, and display good content validity. The PET is better able to evaluate the problems specific to the individual patient and is therefore measuring a somewhat different construct to the fixed-item questionnaires. Ó 2008 Elsevier Ltd. All rights reserved.

Keywords: Neck pain Disability Outcome assessment Questionnaires

1. Introduction Neck pain has a lifetime prevalence of about 70% in the general population (Makela et al., 1991; Bovim et al., 1994). Although acute neck pain often resolves, about 19% of the population may suffer from chronic neck pain at any given time (Bovim et al., 1994; Guez et al., 2002). Measurement of the impact of neck pain on the sufferer presents a challenge due to the variability between patients in pain intensity, and the effect of the disorder on physical and psychological functions (Clair et al., 2004). Measures of pain intensity and tissue sensitivity have been used to quantify the sensory dimension of neck pain disorders (Hubka and Phelan, 1994; Olson et al., 2000), while range of motion and muscle function has been used to measure impairments of physical function (Falla et al., 2004; Jull et al., 2004; Hoving et al., 2005; O’Leary et al., 2007). However, recent recommendations place greater emphasis on functional status and quality of life more broadly, in the evaluation of neck pain disorders (Philadelphia Panel, 2001; Pietrobon et al., 2002). Measurement of function has been a developing theme in neck pain research as this shifts the focus away from signs and symptoms towards the specific effects of the symptoms on patient * Corresponding author. Tel.: þ61 8 9266 3665; fax: þ61 8 9266 3699. E-mail address: [email protected] (S.J. Edmondston). 1356-689X/$ – see front matter Ó 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.math.2008.07.005

function (Pietrobon et al., 2002). In relation to neck pain, this includes neck function, physical function more generally, and psychological function. A range of neck pain-specific questionnaires have been developed for this purpose, and have been incorporated into recent clinical studies (Vernon and Mior, 1991; Leak et al., 1994; Jordan et al., 1998; Westaway et al., 1998; Wheeler et al., 1999). The value of questionnaires is dependent on a range of factors but of primary importance is the validity, particularly in relation to construct and content. A recent review of neck pain-specific questionnaires concluded that most have not been extensively validated, and recommended a comparative study to better define the psychometric properties of the commonly used instruments (Pietrobon et al., 2002). The Neck Disability Index (NDI) is the most commonly used questionnaire for the measurement of neck pain disability. It was originally developed to evaluate the activities of daily living in patients with disabling neck pain, particularly that resulting from whiplash trauma (Vernon and Mior, 1991). The NDI includes 10 questions of which 7 examine functional activities, 2 ask about symptoms and the final question considers concentration. The Neck Pain and Disability Scale (NPAD) was developed to provide clinicians with a tool to assess the multi-dimensional effects of the neck pain disorder (Wheeler et al., 1999). The scale consists of 20 questions relating to 4 domains (neck function, pain intensity, emotion/ cognition and activities of daily living) which look at the effects of

434

M. Chan Ci En et al. / Manual Therapy 14 (2009) 433–438

the neck pain disorder on patients’ physical and emotional functions. The potential limitation of these questionnaires, and others with fixed questions, is that they constrain the scope of the evaluation to the specific issues included. Therefore, the questionnaire may include questions not relevant to some patients, and may not include issues of importance. An alternative to the fixed-item questionnaire are the patientspecific techniques which require patients to generate their own, possibly unique, set of problems or items. The patient-specific methods offer the advantage of identifying the problems or issues relevant to each individual, and are therefore consistent with the approach to patient evaluation commonly employed in clinical practice (Jolles et al., 2005). Two patient-specific techniques which have been used to evaluate neck pain are the Problem Elicitation Technique (PET) and the Patient-Specific Functional Scale (PSFS) (Buchbinder et al., 1995; Westaway et al., 1998). The disadvantage of this approach, particularly in research, is that without standardisation of content, the scale is different for each patient (Jolles et al., 2005). The level of statistical correlation between patientspecific scales and fixed-item questionnaires has been found to be only moderate (Westaway et al., 1998; Hoving et al., 2003). Clinical studies in which multiple neck pain questionnaires are applied simultaneously in the same patient population have been identified as an important focus for research in this area (Pietrobon et al., 2002). Specifically, patient-specific questionnaires will help identifying the problems which are most common and relevant to specific sub-groups of patients with neck pain. This may assist the development or modification of fixed-item questionnaires, and enhance the psychometric evaluation, particularly the content and construct validity. Hoving et al. (2003) used the PET to evaluate the validity of the NDI and Northwick Park Neck Pain questionnaire in patients with neck pain associated with whiplash injury. They found that the two fixed-item questionnaires did not fully cover the problems considered important in patients with whiplash injury, especially those concerning emotional and social functions. More recently, the NDI has been shown to have poor construct validity and to be less responsive to change than the PSFS, in patients with cervical radiculopathy (Cleland et al., 2006). Further, the NDI has been shown to be less responsive to change than previously reported in patients with non-traumatic neck pain (Cleland et al., 2008). These findings suggest that analysis of the psychometric properties of the NDI and other fixed-item questionnaires in different groups of patients with neck pain should be an on-going process. To date there has been little evaluation of the common problems associated with chronic, non-traumatic neck pain, particularly in older patients. Recent studies of this patient group have examined patient variability and treatment dose (Clair et al., 2004; Clair et al., 2006), but an evaluation of the common functional problems and validity of fixed-item questionnaires has not been conducted. The purpose of this study was to examine the content and construct validity of two fixed-item questionnaires as measures of disability in patients with chronic, non-traumatic neck pain. This was achieved through a comparison of the NDI and NAPD questionnaires in a cohort of patients with this disorder, and comparison of the responses to the fixed-item questionnaires with a patient-specific questionnaire, the PET. 2. Methods 2.1. Study population A cross-sectional survey of 20 subjects with chronic, non-traumatic neck pain was performed. Subjects were recruited from two public hospitals in Perth, Western Australia. Nineteen subjects were referred to the physiotherapy department for treatment of chronic,

non-traumatic neck pain. The remaining subject was a physiotherapist who had a long-standing history of non-traumatic neck pain. All subjects had a current episode of neck pain of greater than three months duration, with pain predominantly located in the somatic referral zones of the cervical spine (Grubb and Kelly, 2000). Participants were excluded if the symptom duration did not exceed three months, or were unable to read or comprehend the questionnaires. Subjects with a specific diagnosis such as cervical radiculopathy, ankylosing spondylitis and rheumatoid arthritis, or who had a history of neck trauma were not included in the study. The institutional Human Research Ethics Committee granted approval for this study. 2.2. Procedure Upon consenting to be involved in the study, participants were asked general questions in relation to age, symptom duration, current medication intake, and average pain intensity over the previous week. The average pain intensity was measured using a 10 cm visual analogue scale (VAS). The PET was firstly performed for all subjects, and was administered by a skilled interviewer. After PET, all participants completed the NDI and NPAD in a random order. 2.3. Patient questionnaires The PET is a disability questionnaire designed to help clinicians to identify the most significant problems experienced by each individual patient (Bakker et al., 1995; Buchbinder et al., 1995). A skilled interviewer carries out the interview where subjects are allowed to spontaneously identify problems associated with their neck pain disorder. The interviewer then assists the subject with a series of open-ended questions, which cover areas such as selfcare and work, mobility, leisure activities, social activities, emotion, communication and sleep. Subjects are allowed to identify a maximum of 15 problems. The subject is then asked to rate the severity of each problem on an 11-point numerical scale (no severity ¼ 0 to extremely severe ¼ 10). Finally, the subject ranks the problems according to their importance, from the most to least significant. In this study, the overall PET score was defined as the sum of the severity scores for all problems divided by the number of problems identified. The PET score had a possible range between 0 and 10. The NDI is a 10-item questionnaire which asks patients about their symptoms and the effect of their neck pain on a range of functional activities (Vernon and Mior, 1991). The items in the questionnaire are pain intensity, personal care, lifting, reading, headache, concentration, work, driving, sleeping and recreation. The subject is instructed to circle one of the six options which describes the severity of each item (0–5). The NDI score is calculated as the sum of the scores for each question multiplied by two (range ¼ 0–100). A higher score is indicative of greater disability associated with the neck disorder (Vernon and Mior, 1991). The NPAD is a 20-item questionnaire which examines the effects of the neck pain disorder and covers four factors of neck function, pain intensity, cognitive and emotional affects and activities of daily living (Wheeler et al., 1999). The subject responds to each question on a 10 cm visual analogue scale where the subject indicates the severity or frequency of each item. Each question was measured using a ruler to provide each individual score. The total NPAD score is the sum of the scores of all 20 questions divided by 2, where the maximum score is 100 and the minimum is 0. Like the NDI, a higher score will indicate greater disability. The NPAD has the potential advantage as the four factors examine the various aspects of the effects of the neck pain disorder on patients’ physical, cognitive or emotional function (Goolkasian et al., 2002; Clair et al., 2004).

M. Chan Ci En et al. / Manual Therapy 14 (2009) 433–438

2.4. Statistical analyses

Table 2 Disability profile of the study population according to the PET dimensions.

The content validity of the NDI and NPAD was studied by comparing the items in the questionnaires with the problems identified by the PET. The frequency of each problem and ranking by importance were determined. The construct validity of the NDI and NPAD was examined by determining the association between the questionnaire scores and the PET score, using Pearson’s correlation coefficients. The correlation between the NDI and NPAD scores was also examined. 3. Results Twenty subjects (7 males, 13 females) with a mean age of 64.5 years (SD ¼ 12.8) were recruited over a three-month period. The mean symptom duration was 115.6 months (SD ¼ 119.5). The characteristics of the study population are summarised in Table 1. The average pain intensity over the week prior to the interview was 5.2 (SD ¼ 1.9). The mean NDI and NPAD scores were 33.1 (SD ¼ 17.2) and 47.7 (SD ¼ 23.7), respectively. The mean PET score was 6.6 (SD ¼ 1.7). The PET identified an average of 9.2 problems per patient (SD ¼ 3.7). 3.1. Content validity Table 2 presents the disability profile of the study population according to the PET dimensions. Over 80% of the patients identified one or more problems within the dimensions of sleep (90.0%), mobility (90.0%) and role activity (85.0%). There were 23, 30 and 53 individual problems identified within these three dimensions, respectively. More than half of the subjects identified problems within the dimensions of emotion (75.0%) and symptoms (75.0%). The individual patient problems identified by the PET and their mean severity are presented in Table 3. Only problems identified by two or more patients are presented. Of the individual problems identified by the PET, sleep disturbance had the highest prevalence (75.0%). More than half of the subjects identified frustration (65.0%), driving (60.0%) and lifting (60.0%). Other common problems (more than 33%) were looking into cupboards (45.0%), gardening (40.0%), headaches (40.0%), housework (35.0%), working overhead (35.0%), and general exercise (35.0%). The highest mean

Role activity Emotional Sleep Mobility Sport and leisure Social activity Personal care Communication Symptoms

Gender (%) Male Female Symptom duration (%) 3–6 months 6–12 months 12–24 months >24 months Medication (%) None Analgesics only Analgesics and NSAIDS Anti-depressants Outcome measures [mean (SD)] Average pain intensity over past week NDI percentage score (0–100) NPAD percentage score (0–100) Overall PET score (0–10) Mean number of problems elicited

64.5 12.8 22.0–83.0

(15.0) (5.0) (15.0) (65.0)

4 6 9 1

(20.0) (30.0) (45.0) (5.0)

5.2 33.1 47.7 6.6 9.2

(1.9) (17.2) (23.7) (1.7) (3.7)

Overall number of problems identified in each dimension

Number of different problems within each dimension

Mean severity per dimension (possible range 0–10)

17 15 18 18 6

(85.0) (75.0) (90.0) (90.0) (30.0)

53 19 23 30 6

12 4 2 5 1

6.8 7.1 7.1 6.8 6.5

10 8 10 15

(50.0) (40.0) (50.0) (75.0)

10 10 11 19

3 2 3 3

6.9 6.5 6.9 7.4

severity scores were found in depression (9.3), cooking (8.3), and sitting upright (8.0). The individual problems that were ranked most important by the subjects were driving (30.0%), sleep disturbance (30.0%) and frustration (20.0%). Of the 11 problems identified by most subjects, 6 of these are included in the NDI and 7 are included in the NPAD (Table 4). 3.2. Construct validity The PET was moderately correlated with the NDI (r ¼ 0.62, p < 0.01) and NPAD (r ¼ 0.71, p < 0.01). There was a stronger correlation between the NDI and NPAD (r ¼ 0.86, p < 0.01). The Table 3 Problems identified by patients using the PET, and the mean severity and ranking of importance. Dimension

Problem

Role Activity

Looking into cupboards Gardening Housework Working overhead Hanging up washing Work for wages Vacuuming Cooking

Number Mean of subjects (SD) (%) severity (0–10) 9 8 7 7 6 4 4 4

Number of subjects ranking problem as most important (%)

(45) (40) (35) (35) (30) (20) (20) (20)

5.3 6.6 7.7 6.6 7.8 6.5 7.0 8.3

0 0 3 0 2 3 1 1

(0) (15) (10) (15) (5) (5)

Emotional

Frustration Depression

13 (65) 4 (20)

6.5 9.3

4 (25) 0

Sleep

Affected sleep Fatigue throughout day

15 (75) 8 (40)

7.1 7.1

6 (30) 0

Mobility

Driving Lifting Crossing the road

12 (60) 12 (60) 3 (15)

6.2 7.6 7.0

6 (30) 2 (10) 1 (5)

Social Activities

Socialising with friends

6 (30)

6.5

0

Sports and leisure

General exercise Non athletic leisure activities

7 (35) 2 (10)

7.3 7.0

0 0

Personal Care

Dressing Hair care

6 (30) 4 (20)

6.0 7.3

2 (10) 0

Communication Computer use Reading

6 (30) 4 (20)

6.3 7.3

1 (5) 1 (5)

Symptoms

8 (40) 6 (30) 5 (25)

7.6 7.8 6.6

2 (10) 1 (5) 0

7 (35.0) 13 (65.0) 3 1 3 13

Number of patients who identified problems within each dimension (%)

Table shows the number of problems, the number of patients who identified problems within each dimension, and the mean severity of problems per dimension.

Table 1 Demographic and clinical characteristics of the study population (n ¼ 20). Age (years) Mean SD Range

435

Headaches Concentration Neck movements

436

M. Chan Ci En et al. / Manual Therapy 14 (2009) 433–438

Table 4 Comparison of commonly identified problems elicited by the PET with items in NDI and NPAD. Identified problem

Item included in NDI

Item included in NPAD

Affected sleep Frustration Driving Lifting Looking into cupboards Gardening Headaches Fatigue throughout day Housework Working overhead

Yes

Yes Yes Yes

Yes Yes

Yes Yes Yes

Yes Yes

association between the three questionnaires is presented graphically in Fig. 1. 4. Discussion In the development of neck pain questionnaires, assumptions were made as to the nature of the functional limitations associated with neck pain. However, patient input is considered paramount in the development and evaluation of an outcome measure (Guyatt et al., 1993). While there was some input from patients in the

construction of the NDI and NPAD, some questionnaires have been developed with little or no input from patients with neck pain (Pietrobon et al., 2002). Establishing the validity of a questionnaire is important to ensure that it reflects the nature and spectrum of the problems experienced by the majority of patients. The validity of the NDI and NPAD has been investigated in patients with whiplash-related neck pain (Goolkasian et al., 2002; Hoving et al., 2003), but this is the first study to specifically examine the validity of two commonly used disability questionnaires in an older cohort of patients with chronic, non-traumatic neck pain. The PET simulates clinical practice by asking patients to identify physical, emotional and cognitive problems specific to their neck pain disorder. It serves to elicit the problems specific to each patient, thereby reducing the ‘noise’ created when items not relevant to the patient are included (Buchbinder et al., 1995; Jolles et al., 2005). The most commonly reported functional problems in the present study were disturbed sleep, driving, and lifting, while frustration was the most common emotional problem. In patients with whiplash, who were significantly younger than the patients in the current study, the most common functional problems identified were work for wages, fatigue during the day, participation in sports, and driving, while the most common emotional problem was depression (Hoving et al., 2003). Driving or riding in a car was the only common functional problem experienced by both patient

80

100 r = 0.71

r = 0.86 80

NPAD Score

40

60

40

20 20

0

0 0

20

40

60

80

100

3

4

5

NPAD Score

6

7

PET Score

80 r = 0.62

60

NDI Score

NDI Score

60

40

20

0 3

4

5

6

7

8

9

PET Score Fig. 1. Scatter plots showing the relationships between total scores for the NDI, NPAD and PET questionnaires.

8

9

M. Chan Ci En et al. / Manual Therapy 14 (2009) 433–438

groups. This finding suggests that the impact of neck pain on physical and emotional functions may be somewhat different in older patients with non-traumatic neck pain compared to younger patients who have whiplash-related neck pain. This is consistent with the studies which have reported differences in the nature and severity of physical impairments in patients with neck pain of traumatic origin, compared to those with a non-traumatic onset (Dumas et al., 2001; Drottning et al., 2002). Of the 10 most common problems identified by the PET, 6 were included in the NDI and 7 were included in the NPAD, which supports the content validity of both questionnaires for this patient population. Sleep disturbance, driving and frustration were ranked as the three most commonly reported problems. All are included in the NPAD, while frustration is not addressed in the NDI. A significant impact on psychological function has been described in patients with chronic, non-traumatic neck pain, which was found to improve with improvements in pain intensity and functional limitation (Clair et al., 2006). The NPAD differs from the NDI in that it includes questions which relate specifically to emotion and social function. The NPAD uses sub-domains to identify more specifically the areas of physical and psychological functions most commonly indicated by the patient as being affected. While most of the common functional problems relevant to this patient group are included in the NDI and NPAD, the greater scope of the latter questionnaire may provide better information about the impact of the disorder on the patient more broadly. The results of the present study suggest that neck pain questionnaires should have a greater emphasis on neck function, activities of daily living and psychological function, and limited emphasis on symptoms such as pain intensity, tissue tenderness and movement restrictions which can be measured in other ways. The high correlation between the NDI and NPAD scores suggests that they measure the same construct in this patient group. While the questionnaire format and scoring systems are different, the questionnaires address a range of common items, which relate to function rather than symptoms. Previous studies have shown high correlation between the NDI and other fixed-item neck pain questionnaires, where the items in the questionnaires were very similar (Hoving et al., 2003; Wlodyka-Demaille et al., 2004; Gay et al., 2007). The only previous study to directly compare the NDI and NPAD found a moderate correlation (r ¼ 0.72) between questionnaires, in younger patients with neck pain of both traumatic and atraumatic origins (Goolkasian et al., 2002). Consistent with the study of Hoving et al. (2003), there was a moderate correlation between the PET and both fixed-item questionnaires, which suggests that the PET measures a somewhat different construct. This may be due to the PET only scoring items of relevance to each patient. For this reason, the PET reflects clinical practice as it identifies problems relevant to the individual, which may include issues not addressed in fixed-item questionnaires. A recommendation based on the results of the present study is that the PET should be used in conjunction with a fixed-item questionnaire, as each provides different information about the study population. While fixed-item questionnaires are relatively simple to administer, the PET requires some training and experience of the interviewer to ensure consistency in its application and adherence to the target concept (Jolles et al., 2005). During the study the relevance of the driving item in the NDI was raised by some patients, as many of the subjects did not drive, either due to their age or the neck pain disorder. The applicability of the driving item in the NDI to non-drivers has not previously been considered and may be an area for review. Previous studies have chosen to modify the NDI in an attempt to improve the relevance of the questionnaire to the specific study population (Hains, 1998; Riddle and Stratford, 1998). In the present study, the item was answered by either a driver or a passenger, consistent with the driving question in the NPAD. This modification to the NDI should

437

be considered if the questionnaire is used in future studies with this patient group. A potential limitation of this study is the relatively small study population. However, the characteristics of the patients were consistent with those of larger studies of patients with chronic, non-traumatic neck pain (Clair et al., 2004, 2006). The higher proportion of women in this study (65%) is consistent with gender ratios in other neck pain studies where the proportion of female subjects has been between 60 and 70% (Guez et al., 2003; Clair et al., 2004; Gay et al., 2007). The age and symptom duration suggest that degenerative pathology and the effects of aging may be important in the development of the symptoms, however, a review of radiological examinations and correlation with symptoms were not part of the present study. The patients in the study were receiving treatment in a public health system, and the problems identified may not reflect those of all patients with non-traumatic neck pain in the broader community. In conclusion, the NDI and NPAD both identified the common problems considered important by the patients, and the NPAD included all problems ranked as most important. Both questionnaires have good content validity and are therefore equally relevant for use in this patient group. The broader scope of the NPAD, particularly in relation to emotional and social functions, may be an advantage in future studies. In future research involving chronic, non-traumatic neck pain, it is recommended that a patient-specific questionnaire such as the PET should be used in conjunction with a fixed-item neck pain questionnaire, as each seems to measure a somewhat different construct. Acknowledgements The authors acknowledge the support of the staff of the Physiotherapy Outpatient Department, Osborne Park Hospital, Perth, Western Australia, for their assistance in recruitment of subjects for this study. References Bakker C, van der Linden S, van Santen-Hoeufft M, Bolwijn P, Hidding A. Problem elicitation to assess patient priorities in ankylosing spondylitis and fibromyalgia. Journal of Rheumatology 1995;22:1304–10. Bovim G, Schrader H, Sand T. Neck pain in the general population. Spine 1994;19:1307–9. Buchbinder R, Bombardier C, Yeung M, Tugwell P. Which outcome measures should be used in rheumatoid arthritis clinical trials? Arthritis & Rheumatism 1995;38:1568–80. Clair D, Edmondston S, Allison G. Variability in pain intensity, physical and psychological function in non-acute, non-traumatic neck pain. Physiotherapy Research International 2004;9:43–54. Clair DA, Edmondston SJ, Allison GT. Physical therapy treatment dose for nontraumatic neck pain: a comparison between 2 patient groups. Journal of Orthopaedic and Sports Physical Therapy 2006;36:867–75. Cleland JA, Childs JD, Whitman JM. Psychometric properties of the neck disability index and numeric pain rating scale in patients with mechanical neck pain. Archives of Physical Medicine and Rehabilitation 2008;89:69–74. Cleland JA, Fritz JM, Whitman JM, Palmer JA. The reliability and construct validity of the neck disability index and patient specific functional scale in patients with cervical radiculopathy. Spine 2006;31:598–602. Drottning M, Staff PH, Sjaastad O. Cervicogenic headache (CEH) after whiplash injury. Cephalalgia 2002;22:165–71. Dumas J-P, Arsenault AB, Boudreau G, Magnoux E, Lepage Y, Bellavance A, Loisel P. Physical impairments in cervicogenic headache: traumatic vs. non-traumatic onset. Cephalalgia 2001;21:884–93. Falla D, Bilenkij G, Jull G. Patients with chronic neck pain demonstrate altered patterns of muscle activation during performance of a functional upper limb task. Spine 2004;29:1436–40. Gay RE, Madson TJ, Cieslak KR. Comparison of the neck disability index and the Bournemouth neck questionnaire in a sample of patients with chronic uncomplicated neck pain. Journal of Manipulative and Physiological Therapeutics 2007;30:259–62. Goolkasian P, Wheeler AH, Gretz SS. The neck pain and disability scale: test–retest reliability and construct validity. Clinical Journal of Pain 2002;18:245–50. Grubb SA, Kelly CK. Cervical discography: clinical implications from 12 years of experience. Spine 2000;25:1382–9.

438

M. Chan Ci En et al. / Manual Therapy 14 (2009) 433–438

Guez M, Hildingsson C, Nilsson M, Toolanen G. The prevalence of neck pain. Acta Orthopaedica Scandinavica 2002;73:455–9. Guez M, Hildingsson C, Stegmayr B, Toolanen G. Chronic neck pain of traumatic and non-traumatic origin. Acta Orthopaedica Scandinavica 2003;74:576–9. Guyatt G, Feeney DH, Patrick DL. Measuring health-related quality of life. Annals of Internal Medicine 1993;118:622–9. Hains F. Psychometric properties of the neck disability index. Journal of Manipulative and Physiological Therapeutics 1998;21:75–80. Hoving JL, O’Leary EF, Niere KR, Green S, Buchbinder R. Validity of the neck disability index, Northwick Park neck pain questionnaire, and problem elicitation technique for measuring disability associated with whiplash-associated disorders. Pain 2003;102:273–81. Hoving JL, Pool JJ, van Mameren H, Deville WJ, Assendelft WJ, de Vet HC, de Winter AF, Koes BW, Bouter LM. Reproducibility of cervical range of motion in patients with neck pain. BMC Musculoskeletal Disorders 2005;6:59–67. Hubka MJ, Phelan SP. Inter-examiner reliability of palpation for cervical spine tenderness. Journal of Manipulative and Physiological Therapeutics 1994;17:591–5. Jolles BM, Buchbinder R, Beaton DE. A study compared nine patient-specific indices for musculoskeletal disorders. Journal of Clinical Epidemiology 2005;58:791–801. Jordan A, Manniche C, Mosdal C, Hindsberger C. The Copenhagen neck functional disability scale: a study of reliability and validity. Journal of Manipulative and Physiological Therapeutics 1998;21:520–7. Jull G, Kristjansson E, Dall’Alba P. Impairment in the cervical flexors: a comparison of whiplash and insidious onset neck pain patients. Manual Therapy 2004;9:89–94. Leak AM, Cooper J, Dyer S, Williams KA, Turner-Stokes L, Frank AO. The Northwick Park neck pain questionnaire, devised to measure neck pain and disability. British Journal of Rheumatology 1994;33:469–74.

Makela M, Heliovarra M, Sievers K, Impivaara O, Knekt P, Aromaa A. Prevalence, determinants, and consequences of chronic pain in Finland. American Journal of Epidemiology 1991;124:1356–67. O’Leary S, Jull G, Kim M. Vicenzino B Cranio-cervical flexor muscle impairment at maximal, moderate, and low loads is a feature of neck pain. Manual Therapy 2007;12:34–9. Olson SL, O’Connor DP, Birmingham G, Broman P, Herrera L. Tender point sensitivity, range of motion and perceived disability in subjects with neck pain. Journal of Orthopaedic and Sports Physical Therapy 2000;30:13–20. Philadelphia Panel. Philadelphia Panel evidence-based clinical practice guidelines on selected rehabilitation interventions for neck pain. Physical Therapy 2001;81:1701–17. Pietrobon R, Coeytaux RR, Carey TS, Richardson WJ, DeVellis RF. Standard scales for measurement of functional outcome for cervical pain or dysfunction. Spine 2002;27:515–22. Riddle DL, Stratford PW. Use of generic versus region-specific functional status measures on patients with cervical spine disorders. Physical Therapy 1998;79:951–63. Vernon H, Mior S. The neck disability index: a study of reliability and validity. Journal of Manipulative and Physiological Therapeutics 1991;14:409–15. Westaway MD, Stratford PW, Binkley JM. The patient-specific functional scale: validation of its use in persons with neck dysfunction. Journal of Orthopaedic and Sports Physical Therapy 1998;27:331–8. Wheeler AH, Goolkasian P, Baird AC, Darden BV. Development of the neck pain and disability scale: item analysis, face, and criterion-related validity. Spine 1999;24:1290. Wlodyka-Demaille S, Poiraudeau S, Catanzariti J-F, Rannou F, Fermanian J, Revel M. The ability to change of three questionnaires for neck pain. Joint Bone Spine 2004;71:317–26.

Manual Therapy 14 (2009) 439–443

Contents lists available at ScienceDirect

Manual Therapy journal homepage: www.elsevier.com/math

Original Article

The use of osteopathic manipulative treatment as adjuvant therapy in patients with peripheral arterial disease Rita Lombardini a, *, Simona Marchesi a, Luca Collebrusco b, Gaetano Vaudo a, Leonella Pasqualini a, Giovanni Ciuffetti a, Matteo Brozzetti a, Graziana Lupattelli a, Elmo Mannarino a a b

Internal Medicine, Angiology and Atherosclerosis, University of Perugia, Italy Division of Physiotherapy, ‘‘S. Maria della Misericordia’’ Hospital, Perugia, Italy

a r t i c l e i n f o

a b s t r a c t

Article history: Received 20 November 2007 Received in revised form 16 July 2008 Accepted 2 August 2008

Peripheral arterial disease (PAD) is a manifestation of systemic atherosclerosis associated with impaired endothelial function and intermittent claudication is the hallmark symptom. Hypothesizing that osteopathic manipulative treatment (OMT) may represent a non-pharmacological therapeutic option in PAD, we examined endothelial function and lifestyle modifications in 15 intermittent claudication patients receiving osteopathic treatment (OMT group) and 15 intermittent claudication patients matched for age, sex and medical treatment (control group). Compared to the control group, the OMT group had a significant increase in brachial flow-mediated vasodilation, ankle/brachial pressure index, treadmill testing and physical health component of life quality (all p < 0.05) from the beginning to the end of the study. At univariate analysis in the OMT group there was a negative correlation between changes in brachial flow-mediated vasodilation and IL-6 levels (r ¼ 0.30; p ¼ 0.04) and a positive one between claudication pain time and physical function score (r ¼ 0.50; p ¼ 0.05). In conclusion, despite the relatively few patients in our study, these results suggest that OMT significantly improves endothelial function and functional performance in intermittent claudication patients along with benefits in quality of life. This novel treatment combined with drug and lifestyle modification might be an effective alternative to traditional training based on exercise. Ó 2008 Elsevier Ltd. All rights reserved.

Keywords: Intermittent claudication Osteopathic manipulative treatment Endothelial function

1. Introduction Peripheral arterial disease (PAD) is a manifestation of systemic atherosclerosis. Intermittent claudication, the predominant clinical symptom, is characterized by cramping, aching or fatigue, which typically involves the calf muscles, thighs and buttocks. PAD is associated with impaired endothelium-dependent vasodilation of peripheral vessels (Weiss et al., 2002) and by an increased expression of adhesion molecules and release of proinflammatory circulating molecules (Fiotti et al., 1999; Brevetti et al., 2001a). In patients with mild to moderate symptoms (Fontaine stage II; European Working Group on Critical Leg Ischemia, 1991) conservative treatment improves endothelial function and peripheral circulation. Results of functional tests improve and quality of life is better (Shepard and Balady, 1999; Brendle et al., 2001). Conservative treatment includes dietary and pharmacological risk factor

* Corresponding author. Medicina Interna, Angiologia e Malattie da Arteriosclerosi, Universita` di Perugia, Ospedale ‘‘S. Maria della Misericordia’’, Loc. S. Andrea delle Fratte, 06156 Perugia, Italy. E-mail address: [email protected] (R. Lombardini). 1356-689X/$ – see front matter Ó 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.math.2008.08.002

modification, mild invasive strategies such as spinal cord stimulation (SCS) and exercise therapy. Risk factor modification (lipid lowering drugs, glycaemic control, antihypertensive and antiplatelet treatment), is recognized to be beneficial. SCS is a neuromodulation technique using electricity which has shown beneficial impacts on quality of life in patients with ischemic heart disease and PAD (Ubbink et al., 2004; De Vries et al., 2007). Exercise training is a critical and debated part of PAD patient treatment. It is often plain advice to ‘‘walk more’’ rather than follow a closely supervised exercise programme. In fact, there is no consensus on the most effective form of exercise therapy. The kind and frequency of exercise, level of supervision and improvement scales have not yet been fully defined, and few trials have addressed these specific points (Bartelink et al., 2004). Several recent studies showed exercise training confers benefits (Gardner and Poehlman, 1995; Collins et al., 2007), but could enhance endothelial injury (Hickman et al., 1994; Brevetti et al., 2001b). During rest intervals in an exercise programme claudicant patients suffered repeated episodes of leg muscle ischaemia- and reperfusion injury which led to development of a systemic inflammatory response that was hypothesized to injure the endothelium, accelerate atherosclerosis and increase thrombotic risk

440

R. Lombardini et al. / Manual Therapy 14 (2009) 439–443

(Silvestro et al., 2002; Burns et al., 2003). Given these findings, along with few manageable supervised exercise programmes and major financial and logistical obstacles to improving exercise programmes, we turned our attention to a more cost-effective approach. Osteopathic manipulative therapy (OMT) is one of the hallmarks of osteopathic medicine, which was founded in 1874 by Andrew Taylor Still (Still, 1992). Osteopathic medicine focuses on the union of all body components and identifies the muscoloskeletal system as a key to health. In osteopathic manipulation, bones, muscles, and tendons are manipulated to promote blood flow through tissues and thus enhance the body’s own healing powers (DiGiovanna et al., 1997). The osteopathic manipulative techniques used may allow for a normalization of imbalances between the sympathetic and parasympathetic nervous systems and improved vascular motion which would result in a more balanced homeostatic mechanism. A recent report by Salamon et al. (2004) used their own findings in the area of nitric oxide (NO) research to explain the therapeutic vascular effects of OMT. Until recently, there was little scientific evidence to support these claims. Despite reports that OMT is successful in low back pain and in patients with cardiovascular diseases, including hypertension (Rogers and Rogers, 1976; Spiegel et al., 2003), there are no reports of its use in PAD. The present pilot study investigated whether OMT, when combined with lifestyle modifications and pharmacological therapy, could be of benefit to patients with intermittent claudication. 2. Patients and methods The study was carried out from January 2005 to February 2006. Consecutive eligible male patients with PAD were recruited from those attending the Unit of Internal Medicine, Angiology and Atherosclerosis at the University of Perugia – ‘‘Santa Maria della Misericordia’’ Hospital (Perugia, Italy). PAD was diagnosed by highresolution ultrasound screening of femoral tracts (HDI 3500, Advanced Technology Laboratories, USA, which is equipped with a linear multifrequency 5–12 MHz transducer) and Doppler velocimetry (including the treadmill test) and measurement of ankle/ brachial pressure index (ABPI) <0.90 at rest, as measured by Doppler ultrasound (Ultrasound SRL, Italy). Inclusion criteria were: Fontaine stage II monolateral intermittent claudication, male sex, clinical onset of PAD of less than 1 year, low compliance with a physical training programme, a recent lower limb duplex scan (ABPI <0.90 at rest that decreased to at least 0.75 after exercise) and stable maximum walking time of 170–250 s at the last two monthly check-ups (reconfirmed immediately before enrolment) during a standard treadmill test. Exclusion criteria were: female gender so as to avoid genderrelated differences in endothelial response which could confound results, smoking habit because it alters endothelial function, rest pain so as to avoid the confounding effects of antalgic or other pharmaceutical therapy on results, vascular or endovascular surgery for coronary or peripheral arterial disease in the previous 6 months, unstable angina, myocardial infarction or stroke, heart failure, neoplasia, significant liver or renal impairment and inflammatory diseases. All patients provided written informed consent to the study which was approved by the Local Ethics Committee.

The others, were considered controls. Groups were matched for age and medical treatment (see Section 3). This non-randomized clinical trial lasted for 6 months: Months 1 and 2 One OMT session every 2 weeks. Assessments at baseline before starting OMT (see Section 2.1.1) and 2 months after finishing the OMT session. Month 3 Interval to allow assessment of response to OMT techniques, adjustments to techniques as required as well as evaluation of partial response to OMT protocol (Maigne, 1996). Months 4, 5 and 6 One OMT session every 3 weeks. Assessments before starting OMT of month 4 and at end of month 6 after finishing the OMT session. All control patients underwent laboratory and clinical assessments at the same time-points as patients in the OMT group after 30-min rest in the supine position. The study was conducted in two adjacent medical rooms in the hospital unit. OMT treatment was delivered by an osteopath (DO) (who was blinded to the clinical results) and controls rested in the first room. In the second room physicians and nurses who were blinded to the OMT results collected blood samples and assessed vascular parameters at all assessment time-points. OMT patients and controls filled in the self-administered quality of life questionnaire (SF-36) at baseline and after 6 months. 2.1.1. OMT protocol At the start of each OMT session, the DO performed a structural examination of each patient to identify areas of somatic dysfunction, defined as ‘‘impaired or altered function of related components of the somatic (body framework) system; skeletal, arthrodial, and myofascial structures; and related vascular, lymphatic and neural elements.’’ (Rumney, 1971; Lesho, 1999). The somatic dysfunction of each patient was treated using one or more of the following OMT techniques: myofascial release, strain/counterstrain, muscle energy, soft tissue, high-velocity lowamplitude (thoracolumbar region, typically T10–L1), lymphatic pump and craniosacral manipulation. All these techniques were considered safe even in view of the patients’ mean age and associated pathophysiology. Techniques were chosen and listed by the DO at each session which lasted for approximately 30 min. During the study no patient referred to any common side effects except for three who had transient muscle tenderness; no new pathologies developed. At all assessment time-points the following measurements were performed:

2.1. Study design

2.1.2. Blood tests Fasting venous blood samples were taken from all patients before performing the treadmill walking test. Routine lipid parameters, listed in Table 1, were determined immediately after blood collection by standard techniques. Soluble vascular cell adhesion molecule-1 (sVCAM-1), soluble intercellular adhesion molecule-1 (sICAM-1) and interleukin-6 (IL-6) were assessed with commercial ELISA kits (Cytopass – Technogenetics s.r.l., Sesto S.Giovanni, MI, Italy) equipped with a positive control sample of known sVCAM-1, sICAM-1 and IL-6 concentrations. If values of the positive control sample were not within the expected range, the assay results were considered invalid. All assays were performed in duplicate for each sample. The inter-assay coefficients of variations were for sVCAM-1 5.2%, sICAM-1 7.6% and IL-6 5.2%.

In this case-control pilot study 50% of recruited patients were assigned to osteopathic treatment as well as their usual pharmacological therapy (OMT group) according to the patients’ willingness to undergo osteopathic therapy within the study time-frame.

2.1.3. Brachial artery flow-mediated vasodilation (FMV) FMV was assessed by two-dimensional ultrasonography of the brachial artery using a 10 MHz probe (HDI 3500, Advanced Technology Laboratories, USA), as previously reported (Marchesi et al.,

R. Lombardini et al. / Manual Therapy 14 (2009) 439–443 Table 1 Effects of OMT on ABPI and functional performance. Month 2

Month 4

Month 6

OMT patients (n ¼ 15) ABPI Rest 0.78  0.50 After exercise 0.57  0.02 CPT (min) 2.8  0.3 TWT (min) 4.5  0.8

Baseline

0.78  0.04 0.58  0.02 2.9  0.3 4.6  0.9

0.79  0.05 0.58  0.03 2.9  0.4 4.6  0.8

0.87  0.05a 0.79  0.06a 3.7  0.4a 4.7  0.4a

Controls (n ¼ 15) ABPI Rest After exercise CPT (min) TWT (min)

0.76  0.04 0.57  0.04 2.8  0.1 4.4  0.8

0.78  0.02 0.57  0.02 2.9  0.4 4.5  0.6

0.78  0.05 0.57  0.04 2.9  0.3 4.5  0.8

441

and Spearman’s correlation coefficients tested univariate association between variable changes (D) at the study time-points. Our sample size was calculated on the basis of our previous experiences in young healthy subjects and conservatively assuming a 15% of FMV and a 6% SD to detect a difference with an a of 0.05 (Marchesi et al., 2000). p levels <0.05 were considered statistically significant. Data were stored by SPSS statistical package, release 13.0 (SPSS Inc, Chicago, IL, USA). 3. Results

0.76  0.05 0.55  0.02 2.8  0.3 4.4  0.7

ABPI, ankle-brachial pressure index; CPT, claudication pain time; TWT, total walking time. a p < 0.05 vs baseline.

2000). The ultrasonographic procedure conformed to the recommendations of the International Brachial Artery Reactivity Task Force (Corretti et al., 2002). FMV was expressed as the percentage difference between baseline brachial artery diameter and diameter 45–60 s after 4-min upper-arm occlusion by a pneumatic cuff at 230–250 mmHg. Tracings were recorded using a totally automated computerized system (Artery Measurement System AMS, Go¨teborg, Sweden) and read by one investigator who was blinded to the subject’s clinical data and study time-point. The between-occasion intra-observer reproducibility of FMV in our laboratory was assessed in 21 subjects examined 2 days apart. The mean  SD difference between the two examinations was 0.8  0.5%. 2.1.4. ABPI The ABPI is a ratio of ankle to arm systolic blood pressure. Dorsalis pedis and posterior tibial pressures were measured in both ankles, and brachial pressures in both arms using a 4- or 8-MHz Doppler probe. To calculate the ABPI, the highest ankle pressure in each leg was divided by the higher brachial pressure. An ABPI 0.9 was diagnostic of PAD with 95% sensitivity and almost 100% specificity (Yao et al., 1969). A normal ABPI ranges from 0.9 to 1.3. Measurements were performed while the patient was at rest (supine), before and after the treadmill test.

Thirty consecutive male patients (mean age 69  8 years) were recruited to the study. All were non-smokers (12 never smoked, 18 had stopped smoking for at least 1 year) and were under treatment with aspirin (100 mg daily), 24 were also receiving ACE inhibitors and 22 statins. All pharmacological treatment was continued throughout the study. OMT and control groups were matched for age (69  5 years vs 68  7 years) and medical treatment (OMT group 12 were receiving ACE inhibitors and 11 statins versus 12 and 11, respectively in controls). In the control group, no changes were observed in any parameter at any time-point. In the OMT group, significant improvements were observed only after 6 months vs baseline (Table 1). The 15 patients had a significant increase in ABPI, at rest and after exercise CPT and TWT were significantly longer (all p < 0.05) (Table 1). Brachial FMV increased significantly at months 2, 4 and 6 vs baseline. Expression of sICAM, sVCAM and IL-6 were significantly reduced at all time-points vs baseline (all p < 0.05) (Table 2). Questionnaire scores (physical function, role limitations due to physical problems, bodily pain and general health) overlapped in OMT patients and controls at baseline. In the OMT group they were significantly higher at month 6 (p < 0.05 vs baseline; p < 0.05 vs controls month 6). No differences emerged in the mental health component scores at any time-point (Table 3). Univariate analyses showed D changes in brachial FMV at all time-points correlated negatively with changes in IL-6 levels (r ¼ 0.30; p ¼ 0.04) (Fig. 1). D changes in CPT and physical function score between baseline and month 6 correlated positively (r ¼ 0.50, p ¼ 0.05). No other significant correlation emerged. 4. Discussion

2.1.5. Treadmill testing In all patients, times to onset of claudication pain (CPT) when the patient first noticed calf pain during the treadmill test (Hiatt et al., 1990) and until severe claudication pain stopped exercise, i.e. the total walking time (TWT), were measured on the treadmill (Beta Rolling Belt, Reggio Emilia, Italy) at a constant speed of 2.0 km/h, and a 12-degree inclination angle. The treadmill was calibrated before the tests. Walking test data were according to international treadmill protocols (Manfredini et al., 2004). 2.1.6. Health-related quality of life (QoL) Standardized health measures were derived using the Italian version (Apolone et al., 1997), of the Medical Outcomes Study Short Form-36 (SF-36). The self-administered questionnaire, which patients answered after a brief explanation, assesses the physical and mental components of QoL. Answers were transformed using established scoring algorithms to generate standardized health scale scores ranging from 0 to 100, with better health states resulting in higher scores. 2.2. Statistical analysis Data are presented as mean  SD. Student’s t-test was performed to compare parametric variables between groups. Pearson’s

To the best of our knowledge, this is the first study which has attempted to investigate the effect of OMT on endothelial function and functional performance in PAD patients with intermittent claudication. Improvements in brachial FMV, serum inflammatory markers and clinical parameters suggest OMT might be useful in the management of these patients. OMT currently includes a broad spectrum of techniques but does not have any standard application protocols. In our study, DO performed individual treatment and used any technique which appropriately addressed specific somatic dysfunction. OMT, as well as other manual therapies, have been reported to aid circulation, and provide increased blood flow in peripheral vascular tissue; these effects may be partially modulated by nitric oxide (NO) (Backer et al., 2002). Although NO release persists for a short time, it promotes vasodilation, inhibits platelet aggregation, white blood cell adhesion and smooth muscle cell proliferation (Rubanyi, 1993) which may exert profound physiological consequences for a much longer period (Stefano et al., 2000). Salamon et al. (2004) suggested physical manipulation may promote a rapid release of NO by endothelial NO synthase (eNOS) activation (Rizzo et al., 1998; Feron et al., 1998). Indeed, eNOS is enzymatically active in the caveolae and physical manipulation activates eNOS by dissociating it from caveolin, the inhibitory protein, and coupling it

442

R. Lombardini et al. / Manual Therapy 14 (2009) 439–443

Table 2 Effects of OMT on blood tests and endothelial function. Baseline

Month 2

Month 4

Month 6

OMT patients (n ¼ 15) Total cholesterol (mg/dL) Triglycerides (mg/dL) HDL-cholesterol (mg/dL) LDL- cholesterol (mg/dL) Brachial FMV (%) Brachial diameter (mm) sICAM (ng/mL) sVCAM (ng/mL) IL-6 (ng/mL)

198  40 145  128 42  12 127  28 3.6  3.8 3.8  2.1 255  15 1846  298 4.6  0.6

200  37 151  120 43  12 127  27 4.9  3.3a 3.9  3.8 215  15a 1538  203a 3.2  0.4a

204  37 152  128 44  12 129  27 5.2  3.7a 3.9  4.0 214  11a 1546  180a 3.2  0.6a

210  40 156  128 45  13 134  27 5.6  4.2a 3.8  1.8 211  11a 1515  185a 2.9  0.6a

Controls (n ¼ 15) Total cholesterol (mg/dL) Triglycerides (mg/dL) HDL-cholesterol (mg/dL) LDL- cholesterol (mg/dL) Brachial FMV (%) Brachial diameter (mm) sICAM (ng/mL) sVCAM (ng/mL) IL-6 (ng/mL)

198  43 146  78 40  10 127  29 4.6  3.6 3.8  2.1 265  14 1852  235 4.7  0.6

200  34 140  77 42  11 126  22 4.7  3.8 3.7  1.9 258  12 1866  186 5.1  0.7

200  28 143  68 42  10 128  33 4.6  3.7 3.7  2.0 270  11 1820  180 4.9  0.6

197  32 138  82 40  11 129  31 4.8  3.9 3.8  1.8 268  11 1834  185 5.0  0.6

HDL, high-density lipoprotein; LDL, low-density lipoprotein; FMV, flow-mediated vasodilation; sICAM, soluble intercellular adhesion molecule; sVCAM, soluble vascular adhesion molecule; IL-6, human interleukin-6. a p < 0.05 vs baseline.

to calmodulin. Thus mechanical transduction, i.e. physical manipulation, may be linked to eNOS activity. Studies on endothelial dysfunction in PAD focused on two distinct mechanisms: NO dysfunction and imbalance in pro- and anti-inflammatory mechanisms which are evaluated by brachial FMV, soluble adhesion molecules and cytokines (Joannides et al., 1995; Witte et al., 2003). In the present study, OMT gradually improved brachial artery FMV and the improvement reached significance after 6 months. Interestingly, from the second month onwards OMT significantly decreased sICAM-1, sVCAM-1 and IL-6 serum levels providing evidence to support the hypothesis that osteopathic therapy acts via complex NO induced pathways. Functional performance in the present cohort of patients with PAD was assessed by questionnaire scores because intermittent

Table 3 Change in health-related quality of life.

Physical function Baseline Month 6 Role limitations/physical Baseline Month 3 Bodily pain Baseline Month 6 General health Baseline Month 6 Mental health Baseline Month 6 Role limitations/emotional Baseline Month 6 Social function Baseline Month 6 Vitality Baseline Month 6 a b

Score OMT patients (n ¼ 15)

Controls (n ¼ 15)

37.2  3.9 72.8  3.7a,b

39.2  4.6 37.5  4.7

24.7  17.8 60.5  22.6a,b

24.2  18.5 29.3  16.5

0

62.4  17.4 86.5  19.7a,b

67.8  12.4 66.5  15.8

-1

53.5  11.2 67.8  7.6a,b

52.9  10.6 53.2  12.0

71.7  10.4 75.9  9.6

70.6  12.5 73.5  11.3

82.3  11.5 86.4  8.7

80.9  10.3 83.5  11.0

78.9  16.7 82.7  10.4

77.9  9.5 79.0  8.5

60.3  8.7 65.7  10.2

58.9  11.3 60.8  10.6

Significant change from baseline (p < 0.05). Significant difference between groups (p < 0.05).

Δ brachial FMV, %

Components

claudication can interfere with the patient’s QoL by limiting daily physical activities and thus ability to work, personal and social relationships, and independence. To ascertain if the OMT-related improvement in clinical measures was associated with improved perception of QoL, we used the popular SF-36 questionnaire (Beattie et al., 1997; Dumville et al., 2004). After 6 months of OMT, scores indicated a significant improvement in physical health, which was similar to improvements achieved with exercise rehabilitation in PAD patients (Patterson et al., 1997; Menard et al., 2004). OMT was associated with longer walking times in our patients such as were achieved in other cohorts with supervised exercise training programmes (Izquierdo-Porrera et al., 2000, Cheetham et al., 2004). There were differences in duration, frequency and intensity of training. Our OMT programme included one session every 2 weeks for 2 months, and one session every 3 weeks for 3 months and the total therapist contact time per patient was 4 h. Exercise training for patients with PAD usually involved three 45min sessions of intermittent treadmill walking in a supervised environment per week for 20 weeks or more (Bulmer and Coombes, 2004), for a total of 45 h or more therapist–patient

r= -0.30; p= 0.04

-2

-3

-4

-5 0,6

0,8

1,0

1,2

1,4

1,6

1,8

2,0

2,2

2,4

2,6

2,8

Δ IL-6, ng/dL Fig. 1. Correlation between changes in brachial FMV and IL-6 in patients treated with OMT.

R. Lombardini et al. / Manual Therapy 14 (2009) 439–443

contact. From this point of view, our new approach with OMT might offer a different way to increase walking ability in PAD. The OMT programme has the potential advantage to be cost-effective and logistically viable, since frequency and intensity are reduced and OMT does not necessarily need to be performed in a hospital out-patients setting. Enthusiasm must, however, be tempered by the limitations of our findings. First of all, the relatively few patients in our study, that was not a randomized clinical trial comparing the effects of OMT and traditional exercise training. Finally, these results do not take into account the fact that OMT is a holistic approach involving patient–doctor interaction with potential benefits from hands-on contact. Against this, there was no improvement in the mental health component of the QoL questionnaire. In conclusion, despite these limitations, the OMT capacity to regulate vascular endothelial function and improve functional performance in patients with claudication suggests that when combined with other approaches such as lifestyle modification and drug treatment OMT might be an effective alternative to exercise training and certainly warrants further investigation. Acknowledgements The authors would like to thank Dr Geraldine Anne Boyd for editing this paper. References Apolone G, Mosconi P, Ware Jr. JE. Questionario sullo stato di salute SF-36. Manuale d’uso e guida all’interpretazione dei risultati. Guerini e Associati eds, 1997. Backer M, Hammes MG, Valet M, Deppe M, Conrad B, Tolle TR, et al. Different modes of manual acupuncture stimulation differentially modulate cerebral blood flow velocity, arterial blood pressure and heart rate in human subjects. Neuroscience Letters 2002;333:203–6. Bartelink ML, Stoffers HEJH, Biesheuvel CJ, Hoes AW. Walking exercise in patients with intermittent claudication. Experience in routine clinical practice. British Journal of General Practice 2004;54:196–200. Beattie DK, Golledge J, Greenhalgh RM, Davis AH. Quality of life assessment in vascular disease: towards a consensus. European Journal of Vascular and Endovascular Surgery 1997;13:9–13. Brendle DC, Joseph LJO, Corretti MC, Gardner AW, Katzel LI. Effects of exercise rehabilitation on endothelial reactivity in older patients with peripheral arterial disease. American Journal of Cardiology 2001;87:324–9. Brevetti G, Martone VC, De Cristoforo T. High levels of adhesion molecules are associated with impaired endothelium-dependent vasodilation in patients with peripheral arterial disease. Thrombosis and Haemostasis 2001a;85:63–6. Brevetti G, De Caterina M, Martone VD. Exercise increases soluble adhesion molecules ICAM-1 and VCAM-1 in patients with intermittent claudication. Clinical Hemorheology and Microcirculation 2001b;24:193–9. Bulmer AC, Coombes JS. Optimising exercise training in peripheral arterial disease. Sports Medicine 2004;34:983–1003. Burns P, Wilmink T, Fegan C, Bradbury AW. Exercise in claudicants is accompanied by excessive thrombin generation. European Journal of Vascular and Endovascular Surgery 2003;26:150–5. Cheetham DR, Burgess L, Ellis M, Greenhalgh RM, Davies AH. Does supervised exercise offer adjuvant benefit over exercise alone for the treatment of intermittent claudication? A randomised trial. European Journal of Vascular and Endovascular Surgery 2004;27:17–23. Collins TC, Johnson SL, Souchek J. Unsupervised walking therapy and atherosclerotic risk factor management for patients with peripheral arterial disease. Annals of Behavioral Medicine 2007;33:318–24. Corretti MC, Anderson TJ, Benjamin EJ, Celermajer D, Charbonneau F, Creager MA, Deanfield, et al. International Brachial Artery Reactivity Task Force. Guidelines for the ultrasound assessment of endothelial-dependent flow-mediated vasodilation of the brachial artery. A report of the International Brachial Artery Reactivity Task Force. Journal of the American College of Cardiology 2002;39:257–65. De Vries J, De Jongste MJ, Spincemaille G, Staal MJ. Spinal cord stimulation for ischemic heart disease and peripheral vascular disease. Advances and Technical Standards in Neurosurgery 2007;32:63–89. DiGiovanna EL, Marinke DJ, Dowling DJ. Introduction to osteopathic medicine. In: DiGiovanna EL, Schiowitz S, editors. An osteopathic approach to diagnosis and treatment. 2nd ed. New York: Lippincott-Raven; 1997. p. 2–31.

443

Dumville JC, Lee AJ, Smith FB, Fowkes FGR. The health-related quality of life of people with peripheral arterial disease in the community: the Edinburgh Artery Study. British Journal of General Practice 2004;54:826–31. European Working Group on Critical Leg Ischemia. Second European consensus document on chronic critical leg ischaemia. Circulation 1991;84(suppl. IV): IV-1–26. Feron O, Saldana F, Michel JB, Michel T. The endothelial nitric-oxide synthasecaveolin regulatory cycle. Journal of Biological Chemistry 1998;273:3125–8. Fiotti N, Giansante C, Ponte E, Delbello C, Calabrese S, Zacchi T, et al. Atherosclerosis and inflammation. Patterns of cytokine regulation in patients with peripheral vascular disease. Atherosclerosis 1999;145:51–60. Gardner AW, Poehlman ET. Exercise rehabilitation programs for the treatment of claudication pain. A meta-analysis. JAMA 1995;274:975–80. Hiatt WR, Regensteiner JG, Hargarten ME, Wolfel EE, Brass EP. Benefit of exercise conditioning for patients with peripheral arterial disease. Circulation 1990;81:602–9. Hickman P, Harrison DK, Hill A, McLaren M, Tamei H, McCollum PT, et al. Exercise in patients with intermittent claudication results in the generation of oxygen derived free radicals and endothelial damage. Advances in Experimental Medicine and Biology 1994;361:565–70. Izquierdo-Porrera AM, Gardner AW, Powell CC, Katzel LI. Effects of exercise rehabilitation on cardiovascular risk factors in older patients with peripheral arterial occlusive disease. Journal of Vascular Surgery 2000;31:670–7. Joannides R, Haefeli WE, Linder L, Richard V, Bakkali EH, Thuillez C, et al. Nitric oxide is responsible for flow-dependent dilatation of human peripheral conduit arteries in vivo. Circulation 1995;91:1314–9. Lesho EP. An overview of osteopathic medicine. Archives of Family Medicine 1999;8:477–84. Maigne R. Medicina manuale. Diagnosi e terapia per le patologie di origine vertebrale. Ed. UTET, 1996; part V, ch. 1, p. 194–195. Manfredini F, Conconi F, Malagoni AM, Manfredini R, Mascoli F, Liboni A, et al. Speed rather than distance: a novel graded treadmill test to assess claudication. European Journal of Vascular and Endovascular Surgery 2004;28:303–9. Marchesi S, Lupattelli G, Schillaci G, Pirro M, Siepi D, Roscini AR, et al. Impaired flow-mediated vasoactivity during post-prandial phase in young healthy men. Atherosclerosis 2000;153:397–402. Menard JR, Smith HE, Riebe D, Braun CM, Blissmer B, Patterson RB. Long-term results of peripheral arterial disease rehabilitation. Journal of Vascular Surgery 2004;39:1186–92. Patterson RB, Pinto B, Marcus B, Colucci A, Braun T, Roberts M. Value of a supervised exercise program for the therapy of arterial claudication. Journal of Vascular Surgery 1997;25:312–9. Rizzo V, McIntosh DP, Oh P, Schnitzer JE. In situ flow activates endothelial nitric oxide synthase in luminal caveolae of endothelium with rapid caveolin dissociation and calmodulin association. Journal of Biological Chemistry 1998;273:34724–9. Rogers JT, Rogers JC. The role of osteopathic manipulative therapy in the treatment of coronary heart disease. Journal of the American Osteopathic Association 1976;76:21–31. Rubanyi GM. The role of endothelium in cardiovascular homeostasis and diseases. Journal of Cardiovascular Pharmacology 1993;22:S1–14. Rumney IC. Basic terminology for osteopathic procedures. Journal of the American Osteopathic Association 1971;70:1275–88. Salamon E, Zhu W, Stefano GB. Nitric oxide as a possible mechanism for understanding the therapeutic effects of osteopathic manipulative medicine (Review). International Journal of Molecular Medicine 2004;14:443–9. Shepard RJ, Balady GJ. Exercise as cardiovascular therapy. Circulation 1999;99: 963–72. Silvestro A, Scopacasa F, Oliva G, De Cristofaro T, Iuliano L, Brevetti G. Vitamin C prevents endothelial dysfunction induced by acute exercise in patients with intermittent claudication. Atherosclerosis 2002;165:277–83. Spiegel AJ, Capobianco JD, Kruger A, Spinner WD. Osteopathic manipulative medicine in the treatment of hypertension. An alternative, conventional approach. Heart Disease 2003;5:272–8. Stefano GB, Goumon Y, Bilfinger TV, Welters ID, Cadet P. Basal nitric oxide limits immune, nervous and cardiovascular excitation: human endothelia express a mu opiate receptor. Progress in Neurobiology 2000;60:513–30. Still AT. Osteopathy: research and practice. Seattle, WA: Eastland Press; 1992 (print version from the original 1910 edition). Ubbink DT, Vermeulen H, Spincemaille GHJJ, Gersbach PA, Berg P, Amann W. Systematic review and meta-analysis of controlled trials assessing spinal cord stimulation for inoperable critical leg ischaemia. British Journal of Surgery 2004;91:948–55. Weiss T, Fischer D, Hausmann D, Weiss C. Endothelial function in patients with peripheral vascular disease: influence of prostaglandin E1. Prostaglandins, Leukotrienes and Essential Fatty Acids 2002;67:277–81. Witte DR, Broekmans WMR, Kardinaal AFM, Klopping-Ketelaars IAA, van Poppel G, Bots ML, et al. Soluble intercellular adhesion molecule 1 and flow-mediated dilatation are related to the estimated risk of coronary heart disease independently from each other. Atherosclerosis 2003:147–53. Yao ST, Hobbs JT, Irvine WT. Ankle systolic pressure measurements in arterial disease affecting the lower extremities. British Journal of Surgery 1969;56: 676–9.

Manual Therapy 14 (2009) 444–447

Contents lists available at ScienceDirect

Manual Therapy journal homepage: www.elsevier.com/math

Case Report

Long-term effects of a stabilization exercise therapy for chronic low back pain Ney Meziat Filho*, Sonia Santos, Ricardo Mourilhe Rocha ´-Cardı´aco/PROCEP, Rio de Janeiro, RJ, Brazil Clinical Research Post-graduation Program, Hospital Pro

a r t i c l e i n f o Article history: Received 30 July 2007 Received in revised form 8 January 2008 Accepted 2 October 2008 Keywords: Low back pain Exercise therapy Disability Long-term

1. Introduction Low back pain is a common problem which affects the majority of adults at least once in a life time. The majority of them feel relief of symptoms spontaneously. However, a significant minority has recurrences and a small number suffer with persistent symptoms (Dunn and Croft, 2004). According to the experimental model from Hodges and Richardson (1996), individuals with a history of low back pain show a delay in contraction of the transversus abdominis muscle during a trunk disturbance leading to an inappropriate stabilization pattern which causes recurrences. Within 24 hours after the first low back pain episode the lumbar multifidus muscle shows an ipsilateral pain related decrease in muscle bulk and this loss of bulk is not recovered even after the symptoms have resolved (Hides et al., 1994, 1996). Spinal segmental stabilization exercises were developed (Richardson and Jull, 1995) with the aim of correcting the transversus abdominis contraction delay and also to recover the activation of lumbar multifidus muscle. This motor control approach focuses on an isolated cocontraction of such muscles while keeping the lumbar spine in a neutral position (Richardson et al., 2004). Using ultra-sound and magnetic resonance imaging (Ferreira et al., 2004; Hides et al., 2006) it was possible to observe the appropriate pattern of transversus abdominis muscle contraction during the draw-in movement of the lower abdominal wall. In order to test the efficacy of this specific approach some authors undertook randomized controlled trials. Hides et al. (2001) conducted research which demonstrated a long-term * Corresponding author. E-mail address: [email protected] (N. Meziat Filho). 1356-689X/$ – see front matter Ó 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.math.2008.10.002

decrease of low back pain recurrences in patients treated with specific stabilization exercises after the first episode. O’Sullivan et al. (1997) concluded that specific stabilization exercises seem to be more effective than the commonly prescribed conservative treatment programs for chronic low back pain patients with spondylolisis or spondylolisthesis diagnoses. Ferreira et al. (2006) conducted a systematic review of randomized clinical trials which demonstrated the efficacy of such exercises in relation to the chronic pain and recurrences in patients with low back pain and pelvic pain. Other authors (Comerford and Mottram, 2001a; Sahrmann, 2002) defend the training of the global system of stabilization muscles with the aim of correcting the movement impairment. This approach focuses on correcting the movement patterns as well as inappropriate postures and also the lumbopelvic imbalance through therapeutic exercises (Sahrmann, 2002). Van Dillen et al. (2003b) created a system of classification for mechanical low back pain that takes into account not only the lumbar movement but also the limb movement in the reproduction and decrease of the symptoms. These authors concluded that the substitution of a symptomatic movement which reproduces the symptom by another movement with the same aim, yet without reproducing any symptom, is important to decrease the pain and disability. Three case reports utilising this approach where patients received therapeutic exercises according to their impairments demonstrated a decrease in both pain and disability in patients classified as rotation with flexion syndrome, rotation with extension syndrome and extension syndrome respectively (Maluf et al., 2000; Harris-Hayes et al., 2005; Van Dillen et al., 2005). The current case report deals with the assessment, classification, intervention and long-term follow-up of an individual with chronic

N. Meziat Filho et al. / Manual Therapy 14 (2009) 444–447

low back pain. The treatment was based on spinal segmental stabilization exercise combined with therapeutic exercises related to the movement impairment syndrome and the global system dysfunction. 2. History A 41-year-old female patient with a medical diagnosis of degenerative disc disease reported a 15-year history of unilateral low back pain (see Fig. 1). The patient had been a cyclist and a runner. The pain had increased through the years before treatment. Besides the chronic pain the patient had had some acute episodes in the past few years. The average intensity of the symptoms in the week before the assessment was 5 on a numeric pain scale from 0 to 10 (Bolton, 1999). During acute episodes the pain usually reached 10 on the same scale. The patient reported increased pain when in trunk flexion activities such as putting pants on and prolonged sitting. 3. Examination In the physical assessment, the forward bending test (Van Dillen et al., 2003a, 2003b) was performed and increased the symptoms in the beginning of the movement although there was no limitation in the range of lumbar motion. In the corrected forward bending test (Van Dillen et al., 2003a, 2003b), when the patient was taught to perform the same movement but bending via the hips keeping the neutral lordosis, there were no increased symptoms after the test. The slump and straight leg raise (SLR) tests were both negative. The patient was then classified as in flexion syndrome (Van Dillen et al., 2003b). Palpation of the L4 spinous process was painful. Assessing the lumbar segmental stabilizers and lumbopelvic dynamic stabilizers by manual palpatory tests (Richardson and Jull, 1995; Comerford and Mottram, 2001b), the transversus abdominis muscle showed a substitution pattern by oblique abdominis (Richardson et al., 2004). The lumbar multifidus showed an

445

ipsilateral asymmetry to symptoms in the prone position palpatory test (Richardson et al., 2004). Using ultra-sound imaging (Hides et al., 1994), wasting in the lumbar multifidus ipsilateral to the symptoms in acute/sub acute low back pain was confirmed. In the gluteus maximus and hamstring muscle imbalance test (Goecking et al., 2006) the unilateral hip extension in prone position caused an anterior glide of the greater trochanter of the femur and a very delayed and weak contraction of gluteus maximus on both sides. The hamstring bulk was perceived at the beginning of the movement. (Bullock-Saxton, 1994; Bullock-Saxton et al., 1994). Other muscle imbalance tests were negative (Goecking et al., 2006). 4. Intervention and results Regarding the intervention, isolated contraction of the transversus abdominis muscle in supine and in co-contraction with lumbar multifidus in prone as well as an isometric contraction exercise to gluteus maximus in prone were taught (Richardson and Jull, 1995; Sahrmann, 2002; Richardson et al., 2004) ensuring trunk flexion via the hips kept a neutral lumbar lordosis (Sahrmann, 2002). To partially restrict the lumbar flexion in the first 48 hours, tape was placed on the patient’s lower back only at the first visit (Sahrmann, 2002). The patient was instructed to practice the exercises 3 times a day and to send an e-mail or take notes of everything she had done. The exercise program was modified by the physical therapist according to the progress of the patient (see Table 1). Eleven visits were made over a period of 6 months. There was an e-mail follow-up after 1 year and 6 months and another after 2 years and 5 months. At each visit and also in the follow-ups, data was obtained about the pain average and disability of the previous week via a numeric pain rating scale and the Oswestry Disability Index Questionnaire respectively. Magnetic Resonance Imaging (MRI) was done after the last follow-up in order to compare with that done prior to the beginning of the treatment. On the day of the initial assessment, the patient reported that the previous week average pain was 5 on the numeric scale and disability was 58% according to the Oswestry Disability Index Questionnaire. At the second visit, one week later, the pain was rated at 4 and the disability had dropped to 28% which, according to the patient, was due to the training of a new trunk forward bending pattern done with the help of the lumbar tape. It is possible to observe in the graphics the decrease in the pain rating (see Graph 1) and disability over the weeks (see Graph 2). There was only one acute episode which raised the levels of pain and disability and this was during the fourth week of treatment. However, these acute symptoms decreased in less than 2 weeks. After two months of treatment, the patient was almost symptom Table 1 Exercises and duration.

Fig. 1. Body chart illustrating pain presentation.

Exercises

Duration

Transversus abdominis/abdominal drawing in action - supine co-contraction of transverses and multifidus - prone Forward bending from the hips with hand support - standing Isometric gluteus maximus contraction - prone Forward bending from the hips without hand support - standing Hip extension with knee extended - prone Hip/knee extension with shoulder flexion - quadruped position Transversus abdominis/abdominal drawing in action - sitting

Since assessment until last follow-up Since assessment until last follow-up Since assessment until visit 3 Since assessment until visit 4 Since visit 3 until last follow-up Since visit 4 until visit 7 Since visit 7 until last follow-up Since visit 8 until last follow-up

446

N. Meziat Filho et al. / Manual Therapy 14 (2009) 444–447

Graph 1. Numeric Pain Scale over the weeks.

free and this condition lasted up to the last e-mail follow-up done 2 years and 5 months after the beginning of the treatment. By this stage the patient was able to play tennis without worrying about her lumbar spine. During the treatment, the patient undertook the exercises 5 times a week, twice or 3 times a day and in the last visit it was possible to perceive an improvement in transversus abdominis muscle activation, the lumbar multifidus asymmetry correction and also the correction of the gluteus maximus and hamstrings imbalance in their respective tests. In the e-mail follow-up period, the stabilization exercises were done at least 3 times a week. It was also possible to notice the recovery of physiologic lumbar lordosis via MRI (see Figs. 2 and 3), in spite of the evolution of lumbar osteoarthrosis. Fig. 2. Before the treatment.

5. Discussion Despite the study limitation in relation to clinical tests and exercise practice without the use of ultra-sound imaging and electromyography, this case report was important because it individualized the treatment based on an assessment in which it was possible to classify the patient in lumbar flexion syndrome via the forward bending test and the corrected forward bending test (Van Dillen et al., 2003a, 2003b). Although there had been a series of tests to determine the movement impairment syndrome (Sahrmann, 2002; Van Dillen et al., 2003a, 2003b), these tests were selected according to the anamnesis when the patient mentioned an increase of pain in activities of trunk flexion pattern. The approach

Graph 2. Disability over the weeks.

Fig. 3. After the last follow-up.

N. Meziat Filho et al. / Manual Therapy 14 (2009) 444–447

of using the corrected forward bending test as a home exercise associated to the tape application to restrict the lumbar flexion stimulated the patient to exchange the excessive lumbar motion to a safer motion on the hips in her daily activities. This led to the fact that the main pain provoking movement stopped stimulating the chronic low back pain and it was possible to contribute to the improvement of pain as well as the disability even in the long-term. The recovery of lumbar lordosis observed in the MRI after the last follow-up was probably due to the new postural and movement habit changes which kept the lumbar spine more stable near the neutral position. This change was reached without any passive exercise focusing on the lumbar extension. The improvement of the transversus abdominis muscle activity happened a few months after the beginning of the treatment due to the patient’s great difficulty in initially relaxing the external oblique muscle. It is likely that this inappropriate transversus abdominis activation with an external oblique substitution pattern had led to an acute episode at the end of the first month of treatment. However, as the patient achieved the isolated segmental stabilizing muscle control, the levels of pain and disability were kept low even after several months. The correction of the gluteus maximus and hamstring muscle imbalance decreased the excess hamstring activation during the hip extension movement. It is also probable that this change may have contributed to a wider range of forward bending movement from the hip as a result of the decreased hamstrings resistance towards this movement. This case study was undertaken with only one patient and therefore randomized clinical trials with an adequate number of subjects and control group will be necessary in order to demonstrate the efficacy of the stabilization exercises in this clinical classification. 6. Conclusion This case report elicited the long-term benefits of a stabilization exercise therapy for a patient with chronic low back pain. The exercises were selected according to an assessment of the segmental stabilizing and global lumbopelvic muscles and also according to the classification of lumbar flexion syndrome. There was an important decrease in pain and disability which was kept after a long period and the recovery of lumbar lordosis was able to be observed in MRI after the treatment. Acknowledgements This study was carried out as a course conclusion work for the Clinical Research Post-graduation Program from Pro´-Cardı´aco Hospital/PROCEP.

447

References Bolton JE. Accuracy of recall of usual pain intensity in back pain patients. Pain 1999;83(3):533–9. Bullock-Saxton JE. Local sensation changes and altered hip muscle function following severe ankle sprain. Phys Ther 1994;74(1):17–28 [discussion 28–31]. Bullock-Saxton JE, Janda V, Bullock MI. The influence of ankle sprain injury on muscle activation during hip extension. Int J Sports Med 1994;15(6):330–4. Comerford MJ, Mottram SL. Functional stability re-training: principles and strategies for managing mechanical dysfunction. Man Ther 2001;6(1):3–14. Comerford MJ, Mottram SL. Movement and stability dysfunction–contemporary developments. Man Ther 2001;6(1):15–26. Dunn KM, Croft PR. Epidemiology and natural history of low back pain. Eur Med 2004;40(1):9–13. Ferreira PH, Ferreira ML, Hodges PW. Changes in recruitment of the abdominal muscles in people with low back pain: ultrasound measurement of muscle activity. Spine 2004;29(22):2560–6. Ferreira PH, Ferreira ML, Maher CG, Herbert RD, Refshauge K. Specific stabilisation exercise for spinal and pelvic pain: a systematic review. Aust J Physiother 2006;52(2):79–88. Goecking BJL, Fonseca ST, Aquino CF, Silva AA. Confiabilidade de exames fı´sicos para identificaça˜o de desequilı´brios musculares na regia˜o lombope´lvica. Fisioterapia em Movimento 2006;19(2):57–66. Harris-Hayes M, Van Dillen LR, Sahrmann SA. Classification, treatment and outcomes of a patient with lumbar extension syndrome. Physiother Theory Pract 2005;21(3):181–96. Hides J, Wilson S, Stanton W, McMahon S, Keto H, McMahon K, et al. An MRI investigation into the function of the transversus abdominis muscle during ‘drawing-in’ of the abdominal wall. Spine 2006;31(6):E175–8. Hides JA, Jull GA, Richardson CA. Long-term effects of specific stabilizing exercises for first-episode low back pain. Spine 2001;26(11):E243–8. Hides JA, Richardson CA, Jull GA. Multifidus muscle recovery is not automatic after resolution of acute, first-episode low back pain. Spine 1996;21(23):2763–9. Hides JA, Stokes MJ, Saide M, Jull GA, Cooper DH. Evidence of lumbar multifidus muscle wasting ipsilateral to symptoms in patients with acute/subacute low back pain. Spine 1994;19(2):165–72. Hodges PW, Richardson CA. Inefficient muscular stabilization of the lumbar spine associated with low back pain. A motor control evaluation of transversus abdominis. Spine 1996;21(22):2640–50. Maluf KS, Sahrmann SA, Van Dillen LR. Use of a classification system to guide nonsurgical management of a patient with chronic low back pain. Phys Ther 2000;80(11):1097–111. O’Sullivan PB, Twomey LT, Allison GT. Evaluation of specific stabilizing exercise in the treatment of chronic low back pain with radiologic diagnosis of spondylolysis or spondylolisthesis. Spine 1997;22(24):2959–67. Richardson C, Hodges PW, Hides J. Therapeutic exercise for lumbopelvic stabilization: a motor control approach for the treatment and prevention of low back pain. Edinburgh; New York: Churchill Livingstone; 2004. Richardson CA, Jull GA. Muscle control-pain control. What exercises would you prescribe? Man Ther 1995;1(1):2–10. Sahrmann S. Diagnosis and treatment of movement impairment syndromes. St. Louis, Mo.; London: Mosby; 2002. Van Dillen LR, Sahrmann SA, Norton BJ, Caldwell CA, McDonnell MK, Bloom N. The effect of modifying patient-preferred spinal movement and alignment during symptom testing in patients with low back pain: a preliminary report. Arch Phys Med Rehabil 2003;84(3):313–22. Van Dillen LR, Sahrmann SA, Norton BJ, Caldwell CA, McDonnell MK, Bloom NJ. Movement system impairment-based categories for low back pain: stage 1 validation. J Orthop Sports Phys Ther 2003;33(3):126–42. Van Dillen LR, Sahrmann SA, Wagner JM. Classification, intervention, and outcomes for a person with lumbar rotation with flexion syndrome. Phys Ther 2005;85(4):336–51.

Manual Therapy 14 (2009) 448–451

Contents lists available at ScienceDirect

Manual Therapy journal homepage: www.elsevier.com/math

Case Report

Treatment of upper cervical subluxation in pediatric patients Sen-Wei Tsai a, b, c, *, Jin-Deng Zhong a, Yan-Wen Chen a, Shyi-Kuen Wu c, Yi-Wen Lin d a

Department of Physical Medicine and Rehabilitation, Taichung Veterans General Hospital, No.160, Chungkang Rd., Taichung 407, Taiwan Department of Life Science, National Chung-Hsing University, No.250, Guoguang Rd., South District, Taichung City 402, Taiwan c Department of Physical Therapy, HungKuang University, No.34, Chung-Chie Rd., Salu, Taichung County 433, Taiwan d Department of Internal Medicine, Gaomay Clinics, 5F., No.94, Jhongshan Rd., Cingshuei Township, Taichung County 436, Taiwan b

a r t i c l e i n f o Article history: Received 21 September 2008 Received in revised form 31 December 2008 Accepted 6 January 2009 Keywords: Manual therapy Atlantoaxial rotatory subluxation

1. Introduction Sudden onset of acute torticollis is a rare condition in children and is usually diagnosed as atlantoaxial rotatory subluxation (Wurm et al., 2004; Tsai and Chou, 2005; Tonomura et al., 2007; Weigel et al., 2007). It is usually symptomatic, but without neurologic disturbances. Treatment procedures include medication such as anti-inflammatory drugs (NSAIDs), stretch or mobilization technique of physiotherapy, hard collar immobilization, or cervical traction in a halo cast (Grieve, 1988; Subach et al., 1998; Park et al., 2005). Surgery such as C1-C2 posterior fusion is usually used as a final treatment procedure especially in patients with longstanding subluxation (Parisini et al., 2003; Rahimi et al., 2003; Park et al., 2005). Rarely are conservative treatment options such as manual therapy reported in the literature (Tsai and Chou, 2005). The mechanism that causes atlantoaxial rotatory subluxation is not clearly understood. Trauma, a history of previous upper respiratory infection, intra-articular derangement, or simple muscle spasm have been described in the etiology of this condition (Grieve, 1988; Dhaon et al., 2003; Al Kaissi et al., 2007). Long-standing cases may have chronic ligamentous changes that cause recurrence of the subluxation and require surgical fusion to restore alignment (Lin et al., 1995; Harma and Firat, 2008). Two cases with AARS and one

* Correspondence author. Department of Physical Medicine and Rehabilitation, Taichung Veterans General Hospital, No.160, Chungkang Rd., Taichung 407, Taiwan. Tel.: þ886 4 23592525x3506; fax: þ886 4 230129615. E-mail address: [email protected] (S.-W. Tsai). 1356-689X/$ – see front matter Ó 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.math.2009.01.002

case with C2-3 subluxation which were all treated successfully with manual therapy are presented. 2. Case 1 One week ago, an 11-year-old girl had a sore throat, headache and neck stiffness in the early morning when she woke up. She was brought to her family physician later where a physical examination showed left neck swelling with the head deviated to the left side and an upper respiratory tract infection. Although hot pack and medications (NSAIDs) were prescribed, the neck stiffness persisted. She was sent to the hospital. On clinical examination her neck was tilted 30 degrees to the left side with 10 degrees of left side rotation. Active neck motion was limited in all directions. There was no neurological sign, but pain when she tried to rotate her head, and tried to extend her head. The cervical side bending and side gliding motions from C3-4 to C6-7 were normal. On palpation, a protective muscle spasm on the right sternocleidomastoid muscle and left neck lymphadenopathy was found. Radiographs (open mouth view, lateral and posterior–anterior(PA)) showed that her neck was in a tilted to the left position (see Fig. 1). Flexion and extension view were taken for evaluation of the neck motion. There was a C2-3 anterior subluxation when neck was in flexion position (Fig. 2), and improved when neck was in extension position. Three-dimensional CT scan (3D CT) was arranged, and it showed that there was no bony fracture, but a right C2-3 facet joint subluxation.

S.-W. Tsai et al. / Manual Therapy 14 (2009) 448–451

449

Fig. 1. Case 1. Open mouth radiographic view. The head is fixed in left side-tilted position without atlantoaxial deformity.

Treatment: A slightly modified Saunders’s upper cervical rotation maneuver was used (Saunders, 1995). The patient lay supine on the treatment table and her pelvis was held by an assistant. With both hands cradled to hold the patient’s head and neck, and both thenar eminence areas in contact with the occiput, an axial traction force was applied to the occiput. With the right index finger in contact with the right C3 pillar, a small amplitude C2-3 right rotation mobilization (grade IV mobilization) was performed. The mobilization session was about 1–2 min and repeated several times. After each session, the patient was asked to sit up and the active and passive neck motion assessed. The treatment procedure continued for about 15 min until she could hold her head in an upright position. The PROM was full but there was still some pain sensation at end range of passive right rotation. By the next day, she had full recovery and was discharged. Three days later, she came back to the clinic for follow-up. Physical examination showed that there was no limitation in active and passive neck motion, and no pain was complained of. X-ray for posterior–anterior, lateral, and lateral flexion-extension was done and there was no abnormal finding (Fig. 3). There was no recurrence after six months’ follow-up. 3. Case 2 A 8-year-old boy had suffered for 2 weeks from a neck fixed in a right-rotated position. Medicines (NSAIDs) and a hot pack were prescribed by his physician, but in vain. Due to the persistent neck pain and torticollis, he was sent to the pediatric emergency department for evaluation. Physical examination showed there was no neurological defect but limitation of neck motion in all directions. The head was fixed in a right-rotated position of 45 degrees. Palpation found no lymphadenopathy but there was protective guarding of the right sternocleidomastoid muscle. Radiographic

Fig. 2. Case 1. Cervical lateral flexion radiographic view. C2-3 anterior subluxation was noted when neck in flexion position.

examination was undertaken and showed the head was fixed in right-rotated position (Fig. 4). A computer tomography scan was arranged and left side C1-2 atlantoaxial rotatory subluxation was observed. The manual therapy procedure was slightly modified from that described by Schneider et al. (1988) and Tsai and Chou (2005). The patient lay supine with his pelvis held by an assistant. The therapist’s left hand cradled the patient’s occiput while a sustained axial traction force was applied. The right index finger was fixed on the right C1-C2 articular pillar, a gentle rotary mobilisation was applied to the left side for 30–60 s. Hold–relax and contract–relax stretching of the right neck muscles were done after this technique and prior to checking the range of motion in sitting. This process was repeated several times over approximately 15 min until the young boy could hold his head in an upright position and actively rotate 45 degrees to the left side. Two days later the range of motion was full and painfree. There had been no recurrence when he was reviewed at a 1 year follow-up. 4. Case 3 A 10-year-old boy had acute torticollis and neck stiffness after he suddenly rotated his head back to say hello to his friend. One hour after the onset he was sent to the clinic. There was no history of previous upper respiratory tract infection or lymphadenopathy. Clinical assessment showed his head was fixed in a left sidetilted position of 25 degrees with his chin inclined to the right. Only

450

S.-W. Tsai et al. / Manual Therapy 14 (2009) 448–451

The manual therapy procedure was as reported in Case 2. The patient was positioned as before and a sustained axial traction force was applied by the thenar eminences with manual contact to the occiput. At the same time, with the left index finger fixed on the left C2 pillar, a gentle left rotary force was applied to the right C1-2 for about 30 s. After each session the passive range of motion for C01-2 was checked. After the left C1-2 rotation was regained, a right rotation force was then applied to the C1-2 joint. The procedure lasted approximately 20 min. After this procedure, the head could be held in an upright position although there was still some pain and a 10 degree limitation active right rotation. By the next day the boy had achieved full recovery and there had been no recurrence when reviewed at the six months’ follow-up. 5. Discussion

15 degrees of active flexion was retained. There were no neurological signs. A palpable protective spasm was found in the lateral posterior aspect of the left C1-2 paraspinal muscle. Several radiographs were taken (see Fig. 5 open mouth view) and the diagnosis of left side atlantoaxial rotatory subluxation (AARS) was made.

Treatment methods for AARS include conservative managements such as NSAIDs, physiotherapy, hard collar, or more invasive methods such as cervical halo or surgical fixation. Manual therapy is a widely used method for spinal pain management but there are few reports in the literature that focus on manual therapy for cervical subluxation. Furthermore, there were no reports about how quickly the symptoms of AARS can be relieved. Previously the authors (Tsai and Chou, 2005) described four cases with AARS successfully managed under anesthesia with manual therapy. This time, the cases with AARS and C2-3 subluxation were successfully managed only by manual therapy. Some special aspects of children’s spines may predispose them to atlantoaxial rotatory subluxation. Fesmire and Luten (1989) have described the developmental anatomy of children and pointed out

Fig. 4. Case 2. Posterior-anterior radiographic view. The head was fixed in a rightrotated position.

Fig. 5. Case 3. Open mouth radiographic view. The head is fixed in a left side-tilted position with atlantoaxial subluxation.

Fig. 3. Case 1. Cervical lateral flexion radiographic view 2 days after treatment. There was no limitation when compared to Fig. 2.

S.-W. Tsai et al. / Manual Therapy 14 (2009) 448–451

that their ligaments and joint capsules possess sufficient elasticity to allow hypermobility without disruption. Kawabe and associates concluded that the dens-facet angle of the axis was steeper in children than in adults, and meniscus-like synovial folds were found in the C0/1 and C1/2 facet joints of the spines of children but not in those of adults (Kawabe et al., 1989; Mercer and Bogduk, 1993). In Lin et al.’s study (1995), in the supine position, they showed that a 5 kg traction force increased both the intervertebral foraminal area and disc height. In the three cases presented, the main components of the treatment procedure are cervical axial traction force and mobilization of subluxed joints. The axial maneuver force may have the effect to make some separation of the cervical joints, which could relieve the symptoms. In these three cases muscle spasm was found. According to Grieve (1988), radiological examination is preferable before manipulation to exclude organic disease. However, Maitland (2001) states that ‘‘It is a cardinal rule that movements must never be forcibly thrust through protective spasm’’ and for that reason, grade V or thrust techniques on these patients was prohibited. Other absolute contradictions on cervical spine manipulation such as bone disease, inflammatory arthritis, positive Lhermitte’s sign etc. should be carefully considered before manipulation. In adult patients, a provocative test for vertebro-basilar insufficiency(VBI) is usually done before cervical spine manipulation (APA, 1988). In the three cases reported, the neck rotation and extension test is not possible due to the locked neck position. To avoid any adverse effect of cervical spine treatment, muscle palpation, X-ray or even CT scan is the pre-treatment screening protocol to see if there are any inflammation or bony abnormalities. Instead of the conventional static radiographic imaging for the cervical spine, studies (Bouillot et al., 1999; McGuire et al., 2002; Gradl et al., 2005; Been et al., 2007) show that three-dimensional CT scan (3D CT) may offer a more accurate image for diagnosis of atlantoaxial rotatory subluxation. However, in Case 1, the lateral flexion and extension view, especially the flexion view, provided a more accurate image for diagnosis of C2-3 anterior subluxation than in 3D CT, which only showed right C2-3 subluxaton. The reason for this is that when the CT was done, the patient was always put in a supine position, which mimics the position when taking an extension view. It is suggested that when AARS is suspected, dynamic cervical flexion and extension lateral views should be done. In this report, the three cases all got improvement immediately after manual therapy and achieved complete recovery within two days. A cervical orthosis was not prescribed. Although cervical collar is usually recommended for AARS, (Subach et al., 1998; Rahimi et al., 2003) it may need further investigation to disclose it’s effect if any on AARS. Physicians agree that once atlantoaxial rotatory subluxation has been diagnosed, treatment should be undertaken as soon as possible. Subach et al. (1998) concluded that there would be no recurrence if reduction were achieved within 21 days of symptoms. The length of time before reduction seems to correlate with the likelihood of recurrence and failure of closed reduction techniques. In this report, the three cases were seen within 14 days, which may predict a better prognosis.

451

6. Conclusion A mobilization maneuver with traction force may be an effective and quick management procedure for cases of pediatric upper cervical subluxation. Open mouth radiographic view, dynamic flexion and extension cervical views are important when interpreting upper cervical problems.

References Al Kaissi A, Zwettler E, Roetzer KM, Haller J, Varga F, Klaushofer K, et al. Vertebral hyperostosis, ankylosed vertebral fracture and atlantoaxial rotatory subluxation in an elderly patient with a history of infantile idiopathic scoliosis; a case report. J Med Case Reports 2007;1:25. Australian Physiotherapy Association (APA). Protocol for premanipulative testing of the cervical spine. Aust. J. Physiother. 1988;34:97–100. Been HD, Kerkhoffs GM, Maas M. Suspected atlantoaxial rotatory fixation-subluxation: the value of multidetector computed tomography scanning under general anesthesia. Spine 2007;32(5):E163–7. Bouillot P, Fuentes S, Dufour H, Manera L, Grisoli F. Imaging features in combined atlantoaxial and occipitoatlantal rotatory subluxation: a rare entity. Case report. J Neurosurg 1999;90(2 Suppl.):258–60. Dhaon BK, Jaiswal A, Nigam V, Jain V. Atlantoaxial rotatory fixation secondary to tuberculosis of occiput: a case report. Spine 2003;28(11):E203–5. Fesmire FM, Luten RC. The pediatric cervical spine: developmental anatomy and clinical aspects. J Emerg Med 1989;7(2):133–42. Gradl G, Maier-Bosse T, Penning R, Stabler A. Quantification of C2 cervical spine rotatory fixation by X-ray, MRI and CT. Eur Radiol 2005;15(2):376–82. Grieve GP. Common vertebral joint problems. 2nd ed. Edinburgh: Churchill Livingstone; 1988 [chapter 7], p. 343–5, 367 and 648. Harma A, Firat Y. Grisel syndrome: nontraumatic atlantoaxial rotatory subluxation. J Craniofac Surg 2008;19(4):1119–21. Kawabe N, Hirotani H, Tanaka O. Pathomechanism of atlantoaxial rotatory fixation in children. J Pediatr Orthop 1989;9(5):569–74. Lin CJ, Lin RM, Wu MH. Atlantoaxial rotatory instability secondary to odontoid hypoplasia as a cause of acute torticollis in children: report of one case. Zhonghua Min Guo Xiao Er Ke Yi Xue Hui Za Zhi 1995;36(6):448–51. Maitland GD. Maitland’s vertebral manipulation. 6th ed. Butterworth-Heinemann; 2001 [chapter 9], p. 220. Mercer S, Bogduk N. Intra-articular inclusions of the cervical synovial joints. Br J Rheumatol 1993;32(8):705–10. McGuire KJ, Silber J, Flynn JM, Levine M, Dormans JP. Torticollis in children: can dynamic computed tomography help determine severity and treatment. J Pediatr Orthop 2002;22(6):766–70. Parisini P, Di Silvestre M, Greggi T, Bianchi G. C1–C2 posterior fusion in growing patients: long-term follow-up. Spine 2003;28(6):566–72. discussion 572. Park SW, Cho KH, Shin YS, Kim SH, Ahn YH, Cho KG, et al. Successful reduction for a pediatric chronic atlantoaxial rotatory fixation (Grisel syndrome) with longterm halter traction: case report. Spine 2005;30(15):E444–9. Rahimi SY, Stevens EA, Yeh DJ, Flannery AM, Choudhri HF, Lee MR. Treatment of atlantoaxial instability in pediatric patients. Neurosurg Focus 2003;15(6). ECP1. Saunders D. Evaluation, treatment and prevention of musculoskeletal disorders. In: Spine. 3rd ed., vol. 1. Educational Opportunities, A Saunders Group Company; 1995 [chapter 4]51–53. Schneider W, Dvorak J, Dvorak V, et al. Manual medicine: therapy. Kreuzlingen and Berne, Switzerland: Thieme; 1988 [chapter 5], p. 28–36. Subach BR, McLaughlin MR, Albright AL, Pollack IF. Current management of pediatric atlantoaxial rotatory subluxation. Spine 1998;23(20):2174–9. Tonomura Y, Kataoka H, Sugie K, Hirabayashi H, Nakase H, Ueno S. Atlantoaxial rotatory subluxation associated with cervical dystonia. Spine 2007;32(19):E561–4. Tsai SW, Chou CS. A case report of manipulation under anesthesia of posttraumatic type II occipital-atlantoaxial rotatory subluxation in a 4-year-old girl. J Manipulative Physiol Ther 2005;28(5):352–5. Weigel RM, Capelle HH, Krauss JK. Cervical dystonia in Bechterev disease resulting in atlantoaxial rotatory subluxation and cranio-cervical osseous fusion. Spine 2007;32(25):E781–4. Wurm G, Aichholzer M, Nussbaumer K. Acquired torticollis due to Grisel’s syndrome: case report and follow-up of non-traumatic atlantoaxial rotatory subluxation. Neuropediatrics 2004;35(2):134–8.

Manual Therapy 14 (2009) 452–455

Contents lists available at ScienceDirect

Manual Therapy journal homepage: www.elsevier.com/math

Technical and Measurement Report

Stability and intra-tester reliability of an in vivo measurement of thoracic axial rotation using an innovative methodology Nicola R. Heneghan a, *, Alison Hall b, Mark Hollands a, George M. Balanos a a b

School of Sport and Exercise Sciences, University of Birmingham, Birmingham, B15 2TT, UK Medical Ultrasound, Mid Staffs NHS Trust, UK

a r t i c l e i n f o

a b s t r a c t

Article history: Received 1 May 2008 Received in revised form 19 August 2008 Accepted 9 October 2008

The aim of this study was to evaluate the stability and intra-tester reliability of an innovative approach to measure active thoracic spine axial rotation. Ultrasound imaging of a thoracic vertebra in conjunction with Polhemus motion analysis of the transducer was used to measure axial thoracic spine rotation in a functional position. The range of motion in a convenience sample of asymptomatic subjects (n ¼ 24) was calculated across ten repetitions of a single trial to evaluate stability. The protocol was repeated the same day and 7–10 days later to provide data for within and between day intra-tester reliability. Mean total range of axial rotation was 85.15 degrees across a single trial with SD ¼ 14.8, CV ¼ 17.4, SEM ¼ 3.04. SEM ranged 0.63–3.37 for individual subjects and 2.60–3.64 across repetitions. Stability of performance occurred at repetitions 2–4. Intra-tester reliability (ICC2,1) was excellent within day (0.89–0.98) and good/excellent between days (0.72 0.94). Bland–Altman plots however suggest that agreement may range from 0 to 10% for within day measures and from 0 to 15% for between day measures. Whether this combined approach has sufficient precision and accuracy as a clinical research tool has yet to be fully evaluated. Ó 2008 Elsevier Ltd. All rights reserved.

Keywords: Thoracic spine Axial rotation Reliability Motion analysis

1. Introduction There are a range of measurement tools available to evaluate motion in the thoracic spine, although many have not been investigated for their reliability. This compromises their suitability for use in evaluating the effectiveness of clinical interventions on thoracic spine motion. It is however widely accepted that the use of reliable outcomes measures as a means of evaluation is fundamental to evidence based practice. Ultrasound imaging is widely used as a clinical and research tool because it is relatively inexpensive and safe. Positional rotation of vertebrae in subjects with idiopathic scoliosis has been measured using ultrasound imaging (Suzuki et al., 1989; Burwell et al., 1999; Kirby et al., 1999). Using an inclinometer attached to the ultrasound transducer with the subject in prone position, a measure in degrees of the vertebral position relative to the horizontal plane was acquired using an image of the laminae of each level in the spine (representative of vertebra position) (Suzuki et al., 1989; Burwell et al., 1999; Kirby et al., 1999). Use of ultrasound enhanced the sensitivity and specificity for the detection of scoliosis by 16% and 23% respectively (Burwell et al., 1999). Furthermore, measures of

* Corresponding author. Tel.: þ11 (0)121 4148739. E-mail address: [email protected] (N.R. Heneghan). 1356-689X/$ – see front matter Ó 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.math.2008.10.004

laminar rotation obtained correlated significantly at 95% confidence with the vertebral rotation obtained using x-ray, despite different test positions being used. Vertebral rotation using this approach was shown to be measured to within 3.1 (Kirby et al., 1999). Stability considers how consistent a tool is at producing a result whilst measuring the same entity on repeated occasions (Sim and Wright, 2000). Owing to the viscoelastic properties of tissues (stress relaxation and hysteresis) range of motion may change with increasing repetitions. Once stability is established, intra-tester reliability may be investigated. Within and between day intratester reliability provide an indication of how useful a method is in detecting change in motion following clinical interventions. The purpose of this study was to evaluate the stability of data over repeated measures of the axial rotation in the thoracic spine and the intra-tester reliability (within and between days) using ultrasound imaging combined with Polhemus motion analysis in a seated position. 2. Materials and methods A prospective test–retest design combined an evaluation of stability with a within and between day intra-tester reliability. A convenience sample of asymptomatic subjects (n ¼ 24) was recruited based on a power calculation where n  13 for a 5%

N.R. Heneghan et al. / Manual Therapy 14 (2009) 452–455

significance level with reliability (ICC2,1) powered at >0.7 (Walter et al., 1998). The sample included 9 males, 15 females, age range 18– 32 years with a mean (SD) age of 24.96 years (2.6), weight 70.8 kg (14.35), height 170.2 cm (8.7). Ethical approval was obtained from the School of Sports and Exercise Sciences, University of Birmingham. Subjects were excluded if they had a current or previous neuromusculoskeletal spine condition, pre-existing systemic rheumatological condition, undergone abdominal surgery, were pregnant, or affected by a current or chronic respiratory condition. 2.1. Equipment Ultrasound images were acquired using a Phillips Sonos 5500, a 26 mm linear array transducer and frequency range of 3–11 MHz. Polhemus (LibertyÔ) (Colchester, VT, USA) was used for the motion analysis. This system records sensor position and movement relative to a source transmitter with 6 degrees of freedom. The static accuracy is 0.03 inch in RMS for the x, y or z position and 0.15 degrees RMS for sensor orientation (Polhemus Specifications, 2007). The sensor was fixed via a plastic extended arm to the transducer to avoid interference and the source transmitter was mounted in front of the subjects (Fig. 1). 2.2. Procedure Subjects were seated in a standardized position with their lumbar spine in the neutral position, legs fully supported and arms in mid abduction (Fig. 1). A fully adjustable wooden bar was positioned at the level of L1 to minimize movement occurring in the lumbar spine. This experimental set up was used to standardise thoracic spine posture across repetitions and trials, as posture has been shown to influence thoracic motion (Edmonston et al., 2006).

453

The spinous process of the C7 vertebra was palpated by an experienced physiotherapist in neutral and the skin marked. An ultrasound image of the T1 spinal laminae was acquired in the horizontal plane using reference lines on the monitor (Fig. 2) (Kirby et al., 1999). The coordinate position of the transducer was then recorded. The subjects actively moved to a position of maximum axial rotation and a ‘new’ image of T1 laminae acquired, then the ‘new’ transducer position was recorded. The minimal acceptable criteria for each image were that the C7 spinous process and T1 laminar had to be clearly visible and consistent on each occasion with respect to their position on the monitor. This procedure was done sequentially in neutral, right rotation, neutral, left rotation for ten repetitions. Although data was captured for transducer movement about the x- and z-axes, this study only used data from the y-axes to calculate thoracic axial rotation. The transducer was removed from the skin following each data point to avoid any influence on the subjects’ motion. Expertise in image acquisition was required to minimize time spent and stress relaxation at end-range positions. Each subject attended on two occasions, trials 1 and 2 taking place on the first occasion with subjects getting up between trials, and trial 3 taking place 7–10 days later with environmental and diurnal variables being controlled for. The researcher was blinded to the results of each trial with data analysis occurring on completion of trial 3. Accuracy of the Polhemus system was evaluated using a ‘mock’ spine (Koerhuis et al., 2003). A rod with the transducer fixed at the top of the unit was mounted on a stand. The transducer with sensor attached was then axially rotated (across a 180 degree range), including varying positions of tilt (up to 15 degrees) to simulate out of plane motion across 120 trials. Accuracy was calculated using means and standard deviation. 2.3. Data analysis The individual and group data were analyzed to derive the range of rotation to the left and to the right from the neutral position. A composite measure of axial rotation for each repetition was then calculated from the raw data and used for subsequent analysis. All data analysis was performed using SPSS version 14.00, where p < 0.05. To assess the stability of measures, the data from trial 1 was used, with the individual and group means being analyzed descriptively across the ten repetitions. Stability was analyzed using means for accuracy, standard deviation (SD) for precision, standard error of the mean (SEM) as a measure of the sampling

Fig. 1. Experimental set up.

Fig. 2. Ultrasound image of spinal laminae.

454

N.R. Heneghan et al. / Manual Therapy 14 (2009) 452–455

error, and the coefficient of variation (CV) to calculate the variability of repeated measures relative to the mean (Sim and Wright, 2000). In order to derive a measure for subsequent inferential analysis of reliability, repeated measures one-way ANOVA on successive triads of repetitions across trial 1 was used (repetitions 1–3, 2–4, 3–5, 4–6 etc.). The triad of data where there was least variability within the trial data set was analyzed using a confidence interval of 95%. Intra-tester reliability analysis using intra-class correlation coefficients (ICC2,1) from a repeated measures ANOVA test were calculated using 95% confidence intervals (CI) to determine the within day (trials 1 and 2) and between day (trials 1 and 3) reliability. Reliability is deemed to be ‘good’ where values range from 0.61 to 0.80 and ‘excellent’ for values between 0.81 and 1.00 (Shrout, 1998). Limits of agreement analysis (95%) were derived using Bland–Altman plots for trials 1 and 2 (within day) and trials 1 and 3 (between day) (Bland and Altman 1986). Repeatability analysis across all three trials was performed using repeated measures ANOVA on the mean value of the triad with the least variability from trial 1 and the middle value from this derived triad.

2.6. Reliability Using group mean data from repetitions 2–4 for each trial, the reliability (ICC2,1) was shown to be ‘excellent’ for within day measures (0.89–0.98) and ‘good/excellent’ for between days measures (0.72–0.94). Furthermore analysis of reliability (ICC2,1) using only the value from the third repetition was also ‘excellent’ (0.80–0.96) and ‘good/excellent’ (0.76–0.95) for within and between days respectively. Bland–Altman plots illustrate at 95% CI agreement for within and between day reliability, trials 1 and 2, and trials 1 and 3, with the mean value from the 2–4 repetition for each trial (Fig. 4). 2.7. Repeatability across trials The repeatability across all trials using the mean value for repetitions 2–4 or the third value was shown to be good (repeated measures ANOVA, p > 0.9). The mean values were 84.5, 83.8 and 84.0 degrees for trials 1, 2 and 3, respectively using the mean of 2–4, and 84.5, 83.8, 83.7 degrees for trials 1, 2 and 3, respectively using the value from the third repetition. 3. Discussion

2.4. Results The accuracy of the measurement system using the ‘mock’ spine was calculated to be 1.72 degrees, SD 0.52 over a 180degree range. The mean composite range of axial rotation was 85.15 degrees across a single trial (SD ¼ 14.8, SEM ¼ 3.04, CV ¼ 17.4).

The findings suggest that this innovative approach provides a stable and reliable method for the measurement of thoracic rotation in a functional seated position. The basis for the development of this technique primarily relates to the need for reliable evaluation of the effectiveness of clinical interventions in this relatively under-researched region of the spine. Surface markers/

2.5. Stability The data was normally distributed across trial 1, using a Kolmogrov–Smirnov test (p > 0.05). Fig. 3 illustrates mean values (SEM) for each successive repetition across trial 1. Although between subject variability is considerable (mean SEM ¼ 3.23), within subject variability was small across the ten repetitions (mean SEM ¼ 1.70). Statistical analysis was performed on the data to determine the triad with the least variability across trial one. Repeated measures ANOVA for repetitions 1–3, 2–4, 3–5, 4–6, etc. showed that repetitions 2–4 had the smallest effect size and least variability (partial Eta squared 0.005 at p ¼ 0.95).

Fig. 3. The mean value for the group for each successive repetition across trial 1.

Fig. 4. Bland–Altman plots for within and between day comparisons.

N.R. Heneghan et al. / Manual Therapy 14 (2009) 452–455

sensors, the preferred approach for motion analysis of this region in vivo (Willems and Jull, 1996; Theodoridis and Ruston, 2002; Edmonston et al., 2006), possibly lack accuracy and reliability due to relative movement between the sensor, skin and bone, despite reporting good stability (Willems and Jull, 1996; Edmonston et al., 2006). The stability of the measurements appeared relatively constant, despite some considerable variation between subjects. This may in part be due to the use of a standardised sitting posture resulting in some individuals performing axial rotation away from their ‘normal’ sitting posture. The experimental set up was used to ensure that subjects’ thoracic posture was consistent across repetitions and trials. Motion in other planes was recorded, however in this instance data was only analysed to determine axial rotation. Although minimal time was spent at the extremes of the range of motion stress relaxation may account for the slight trend for increase in range across each trial. Hysteresis may explain the lack of cumulative increase in range from trial 1 to trial 2 with tissues having a chance to ‘recover’ between trials. Whilst intra-tester reliability was found to be ‘excellent’ and ‘good to excellent’ for within and between day measures respectively, using the data from either the third repetition or the mean of the 2–4 repetitions, some caution should be taken when interpreting the results alongside the Bland–Altman plots. The Bland– Altman plots suggest that the difference between paired measures is up to 10% and 15% for within and between day agreements of measures respectively. Given that the mean range of motion was 85 degrees, this suggests that there could be an error as large as 8–10 degrees size for some subjects. Visual inspection of the graphs however suggests that the approach may be better suited to within day measures, where the percentage difference for the majority of subjects is less than 5%. Further research with a larger sample is indicated to explore the nature and extent of sources of error with this approach and perhaps using images from more than one spinal level. This is the first study that has utilised ultrasound imaging of the spine in dynamic and functional motion analysis as a means of ensuring that the start and end body positions/postures are truly representative of the underlying bony anatomy. As image acquisition of sufficient quality is operator dependent and the equipment reasonably expensive, there is currently little prospect for this as a clinical practice measurement tool. However, as our current understanding of biomechanics and effects of interventions in this region is considerably underdeveloped compared to other areas of the spine the need to consider alternative non-invasive measurement approaches remains. Future research could, with sufficient imaging expertise, explore regional

455

or segmental motion analysis using data from all three coordinates (x, y, z) or vertebral coupling which has in this region been debated. 4. Conclusion Ultrasound imaging combined with a motion analysis system has been shown to be a reliable method as a measure of active thoracic axial rotation in a seated position. Although the intratester reliability was shown to be ‘‘good to excellent’’ further work is indicated to explore the use of this approach in the evaluation biomechanics and clinical interventions in this region. Acknowledgments Alison Rushton, School of Health Sciences, University of Birmingham, Doug Carroll, School of Sport and Exercise Sciences, Chris Wright, School of Health Sciences, University of Birmingham. The Manipulative Association of Chartered Physiotherapists Doctoral Award 2007. References Bland CJ, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986;327:307–10. Burwell RG, Kirby AS, Aujla RK, Krik EL, Pratt RK, Bailey MA, et al. Evaluation of vertebral rotation by ultrasound for the early detection of adolescent idiopathic scoliosis. In: Research into Spinal Deformities 2. Health Technology & Information 59. Amsterdam: IOS Press; 1999. Edmonston S, Aggerholm M, Elfving S, Flores N, Ng C, Smith R, et al. Influence of posture on the range of axial rotation and coupled lateral flexion of the thoracic spine. Journal of Manipulative and Physiological Therapeutics 2006;30(3): 193–9. Koerhuis CL, Winters JC, van der Helm FCT, Hof AL. Neck mobility measurement by means of the Flock of Birds electromagnetic tracking system. Clinical Biomechanics 2003;18:14–8. Kirby A, Burwell R, Cole A, Pratt R, Webb J, Moulton A. Evaluation of a new real-time ultrasound method for measuring segmental rotation of vertebrae and ribs in scoliosis. In: Research into Spinal Deformities 2. Health Technology & Information 59. Amsterdam: IOS Press; 1999. Polhemus Specifications. Available at http://www.polhemus.com/?page¼Motion_ Liberty (accessed November 22nd, 2007). Sim J, Wright C. (2000) Research in Health Care. Concepts, Designs, and Methods. Gloucester, UK: Stanley Thornes. Shrout PE. Measurement reliability and agreement in psychiatry. Statistical Methods in Medical Research 1998;7:301–17. Suzuki S, Yamamuro T, Shikata J, Shimizu K, Iida H. (1989) Ultrasound measurement of vertebral rotation in idiopathic scoliosis. Journal of Bone and Joint Surgery 71-B(2):252–5. Theodoridis D, Ruston S. The effect of shoulder movements on thoracic spine 3D motion. Clinical Biomechanics 2002;17:418–21. Walter SD, Eliasziw M, Donner A. Sample size and optimal designs for reliability studies. Statistics in Medicine 1998;17:101–10. Willems JM, Jull GA. An in vivo study of the primary and coupled rotations of the thoracic spine. Clinical Biomechanics 1996;11(6):311–6.

Manual Therapy 14 (2009) 456–459

Contents lists available at ScienceDirect

Manual Therapy journal homepage: www.elsevier.com/math

Professional Issue

Bibliometrics, impact factors and manual therapy: Balancing the science and the art Derek R. Smith*, Darren A. Rivett School of Health Sciences, Faculty of Health, The University of Newcastle, Australia

a r t i c l e i n f o Article history: Received 8 November 2008 Accepted 28 November 2008 Keywords: Impact factors Publishing Manual therapy Physiotherapy Bibliometrics Rehabilitation

Bibliometrics can be defined as a field of research that examines bodies of knowledge both within and across disciplines (Holden et al., 2005). Although many methods are commonly used, perhaps the most widely known is citation analysis; that is, tracking published articles to see whether they are subsequently cited by others (Smith, 2008a). Much of contemporary bibliometrics can be traced back to a seminal publication known as Shepard’s Citations, a tool first used by American lawyers in 1873 to establish whether a previous legal judgment had been referred to, overruled, or made invalid in some other way (Adair, 1955). By the early 20th century, citation analysis had attracted the attention of various scientific scholars, although most of their early work simply involved the counting and sorting of reference lists. Nevertheless, some trends were noticed early on in the journals of chemistry (Gross and Gross, 1927), engineering (Bradford, 1934) and physiology (Brodman, 1944). Perhaps the most striking observation was that not all journals were being equally cited; rather, only a few core periodicals appeared to be attracting the majority of all citations. On the other hand, while larger journals tended to gather more citations than smaller ones, some of the smaller periodicals still appeared to be performing well, relative to their actual size and circulation.

* Correspondence to: Derek R. Smith, WorkCover New South Wales Research Centre of Excellence, School of Health Sciences, Faculty of Health, University of Newcastle, Ourimbah, New South Wales 2258, Australia. Tel.: þ61 2 4348 4021; fax: þ61 2 4348 4013. E-mail address: [email protected] (D.R. Smith). 1356-689X/$ – see front matter Ó 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.math.2008.11.004

With this phenomenon in mind, an information scientist named Eugene Garfield proposed calculating a journal’s relative ‘impact’ in 1955 (Garfield, 1955), whereby the number of citations received by a particular journal in a particular time period could be divided by the number of articles it had actually published during that time. The concept was refined with experience to include only ‘citable items’ in the calculation, that is, substantial types of articles that were most likely to be cited by others (Garfield, 1986). Garfield then founded a company known as the Institute for Scientific Information (ISI), and began publishing impact factors in the yearly Journal Citation ReportsÒ (JCRÒ) during the early 1970s (Smith, 2008c). Data from ISI-listed journals was initially collated in the Science Citation IndexÒ (SCIÒ), although by 1972 the scope of material had expanded to also encompass the social sciences, which led to formation of the Social Sciences Citation IndexÒ (SSCIÒ) (Garfield, 1972). Garfield’s idea caught on over time, the ISI was acquired by Thomson Scientific (later Thomson Reuters), and impact factors now represent a powerful influence in the world of modern publishing (Smith, 2006). Given that the impact factor calculation is fundamentally based on citation counts, citations themselves have now risen to become the ‘currency’ of modern scientific research (Joseph, 2003). Impact factors tend to change over time and are generally believed to be increasing in recent years. Such trends have been quantitatively demonstrated in both the larger medical journals (Chew et al., 2007) and also in some of the smaller health (Smith, 2008b) and medical (Boldt et al., 2000) sub-disciplines. There are a few potential reasons for this phenomenon. Firstly and fundamentally, although the basic calculation itself has not changed for

D.R. Smith, D.A. Rivett / Manual Therapy 14 (2009) 456–459

over 50 years (i.e. citations received divided by articles published), the actual number of citations being made each year is steadily increasing. This is probably due to various factors, such as an increased use of automatic referencing software, thereby making it easier to include a larger number of references in a journal article than before (Smith and Hazelton, 2008). This may reflect an increasing tendency for authors to cite their peers whenever possible, and thus increase the chances of their article being accepted. There is also increasing pressure on authors to include more references per article to help demonstrate a more comprehensive understanding of the topic. Secondly, journal impact factors are now being increasingly influenced by deliberate editorial practices in recent years, as more and more journals learn to play the ‘impact factor game’ (Tse, 2008). Citation analysis represents a subset of bibliometric research with which most researchers and academics are becoming increasingly familiar. Although relevancy of a published article is highly desirable, exactly how well it will achieve this goal is not predictable at the time of publication (Balon, 2005). Given time, however, citation trends will emerge and many scholars now consult the various electronic databases to see where and when their articles have been cited, and by whom. There are three main types of scientific article; those that present data, those that teach, and those which analyze, speculate or comment (D’Auria, 1999). Literature reviews tend to attract more citations than original articles, which themselves tend to attract more citations than editorials or letters. Even so, all forms of academic scholarship are important for clinicians in manual therapy, as evidenced in this journal. Although clinical experience may suggest that various kinds of physical therapies are worthwhile, evaluations still need to be based on research findings, especially empirical clinical research (Michels, 1982). Academic publishing is not just for scientists. Indeed, ten years ago D’Auria suggested that scholarship itself is the ‘intellectual counterpart of manual dexterity in clinical work’ (D’Auria, 1999, p. 277). Some of the earliest citation analysis in physical therapy journals appears to have been conducted around 25 years ago. Although it was not citation analysis as such, in 1982 Michels explored the issue of research evaluation in physical therapy, asserting the importance of a sound research base for clinical practice (Michels, 1982). In one of the first bibliometric investigations in our field, Dean and Davies (Dean and Davies, 1986) investigated the frequency of citations combined with a ‘Reputational Assessment’ of contributors in physical therapy. In their article, the authors performed a citation analysis of the journals Physical Therapy and Physiotherapy Canada between 1981 and 1982, concluding that the perception of therapists regarding the impact of eminent individuals was comparable to ratings of those individuals made by objective measures, such as citation analysis. Five individuals who were nominated as being ‘eminent’ in the profession also appeared regularly in the citation lists of these two journals. In the same year, 1986, Bohannon and Gibson published their analysis of journals cited in Physical Therapy, finding that Physical Therapy itself was the most highly-cited periodical, followed by the Archives of Physical Medicine and Rehabilitation (Bohannon and Gibson, 1986). Interestingly, the second ranked journal on the list had attracted less than half the number of citations as the first. In 1987, Bohannon proposed one of the first ‘core’ lists of physiotherapy journals by examining citation frequency in Physical Therapy, Physiotherapy, Physiotherapy Canada and Physiotherapy Practice (Bohannon, 1987). In 1989, Bohannon and Tiberio examined the medical index coverage of journals cited in physiotherapy periodicals, finding that it was neither complete nor consistent. The following year, 1990, an article looking at information accessing

457

behavior of physical therapists was published by Bohannon, who established that while books and journals were certainly being read, physical therapists also utilized patient protocols, medical communications, course notes and materials from company representatives (Bohannon, 1990). In 1991, another citation analysis was conducted by Bohannon and Roberts to help establish a ‘core list’ of rehabilitation journals, during which the authors found that information relating to rehabilitation was actually being published across a large number of different journals (Bohannon and Roberts, 1991). In 1992, Roberts conducted a study of the journal literature and its quality in relation to physiotherapy. His results suggested that while Medline was an excellent source of supplementary material relevant to physiotherapy, coverage was not complete and other information sources also needed to be consulted (Roberts, 1992b). In the following year, Roberts then looked at the coverage of core journals in rehabilitation and related topics by various online databases. In his study, it was revealed that although the number of core journals was very large, their actual coverage by information services was still very selective (Roberts, 1992a). In the same year, Kuhlemeier published a bibliometric analysis of the Archives of Physical Medicine and Rehabilitation, reporting that although it sat near the top of impact factor rankings for rehabilitation periodicals, its score was lower than for most medical journals (Kuhlemeier, 1992). In 1995, Tesio and colleagues investigated, from a bibliometric and citationist perspective, the drive of the neurology profession toward rehabilitation. In their study, the authors found that rehabilitation literature suffered from having a relatively small number of articles published, having a greater proportion of the literature being published in journals without impact factors, and having lower impact factors even when published in ISI-listed periodicals (Tesio et al., 1995). In a 1997 article, the literature of physical therapy was ‘mapped’ by examining citations in two established journals, Physical Therapy and the Archives of Physical Medicine and Rehabilitation (Wakiji, 1997). A skewed distribution of citations was clearly demonstrated, with only 14 journals being responsible for one-third of all references, whereas the next one-third came from 95 other journals. In 1999, Bohannon proposed another list of core physiotherapy journals established by means of citation analysis. In his study, the author looked at 5534 citations from 973 journals, finding that 48 journals had been cited 20 times or more in the time period 1997–1998. Over half of all citations received by this core group were from only 10 journals with the highest number of citations (Bohannon, 1999). In 2001, the issue of impact factors and their relationship with rehabilitation journals was explored by Lankhorst and Franchignoni, with the authors finding that Clinical Rehabilitation, was placed second in impact factor rankings among journals specifically dedicated to rehabilitation medicine (Lankhorst and Franchignoni, 2001). Although citation analysis has not yet been performed for Manual Therapy, a recent editorial pointed out that Masterclasses themselves now represent some of the most popular downloads (Beeton, 2008). Future bibliometric analysis of Manual Therapy would certainly be interesting to conduct, if only to establish whether article download trends in manual therapy mirror citation behavior. In the title of this paper we have called for a balance between the science and the art in manual therapy, and there are a few reasons why such an approach is necessary. Firstly, given the current obsession with journal impact factors, it is often forgotten that this measure contains various intrinsic limitations, and citation analysis itself is by no means perfect (Smith, 2008c). While the JCRÒ is known to be useful for those in the field of physical therapy (Bohannon, 1986), the impact factor calculation has not changed since it was first invented, and the two year ‘citation window’ may not be appropriate for every research

458

D.R. Smith, D.A. Rivett / Manual Therapy 14 (2009) 456–459

field, particularly some of the smaller sub-disciplines with longer publication lag-times (Smith, 2007). The impact factor itself may be due for an overhaul, perhaps as part of a general reflection and debate on where citation indexing is heading in future. From a social science perspective, Holden and colleagues have also suggested that ‘any research area worthy of investigation needs to have its methods continuously and critically reviewed’ (Holden et al., 2005, p. 4). Either way, most critics agree that impact factors tell only part of the story and the future use of alternative measures such as article download counts and internet-based journal sessions may offer more tangible alternatives (Favaloro, 2008). Secondly, there is the issue of journal coverage. While bibliographic databases clearly provide a means of assimilating common threads of ideas and data (Ebrahim, 2006), this can only be achieved if journals relevant to our field are actually being included. Recent years have also witnessed a large expansion of the various complementary disciplines in health care, although less than half of the studies in this area tend to be published in journals with impact factors (Raschetti et al., 2005). Rehabilitation research is published in a large number of different journals (Bohannon and Roberts, 1991), and not all will be included in citation-tracking databases. Not all physiotherapy journals have impact factors either. The longstanding journal Physiotherapy for example, was not ISI-listed until 2005 and thus did not receive an impact factor score until 2007 (Harms, 2006). As such it can be suggested that the publication of modern physiotherapy research may be a relatively new phenomenon in the scientific literature. The broader field of rehabilitation itself has a shorter tradition than some of the other medical disciplines, and as such, fewer human and financial resources are probably being dedicated to it (Tesio et al., 1995). Thirdly, there is the issue of ‘crossover’ given the fact that authors in multidisciplinary fields often have many different choices as to where their findings may be published. Manual therapy is well established as a multidisciplinary field and this is reflected in journal readership and article authorship. With regard to Manual Therapy journal itself, while most readers and authors are physiotherapists working in the field of manual therapy, the journal has also published work written by physicians, scientists, chiropractors and osteopaths (Beeton, 2008). Being interdisciplinary is not easy however, as individuals need to be familiar with other disciplines, and this often takes a great deal of time and effort (Lynch, 2006). Fourthly, there is the philosophical issue of why we conduct research in the first place. Achieving a professional balance between the science and the art of manual therapy is clearly important for the practitioner, and this point also needs to be remembered when publishing. Blind obsession with journal performance indicators is not always good for a journal and its readers, and it has been suggested that having a purely ‘impact factor centered approach’ can easily lead to a situation where everything practical, readable and entertaining is cut, in favor of material that will be cited (Smith, 2006). A periodical can easily fall into the trap of focusing more on those who might cite it, rather than those who will actually read it. On the other hand, journals still need to attract quality submissions on a regular basis, which is where having a high impact factor can be very useful. It is important for editors and readers to keep this balance in mind. Finally, and perhaps most critically, it is important to remember that while publication usually stems from research, becoming purely focused on research can be detrimental for the individual practitioner, as clinical skills might easily be forgotten. Although he was referring to dermatology, in a 1999 editorial that is equally relevant to modern manual therapy, Marks suggested that there

will always be a need to keep a balance between the art and science ‘lest we rely too much on the modern reductionist approach to defining clinical skills and rely too little on the lessons learned from history on the value of the bedside’ (Marks, 1999, p. 344). References Adair WC. Citation indexes for scientific literature? American Documentation 1955;6:31–2. Balon R. Reflections on relevance: psychotherapy and Psychosomatics in 2004. Psychotherapy and Psychosomatics 2005;74(1):3–9. Beeton K. Masterclass editorial. Manual Therapy 2008;13(5):373–4. Bohannon R. Core journals of physiotherapy. Physiotherapy Practice 1987;3(3): 126–8. Bohannon RW. Journal citation reports. Physical Therapy 1986;66(8):1275. Bohannon RW. Information accessing behaviour of physical therapists. Physiotherapy Theory and Practice 1990;6(4):215–25. Bohannon RW. Core journals of physiotherapy. Physiotherapy 1999;85(6):317–21. Bohannon RW, Gibson DF. Citation analysis of physical therapy. A special communication. Physical Therapy 1986;66(4):540–1. Bohannon RW, Roberts D. Core journals of rehabilitation: identification through index analysis. International Journal of Rehabilitation Research 1991;14(4): 333–6. Boldt J, Haisch G, Maleck WH. Changes in the impact factor of anesthesia/critical care journals within the past 10 years. Acta Anaesthesiologica Scandinavica 2000;44(7):842–9. Bradford SC. Sources of information on specific subjects. Engineering 1934;137:85–6. Brodman E. Methods of choosing physiology journals. Bulletin of the Medical Library Association 1944;32:479–83. Chew M, Villanueva EV, van der Weyden MB. Life and times of the impact factor: retrospective analysis of trends for seven medical journals (1994–2005) and their Editors’ views. Journal of the Royal Society of Medicine 2007;100(3):142– 50. D’Auria D. Occupational Medicine, publishing and the new millenium. Occupational Medicine (London) 1999;49(5):277. Dean E, Davies J. Frequency of citation and reputational assessment of contributors in physical therapy. Physical Therapy 1986;66(6):961–6. Ebrahim S. Entelechy, citation indexes, and the association of ideas. International Journal of Epidemiology 2006;35(5):1117–8. Favaloro EJ. Measuring the quality of journals and journal articles: the impact factor tells but a portion of the story. Seminars in Thrombosis and Hemostasis 2008;34(1):7–25. Garfield E. Citation indexes for science; a new dimension in documentation through association of ideas. Science 1955;122(3159):108–11. Garfield E. The new social sciences citation index (SSCI) will add a new dimension to research on man and society. Essays of an Information Scientist 1972;1: 317–9. Garfield E. Which medical journals have the greatest impact? Annals of Internal Medicine 1986;105(2):313–20. Gross PL, Gross EM. College libraries and chemical education. Science 1927;66(1713):385–9. Harms M. Impact factors. Physiotherapy 2006;92:73–4. Holden G, Rosenberg G, Barker K. Tracing thought through time and space: a selective review of bibliometrics in social work. Social Work in Health Care 2005;41(3–4):1–34. Joseph KS. Quality of impact factors of general medical journals. British Medical Journal 2003;326(7383):283. Kuhlemeier KV. A bibliometric analysis of the archives of physical medicine and rehabilitation. Archives of Physical Medicine and Rehabilitation 1992;73(2):126–32. Lankhorst GJ, Franchignoni F. The impact factor – an explanation and its application to rehabilitation journals. Clinical Rehabilitation 2001;15(2):115–8. Lynch J. It’s not easy being interdisciplinary. International Journal of Epidemiology 2006;35(5):1119–22. Marks R. The art, the science, and the practice of dermatology in the next millennium. International Journal of Dermatology 1999;38(5):343–4. Michels E. Evaluation and research in physical therapy. Physical Therapy 1982;62(6):828–34. Raschetti R, Menniti-Ippolito F, Forcella E, Bianchi C. Complementary and alternative medicine in the scientific literature. Journal of Alternative and Complementary Medicine 2005;11(1):209–12. Roberts D. Coverage by four information services of the core journals of rehabilitation and related topics. Scandinavian Journal of Rehabilitation Medicine 1992a;24(4):167–73. Roberts D. The journal literature of physiotherapy: quality through peer review and access through MEDLINE. Physiotherapy 1992b;78(1):29–33. Smith DR. Historical development of the journal impact factor and its relevance for occupational health. Industrial Health 2007;45(6):730–42. Smith DR. Bibliometrics, dermatology and contact dermatitis. Contact Dermatitis 2008a;59(3):133–6. Smith DR. Citation analysis and impact factor trends of 5 core journals in occupational medicine, 1985–2006. Archives of Environmental & Occupational Health 2008b;63(3):114–22.

D.R. Smith, D.A. Rivett / Manual Therapy 14 (2009) 456–459 Smith DR. Citation indexing and the development of academic journals in tropical medicine. Memorias do Instituto Oswaldo Cruz 2008c;103(3):310–2. Smith DR, Hazelton M. Bibliometrics, citation indexing, and the journals of nursing. Nursing & Health Sciences 2008;10(4):260–5. Smith R. Commentary: the power of the unrelenting impact factor – is it a force for good or harm? International Journal of Epidemiology 2006;35(5):1129–30.

459

Tesio L, Gamba C, Capelli A, Franchignoni FP. Rehabilitation: the Cinderella of neurological research? A bibliometric study. Italian Journal of Neurological Sciences 1995;16(7):473–7. Tse H. A possible way out of the impact-factor game. Nature 2008;454(7207):938–9. Wakiji EM. Mapping the literature of physical therapy. Bulletin of the Medical Library Association 1997;85(3):284–8.

Available online at www.sciencedirect.com

Manual Therapy 14 (2009) 460 www.elsevier.com/math

Diary of events NZMPA biennial scientific conference, Heritage Hotel, Rotorua, New Zealand 28, 29 & 30 August 2009. The theme is ‘Striving for Excellence in OMT’ & also celebrating 40 years of Manual Therapy in New Zealand. The conference co-coordinator is Vicki Reid, Phone 0800 646 000 or 09 476 5353 Fax 09 476 5354 e-mail: [email protected] Website: www.nzmpa.org.nz

NOI International conference UK and Ireland Nottingham UK e April 15e17, 2010 Dublin IRELAND April 21e23, 2010 For further details www.noi2010.com Fax þ 3906 51882443

APA Conference Week, Sydney Convention Centre, Sydney, Australia 1-5 October 2009 For more information please visit www.apaconferenceweek09. asn.au

Janet G. Travell, MD Seminar Series, Bethesda, USA For information, contact: Myopain Seminars, 7830 Old Georgetown Road, Suite C-15, Bethesda, MD 20814-2432, USA. Tel.: þ1 301 656 0220; Fax: þ1 301 654 0333; website: www.painpoints.com/seminars.htm E-mail: [email protected]

Aaron Mattes 4 Day Active Isolated Stretching & Strengthening Seminar The Renaissance Hotel, Heathrow 15e18 October 2009 First ever UK Seminar. Further details regarding the contents of the seminar can be found at www.stretchingusa.com & registration details can be found at www.stretchinggb.com. Phone: 020 8897 0377 / 07984 005366. Email: [email protected]

The Diary of events will soon be moving to http://www.elsevier. com/math (click on the journal news tab).

1356-689X/$ - see front matter doi:10.1016/S1356-689X(09)00074-5

MUSCULOSKELETAL DISORDERS IN PRIMARY CARE RESEARCH CONGRESS 2010 Rotterdam 11-13 October 2010 For further information visit www.bsl.nl/primarycare

If you wish to advertise a course/conference, please contact: Karen Beeton, Associate Head of School (Professional Development), School of Health and Emergency Professions, University of Hertfordshire, College Lane, Hatfield, Herts AL10 9AB, UK. E-mail: [email protected] There is no charge for this service.

Manual Therapy 14 (2009) e1

Contents lists available at ScienceDirect

Manual Therapy journal homepage: www.elsevier.com/math

Letter to the Editor

Scha¨fer A, Hall T, Briffab K. Classification of low back-related leg pain – A proposed patho-mechanism-based approach. Manual Therapy (2007) doi:10.1016/j.math.2007.10.003

Dear Editor We welcome the contribution to the debate about low backrelated leg pain, it’s origins and a possible means of sub-classification by Schafer et al. (2009) Whilst they raise many important and salient points, we do have some concerns about the model they propose and the evidence used to support it. Their classification of central sensitisation is characterised (see Figure 1) by a score of 12 or more on the Leeds Assessment of Neuropathic Symptoms and Signs (LANSS) a scale designed to identify pain of a probable neuropathic origin in a variety of neurological conditions (Bennett, 2001). A more appropriate tool to assess low back-related leg pain may be a questionnaire recently validated for low back and leg pain, painDETECT (Freynhagen et al., 2006). With regard to the interpretation of these and other neuropathic pain scales (Bennett et al., 2007), they have been designed to distinguish between neuropathic and non-neuropathic pain, and do not have the sensitivity or selectivity to distinguish between neuropathic pain of peripheral or central origin. Secondly, the proposal that Paraesthesia is a unique distinguisher between central and peripheral neuropathic pain lacks supporting evidence. Whilst Paraesthesia may well result from changes in central sensitivity it is commonly associated with irritation of peripheral nerves, nerve roots, and dorsal root ganglion (Devor, 2006). Furthermore, vascular compromise and the resultant ischaemia is a potent causative factor in the development of Paraesthesia, which occurs more commonly in peripheral nerves (Lundborg, 2005). Thirdly these authors propose that distal pain is a central phenomena and pain anywhere in the leg is peripherally mediated (Table 2). Such a distinction has not been demonstrated or reported. Finally it is proposed that Denervation is identified by ‘‘negative’’ symptoms. In fact positive symptoms including Paraesthesia are often present (a) in the first 6–12 weeks when Wallerian degeneration is an active process (Lundborg, 2005) and (b) during the recovery stages (tingling is the positive finding of the Tinel’s test for a recovering median nerve (Stewart and Eisen, 1978)). The algorithm summarising Schafer et al’s proposed model (Figure 1) is attractive and could easily be adopted in clinical practice. For the above reasons we have reservations about the model’s applicability, as the symptoms and signs used are not unique to any of the three neuropathic groupings suggested. It is our opinion that

1356-689X/$ – see front matter Ó 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.math.2008.07.001

these inaccuracies are likely to lead to errors in clinical reasoning and at least some inappropriate management. Whilst the authors acknowledge that there is ‘‘significant overlap and causal interrelation between the different mechanisms’’ they suggest it is ‘‘feasible that there is a predominant mechanism primarily responsible for a patient’s complaints’’ which the proposed algorithm may help to detect. What little we currently do know about neuropathic pain and its management does not justify (a) the proposed groups included in this algorithm, or (b) the idea of a predominant mechanism. The suggested groupings of nervous system anatomy and physiological function/dysfunction may become valid as our knowledge increases, albeit with different underlying reasoning. The current model, however, is premature and may lead to errors in reasoning, and inappropriate management of this debilitating situation. References Bennett M. The LANSS pain scale: the Leeds assessment of neuropathic symptoms and signs. Pain 2001;92(1–2):147–57. Bennett MI, Attal N, Backonja MM, Baron R, Bouhassira D, Freynhagen R, et al. Using screening tools to identify neuropathic pain. Pain 2007;127(3):199–203. Devor M. Centralization, central sensitization and neuropathic pain. Focus on sciatic chronic constriction injury produces cell-type-specific changes in the electrophysiological properties of rat substantia gelatinosa neurons. J Neurophysiol 2006;96:522–3. Freynhagen R, Baron R, Gockel U, Tolle TR. painDETECT: a new screening questionnaire to identify neuropathic components in patients with back pain. Curr Med Res Opin 2006;22(10):1911–20. Lundborg G. Nerve and nerve injuries. 2nd ed. Oxford: Elsevier, 2005. Schafer A, Hall T, Briffa K. Classification of low back-related leg pain – a proposed pathomechanism-based approach. Manual Therapy 2009;14(2):222–30. Stewart JD, Eisen A. Tinel’s sign and the carpal tunnel syndrome. BMJ 1978;2(6145): 1125–6.

Iain Beith, PhD, MSc, MCSP, DipTP (Cert Ed)* Mick Thacker, MSc, MMACP, MCSP King’s College London, Division of Applied Biomedical Research, School of Biomedical & Health Sciences, Guy’s Campus, London SE1 1UL, UK  Corresponding author. E-mail address: [email protected] (I. Beith) 5 February 2008

Manual Therapy 14 (2009) e2

Contents lists available at ScienceDirect

Manual Therapy journal homepage: www.elsevier.com/math

Letter to the Editor

Letter to Editor Response: Scha¨fer A, Hall T, Briffa NK. Classification of low back-related leg pain – A proposed patho-mechanism-based approach. Manual Therapy (2007)

Dear Editor, We would like to thank Beith and Thacker for their interest in our paper (Scha¨fer et al., 2009). We cannot agree that the presentation of the classification system is premature but recognize it may need to be refined as evaluation studies progress. It is well recognized that a different treatment approach is required for patients with a predominant neuropathic pain component. For example the presence of sensory hypersensitivity, which is commonly associated with neuropathic pain, has been linked to a poor response to physiotherapy intervention (Sterling et al., 2003; Jull et al., 2007). For the purposes of our classification system, we selected a score of 12 on the Leeds Assessment of Neuropathic Symptoms and Signs (LANSS) scale as the criterion for the first level. This scale has demonstrated validity and reliability to identify neuropathic pain (Bennett, 2001) and has been used to detect neuropathic components in patients with low back-related leg pain (Kaki et al., 2005). Items within the LANSS scale are primarily concerned with identifying positive features of neuropathic pain arising from a mix of peripheral and central mechanisms. As there is ample evidence to support the role of central mechanisms in neuropathic pain (Woolf and Mannion, 1999; Jensen and Baron, 2003) we elected to use the term Central Sensitization when referring to the group with positive LANSS scores. At the time our classification system was developed, the painDETECT screening questionnaire was not yet published (Freynhagen et al., 2006). It is of interest, but not surprising, that the LANSS scale and the painDETECT questionnaires incorporate similar items. A strength of our classification system is that none of the subgroups are defined by a single clinical feature. In Table 2 we listed features that can be used clinically to align patients with categories but we did not mean to suggest that any individual features were unique to one diagnostic category. Although the classification system is based on the premise there will be a predominant underlying mechanism, the potential for overlap and causal interrelation between the different mechanisms acting in each patient was emphasized. Moreover, the hierarchical nature of the algorithm (Figure 1) does not preclude patients with features of a lower level classification being classed at a higher level. Beith and Thacker suggest the current understanding of pain mechanisms does not justify the concept of a predominant mechanism, yet it is widely acknowledged that neuropathic pain can rarely be attributed to a single mechanism (Finnerup et al., 2005). 1356-689X/$ – see front matter Ó 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.math.2008.07.002

Where mixed mechanisms may be contributing, clinical judgment is required to establish the predominant mechanism (Bennett et al., 2006) to provide a foundation for clinical reasoning and treatment. We acknowledge the model may appear rudimentary in the complex framework of events occurring with nerve injury. If it is accepted that the predominant pathomechanisms active in each group differ, it follows that the classification system for the optimal treatment approach for each group is also likely to differ. Further evaluation will determine whether this form of classification has clinical utility and in time, we anticipate refinement of this classification system will occur.

References Bennett M. The LANSS pain scale: the Leeds assessment of neuropathic symptoms and signs. Pain 2001;92:147–57. Bennett MI, Smith BH, Torrance N, Lee AJ. Can pain can be more or less neuropathic? Comparison of symptom assessment tools with ratings of certainty by clinicians. Pain 2006;122:289–94. Finnerup NB, Otto M, McQuay HJ, Jensen TS, Sindrup SH. Algorithm for neuropathic pain treatment: an evidence based proposal. Pain 2005;118:289–305. Freynhagen R, Baron R, Gockel U, To¨lle TR. painDETECT: a new screening questionnaire to identify neuropathic components in patients with back pain. Curr Med Res Opin 2006;22:1911–20. Jensen TS, Baron R. Translation of symptoms and signs into mechanisms in neuropathic pain. Pain 2003;102:1–8. Jull G, Sterling M, Kenardy J, Beller E. Does the presence of sensory hypersensitivity influence outcomes of physical rehabilitation for chronic whiplash? A preliminary RCT. Pain 2007;129:28–34. Kaki AM, El-Yaski AZ, Youseif E. Identifying neuropathic pain among patients with chronic low-back pain: use of the Leeds assessment of neuropathic symptoms and signs pain scale. Reg Anesth Pain Med 2005;30:422–8. Scha¨fer A, Hall T, Briffa K. Classification of low back related leg pain – a proposed pathomechanism based approach. Manual Therapy 2009;14(10):222–33. Sterling M, Jull G, Vicenzino B, Kenardy J. Sensory hypersensitivity occurs soon after whiplash injury and is associated with poor recovery. Pain 2003;104:509–17. Woolf CJ, Mannion RJ. Neuropathic pain: aetiology, symptoms, mechanisms, and management. Lancet 1999;353:1959–64.

Axel Scha¨fer* Toby Hall Kathy Briffa Ruckenzentrum am Michel, Ludwig Erhart Strasse 18, 20459 Hamburg, Germany  Corresponding author. Tel.: þ49 40 43280274. E-mail address: [email protected] (A. Scha¨fer) 22 July 2008

Manual Therapy 14 (2009) e3

Contents lists available at ScienceDirect

Manual Therapy journal homepage: www.elsevier.com/math

Book Review Whiplash, Headache, and Neck Pain: Research Based Directions for Physical Therapies, G. Jull, M. Sterling, D. Falla, J. Treleaven, S. O’Leary. Churchill Livingstone (2008). 242 pp., ISBN: 9780443100475 The textbook by Gwendolen Jull et al. provides a contemporary overview of whiplash, headache and neck pain. As the authors note in the preface, research on cervical spine disorders ‘‘played second fiddle to low-back pain for a large part of the 20th century’’. In the last few decades, however, more extensive knowledge has become available on the cervical spine, and this book is designed to bring both students and clinicians up-to-date on recent developments. This book consists of 15 chapters covering three main areas, basic clinical science (including an analysis of psychosocial factors, sensory manifestations of neck pain, and disturbances in postural stability and movement control), differential diagnosis for the three conditions presented, and a discussion on clinical assessment (history, physical examination, and principles of management). The language used is reader friendly without sacrificing the science. Although ‘‘evidence-based medicine’’ has become a rallying cry in the last decade by clinicians of all disciplines, relatively few are able to bring this into practice because they lack the ability to dissect or understand the literature and apply it. This textbook attempts to tie the two together. As is appropriate or expected in

doi:10.1016/j.math.2009.04.002

such a textbook, there are some nice tables delineating, for example, the classification of whiplash as well as a proposed adaptation of this classification based on physical and psychological factors. However, I would have liked to have seen more tables scattered throughout the chapters. The authors demonstrate not only an appreciation of the scientific literature, but an understanding of the clinical aspect as well. For example, provocative testing is presented in many, even recent orthopaedic textbooks as rather matter of fact (i.e. a positive test indicates the presence of the disease while a negative test indicates its absence) without discussing the actual strategy or diagnostic accuracy of the test. Here Jull et al. report the latest information based upon a systematic review and discuss the nuances (or implications) of orthopaedic testing. In short, students as well as experienced clinicians who want a good overview of the vast body of literature in this area will appreciate this textbook. Sidney M. Rubinstein Department of Epidemiology and Biostatistics, EMGO Institute for Health and Care Research, VU University Center, Amsterdam, The Netherlands E-mail address: [email protected] 2 April 2009

Manual Therapy 14 (2009) e4

Contents lists available at ScienceDirect

Manual Therapy journal homepage: www.elsevier.com/math

Book Review Clinical Sports Medicine – Medical Management and Rehabilitation, W.R. Frontera, S.A. Herring, L.J. Micheli, J.K. Silver. Saunders Elsevier (2007). Hard Cover, 512 pp., 336 ills, CD-ROM, $155.00 CAN, ISBN: 978-1-4160-2443-9 Advancement in sports medicine is an international effort shared by experts around the world. Clinical Sports Medicine – Medical Management and Rehabilitation is the culmination of the editors’ effort to bring a large group of international experts together within a 34 chapter, multi-authored (61 contributors), comprehensive sports medicine text. Within the preface, the editors state their intent to provide a comprehensive text for the medical and non-medical sports medicine professional with an emphasis on the rehabilitation management of the injured athlete. What we find, of their final product, is a commendable text that may have slightly missed its mark on its intended purpose. The text is organized into 3 sections. The first section, ‘‘General Scientific and Medical Concepts’’, overviews general topics in sports medicine such as physiology, conditioning, and nutrition. Following these topics are several chapters detailing management considerations for special populations in sports medicine. The second section, ‘‘Principles of Injury Care and Rehabilitation’’, reviews everything from pre-participation evaluations to prescribing medications. The last section is comprised of a discussion of common athletic injuries topographically organized into chapters. Accompanying the text is a CD-ROM that contains patient handouts and PDF files of diagrams taken directly from the textbook. Praiseworthy to the editors, considering 61 authors contributed to this text, each chapter is written with a consistent structure, and the multiple writing styles of each author blend seamlessly into the next chapter. The narration is well-written and easy to follow; however, there is a noticeable lack of visual aids to explain technically demanding concepts. As an intended comprehensive sports medicine text, many chapters lack sufficient information and detail to be classified as being comprehensive. For instance, there is an absence of review of SLAP lesions, internal impingement, and rotator interval tears within the shoulder chapter. The book is subtitled Medical Management and Rehabilitation. Of the 61 contributing authors, only 2 are not physicians. It is not surprising that the text is written from a physician’s perspective. The advantage of this authorship is the excellent content relating

doi:10.1016/j.math.2009.03.001

to the medical management of the athlete. Highlighting this strength are exceptional chapters on doping and sports, special populations in sports medicine, laboratory tests and diagnostic imaging, and prescribing medications for pain and inflammation. Where this text falters is in its lack of emphasis on rehabilitation. This is inconsistent with the text’s original intention. Considering the book is subtitled Medical Management and Rehabilitation, the textbook offers little to the practical application of rehabilitation for diagnosis-specific conditions. Reference is made consistently within chapters to the conservative management of the athlete. However, detailed diagnosis-specific protocols for rehabilitation are not mentioned in sufficient detail for in-office application. The reason for this omission may potentially be the result of not including experts from other rehabilitation and manual therapy health professions as contributing authors. Perhaps, the inclusion of such experts in a future edition may resolve this weakness. Ironically, due to the wealth of expert medical knowledge that contributed to the development of this text, this book is an excellent resource for the manual therapy practitioner seeking a textbook to further their understanding of the medical management of the injured athlete. When balancing the strengths and weaknesses of this resource, we have to keep in mind multi-authored textbooks will always have both excellent and mediocre chapters. While this text may not be detailed and broad enough to be considered a comprehensive reference text, it does have some chapters that are clinical ‘‘gems’’. While perhaps missing the mark on its intended focus on rehabilitation, the text is an admirable sports medicine resource written for the primary care sports medicine setting. From the perspective of a manual therapy health care professional, this text is a great resource to be included in the manual therapist’s sports medicine library. Alex Lee Graduate Education and Research Programs, Canadian Memorial Chiropractic College, 6100 Leslie Street, Toronto, Ontario, Canada M2H 3J1 Tel.: þ1 416 482 2340; fax: þ1 416 482 2560. E-mail address: [email protected] 24 March 2009