Important Announcement
PubHTML5 Scheduled Server Maintenance on (GMT) Sunday, June 26th, 2:00 am - 8:00 am.
PubHTML5 site will be inoperative during the times indicated!

Home Explore Australian Physiotherapy Journal

Australian Physiotherapy Journal

Published by Horizon College of Physiotherapy, 2022-07-24 13:22:28

Description: Vol. 57 Jul 2011

Search

Read the Text Version

Appraisal Media A mechanisms approach to physical therapy management of pain 4MVLB,\" &E   .FDIBOJTNTBOENBOBHFNFOUPGQBJOGPSUIFQIZTJDBMUIFSBQJTU *\"411SFTT QBHFT Kathleen Sluka is a well regarded educator and researcher presentation of common tools of pain assessment and who has published over 100 peer-reviewed papers. She has treatment is well done, although the application of these provided a voice for the role of physical therapy in pain may be enhanced by reintroducing the models of pain through national (USA) and international professional described in earlier sections e.g. as per the ICF in the IASP- bodies including the International Association for the recommended curricula. Study of Pain (IASP). This book draws on material that she has prepared for a doctoral course titled ‘Mechanisms It was somewhat disappointing that the consideration and Management of Pain’; as such Dr Sluka edits the text of the more physical therapy modalities did not include and is the first author on the large majority of chapters. analysis of their psychological or neuroplastic potential. Other contributions are provided by a mix of American, Once we understand the variability of pain (Chapter 4), European, and Australasian authors. it is improbable that an intimate treatment interaction or particular modality of treatment will not influence non- The target audience of the book is students of physical specific treatment effects. For example, focusing on the therapy and physical therapists who treat people with pain. hypoalgesic effects of exercise without incorporating the While other health professionals would also find the book potential for learning (ie, challenging concepts of re-injury) useful, the text is quite discipline-specific in parts. and fear-reduction through physical activity seems not to align with some of the earlier sentiments of the book. Dr Sluka’s Preface is informative. She summarises the human pain experience as involving three mechanism- The final section of the book considers pain ‘syndromes’ based categories: 1) peripheral mechanisms that drive pain, and some case studies. These are valuable as they present ie, acute pain, 2) central mechanisms that drive pain, ie, the complexity of some common pain conditions and chronic pain, and 3) a combined category, ie, subacute/ also illustrate how some of the assessment and treatment chronic. approaches might be applied. The opening section (the book is divided into four In summary, this book is an ambitious attempt to capture parts) provides definitions of common terms and a brief the complexity of the human pain experience and explain introduction to important explanatory theories and how physical therapists can apply an evidence-based models, including the useful International Classification of approach to manage pain. It is well structured and well Functioning, Disability and Health (ICF). This is followed researched and, for the most part, is likely to be valuable by extensively referenced chapters on pain mechanisms, for its intended target audience. This reviewer considers the using human and animal research evidence to support categorisation of pain into acute, subacute, and chronic (and description of peripheral and central processes. associating pain mechanisms to them) as inadequate, as it often simplifies clinical reasoning to chronology. That is, A highlight is the well worked chapter on pain variability, it can promote the untimely management of complex pain which reminds us that we cannot embed our personal pain presentations in a person with frank acute tissue damage, experiences in our interpretation of the pain experience and discourage the proper somato-visceral evaluation and of others. This emphasises that the complexity of the pain management where pain persists and tissue damage is not experience might be more important to assess than duration apparent; but this is not the common view. Maintaining the of the pain. This perhaps contradicts the simplistic – but focus on pain mechanisms – without the categorisation – well accepted – categorisation of pain based on duration would be a preferred approach, and the main elements of proposed by Dr Sluka in the preface. this book could easily facilitate this. In light of this, and given the evidence of inadequate pain education in physical The middle sections of the book address assessment and therapy programs, Dr Sluka’s book has the potential to treatment including a section devoted to interdisciplinary extend and enhance physiotherapists’ management of pain. management. The chapters include exercise, transcutaneous electrical nerve stimulation and interferential therapy Lester E Jones (reflecting Dr Sluka’s research interests), manual therapy, La Trobe University, Australia medical management, and psychological approaches. The 134 Journal of Physiotherapy 2011 Vol. 57 – © Australian Physiotherapy Association 2011

Appraisal Correspondence Author response We thank Dr English for her thoughtful comments on our examination of transitions provides the best insight into the paper (Alzahrani 2011). We agree that – given the similarly differences between stroke survivors and healthy controls large amount of time spent in sedentary behaviours by in terms of bouts of activity. The transitions we recorded both the stroke survivors and the age-matched controls included lie to sit, sit to lie, recline to sit, sit to recline, – both may be at risk of cardiovascular events including recline to stand, stand to recline, sit to stand and stand to sit. stroke. Our point, which we stand by, was that stroke Despite this comprehensive measurement of transitions, the survivors appear to be no more at risk of recurrent stroke amount of time spent making transitions was very small in and cardiovascular events due to the amount of activity both groups, with a mean of 1 min in the stroke group and they do. This is reflected in our statement that, ‘This would 2 min in the control group. Although this difference was mean that they were no more at risk of recurrent stroke and statistically significant (mean between-group difference 1 cardiovascular events due to low levels of physical activity min, 95% CI 0.3 to 2), this difference was also very small. than their healthy peers.’ It is certainly possible that they This suggests that the sedentary behaviour was likely are more at risk due to the pattern in which that activity to be accumulated in long bouts by both groups, putting is accumulated, but we refrained from making strong both groups at risk of cardiovascular disease. We strongly statements about this possibility for two reasons. First, we agree with Dr English that further research is needed to did not measure the pattern of accumulation of sedentary understand the influence of the pattern of accumulation time and can therefore only make indirect estimates about of sedentary time in stroke survivors. We welcome future such patterns from our data about transitions. Second, the findings in this important area. data about activity pattern and risk is from people without stroke and may not extrapolate to people with stroke. Catherine Dean, Louise Ada and Matar Alzahrani Discipline of Physiotherapy, Faculty of Health Sciences, We agree, nevertheless, with Dr English’s interpretation of how the evidence about sedentary behaviour might apply to The University of Sydney, Australia our data. It is therefore interesting to consider what our data can reveal about this issue. Without reanalysis of the data, Reference Alzahrani A et al (2011) J Physiother 57: 47–51. Journal of Physiotherapy 2011 Vol. 57 – © Australian Physiotherapy Association 2011 133

Bruder et al: Exercise for upper limb fractures Exercise reduces impairment and improves activity in people after some upper limb fractures: a systematic review Andrea Bruder, Nicholas F Taylor, Karen J Dodd and Nora Shields School of Physiotherapy, and Musculoskeletal Research Centre, La Trobe University Australia Question: What is the effect of exercise on reducing impairment and increasing activity in the rehabilitation of people with upper limb fractures? Design: Systematic review of controlled trials. Participants: Adults following an upper limb fracture. Intervention: Any exercise therapy program, including trials where exercise was delivered to both groups providing there was an expectation of different amounts of exercise. Outcome measures: Body structure and function, and activity limitations. Results: 13 relevant trials involving 781 participants with an upper limb fracture were identified. 12 of the 13 trials included exercise of different duration and administration in both intervention and comparison groups. In support of the role of exercise there is evidence that: exercise and advice compared to no intervention reduce pain and improve upper limb activity in the short term after distal radius fracture; starting exercise earlier after conservatively managed proximal humeral fractures can reduce pain and improve shoulder activity; and physiotherapy that included supervised exercise and home exercise increased wrist movement after distal radius fracture when compared to home exercise alone. There is contrary evidence from two trials one after distal radius fracture and one after proximal humeral fracture that a home exercise program was superior to a supervised plus home exercise program. Only a single meta-analysis was conducted due to clinical heterogeneity and a lack of common outcome measures among the included trials. Conclusion: There is evidence to support the role of specific exercise regimens in reducing impairments and improving upper limb function following specific upper limb fractures. [Bruder A, Taylor /' %PEE,+ 4IJFMET/  &YFSDJTFSFEVDFTJNQBJSNFOUBOEJNQSPWFTBDUJWJUZJOQFPQMFBGUFSTPNFVQQFSMJNC GSBDUVSFTBTZTUFNBUJDSFWJFXJournal of Physiotherapyo> ,FZXPSET: Upper limb fracture, Exercise, Rehabilitation, Physiotherapy Introduction Exercise is a common intervention after upper limb fracture. For example, Michlovitz et al (2001) found that exercise was Upper limb fractures are common and affect all age groups prescribed to at least 90% of patients receiving rehabilitation (Bradley and Harrison 2004, Court-Brown et al 2001, after distal radius fracture. The application of exercise Larsen and Lauritsen 1993). In younger adults, upper limb is also consistent with the third key principle of fracture fractures are usually sustained from high-energy trauma management – movement (Adams and Hamblen 1995). such as a motor vehicle accident, whereas in older adults Previous research has identified that therapeutic exercise is with osteoporotic changes these fractures are usually beneficial across a broad range of health conditions (Taylor sustained from a fall (Bradley and Harrison 2004, Court- et al 2007). However, previous systematic reviews of trials Brown et al 2001, Kelsey et al 1992, Larsen and Lauritsen of upper limb fracture management have not focused 1993). Due to an ageing population, the number of the most on the effect of exercise (Handoll et al 2003, Handoll et common upper limb fractures – proximal humeral fractures al 2006). In addition, clinical practice guidelines for the and distal radius fractures – are expected to increase by treatment of distal radius fractures concluded that there about 10% every five years to 2036 (Sanders et al 1999). was weak evidence to support the use of a home exercise program (Lichtman et al 2010). New trials of physiotherapy Following an upper limb fracture, patients are often referred rehabilitation have been published since the two reviews to physiotherapy for rehabilitation to reduce pain, improve were completed in 2003 and 2006. range of movement and strength, and to regain function (AIHW 2008). Even though the aims of physiotherapy are Physiotherapists need current evidence about the clear, the interventions used during the rehabilitation phase effectiveness of treatment techniques to help them make can vary greatly. These interventions can include thermal clinical decisions about patient care and to allocate limited modalities, ultrasound, electrical stimulation, continuous therapy resources for people with upper limb fractures. passive movement, electromyographic biofeedback, soft Therefore, the specific research question for this systematic tissue mobilisation, mobilising and strengthening exercises, review was: application of resting or dynamic splints, advice, and education (Bertoft et al 1984, Clifford 1980, Lundberg et al What is the effect of exercise on reducing impairment 1979, Michlovitz et al 2001). and increasing activity in the rehabilitation of people with upper limb fractures? Journal of Physiotherapy 2011 Vol. 57 – © Australian Physiotherapy Association 2011 71

Research Method quality because it was thought that setting a cut-off value to exclude studies of lesser quality could potentially bias the Identification and selection of studies results of the systematic review (Juni et al 1999). Relevant randomised and quasi-randomised controlled Participants: Age, sex, and type of fracture were recorded trials were identified using a search strategy (See Appendix to enable comparisons of participants between trials. 1 on the eAddenda for full search strategy) from the earliest date possible until January 2011 in the following electronic Intervention: A description of the exercise therapy program databases: CINAHL, MEDLINE, Embase, AMED, SPORT (including timing, intensity, frequency, duration, exercises Discus, PubMed, PEDro and the Cochrane Central Register performed, equipment, total time of each session, number of Controlled Trials. To ensure all relevant studies were of sets and repetitions), the setting in which the program captured, manual reference list checks and citation tracking was performed, and the qualifications of the person of included studies using Web of Science were performed. administering the intervention were recorded. One reviewer examined the study titles and abstracts to determine if they satisfied the inclusion criteria. A second Outcome measures: Outcome measures that assessed reviewer was consulted if the primary reviewer had doubts body structure and function, activity limitations, and about inclusion. Where eligibility was not clear, the full text participation restrictions were examined in accordance with was obtained for more detailed assessment. Studies that the International Classification of Functioning, Disability clearly did not meet the inclusion criteria were eliminated at and Health (ICF) framework (World Health Organisation this point. Titles of journals, names of authors, or supporting 2001). This framework defines functioning and disability as institutions were not masked during the selection process. a multi-dimensional concept according to body functions (eg, loss of muscular strength) and structures (eg, change The inclusion criteria for studies are presented in Box 1. to the skeletal system such as a fracture), activities (eg, The exercise therapy program did not need to be carried unable to dress self), and social participation (eg, unable to out by a physiotherapist provided that the program could be continue employment). regarded as one that a physiotherapist might employ. Trials that were not published in full were excluded. Trials that Data analysis: Summary data for each study, including examined interventions for major complications of fractures means and standard deviations of the post-intervention such as non-union or delayed union were excluded on the group, were extracted independently by two reviewers. basis that these interventions aimed to treat the fracture Study characteristics, patient demographic data, and results itself rather than rehabilitate the individual. were summarised and presented in a tabulated format. For continuous data, standardised mean differences (otherwise #PY Inclusion criteria known as effect sizes), with 95% CIs were calculated by dividing the post-intervention means by the pooled Design standard deviation (Hedges g). Where means and standard deviations were not reported, data were estimated according š Published randomised or quasi-randomised to recommendations outlined by Higgins and Deeks (2009) controlled trial (see Appendix 2 on the eAddenda for statistical equations). A meta-analysis was conducted where a minimum of Participants two trials were clinically homogenous. To account for clinical, methodological, or statistical heterogeneity, a š Participants who had reached skeletal maturity pooled random effects model was applied using RevMan š Participants who had sustained any degree of upper 5a. Statistical heterogeneity was examined by calculating the quantity I² where a value of 0% indicates no observed limb fracture (scapula, clavicle, humerus, radius, heterogeneity, less that 25% is considered to have low levels, ulna, carpal, phalanx) and a value of 100% indicates a completely heterogeneous sample (Higgins et al 2003). š Human Results Intervention 'MPXPGTUVEJFTUISPVHIUIFSFWJFX š Any exercise therapy program The search strategy identified 2375 papers. Following Outcome measures removal of duplicates, screening of titles and abstracts, and the inclusion of one paper identified through citation š Any outcome measure (classified by World Health tracking and one through hand searching of reference Organization 2001) lists, 29 potentially relevant papers remained. After re- application of inclusion criteria to full-text copies of these Comparisons 29 papers, 14 papers remained (Figure 1). These 14 papers represented 13 separate trials because two papers reported š Exercise therapy program versus no exercise data from the same trial at different time points. The other j^[hWfofhe]hWc%fbWY[Xe 15 studies obtained as full text were excluded. Five were not randomised or quasi-randomised controlled trials (Altissimi š Exercise therapy program plus other therapy versus et al 1986, Amirfeyz and Sarangi 2008, Clifford 1980, Liow other therapy et al 2002, MacDermid et al 2001), one was not available in English (Grønlund et al 1990), one was published only š Exercise therapy program versus an alternative as an abstract (Bache et al 2000), and eight had insufficient therapy program that differs in duration, frequency, intensity or method of administration Assessment of characteristics of studies Quality: All included studies were assessed for quality by two reviewers independently using the PEDro scale. The PEDro scale has demonstrated moderate levels of inter-rater reliability (ICC = 0.68, 95% CI 0.57 to 0.76) (Maher et al 2003), and demonstrated evidence of construct reliability in evaluating the methodological quality of clinical trials (de Morton 2009). Studies were not excluded on the basis of 72 Journal of Physiotherapy 2011 Vol. 57 – © Australian Physiotherapy Association 2011

Bruder et al: Exercise for upper limb fractures Titles and abstracts screened (n = 2375) Point estimate Total Y7 Y5 and variability (0 to 10) Potentially relevant papers retrieved for evaluation of full text (n = 29) reported Y6 Papers excluded after screening titles and Y6 abstracts (n = 2346) N4 Y7 Papers excluded after evaluation Y8 of full text (n =15) Y5 š Research design not a Y8 Y2 randomised or quasi-randomised Y7 clinical trial (n = 5) Y2 š Full text not available in English N5 (n = 1) š Not published as full text (n = 1) Between-group N Y N NYY Y Y N Y N NYNN Y š Did not provide sufficient difference information about physiotherapy reported intervention (n = 1) Y š Intervention compared different N periods of mobilisation and Y immediate post fracture Y management (n = 7) Y Y Papers included in review (n = 14) Y N 'JHVSF Flow of studies through the review. N N Y Intention- to-treat analysis N N N N Y N Y N Y N N < 15% dropouts N Y N Y Y Y Y N Y N N Therapist Assessor blinding blinding NY NY NN NY NY NN NY NN NY NN NY Participant blinding N N N N N N N N N N N information about the exercise therapy intervention (Davis Groups and Buchanan 1987, de Bruijn 1987, Dias et al 1987, Gaine similar at et al 1998, Lozano Calderón et al 2008, McAuliffe et al baseline 1987, Millett and Rushton 1995, Oskarsson et al 1997). Y Characteristics of included studies Y Y Design: A single trial evaluated the effects of exercise and Y home advice compared to a no-intervention control group Y in patients with a distal radius fractures (Kay et al 2008). In Y the remaining 12 trials, differing amounts of exercise and Y advice were incorporated in both control and intervention N groups. Three trials compared exercise introduced earlier Y in rehabilitation with delayed introduction of exercise N following a proximal humeral fracture (Agorastides et Y al 2007, Hodgson et al 2003, Lefevre-Colau et al 2007), while in four trials patients received supervised exercise in 5BCMFPEDro scores of included studies (n = 13). Concealed addition to a home exercise program compared to simply allocation a home exercise program (Christensen et al 2001, Maciel et al 2005, Pasila et al 1974, Revay et al 1992). Five trials Y compared physiotherapy, which included supervised Y exercise plus a home exercise program, with a home exercise Y program (Bertoft et al 1984, Krischak et al 2009, Lundberg Y et al 1979, Wakefield and McQueen 2000, Watt et al 2000). Y N Quality: Quality assessment PEDro scores ranged from 2 Y to 8 out of 10 with a median score of 6 (see Table 1). Due to N Y Journal of Physiotherapy 2011 Vol. 57 – © Australian Physiotherapy Association 2011 N Y Random allocation Y Y Y Y Y Y Y Y Y Y Y Y Y Study Agorastides et al (2007) Bertoft et al (1984) Christensen et al (2007) Hodgson et al (2003, 2007) Kay et al (2008) Krischak et al (2009) Lefevre-Colau et al (2007) Lundberg et al (1979) Maciel et al (2005) Pasila et al (2000) Revay et al (1992) MWa[Å[bZCYGk[[d (2000) Watt et al (2000) 73

74 JournalofPhysiotherapy2011 Vol.57 – ©AustralianPhysiotherapyAssociation2011 5BCMFSummary of included studies (n = 13) Study Design Par ticipants Inter vention RCT Agorastides n = 49 Exp = Advice et al (2007) Age (yr) = Exp 72 (SD 12), Con 67 (SD 14) immob Gender = 10 M, 39 F progres Bertoft et al RCT Diagnosis = Proximal humerus fracture (1984) Fracture Type = 3-part Neer: 9, 4-part Neer: 39 Con = Advice immob Christensen RCT n = 18 progre et al (2001) Age (yr) = Exp 62, Con 66, range 50 –75 Gender = Not avail Exp = Exercis Hodgson RCT Diagnosis = Proximal humerus fracture joint m et al (2003, RCT Fracture Type = non-displaced or slightly sling 10 2007) displaced n = 30 Con = 3 x Ins Kay et al Age (yr) = Exp 66 (range 46 –82), Con 66 (range 5–10 m (2008) 57–79) _d`kh o\" Gender = 3 M, 27 F Krischak et al RCT Diagnosis = Distal radius fracture Exp = Exercis (2009) Fracture Type = Older Type I: 2, Type II: 11, HEP), h Type III: 9, Type IV: 4, unclassified: 3 eYYkfW n = 86 Age (yr) = Exp 71 (SD 13), Con 67 (SD 12) Con = 1 x Ins Gender = 16 M, 70 F _dj[h l[ Diagnosis = Proximal humerus fracture Fracture Type = Neer group 1 Exp = Exercis mobilis n = 56 within 1 Age (yr) = Exp 55 (SD 20), Con 56 (SD 20) Gender = 17 M, 39 F Con = Exercis Diagnosis = Distal radius fracture immob Fracture Type = AO system: ea: 20, pa: 8, ca: comme 11 n = 48 Exp = 1 x Ins Age (yr) = Exp 56 (SD 11), Con 54 (SD 18) compre Gender = 16 M, 30 F Diagnosis = Distal radius fracture Con = Usual Fracture Type = AO system: ea: 14, pa: 1, ca: 31 Exp = Exercis Maciel et al RCT n = 41 inter ve (2005) Age (yr) = Exp 56 (SD 18), Con 56 (SD 19) wk pos Gender = 10 M, 31 F Diagnosis = Distal radius fracture Con = Exercis Fracture Type = AO system: ea: 29, pa: 1, ca: Yecc[ 10, unclassified: 1 for 6 w Exp = Exercis therapy treatme Con = 1–2 x I with ad

e and exercise program (supervised and HEP), arm Outcome measures Research bilised (sling) x 2 wk, then exercises commenced and š Constant Shoulder ssed every 2 wk after commencement for 10 –12 wk Assessment Score, Oxford Score e and exercise program (supervised and HEP), arm bilised (sling) x 6 wk, then exercises commenced and š Follow-up = 24, 52 wk essed every 2 wk after commencement for 10 –12 wk (from injury) se program (supervised and HEP), advice, passive š Shoulder ROM, isometric mobilisation, commenced following immobilisation in strength, subjective assessment ADLs 0 –12 days post injury, 9 x over 10 –12 wk š Follow-up = 3, 8, 16, 24, structed to perform same HEP as intervention group, 52 wk (from injury) min following immobilisation in sling 10-12 days post * ¸ +n %ZWo š CeZIeb]WWhZM[hb[o Functional Score, grip se program (supervised by occupational therapist and strength heat, advice commenced following removal of POP, Wj_edWbj^[hWfo(%ma\">;F)%ZWo\"ZkhWj_eddejif[Y_Å[Z š Follow-up = 0, 7, 31 wk (from removal of POP) structed to perform same HEP (with heat) as [dj_ed]hekf\")%ZWo\"ZkhWj_eddejif[Y_Å[Z se program (supervised and HEP), advice, immediate š Constant Shoulder sation and exercise program with physiotherapist Score, Short Form 36, 1 x wk of injury Croft shoulder disability questionnaire se program (supervised and HEP), advice, bilised (collar and cuff sling) for 3 wk from injury then š Follow-up = 8 wk, 16 wk, 1 yr, 2 yr (from injury) enced exercise program with physiotherapist š Wrist ROM, grip strength, structed to perform exercise program (HEP), advice, PRWE, QuickDASH ession commenced after removal of POP š Follow-up = 0, 3, 6 wk care. No physiotherapy intervention (natural recovery) (from removal of POP) se program (supervised and HEP), advice, other š Wrist ROM, grip strength, entions (at discretion of physiotherapist), commenced 1 PRWE st volar plating, 20 –30 min x 12 for 6 wk š Follow-up = 1, 7 wk (from se program (detailed HEP and guidance booklet), injur y) [dY[Z'mafeijlebWhfbWj_d]\">;F(&c_dn(n %ZWo wk provided by a physician š Wrist ROM, grip strength, PRWE se program (supervised and HEP), advice, manual y commenced after removal of POP, regular š Follow-up = 0, 6, 24 wk ents for 6 x wk (from removal of POP) Instructed to perform same HEP as intervention group dvice, commenced after removal of POP

JournalofPhysiotherapy2011 Vol.57 – ©AustralianPhysiotherapyAssociation2011 Lefevre- RCT n = 74 Exp = Exercis Colau et al Age (yr) = Exp 63 (SD 18), Con 63 (SD 18) massa (2007) Gender = 20 M, 54 F session Diagnosis = Proximal humerus fracture \\hWYjkh[ Lundberg et RCT Fracture Type = 1-part Neer: 34, 2-part Neer: al (1979) 16, 3-part Neer: 24 Con = Exercis ice, ma Pasila et al RCT n = 42 betwee (1974) Age (yr) = 65 (range 30 –89) _cceX Gender = 5 M, 37 F freque Revay et al RCT Diagnosis = Proximal humerus fracture (1992) Fracture Type = Neer group 1 Exp = Exercis n = 96 joint m Age (yr) = Not specified f^oi_ej Gender = 7 M, 89 F Diagnosis = Distal radius fracture 9ed3( n?di Fracture Type = Older Type II: 9, Type III: 66, with ad Type IV: 18 n = 48 Exp = Exercis Age (yr) = 62 comme Gender = 9 M, 39 F averag Diagnosis = Proximal humerus fracture Fracture Type = Neer group 1 Con = 1 x Ins physic Exp = Exercis advice, cWn(& Con = 2 x Ins ma\">; Wakefield et RCT n = 96 Exp = Exercis al (2000) Age (yr) = Exp 72 (SD 10), Con 74 (SD 9) _dj[h l[ Gender = 9 M, 87 F duratio Watt et al RCT Diagnosis = Distal radius fracture (2000) Fracture Type = AO system: predominantly ea 9ed3'n?di n = 18 Age (yr) = Exp 74 (SD 10), Con 77 (SD 5) Exp = Exercis Gender = 1 M, 17 F joint m Diagnosis = Distal radius fracture Fracture Type = Frykman I–III: 6, Frykman IV– Con = 1 x Ins VI: 5, Frykman VII–VIII: 7 ADL = activities of daily living, Con = control group, Exp = experimental group, HEP = home exercise pr PRWE = patient rated wrist evaluation, SF36 = short form 36 75

se program (supervised and HEP), advice, ice, š Pain, shoulder ROM age, passive joint mobilisation and sling between š Follow-up = 6, 12, 24 wk ns (4– 6 wk), treatment commenced 72 hrs post (from injury) [\"(^hin+n %ma\"\\h[gk[dYoh[ZkY[Zel[h)cedj^i š Pain, shoulder ROM, se program (supervised and HEP), advice, strength (grip and assage, passive joint mobilisation, and sling shoulder lifting power) en sessions (1–3 wk), treatment commenced after X_b_iWj_ed_dib_d]\\eh)ma\"(^hin*n %ma\\eh*ma š Follow-up = 4, 12 wk ency reduced until 6 mths (from injury) se program (supervised and HEP), advice, passive š Wrist ROM, grip strength, mobilisation, immobilised in sling 1 wk then commenced oedema, subjective questions j^[hWfo(& ¸)&c_dn'¸(n%man.¸'(ma$ š Follow-up = 5, 8, 12 wk ijhkYj[Zjef[h\\ehciWc[>;F+ ¸'&c_d*¸+n%ZWo (from injury) dvice, immobilised in sling 1 wk. se program (supervised and HEP) and advice enced during immobilisation period 1–12 x (4 on ge), discharge at discretion of treating physiotherapist structed to perform same HEP and advice provided by cian se program (supervised hydrotherapy and HEP) and š Pain, shoulder ROM, Bruder et al: Exercise for upper limb fractures e, immobilised in sling 1 wk, hydrotherapy 30 min x subjective assessment of 9 ADL items and &WdZ>;F'& ¸'+c_dn*n%ZWo 4 functional tests structed to perform same HEP, immobilised in sling 1 measured ad modum ;F'& ¸'+c_dn*n%ZWo Ber tof t-Solem se program (supervised and HEP), advice and passive š Follow-up = 4, 8, 12, 52 [dj_ediWjZ_iYh[j_ede\\f^oi_ej^[hWf_ij\">;F)%ZWo wk (from injury) on and frequency at the discretion of physiotherapist ijhkYj[Zjef[h \\ehciWc[>;F)%ZWo š Pain, wrist ROM, grip strength, scoring system se program (supervised and HEP), advice and passive to assess ADL, SF 36 mobilisation, attended ~5 times š Follow-up = 6, 12, 24 wk structed by orthopaedic surgeon to perform same HEP (from injury) š Wrist ROM, grip strength š Follow-up = 0, 6 wk (from removal of POP) rogram, RCT = randomised controlled trial, ROM = range of motion, POP = plaster of paris,

Research Study SMD (95% CI) –1 –0.5 SMD 0.5 1 0 Christensen et al (2001) n = 30 Insufficient data to calculate SMD Maciel et al (2005) n = 41 –0.17 (–0.83, 0.50) Wrist Ext ROM 6 wk –0.35 (–1.01, 0.33) Wrist Flex ROM 6 wk 0.07 (–0.60, 0.73) Grip Strength 6 wk 0.04 (–0.63, 0.70) Pain (PRWE) 6 wk –0.24 (–0.90, 0.43) Activity (PRWE) 6 wk 0.15 (–0.54, 0.84) Wrist Ext ROM 24 wk –0.04 (–0.73, 0.65) Wrist Flex ROM 24 wk –0.14 (–0.83, 0.55) Grip Strength 24 wk –0.11 (–0.80, 0.58) Pain (PRWE) 24 wk –0.18 (–0.87, 0.52) Activity (PRWE) 24 wk 0.13 (–0.27, 0.53) Pasila et al (1999) n = 96 –0.22 (–0.62, 0.18) Wrist Ext ROM 5 wk –0.37 (–0.78, 0.03) Wrist Flex ROM 5 wk 0.16 (–0.25, 0.55) Grip Strength 5 wk 0.17 (–0.24, 0.57) Wrist Ext ROM 8 wk –0.10 (–0.50, 0.30) Wrist Flex ROM 8 wk 0.07 (–0.33, 0.47) Grip Strength 8 wk 0.00 (–0.40, 0.40) Wrist Ext ROM 12 wk –0.11 (–0.51, 0.29) Wrist Flex ROM 12 wk Grip Strength 12 wk Favours control Favours experimental Figure 2. SMD (95% CI) of effect of supervised exercise plus a home exercise program compared with home exercise program alone after distal radius fracture. Ext = extension, Flex = flexion, PRWE = patient rated wrist evaluation, ROM = range of motion, SMD = standardised mean difference the nature of the interventions, none of the trials was able to Participants: The 13 trials included in the analysis provided blind the participants or therapists to the intervention. Eight data from 781 participants aged from 32 to 82 years, of trials blinded the assessor, four trials used intention-to-treat whom about 80% were female (see Table 2). Participants analysis, and eight trials concealed allocation. had sustained either a distal radius fracture (7 trials) or a proximal humeral fracture (6 trials) (see Table 2). No other Sufficient data in the form of means and standard deviations upper limb fractures were included. were provided in six trials to allow calculation of effect sizes (Agorastides et al 2007, Bertoft et al 1984, Hodgson et Synthesis: Only one meta-analysis could be performed. al 2003, Kay et al 2008, Lefevre-Colau et al 2007, Maciel Clinical heterogeneity between trials precluded further et al 2005). For an additional trial, the mean and standard meta-analysis. The results are presented according to the deviations were imputed from a graph (Pasila et al 1974). interventions being compared and the type of fracture. Five trials provided adequate data to estimate means and standard deviations by providing median and interquartile Effect of advice and exercise versus no ranges (Krischak et al 2009, Watt et al 2000), means with p intervention values (Revay et al 1992), and means with standard errors (Lundberg et al 1979, Wakefield and McQueen 2000). Two Distal radius fractures: There is preliminary evidence trials provided insufficient data to calculate standardised from a single trial that exercise combined with advice mean differences (Christensen et al 2001, Hodgson et al can improve upper limb activity and reduce pain in the 2007). short term after distal radius fracture. A single session of advice and exercise compared to no intervention found 76 Journal of Physiotherapy 2011 Vol. 57 – © Australian Physiotherapy Association 2011

Bruder et al: Exercise for upper limb fractures Study SMD (95% CI) SMD 0 Conservatively Managed –1 – 0.5 0.5 1 Hodgson et al (2003) n = 83 Pain (SF-36) 8 wk 0.81 (0.36, 1.25) Constant Shoulder Score 8 wk 0.13 (–0.31, 0.57) Role Limitation-phys (SF-36) 8 wk 0.60 (0.15, 1.03) Pain (SF-36) 16 wk 0.60 (0.14, 1.04) Constant Shoulder Score 16 wk 0.78 (0.32, 1.22) Role Limitation-phys (SF-36) 16 wk 0.53 (0.08, 0.96) Pain (SF-36) 52 wk 0.13 (–0.31, 0.57) Constant Shoulder Score 52 wk 0.29 (–0.15, 0.72) Role Limitation-phys (SF-36) 52 wk 0.13 (–0.31, 0.56) Lefevre-Colau et al (2007) n = 64 0.61 (0.10, 1.11) Constant Shoulder Score 6 wk 0.10 (–0.39, 0.59) Change of Pain Intensity 6 wk 0.60 (0.10, 1.10) Difference in Abd ROM 6 wk 0.67 (0.16, 1.17) Difference in Flex ROM 6 wk 0.62 (0.12, 1.12) Constant Shoulder Score 12 wk 0.51 (0.00, 1.00) Change of Pain Intensity 12 wk 0.52, (0.01, 1.01) Difference in Abd ROM 12 wk 0.59 (0.08, 1.08) Difference in Flex ROM 12 wk 0.47 (–0.03, 0.96) Constant Shoulder Score 24 wk 0.01 (–0.48, 0.50) Change of Pain Intensity 24 wk 0.36 (–0.14, 0.85) Difference in Abd ROM 24 wk 0.28 (–0.22, 0.77) Difference in Flex ROM 24 wk Hodgson et al (2007) n = 86 –0.07 (–0.63, 0.50) Insufficient data to calculate SMD –0.23 (–0.79, 0.34) –0.19 (–0.75, 0.38) Surgically Managed –0.31 (–0.87, 0.26) Agorastides et al (1999) n = 49 Constant Shoulder Score 24 wk Oxford Score 24 wk Constant Shoulder Score 52 wk Oxford Score 52 wk Favours control Favours experimental Figure 3. SMD (95% CI) of effect of early exercise and mobilisation compared with delayed exercise and immobilisation after proximal humeral fracture. Abd = abduction, Flex = flexion, phys = physical domain, ROM = range of motion, SF-36 = Short Form 36, SMD = standardised mean difference Journal of Physiotherapy 2011 Vol. 57 – © Australian Physiotherapy Association 2011 77

Research Study SMD (95% CI) SMD 1 1.5 2 2.5 –2.5 – 2 –1.5 –1 – 0.5 0 0.5 Conservatively Managed 0.14 (–0.26, 0.54) Wakefield & McQueen (2000) n = 96 0.00 (–0.41, 0.41) Pain 6 wk 0.03 (–0.38, 0.44) Pain 12 wk 0.23 (–0.19, 0.64) Grip strength 12 wk 0.07 (–0.34, 0.48) Wrist flex/ext ROM 12 wk 0.07 (–0.42, 0.55) Functional score for ADL12 wk 0.03 (–0.46, 0.51) Pain 24 wk 0.72 (0.21, 1.21) Grip strength 24 wk 0.03 (–0.45, 0.51) Wrist flex/ext ROM 24 wk Functional score for ADL 42 wk 1.56 (0.44, 2.53) 1.33 (0.26, 2.28) Wa tt et al (2007) n = 18 Wrist ext ROM 6 wk Grip strength 6 wk Surgically Managed –1.70 (–2.35, –1.00) Krischak et al (2009) n = 46 – 0.95 (–1.54, –0.32) Grip strength 6 wk –1.18 (–0.53, –1.78) Wrist flex/ext ROM 6 wk PRWE 6 wk Favours control Favours experimental Figure 4. SMD (95% CI) of effect of physiotherapy including supervised exercise and a home exercise program compared to a home exercise program after distal radius fracture. ADL = activities of daily living, Ext= extension, Flex = flexion, ROM = range of motion, SMD = standardised mean difference improvements in upper limb activity at 3 weeks (SMD 0.61, Wakefield 95% CI 0.03 to 1.19), and reduced pain at 3 weeks (SMD Watt 0.77, 95% CI 0.18 to 1.36) and 6 weeks (SMD 0.63, 95% CI 0.04 to 1.04) (Kay et al 2008). There were no other -2 -1 0 1 2 statistically significant between-group differences for the primary outcome measure of wrist extension or for the 'JHVSF SMD (95%) CI) of effect of physiotherapy secondary outcomes of other ranges of motion and grip (including supervised exercise plus home exercise strength at weeks three or six. program) on grip strength compared with a home exercise program at 12 weeks after immobilisation. HEP = home Proximal humeral fractures: No trials examined exercise exercise program. and advice compared to no intervention after proximal humerus fracture. Proximal humeral fractures: There is preliminary evidence from a single trial that adding supervised exercise Supervised and home exercise versus home to a home exercise program may reduce upper limb activity, exercise and increase impairment in the short term after proximal humeral fracture when compared with home exercise alone. Distal radius fractures: There is no evidence to support Compared to supervised exercise in a swimming pool adding supervised exercise to a home exercise program after (20 classes of 30 minutes duration) plus home exercise, a distal radius fracture (Figure 2). None of the three trials that control group performing home exercise only demonstrated investigated the effect of physiotherapy-supervised exercise improvement at two months in self-reported assessments plus a home exercise program compared to a home exercise including taking an object from a shelf (SMD –1.02, 95% program alone reported statistically significant between- CI –1.61 to –0.40), hanging the laundry (SMD –0.65, 95% group differences for any impairment or activity outcome CI –1.22 to –0.06), washing the opposite axilla (SMD –0.70, measures (Christensen et al 2001, Maciel et al 2005, Pasila 95% CI –1.27 to –0.10) and making a bed (SMD –0.78, et al 1974). 95% CI –1.35 to –0.18) (Revay et al 1992). The control group also had greater improvements in active shoulder Two trials were similar in that the supervised therapy abduction, flexion, and internal rotation at 2 months, and program commenced six weeks following distal radius active shoulder abduction and internal rotation at 3 months fracture and involved activity focused exercises, however were also reported. There were no significant between- no common outcome measures were used (Christensen et group differences at one year follow up. al 2001, Maciel et al 2005). The third trial (Pasila et al) was not comparable to the other two trials as the intervention was implemented to non-splinted joints during the immobilisation period. 78 Journal of Physiotherapy 2011 Vol. 57 – © Australian Physiotherapy Association 2011

Bruder et al: Exercise for upper limb fractures Study SMD (95% CI) SMD 0.45 (–0.51, 1.37) –1 –0.5 0 0.5 1 1.5 Bertoft et al (1984) n = 18 Pain (hand on neck) 3 wk Pain (hand in back) 3 wk 1.07 (0.03, 2.00) Sh AROM-hand on neck 3 wk –0.30 (–1.22, 0.65) Sh AROM-hand in back 3 wk –0.67 (–1.59, 0.32) Pain (hand on neck) 8 wk –0.19 (–1.11, 0.75) Pain (hand in back) 8 wk 0.63 (–0.35, 1.55) Sh AROM-hand on neck 8 wk 0.45 (–0.51, 1.37) Sh AROM-hand in back 8 wk –0.32 (–1.24, 0.63) Pain (hand on neck) 16 wk 0.18 (–0.78, 1.13) Pain (hand in back) 16 wk 0.20 (–0.76, 1.15) Sh AROM-hand on neck 16 wk 0.27 (–0.68, 1.19) Sh AROM-hand in back 16 wk –0.50 (–1.42, 0.46) Pain (hand on neck) 24 wk 0.22 (–0.81, 1.22) Pain (hand in back) 24 wk –0.13 (–1.14, 0.89) Sh AROM-hand on neck 24 wk –0.23 (–1.15, 0.72) Sh AROM-hand in back 24 wk –0.15 (–1.07, 0.79) Lundberg et al (1979) n = 42 0.34 (–0.27, 0.95) Grip Strength 4 wk 0.47 (–0.15, 1.08) Grip Power 4 wk –0.02 (–0.63, 0.59) Sh Flex AROM 4 wk 0.35 (–0.27, 0.95) Sh Flex PROM 4 wk –0.20 (–0.81, 0.41) Grip Strength 12 wk –0.25 (–0.86, 0.36) Grip Power 12 wk 0.20 (–0.41, 0.80) Sh Flex AROM 12 wk 0.38 (–0.24, 0.99) Sh Flex PROM 12 wk Favours control Favours experimental Figure 7. SMD (95% CI) of effect of physiotherapy including supervised exercise and a home exercise program compared to a home exercise program after proximal humeral fractures. AROM = active range of motion, Flex = flexion, Sh = shoulder, SMD = standardised mean difference Early versus delayed exercise and home exercise program in both the intervention and control groups. At one year follow-up, total shoulder Distal radius fractures: No trials examined starting disability as measured on the Croft Shoulder Disability exercise earlier after immobilisation compared with delayed Questionnaire was 43% compared to 73% in the early exercise after distal radius fracture. exercise group compared to the delayed exercise group (Hodgson et al 2007). Proximal humeral fractures: There is evidence that starting exercise earlier after conservatively managed In one trial involving surgically managed proximal humeral proximal humeral fractures can reduce pain in the short fractures, starting exercise earlier did not improve shoulder term and improve shoulder activity in the short and medium activity (Figure 3). Agorastides et al (2007) included more term (Figure 3). The trials by Hodgson et al (2003) and severe fracture types (Neer 3- and 4-part fractures) managed Lefevre-Colau et al (2007) started exercise within the by hemiarthroplasty, comparing exercises started at 2 first week after fracture compared to starting exercise at 3 weeks with exercises started after 6 weeks immobilisation. weeks. Meta-analysis was not conducted as the two trials There were no significant between-group differences on the differed in that Lefevre-Colau et al (2007) included other Constant Shoulder Assessment Score or Oxford Score. physiotherapy modalities in addition to supervised exercise Journal of Physiotherapy 2011 Vol. 57 – © Australian Physiotherapy Association 2011 79

Research Physiotherapy with supervised and home the rehabilitation of people with upper limb fractures exercise versus home exercise (Michlovitz et al 2001). Prescription of exercise after upper limb fracture is also consistent with the key principle of Distal radius fractures: Two trials found that adding fracture management, movement (Adams and Hamblen supervised exercise to a home exercise program as part 1995), and adherence to prescribed home exercise has been of physiotherapy for conservatively managed distal radius found to be moderately-to-strongly associated with short- fractures can improve wrist range of motion in the short term term outcomes of impairment and activity after distal (Figure 4). In contrast, a meta-analysis did not demonstrate radius fracture (Lyngcoln et al 2005). Despite this there any effect of physiotherapy including supervised exercise are currently no high quality trials that have evaluated plus a home exercise program on grip strength following the effects of exercise alone on rehabilitation outcomes. distal radius fracture (d = 0.55, 95% CI –0.65 to 1.75, I² For this reason it is not possible to strongly advocate the = 79%) (Wakefield and McQueen 2000, Watt et al 2000) routine use of exercise for all upper limb fractures. Having (Figure 5, see also Figure 6 on the eAddenda for detailed said that, there is preliminary evidence to support the role forest plot). No further meta-analyses could be conducted of exercise in the rehabilitation of specific upper limb due to the use of different outcome measures. fractures, which provides support for particular protocols. Exercise and advice was found to be beneficial compared One trial reported that adding supervised exercise to a home to no intervention in the short term in the management of exercise program as part of physiotherapy after surgically patients with a distal radius fracture (Kay et al 2008); early managed distal radius fractures reduces upper limb function commencement of exercise was found to be beneficial in and increases impairment in the short term when compared patients with conservatively managed proximal humeral with home exercise alone (Krischak et al 2009) (Figure 4). fractures (Hodgson et al 2007, Lefevre-Colau et al 2007); Krischak et al (2009) commenced mobilisation of patients and supervised exercise in addition to home exercise as two weeks after volar plating for a distal radius fracture. part of physiotherapy was found to increase wrist range of Patients randomised to the control group received detailed movement in patients with conservatively managed distal instructions and a home exercise program. radius fractures (Wakefield and McQueen 2000, Watt et al 2000). In contrast, however, a program of supervised Proximal humeral fractures: There is no available evidence exercise in addition to home exercise was found to result that adding supervised exercise to a home exercise program in poorer short-term outcomes of range of movement and as part of physiotherapy compared to a home exercise upper limb activity after surgically managed distal radius program alone can improve upper limb activity, or reduce fractures (Krischak et al 2009) and proximal humeral impairment after proximal humeral fracture (Figure 7). Two fractures (Revay et al 1992). trials investigated physiotherapy which included supervised exercise plus a home exercise program compared with One factor that makes interpretation of the results of this a home exercise program on patients with conservatively review difficult is the use of co-interventions in the designs managed proximal humeral fractures, with removal of of the included trials. Apart from one trial that found sling between days 7 to 12 (Bertoft et al 1984, Lundberg exercise and advice compared to no intervention beneficial et al 1979). No significant between-group differences were (Kay et al 2008), all trials included exercise in both the identified on any impairment (shoulder range of movement, intervention and control group, albeit with differences in muscle strength, pain) or activity measure (activities of the duration or number of supervised sessions. Further daily living) in the short or medium term (Bertoft et al investigation with controlled trials that investigate exercise 1984, Lundberg et al 1979). as the only intervention versus a no-intervention control group is warranted to explore the role of exercise in upper Adherence to an exercise program: Three of the 13 trials limb fracture rehabilitation. reported adherence to the supervised exercise sessions or to the prescribed home exercise program. Adherence The evidence demonstrating short- and medium-term was reported for the entire study cohort in one trial (70% improvement in upper limb function and reduced impairment attended the supervised exercise sessions) (Lefevre-Colau et with early commencement of exercise after fracture, is al 2007), the intervention group in one trial (85% completed an example of how the use of co-interventions can make their exercises at least once a day) (Kay et al 2008), and interpretation difficult (Hodgson et al 2003, Lefevre-Colau the control group in one trial (97% rated the home exercise et al 2007). One explanation could be that the benefits may program as being completed) (Krischak et al 2009). be attributable to exercising for a longer duration. However, an alternative explanation for the positive outcomes could Adverse events: In general, adverse events were not reported be that the participants benefited from the reduced time systematically. One trial explicitly stated that no adverse of immobilisation rather than from implementing exercise events were related to the intervention (Maciel et al 2005). earlier. Several trials indicate that reducing immobilisation Another trial did report complications associated with the time alone after an upper limb fracture without therapy wrist fracture, but most of these were noted at the time of intervention could be beneficial (Davis and Buchanan 1987, initial assessment (Kay et al 2008), and another reported Dias et al 1987, McAuliffe et al 1987). complications but these related more to the surgical approach than the physiotherapy interventions (Agorastides A theme that emerged from the review was that the trials et al 2007). that reported contrary findings or lack of effect included more severe fractures that had been surgically managed Discussion (Agorastides et al 2007, Krischak et al 2009). In these trials the group that received more exercise (ie, supervised Exercise (often in conjunction with other interventions) exercise in addition to home exercise program or earlier is one of the most common physiotherapy interventions commencement of exercise) had poorer observed outcomes used to reduce impairment and increase activity in 80 Journal of Physiotherapy 2011 Vol. 57 – © Australian Physiotherapy Association 2011

Bruder et al: Exercise for upper limb fractures than the group that received less exercise (ie, home exercise In conclusion, this systematic review suggests there is program alone or delayed exercise). These results lead to inconclusive evidence to support the role of exercise during the speculation that the amount of inflammation and tissue rehabilitation following an upper limb fracture. This is damage from the severity of the fracture and surgery might not consistent with previous research demonstrating the mean that a period of relative rest or controlled movement effectiveness of exercise in other conditions. There is some may be an important part of recovery during rehabilitation. evidence that conservatively managed fractures of the distal However, further research that controls for co-interventions radius and the proximal humerus may benefit from exercise, and closely monitors the amount of exercise completed which is consistent with the theoretical benefits associated would be needed to confirm this. with movement. However, the use of co-interventions in the trials makes a more definite conclusion difficult. Given that Another theme that emerged was that exercise may be more exercise is a common intervention used after an upper limb likely to lead to reduction in impairment, particularly range fracture, controlled trials are needed to provide stronger of movement, than improvements in activity limitations. evidence about the role of exercise in upper limb fracture A number of trials reported short-term improvements in rehabilitation. Q range of movement in the group receiving more exercise (Lefevre-Colau et al 2007, Wakefield and McQueen 2000, Correspondence: Andrea Bruder, School of Physiotherapy Watt et al 2000), but there were few examples where the and Musculoskeletal Research Centre, La Trobe University, improvements carried over into an improved ability to Australia. Email: [email protected] complete daily activities. Given the principle of specificity of training, it is perhaps not surprising that exercises for Footnotes: aReview Manager 5 (2008) upper limb fracture rehabilitation that focus on repeated http://ims.cochrane.org/revman. movements or repeated contractions might lead, when effective, to increased range of movement and increased eAddenda: Appendix 1 and 2, and Figure 6 available at strength. A couple of trials attempted to address this jop.physiotherapy.asn.au possible limitation by implementing ‘activity-focused’ exercises, but the content of the interventions were not well References described and the investigators did not detect any beneficial effect (Christensen et al 2001, Maciel et al 2005). Adams JC, Hamblen DL (1995) Outline of Orthopaedics (12th edn). Edinburgh: Churchill Livingston. The findings of this review are similar to two previously published systematic reviews that concluded there was 7]ehWij_Z[i ?\" I_def_Z_i 9\" ;b C[b_]o C\" O_d G\" 8hemdied insufficient evidence to determine which rehabilitation P, Frostick SP (2007) Early versus late mobilization after interventions may be useful for the management of distal hemiarthroplasty for proximal humeral fractures. Journal of radial fractures (Handoll et al 2006) and proximal humeral Shoulder and Elbow Surgery 16: S33–S38. fractures (Handoll et al 2003). The current systematic review adds to the literature by focusing on exercise and Altissimi M, Antenucci R, Fiacca C, Mancini GB (1986) Long- including recently published studies (Agorastides et al 2007, term results of conservative treatment of fractures of the Hodgson et al 2007, Kay et al 2008, Krischak et al 2009). distal radius. Clinical Orthopaedics and Related Research 206: 202–210. A strength of this systematic review was its comprehensive search strategy which included eight electronic databases, Amirfeyz R, Sarangi P (2008) Shoulder hemiarthroplasty for citation tracking, and manual reference list checks with fracture with a conservative rehabilitation regime. Archives no included trials identified outside the database searches. of Orthopaedic and Trauma Surgery 128: 985–988. A strict inclusion criterion was used to include only randomised controlled trials or quasi randomised controlled Australian Institute of Health and Welfare (2008) Arthritis and trials as they are less subject to bias than other designs osteoporosis in Australia 2008. Canberra. (Khan et al 2001). Bache SJ, Ankorn L, Hiller L, Gaffey A (2000) Two different A limitation of this systematic review is that only a single approaches to physiotherapeutic management of patients meta-analysis could be conducted. No other meta-analyses with distal radial fractures. Physiotherapy 86: 383. were conducted due to clinical heterogeneity and a lack of common outcome measures among the included trials. We Bertoft ES, Lundh I, Ringqvist I (1984) Physiotherapy after may have missed some trials due to language restrictions. fracture of the proximal end of the humerus. Comparison Incomplete data required the authors to interpret data from between two methods. Scandinavian Journal of Rehabilitation Figures in some trials, which could have been a source of Medicine 16: 11–16. error. Methodological flaws were also identified among the included trials. Some trials consisted of small sample Bradley C, Harrison J (2004) Descriptive epidemiology sizes, there was lack of use of reliable and valid outcome of traumatic fractures in Australia. Injury Research and measures, and a lack of blinding. Trial reports frequently did Statistics Series Number 17. Canberra: Australian Institute not clearly define the exercises included in the interventions of Health and Welfare. and the prescribed regimen. From the trials that did outline the intensity of the program, adherence to the protocols Christensen OM, Kunov A, Hansen FF, Christensen TC, was poorly reported. Further research is needed that is Krasheninnikoff M (2001) Occupational therapy and Colles’ methodologically sound and clearly describes the exercise fractures. International Orthopaedics 25: 43–45. program to allow for study comparison including reporting of exercise adherence. Clifford PC (1980) Fractures of the neck of the humerus: a review of the late results. Injury 12: 91–95. 9ekhj#8hemd 9C\" =Wh] 7\" CYGk[[d CC (&&' J^[ epidemiology of proximal humeral fractures. Acta Orthopaedica Scandinavica 72: 365–371. Davis TR, Buchanan JM (1987) A controlled prospective study of early mobilization of minimally displaced fractures of the distal radial metaphysis. Injury 18: 283–285. de Bruijn HP (1987) Functional treatment of Colles fracture. Journal of Physiotherapy 2011 Vol. 57 – © Australian Physiotherapy Association 2011 81

Research Acta Orthopaedica Scandinavica 223: 1–95. desirable. A prospective randomised study of two protocols. Injury 33: 801–806. de Morton NA (2009) The PEDro scale is a valid measure of the methodological quality of clinical trials: a demographic Lozano Calderón SA, Souer S, Mudgal C, Jupiter JB, Ring study. Australian Journal of Physiotherapy 55: 129–133. D (2008) Wrist mobilization following volar plate fixation of fractures of the distal part of the radius. Journal of Bone and Dias JJ, Wray CC, Jones JM, Gregg PJ (1987) The value of early Joint Surgery–American Version 90: 1297–1304. mobilisation in the treatment of Colles’ fractures. Journal of Bone and Joint Surgery–British Volume 69-B: 463–467. Lundberg BJ, Svenungson-Hartwig E, Wikmark R (1979) Independent exercises versus physiotherapy in nondisplaced Gaine WJ, Beardsmore J, Fahmy N (1998) Early active proximal humeral fractures. Scandinavian Journal of mobilisation of volar plate avulsion fractures. Injury 29: 589– Rehabilitation Medicine 11: 133–136. 591. Lyngcoln A, Taylor N, Pizzari T, Baskus K (2005) The Grønlund B, Harreby MS, Kofoed R, Rasmussen L (1990) The relationship between adherence to hand therapy and short- importance of early exercise therapy in the treatment of term outcome after distal radius fracture. Journal of Hand Colles’ fracture. A clinically controlled study. Ugeskr Laeger Therapy 18: 2–8. 152: 2491–2493. MacDermid JC, Richards RS, Roth JH (2001) Distal radius Handoll HH, Gibson JN, Madhok R (2003) Interventions for fracture: a prospective outcome study of 275 patients. treating proximal humeral fractures in adults.[update of Journal of Hand Therapy 14: 154–169. Cochrane Database Syst Rev. 2002;(2):CD000434; PMID: 12076396]. Cochrane Database of Systematic Reviews: Maciel JS, Taylor NF, McIlveen C (2005) A randomised clinical CD000434. trial of activity-focussed physiotherapy on patients with distal radius fractures. Archives of Orthopaedic Trauma Surgery Handoll HH, Madhok R, Howe TE (2006) Rehabilitation for 125: 515–520. distal radial fractures in adults[update in Cochrane Database Syst Rev. 2006;3:CD003324; PMID: 16856004]. Cochrane Maher CG, Sherrington C, Herbert RD, Moseley AM, Elkins Database of Systematic Reviews: CD003324. M (2003) Reliability of the PEDro scale for rating quality of randomized controlled trials. Physical Therapy 83: 713–721. Higgins J, Thompson S, Deeks J, Altman D (2003) Measuring inconsistency in meta-analysis. BMJ 327: 557–560. McAuliffe TB, Hilliar KM, Coates CJ, Grange WJ (1987) Early mobilisation of Colles’ fractures. A prospective trial. Journal Hodgson SA, Mawson SJ, Saxton JM, Stanley D (2007) of Bone and Joint Surgery–British Volume 69-B: 727–729. Rehabilitation of two-part fractures of the neck of the humerus (two-year follow-up). Journal of Shoulder and Michlovitz SL, LaStayo PC, Alzner S, Watson E (2001) Distal Elbow Surgery 16: 143–145. radial fractures: therapy practice patterns. Journal of Hand Therapy 14: 249–257. Hodgson SA, Mawson SJ, Stanley D (2003) Rehabilitation after two-part fractures of the neck of the humerus. Journal Millett PJ, Rushton N (1995) Early mobilization in the treatment of Bone & Joint Surgery–British Volume 85-B: 419–422. of Colles’ fracture: a 3 year prospective study. Injury 26: 671–675. Juni P, Witschi A, Bloch R, Egger M (1999) The hazards of scoring the quality of clinical trials for meta-analysis. JAMA Oskarsson GV, Hjall A, Aaser P (1997) Physiotherapy: an 282: 1054–1060. overestimated factor in after-treatment of fractures in the distal radius? Archives of Orthopaedic Trauma Surgery 116: Kay S, McMahon M, Stiller K (2008) An advice and exercise 373–375. program has some benefits over natural recovery after distal radius fracture: a randomised trial. Australian Journal of Pasila M, Karaharju EO, Lepisto PV (1974) Role of physical Physiotherapy 54: 253–259. therapy in recovery of function after Colles’ fracture. Archives of Physical Medicine Rehabilitation 55: 130–134. Kelsey JL, Browner WS, Seeley DG, Nevitt MC, Cummings SR (1992) Risk factors for fractures of the distal forearm and Revay S, Dahlstrom M, Dalen N (1992) Water exercise versus proximal humerus. American Journal of Epidemiology 135: instruction for self-training following a shoulder fracture 477–489. International Journal of Rehabilitation Research 15: 327–333. Khan KS, ter Riet G, Popay J, Nixon J, Kleijnen J (2001) Stage II Sanders KM, Nicholson GC, Ugoni AM, Pasco JA, Seeman E, Conducting the Review. Phase 5–Study quality assessment. Kotowicz MA (1999) Health burden of hip and other fractures In, Khan KS, ter Riet G, Glanville H, Sowden AJ, Kleijnen J in Australia beyond 2000: Projections based on the Geelong (Eds) CRD Report Number 4. York: NHS Centre for Reviews Osteoporosis Study. The Medical Journal of Australia 170: and Dissemination, University of York. 467–470. Krischak GD, Krasteva A, Schneider F, Gulkin D, Gebhard Taylor NF, Dodd KJ, Shields N, Bruder A (2007) Therapeutic F, Kramer M (2009) Physiotherapy after volar plating of exercise in physiotherapy practice is beneficial: a summary wrist fractures is effective using a home exercise program. of systematic reviews 2002–2005. Australian Journal of Archives of Physical Medicine Rehabilitation 90: 537–544. Physiotherapy 53: 7–16. Larsen CF, Lauritsen J (1993) Epidemiology of acute wrist MWa[Å[bZ7;\"CYGk[[dCC(&&&J^[heb[e\\f^oi_ej^[hWfo trauma. International Journal of Epidemiology 22: 911–916. and clinical predictors of outcome after fracture of the distal radius. Journal of Bone and Joint Surgery–British Volume 82- Lefevre-Colau MM, Babinet A, Fayad F, Fermanian J, Anract P, B: 972–976. Roren A, et al (2007) Immediate mobilization compared with conventional immobilization for the impacted nonoperatively Watt CF, Taylor NF, Baskus K (2000) Do Colles’ fracture patients treated proximal humeral fracture. Journal of Bone and Joint benefit from routine referral to physiotherapy following cast Surgery–American Version 89-A: 2585–2590. removal? Archives of Orthopaedic Trauma Surgery 120: 413–415. Lichtman DM, Bindra RR, Boyer MI, Putnam MD, Ring D, Slutsky DJ, et al (2010) American Academy of Orthopaedic World Health Organisation (2001) ICF: international Surgeons clinical practice summary: treatment of distal classification of functioning, disability and health. Geneva: radius fractures. Journal of the American Academy of World Health Organization. Orthopaedic Surgeons 18: 180–189. Websites Liow RY, Cregan A, Nanda R, Montgomery RJ (2002) Early mobilisation for minimally displaced radial head fractures is www.pedro.org.au 82 Journal of Physiotherapy 2011 Vol. 57 – © Australian Physiotherapy Association 2011

Appraisal Clinical Practice Guidelines Falls prevention 1SFWFOUJPOPG'BMMTJO0MEFS1FSTPOT Description: These guidelines present evidence for the screening and assessment of older persons for falls risk, Latest update: 2010. Next update: Not indicated. and provide evidence-based guidelines for intervention Patient group: Older adults living in the community and to prevent falls in older persons living in the community residential aged care. Intended audience: Clinicians in or residential aged care facilities, and in those with contact with older persons. Additional versions: This is an cognitive impairment. A clinical algorithm is presented update of the 2001 guidelines. Patient education resources describing a systematic process of decision-making and and summary documents are available at the website intervention that should occur in the management of older below. Expert working group: The working party of 12 persons who present in a clinical setting with recurrent consisted of representatives from: the American Academy falls, difficulty walking, or in the emergency department of Orthopaedic Surgeons (AAOS), the American Board of following a fall. Latest evidence for screening of falls Internal Medicine, the American College of Emergency risk is presented. Multifactorial falls risk assessment is Physicians, the American Geriatrics Society, the American advocated, with updated recommendations presented for Medical Association (AMA), the American Occupational assessment. Evidence for multifactorial/multicomponent Therapy Association, the American Physical Therapy interventions are outlined, including recommendations that Association (APTA), the American Society of Consultant all interventions for community-residing persons include an Pharmacists, the British Geriatrics Society, the John A exercise component. The website contains detailed clinical Hartford Foundation Institute for Geriatric Nursing at practice guidelines, an accompanying summary article, a New York University, and the National Association for 4-page summary of recommendations, and two patient Home Care and Hospice. Funded by: American Geriatrics education resources. Society. Consultation with: Representatives of over 20 British and American medical societies, including the Sandra Brauer APTA and the Chartered Society of Physiotherapists. The University of Queensland, Australia Approved by: Several societies including American Geriatrics Society, British Geriatrics Society, APTA, AMA, and the AAOS. Location: All material related to the guidelines are available at: http://www.americangeriatrics. org/health_care_professionals/clinical_practice/clinical_ guidelines_recommendations/2010/ Knee pain ,OFFQBJO Latest update: June 2010. Next update: To be considered Description: This 35-page document presents evidence- for review in 2014. Patient group: Patients presenting based clinical practice guidelines on the clinical course, with knee pain and mobility impairments associated risk factors, diagnosis, classification, outcome measures, with meniscal and articular cartilage lesions. Intended activity limitation measures, and physical therapy audience: Orthopaedic physical therapy clinicians who interventions for people presenting with knee pain. diagnose and manage patients with knee pain, academic The guidelines are presented within an International and clinical instructors, policy makers, payers, and claims Classification of Functioning Disability and Health (ICF) reviewers. Additional versions: Nil. Expert working framework. It begins with a 1-page summary of all guideline group: The guidelines were produced by 4 authors and 14 recommendations. The prevalence and pathoanatomical content experts. They consisted of 14 physiotherapists and features are presented. Signs, symptoms and potential 4 doctors from the USA appointed as content experts by conditions to consider in the differential diagnosis are also the Orthopaedic section of the American Physical Therapy outlined. Measurement properties and details of tools to Association. Funded by: Not indicated. Consultation with: measure physical impairments, activity restriction and Consultants from a variety of fields such as epidemiology, participation limitations specific to a person with knee pain orthopaedic surgery, and sports physical therapy served as are presented. Evidence for the efficacy of physical therapy reviewers of early drafts of the guideline. Approved by: interventions are detailed and include progressive knee Orthopaedic section of the American Physical Therapy motion, weightbearing, return to activity, rehabilitation Association. Location: Logerstedt DS et al (2010) Knee programs, therapeutic exercises, and neuromuscular pain and mobility impairments: meniscal and articular electrical stimulation. All 144 cited references are listed at cartilage lesions. J Orthop Sports Phys Ther 40: A1–35. the end of the document. http://www.jospt.org/issues/id.2459/article_detail.asp Sandra Brauer The University of Queensland, Australia 130 Journal of Physiotherapy 2011 Vol. 57 – © Australian Physiotherapy Association 2011

Appraisal Clinimetrics International Standards for the Neurological Classification of Spinal Cord Injury Description the sensory and motor tests are variable. One study reported a median (interquartile) Kappa value for assigning sensory The International Standards for Classification of Spinal and motor scores of 0.59 (0.48 to 0.70) and 0.65 (0.57 to Cord Injury (ISCSCI) are widely used to classify the 0.69), respectively (Jonsson et al 2000) while another type and extent of a spinal cord injury (SCI) (American study reported inter-reliability coefficients (ICCs) (95% Spinal Injury Association 2003). The standards are based CI) ranging from 0.69 to 1.00 (0.25 to 1.00) (Marino et al on comprehensive sensory and motor tests and are used 2008). The validity of the motor scores have been verified to derive right and left sensory and motor levels. Sensory in studies which have found that these scores can predict and motor deficits can be summarised by tallying scores in motor Functional Independence Measure scores reasonably different ways. For example, strength deficits in the upper well provided the upper and lower limbs scores are treated limbs can be summarised by tallying the results of the separately (R2 = 0.71) (Marino et al 2004). upper limb motor tests (maximal score is 50). Importantly, the sensory and motor tests are also used to classify the The reliability of correctly classifying patients using the type of spinal cord injury using the American Spinal Injury AIS has also been investigated (Cohen et al 1994, Cohen et Association Impairment Scale (AIS). The important feature al 1996). ICC for assigning total motor and sensory scores of the AIS is its definitions of complete and incomplete is very high (0.91 to 0.99) with little variability due to raters’ SCI. An SCI is only classified as incomplete if there is profession or years of experience. The inter-reliability of some sensory or motor function in the S4/5 segments, ie, correctly classifying patients is more variable with higher if a person has anal sensation or the ability to voluntarily reliability for complete paraplegia (1.00) than incomplete contract the anal sphincter. tetraplegia (0.91). Another recent study indicated an overall 11% error rate in assigning AIS classifications from trained Validity and Reliability: The ISCSCI has good face validity staff, with a particularly high 46% error rate in correctly because they were developed by expert and international assigning an AIS D classification (Chafetz et al 2008). consensus over a 20-year period. The Standards have two components: the physical examination and the potential for recovery especially as this is primarily of classification. Reports on the inter-reliability of performing relevance in the first two years after injury when recovery is greatest. Instead, they argue that a classification system Commentary should readily convey a person’s level of disability, which is best gauged by looking at the overall sensory and motor While the ICSCSI are primarily of interest to clinicians deficits. Of course, the tallied sensory and motor scores can working in the area of spinal cord injuries, the sensory be used for this purpose. However, tags of ‘incomplete’ or and motor tests could be relevant to musculoskeletal ‘complete’ SCI which are reliant on S4/5 sensory and motor physiotherapists. The sensory and motor tests provide a function are often misunderstood outside professional concise way of testing each dermatome and myotome. For spheres. example, a three-point testing system is used to test light touch and pinprick for each of the 28 dermatomes on each Lisa Harvey side of the body spanning from C2 to S4/5. In addition, one The University of Sydney, Australia key muscle is tested using standard manual muscle testing procedures to evaluate ten important myotomes, namely the Daniel Graves C5 to T1 and L2 to S1 myotomes. Baylor College of Medicine, Houston, USA An AIS assessment form is freely available in a one References page document (http://www.asia-spinalinjury.org/ publications/2006_Classif_worksheet.pdf). This makes the American Spinal Injury Association (2003) Reference manual assessment appear misleadingly simple. In reality, there for the international standards for neurological classification are many complexities involved in correctly testing and of spinal cord injury. Chicago: American Spinal Injury defining a person’s AIS which leads to confusion and a high Association. error rate especially in untrained staff (Chafetz et al 2008). There are also a number of anomalies and ambiguities Chafetz RS et al (2008) J Spinal Cord Med 31: 538–542. which are yet be resolved (Waring III et al 2010). There is a comprehensive online training module put out by the Cohen M, Bartko J (1994) In Ditunno J et al (Eds): Reference American Spinal Injuries Association but it is not freely manual for the Neurological and Functional Classification available. of Spinal Cord Injury. Chicago: American Spinal Injury Association. It is unfortunate that classification by the AIS requires S4/5 sensory and motor tests. These tests are intrusive and involve Cohen M et al (1996) Top Spinal Cord Inj Rehabil 1: 15–31. an assessment of deep anal sensation. The rationale for the reliance on S4/5 is debated in SCI international spheres. Jonsson M et al (2000) Spinal Cord 38: 675–679. Advocates argue that S4/5 sensation or motor function is a strong predictor of future recovery and therefore essential Marino R, Graves D (2004) Arch Phys Med Rehabil 85: 1804– to the classification standards. Others argue that the ISCSCI 1810. should not be solely concerned with capturing people’s Marino RJ et al (2008) J Spinal Cord Med 31: 166–170. Waring III WP et al (2010) J Spinal Cord Med 33: 346–352. Journal of Physiotherapy 2011 Vol. 57 – © Australian Physiotherapy Association 2011 129

Appraisal Clinical Practice Guidelines Falls prevention 1SFWFOUJPOPG'BMMTJO0MEFS1FSTPOT Description: These guidelines present evidence for the screening and assessment of older persons for falls risk, Latest update: 2010. Next update: Not indicated. and provide evidence-based guidelines for intervention Patient group: Older adults living in the community and to prevent falls in older persons living in the community residential aged care. Intended audience: Clinicians in or residential aged care facilities, and in those with contact with older persons. Additional versions: This is an cognitive impairment. A clinical algorithm is presented update of the 2001 guidelines. Patient education resources describing a systematic process of decision-making and and summary documents are available at the website intervention that should occur in the management of older below. Expert working group: The working party of 12 persons who present in a clinical setting with recurrent consisted of representatives from: the American Academy falls, difficulty walking, or in the emergency department of Orthopaedic Surgeons (AAOS), the American Board of following a fall. Latest evidence for screening of falls Internal Medicine, the American College of Emergency risk is presented. Multifactorial falls risk assessment is Physicians, the American Geriatrics Society, the American advocated, with updated recommendations presented for Medical Association (AMA), the American Occupational assessment. Evidence for multifactorial/multicomponent Therapy Association, the American Physical Therapy interventions are outlined, including recommendations that Association (APTA), the American Society of Consultant all interventions for community-residing persons include an Pharmacists, the British Geriatrics Society, the John A exercise component. The website contains detailed clinical Hartford Foundation Institute for Geriatric Nursing at practice guidelines, an accompanying summary article, a New York University, and the National Association for 4-page summary of recommendations, and two patient Home Care and Hospice. Funded by: American Geriatrics education resources. Society. Consultation with: Representatives of over 20 British and American medical societies, including the Sandra Brauer APTA and the Chartered Society of Physiotherapists. The University of Queensland, Australia Approved by: Several societies including American Geriatrics Society, British Geriatrics Society, APTA, AMA, and the AAOS. Location: All material related to the guidelines are available at: http://www.americangeriatrics. org/health_care_professionals/clinical_practice/clinical_ guidelines_recommendations/2010/ Knee pain ,OFFQBJO Latest update: June 2010. Next update: To be considered Description: This 35-page document presents evidence- for review in 2014. Patient group: Patients presenting based clinical practice guidelines on the clinical course, with knee pain and mobility impairments associated risk factors, diagnosis, classification, outcome measures, with meniscal and articular cartilage lesions. Intended activity limitation measures, and physical therapy audience: Orthopaedic physical therapy clinicians who interventions for people presenting with knee pain. diagnose and manage patients with knee pain, academic The guidelines are presented within an International and clinical instructors, policy makers, payers, and claims Classification of Functioning Disability and Health (ICF) reviewers. Additional versions: Nil. Expert working framework. It begins with a 1-page summary of all guideline group: The guidelines were produced by 4 authors and 14 recommendations. The prevalence and pathoanatomical content experts. They consisted of 14 physiotherapists and features are presented. Signs, symptoms and potential 4 doctors from the USA appointed as content experts by conditions to consider in the differential diagnosis are also the Orthopaedic section of the American Physical Therapy outlined. Measurement properties and details of tools to Association. Funded by: Not indicated. Consultation with: measure physical impairments, activity restriction and Consultants from a variety of fields such as epidemiology, participation limitations specific to a person with knee pain orthopaedic surgery, and sports physical therapy served as are presented. Evidence for the efficacy of physical therapy reviewers of early drafts of the guideline. Approved by: interventions are detailed and include progressive knee Orthopaedic section of the American Physical Therapy motion, weightbearing, return to activity, rehabilitation Association. Location: Logerstedt DS et al (2010) Knee programs, therapeutic exercises, and neuromuscular pain and mobility impairments: meniscal and articular electrical stimulation. All 144 cited references are listed at cartilage lesions. J Orthop Sports Phys Ther 40: A1–35. the end of the document. http://www.jospt.org/issues/id.2459/article_detail.asp Sandra Brauer The University of Queensland, Australia 130 Journal of Physiotherapy 2011 Vol. 57 – © Australian Physiotherapy Association 2011

Scrivener et al: Quantifying exercise dosage in rehabilitation Many participants in inpatient rehabilitation can quantify their exercise dosage accurately: an observational study Katharine Scrivener1,2, Catherine Sherrington2, Karl Schurr1 and Daniel Treacy1 1Bankstown-Lidcombe Hospital, 2The George Institute for Global Health, The University of Sydney Australia Question: Are inpatients undergoing rehabilitation who appear able to count exercises able to quantify accurately the amount of exercise they undertake? Design: Observational study. Participants: Inpatients in an aged care rehabilitation unit and a neurological rehabilitation unit, who appeared able to count their exercises during a 1–2 min observation by their treating physiotherapist. Measurements: Participants were observed for 30 min by an external observer while they exercised in the physiotherapy gymnasium. Both the participants and the observer counted exercise repetitions with a hand-held tally counter and the two tallies were compared. Results: Of the 60 people admitted for aged care rehabilitation during the study period, 49 (82%) were judged by their treating therapist to be able to count their own exercise repetitions accurately. Of the 30 people admitted for neurological rehabilitation during the study period, 20 (67%) were judged by their treating therapist to be able to count their repetitions accurately. Of the 69 people judged to be accurate, 40 underwent observation while exercising. There was excellent agreement between these participants’ counts of their exercise repetitions and the observers’ counts, ICC (3,1) of 0.99 (95% CI 0.98 to 0.99). Eleven participants (28%) were in complete agreement with the observer. A further 19 participants (48%) varied from the observer by less than 10%. Conclusion: Therapists were able to identify a group of rehabilitation participants who were accurate in counting their exercise repetitions. Counting of exercise repetitions by therapist-selected patients is a valid means of quantifying exercise dosage during inpatient rehabilitation. <4DSJWFOFS, 4IFSSJOHUPO$ 4DIVSS ,  5SFBDZ %   .BOZ QBSUJDJQBOUT JO JOQBUJFOU SFIBCJMJUBUJPO DBO RVBOUJGZ UIFJS FYFSDJTF EPTBHF BDDVSBUFMZ BO PCTFSWBUJPOBMTUVEZJournal of Physiotherapyo> ,FZXPSET: Rehabilitation, Physiotherapy, Dosage, Exercise Introduction 2001) and establishing the impact of a higher dose versus lower doses of rehabilitation intervention is an important Accurate quantification of the nature and dose of the aspect of this investigation (Kwakkel et al 2004). Guidelines interventions provided in rehabilitation settings is an for complex interventions suggest that a clear description important challenge for both clinicians and researchers. of the intervention needs to be provided to enable others For rehabilitation participants to reacquire skilled motor to replicate the intervention clinically, replicate the performance, a significant amount of repetitive task practice study, and combine evidence (Craig et al 2008). To date, is required (Butefisch et al 1995, Classen et al 1998). the standard method used to quantify exercise dosage Studies of neural plasticity have shown that repetitive task is the time rehabilitation participants spend in therapy training can change cortical organisation (Plautz et al 2003) (Cooke et al 2010, French et al 2008, Galvin et al 2008, however, the dose of repetitive task practice often available Kwakkel et al 2004). However, time spent in therapy is in therapy sessions is unlikely to be sufficient to induce unlikely to be an accurate measure of the actual exercise cortical changes (Lang et al 2009). Some rehabilitation units dose for individual rehabilitation participants. The dose seek to maximise the dose of repetitive task practice by the and intensity of exercise each participant completes in a prescription of task-related exercises to be undertaken daily set time can vary significantly. In addition, measurement during the inpatient stay in the rehabilitation gymnasium of total time spent in therapy may not take into account (Olivetti et al 2007, Sherrington et al 2003). Unfortunately, rests and other interruptions to therapy sessions. In fact, an therapists’ estimates of the amount of exercise that occurs observational study of activity levels in rehabilitation found in rehabilitation have been shown to be poor (Bagley et that rehabilitation participants complete relevant activities al 2009, Collier and Bernhardt 2008, Lang et al 2007). only 45% of the time they are in a therapy area (Mackey et More accurate knowledge of exercise dosage may assist al 1996). This suggests that studies using time as a measure in intervention prescription and assessment of goal of exercise dosage may be overestimating actual exercise achievement. Thus a method for objectively recording the substantially. amount of exercise that participants complete is required. A count of each repetition of exercise the participant Establishing the effectiveness of different components completes may be a more accurate measure of exercise of rehabilitation or ‘unpacking the black box’ has been dosage. This would capture the work the participant identified as a key research area (Langhorne and Duncan completes and not any accessory activities nor resting Journal of Physiotherapy 2011 Vol. 57 – © Australian Physiotherapy Association 2011 117

Research time. Several published studies have used repetitions to exercise dosage. So, the focus of the study was whether measure dosage (Lang et al 2009, Lang et al 2007, Nugent those patients who seem able to count accurately and were et al 1994). These studies have used either a therapist or an left to count exercises independently for extended periods, external observer to record repetitions of exercise. External were truly accurate when observed closely. observation is a labour-intensive process that would be impractical for studies with large cohorts or for daily The participants who were observed were chosen randomly clinical practice. An alternative strategy is for rehabilitation from all patients admitted to the two rehabilitation units participants to count their own exercise repetitions while during the study period and who were judged by therapists completing their prescribed exercise. This method has to be able to count accurately (based on a short period been implemented in several rehabilitation units including of observation). Random selection was achieved using Bankstown-Lidcombe Hospital in Sydney, Australia. It is a random number generator on a computer. A research usual clinical practice at Bankstown-Lidcombe Hospital assistant who did not work clinically on the rehabilitation for rehabilitation patients to count their own exercise units completed this process. This research assistant repetitions with a hand-held tally counter if they are able scheduled the observation sessions based on observer and to do this. These exercise totals are recorded and used for participant availability. When scheduling the sessions she clinical decision-making and documentation. ensured that the observer was not the participant’s treating therapist. The aim of this study was to determine if rehabilitation participants assessed by their therapist as being able to count Participants were unaware of their inclusion in the study their repetitions of exercise accurately (based on a short and did not know they were being observed. The treating period of observation) are able to count exercise repetitions therapists did not know the timing of observations and accurately when observed more closely over a longer period were also unaware which aged care rehabilitation patients of time. The validity of exercise dose quantification by had been selected for the study. This was to ensure that therapist-selected rehabilitation participants was determined increased therapist time was not devoted to the participant by comparing the number of exercise repetitions counted by during the observation period. participants to the number counted by an external observer. Therefore, the research question for this study was: Measurement Can therapist-identified rehabilitation participants Prior to inclusion into the study, the treating physiotherapist accurately quantify their exercise dosage during collected eligible participants’ demographic data. The inpatient rehabilitation? Mini-Mental State Examination was completed as part of usual practice on admission to each rehabilitation unit but Method two participants were unable to complete this test due to limited English language skills. The treating therapist also Design rated the participants’ level of disability with the Modified Rankin Scale. An observational study was conducted involving people admitted to inpatient rehabilitation at Bankstown-Lidcombe An observer, who was a physiotherapist but not the Hospital, Sydney during the six-week study period participant’s treating therapist, covertly counted each beginning in November 2009. Participants were included participant’s exercise repetitions via direct observation from two rehabilitation units: aged care rehabilitation and in the rehabilitation gymnasium. Direct observation was stroke/neurological rehabilitation. We sought to observe chosen as it is considered the gold standard means of 20 participants from each unit who were deemed likely to observing behaviour over time (Ryan et al 1995). The covert be able to count exercise repetitions accurately while they observation of the participant’s exercise was for a period of exercised. 30 minutes. The observer and the participant each counted the exercise repetitions using a hand-held tally counter. Participants Participants were instructed to count all repetitions of their exercise accurately. At the end of the 30-minute observation Patients were included in the observational phase of the session, the observer recorded the two tallies: the observer’s study if they were a current rehabilitation inpatient and tally and the participant’s tally. their treating physiotherapist judged that they were not obviously inaccurate in counting their exercise repetitions. Participants were observed in the rehabilitation gymnasium, The therapists’ decision regarding ability to count was used located adjacent to the two rehabilitation wards. Most clinically to determine which patient’s results were trusted participants attended the gym twice daily and participated and therefore documented. Therapists observed the patients in a variety of exercise groups, eg, the Upper Limb Group counting their exercise repetitions during semi-supervised or Standing Balance Group. Observations occurred at or group sessions for a short period, normally 1–2 minutes. different times of day and in a variety of therapy contexts This was to determine if there was any obvious inaccuracy including the exercise groups. Different exercises were in the patient’s counting ability. Common inaccuracies are observed in the study including task-related upper limb counting multiple times for each exercise, or inconsistent practice (eg, reaching and manipulation) or lower limb counting of each repetition of exercise, meaning that practice (eg, sit-to-stand and walking), balance training, patients miss repetitions. and strength exercises. The number of exercises completed by participants varied depending on the participants’ This study aimed to reflect clinical practice. Therefore physical abilities and the exercise type. Some participants those patients who were obviously inaccurate in counting were observed in an exercise circuit where they changed were excluded from the study. Clinically, these individuals exercises every six minutes, and others carried out the same are not asked to count their exercise independently. Instead exercise for the 30-minute period. therapists, therapy assistants, or family members tally 118 Journal of Physiotherapy 2011 Vol. 57 – © Australian Physiotherapy Association 2011

Patients admitted to aged care Scrivener et al: Quantifying exercise dosage in rehabilitation rehabilitation unit (n = 60) FWj_[djiWZc_jj[Zjeijhea[%d[khebe]_YWb rehabilitation unit (n = 30) Patients deemed able to count repetitions Patients deemed able to count repetitions after brief observation by the treating after brief observation by the treating physiotherapist (n = 49) physiotherapist (n = 20) Randomly selected for inclusion (n = 20) Included (n = 20) Measured ability to tally exercise repetitions (n = 40) 'JHVSF Flow of participants through the study. Data analysis 5BCMF Characteristics of the study participants. Criterion-related validity was assessed by investigating the Characteristic n = 40 level of agreement of the participant-and observer-counted exercises using the intraclass correlation coefficient (ICC). Age (years), mean (SD) 73 (12) The 3,1 form was used as we considered it to be the most Gender n female (%) 17 (43) appropriate form for this research question. An ICC of Diagnosis greater than 0.75 is generally considered to represent excellent agreement (Fleiss 1986). The level of agreement Stroke, n (%) 17 (43) of participants with the observer was also calculated Other neurological, n (%) 5 (12) by tallying the proportion of participants in complete Orthopedic, n (%) 9 (23) agreement with the observer. The proportion of participants Frailty, n (%) 6 (15) in close agreement with the observer (ie, absolute percentage Amputee, n (%) 3 (8) error up to 5%, 10%, 20%, and 30%) was also calculated. Cognition (MMSE), mean, (SD) 26 (5) In addition, Pearson’s r was used to assess the degree of Disability (mRS), mean (SD) 3.2 (0.8) correlation between each participant’s counting ability Able to sit independently, n (%) 39 (98) (calculated by the percentage agreement for their count Able to sit-to-stand independently, n (%) 23 (58) compared to the observer) and their cognition (assessed by the Mini-Mental State Examination), their age, and their MMSE = Mini-Mental State Examination, mRS = Modified Rankin disability level (as assessed by the Modified Rankin Scale). Scale Results count exercise repetitions. All 20 were included in the 30-minute observation component of the study (see 'MPXPGQBSUJDJQBOUTBOEUIFSBQJTUTUISPVHIUIF Figure 1). Although there was no random selection of study the neurological rehabilitation participants, blinding of therapists was maintained as the research assistant was the Ninety people were admitted to the rehabilitation units only person aware of the number of included participants. during the study period: 60 to the aged care rehabilitation All participants were observed within five days of inclusion. unit and 30 to neurological rehabilitation unit. Of the 60 patients admitted for aged care rehabilitation, 49 (82%) As shown in Table 1, the participants had a range of were judged by their treating therapist to be able to count diagnoses, with stroke (43%) being the most common their own exercise accurately. Twenty of these patients diagnosis. Participants had reasonable cognition as were randomly selected for inclusion in the 30-minute measured by the Mini Mental State Examination, with an observation component of the study. Of the 30 patients average score of 26 out of a possible 30 points, although admitted for neurological rehabilitation, 20 (67%) were judged by their treating therapist to be able to accurately Journal of Physiotherapy 2011 Vol. 57 – © Australian Physiotherapy Association 2011 119

Research 5BCMF Number of participants in complete agreement with the observer, and the number and cumulative total of participants in agreement with the observer to within the specified range of percentage error. Error of participants Aged care Neurological All participants compared to observer participants participants Absolute n (%) n (%) n (%) Cumulative percentage percentage 6 (30) 5 (25) 11 (27.5) 0% 6 (30) 6 (30) 12 (30) 27.5 1–5% 3 (15) 4 (20) 7 (17.5) 57.5 6–10% 3 (15) 0 (0) 3 (7.5) 75 11–20% 1 (5) 3 (15) 4 (10) 82.5 21–30% 1 (5) 2 (10) 3 (7.5) 92.5 >30% 100 'JHVSF The relationship between participant and observer exercise count values (n = 40). scores ranged from 13 to 30. The average Modified Rankin observation periods, ranged from a minimum of 4 to a Scale score was 3.2 out of 6 points, indicating that typically maximum of 369 repetitions. The average number of the participants were limited by their disability but did repetitions observed was 113 (SD 100). not need assistance to walk. Participants were observed at different time points in their rehabilitation, with time from Validity of participant counting admission to inclusion in the study varying from 2 to 46 days. The intraclass correlation coefficient (ICC) (3,1) between participant and observer exercise counts was 0.99 (95% CI The therapists determining the accuracy of participant 0.98 to 0.99). This suggests that there is excellent agreement counting varied in clinical experience from 0.5 years between the two counts of exercise repetitions. to greater than 20 years of experience. The number of exercise repetitions, which were counted in the 30-minute The level of agreement for neurological rehabilitation participants was ICC (3,1) 0.99 (95% CI 0.98 to 1.00). The 120 Journal of Physiotherapy 2011 Vol. 57 – © Australian Physiotherapy Association 2011

Scrivener et al: Quantifying exercise dosage in rehabilitation agreement for aged care rehabilitation participants was ICC A strength of this study is the blinding of both participant (3,1) 0.98 (95% CI 0.95 to 0.99). and therapist to when the covert observation was occurring. In addition, a variety of therapy contexts were observed, The accuracy in counting varied between the participants, meaning that the results are representative of daily therapy as shown in Table 2, with 11 participants (28%) being in practice. The participants were also observed at various complete agreement with the observer. Moreover a further time points in their rehabilitation. Another strength is that 19 participants (48%) were within 10% of the observer’s the method used to identify patients who are able to count is total. There were 3 participants (8%) with more than a 30% simple and efficient so it can be replicated clinically. differential. The most inaccurate participant underestimated the exercise tally by 47% (17 repetitions). Again there was A limitation of this study could be the 30-minute observation minimal difference in error rates between neurological period. This represents a small proportion of time the and aged care participants. The relationship between the participant would be in therapy each day at Bankstown- observer and participant counts can be seen more clearly Lidcombe Hospital. However, for pragmatic reasons a in Figure 2. substantial yet not exhaustive time period was chosen. It is reasonable to believe that if a participant is able to count in The participants’ ability to count exercise repetitions did this period, that skill would be transferable to other times. It not correlate with their cognition (r = 0.16, p = 0.35), age (r is also important to consider that other studies investigating = 0.12, p = 0.46), or level of disability (r = 0.16, p = 0.34). activity levels in rehabilitation have found that on average participants spend 39 minutes each day in physiotherapy Discussion and occupational therapy combined (Kwakkel et al 2004), which would indicate that the 30-minute timeframe is This study provides evidence that therapist-selected typical of usual clinical practice at many sites. rehabilitation patients are able to count their repetitions of exercise accurately. The high level of agreement (ICC = 0.99, All observations were completed in the rehabilitation 95% CI 0.98 to 0.99) between therapist-selected participant gymnasium with therapy staff present. The exercise count data and the data from an external observer, and the observed was semi-supervised meaning therapists may low percentage errors suggest that therapist-selected patient sometimes provide feedback and check on progress count data may be used in place of observer data in future including current participant exercise tally. No independent research. exercise, eg, exercise that occurred outside the therapy setting, was observed. However, due to the nature of the The 30-minute observation component of this study included gymnasium environment and the fact that participants were only those participants deemed by their treating therapist exercising alone but in the presence of others, it is possible likely to be accurate in counting exercise repetitions. The that the results may be extrapolated to home/room based high level of agreement found by this study suggests that programs. therapists demonstrate good judgement regarding the ability of rehabilitation patients to count exercise repetitions Another limitation of the study is the low power to detect accurately. The observation of a patient counting for a small factors that influence the accuracy of exercise repetition period (1–2 minutes) to look for obvious errors in counting counting. We did not find strong correlations between can be used by therapists to determine if the patient is able accuracy of exercise repetition counting and cognition, age, to count accurately. or disability level. Future research with a larger sample could further investigate predictors of accurate exercise It is often perceived by clinicians that rehabilitation repetition counting. patients with neurological diagnoses have less ability to concentrate and multi-task. The results of this study In conclusion, this study indicates that therapist-identified indicate that patients with neurological diagnoses can be rehabilitation participants are able to count their repetitions accurate in counting their exercises repetitions. However, of exercise accurately. This method can be used clinically a lower percentage of participants with neurological or in future research. Q diagnoses met this study’s inclusion criteria (67% for people admitted to the neurological rehabilitation unit vs 82% of Ethics: The Human Research Ethics Committee (Western people admitted to the aged care rehabilitation unit were Zone) of the Sydney South West Area Health Service included). Therefore there were more rehabilitation patients approved this study on the 13th August 2008. Project with neurological diagnoses excluded from the study number QA2008/049. All patients consent to the counting because they were obviously unable to count their exercise and documenting of exercise repetitions as part of their repetitions accurately. usual care on the rehabilitation units. This appears to be the first observational study to analyse Competing interests: Nil. the accuracy of quantification of exercise dosage by patients undertaking rehabilitation. Previous methods of analysing Support: This study was supported by an infrastructure exercise dosage include the use of time in therapy and grant (number 07-08/007) from the Ingham Health behaviour mapping (Kwakkel et al 2004, Mackey et al Research Institute. 1996). Both methods were based on time rather than dosage of exercise. In this study the number of exercise repetitions Acknowledgements: Dharani Khandasamy assisted with observed in the 30-minute sessions varied greatly, with completing observations and data entry. Bankstown- a range of 4 to 369 repetitions. Those studies that only Lidcombe Hospital physiotherapy staff and students assisted consider time will not take into account the rate and with observations including significant contributions from therefore the intensity of exercise. Journal of Physiotherapy 2011 Vol. 57 – © Australian Physiotherapy Association 2011 121

Research Simone Dorsch, Susan Mayo, Lily Jian, James Ruddell, and Kwakkel G, van Peppen R, Wagenaar R, Wood Dauphinee S, Dimyana Tanyous. Richards C, Ashburn A, et al (2004) Effects of augmented exercise therapy time after stroke: a meta–analysis. Stroke Correspondence: Katharine Scrivener, Department of 35: 2529–2539. Physiotherapy, The Bankstown-Lidcombe Hospital, NSW, Australia. Email: [email protected] Lang CE, MacDonald JR, Gnip C (2007) Counting repetitions: an observational study of outpatient therapy for people with References hemiparesis post-stroke. Journal of Neurologic Physical Therapy 31: 3–10. Bagley P, Hudson M, Green J, Forster A, Young J (2009) Do physiotherapy staff record treatment time accurately? An Lang CE, MacDonald JR, Reisman DS, Boyd L, Jacobson observational study. Clinical Rehabilitation 23: 841–845. Kimberley T, Schindler-Ivens SM, et al (2009) Observation of amounts of movement practice provided during stroke Butefisch C, Hummelsheim H, Denzler P, Mauritz K-H (1995) rehabilitation. Archives of Physical Medicine & Rehabilitation Repetitive training of isolated movements improves the 90: 1692–1698. outcome of motor rehabilitation of the centrally paretic hand. Journal of the Neurological Sciences 130: 59–68. Langhorne P, Duncan P (2001) Does the organization of postacute stroke care really matter? Stroke 32: 268–274. Classen J, Liepert J, Wise S, Hallett M, Cohen L (1998) Rapid plasticity of human cortical movement representation Mackey F, Ada L, Heard R, Adams R (1996) Stroke induced by practice. Journal of Neurophysiology 79: 1117– rehabilitation: are highly structured units more conducive 1123. to physical activity than less structured units? Archives of Physical Medicine and Rehabilitation 77: 1066–1070. 9ebb_[h @C\" 8[hd^WhZj @ (&&. J^[ j^[hWfo ¼f_bb½0 WY^_[l_d] treatment dose within a rehabilitation trial. Brain Impairment Nugent JA, Schurr KA, Adams RD (1994) A dose-response 9: 191–197. relationship between amount of weight-bearing exercise and walking outcome following cerebrovascular accident. Cooke EV, Mares K, Clark A, Tallis RC, Pomeroy VM (2010) Archives of Physical Medicine and Rehabilitation 75: 399– The effects of increased dose of exercise-based therapies 402. to enhance motor recovery after stroke: a systematic review and meta-analysis. BMC Medicine 8: 60. Olivetti L, Schurr K, Sherrington C, Wallbank G, Pamphlett P, Kwan M, et al (2007) A novel weight-bearing strengthening Craig P, Dieppe P, Macintyre S, Michie S, Nazareth I, Petticrew program during rehabilitation of older people is feasible M (2008) Developing and evaluating complex interventions: and improves standing up more than a non-weight-bearing the new Medical Research Council guidance. BMJ 337: strengthening program: a randomised trial. Australian 979–983. Journal of Physiotherapy 53: 147–153. Fleiss J (1986) The design and analysis of clinical experiments Plautz E, Barbay S, Frost S, Friel K, Dancause N, Zoubina E, (1st ed.) New York: John Wiley. et al (2003) Post-infarct cortical plasticity and behavioural recovery using concurrent cortical stimulation and French B, Leathley M, Sutton C, McAdam J, Thomas L, Forster rehabilitative training: a feasibility study in primates. A, et al (2008) A systematic review of repetitive functional Neurological Research 25: 801–810. task practice with modelling of resource use, costs and effectiveness. Health Technology Assessment 12: 1–140. Ryan AM, Daum D, Bauman T, Grisez M, Mattimore K, Nalodka T, et al (1995) Direct, indirect and controlled observation and Galvin R, Murphy B, Cusack T, Stokes E (2008) The impact rating accuracy. Journal of Applied Psychology 80: 664–670. of increased duration of exercise therapy on functional recovery following stroke–what is the evidence? Topics in Sherrington C, Lord SR, Herbert RD (2003) A randomised Stroke Rehabilitation 15: 365–377. trial of weight-bearing versus non-weight-bearing exercise for improving physical ability in inpatients after hip fracture. Australian Journal of Physiotherapy 49: 15–22. 122 Journal of Physiotherapy 2011 Vol. 57 – © Australian Physiotherapy Association 2011

Editorial New initiatives in Journal of Physiotherapy Mark Elkins Scientific Editor on behalf of the Editorial Board of Journal of Physiotherapy The Editorial Board of Journal of Physiotherapy endeavours and retrospective registrations. Therefore, the Journal of to publish an informative journal featuring scientifically Physiotherapy is instituting the requirement that systematic rigorous research with clear implications for the clinical reviews be registered, just as we have done with clinical practice of physiotherapy. We also seek to promote the trial registration. At some point in the future, we will journal and to acknowledge the contribution of those who mandate that these registrations are prospective. Therefore support it. In keeping with these aims, the members of we encourage all potential authors to register their clinical the Editorial Board are introducing several changes to the trials and systematic reviews as early as possible. journal. Some changes will facilitate use of the journal by readers. Other changes are most relevant to authors who Cochrane systematic reviews are considering submitting a manuscript to the journal. The remaining changes acknowledge the contribution of The Editorial Board has also changed its policy regarding supporters of the journal. Cochrane systematic reviews. Although the publisher of Cochrane reviews allows them to be co-published in another ScienceDirect journal, Cochrane reviews have not been accepted by the Journal of Physiotherapy in the past. We have now reversed One important change is that the journal has been that policy. Cochrane reviews, if suitably condensed, will made available in digitised form from ScienceDirect to be considered for co-publication. However, publication institutional subscribers. This will enhance the visibility in the Cochrane Library does not guarantee acceptance of existing and future papers in the journal. It will also and priority will still be given to reviews that identify facilitate use of the journal, by providing such facilities as substantial data and draw important clinical implications hyperlinks within the text, automated export of citations, from the results. links to articles cited in the paper, links to other related articles and textbooks, and automated emailing of selected Appraisal index articles. Another benefit to readers is the RSS feed facility, which provides timely updates about the journal content Another change that will benefit readers of both print and that can be read by web-based, desktop-based, or mobile- electronic versions of the journal is the introduction of device-based software. an annual index of items in the Appraisal section of the journal. These include items such as critically appraised Registration of trials and systematic reviews papers, clinimetric appraisals, and appraisals of clinical practice guidelines, books and websites. The annual index The next changes relate to the submission of manuscripts to will appear in the last issue of each calendar year. the journal. Since 2008, the journal has required that trials submitted for publication provide evidence of registration on Paper of the Year a publicly accessible register (Askie et al 2006). This policy had produced some benefits. By appearing on a register, In recognition of the high standard of work performed by our trials are less likely to be overlooked by the authors submitting authors, the Editorial Board has introduced a of systematic reviews and by researchers planning future Paper of the Year award. The winning paper is chosen by trials. However, the majority of benefits of registration a panel of members of our International Advisory Board occur when trials are registered prospectively: researchers who do not have a conflict of interest with any of the papers are obliged to publish completed trials, any selective under consideration. The paper will be chosen from those reporting of outcomes (eg, only favourable outcomes) is published in a given calendar year and will be announced in easily identifiable, and other researchers can know that a the June issue of the following year. The Paper of the Year trial is underway so that it is not duplicated unnecessarily for 2010 has been awarded to the paper entitled Mobility- (World Health Organization 2009). Therefore, in 2012, related disability three months after aged care rehabilitation the journal will begin accepting trials only if they are can be predicted with a simple tool: an observational prospectively registered. Clinical trials are not the only study by Catherine Sherrington and colleagues from type of research for which prospective registration has been Sydney (Sherrington et al 2010). This study found that, recommended. Registration of systematic reviews has also in people who have undergone inpatient rehabilitation, been recommended in the Preferred Reporting Items for ongoing mobility-related disability is common and can be Systematic reviews and Meta-analyses (PRISMA) statement predicted with a high degree of accuracy with a simple tool. (Moher et al 2009). Soon after the PRISMA statement was This information can be used to identify need for service released, its recommendations were adopted by the Journal provision and to tailor intervention to minimise disability. of Physiotherapy (Elkins and Ada 2010). However, the We congratulate Dr Sherrington and her co-authors. recommendation to register systematic reviews has not been achievable due to the absence of a publicly available register. Review process This year, a free, publicly available register for systematic review protocols – known as PROSPERO – has been The final two changes relate to the review process. We are established by the Centre for Reviews and Dissemination in extremely grateful to all the external reviewers for their York, UK. Currently, PROSPERO accepts both prospective evaluations of manuscripts we receive. In recognition of their invaluable support of the journal, we will list the Journal of Physiotherapy 2011 Vol. 57 – © Australian Physiotherapy Association 2011 67

Editorial reviewers – if they agree to be identified – in an annual term. Professor Herbert was elected as Chair by the other list on the journal’s website. This will include reviewers of members of the Editorial Board at the first meeting this both published and rejected papers from the previous year. year. Reviewers will not be linked to the paper or papers they have reviewed. The other change to the review process is We are confident that these changes will improve the that submitting authors will be given an opportunity to interest and accessibility of the Journal of Physiotherapy nominate individuals whom they believe may not provide and look forward to its continued growth and increasing an unbiased review of their manuscript. Up to three non- international presence. reviewers can be identified. References Changes to the Editorial Board Askie L et al (2006) Aust J Physiother 52: 237–239. It is also timely to note recent changes in the membership of the Editorial Board. We acknowledge the contribution Elkins M, Ada L (2010) J Physiother 56: 69. of Professor Kim Bennell, who decided to step down from the Editorial Board this year. Professor Bennell was Moher D et al (2009) J Clin Epidemiol 62: 1006–1012. appointed to the Editorial Board in January 2008 and she became Chair in February 2010. During this time, she has Sherrington C et al (2010) J Physiother 56: 121–127. been a strong advocate for the journal and for the Editorial Board in many forums. We are grateful for her substantial World Health Organization. International Clinical Trials Registry contribution. Professor Rob Herbert was successful in FbWj\\ehc$QkfZWj[Z',Del[cX[h(&&/S$7lW_bWXb[\\hec0^jjf0%% being re-appointed to the board and, at this time, Associate mmm$m^e$_dj%_Yjhf%jh_WbUh[]%[d%_dZ[n$^jcb Professor Michelle Sterling was reappointed for a further Websites www.prisma-statement.org mmm$YhZ$oeha$WY$ka%fheif[he Finding papers from Journal of Physiotherapy From Volume 57 No 1 March 2011, papers from Journal of Physiotherapy are hosted by Elsevier and available from ScienceDirect for institutional subscribers. Back issues (to 2005) are also available on ScienceDirect for these subscribers. Papers are also available to members and individual subscribers from the APA website at JoP.physiotherapy.asn.au. 68 Journal of Physiotherapy 2011 Vol. 57 – © Australian Physiotherapy Association 2011

Appraisal Critically Appraised Papers Patients with chronic obstructive pulmonary disease (COPD) who are not hypoxaemic at rest do not benefit from home oxygen Synopsis activities that caused dyspnoea. Cylinders were identical in appearance and weighed 4.2 kg when full. Outcome Summary of: Moore RP et al (2011) A randomised trial measures: The primary outcome was the dyspnoea domain of domiciliary, ambulatory oxygen in patients with COPD of the Chronic Respiratory Disease Questionnaire (CRDQ). and dyspnoea but without resting hypoxaemia. Thorax 66: Secondary outcomes included dyspnoea measured by the 32–37. [Prepared by Kylie Hill, CAP Editor.] Baseline/Transitional Dyspnoea Index, health-related quality of life measured by the CRDQ and Assessment of Question: In patients with COPD and exertional dyspnoea, Quality of Life Utility Index, mood disturbance measured but without severe hypoxaemia at rest, does domiciliary by the Hospital Anxiety and Depression Scale, functional ambulatory oxygen change dyspnoea, health-related quality exercise capacity measured by the six-minute walk distance, of life, mood, or functional status? Design: Randomised and physical activity measured using a pedometer and self- controlled trial in which the investigators and participants report. Results: The primary outcome was available for were blinded to group allocation and the randomisation 139 of the enrolled patients. No between-group differences sequence was concealed prior to allocation. Setting: In the were demonstrated for any outcome. At 12 weeks dyspnoea, patient’s home with assessments at a tertiary hospital in mean difference 1.1 units (95% CI –0.9 to 3.1), did not Victoria, Australia. Participants: People with stable COPD differ significantly between groups. Using domiciliary who: (i) were ex-smokers on optimal medical treatment, (ii) oxygen for participants with exertional desaturation was had a partial pressure of oxygen in arterial blood > 55mmHg not more predictive of changes in dyspnoea than using air. at rest, and, (iii) reported moderate to severe functional Conclusion: Patients with chronic obstructive pulmonary limitation from dyspnoea. Randomisation of 143 patients disease (COPD) who are not hypoxaemic at rest do not allocated 68 to the cylinder oxygen group and 75 to the benefit from home oxygen. cylinder air group. Interventions: Participants received 12 weeks of either cylinder oxygen (intervention) or cylinder [Mean difference and 95% CIs calculated by the CAP air (control) set at 6 L/min for use during activities of daily Editor] living. Both groups were provided with a trolley/stroller to transport cylinders as well as verbal and written instruction by-patient basis using rigorous methods based on N-of-1 to use the cylinders inside and outside the home during research designs. The cost of such an approach would be offset by the savings associated with providing AOT only to Commentary those who benefit from it and use it. Six previous studies that investigated long-term ambulatory Mika L Nonoyama oxygen therapy (AOT) for patients with COPD demonstrated Lawrence S Bloomberg Faculty of Nursing and that, on average, AOT did not improve patient outcomes Department of Physical Therapy, University of Toronto, (Liker et al 1975, McDonald et al 1995, Eaton et al 2002, Lacasse et al 2005, Nonoyama et al 2007, Sandland et Canada al 2008). Even after increasing the sample size, Moore et al (2010) showed a similar lack of benefit. Is AOT an References ineffective treatment or have we yet to identify those who benefit? Earnest MA (2002) J Gen Intern Med 17: 749–755. A proportion of patients may ‘respond’ to AOT. However, Eaton T et al (2002) Eur Respir J 20: 306–312. as the consistent definition of a ‘responder’ has not been established, the range of responders within study samples Lacasse Y et al (2005) Eur Respir J 25: 1032–1038. is large: 56% in Eaton et al (2002) and 7% in Nonoyama et al (2007). Predictors of benefit remain unknown; due partly Liker ES, et al (1975) Chest 68: 236–241. to small sample sizes, but also because psychological and behavioural barriers (Earnest 2002) potentially outweigh McDonald CF et al (1995) Am J Respir Crit Care Med 152: any physiologic benefit of AOT. A low average duration of 1616–1619. AOT use (ie, < 2 hours/day) is a common finding. Nonoyama ML et al (2007) Am J Respir Crit Care Med 176: Until the characteristics of ‘responders’ to AOT are 343–349. established, clinicians should be sceptical of its benefit. The best course of action may be to assess on a patient- Sandland CJ et al (2008) Chest 134: 753–760. Journal of Physiotherapy 2011 Vol. 57 – © Australian Physiotherapy Association 2011 127

Appraisal Correspondence Publications are welcome but mechanisms underlying the effects of new technologies have yet to be delineated We note with interest two recent articles in the Journal the craze’ by the Editorial in the same issue of the journal of Physiotherapy regarding the use of new technologies (Russell and Jones 2011), which reviews the medico- in clinical practice. We think this is an exciting field of legal implications of the use of new technologies in both research, illustrated by the growing number of published clinical practice and research. This is particularly timely studies in this area (Piron et al 2009, Yavuzer et al 2008, as preliminary research highlights possible adverse effects Yang et al 2008, Chuang et al 2006). Results from several of long-term use of these types of devices, such as fatigue trials indicate that use of these technologies might improve (Carey et al 2007) and shoulder pain (Hijmans et al in physical outcomes when compared to conventional clinical press). We encourage the international readership of the rehabilitation (Piron et al 2009, Yavuzer et al 2008, Yang et journal to investigate the relevant regulations in their own al 2008, Chuang et al 2006). jurisdiction. We caution that the introduction of these new technologies into clinical practice should be judicious, Many research publications in this area are now freely as the mechanisms underlying their effects have yet to available on websites. For example, www.wiihabilitation. be delineated and possible adverse effects are yet to be co.uk has indexed over 80 articles published since the examined using robust research designs. website was created in 2010. Whilst the amount of research activity in this area is impressive, recommendations about Cath Smith and Leigh Hale the clinical usefulness of these interventions should be University of Otago, New Zealand interpreted with caution. Of all the abstracts of research articles indexed on the Wiihabilitation website, only two /PUF state they have used a randomisation process (Saposnik et al 2010, Wuang et al 2010). It is heartening to see trials, Associate Professor Leigh Hale is Editor of The New such as the one by Kuys and colleagues in the latest issue Zealand Journal of Physiotherapy. of Journal of Physiotherapy, using robust research designs (Kuys et al 2011). In addition, it is reassuring to see that References a small number of randomised trials investigating clinical applications of gaming consoles have been registered on Carey JR et al (2007) Neurorehabil Neural Repair 21: 216–232. sites such as www.clinicaltrials.gov and www.anzctr.org. au. We look forward to publication of these trials. We Chuang TY et al (2006) Phys Ther: 86: 1369–1377. encourage readers who are interested in the clinical effects of technology-related interventions to consider the research Hijmans JM et al (in press) J Rehabil Res. designs used in the studies they read. Furthermore, readers might consider searching for trials on sites such as PubMed Kuys SS et al (2011) J Physiother 57: 35–40. and PEDro, where searches can be restricted to studies of appropriate research design such as randomised controlled Piron L et al (2009) J Rehabil Med 41: 1016–1020. trials. Russell TG, Jones AF (2011) J Physiother 57: 5–7. Kuys and colleagues (2011) acknowledge that their assessment of the clinical effects of exercise with and Saposnik G et al (2010) Stroke 41: 1477–1484. without the use of a gaming console was limited to immediate cardiovascular demand and caution that further Wuang YP et al (2011) Res Dev Disabil 32: 312–321. research into the use of this device for maintenance exercise is appropriate. It is also good to see some ‘tempering of Yang YR et al (2008) Gait Posture 28: 201–206. Yavuzer G et al (2008) Eur J Phys Rehabil Med 44: 237–244. Websites www.wiihabilitation.com www.healthgamers.com www.wiihabtherapy.blogspot.com Journal of Physiotherapy 2011 Vol. 57 – © Australian Physiotherapy Association 2011 131

Weon et al: Visual feedback for scapular winging Real-time visual feedback can be used to activate scapular upward rotators in people with scapular winging: an experimental study Jong-Hyuck Weon1,2, Oh-Yun Kwon2, Heon-Seock Cynn2, Won-Hwee Lee2, Tae-Ho Kim3 and Chung-Hwi Yi2 1Wonju Christian Hospital, 2Yonsei University, 3Daegu Health College Republic of Korea Question: Can real-time visual feedback facilitate the activity of serratus anterior in individuals with scapular winging during shoulder flexion? Design: Comparative, repeated-measures experimental study. Participants: Nineteen subjects with scapular winging. Intervention: Participants performed isometric shoulder flexion at 60° and 90° with and without real-time visual feedback using a video camera to monitor scapular winging. Outcome measures: Activity in the upper trapezius, lower trapezius, and serratus anterior muscles was measured using surface electromyography. A video motion analysis system measured the displacement of a marker attached to the acromion in the frontal and sagittal planes. Results: Visual feedback significantly increased activity in the upper trapezius at 60° of shoulder flexion by 2.3% of maximum voluntary isometric contraction (95% CI 0.7 to 4.0). Visual feedback also significantly increased activity in the serratus anterior at 60° and 90° of shoulder flexion, by 3.0% (95% CI 2.3 to 3.6) and 5.9% (95% CI 3.3 to 8.5) of maximum voluntary isometric contraction respectively. These effects equated to effect sizes from 0.29 to 0.46. Visual feedback also significantly improved movement of the acromion superiorly at 60° of shoulder flexion and anteriorly at 60° and 90° of shoulder flexion. Conclusion: Real-time visual feedback can be used to activate the upper trapezius and serratus anterior muscles and to improve movement of the scapula during shoulder flexion in people with scapular winging. [Weon +) ,XPO O-Y, Cynn )4 -FF8) ,JN T-H, Yi C-H  3FBMUJNFWJTVBMGFFECBDLDBOCFVTFEUPBDUJWBUFTDBQVMBSVQXBSESPUBUPSTJOQFPQMFXJUITDBQVMBSXJOHJOH BOFYQFSJNFOUBMTUVEZJournal of Physiotherapyo> ,FZXPSET: Electromyography, Scapular winging, Serratus anterior, Visual feedback Introduction and winging of the scapula (Ludewig and Cook 2000, Lukasiewicz et al 1999). In particular, over-activation of the Shoulder pain is a common problem. The incidence is upper trapezius and reduced activity in the lower trapezius 11.6 per 1000 person-years in Dutch general practice and serratus anterior muscles during shoulder flexion may (Bot et al 2005), with reports of the prevalence in various contribute to abnormal scapulohumeral rhythm and scapular populations ranging from 7% to 67% (Adebajo and winging (Cools et al 2004, Cools et al 2007, Ludewig and Hazleman 1992, Cunningham and Kelsey 1984, Meyers Cook 2000). Kendall and colleagues (1993) and Sahrmann et al 1982, Reyes Llerena et al 2000). Abnormal scapular (2002) also emphasise weakness of serratus anterior as an position and movement are associated with shoulder pain etiological factor for aberrant scapular mechanics. and glenohumeral joint impingement syndrome (Cools et al 2003, Kibler 1998). Scapular dysfunction may arise from Several pushup and wall sliding exercises have been musculoskeletal factors – including sustained abnormal developed for rehabilitation and in the sports field to posture (Rempel et al 2007), repetitive movements that activate serratus anterior (Hardwick et al 2006, Ludewig deviate from normal movement patterns (Madeleine et et al 2004). However, because the scapula is located behind al 2008), or glenohumeral and scapulothoracic muscle the rib cage, it is not possible for the patient to monitor imbalance (Cools et al 2004, Hallstrom and Karrholm scapular movement visually during these exercises. Thus, 2006) – or from neurological abnormalities. Co-ordinated for effective training of serratus anterior, the exercise must activation of the scapular upward rotators is essential for be supervised to ensure that the load applied to the upper normal scapulohumeral rhythm. limb is appropriate and does not cause scapular winging. Scapular winging is a specific type of scapular dysfunction To our knowledge, none of the studies that have investigated that has two common causes. One is the denervation of the exercises to strengthen serratus anterior in people with long thoracic nerve leading to difficulty flexing the shoulder scapular winging have used real-time visual feedback actively above 120º. The second cause is weakness of the with a video camera to monitor scapular movement during serratus anterior muscle. Serratus anterior is one of the shoulder flexion exercise. We hypothesised that real-time prime muscles used to rotate the scapula upward (Dvir and visual feedback would enable neurologically intact people Berme 1978), and weakness of serratus anterior is a major with scapular winging to activate the scapular upward cause of shoulder pain, glenohumeral joint impingement, rotators, particularly the serratus anterior muscle, during Journal of Physiotherapy 2011 Vol. 57 – © Australian Physiotherapy Association 2011 101

Research shoulder flexion. Therefore the specific research question 'JHVSF Scapulometer used to measure scapular for this study was: winging. A, Stationary board. B, Guide board. C, Sliding board. D, Pad. E, Body. F, Ruler. G, Grip handle. Can real-time visual feedback using a video camera facilitate activation of serratus anterior in people with the scapula, with the sliding board at the level of the inferior scapular winging during shoulder flexion? angle of the scapula. Holding the scapulometer in place with one hand, the examiner moves the sliding board anteriorly Method until it touches the inferior angle of the scapula. A ruler on the fixed board measures the posterior displacement of the Design inferior angle of the scapula from the thoracic wall (Figure 2). A within-participant, repeated measures experimental Several methods could be used to elicit scapular winging study of shoulder muscle activation and scapular alignment for measurement, such as applying a load to the patient’s was carried out in people with scapular winging as they flexed shoulder. Even if the amount of shoulder flexion performed isometric shoulder flexion with and without was fixed, however, the position of the inferior angle of the visual feedback. Electrodes for electromyography were scapula would vary according to the strength of the upward applied over serratus anterior and upper and lower trapezius. rotators of the scapula and the scapulohumeral movement Scapular winging was measured with a scapulometer. pattern. A further problem with this method in the present Initially, scapular winging was measured in a neutral study would be the inability of the participants to maintain shoulder position. Participants then flexed their shoulder a stable position of shoulder flexion, due to weakness of isometrically at 60º and 90º, during which muscle activity serratus anterior. We therefore positioned participants in and scapular winging were measured. standing with the shoulder in the neutral position, the elbow flexed at 90º, and the forearm in neutral rotation. A cuff Participants weighing 5% of the patient’s body weight was placed on the wrist (Figure 3). In this position, a wrist weight provides a Participants were recruited from the Department of Physical load in a direction that tends to induce scapular winging, Therapy, Yonsei University, Korea. A physical examination tilting, and depression. Participants were advised to keep was carried out to determine subject eligibility. Adults were their hand relaxed in a loose fist because hand activity eligible to participate in the study if they had weakness increases shoulder girdle muscle activity (Sporrong et al of serratus anterior and scapular winging. Weakness of 1998). serratus anterior was confirmed by a grade of ‘fair minus’ Before the study commenced, the test-retest reliability of or lower on manual muscle testing (Hislop and Montgomery this method of measuring scapular winging was examined 1995). Scapular winging was confirmed by a distance of at least 2 cm between the thoracic wall and the inferior 'JHVSF. Superior view of the measurement of scapular angle of the scapula, measured using a scapulometer – winging. described in detail below. The exclusion criteria were past or present musculoskeletal conditions affecting the cervical or thoracic spine, glenohumeral joint, rotator cuff muscles, or acromioclavicular joints, as well as neurological and cardiopulmonary disorders that could interfere with shoulder flexion. The scapulometer was modified from the Perry Tool, developed by Plafcan and colleagues (1997). The Perry Tool measures the angle between the transverse plane and a line joining the spinous process and the inferior angle of the scapula. This angle increases as scapular winging increases. However, the angle is also influenced by the amount of scapular abduction, so it does not provide a valid measure of scapular winging. Therefore we developed the scapulometer to measure the posterior displacement of the inferior angle of the scapula from the posterior thoracic wall directly. The body of the scapulometer is a vertical board 20 cm high with an upper width of 14 cm and a lower width of 11 cm, and a thickness of 1.8 cm. Circular pads (2 cm in diameter and 2 cm high) near each corner of the scapulometer allow it to be applied comfortably to the posterior wall of the thorax. A handle on the opposite surface of the scapulometer allows it to be held in place easily. Extending posteriorly from the superior edge of the scapulometer body is a fixed board, mounted with two parallel guides, which allow a horizontal sliding board to move anteroposteriorly between them (Figure 1). To measure scapular winging, the examiner stands behind the patient and places the four pads of the scapulometer on the posterior thoracic wall medial to the vertebral border of 102 Journal of Physiotherapy 2011 Vol. 57 – © Australian Physiotherapy Association 2011

Weon et al: Visual feedback for scapular winging 'JHVSF The marker measuring acromion displacement is illustrated during the 90° shoulder flexion in the visual feedback condition. This view illustrates assessment of movement of the marker superiorly or inferiorly in the frontal plane. A, Starting position. B, Shoulder flexion position. 'JHVSF. Lateral view of the measurement of Measurement scapular winging. The primary measure in the study was muscle activity in in 10 people over five days. The interclass correlation the scapular upward rotators. Surface electromyographic coefficient (ICC 2,1) was 0.97 (95% CI 0.87 to 0.99). The data were collected from the upper and lower trapezius standard error of the measurement was 0.1 cm. and serratus anterior, using a standard data acquisition systema. Preparation of the electrode sites involved shaving Intervention and cleaning the skin with rubbing alcohol (Cram et al 1998). Disposable silver/silver chloride surface electrodesb Each participant was seated on a chair with the cervical were positioned at an inter-electrode distance of 2 cm. The spine in a neutral position. Participants were asked to flex reference electrode was attached to the styloid process of the affected shoulder to two angles (60º and 90º), either with the ulna of the upper limb being tested. Electromyographic or without real-time visual feedback. The order of the two data were collected for the upper trapezius muscle angles and the two feedback conditions were randomised by (using electrodes placed 2 cm lateral to the midpoint drawing a sealed envelope from a box. of a line drawn between the C7 spinous process and the posterolateral acromion), lower trapezius muscle (placed on Participants were instructed to lift the upper limb being an oblique vertical angle with one electrode superior and tested slowly with the elbow extended, the forearm and one inferior to a point 5 cm inferomedial from the root of wrist in a neutral position, and a loose fist, and to hold the spine of the scapula), and the serratus anterior muscle the position for 5 sec at the flexion angle of 60º or 90º. A (placed vertically along the midaxillary line at rib levels universal goniometer was used to determine the flexion 6–8) (Cram et al 1998, Nieminen et al 1993). Each pair of angle, and a horizontal target bar was positioned at each electrodes was aligned parallel to the line of underlying angle in the sagittal plane. The shoulder level and scapular muscle fibres. Electromyographic data were sampled at movement in the lateral and posterior view were recorded 1000 Hz. The signals were amplified and digitisedc. A band- on two video cameras connected to a personal computer. pass filter (20–450 Hz) was used. The root mean square was The computer screen was positioned at the participant’s eye calculated from the raw data using a moving window of 50 level and turned on when real-time visual feedback was msec and was converted to ASCII files for analysis. required. For normalisation, 5 sec of reference contraction data were Before the shoulder flexion, the principal investigator recorded while the participant performed three trials of placed the scapula in the normal position (vertebral border maximal voluntary isometric contraction in the manual parallel with spine spacing at approximately 7 cm, scapula muscle testing position for each muscle (Kendall et al 1993). positioned between T2 and T7 and flat on the posterior rib To ensure maximal effort, verbal encouragement was given. cage). The subject was asked to observe the scapular motion To minimise compensation during data collection, subjects through the computer monitor (Figure 4). If shoulder were encouraged to maintain the testing position (Boettcher depression, tilting, or winging were observed during et al 2008). The middle 3 sec of the 5-sec contraction were shoulder flexion, the investigator encouraged the subject used for data analysis. The initial 1 sec was excluded to to protract and elevate the scapula. Participants practised ensure maximal amplitude had been reached, and the final using the visual feedback to maintain the scapula in a 1 sec was discarded to avoid possible fatigue from sustained normal position for 15 min. The shoulder flexion task was maximal muscle contraction (Soderberg and Knutson 2000, performed three times. A 3-min rest period was allowed Dankaerts et al 2004, Tucker et al 2010). A 3-min rest between trials to minimise fatigue. period was provided between trials. The mean root mean square of the three trials was calculated for each muscle. The electromyographic signals collected during each angle of shoulder flexion were expressed as a percentage of the calculated root mean square of maximal voluntary isometric contraction. Journal of Physiotherapy 2011 Vol. 57 – © Australian Physiotherapy Association 2011 103

Research The secondary measure in the study was displacement of 5BCMF Characteristics of the study participants the acromion in the frontal and sagittal planes. A reflective marker 14 mm in diameter was placed on the skin at the Characteristic n = 19 midpoint of the acromion to measure its displacement in the Age (years), mean (SD) 24 (2) frontal and sagittal planes during shoulder flexion (Figure Height (cm), mean (SD) 169 (6) 4). The reflective markerd was not used for visual feedback, Weight (kg), mean (SD) 61 (7) but was used for measuring the displacement of acromion. Gender, n male (%) 10 (53) Two video cameras were placed 1.5 m from the shoulder Side of winging, n right (%) 14 (74) joint; one was located behind the subject to capture the superior and inferior displacement of the marker in the Results frontal plane, and the other was placed to the side of the subject to capture the anterior and posterior displacement 'MPXPGQBSUJDJQBOUTUISPVHIUIFTUVEZ of the marker in the sagittal plane. Two 30-cm-long wooden rods attached to the side and back of a wooden chair were Nineteen participants were recruited from the Department used as reference points to calibrate the motion analysis of Physical Therapy, Yonsei University, Korea. The systeme in the frontal and sagittal planes (Figure 5). Video characteristics of the participants are presented in Table files captured during the shoulder flexion test were used to 1. All participants completed all aspects of the testing calculate the displacement of the marker. The distance of procedure according to the random allocation of testing the acromion movement was measured from the starting conditions. position to the end of the predetermined shoulder flexion position in cm by the video motion analysis system software Muscle activity (Figure 5). For the upper trapezius muscle, the main effects were Data analysis significant for shoulder flexion angle (p < 0.001) and feedback (p = 0.017), as was the interaction effect (p = For each combination of flexion angle and feedback 0.003). Visual feedback increased activation of the upper condition, the average of the three trials was calculated trapezius at both 60º and 90º of shoulder flexion (Table 2). for the data analysis. A two-factor repeated-measures After Bonferroni correction, however, the effect of visual analysis of variance was used to determine differences in feedback was significant only at the 60º shoulder flexion muscle activity and acromion displacement during shoulder angle (p = 0.008). flexion (60º and 90º) between the real-time visual-feedback and no-visual-feedback conditions. A p value ≤ 0.05 was For the lower trapezius muscle, the main effect for shoulder deemed to be statistically significant. A paired t-test with flexion angle was significant (p = 0.001), but neither the Bonferroni correction was used (with p = 0.05/6 = 0.0083) main effect for the visual-feedback condition (p = 0.152) for the pair-wise comparison in muscle activity and marker nor the interaction effect (p = 0.150) was significant. The displacement in the frontal and sagittal planes for the two data are presented in Table 2. feedback conditions. 'JHVSF. Real-time visual feedback using video camera input to the computer monitor. 104 Journal of Physiotherapy 2011 Vol. 57 – © Australian Physiotherapy Association 2011

Weon et al: Visual feedback for scapular winging 5BCMFMean (SD) muscle activation as a percentage of the maximum voluntary isometric contraction at 60° and 90° shoulder flexion with and without visual feedback, and mean (95% CI) difference between feedback conditions (n = 19). Muscle Shoulder Muscle activation (%MVIC) p value Effect flexion size angle Feedback Difference 0.008 0.049 0.57 Visual None Visual minus None 0.679 0.34 0.104 0.04 Upper trapezius 60° 7.9 (5.4) 5.6 (2.3) 2.3 (0.7 to 4.0) 0.000 0.18 Lower trapezius 0.000 0.29 Serratus anterior 90° 8.8 (4.7) 7.5 (2.4) 1.3 (0.0 to 2.5) 0.46 60° 15.0 (11.4) 14.5 (8.4) 0.4 (–1.7 to 2.6) 90° 21.0 (18.8) 24.0 (14.8) –3.0 (–6.8 to 0.7) 60° 18.0 (10.4) 15.1 (10.1) 3.0 (2.3 to 3.6) 90° 31.6 (12.6) 25.7 (13.6) 5.9 (3.3 to 8.5) %MVIC = percentage of maximal voluntary isometric contraction 5BCMFMean (SD) displacement (cm) of the acromion at 60° and 90° shoulder flexion with and without visual feedback, and mean (95% CI) difference between feedback conditions (n = 19). Direction Shoulder Displacement (cm) p value Effect flexion Feedback size Superior angle Anterior Difference 60° 90° Visual None Visual minus None 0.000 2.45 60° 3.6 (1.0) 1.5 (0.7) 2.1 (1.4 to 2.7) 0.024 0.94 90° 3.6 (0.5) 3.0 (0.9) 0.6 (0.1 to 1.2) 0.000 2.64 2.5 (1.3) –0.2 (0.7) 2.7 (1.9 to 3.5) 0.000 2.47 2.4 (1.6) –0.6 (0.9) 3.1 (2.1 to 4.0) For the serratus anterior muscle, the main effects were flexion remained significant at both 60º and 90º of shoulder significant for shoulder flexion angle (p < 0.001) and flexion (p < 0.001). feedback (p < 0.001), as was the interaction effect (p = 0.045). Visual feedback significantly increased activation Discussion of serratus anterior at both 60º and 90º of shoulder flexion (Table 2). After Bonferroni correction, the effect of visual This study found that muscle activity in all scapular upward feedback remained significant at both 60º and 90º of rotators increased significantly as the shoulder flexion angle shoulder flexion (p < 0.001). increased and that upper trapezius and serratus anterior muscle activity increased significantly with real-time visual Scapular movement feedback. Measurement of displacement of the acromial marker in The increase in the activity of the upward rotators of the the frontal plane showed that the average movement was scapula between 60º and 90º of shoulder flexion is similar superior for all combinations of flexion angle and feedback. to the gradual increase in activity of the upper trapezius and The main effects were significant for shoulder flexion angle serratus anterior muscles during arm abduction (Bagg and (p < 0.001) and feedback (p < 0.001), as was the interaction Forrest 1986). In that study, the lower trapezius remained effect (p = 0.001). Visual feedback significantly increased relatively inactive until the arm was abducted 90º. The the superior displacement of the acromion (Table 3). After lower trapezius increased its activity – and therefore its Bonferroni correction, the effect of feedback remained contribution to the upward rotation force couple – as the significant only at 60º of shoulder flexion (p < 0.001). arm was elevated beyond 90º. With increasing abduction, the instantaneous centre of rotation of the scapula moved Measurement of displacement of the acromial marker in toward the acromioclavicular joint from the root of the the sagittal plane showed that the average movement was spine of the scapula, lengthening the moment arm of anterior with feedback and posterior without feedback. The the lower trapezius muscle (Bagg and Forrest 1988). main effect was significant for the visual feedback (p = Similarly, in the current study of flexion, the moment arm 0.000), but neither the main effect for shoulder flexion angle of the lower trapezius lengthens as the amount of shoulder (p = 0.100) nor the interaction (p = 0.268) was significant. flexion increases. This is likely to be responsible the After Bonferroni correction, the effect of visual feedback significant increase in activity of the lower trapezius at 90º on anterior movement of the acromion during shoulder flexion (especially maintaining the isometric contraction) Journal of Physiotherapy 2011 Vol. 57 – © Australian Physiotherapy Association 2011 105

Research compared to at 60º flexion. This finding is consistent with warranted to confirm our findings on an objective and the results of other studies investigating muscle activity in reliable grading system and to further investigate the the scapular upward rotator muscles during arm elevation correlation between scapular winging and serratus anterior (Antony and Keir 2010, Ebaugh et al 2005, Jarvholm et al muscle strength. 1991, Mathiassen and Winkel 1990). The present study had several limitations. First, this was Muscle activity in the upper trapezius increased significantly a cross-sectional study, so it could only assess immediate when the participants maintained 60º of shoulder flexion effects. A longitudinal study is warranted to determine the while simultaneously reducing scapular winging using long-term effect of training with visual feedback by people real-time visual feedback. Sahrmann (2002) stated that with scapular winging. Also, kinematic data of scapular an increase in upper trapezius activation is needed to upward rotation were not collected in this study. Finally, we compensate for the weakened serratus anterior muscle. measured scapular upward rotator muscle activity during Thus the upper trapezius may be supporting the increased isometric shoulder flexion, so the findings of this study activity in the serratus anterior, which was significantly cannot necessarily be generalised to concentric or eccentric greater at both the 60º and 90º angles when visual feedback control of shoulder flexion. was provided. The marker displacement in the frontal plane indicated that scapular elevation increased significantly at Our findings demonstrate that muscle activity increased in the 60º shoulder flexion angle when visual feedback was the upper trapezius, lower trapezius, and serratus anterior provided. This may also be the result of the activity of the as the shoulder flexion angle increased under the visual- upper trapezius at the 60º angle. Anterior movement of the feedback condition and that the activity in the upper acromion in the sagittal plane was significantly greater at trapezius and serratus anterior muscles was significantly both shoulder flexion angles when visual feedback was greater than that measured during the no-visual-feedback provided, which is consistent with the increased activity condition. Thus, visual feedback during shoulder flexion of serratus anterior. These findings indicate that visual can be recommended to increase activation of the upper feedback helped the participants activate appropriate trapezius and serratus anterior muscles. Q musculature during shoulder flexion to control scapular winging. Footnotes: aBiopac MP100WSW System, Biopac Systems Inc, Goleta, CA, USA. bDelsys Inc, Boston, MA, USA. A number of exercises to strengthen serratus anterior have cAcqKnowledge 3.7.2 software, Biopac Systems Inc, been described in the literature (Decker et al 1999, Ekstrom Goleta, CA, USA. dOxford Metrics, Inc. Oxford, UK. et al 2003, Hardwick et al 2006, Ludewig et al 2004). These eTwinner Pro, SIMI Motion 5.0 Reality Motion Systems, exercises should be performed with scapular protraction Unterschleissheim, Germany. to activate the serratus anterior muscle while stabilising the thoracic wall, and they should be carried out with no Ethics: The Yonsei University institutional review board scapular winging. Sahrmann (2002) suggested that the level approved this study. All participants gave written informed of resistance be modified for weakened scapular upward consent before data collection began. rotators. However, no effective means of self-monitoring and correcting scapular winging during shoulder flexion Competing interests: None declared. exercise has been available. Real-time visual feedback using a video provides an immediate and continuous feedback for Correspondence: Oh-Yun Kwon, Department of correcting scapular movement during independent shoulder Rehabilitation Therapy, Yonsei University, Republic of flexion exercise. Therefore this system of visual feedback Korea. Email: [email protected] is a useful way to facilitate serratus anterior activity during shoulder flexion in people with winging of the scapula. References The activity of the lower trapezius was not significantly Adebajo AO, Hazleman BL (1992) Soft-tissue shoulder lesions increased when visual feedback was provided. This finding in the African. British Journal of Rheumatology 31: 275–276. may be related to the verbal instructions given to the participants. Participants were instructed to protract and Antony NT, Keir PJ (2010) Effect of posture, movement elevate the affected scapula. Thus the verbal instructions and hand load on shoulder muscle activity. Journal of may have reinforced the actions of both the serratus anterior Electromyography and Kinesiology 20: 191–198. and the upper trapezius more than the action of the lower trapezius. Bagg SD, Forrest WJ (1986) Electromyographic study of the scapular rotators during arm abduction in the scapular plane. The scapulometer showed high test-retest reliability for American Journal of Physical Medicine 65: 111–124. the measurement of scapular winging in this study. The scapulometer may be utilised in future research as a Bagg SD, Forrest WJ (1988) A biomechanical analysis of screening tool for scapular winging. The threshold of 2 cm scapular rotation during arm abduction in the scapular plane. was used to define scapular winging in this study because American Journal of Physical Medicine & Rehabilitation 67: this is the minimum amount of winging of the inferior angle 238–245. of the scapula we had observed in people with ‘fair minus’ or lower grade of muscle strength of the serratus anterior Boettcher CE, Ginn KA, Cathers I (2008) Standard maximum on manual muscle testing. However, no previous studies isometric voluntary contraction tests for normalizing shoulder have provided normative data for winging or suggested a muscle EMG. Journal of Orthopaedic Research 26: 1591– relationship between the degree of winging and the strength 1597. of the serratus anterior muscle. Thus, future studies are Bot SD, van der Waal JM, Terwee CB, van der Windt DAWM, Schellevis FG, Bouter LM (2005) Incidence and prevalence of complaints of the neck and upper extremity in general practice. Annals of the Rheumatic Diseases 64: 118–123. Cools AM, Declercq GA, Cambier DC, Mahieu NM, Witvrouw 106 Journal of Physiotherapy 2011 Vol. 57 – © Australian Physiotherapy Association 2011

Weon et al: Visual feedback for scapular winging EE (2007) Trapezius activity and intramuscular balance Ludewig PM, Cook TM (2000) Alterations in shoulder kinematics during isokinetic exercise in overhead athletes with and associated muscle activity in people with symptoms of impingement symptoms. Scandinavian Journal of Medicine shoulder impingement. Physical Therapy 80: 276–291. & Science in Sports 17: 25–33. Ludewig PM, Hoff MS, Osowski EE, Meschke SA, Rundquist Cools AM, Witvrouw EE, Declercq GA, Danneels LA, Cambier PJ (2004) Relative balance of serratus anterior and upper DC (2003) Scapular muscle recruitment patterns: trapezius trapezius muscle activity during push-up exercises. The muscle latency with and without impingement symptoms. American Journal of Sports Medicine 32: 484–493. The American Journal of Sports Medicine 31: 542–549. Lukasiewicz AC, McClure P, Michener L, Pratt N, Sennett B Cools AM, Witvrouw EE, Declercq GA, Vanderstraeten GG, (1999) Comparison of 3-dimensional scapular position and Cambier DC (2004) Evaluation of isokinetic force production orientation between subjects with and without shoulder and associated muscle activity in the scapular rotators during impingement. Journal of Orthopaedic & Sports Physical a protraction–retraction movement in overhead athletes with Therapy 29: 574–583. impingement symptoms. British Journal of Sports Medicine 38: 64–68. Madeleine P, Mathiassen SE, Arendt-Nielsen L (2008) Changes in the degree of motor variability associated with Cram JR, Kasman GS, Holtz J (1998) Introduction to Surface experimental and chronic neck-shoulder pain during a Electromyography. Gaithersburg: Aspen pp 273–283. standardized repetitive arm movement. Experimental Brain Research 186: 689–698. Cunningham LS, Kelsey JL (1984) Epidemiology of musculoskeletal impairments and associated disability. Mathiassen SE, Winkel J (1990) Electromyographic activity American Journal of Public Health 74: 574–579. in the shoulder-neck region according to arm position and glenohumeral torque. European Journal of Applied Dankaerts W, O’Sullivan PB, Burnett AF, Straker LM, Physiology and Occupational Physiology 61: 370–379. Danneels LA (2004) Reliability of EMG measurements for trunk muscles during maximal and sub-maximal voluntary Meyers OL, Jessop S, Klemp P (1982) The epidemiology of isometric contractions in healthy controls and CLBP patients. rheumatic disease in a rural and an urban population over Journal of Electromyography and Kinesiology 14: 333–342. the age of 65 years. South African Medical Journal 62: 403– 405. Decker MJ, Hintermeister RA, Faber KJ (1999) Serratus anterior muscle activity during selected rehabilitation exercises. The Nieminen H, Takala EP, Viikari-Juntura E (1993) Normalization American Journal of Sports Medicine 27: 784–791. of electromyogram in the neck-shoulder region. European Journal of Applied Physiology and Occupational Physiology Dvir Z, Berme N (1978) The shoulder complex in elevation of 67: 199–207. the arm: A mechanism approach. Journal of Biomechanics 11: 219–225. Plafcan DM, Turczany PJ, Guenin BA, Kegerreis S, Worrell TW (1997) An objective measurement technique for posterior Ebaugh DD, McClure PW, Karduna AR (2005) Three- scapular displacement. Journal of Orthopaedic & Sports dimensional scapulothoracic motion during active and Physical Therapy 25: 336–341. passive arm elevation. Clinical Biomechanics 20: 700–709. Rempel DM, Wang PC, Janowitzl I, Harrison RJ, Yu F, Ritz BR Ekstrom RA, Donatelli RA, Soderberg GL (2003) Surface (2007) A randomized controlled trial evaluating the effects of electromyographic analysis of exercises for the trapezius new task chairs on shoulder and neck pain among sewing and serratus anterior muscles. Journal of Orthopaedic & machine operators: the Los Angeles garment study. Spine Sports Physical Therapy 33: 247–258. 32: 931–938. Hallstrom E, Karrholm J (2006) Shoulder kinematics in Reyes Llerena GA, Guibert Toledano M, Hernandez Martinez 25 patients with impingement and 12 controls. Clinical AA, Gonzalez Otero ZA, Alcocer Varela J, Cardiel MH et Orthopaedics and Related Research 448: 22–27. al (2000) Prevalence of musculoskeletal complaints and disability in Cuba. A community-based study using the Hardwick DH, Beebe JA, McDonnell MK, Lang CE (2006) A COPCORD core questionnaire. Clinical and Experimental comparison of serratus anterior muscle activation during a Rheumatology 18: 739–742. wall slide exercise and other traditional exercises. Journal of Orthopaedic & Sports Physical Therapy 36: 903–910. Sahrmann SA (2002) Diagnosis and Treatment of Movement Impairment Syndrome. St Louis: Mosby, pp 193–245. Hislop HJ, Montgomery J (1995) Daniels and Worthingham’s Muscle Testing: Techniques of Manual Examination (6th ed). Soderberg GL, Knutson LM (2000) A guide for use and Philadelphia: Saunders, pp 57–75. interpretation of kinesiologic electromyographic data. Physical Therapy 80: 485–498. Jarvholm U, Palmerud G, Karlsson D, Herberts P, Kadefors R (1991) Intramuscular pressure and electromyography in Sporrong H, Palmerud G, Kadefors R, Herberts P (1998) The four shoulder muscles. Journal of Orthopaedic Research 9: effect of light manual precision work of shoulder muscles an 609–619. EMG analysis. Journal of Electromyography and Kinesiology 8: 177–184. Kendall FP, McCreary EK, Provance PG (1993) Muscles: Testing and Function with Posture and Pain (4th ed). Tucker WS, Armstrong CW, Gribble PA, Timmons MK, Yeasting Baltimore: Williams and Wilkins, pp 286–293. RA (2010) Scapular muscle activity in overhead athletes with symptoms of secondary shoulder impingement during Kibler WB (1998) The role of the scapula in athletic shoulder closed chain exercises. Archives of Physical Medicine and function. The American Journal of Sports Medicine 26: 325– Rehabilitation 91: 550–556. 337. Journal of Physiotherapy 2011 Vol. 57 – © Australian Physiotherapy Association 2011 107

Lewis et al: Strain-Counterstrain therapy for low back pain Strain-Counterstrain therapy combined with exercise is not more effective than exercise alone on pain and disability in people with acute low back pain: a randomised trial Cynan Lewis1,2, Tina Souvlis3 and Michele Sterling2,4 1Stanthorpe Health Services, Queensland Health, 2Division of Physiotherapy and NHMRC Centre for Clinical Research Excellence in Spinal Pain, Injury and Health, The University of Queensland, 3Research Co-ordination Clinical Education and Training Unit, Queensland Health, 4Centre of National Research on Disability and Rehabilitation Medicine (CONROD), The University of Queensland Australia Question: Is Strain-Counterstrain treatment combined with exercise therapy more effective than exercise alone in reducing levels of pain and disability in people with acute low back pain? Design: Randomised trial with concealed allocation, assessor blinding, and intention-to-treat analysis. Participants: 89 (55 female) participants between 18 and 55 years experiencing acute low back pain were randomised to experimental (n = 44) and control (n = 45) groups. Intervention: Participants attended four treatments in two weeks. The experimental group received Strain-Counterstrain treatment and review of standardised exercises (abdominal bracing, knee to chest, and lumbar rotation). The control group performed the standardised exercises under supervision. Following the intervention period, all participants received exercise progression, manual therapy, and advice. Outcome measures: The primary outcome was the modified Oswestry low back pain disability questionnaire, measured at 2 weeks (ie, end of treatment), 6 weeks, and 28 weeks. Secondary outcome measures included the SF-36, visual analogue scale pain ratings, and a 7-point global rating of change. Results: The experimental intervention was not more effective than exercise alone in reducing levels of pain and disability. Mean between-group differences in change from baseline for the Oswestry Disability Index were 0 (95% CI –6 to 7) after treatment, –1 (95% CI –7 to 6) at 6 weeks, and 2 (95% CI –4 to 8) at 28 weeks. Other outcomes did not differ significantly between groups. Conclusion: There is no advantage in providing Strain- Counterstrain treatment to patients with acute low back pain, although further studies could examine whether a subset of these patients can benefit from the treatment. Trial registration: ACTRN 12609000084280. <-FXJT$ 4PVWMJT5 4UFSMJOH.   Strain-Counterstrain therapy combined with exercise is not more effective than exercise alone on pain and disability JOQFPQMFXJUIBDVUFMPXCBDLQBJOBSBOEPNJTFEUSJBMJournal of Physiotherapyo> ,FZXPSETStrain-Counterstrain, Manual therapy, Spinal manipulative therapy, Exercise Introduction effective than recommended first-line intervention for treatment of non-specific acute low back pain (Assendelft Low back pain remains a common disabling condition et al 2003, Ferreira et al 2003) and chronic low back pain (Bogduk and McGuirk 2002, Walker et al 2004) that is (Assendelft et al 2003). However, many physiotherapists immensely costly in Australia (Rahman et al 2005) and the use spinal manipulative therapy simultaneously with United States of America (Luo et al 2004). There is evidence recommended first-line intervention for treatment of non- that many individuals with acute low back pain develop specific acute low back pain (Reid et al 2002). persistent or recurrent low back pain (Henschke et al 2008, Pengel et al 2003, Abbott and Mercer 2002). The cause Strain-Counterstrain is a manual therapy intervention of acute low back pain is ‘non-specific’ in approximately involving passive positioning of the body or limbs. It has 95% of cases (Hollingworth et al 2002). Nevertheless, been proposed as a treatment for musculoskeletal pain and physiotherapists have developed various algorithms for dysfunction (Jones et al 1995). When used to treat acute diagnosis of the condition (Deyo 1993, Winkel et al 1996) low back pain, this intervention can be considered as a form and many clinical interventions have been proposed and are of spinal manipulative therapy because the pelvis, sacrum, used for the treatment of acute low back pain (Deyo 1993, and lower limbs are used to position the lumbar and sacral March et al 2004, Reid et al 2002). regions passively in degrees of flexion, extension, lateral flexion, and rotation. The rationale for Strain-Counterstrain Recent guidelines assert that there is ‘fair’ evidence that treatment is unclear. A proprioceptive model (Korr 1975), spinal manipulative therapy provides a small to moderate which has not been experimentally tested, provides the benefit (a 5 to 20 point reduction in Oswestry Disability hypothetical basis for the Strain-Counterstrain assessment Index score) in the treatment of acute low back pain (Chou and treatment using digitally tender points (Jones et al 1995, et al 2007). However, most international guidelines for Kusunose 1993). To our knowledge, there is no experimental treatment of non-specific acute low back pain recommend evidence to support the use of Strain-Counterstrain for the spinal manipulative therapy as a second-line intervention treatment of acute low back pain, although reductions in after first-line treatment of simple analgesics and advice pain and disability following Strain-Counterstrain treatment (van Tulder et al 2006, Koes et al 2001) and this position for low back pain have been reported in case studies (Lewis is supported by contemporaneous meta-analyses, which and Flynn 2001). This randomised trial was intended to concluded that spinal manipulative therapy was not more Journal of Physiotherapy 2011 Vol. 57 – © Australian Physiotherapy Association 2011 91

Research investigate the effect of Strain-Counterstrain treatment for of side-lying abdominal bracing was assessed clinically by acute low back pain in a clinical setting. observing for a slight drawing-in of the lower abdominal wall below the umbilicus which is consistent with activation The research questions for this study were: of the transversus abdominis muscle (Richardson et al 1. Is a combination of Strain-Counterstrain and exercise 1999). Participants were asked to perform the exercises in more effective than exercise alone in reducing levels a range that did not increase their pain, twice a day during of pain, disability, and dysfunction in participants the intervention period. The exercises were not progressed with acute low back pain after 2 weeks? during the intervention period. 2. Is there any residual effect of Strain-Counterstrain on levels of pain, disability, and dysfunction at 6 weeks Participants in the experimental group attended twice (short term) and 28 weeks (medium term)? a week for two consecutive weeks and received Strain- Counterstrain treatment and review of the standardised Method exercises. Strain-Counterstrain treatment involved passive positioning of a participant, with varying degrees of Design spinal flexion/extension, lateral flexion and rotation, such that there was a two-thirds reduction in tenderness at a A single-centre, randomised controlled trial was conducted monitored digitally tender point (Jones et al 1995). This was at the physiotherapy outpatient department of a rural public determined by having participants rate their tenderness to hospital in Australia. Participants were referred by public palpation at digitally tender points on a numerical pain scale and private medical practitioners for treatment of acute low where 10 represented initial tenderness and 0 no tenderness. back pain or were recruited through posted notices and In addition to reported tenderness with intermittent probing, advertisement in local papers. Randomisation was achieved perceived tissue tension was used to guide the experimenter by having the participant select one of 100 sealed opaque to the appropriate passive position (Jones et al 1995). envelopes, each containing a group allocation, which had The participant was passively maintained at this point been prepared and shuffled by an independent investigator. by the experimenter for approximately 90 seconds, with The experimental group received a combination of Strain- intermittent probing at 30-sec intervals to ensure correct Counterstrain and exercise, while the control group received positioning, before being slowly and passively returned to a only the exercises. The interventions were provided at neutral position (Jones et al 1995, Kusunose and Wendorff four visits occurring over two weeks. Measurements 1990, Kusunose 1993). Treatment of a digitally tender point were recorded at baseline, at 2 weeks (immediately after was considered successful if tenderness reduced by 70% or the intervention), at 6 weeks, and at 28 weeks. The 28- more (Kusunose 1993, Kusunose and Wendorff 1990). week follow-up was expected to capture the majority of participants who would develop persistent low back pain or Participants in the control group also attended twice a week recurrence of low back pain within 12 months (Philips and for two consecutive weeks for revision and supervised Grant 1991, Von Korff and Saunders 1996). performance of the standardised exercises. After the intervention period, both experimental and control group Participants participants received similar additional interventions deemed appropriate by the treating physiotherapist with People entering the trial had to meet the following inclusion neither group receiving Strain-Counterstrain treatment. criteria: aged between 18 and 55 years; experiencing pain These included progression of home exercise program, in the lumbar and/or sacral regions (Merskey 1994) that ergonomic instruction, soft-tissue mobilisation, and joint had been present for less than three months following a mobilisation. period of a month without pain in that region; able to lie either prone or supine for up to 10 minutes; and displaying Outcome measures a minimum of 4 digitally tender points at examined sites (Kusunose and Wendorff 1990) at their initial assessment. The primary outcome was disability measured by the The exclusion criteria were: Oswestry Disability Index modified Oswestry low back pain disability questionnaire score less than 10, history of spinal surgery or fracture or (Fritz and Irrgang 2001). This measure has been shown to diagnosis with an inflammatory disorder or fibromyalgia. be valid and reliable (Fairbank et al 1980) and its properties Patients were also excluded if assessment suggested that have been studied rigorously (Beurskens et al 1996, Fritz and they were experiencing lumbar radiculopathy (Wilk 2004). Irrgang 2001, Davidson and Keating 2002). The secondary outcomes included quality of life, pain, interference with Intervention work, satisfaction with symptoms, satisfaction with the intervention, a global rating of change, and the number of All participants were given the same general advice, which treatments post-intervention and adverse events. was to continue using medication as prescribed by their medical practitioner and to remain active (March et al Quality of life was measured with the SF-36 questionnaire 2004), but to avoid activities that aggravated their low back and calculated using all subscales (Ware and Sherbourne pain. All participants were instructed in a standardised 1992). This health-related quality of life questionnaire has exercise program and issued with a printed handout to been studied with low back pain populations and shown reinforce the verbal instructions. The handout is available to have good validity, reliability, and responsiveness for as an e-addendum (see Appendix 1). The exercise program most subscales (Taylor et al 2001) and has sufficient scale consisted of three exercises that are commonly prescribed width to detect change in most people with low back pain by physiotherapists for clients with low back pain: side- (Davidson and Keating 2002). lying abdominal bracing (intended to activate deep abdominal stabilisers) (Richardson et al 1999), alternate Pain was rated by participants on a 10-cm visual analogue knee-to-chest holds (Nicholas et al 2007), and side-to- scale, which has been shown to be valid and reliable (Price side lumbar rotation (Olson 2007). Correct performance 92 Journal of Physiotherapy 2011 Vol. 57 – © Australian Physiotherapy Association 2011

Lewis et al: Strain-Counterstrain therapy for low back pain 5BCMF Baseline characteristics of participants. Randomised Lost to follow-up Characteristic (n = 89) (n = 5) Gender, n female (%) Exp Con Exp Con Age (yr), mean (SD) (n = 44) (n = 45) (n = 3) (n = 2) Height (m), mean (SD) Weight (kg), mean (SD) 25 (57) 30 (67) 1 (33) 1 (50) Duration of low back pain (wk), mean (SD) 40 (10) 40 (11) 41 (7) 41 (4) Using medication for low back pain, n (%) 1.70 (0.09) 1.69 (0.11) 1.79 (0.05) 1.75 (0.14) Exp = experimental group, Con = control group 78 (18) 80 (18) 74 (9) 79 (4) 4.2 (3.5) 4.3 (3.8) 5.7 (5.7) 1.5 (0.7) 22 (50) 16 (36) 2 (66) 2 (100) et al 1983, Duncan et al 1989, Price et al 1994). Each initial loss to follow-up was very low, study recruitment was participant’s pain was summarised as the mean of three closed at 89 participants. ratings on the visual analogue scale: minimum pain in the last 24 hours, current pain, and maximum in the last 24 Analyses were conducted using the intention-to-treat hours. principle including data from all randomised participants wherever it could be obtained. Significance for analyses The degree to which pain interfered with normal work, was set at p < 0.05. Data samples were examined for including both work outside the home and housework, was normality using the Kolmogorov-Smirnov test and Q-Q rated from 1 (not at all) to 5 (extremely). The degree to plots. Repeated measures ANOVA was used to examine which the participant would be satisfied to spend the rest for differences between groups for Oswestry Disability of their lives with their current symptoms was rated from Index score, VAS, SF-36, and ratings of interference with 1 (very dissatisfied) to 5 (very satisfied). The participants’ work and satisfaction with life, with Bonferroni adjustment satisfaction with their overall physiotherapy care during used for multiple comparisons. Student t-tests were used to the period of intervention was also rated from 1 (very compare global rating of change and satisfaction with the dissatisfied) to 5 (very satisfied). These outcomes have been intervention between treatment and control groups. The recommended for low back pain research by an international Wilcoxon signed ranks test was used to compare the number group of researchers (Deyo et al 1998). of physiotherapy treatments following the intervention period between groups. Pearson’s chi-square test was used Participants provided a ‘global-rating-of-change’ following to compare groups for the number of participants who were the initial two-week intervention period, on a 7-point scale able to manage their acute low back pain without the need where response 1 = ‘completely gone’, 2 = ‘much better’, 3 = to take medication. ‘better’, 4 = ‘a little better’, 5 = ‘about the same’, 6 = ‘a little worse’ and 7 = ‘much worse’ (Patrick et al 1995). A global- Results rating-of-change response of 3 or less was considered to represent improvement (Patrick et al 1995). 'MPXPGQBSUJDJQBOUTBOEUIFSBQJTUTUISPVHIUIF trial The number of treatments received after the 2-week allocated intervention period, the number of adverse events, Between January 2009 and April 2010, 101 volunteers and the number of participants using medication for low were screened for eligibility. Of these 89 were deemed back pain at Week 2 and Week 6 were recorded from patient eligible, gave informed consent, and were randomised: 44 records. to the experimental group and 45 to the control group. The flow of participants through the trial, including reasons for Data analysis exclusion and loss to follow-up, is presented in Figure 1. We sought to power the study to identify a between-group The baseline characteristics of participants are shown in difference of 6 points on the modified Oswestry scale, Table 1 and the first two columns of Table 2. No important because this corresponds to the threshold at which people differences in these characteristics were noted between the with low back pain typically rate the change in their condition experimental and control groups. A single physiotherapist as ‘moderately’ different, as opposed to ‘hardly’, ‘a little’, with a postgraduate degree in manual therapy and 15 or ‘somewhat’ different (Fritz and Irrgang 2001, Jaeschke years of experience using Strain-Counterstrain treatment et al 1989). We considered that a ‘moderate’ improvement provided all interventions to both experimental and control would be enough for typical patients to consider that the groups and remained blind to primary and secondary intervention in this study is worthwhile. outcome measures throughout the trial. A total of 90 participants would provide 80% power to detect Compliance with the trial method a difference between groups of 6 points on the modified Oswestry scale as significant at a two-sided significance In each group, all participants attended two 30-min level, assuming a standard deviation of 10 points (Fritz et al intervention sessions per week for two consecutive 2005, Childs et al 2004). To allow for some loss to follow- weeks. All participants received the study intervention up, we increased the original sample to 100. However, since as originally allocated. By 28 weeks, 3 (7%) participants Journal of Physiotherapy 2011 Vol. 57 – © Australian Physiotherapy Association 2011 93

Research Volunteers with acute low back pain completed screening questionnaires (n = 101) Excluded (n = 4) š Oswestry score less than 10 (n = 3) š unable to attend scheduled visits (n = 1) Volunteers with acute low back pain screened physically (n = 97) Excluded (n = 8) š less than 4 digitally tender points (n = 6) š radicular signs and symptoms (n = 2) Measured disability, quality of life, pain, interference with work, satisfaction with life Week 0 and use of medication for low back pain Randomised (n = 89) (n = 44) (n = 45) Lost to follow up Experimental group Control group Lost to follow up š family reasons š 4 x 30-min sessions š 4 x 30-min sessions š did not (n = 1) in 2 weeks in 2 weeks attend (n = 1) š 3 exercises with š 3 exercises with handout handout š supervision of exercises š Strain-Counterstrain š advice to use analgesia š advice to use analgesia as prescribed, remain as prescribed, remain active and avoid active and avoid aggravating activities aggravating activities Week 2 Measured disability, quality of life, pain, interference with work, satisfaction with life, satisfaction with treatment, global rating of change and medication use (n = 43) (n = 44) Lost to follow up Lost to follow up š unreturned š unreturned forms (n = 1) forms (n = 1) Measured disability, quality of life, pain, interference with work, satisfaction with life, Week 6 and medication use (n = 42) (n = 43) Lost to follow up š unreturned forms (n = 1) Measured disability, quality of life, pain, interference with work, satisfaction with life, Week 28 and adverse events (n = 41) (n = 43) 'JHVSF Design and flow of participants through the trial. 94 Journal of Physiotherapy 2011 Vol. 57 – © Australian Physiotherapy Association 2011

Lewis et al: Strain-Counterstrain therapy for low back pain Difference between groups Week 28 2 0.0 –0.5 (–0.6 to 0.5) (–1.1 to 0.2) (–1.2 to 0.2) in the experimental group and 2 (4%) participants in the minus (–4 to 8) (–0.9 to 1.1) control group had been lost to follow-up (see Figure 1). A Week 0 small number of participants failed to complete the study 3 0.1 questionnaires at isolated measurement points, as presented Exp minus (–4 to 11) (–0.5 to 0.7) in Tables 2 and 3. Con Effect of intervention 5BCMF Mean (SD) for outcomes reported at all study visits for each group, mean (SD) difference within groups, and mean (95% CI) difference between groups. Week 6 –1 –0.4 –0.5 minus (–7 to 6) (–1.3 to 0.6) At the end of the 2-week intervention period, the Week 0 experimental and control groups did not have significantly Exp minus 2 0.1 different scores on the modified Oswestry Disability Con (–6 to 11) (–0.5 to 0.7) Index, with a mean between-group difference in change from baseline of 0 points (95% CI –6 to 7). Also at this Week 2 0 0.3 0.0 time, the groups did not differ significantly on the any of minus (–6 to 7) (–0.6 to 1.2) the secondary outcomes, as presented in Tables 2 and 3 Week 0 (individual data are presented in Table 4 on the eAddenda). Exp minus 3 0.0 The percentage of the experimental group using medication Con (–3 to 9) (–0.5 to 0.5) for their low back pain at the end of the 2-week intervention (88%, 38/43) was not significantly different from the control Week 28 Exp Con Exp Con Exp Con Exp Con –12 –14 –1.7 –1.6 –1.1 group (73%, 32/44), relative risk 1.22 (95% CI 0.98 to 1.50). minus (12) (15) (2.2) (1.6) (1.5) Week 0 16 13 –1.0 A significant difference was found in global rating of change (18) (18) (1.5) between groups immediately following the intervention. The experimental group had a mean rating of 2.9 points (SD Difference within groups –1.7 1.1) while the control group had a mean of 3.5 points (SD (2.0) 1.4). The mean between-group difference was 0.6 points in –1.0 favour of the experimental group (95% CI 0.1 to 1.1). (1.1) At the 6-week and 28-week follow-up points, no statistically Week 6 –13 –12 significant differences were identified for any outcomes, minus (12) (17) even before Bonferroni correction, as presented in Tables 2 Week 0 13 11 and 3. There was no significant difference in the number of (19) (19) treatments received after the 2-week allocated intervention –1.9 –1.5 period. The percentage of the experimental group using (2.0) (2.3) medication for their low back pain at 6 weeks (83%, 34/41) –1.0 –1.1 was not significantly different from the control group (1.3) (1.4) (73%3, 0/41), relative risk 1.13 (95% CI 0.90 to 1.43). There –1.3 –0.9 were no adverse effects reported during the trial in either (1.5) (1.6) group. Week 2 16 16 –10 –11 97 (22) (22) (13) (15) –1.3 –1.6 2.0 2.1 –1.0 –1.0 2.8 3.2 –0.7 –0.7 Exp = experimental group, Con = control group, SF-36 = short form 36, VAS = visual analogue scale for pain Discussion minus (15) (16) (12) (17) (2.1) (2.0) (1.1) (1.3) (1.1) (1.2) (1.4) (1.4) (1.3) (1.2) Week 0 This study was the first to examine the treatment of acute low back pain using Strain-Counterstrain techniques. Week 28 (n = 44) (n = 45) (n = 43) (n = 44) (n = 41) (n = 41) (n = 41) (n = 43) 70 69 2.0 2.1 Adding the Strain-Counterstrain intervention did not (2.0) (2.0) substantially improve outcomes over exercise therapy alone. The best estimates of the effect of the intervention Week 6 Exp Con 15 18 66 67 1.8 2.3 1.9 2.0 3.0 3.4 at the three outcome assessment points were only 2 points (12) (16) (22) (21) (1.8) (2.1) (1.1) (1.1) (1.5) (1.5) or less on a 100-point scale. However, the upper limits of the 95% CIs around these estimates all still included the Groups Exp Con 17 20 63 62 2.4 2.2 2.0 2.2 (1.0) (1.2) 3.6 3.6 pre-specified minimum clinically important difference of (12) (15) (18) (19) (1.8) (1.7) (1.2) (1.3) 6 points. Therefore it is possible, although unlikely, that Week 2 (n = 42) further research could identify a clinically worthwhile difference by further refining these estimates. Week 0 Exp Con 27 31 54 55 3.6 3.8 2.9 3.2 interference (1.1) (1.2) (n = 44) 4.3 (12) (17) (19) (17) (1.6) (1.6) (0.8) We consider Strain-Counterstrain to be a form of spinal (n = 43) manipulative therapy, because the pelvis, sacrum, and lower 4.4 limbs are used to position the lumbar and sacral regions (0.6) passively in degrees of flexion, extension, lateral flexion, and rotation. A systematic review of spinal manipulative Outcome (range) Oswestry SF-36 VAS cm Work (1–5) Life therapy, which the reviewers defined as both joint (0–100) (0–100) (0–10) satisfaction mobilisation and high velocity thrust techniques, found (1–5) that it was not more effective for the treatment of acute low back pain than ‘physical therapy exercise’ (Assendelft et al 2003). Therefore, acknowledging the differences in the definition of spinal manipulative therapy, our findings are consistent with the results of this review. Journal of Physiotherapy 2011 Vol. 57 – © Australian Physiotherapy Association 2011 95

Research 5BCMF Mean (SD) for outcomes reported at one study visit for each group, mean (SD) difference within groups, and mean (95% CI) difference between groups. Outcome Groups Difference between groups Exp Con Exp minus Con (n = 43) (n = 45) Global rating of change at Week 2 2.9 (1.1) 3.5 (1.4) –0.6 (–1.1 to –0.1) (0–7 scale) (n = 42) –0.2 (–0.7 to 0.3) 1.7 (1.2) 1.9 (1.2) 0.2 (–1.0 to 1.4) Patient satisfaction with the intervention at (n = 44) Week 2 2.5 (3.1) — (1–5 scale) 2.3 (2.5) Treatments received after the intervention 0 (0) 0 (0) and before Week 28 (number/participant) Adverse events by Week 28 (number/participant) Exp = experimental group, Con = control group The finding that those provided with Strain-Counterstrain at anterior and posterior sites claimed to be associated with treatment registered a significantly greater improvement in low back pain; pain localised to the lumbosacral region; and global rating of change at the end of the intervention period less than 45 years of age. is unlikely to be clinically relevant because the difference between groups was only 0.5. Our findings should be considered within the context of the limitations of the study design. Since this trial was conducted Approximately 40% of individuals with acute low back pain in a clinical setting with the majority of participants referred are likely to recover rapidly without intervention or with by medical practitioners for physiotherapy treatment, it was first-line intervention of simple analgesia and advice (Pengel not possible to incorporate a control group that did not receive et al 2003). This may be one reason for the small effects of physiotherapy intervention. Consequently, we were unable additional treatments such as Strain-Counterstrain and other to determine the degree to which significant improvements spinal manipulative therapies (Hancock et al 2008). This in outcome measures for both experimental and control may also have clinical implications for provision of spinal groups were due to the natural history of acute low back manipulative therapy to patients with acute low back pain. pain. Due to the type of intervention, it was not possible For trials to demonstrate substantial effect sizes for acute to blind the physiotherapist who provided interventions. low back pain treatments, it may be necessary to exclude Because no sham-experimental intervention was included individuals with a highly favourable prognosis regardless in the study design, it was not possible to determine the of treatment (Hancock et al 2008). Clinically, it would be degree to which the manual contact in the experimental reasonable to withhold relatively expensive treatments group influenced outcome measures. No attempt was made such as Strain-Counterstrain from these individuals while to control for medications taken by participants, which providing adequate analgesia and advice knowing that they included opioid and non-opioid analgesics and non-steroidal are likely to recover quickly (Hancock et al 2008). anti-inflammatory drugs. However, medication use was similar at baseline and no significant difference was found Another consideration for sampling in studies of treatments between the groups for number of participants who were for non-specific acute low back pain is that the condition managing their pain with medication immediately after is unlikely to be homogenous within a sample (Brennan et the 2-week intervention or at 6 weeks. This suggests that al 2006, Kent and Keating 2004). While all participants medication use was unlikely to be a confounding factor for in this study had a minimum of 4 digitally tender points our comparisons between intervention groups. identified using Strain-Counterstrain procedures, this does not confirm that they were a homogenous sample and This study had several strengths, including that it was it is likely that the source of acute low back pain varied analysed using the intention-to-treat principle and that among the participants. A possible strategy to manage participants were assigned randomly to experimental and sample heterogeneity in future studies assessing Strain- control groups. Also, interventions were provided by the Counterstrain treatment for acute low back pain would be to same experienced physiotherapist who remained blind to develop an algorithm, specifically for Strain-Counterstrain outcome measures, which were administered by the same treatment, to identify individuals more likely to respond assistant who was blind to group allocation. Additionally, to this form of treatment. Such algorithms have previously participants in both intervention groups received the same been shown to improve outcomes for non-specific acute/ number of interventions and had comparable contact subacute low back pain (Brennan et al 2006, Childs et al time with the physiotherapist who provided interventions. 2004). Personal clinical experience suggests that for such an A further merit of the study was the high follow-up rate algorithm, factors favouring Strain-Counterstrain treatment (greater than 90%). might include: recent and sudden onset of symptoms; no more than one previous episode of acute low back pain; Several features of the study design mean that the findings more than 4 but less than 10 digitally tender points identified of this study are immediately relevant to the clinical use 96 Journal of Physiotherapy 2011 Vol. 57 – © Australian Physiotherapy Association 2011

Lewis et al: Strain-Counterstrain therapy for low back pain of Strain-Counterstrain treatment for acute low back pain. References Approximately 60% of the participants were referred by medical practitioners to the physiotherapy department Abbott JH, Mercer SR. (2002) The natural history of acute low for treatment of acute low back pain. The single treating back pain. NZ Journal of Physiotherapy 30: 8–16. physiotherapist had 15 years of experience providing Strain-Counterstrain treatment and was able to treat Assendelft WJJ, Morton SC, Yu EI, Suttorp MJ, Shekelle PG freely monitoring anterior and posterior digitally tender (2003) Spinal manipulative therapy for low back pain–a points according to clinical protocols (Jones et al 1995, meta-analysis of effectiveness relative to other therapies. Kusunose 1993). The exercises chosen for the study are Annals of Internal Medicine 138: 871–881. commonly used by physiotherapists for treatment of low back pain (Nicholas et al 2007, Olson, 2007, Richardson Beurskens AJHM, de Vet HCW, Koke AJA (1996) et al 1999) and were reinforced with a detailed written Responsiveness of functional status in low back pain: a hand-out. Although it could be argued that exercise therapy comparison of different instruments. Pain 65: 71–76. is not supported for treatment of acute low back pain (Hayden et al 2005, van Tulder et al 2000) and therefore Bogduk N, McGuirk B (2002) Medical management of acute represents minimal or no treatment, it can be justified as a and chronic low back pain. An evidence based approach. valid clinical intervention in our study since many of our Amsterdam: Elsevier. participants had experienced their low back pain for more than 6 weeks (for both groups the average was over 4 weeks) Brennan GP, Fritz JM, Hunter SJ, Thackeray A, Delitto A, and the optimal time to start exercise therapy after the onset Erhard RE (2006) Identifying subgroups of patients with of symptoms of low back pain is unclear (Chou et al 2007). WYkj[%ikXWYkj[¼dedif[Y_ÅY½bemXWYafW_d$Spine 31: 623– Interventions were provided over 30 minutes twice a week 631. for two consecutive weeks, which is likely to correspond to typical physiotherapy intervention for acute low back pain. Childs JD, Fritz JM, Flynn TW, Irrgang JJ, Johnson KK, Majkowski GR, et al (2004) A clinical prediction rule to In summary, for non-specific acute low back pain there does identify patients with low back pain most likely to benefit from not appear to be any short-term or medium-term advantage spinal manipulation: a validation study. Annals of Internal from the addition of Strain-Counterstrain treatment to Medicine 141: 920–928. appropriate analgesic medication, advice, range of motion exercises, and transversus abdominis exercises. Further 9^ek H\" GWi[[c 7\" Idem L\" 9Wi[o :\" 9heii @J\" I^[a[bb[ studies could examine whether a subgroup of individuals P, et al (2007) Diagnosis and treatment of low back pain: a with non-specific acute low back pain are more likely to joint clinical practice guideline from the American college of benefit from Strain-Counterstrain treatment. Q physicians and the American pain society. Annals of Internal Medicine 147: 478–491. eAddenda: Table 4 and Appendix 1 available at: Davidson M, Keating JI (2002) A comparison of five low back jop.physiotherapy.asn.au disability questionnaires: reliability and responsiveness. Physical Therapy 82: 8–23. Ethics: Ethical approval for the study was given by the Toowoomba and Darling Downs Health Service District Deyo RA, Battie M, Beurskens AJHM, Bombardier C, Croft P, Human Research Ethics Committee and The University Koes B, et al (1998) Outcome measures for low back pain of Queensland Medical Research Ethics Committee. All research: a proposal for standardized use. Spine 23: 2003– participants gave written informed consent before data 2013. collection began. Deyo RA (1993) Practice variations, treatment fads, rising Competing interests: None declared. disability. Do we need a new research paradigm? Spine 18: 2153–2162. Acknowledgements: Thanks to Deborah Davis, Administrative Officer, Stanthorpe Health Services, Duncan GH, Bushnell C, Lavigne GJ (1989) Comparison of for assistance in administering self-report outcome verbal and visual analogue scales for measuring the intensity questionnaires and randomisation of participants. Thanks to of experimental pain. Pain 37: 295–303. Stephanie Valentin, Physiotherapist, for research assistance at The University of Queensland. Thanks to Alexandra Fairbank JCT, Couper J, Davies JB, O’Brien JP (1980) Newcombe, Senior Physiotherapist Warwick Health The Oswestry low back pain disability questionnaire. Services, for pre-study discussion and input. Thanks to Dr Physiotherapy 66: 271–273. Asad Khan, Senior Lecturer in Statistics, The University of Queensland, for statistical analysis guidance. Ferreira ML, Ferreira PH, Latimer J, Herbert RD, Maher CG (2003) Efficacy of spinal manipulative therapy for low Correspondence: Cynan Lewis, Department of back pain of less than three months’ duration. Journal of Physiotherapy, Stanthorpe Health Services, PO Box 273 Manipulative Physiology and Therapeutics 26: 593–601. Stanthorpe QLD 4380, Australia. Email: Cynan_Lewis@ health.qld.gov.au Fritz JM, Irrgang JJ (2001) A comparison of a modified Oswestry bemXWYafW_dZ_iWX_b_jogk[ij_eddW_h[WdZj^[Gk[X[YXWYa pain disability scale. Physical Therapy 81: 776–788. Fritz JM, Whitman JM, Childs JD (2005) Lumbar spine segmental mobility assessment: an examination of validity for determining intervention strategies in patients with low hack pain. Archives of Physical Medicine and Rehabilitation 86: 1745–1752. Hancock MJ, Maher CG, Latimer J (2008) Spinal manipulative therapy for acute low back pain: a clinical perspective. Journal of Manual and Manipulative Therapy 16: 198–203. Hayden JA, Van Tulder MW, Malmivaara AV, Koes BW (2005) Meta-analysis: exercise therapy for nonspecific low back pain. Annals of Internal Medicine 142: 765–775. Henschke N, Maher CG, Refshauge KM (2008) Prognosis in patients with recent onset low back pain in Australian primary care: inception cohort study. BMJ 337: a171. Hollingworth W, Todd CJ, King H, Males T, Dixon AK, Karia KR, et al (2002) Primary care referrals for lumbar spine Journal of Physiotherapy 2011 Vol. 57 – © Australian Physiotherapy Association 2011 97

Research radiography: diagnostic yield and clinical guidelines. British Price DD, Bush FM, Long S, Harkins SW (1994) A comparison Journal of General Practice 52: 475–480. of pain measurement characteristics of mechanical visual analogue and simple numerical rating scales. Pain 56: 217– Jaeschke R, Singer J, Guyatt GH (1989) Measurement of 226. health status: ascertaining the minimal clinically important difference. Controlled Clinical Trials 10: 407–415. Price DD, McGrath PA, Rafii A, Buckingham B (1983) The validation of visual analogue scales as ratio scale measures Jones LH, Kusunose RS, Goering EK (1995) Jones Strain- for chronic and experimental pain. Pain 17: 45–56. Counterstrain. Boise: Jones Strain-Counterstrain Inc. Rahman N, Penm E, Bhatai K (2005) Arthritis and Koes BW, van Tulder MW, Ostelo R, Kim Burton A, Wadell G musculoskeletal conditions in Australia. Canberra: Australian (2001) Clinical guidelines for the management of low back Institute of Health and Welfare. pain in primary care: An international comparison. Spine 26: 2504–2513. Reid D, Larmer P, Robb G, Hing W, McNair P (2002) Use of a vignette to investigate the physiotherapy treatment of Korr IM (1975) Proprioceptors and somatic dysfunction. Journal an acute episode of low back pain: report of a survey of of the American Osteopathis Association 74: 638–650. New Zealand physiotherapists. New Zealand Journal of Physiotherapy 30: 26–32. Kusunose RS (1993) Strain and Counterstrain. In: Basmajian JV, Nyberg R (Eds) Rational Manual Therapies. Baltimore: Richardson C, Jull G, Hodges P, Hides J (1999) Therapeutic Williams and Wilkins. exercise for spinal segmental stabilisation in low back pain. Edinburgh: Churchill Livingstone. Kusunose RS, Wendorff R (1990) Strain and Counterstrain syllabus. Encinitas: Jones Institute. Stanton MR, Henschke N, Maher CG, Refshauge KM, Latimer J, McAuley JH (2008) After an episode of acute low back Lewis C, Flynn TW (2001) The Use of Strain-Counterstrain pain, recurrence is unpredictable and not as common as in the treatment of patients with low back pain. Journal of previously thought. Spine 33: 2923–2928. Manual and Manipulative Therapy 9: 92–98. Taylor SJ, Taylor AE, Foy MA, Fogg AJB (2001) Responsiveness Luo X, Pietrobon R, Sun SX, Liu GG, Hey L (2004). Estimates of common outcome measures for patients with low back and patterns of direct health care expenditures among pain. Spine 24: 1805–1812. individuals with back pain in the United States. Spine 29: 79–86. van Tulder M, Becker A, Bekkering T (2006) European guidelines for the management of acute nonspecific low March L, Trevena L, French S, Rebbeck T, Blyth F, Bellamy back pain in primary care. European Spine Journal 15 Suppl N (2004) Acute low back pain. In: Evidence-based 2: S169–S191. management of acute musculoskeletal pain: A Guide for Clinicians. Brisbane: Australian Academic Press, pp 19–29. van Tulder M, Malmivaara A, Esmail R, Koes B (2000) Exercise therapy for low back pain: a systematic review within the Merskey H (1994) Pain terms: a list with definitions and notes framework of the cochrane collaboration back review group. on usage recommended by the IASP Subcommittee on Spine 25: 2784–2796. taxonomy Pain 6: 249–252. Von Korff M, Saunders K (1996) The course of back pain in Nicholas M, Molloy A, Tonkin L, Beeston L (2007) Manage your primary care. Spine 21: 2833–2837. pain: practical and positive ways of adapting to chronic pain. Sydney: ABC Books. Walker BF, Muller R, Grant WD (2004) Low back pain in Australian adults. Prevalence and associated disability. Ebied I (&&- Ebi[d½i Ehj^efW[Z_Y [n[hY_i[  h[^WX_b_jWj_ed Journal of Manipulative and Physiological Therapeutics 27: fh[iYh_fj_ed a_j L>? ;n[hY_i[ 9ebb[Yj_edi$ mmm$[nhn$d[j% 238–244. Ijeh[%L>?%A_ji%Ehj^ef[Z_YH[^WX_b_jWj_ed$^jcb Ware JF, Sherbourne C (1992) The MOS 36-item short-form Patrick DL, Deyo RA, Atlas SJ, Singer DE, Chapin A, Keller RB survey (SF-36): I. Conceptual framework and item selection. (1995) Assessing health-related quality of life in patients with Medical Care 30: 473–483. sciatica. Spine 20: 1899–1908. Wilk V (2004) Acute low back pain: assessment and Pengel LHM, Herbert RD, Maher CG, Refshauge KM (2003) management. Australian Family Physician 33: 403–407. Acute low back pain: systematic review of its prognosis. BMJ 327: 323–327. Winkel D, Aufdemkampe G, Matthijs O, Meijer OG, Phelps V (1996) Diagnosis and treatment of the spine. Gaithersburg: Philips HC, Grant L (1991) The evolution of chronic back pain Aspen. problems: a longitudinal study. Behavior Research and Therapy 29: 435–441. 98 Journal of Physiotherapy 2011 Vol. 57 – © Australian Physiotherapy Association 2011

Appraisal Critically Appraised Papers Supervised aerobic and resistance exercise improves glycaemic control and modifiable cardiovascular risk factors in people with Type 2 diabetes mellitus Synopsis by an exercise specialist. Outcome measures: The primary outcome was the reduction in HbA1c (glycosylated Summary of: Balducci S et al (2010) Effect of an intensive haemoglobin) at 12 months. Secondary outcome measures exercise intervention strategy on modifiable cardiovascular were physical activity, and a range of cardiovascular risk risk factors in subject s with Type 2 diabetes mellitus. Arch factors including waist circumference, blood pressure, and Intern Med 170: 1794–1803. [Prepared by Nicholas Taylor, coronary heart disease risk scores. Results: 563 participants CAP Co-ordinator.] (93%) completed the study. The median exercise training attendance was 80%. At 12 months, the reduction in Question: Does an intensive exercise program improve HbA1c was significantly more in the exercise group by glycaemic control, physical activity, and modifiable 0.30% (95% CI 0.10 to 0.49). At 12 months, total physical cardiovascular risk factors in patients with Type 2 diabetes activity improved significantly more in the exercise group mellitus? Design: Randomised, controlled trial with than in the comparison group by 10 MET-h/wk (95% CI concealed allocation and blinded outcome assessment. 8.6 to 11.6). Waist circumference had decreased more in Setting: 22 diabetic outpatient clinics in Italy. Participants: the exercise group than in the comparison group by 3 cm The trial included sedentary patients with Type 2 diabetes. (95% CI 2.9 to 4.4), systolic blood pressure had reduced Any conditions limiting or contraindicating physical more in the exercise group than the comparison group by activity were exclusion criteria. Randomisation of 606 4.2 mm Hg (95% CI 1.6 to 6.9), and the coronary heart participants allocated 303 to the intervention group and disease risk score had reduced more in the exercise group 303 to a comparison group. Interventions: Both groups than in the comparison group by 3.1 units (95% CI 2.0 to received structured individual counselling every 3 months 4.0). Conclusion: Exercise was effective in improving over 12 months, which consisted of encouragement and glycaemic control, increasing physical activity, and strategies to achieve recommended levels of physical improving cardiovascular risk profile in sedentary people activity. In addition, the intervention group participated with Type 2 diabetes mellitus, providing benefits over and in an intensive exercise program. The 12 month exercise above individual counselling. program consisted of 150 minutes per week in 2 sessions of progressive aerobic and resistance exercises supervised training consisted of chest press, lateral pull-down, squat/ leg press, and abdominal exercises) over two sessions per Commentary week. Obesity and lack of physical activity are major risk factors The trial provides evidence that education alone is not for the development of Type 2 diabetes, and exercise (along adequate to cause sufficient behavioural change to reduce with medication and diet) has long been recognised as one risk factors related to diabetes and cardiovascular disease. of the three cornerstones of diabetic therapy (Irvine and It is evident that adults also need a practical component to Taylor 2009). This very large randomized controlled trial their learning in order to induce behavioural change that is provides further high quality evidence that high intensity adequate to obtain results. Exercise is a vital component of and progressive exercise can benefit people with Type 2 diabetes management and this trial is further evidence that diabetes. structured, supervised exercise sessions get results. Although the reduction in HbA1c of 0.30% found in this Casey Peiris trial may seem relatively small, any reduction in HbA1c La Trobe University and Eastern Health, Melbourne, is considered clinically significant as it is likely to reduce the risk of diabetic complications (Stratton et al 2000). We Australia also need to consider that the baseline HbA1c values of the participants in this trial were considered to be only slightly References elevated to start with; therefore a reduction of 0.30% in the exercise group allowed participants to achieve the American Diabetes Association (2008) Diabetes Care 31: s12– recommended target HbA1c value of less than 7.0% (ADA s54. 2008). Irvine C, Taylor N (2009) Aust J Physiother 55: 237–246. The combined intervention was replicable and feasible as it was held in community-type gyms using readily available Stratton I et al (2000) BMJ 321: 405–412. equipment (aerobic exercise consisted of either treadmill, step, elliptical, arm or cycle ergometer, and resistance 126 Journal of Physiotherapy 2011 Vol. 57 – © Australian Physiotherapy Association 2011

Appraisal Critically Appraised Papers Telephone-based patient self-management program might be effective in reducing osteoarthritis-related pain Synopsis plans. Participants in the health education group received written and audio materials regarding common health Summary of: Allen KD et al (2010) Telephone-based self- problems, as well as related screening recommendations. management of osteoarthritis: a randomized trial. Ann The health educator also called participants monthly for 12 Intern Med 153: 570–579. [Prepared by Kåre Birger Hagen months to review key points from the educational modules, and Margreth Grotle, CAPs Editors.] and assess whether participants were being screened appropriately. Outcome measures: The main outcome was Question: What are the comparative effects of telephone- the pain subscale of the Arthritis Impact Measurement based self-management support, health education Scales-2 (AIMS2). Secondary outcomes included the materials (attention control), or usual care for primary care AIMS2 physical function and affect subscales, the Arthritis patients with hip or knee osteoarthritis (OA)? Design: A Self-Efficacy Scale (ASES), and a 10-cm pain visual analog randomised clinical trial with equal assignment to three scale (VAS) measured at 12 months follow up. Results: 461 intervention groups. Setting: Primary care clinic, USA. (90%) participants completed the study. The mean AIMS-2 Participants: Men and women with a physician diagnosis pain score (range 0–10) in the self-management group was of hip or knee osteoarthritis, and persistent, current 0.4 points lower (95% CI −0.8 to 0.1) than in the usual care symptoms. Exclusion criteria included other rheumatologic group, and 0.6 points lower (CI −1.0 to −0.2) than in the conditions, psychoses, dementia, or being on a waiting health education group. The only significant differences list for arthroplasty. Randomisation of 523 participants between the groups in secondary outcome measures allocated 174 to self-management, 175 to health education, were for ASES in favour of self-management over health and 174 to usual care. Interventions: The self-management education (0.4 points, 95% CI 0.0 to 0.8) and VAS-pain intervention included two main components: providing in favour of self-management over health education (–1.0 education, and helping participants develop goals and action point, 95% CI –1.5 to –0.5) and usual care (–1.1 point, 95% plans related to osteoarthritis management. Participants CI –1.6 to –0.6). Health care use did not differ across the received written and audio versions of osteoarthritis self- groups. Conclusion: In patients with knee and hip OA, an management educational materials and exercises, and were entirely telephone based self-management support program asked to identify and write down goals and corresponding resulted in modest improvements in pain as compared to action plans related to their osteoarthritis symptoms and general health education and usual care. management. A health educator called participants monthly by telephone for 12 months to discuss key points from the in pain scores was found comparing the self-management educational modules and the participant’s goals and action group with the attention-control group, and not the usual care group. However, this lack of ‘attention effect’ is not Commentary addressed in the discussion. Potentially, the health education interventions increased attention towards screening and Osteoarthritis is a condition characterised by pain, disability awareness of potential health problems resulting in adverse and impaired quality of life. It is one of the leading causes effects. This study includes a relevant, low cost, feasible of pain and disability for the adult population worldwide, self-management support intervention. Telephone-based and the prevalence is increasing mainly due to the growing interventions are particular suitable for trials in rural proportion of elderly and overweight. The present study areas and for older persons with mobility limitations. As represents a timely and important contribution in relation this study mainly included men (93% of sample) who were to this large public health challenge. Self-management is overweight, further studies are warranted before the results recommended as a core treatment for hip and knee OA. can be generalised to a larger population. Recent meta-analyses show significant, but very small, effect sizes in improving pain and function. For telephone Nina Østerås, Rikke H Moe and Linda Fernandes interventions, effect sizes are comparable (Zhang 2010). National Resource Center for Rehabilitation in This trial is well conducted, has sufficient power, and includes an attention-control group with 12 months follow- Rheumatology, Diakonhjemmet Hospital, Norway up. The intervention effects, however, are small. Choosing the AIMS2 pain subscale as primary outcome could be Reference debated. First, as the intervention aimed to enhance self- efficacy, the ASES might have been a more appropriate Zhang W et al (2010): Osteoarthritis Cartilage 18: 476–499. primary outcome. Second, it is a composite score including different constructs (sleep, pain, stiffness). Third, the threshold for clinical important difference for this score is not known. It is interesting that the highest difference Journal of Physiotherapy 2011 Vol. 57 – © Australian Physiotherapy Association 2011 125

de Morton et al: Measuring mobility of patients in the Transition Care Program The de Morton Mobility Index (DEMMI) provides a valid method for measuring and monitoring the mobility of patients making the transition from hospital to the community: an observational study Natalie A de Morton1, Natasha K Brusco1,2, Lauri Wood3, Katherine Lawler2 and Nicholas F Taylor1,2 1La Trobe University, 2Eastern Health, 3Echuca Regional Health Australia Question: Is the de Morton Mobility Index (DEMMI) valid for measuring the mobility of patients making the transition from hospital to the community? Design: Observational cohort study. Participants: 696 consecutive patients admitted to 11 Transition Care Programs for multidisciplinary care in Victoria and Tasmania during a 6-month period. The DEMMI and Modified Barthel Index were administered within 5 working days of admission and discharge from the Transition Care Program. Outcome measures: The DEMMI and Modified Barthel Index. Results: Neither the DEMMI nor the Modified Barthel Index had a floor or ceiling effect. Similar evidence of convergent, discriminant and known-groups validity were obtained for each instrument. The DEMMI was significantly more responsive to change than the Modified Barthel Index using criterion- and distribution- based methods. The minimum clinically important difference estimates represented similar proportions of the scale width for the DEMMI and Modified Barthel Index and were similar using criterion- and distribution-based estimates. Rasch analysis identified the DEMMI as essentially unidimensional in a Transition Care Program cohort and therefore can be applied to obtain interval level measurement. Rasch analysis demonstrated that the DEMMI was administered similarly by physiotherapists and allied health assistants under the direction of a physiotherapist. Conclusion: The DEMMI and Modified Barthel Index are both valid measures of activity limitation for Transition Care Program patients. The DEMMI has a broader scale width, provides interval level measurement, and is significantly more responsive to change than the Modified Barthel Index for measuring the mobility of Transition Care Program patients. [de Morton /\" #SVTDP/, 8PPE - -BXMFS, 5BZMPS/'  5IFEF .PSUPO.PCJMJUZ*OEFY %&..* QSPWJEFTBWBMJENFUIPEGPSNFBTVSJOHBOENPOJUPSJOHUIFNPCJMJUZPGQBUJFOUTNBLJOH UIFUSBOTJUJPOGSPNIPTQJUBMUPUIFDPNNVOJUZBOPCTFSWBUJPOBMTUVEZJournal of Physiotherapyo> ,FZXPSET: Measurement, Older adults, Mobility, Physical ability, Physiotherapy Introduction Morton et al 2008a) and community settings (Davenport et al 2008). There are currently no best practice guidelines The Transition Care Program was established in 2004-05 regarding the optimal method for measuring activity as a jointly funded initiative between the Commonwealth limitation for patients making the transition from hospital and states and territories of Australia. It is provided to to the community. older persons at the end of a hospital stay in the form of a package of services (Department of Health and Ageing Physiotherapy focuses on the assessment and management 2008). Between October 2005 and February 2008 there of problems with movement (Jensen et al 1999). To conduct were 12 573 discharges from the Transition Care Program a rigorous evaluation of the efficacy of physiotherapy nationally (Department of Health and Ageing 2008). A for patients making the transition from hospital to the common component for all Transition Care Programs is community, a tool for measuring activity limitation the provision of allied health services to aid the assessment, that, in particular, measures the construct of mobility treatment and discharge planning of patients. accurately is required. According to the World Health Organisation International Classification of Functioning Across Australia current practice involves a broad range of ‘mobility’ is classified as one of nine domains of ‘activity models of care relating to the provision of Transition Care and participation’ and is defined as ‘moving by changing Program physiotherapy services and the use of allied health body position or location or by transferring from one place assistants. Also, a diverse range of outcome measures are to another, by carrying, moving or manipulating objects, applied. It is a current requirement that all Transition Care by walking, running or climbing, and by using various Programs apply the Modified Barthel Index at admission to forms of transportation’ (WHO 2001). An instrument that and discharge from the program (Department of Health and can be applied in a broad range of environments and that Ageing 2008). However, there is evidence that the Modified will accurately measure and monitor changes in mobility Barthel Index has a ceiling effect in older populations in for all patients in Transition Care Programs without floor or hospital (de Morton et al 2007) and community settings ceiling effects would have many benefits. (Hill et al 2008) and that it measures domains that are not relevant to physiotherapy interventions (de Morton et al In 2008, the de Morton Mobility Index (DEMMI) was 2008c). Systematic reviews have identified drawbacks in developed and validated in an older acute medical the use of other activity limitation measures in hospital (de population (de Morton et al 2008b); it has since been Journal of Physiotherapy 2011 Vol. 57 – © Australian Physiotherapy Association 2011 109

Research validated in subacute hospital (de Morton and Lane 2010) 5BCMFCharacteristics of participants. and community settings (Davenport and de Morton 2010, de Morton et al 2010). However, the DEMMI has not been Characteristic n = 696 validated for use during the transition from hospital to the community. The clinimetric properties of the DEMMI have Age (yr), mean (SD)a 82 (9) been evaluated extensively in a range of clinical populations Gender, female:male 411:280 and it is the first mobility instrument that can accurately Charlson comorbidity score, mean (SD) 1.7 (1.8) measure and monitor the mobility of older adults across Hospital setting before admission to TCP, acute, subacute, and community settings (Belvedere and n (%) 288 (41) de Morton 2010, Davenport et al 2008, de Morton et al 349 (50) 2008a). The DEMMI is a 15-item unidimensional measure Acute hospital of mobility and it appears to have face validity for the needs Sub-acute hospital 12 (2) of physiotherapists and their patients within Transition Care Other Programs. TCP setting, n (%) 259 (37) Hospital 272 (39) Therefore, the aim of this study was to validate the DEMMI Residential care 138 (20) in the Transition Care Program cohort and the secondary Community aim was to investigate whether it is valid for allied health ACAS at admission, n (%) 265 (38) assistants to administer the DEMMI to patients within the High level care 102 (15) Transition Care Program. The specific research questions Low level care 281 (40) of this study were: Other Previous gait aid, n (%) 137 (20) 1. Does the DEMMI have the properties required to None 56 (8) accurately measure and monitor the mobility of  Ij_Ya%YhkjY^[i patients transitioning from the hospital setting to the Frame 441 (63) community? Diagnosis, n (%) Inability to cope 94 (14) 2. Are DEMMI scores valid when it is applied by an  Kff[hb_cX%if_d[%f[bl_i\\hWYjkh[ 51 (7) allied health assistant with patients transitioning from  B em[hb_cX\\hWYjkh[%`e_djh[fbWY[c[dj% 147 (21) the hospital setting to the community? amputation  9e]d_j_l[%fioY^ebe]_YWb 58 (5) Method Neurological 110 (16)  C[Z_YWb%fW_d%fWbb_Wj_l[ 134 (19) Design Surgical 21 (3) Respiratory 74 (11) The mobility of consecutive Transition Care Program patients Modified Barthel Index, mean (SD)b 60 (29) was assessed by usual care physiotherapists or allied health de Morton Mobility Index, mean (SD)c 39 (20) assistants on admission to and prior to discharge from the TCP length of stay (days), mean (SD)d 42 (30) Transition Care Program using the DEMMI (de Morton et F^oi_ej^[hWfo%7>7%]hekf[n[hY_i[ 11 (12) al 2008b). All eligible patients received the Transition Care i[ii_edi%fWhj_Y_fWdj\"c[WdI:e Program’s usual multidisciplinary management. Mobility assessments were conducted within five business days of Missing data account for items with fewer than 696 data points. admission, discharge, or transfer from the Transition Care an = 649, bn = 670, cn = 678, dn = 637, en = 595. ACAS = Aged Program. As the nature of the Transition Care Program is Care Assessment Service, AHA = Allied Health Assistant, TCP = slow stream restorative care, with patients admitted for up Transition Care Program to 18 weeks, it was decided that it was appropriate to allow five business days to complete the assessment. medically contraindicated or if the patient was isolated due to infection or did not consent to the DEMMI mobility Baseline data were collected at initial assessment and assessment. Patients were also excluded if they had only one included age, gender, diagnosis, gait aid use, Transition assessment and it was determined that they did not require Care Program setting, admission Aged Care Assessment physiotherapy intervention as part of their management. Service assessment (ie, assessment related to suitability for high level, low level, or other care), Charlson comorbidity Outcome measures score (Charlson et al 1987), and the Modified Barthel Index (Shah et al 1989). The DEMMI is a mobility outcome measure that was recently developed in an older acute medical population (de Prior to the discharge mobility assessment, patients Morton et al 2008b). It consists of 15 items and is scored were asked, ‘How does your mobility compare to when on an interval level scale from 0 to 100 (de Morton et al you arrived in the Transition Care Program?’ Response 2008b). Eleven items are dichotomous (scored 0 or 1) and choices were based on a 5-point Likert scale (much worse, four items have three response options (scored 0, 1, or 2). A a bit worse, same, a bit better, or much better). Discharge raw ordinal DEMMI score out of 19 is then converted to an assessments followed the same procedures as initial interval-level DEMMI score out of 100 using a conversion assessments and included discharge destination. table. The DEMMI was reported to take an average of 8.8 Participants The 14 Transition Care Programs across Victoria and Tasmania were invited to participate in this study. Patients consecutively admitted to these programs were included. Patients were excluded if mobilisation was 110 Journal of Physiotherapy 2011 Vol. 57 – © Australian Physiotherapy Association 2011

de Morton et al: Measuring mobility of patients in the Transition Care Program 100 Mean = 39 100 Mean = 46 SD = 20 SD = 22 n = 678 n = 502 80 80 Frequency Frequency 60 60 40 40 20 20 0 20 40 60 80 100 0 20 40 60 80 100 0 DEMMI score on admission 0 DEMMI score on discharge 'JHVSFBde Morton Mobility Index 'JHVSFCde Morton Mobility Index (DEMMI) (DEMMI) scores on admission to the scores at discharge from the Transition Care Transition Care Program. 46 participants Program. 35 participants (7%) scored 0 and 6 (7%) scored 0 and 3 participants (0.4%) participants (1%) scored 100. scored 100. 100 100 Mean = 60 Mean = 65 SD = 29 SD = 33 80 n = 670 80 n = 595 Frequency Frequency 60 60 40 40 20 20 0 20 40 60 80 100 0 20 40 60 80 100 0 Barthel score on admission 0 Barthel score on discharge 'JHVSFBModified Barthel Index (MBI) 'JHVSFCModified Barthel Index (MBI) scores scores on admission to the Transition at discharge from the Transition Care Program. Care Program. 16 participants (2%) 43 participants (7%) scored 0 and 50 scored 0 and 16 participants (2%) scored participants (8%) scored 100. 100. minutes (SD 3.9) to complete in an older acute medical Data analysis population (de Morton et al 2008b). Validity: Convergent and discriminant validity for use The modified Barthel Index is an ordinal scale that provides of the DEMMI with this population were investigated by a total score between 0 and 100, where higher scores calculating the correlation between DEMMI and Modified indicate greater independence in the domains of mobility Barthel Index scores using Spearman’s rho and associated and continence (Shah et al 1989). The Barthel Index has 95% confidence bands. A significant, moderate to high been shown to have acceptable levels of inter-observer correlation between measures would provide evidence of and test-retest reliability (Collin et al 1988, Hachisuka and convergent validity. A low correlation of the DEMMI with Ogata 1997). The validity of the Barthel Index has been a measure of a different construct (Charlson Comorbidity widely tested and well established for rehabilitation patients Index) would provide evidence of discriminant validity. (Dewing 1992, Hachisuka and Ogata 1997). Known-groups validity (groups who would be expected to differ in their mobility) was investigated using an Journal of Physiotherapy 2011 Vol. 57 – © Australian Physiotherapy Association 2011 111

Research 5BCMF Convergent and discriminant validity of the de Morton Mobility Index and the Modified Barthel Index. Property Tool n Estimate (95% CI) Convergent validity DEMMI with MBI 655 0.75 (0.71 to 0.78) Discriminant validity DEMMI with Charlson Index 678 –0.11 (–0.18 to –0.04) MBI with Charlson Index 670 0.08 (0.01 to 0.16) DEMMI = de Morton Mobility Index, MBI = Modified Barthel Index 5BCMF Known-groups validity of the de Morton Mobility Index and the Modified Barthel Index. Mean (SD) scores for high and low residential care groups, and mean (95% CI) difference between groups. Tool Groups Difference between groups DEMMI High level Low level High minus low MBI residential care residential care –25 (–30 to –20) 33 (19) 59 (13) 38 (33 to 43) (n = 185) (n = 65) 85 (13) 47 (28) (n = 228) (n = 71) DEMMI = de Morton Mobility Index, MBI = Modified Barthel Index Modified Barthel Index score 100 y = 1.2x + 15.2 90 80 70 60 50 40 30 20 10 0 0 10 20 30 40 50 60 70 80 90 100 DEMMI score 'JHVSF Scatterplot of the de Morton Mobility Index (DEMMI) scores and the Modified Barthel Index scores at admission. independent t-test to compare scores obtained for those who Responsiveness to change: Responsiveness to change were discharged to low level care (eg, hostel) compared to was evaluated using a criterion-based method (Guyatt high level care (eg, nursing home). responsiveness index, Guyatt et al 1987) and a distribution- based method (the Effect Size Index, Kazis et al 1989). Floor and ceiling effects were reported for each measure if Effect size indices of 0.2, 0.5, and 0.8 have been reported 15% or more of the participant population scored the lowest to represent small, moderate and large responsiveness to or highest scale score, respectively. change, respectively (Husted et al 2000). 112 Journal of Physiotherapy 2011 Vol. 57 – © Australian Physiotherapy Association 2011


Like this book? You can publish your book online for free in a few minutes!
Create your own flipbook