Farlie et al: Reporting the intensity of balance exercise Schepens S, Goldberg A, Wallace M (2010) The short version Thompson WR, Gordon NF, Pescatello LS (2010) ASCM’s of the Activities-specific Balance Confidence (ABC) scale: guidelines for exercise testing and prescription (8th edn). Its validity, reliability, and relationship to balance impairment Philadelphia: Wolters Kluwer. and falls in older adults. Archives of Gerontology and Geriatrics 51: 9–12. Tiedemann A, Sherrington C, Close JCT, Lord SR (2011) Exercise and Sports Science Australia Position Statement Sherrington C, Tiedmann A, Fairhall N, Close J, Lord S (2011) on exercise and falls prevention in older people. Journal of Exercise to prevent falls in older adults: an updated meta- Science and Medicine in Sport 14: 489–495. analysis and best practice recommendations. NSW Public Health Bulletin 22: 78–83. Tinetti M (1986) Performance-oriented assessment of mobility problems in elderly patients. Journal of the American Shigematsu R, Okura T, Nakagaichi M, Tanaka K, Sakai T, Geriatrics Society 34: 119–126. Kitazumi S, et al (2008) Square-stepping exercise and fall risk factors in older adults: a single-blind, randomized Tinetti M, Baker D, McAvay G, Claus E, Garrett P, Gottschalk controlled trial. The Journals of Gerontology 63A: 76. M, et al (1994) A multifactorial intervention to reduce the risk of falling among elderly people living in the community. The Silsupadol P, Shumway-Cook A, Lugade V, van Donkelaar P, New England Journal of Medicine 331: 821–827. Chou L-S, Mayr U, et al (2009) Effects of single-task versus dual-task training on balance performance in older adults: Urbscheit NL, Wiegand MR (2001) Effect of two exercise a double-blind, randomized controlled trial. Archives of programs on balance scores in elderly ambulatory people. Physical Medicine and Rehabilitation 90: 381–387. Physical & Occupational Therapy in Geriatrics 19: 49–58. Silsupadol P, Siu K-C, Shumway-Cook A, Woollacott MH van Uffelen JGZ, Chin A Paw MJM, van Mechelen W, Hopman- (2006) Training of balance under single-and dual-task Rock M (2008) Walking or vitamin B for cognition in older conditions in older adults with balance impairment. Physical adults with mild cognitive impairment? A randomised Therapy 86: 269–281. controlled trial. British Journal of Sports Medicine 42: 344– 351. Simpson JM, Worsfold C, Fisher KD, Valentine JD (2009) The CONFbal scale: a measure of balance confidence: a key Wolf SL, Sattin RW, Kutner M, O’Grady M, Greenspan AI, outcome of rehabilitation. Physiotherapy 95: 103–109. Gregor RJ (2003) Intense tai chi exercise training and fall occurrences in older, transitionally frail adults: a randomized, Sjösten NM, Salonoja M, Piirtola M, Vahlberg T, Isoaho R, controlled trial. Journal of the American Geriatrics Society Hyttinen H, et al (2007) A multifactorial fall prevention 51: 1693–1701. programme in home-dwelling elderly people: A randomized- controlled trial. Public Health 121: 308–318. Wolfson L, Whipple R, Derby C, Judge J, King M, Amerman P, et al (1996) Balance and strength training in older adults: Skinner J, Hustler R, Bergsteinova V, Buskirk E (1973) The Intervention gains and tai chi maintenance. Journal of the validity and reliability of a rating scale of perceived exertion. American Geriatrics Society 44: 498–506. Medicine and Science in Sports 5: 94–96. Zhang JG, Ishikawa-Takata K, Yamazaki H, Morita T, Ohta T Thompson B (2007) Capture-recapture inspection template (2006) The effects of tai chi chuan on physiological function GPS &YDFM IUUQMFBOTPGUXBSFFOHJOFFSJOH DPNDBQUVSF and fear of falling in the less robust elderly: An intervention SFDBQUVSFJOTQFDUJPO<\"DDFTTFE0DUPCFS > study for preventing falls. Archives of Gerontology and Geriatrics 42: 107–116. Journal of Physiotherapy 2013 Vol. 59 – © Australian Physiotherapy Association 2013. Open access under CC BY-NC-ND license. 235
Editorial Open access to Journal of Physiotherapy Robert D Herbert1 and Cris Massis2 1Chair, Editorial Board, Journal of Physiotherapy, 2Chief Executive Officer, Australian Physiotherapy Association Australia In 2014, as Journal of Physiotherapy enters its 60th year Publication fees present little problem when the research is of publication, it will undergo one of the most significant supported by large grants, or by a pharmaceutical company, developments in its history. From January 2014 the or by the producer of a medical device, but they constitute Australian Physiotherapy Association will provide open a real impediment to publication for physiotherapy access to Editorials and all research articles published in researchers, many of whom conduct their research with little Journal of Physiotherapy. A unique feature of the new or no funding support. If any of the existing physiotherapy publication model is that access to research content will be journals was to charge a publication fee it would find that the free for readers and its publication will be free for authors. number of manuscripts submitted for publication dropped This initiative is part of the Association’s strategic plan. quickly. Consequently, while some non-core physiotherapy journals have embraced an open access model (www.doaj. For the last 60 years Journal of Physiotherapy has org), and several core physiotherapy journals provide open employed the same publishing model that is used by the access to content that is over one year old, none of the core overwhelming majority of scientific journals: journal physiotherapy journals (Costa et al 2010) has been made content has been made available to those who pay for it. This open access. means that, in addition to being made available to members of the Australian Physiotherapy Association, Journal of The Board of Directors of the Australian Physiotherapy Physiotherapy has been accessible to staff of universities Association has worked with the Editorial Board of and hospitals with institutional subscriptions, individuals Journal of Physiotherapy to create a new model of open with personal subscriptions, and those prepared to pay for access publishing in which (unlike in traditional publishing each article accessed. But that is all. Many potential readers models) content is provided free to readers and (unlike never see the contents of the Journal. existing open access models) publication is free to authors. The Association’s Board of Directors recognises that if The traditional publishing model is unsatisfactory from its flagship journal is to be the world’s best physiotherapy several perspectives. Research funding bodies invest journal it must exploit innovative publishing models. And enormous sums in research, researchers spend years the Association has embraced its role in providing the conducting research, and patients volunteer to participate information infrastructure needed to support evidence- in research, all with the objective of improving clinical based practice. In this way the Australian Physiotherapy practice. But traditional publishing models restrict access Association can build capacity in the physiotherapy to research findings behind pay walls, subscriptions, and profession in Australia, the region, and globally. The user fees, making research findings accessible to only a few. production and wide dissemination of a high quality journal Most research never reaches most of the people who would is the ultimate demonstration to governments and health like to read about it. service providers that physiotherapy is a vibrant, research- based, scientific profession. In the last decade there has been a strong push towards open access publishing – the provision of unrestricted, Journal of Physiotherapy has become one of the world’s free, online access to journal content. Open access has leading physiotherapy journals because it has established been advocated by researchers who want to provide broader a reputation for providing high quality and rapid reviews access to the findings of their research (see, for example, for authors. There is empirical evidence that the quality of http://pkp.sfu.ca/about). Recently open access has been randomised trials of physiotherapy interventions published mandated by several major research funding bodies. The in Journal of Physiotherapy is higher than in any other US National Institutes of Health, the Wellcome Trust, journal (Costa et al 2010). For these reasons the journal has the UK Medical Research Council, and the Australian attracted high quality submissions and is highly cited. The NHMRC all now require that reports of research funded adoption of this new publishing model should see a new by these agencies are given open access within 12 months phase of growth. We hope that researchers will submit of the initial publication. There are compelling ethical their best research knowing that, from 2014, it will be more arguments to prefer open access publishing over traditional accessible and more widely read in Journal of Physiotherapy publishing models (Parker 2013), and there is evidence from than in any other physiotherapy journal. a randomised trial that open access articles are much more widely read (Davis 2010). Now open access publishing has Correspondence: Professor Rob Herbert, Neuroscience become well established in some areas of science. That Research Australia, Australia. Email: [email protected]. is a good thing because it enables wide dissemination of au research findings to the clinicians and researchers and members of the general public who want to read about it. References One major hurdle has so far prevented all core physiotherapy Costa LOP et al (2010) Physical Therapy 90: 1631–1640. journals (Costa et al 2010) from instituting open access policies: someone has to pay, and in open access models that Davis PM (2010) Physiologist 53: 200–191. is usually the author. All major open access journals charge authors a fee to publish, and the fee is usually substantial. Parker M (2013) BMC Medical Ethics 14: 16. Website XXXFMTFWJFSDPNMPDBUFKQIZT Journal of Physiotherapy 2013 Vol. 59 – © Australian Physiotherapy Association 2013. Open access under CC BY-NC-ND license. 217
Coulter et al: Physiotherapy rehabilitation after total hip replacement Physiotherapist-directed rehabilitation exercises in the outpatient or home setting improve strength, gait speed and cadence after elective total hip replacement: a systematic review Corinne L Coulter1,2, Jennie M Scarvell3, Teresa M Neeman4 and Paul N Smith5 1Australian National University Medical School, 2Physiotherapy Department, Canberra Hospital, 3Physiotherapy, University of Canberra, 4Statistical Consulting Unit, Australian National University, 5Trauma and Orthopaedic Research Unit, Canberra Hospital Australia Question: In people who have been discharged from hospital after a total hip replacement, do rehabilitation exercises directed by a physiotherapist improve strength, gait, function and quality of life? Are these exercises as effective in an unsupervised home-based setting as they are in a supervised outpatient setting? Design: Systematic review with meta- analysis of randomised trials. Participants: Adult patients after elective total hip replacement. Intervention: Physiotherapist- directed rehabilitation exercises after discharge from hospital following total hip replacement. Outcome measures: Hip and knee strength, gait parameters, functional measures, and quality of life. Results: Five studies comprising 234 participants were included in the review. Sufficient data for meta-analysis were only obtained for hip and knee strength, gait speed and cadence. Physiotherapy rehabilitation improved hip abductor strength by a mean of 16 Nm (95% CI 10 to 22), gait speed by 6 m/min (95% CI 1 to 11) and cadence by 20 steps/min (95% CI 8 to 32). Favourable but non-significant improvements in strength were noted for other muscle groups at the hip and knee. Function and quality of life could not be meta-analysed due to insufficient data and heterogeneity of measures, but functional measures tended to favour the physiotherapy rehabilitation group. Most outcomes were similar between outpatient and home-based exercise programs. Conclusion: Physiotherapy rehabilitation improves hip abductor strength, gait speed and cadence in people who have been discharged from hospital after total hip replacement. Physiotherapist-directed rehabilitation exercises appear to be similarly effective whether they are performed unsupervised at home or supervised by a physiotherapist in an outpatient setting. <$PVMUFS $- 4DBSWFMM +. /FFNBO 5. 4NJUI 1/ 1IZTJPUIFSBQJTUEJSFDUFE SFIBCJMJUBUJPO FYFSDJTFT JO UIF PVUQBUJFOU PS IPNF TFUUJOH JNQSPWF TUSFOHUI HBJU TQFFE BOE DBEFODF BGUFS FMFDUJWF UPUBM IJQ SFQMBDFNFOU a systematic review. Journal of Physiotherapyo> Key words: Physiotherapy, Rehabilitation, Exercise, Total hip replacement, Physical therapy Introduction physiotherapy is in terms of restoring a patient’s physical health. Osteoarthritis is the most common reason for hip joint replacement surgery in Australia (Australian Orthopaedic Rehabilitation protocols after total hip replacement Association 2011) and, based on current trends, is vary widely in both the specific exercises used and the forecast to become the fourth leading cause of disability timeframes for their delivery (Roos et al 2003). This may worldwide by 2020 (Woolf and Pleger 2003). Osteoarthritis be because they are largely based on clinical experience, causes a substantial burden with impairments not only to physical status and independence but also to quality of What is already known on this topic: Osteoarthritis life. In Australia the pain and disability associated with is a common cause of disability and each year osteoarthritis affect approximately 10% of men and 18% of more total hip replacements are performed. women over 60 years of age (AIHW 2004). Impairments and functional limitations can persist after surgery. Rehabilitation protocols after total hip The rate of hip replacement surgery continues to increase. In replacement vary widely, perhaps because previous Australia, 35 996 hip replacements were performed in 2010, systematic reviews have been unable to make clear an increase of 3.6% compared to 2009. Since 2003, the first recommendations about physiotherapy exercises in year of complete national data collection by the Australian this setting. Orthopaedic Association National Joint Replacement Registry, the number of hip replacements has increased by What this study adds: Physiotherapist-directed 32.4% (Australian Orthopaedic Association 2011). rehabilitation exercises improve hip abductor strength, gait speed, and cadence in people after total hip Traditionally, physiotherapy has been a routine component replacement. The effects on functional measures and of patient rehabilitation following hip replacement surgery. quality of life were less clear, but tended to favour the Impairments and functional limitations remain a year after intervention group. Rehabilitation in the supervised surgery (Minns Lowe 2009, Trudelle-Jackson and Smith outpatient setting or as a home-based program 2004), so it is valid to consider how effective post-discharge seems to provide similar benefits. Journal of Physiotherapy 2013 Vol. 59 – © Australian Physiotherapy Association 2013. Open access under CC BY-NC-ND license. 219
Research surgeon restrictions and anecdotal reports, in the absence of Titles and abstracts screened evidence to direct post-discharge rehabilitation. (n = 3096) One systematic review has examined the extent to which Papers excluded after screening physiotherapy exercise is effective following discharge UJUMFTBCTUSBDUT O after total hip replacement, but this was limited to evidence published in 2004 or earlier (Minns Lowe 2009). This Potentially-relevant papers retrieved for review concluded that ‘insufficient evidence currently exists evaluation of full text (n = 32) to establish the effectiveness of physiotherapy exercise following primary hip replacement for osteoarthritis’. The Papers excluded after evaluation of review considered walking speed, hip abductor strength, full text (n = 27) function, range of motion, and quality of life. However, data t not randomised (n = 16) for only the first two of these outcomes were meta-analysed, t JOUFSWFOUJPOQSPWJEFEJOSFIBC due to variable study quality, clinical heterogeneity, limited data or a combination of these problems. The meta-analytic inpatient setting (n = 10) summaries of the data indicated promise but, as the pooled t comparison with pre-operative results were not statistically significant, definitive answers were unable to be derived from this review. intervention (n = 1) Therefore, we aimed to answer the following research Papers included in review (n = 5) questions: Figure 1. Flow of studies through the review. 1. In people who have been discharged from hospital after a total hip replacement, do rehabilitation (eg, health care delivery, home physiotherapy, home exercises directed by a physiotherapist improve rehabilitation, and self-care). See Appendix 1 on the strength, gait, function and quality of life? eAddenda for the full search strategy. 2. Are these exercises as effective in an unsupervised A single reviewer screened the titles and abstracts of all home-based setting as they are in a supervised the items retrieved by the searches to identify potentially outpatient setting? relevant studies. Full text copies of relevant studies were retrieved and reviewed. The reference lists of these papers Method were then screened for further relevant studies. Each paper obtained in full text was examined for eligibility against the Identification and selection of studies review’s inclusion criteria by two reviewers (CC and JS). Disagreements were resolved by discussion. The inclusion Literature searches were conducted for relevant articles criteria for the review are presented in Box 1. published in English in five databases (MEDLINE, CINAHL, EMBASE, PEDro, and the Cochrane Library) Assessment of study characteristics from the earliest record to March 2012. The search terms Quality: Trials meeting the inclusion criteria were assessed included terms for total hip replacement or arthroplasty, for methodological quality using the PEDro scale (Maher terms for physiotherapy such as rehabilitation or physical et al 2003) by two reviewers (CC and JS). Each assessor therapy, and terms relating to patient discharge (eg, post worked independently. Following assessment, any discharge, after discharge, or outpatient) or home services disagreements were resolved by discussion. The ten internal validity items of the PEDro scale were reported as a total #PY. Inclusion criteria. score (de Morton 2009). The external validity item, which requires both the source of participants and the eligibility Design criteria to be reported, was also determined for each trial. t Randomised trials The PEDro scale scores were used to characterise the trials t English language but were not used to exclude trials from the review or the Participants meta-analyses. t Adults after total hip replacement Interventions Participants and interventions: Interventions involving t Post-discharge physiotherapist-directed early rehabilitation during the hospital inpatient phase, post-acute inpatient rehabilitation, and rehabilitation in rehabilitation exercises (outpatient or home-based) residential care (or comparison to any of these) were not considered by this review. Outcomes measured t Muscle strength t Gait t Function t Quality of life Comparisons t Post-discharge physiotherapist-directed rehabilitation exercises (outpatient or home-based) versus no intervention t Physiotherapist-supervised, outpatient rehabilitation exercises versus physiotherapist-directed, unsupervised, home-based rehabilitation exercises 220 Journal of Physiotherapy 2013 Vol. 59 – © Australian Physiotherapy Association 2013. Open access under CC BY-NC-ND license.
Coulter et al: Physiotherapy rehabilitation after total hip replacement Total Outcomes: The outcomes considered by the review were (0 to 10) muscle strength, gait, function and quality of life. From each trial, data were extracted for these outcome measures, 4 where available, at the beginning of the intervention and at 6 the longest follow-up assessment point. 6 7 Data analysis 6 Data were extracted from each trial regarding sample size, Point estimate population characteristics, details of the interventions, and variability and the effects of interventions. Where outcome measures were reported in two or more trials and were reported by reported population descriptors (mean and standard deviation), Y meta-analyses were performed using standard softwarea. Y Where only one trial reported a particular measure, meta- Y analysis was not used but the data were reported in the text Y as a between-group difference with a 95% CI. Y To determine the effect of intervention, experimental and Between-group control groups were compared. Where a trial employed two difference variations of physiotherapy intervention, the outcomes of reported the two intervention groups within that trial were pooled Y before performing this meta-analysis. Also, to determine Y which mode of post-discharge physiotherapy provides Y better patient outcomes following total hip replacement, we Y meta-analysed the studies in which outpatient and home- Y based exercise programs were compared. Intention-to- Forest plots were created to display effect estimates with treat analysis 95% CIs for individual trials and pooled results. In each case we tested for statistical heterogeneity. This was examined N graphically on the forest plot and statistically through the Y calculation of the I2 statistic. The I2 statistic estimates the Y percentage of the variability in effect estimates that is Y due to heterogeneity rather than sampling error (chance). Y An I2 value greater than 50% was considered substantial heterogeneity and random-effects meta-analysis rather that < 15% Y N Y Y NNN Y N Y N NNY a fixed-effect model was used in these instances. dropouts Y Y N N NYY Y N Y N NNY Results N Flow of studies through the review Assessor blinding The search returned 3096 studies. By screening titles and abstracts, 32 potentially relevant studies were identified N and retrieved in full text. Of these, 27 studies failed to meet the eligibility criteria. Therefore five studies were included Therapist in the review. The flow of studies through the review is blinding presented in Figure 1. N Characteristics of studies Participant Three trials compared an experimental group to a control blinding group (Johnsson et al 1988, Jan et al 2004, Trudelle-Jackson N and Smith 2004), one trial compared two experimental groups (Galea et al 2008), and one trial compared two Groups experimental groups to a control group (Unlu et al 2007). similar at For the comparison of experimental versus control, the baseline outcomes of the two experimental groups in the trial by Unlu et al (2007) were pooled before including this trial in Y the meta-analysis. For the comparison of outpatient versus home-based exercise, the two experimental groups were 5BCMF. PEDro scores of included trials. Random Concealed compared. The quality of the trials is summarised in Table allocation allocation 1 and the characteristics of the participants, interventions and outcome measures are presented in Table 2. YN Quality: The trials included in this review had varying Study internal validity with scores ranging from four to seven Johnsson 1988 Trudelle-Jackson 2004 Jan 2004 Unlu 2007 Galea 2008 Y = yes, N = no. Journal of Physiotherapy 2013 Vol. 59 – © Australian Physiotherapy Association 2013. Open access under CC BY-NC-ND license. 221
222 Journal of Physiotherapy 2013 Vol. 59 – © Australian Physiotherapy Association 2013. Open access under CC BY-NC-ND license. 5BCMF. Characteristics of the five randomised trials included in the review. Trial Par ticipants Inter ventions Galea n = 23 Exp 1 = outpatient exercise program (figure of 8 walk 2008 abd, heel raise, side stepping) Age (yr) = Exp 69 3FHJNFO NB YNJOQFSFY XLYXL QSFTDSJC (SD 10), Con 67 (SD Progression = figure of 8 walk: increase laps, adding 8) combination of the previous; STS, active SLS, steps: raise: increase reps, add weight; side stepping: incre Time since THR = at Exp 2 = same exercise program as Exp but without s hospital discharge &YQ IPNFFYFSDJTFQSPHSBN IJQ'30.CJMBUFSBMMZ Jan 2004 n = 53 bilaterally, 30 min walk Regimen = daily x 12 wk, hip F ROM (10 reps x 2 set Age (yr) = Exp 59 groups with weight on ankle (10 reps x 2 sets), hip ab (SD 12), Con 57 (SD sets) 13) Progression = not reported Con = no intervention Time since THR = > 1.5yrs post THR Exp = outpatient exercise program (supine strengthe hamstrings and quadriceps, SLS, stairs, STS, walkin Johnsson n = 30 3FHJNFO NJOTFTTJPOT XLGPSNP UIFOFJUI 1988 for 2 mo Age (yr) = Exp 70, Progression = not reported Con 66 Con = no intervention Time since THR = 2 Exp = home exercise program (STS, unilateral heel r mo post-op balance in SLS, marching, side and back leg raises i raising and lowering in standing) Trudelle- n = 28 Regimen = 15 reps 3-4 x wk for 8 wks Jackson Progression = 20 reps at 2 wk, 15 reps x 2 sets at 4 w 2004 Age (yr) = Exp 59 Con = 7 basic isometric and active ROM exercises (g (SD 11), Con 60 (SD hamstring sets, ankle pumps, heel slides, hip abduct 12) supine Time since THR = Exp 1 = home exercise program (ROM, isometric an 4–12 mo post-op bilaterally) Exp 2 = outpatient exercise program (as above) Unlu 2007 n = 26 Regimen: 2 x daily, 6 wks Progression: not reported Age = Exp 1 45 (SD Con = no intervention 9), Exp 2 58 (SD 7), Con 53 (SD 10) Time since THR = 12–24 mo post-op abd = abduction, AQoL = Assessment of Quality of Life, Con = control group, E = extension, ER = e item hip questionnaire, IR = internal rotation, QOL = quality of life, reps = repetitions, ROM = Range Osteoarthritis Index, 6MW T = six-minute walk test.
k, STS, active SLS, stairs, hip Primary outcomes Assessments Research Initial: Wk 0 of CFEIPNFFYFSDJTFQSPHSBN (BJUTQFFE DNT program g obstacles, changing surfaces, DBEFODF TUFQNJO TUFQ Final: Wk 8 : increase reps; hip abd, heel length (cm) (completion) ease total distance of program super vision Function: TUG (sec), TUBJSTQPXFS /NT Initial: Wk 0 of Z IJQBCE'&TUSFOHUIFOJOH 6MW T (m), WOMAC program Final: Wk 12 ts), isotonic exercise all muscle QOL: AQoL (completion) bd in SLS (10 reps with 5s hold x 2 of program Strength: Hip abd, E, F (Nm) (BJUTQFFE NNJO Function: HHS (functional part) ening of the abdominals, gluteals, Strength: Hip abd, E, F, Initial: 2 mo ng exercises) Knee E (Nm) post-operative IFS XLGPSNPPSFWFS ZXL Final: 6 mo post-operative raises, partial knee bends, Strength: Hip abd, E, F, Initial: Wk 0 of in standing, unilateral pelvic Knee E (Nm) program wk, 20 reps x 2 sets at 6 wk Function: HQ12 Final: Wk 8 gluteals, quadriceps sets, (completion) tion in supine, hip IR and ER in of program nd eccentric hip exercises Strength: Hip abd (ft.lbs) Initial: Wk 0 of program (BJUTQFFE NNJO DBEFODF TUFQTNJO Final: Wk 6 (completion) of program external rotation, Exp = experimental group, F = flexion, HHS = Harris Hip Score, HQ12 = 12- of Motion, STS = sit to stand, SLS = single-leg stance, WOMAC = Western Ontario McMaster
Coulter et al: Physiotherapy rehabilitation after total hip replacement out of ten. All trials used true random allocation of as presented in Figure 4. See also Figure 5 on eAddenda for participants and had sufficient statistical information to detailed forest plot. The best estimate of the effect on hip make their results interpretable. Only one trial (Unlu et al flexor strength was an improvement of 6 Nm (95% CI –2 to 2007) reported concealment of allocation and blinding of 13) as presented in Figure 6. See also Figure 7 on eAddenda assessors. The PEDro scale criterion that relates to external for detailed forest plot. Two of the three trials included in validity but which does not contribute to the PEDro score these meta-analyses assessed a home-based intervention. was met by all trials. Four of the five trials scored six or more out of the possible ten points. The exercises also did not significantly improve knee extensor strength, although the trend was again favourable Participants: The sample size of the studies ranged from 23 with a mean between-group difference of 42 Nm (95% CI to 53. The time of recruitment of participants varied from at –4 to 89) as presented in Figure 8. See also Figure 9 on discharge from hospital after total hip replacement to 12–24 eAddenda for detailed forest plot. One of the two trials months after the procedure. assessed a home-based intervention. Interventions: The included trials varied in their Gait: Rehabilitation exercises after discharge were effective experimental interventions. One trial assessed a supervised for improving gait speed by 6 m/min (95% CI 1 to 11) as outpatient program (Johnsson et al 1988), three trials presented in Figure 10. See also Figure 11 on eAddenda assessed a home-based exercise program (Jan et al 2004, for detailed forest plot. Rehabilitation exercises also Trudelle-Jackson and Smith 2004, Unlu et al 2007) and significantly improved cadence by a mean of 20 steps/min two trials compared a home-based program to a supervised (95% CI 8 to 32) in the one trial that measured it (Unlu et outpatient program (Galea et al 2008, Unlu et al 2007). al 2007). Note that this result pools the final data from a Three papers included a true control group, who received home-based group and a supervised outpatient group, and no therapeutic intervention (Johnsson et al 1988, Jan et al compares them to the control group. 2004, Unlu et al 2007). The duration of the interventions ranged from six weeks (Unlu et al 2007) to three months Function: The tools used to measure self-reported function (Jan et al 2004, Johnsson et al 1988). varied between the trials. Jan et al (2004) used the Harris Hip Score, which ranges from 0 (lowest function) to 14 Outcomes: All trials recorded outcomes at the end of the (highest function). Although the Harris Hip Score data in intervention (ie, six weeks, eight weeks or three months). this study indicate a statistically significant benefit from Only one trial followed up beyond the intervention period the exercises, the mean between-group estimate equates (Johnsson et al 1998). Most trials measured strength of hip to only 0.9 points (95% CI 0.2 to 1.6). The authors in this abductors (Jan et al 2004, Johnsson et al 1988, Trudelle- study noted that the participants with higher compliance Jackson and Smith 2004, Unlu et al 2007), hip extensors had a greater benefit. Trudelle-Jackson and Smith (2004) (Jan et al 2004, Johnsson et al 1988, Trudelle-Jackson used the 12-item Hip Questionnaire to measure self- and Smith 2004), hip flexors (Jan et al 2004, Johnsson reported function and reported a significant between-group et al 1988, Trudelle-Jackson and Smith 2004), and knee difference in medians of 1.5 points (p = 0.01) on this scale extensors (Johnsson et al 1988, Trudelle-Jackson and Smith from 12 (least difficulties) to 60 (most difficulties) favouring 2004). Gait parameters were included as outcomes in all the experimental group. five trials. Three trials measured gait speed (Galea et al 2008, Jan et al 2004, Unlu et al 2007) and two measured Quality of life: None of the studies comparing rehabilitation cadence (Galea et al 2008, Unlu et al 2007). Although exercise after discharge to a no-intervention control three trials included a self-reported functional measure, measured quality of life. the Western Ontario McMaster Universities Osteoarthritis Index (WOMAC) score (Ehrich et al 2000), the 12-Item Effect of home-based versus outpatient Hip Questionnaire (Dawson et al 1996), and the Harris SFIBCJMJUBUJPOFYFSDJTFT Hip Score (Harris 1969), no two studies used the same measure. Objective functional measures, including stair Strength: Only one trial compared the effect of home- climbing or the 6MWT, varied among the trials. Only one based and supervised outpatient rehabilitation exercises on trial used a generic quality of life measure – the Assessment muscle strength (Unlu et al 2007). Although hip abduction of Quality of Life questionnaire (Hawthorne et al 1999). in both groups improved, the supervised exercise group Because of these differences, function scores and quality of improved by 5.4 Nm more, which the authors reported was life measures were not meta-analysed and are reported as statistically significant. However, there were very large individual results in the text. baseline differences between the groups, which may have influenced their response to the intervention. &GGFDUPGSFIBCJMJUBUJPOFYFSDJTFT Gait: The data from two trials (Galea et al 2008, Unlu et al Strength: Rehabilitation exercises after discharge were 2007) were pooled to compare the effects of home-based and effective for improving hip abductor strength, with a mean supervised outpatient exercises on gait speed and cadence. between-group difference of 16 Nm (95% CI 10 to 22) as Gait speed was not significantly improved by supervision of presented in Figure 2. See also Figure 3 on eAddenda for the exercises, with a mean difference of 8 m/min (95% CI detailed forest plot. For two of the four trials included in –9 to 24), as presented in Figure 12. See also Figure 13 on this meta-analysis, the intervention was home-based. eAddenda for detailed forest plot. Similarly, cadence was not significantly improved by supervision in the same trials The exercises did not, however, have statistically significant (mean difference 2 steps/min, 95% CI –4 to 8), as presented effects on the strength of the hip extensors and flexors. The in Figure 14. See also Figure 15 on eAddenda for detailed best estimate of the effect on hip extensor strength was close forest plot. Galea et al (2008) also measured step length, to significant – an improvement of 21 Nm (95% CI –2 to 44) which did not significantly differ (mean difference 1 cm longer in the supervised exercise group, 95% CI –6 to 7). Journal of Physiotherapy 2013 Vol. 59 – © Australian Physiotherapy Association 2013. Open access under CC BY-NC-ND license. 223
Research Jan 2004 Jan 2004 Johnsson 1998 Unlu 2007 Trudelle-Jackson 2004 Unlu 2007 –50 –25 0 25 50 –50 –25 0 25 50 Favours control Favours experimental Favours control Favours experimental Figure 2. Mean difference (95% CI) of the effect of Figure 10. Mean difference (95% CI) of the effect of rehabilitation exercises on hip abductor strength (Nm) in SFIBCJMJUBUJPOFYFSDJTFTPOHBJUTQFFE NNJO JOUXP four studies (n = 137). studies (n = 79). Jan 2004 Galea 2008 Johnsson 1998 Unlu 2007 Trudelle-Jackson 2004 –50 –25 0 25 50 –50 –25 0 25 50 Favours control Favours experimental Favours control Favours experimental 'JHVSFMean difference (95% CI) of the effect of Figure 12. Mean difference (95% CI) between the effects rehabilitation exercises on hip extensor strength (Nm) in of home-based and supervised outpatient rehabilitation three studies (n = 111). FYFSDJTFTPOHBJUTQFFE NNJO JOUXPTUVEJFT O Jan 2004 Galea 2008 Johnsson 1998 Unlu 2007 Trudelle-Jackson 2004 –50 –25 0 25 50 –50 –25 0 25 50 Favours control Favours experimental Favours control Favours experimental 'JHVSFMean difference (95% CI) of the effect of 'JHVSFMean difference (95% CI) between the effects rehabilitation exercises on hip flexor strength (Nm) in of home-based and supervised outpatient rehabilitation three studies (n =111). FYFSDJTFTPODBEFODF TUFQTNJO JOUXPTUVEJFT O Johnsson 1998 Function: Only the trial by Galea et al (2008) measured Trudelle-Jackson 2004 function, with both self-reported and objective measures being used. The self-reported outcome was the WOMAC –50 –25 0 25 50 score, which has three domains: pain, stiffness, and Favours control Favours experimental function. Although each of the three domains favoured the supervised outpatient exercise group, none was statistically 'JHVSFMean difference (95% CI) of the effect of significant. There were three objective measures of rehabilitation exercises on knee extensor strength (Nm) in function. The Timed Up and Go test was significantly better in the supervised exercise group, by a mean of 1.8 seconds two studies (n = 58). (95% CI 0.1 to 3.5). The time to ascend four stairs did not differ significantly (mean difference 0.2 sec, 95% CI –0.2 to 0.6). Similarly, there were no significant differences in lower limb power (mean difference 26 Nm/s, 95% CI –26 to 78) or the 6-minute walk test (mean difference 31m, 95% CI –54 to 115). 224 Journal of Physiotherapy 2013 Vol. 59 – © Australian Physiotherapy Association 2013. Open access under CC BY-NC-ND license.
Coulter et al: Physiotherapy rehabilitation after total hip replacement Quality of life: Only the trial by Galea et al (2008) measured functioning (stair climbing, the Timed Up and Go test quality of life. They used the Assessment of Quality of and 6-min walk test), and spatiotemporal measures of Life questionnaire, which ranges from 0 (death) to 1 (full gait. The Timed Up and Go test was originally intended health). The two exercise groups did not differ significantly as a functional measure for elderly people (Podsiadlo and (mean between-group difference 0.05 points in favour of Richardson 1991). A case controlled series by Coulter et supervised exercise, 95% CI –0.15 to 0.25). al (2009) reported progressively faster Timed Up and Go test scores at each time interval in the study comparing Discussion home and supervised physiotherapy, displaying results in comparison with community dwelling older adults (Steffen This study pooled data from five eligible papers to conclude et al 2002). Because of the range of different measures used, that post-discharge physiotherapy does provide better this review could not pool the data for function and quality patient outcomes after total hip replacement, in terms of of life measures and the results of the individual studies strength of hip abductor muscles of the operated leg, gait were not in agreement. Therefore, despite some favourable speed, and cadence. Outpatient supervised rehabilitation evidence, it is not yet possible to establish definitively the provided no better results than unsupervised home exercise effectiveness of post-discharge physiotherapy rehabilitation programs for most outcome measures, with the exception in terms of improving function and quality of life following of the Timed Up and Go test, which was faster in the elective total hip replacement. physiotherapist-supervised group. Although this review identified some significant benefits in The studies included in our review found similar results strength and gait speed due to physiotherapy rehabilitation, to other published studies in this area. A non-randomised, it did not demonstrate a difference in outcomes between controlled trial (Sashika et al 1996) showed that a six-week physiotherapist-prescribed home exercises performed home program including hip range of motion exercises, independently and physiotherapist-supervised programs. isometric exercises, and eccentric strengthening increased The positive results in both settings provide an argument for strength of hip abductors, walking speed, and cadence. further studies into these types of rehabilitation intervention Unlu et al (2007) evaluated a six-week program including after hip replacement. Further studies discriminating the same exercises as Sashika et al (1996), though with two between supervised and unsupervised programs would comparison groups: one home based and one supervised provide guidance for clinical practice and resource decisions by a physiotherapist. Both treatment groups showed regarding how to provide post-discharge physiotherapy. In an improvement in isometric hip abductor torque, gait the meantime, home-based exercise programs or supervised speed, and cadence. Di Monaco et al (2009) performed a physiotherapy can be recommended for this patient group. systematic review of controlled trials of physical exercise programs after total hip replacement, which also supported Future studies need to include a longer follow-up period the usefulness of rehabilitation from late phase (> 8wks to identify whether any improvements are maintained and post-operative). This review included some of the studies whether longer term deficits after hip replacement can be in our review (Jan et al 2004, Trudelle-Jackson and Smith addressed. The studies included in this review collected 2004, and Unlu et al 2007), and concluded that for these outcomes at the end of the intervention and none had a programs to be effective they should comprise weight subsequent follow-up period, except Johnsson et al (1988) bearing exercises with hip abductor eccentric strengthening. with a six-month follow up. There is some evidence that weakness persists several months following hip replacement In our systematic review, functional outcomes were (Jan et al 2004) and consequently a 12 or 24 month follow- measured using a wide range of tools. As a consequence up is recommended. meta-analysis of these data was not possible. The review by Minns Lowe (2009) was also unable to meta-analyse these The search strategy used for this review was comprehensive, data and concluded it was not possible to determine whether but was limited to reviews in the English language. The post-discharge physiotherapy is effective due to insufficient limited number of eligible, high quality studies and the evidence. In the absence of meta-analysis, it is worth small sample sizes of those studies prevent a definitive considering some details of the trials that demonstrated answer for all outcomes in this review. This is particularly good outcomes in a range of diverse measures, such as the the case for functional and quality of life outcomes, where Timed Up and Go test and self-perceived function. Jan et the information was so diverse and sparse that meta-analysis al (2004) showed that a 12-week home exercise program could not be performed. The trials in this review spanned a performed for 60 min daily increased bilateral hip muscle period of 21 years and therefore some of the data were more strength, walking speed, and functional score (Harris Hip difficult to extract from the reports, although where data Score). These improvements were significant in a highly were measured from graphs the two independent reviewers compliant patient group (practice ratio > 50%) and patients showed full agreement for all items for all papers. from a low-compliance group compared to the controls. Trudelle-Jackson and Smith (2004) showed an 8-week In conclusion, this review showed that physiotherapy home exercise program, including weight bearing exercises can improve strength and gait speed after total hip aimed at increasing strength and balance, improved self- replacement. The low number of studies limits the evidence perceived function but the control group having isometric to establish the overall effectiveness of post-discharge and range-of-motion exercises did not improve. Galea physiotherapy for patients who have undergone a primary et al (2008) prescribed an 8-week program, again with a total hip replacement. More research is required to establish home and supervised setting, consisting of seven exercises functional and quality of life outcomes, which may be the that focused on functional tasks, daily living tasks, most important to people recovering from the procedure. balance, strength, and endurance and found significant More research is particularly required to compare the improvements within each group in quality of life, physical efficacy of home exercise programs to supervised exercise Journal of Physiotherapy 2013 Vol. 59 – © Australian Physiotherapy Association 2013. Open access under CC BY-NC-ND license. 225
Research programs, especially in regard to relative resource Universities osteoarthritis index questionnaire and global implications. Further well-designed trials are necessary assessments in patients with osteoarthritis. Journal of and researchers are encouraged to continue clinical studies Rheumatology 27: 2635–2641. to evaluate the full range of effects of physiotherapy in this population. Q Galea M, Levinger P, Lythgo N, Cimoli C, Weller R, Tully E, et al (2008) Targeted home- and center-based exercise program Footnotes: aReview Manager 5.1, The Nordic Cochrane for people after total hip replacement: a randomized clinical Centre, Copenhagen: The Cochrane Collaboration, 2011. trial. Archives of Physical Medicine & Rehabilitation 89: 1442–1447. eAddenda: Appendix 1 and 2 available at jop.physiotherapy. asn.au Harris WH (1969) Traumatic arthritis of the hip after dislocation and acetabular fractures: treatment by mold arthroplasty. Ethics: The ACT Health Human Research Ethics Committee An end-result study using a new method of result evaluation. and Australian National University Human Research Ethics Journal of Bone and Joint Surgery (Am) 51: 737–755. Committee approved this study. All participants gave written informed consent before data collection began. Hawthorne G, Richardson J, Osborne R (1999) The Assessment of Quality of Life (AQoL) instrument: a Competing interests: Nil. psychometric measure of health-related quality of life. Quality of Life Research 8: 209–224. Support: Trauma and Orthopaedic Research Unit, and Physiotherapy Department, Canberra Hospital. Jan M-H, Hung J-Y, Lin J, Wang S-F, Liu T-K, Tang P-F (2004) Effects of a home program on strength, walking speed, and Acknowledgements: Australia National University. function after total hip replacement. Archives of Physical Medicine & Rehabilitation 85: 1943–1951. Correspondence: Corinne Coulter, Physiotherapy Department, Canberra Hospital, Canberra, Australia. Johnsson R, Melander A, Onnerfalt R (1988) Physiotherapy Email: [email protected] after total hip replacement for primary arthrosis. Scandinavian Journal of Rehabilitation Medicine 20: 43–45. References Maher CG, Sherrington C, Herbert RD, Moseley AM, Elkins AIHW (2004) Evidence to support the National Action Plan M (2003) Reliability of the PEDro scale for rating quality of for osteoarthritis, rheumatoid arthritis and osteoporosis. randomized controlled trials. Physical Therapy 83: 713–721. Australian Institute of Health and Welfare 2004. Minns Lowe CJ, Barker KL, Dewey ME, Sackley CM (2009) Australian Orthopaedic Association (2011) National Joint Effectiveness of physiotherapy exercise following hip Replacement Registry. Annual Report. Adelaide: Australian arthroplasty for osteoarthritis: a systematic review of clinical Orthopaedic Association. trials. BMC Musculoskeletal Disorders 10: 98. Coulter C, Weber J, Scarvell J (2009) Group physiotherapy 1PETJBEMP% 3JDIBSETPO4 5IFUJNFEA6Q(PBUFTU provides similar outcomes for participants after joint of basic functional mobility for frail elderly persons. Journal replacement surgery as 1-to-1 physiotherapy: a sequential American Geriatric Society 39: 142–148. cohort study. Archives of Physical Medicine & Rehabilitation 90: 1727–1733. Roos E (2003) Effectiveness and practice variation of rehabilitation after joint replacement. Current Opinion in Dawson J, Fitzpatrick R, Carr A, Murray D (1996) Questionnaire Rheumatology 15: 160–162. on the perceptions of patients about total hip replacement. Journal of Bone and Joint Surgery (Br) 78-B: 185–190. Sashika H, Matsuba Y, Watanabe Y (1996) Home program of physical therapy: effect on disabilities of patients with de Morton NA (2009) The PEDro scale is a valid measure of total hip arthroplasty. Archives of Physical Medicine & the methodological quality of clinical trials: a demographic Rehabilitation 77: 273–277. study. Australian Journal of Physiotherapy 55: 129–133. Steffen T, Hacker T, Mollinger L (2002) Age and gender-related Di Monaco M, Vallero F, Tappero R, Cavanna A (2009) test performance in community-dwelling elderly people: Rehabilitation after total hip arthroplasty: a systematic 4JY.JOVUF8BML5FTU #FSH#BMBODF4DBMF 5JNFE6Q(P review of controlled trials on physical exercise programs. Test, and gait speeds. Physical Therapy 82: 128–137. European Journal of Physical and Rehabilitation Medicine 45: 303–317. Trudelle-Jackson E, Smith S (2004) Effects of a late-phase exercise program after total hip arthroplasty: a randomized Ehrich EW, Davies GM, Watson DJ, Bolognese JA, Seidenberg control trial. Archives of Physical Medicine & Rehabilitation BC, Bellamy N (2000) Minimal perceptible clinical 85: 1056–1062. improvement with the Western Ontario and McMaster Unlu E, Eksioglu E, Aydog E, Tolga Aydoo S, Atay G (2007) The effect of exercise on hip muscle strength, gait speed and cadence in patients with total hip arthroplasty: a randomized controlled study. Clinical Rehabilitation 21: 706–711. Woolf A, Pleger B (2003) Burden of major musculoskeletal conditions. Bulletin of World Health Organization (WHO) 81: 646–656. 226 Journal of Physiotherapy 2013 Vol. 59 – © Australian Physiotherapy Association 2013. Open access under CC BY-NC-ND license.
Appraisal Critically Appraised Papers Positive expiratory pressure prevents more exacerbations than high frequency chest wall oscillation via a vest in people with cystic fibrosis Synopsis expiratory resistor creating a back pressure of 10–20 cmH20, for six cycles of 15 breaths, also separated by 2–3 huffs. Summary of: McIlwaine MP, et al (2013) Long-term The allocated airway clearance regimen was prescribed multicenter randomised controlled study of high frequency twice daily for one year. Outcome measures: The primary chest wall oscillation versus positive expiratory pressure outcome was the number of pulmonary exacerbations, mask in cystic fibrosis. Thorax DOI: thoraxjnl-2012-202915. defined as when prespecified symptoms lasted longer [Prepared by Mark Elkins, Journal Editor.] than 3 days and required antibiotics. Secondary outcomes included time to first pulmonary exacerbation and changes Question: What are the relative effects of high frequency in lung function and quality of life. Results: 88 participants chest wall oscillation (HFCWO) and positive expiratory completed the study. At one year, the median number of pressure (PEP) therapy on pulmonary exacerbations, lung pulmonary exacerbations per participant was 2 (IQR 1 to function, and quality of life in people with cystic fibrosis? 3) in the HFCWO, which was significantly higher than in Design: Randomised trial with concealed allocation and the PEP group at 1 (IQR 0 to 2), p = 0.007. Median time blinded outcome assessment. Setting: Eight paediatric and to first exacerbation was 115 days in the HFCWO group, four adult cystic fibrosis centres in Canada. Participants: which was significantly sooner than in the PEP group at People over 6 years old with clinically stable cystic fibrosis 220 days, p = 0.02. Changes in lung function and quality and forced expiratory volume in 1 sec (FEV1) over 45% of life did not significantly differ between the groups. PEP of the predicted value. Uncommon respiratory organisms was rated as significantly better than HFCWO with respect and recent changes in medications were exclusion criteria. to flexibility in where it could be performed (p < 0.001) and Randomisation allocated 56 participants to HFCWO and the duration of each treatment, which differed by a median 51 to PEP. Interventions: All participants used an airway of 10 min (p < 0.001). Self-reported adherence was over clearance method other than HFCWO or PEP for 2 months 90% in both groups. Conclusion: When prescribed as a prior to starting their intervention. The HFCWO group long-term airway clearance therapy, PEP has significantly then used a pneumatic vest system to apply high frequency better outcomes than HFCWO in terms of exacerbations, oscillations with a triangular wave form to the chest wall. A flexibility, and treatment duration. 30-min ramped protocol was used consisting of six 5-min cycles, with the participant performing 2–3 huffs between 2006, Myers & Horn 2006). This may be due to a selection cycles. The PEP group breathed through a facemask with an effect of participation in an airway clearance study and the monthly telephone calls to encourage good adherence. Commentary This study illustrates the importance of obtaining evidence This study is an excellent example of research designed about the effects of therapies that are prescribed for long- to resolve a widespread clinical question. The marked term use. Recent studies of new airway clearance and difference in pulmonary exacerbations in this trial, exercise interventions in CF continue to consider only single alongside equivocal outcomes for lung function and quality doses (Kuys et al 2011, Reix et al 2012), so clinicians should of life, shows clearly the superiority of PEP over HFCWO be wary of prescribing regular use of new interventions as a regular airway clearance therapy for this population. (especially expensive ones) before their long-term effects PEP’s superiority is reinforced by the other characteristics are known. on which it was rated as better than HFCWO by participants: treatment duration, and flexibility of treatment location. Louella O’Herlihy Physiotherapy Department, Poole Hospital, UK The paper does not provide much detail about the standard care received by both groups, apart from baseline References respiratory medication use. Given that ordering and overlapping nebulised and physical therapies in an airway Abbott J et al (1996) Thorax 51: 1233–1238. clearance session can influence the overall session duration (Bishop et al 2011, Dentice et al 2012, Dentice et al 2013), Bishop JR et al (2011) J Physiother 57: 223–229. more information about how nebulised therapies were incorporated into the overall airway clearance sessions Dentice RL et al (2012) J Physiother 58: 33–40. could have been providded.. Dentice RL et al (2013) Cochrane Database Syst Rev: Another crucial consideration is cost, with a HFCWO CD007923. system being about 100 fold more expensive than a PEP mask system. With better outcomes for far less expense, Kuys SS et al (2011) J Physiother 57: 35–40. physiotherapists should strongly recommend PEP over HFCWO. Modi AC et al (2006) J Cyst Fibrosis 5: 177–185. Adherence to the therapies was very high at 94%. Although Myers LB, Horn SA (2006) J Health Psychol 11: 915–926. self-reported adherence can easily be inflated, this is much higher than in other studies using self-report (Modi et al Reix P et al (2012) J Physiother 58: 241–247. Journal of Physiotherapy 2013 Vol. 59 – © Australian Physiotherapy Association 2013 275
Hespanhol Junior et al: Predictive factors and incidence of running injuries Previous injuries and some training characteristics predict running-related injuries in recreational runners: a prospective cohort study Luiz Carlos Hespanhol Junior1,2, Leonardo Oliveira Pena Costa1,3 and Alexandre Dias Lopes1 1Master’s and Doctoral Programs in Physical Therapy, 2São Paulo Running Injury Group (SPRunIG), Universidade Cidade de São Paulo (UNICID), São Paulo, Brazil, 2Department of Public & Occupational Health, EMGO+ Institute for Health and Care Research, VU University Medical Center, Amsterdam, The Netherlands, 3Musculoskeletal Division, The George Institute for Global Health, Sydney, Australia Questions: What is the incidence of running-related injuries (RRIs) in recreational runners? Which personal and training characteristics predict RRIs in recreational runners? Design: Prospective cohort study. Participants: A total of 200 recreational runners answered a fortnightly online survey containing questions about their running routine, races, and presence of RRI. These runners were followed-up for a period of 12 weeks. Outcome measures: The primary outcome of this study was running-related injury. The incidence of injuries was calculated taking into account the exposure to running and was expressed by RRI/1000 hours. The association between potential predictive factors and RRIs was estimated using generalised estimating equation models. Results: A total of 84 RRIs were registered in 60 (31%) of the 191 recreational runners who completed all follow-up surveys. Of the injured runners 30% (n = 18/60) developed two or more RRIs, with 5/18 (28%) being recurrences. The incidence of RRI was 10 RRI/1000 hours of running exposure. The main type of RRI observed was muscle injuries (30%, n = 25/84). The knee was the most commonly affected anatomical region (19%, n = 16/84). The variables associated with RRI were: previous RRI (OR 1.88, 95% CI 1.01 to 3.51), duration of training although the effect was very small (OR 1.01, 95% CI 1.00 to 1.02), speed training (OR 1.46, 95% CI 1.02 to 2.10), and interval training (OR 0.61, 95% CI 0.43 to 0.88). Conclusions: Physiotherapists should be aware and advise runners that past RRI and speed training are associated with increased risk of further RRI, while interval training is associated with lower risk, although these associations may not be causative. [Hespanhol Junior LC, Costa LOP, Lopes \"% 1SFWJPVTJOKVSJFTBOETPNFUSBJOJOHDIBSBDUFSJTUJDTQSFEJDUSVOOJOHSFMBUFEJOKVSJFTJOSFDSFBUJPOBMSVOOFST a prospective cohort study. Journal of Physiotherapyo> Key words: Athletic injuries, Epidemiology, Etiology, Follow-up studies, Incidence, Risk factors, Running, Sports Introduction (Macera et al 1989). We are unaware of prospective observational studies that controlled important aspects of Running is widely known to be beneficial for general health training (duration of training sessions, speed training, and (Marti 1991, Williams 1997, Williams 2007, Williams interval training) and the level of motivation to run in this 2008). However, one of the consequences of running is population. Information about predictive factors for running running-related injuries (RRI), with incidence rates ranging injuries is essential for sports physiotherapists and other from 18.2% to 92.4% (Satterthwaite et al 1999, van Gent et healthcare professionals for the development of prevention al 2007, Van Middelkoop et al 2008a) or 6.8 to 59 injuries strategies for running injuries. Therefore the objectives of per 1000 hours of running exposure (Bovens et al 1989, Buist et al 2010, Lun et al 2004, Lysholm and Wiklander What is already known on this topic: Running- 1987, Rauh et al 2006, Wen et al 1998). This large variability related injuries are common and frequently cause may be explained by differences in the target populations absence from running. Studies among recreational investigated, such as recreational (Lun et al 2004) or ultra- runners have identified previous injuries, running more marathon runners (Scheer and Murray 2011), and in the UIBOLNXFFL BOEMFTTUIBOZFBSTPGSVOOJOH definitions of RRI used (Jacobs and Berson 1986, Lun et al experience as being associated with increased risk 2004, Pileggi et al 2010, van Gent et al 2007). of running-related injury. Most runners run exclusively for fun and often complete What this study adds: Over a 12-week period, 31% of just a few kilometres per training session. Some of them do recreational runners sustained a running-related injury not participate in running races at all. These recreational severe enough to prevent participation in running runners are probably the most common cohort within for at least one usual training session. Predictors of the running community. Few observational studies have increased injury risk included a previous running- investigated prospectively the incidence and risk factors of related injury, higher duration of training (although the RRI in recreational runners who were not enrolled or not increase in risk was very small), and the use of speed training to participate in races (Lun et al 2004, Macera et al training. The use of interval training was predictive of 1989). The risk factors for RRI that have been identified in reduced injury risk. this population are: previous injuries, running more than 64 km/week, and less than three years of running experience Journal of Physiotherapy 2013 Vol. 59 – © Australian Physiotherapy Association 2013. Open access under CC BY-NC-ND license. 263
Research this study were to determine the incidence of RRI in the Invitation sent to recreational runners lower limbs and spine in a sample of recreational runners, (n = 4000) and to determine which training or personal characteristics may be considered predictive factors for RRI in this First runners to agree to participate population. and completed the initial survey Method 1st follow-up period (n = 200) Study design Loss to follow-up This is an observational injury surveillance study with a t personal reasons (n = 4) prospective cohort design that included 200 recreational t no response (n = 2) runners who responded to an online survey with questions related to their running training routine, races and RRI. 2nd follow-up period The recreational runners were followed-up for a period of (n = 194) 12 weeks, during which the online surveys were answered every two weeks. Loss to follow-up t personal reasons (n = 2) Participants 3rd follow-up period To be included in the study, runners had to be at least 18 (n = 192) years old and to have been running for at least six months. Runners were excluded if they had either any medical Loss to follow-up restriction to running or any musculoskeletal injury that t no response (n = 1) could preclude their participation in running training sessions. 4th follow-up period (n = 191) Recruitment and baseline survey 5th follow-up period A total of 4000 runners who were registered on the database (n = 191) of a running promoter were invited by email to participate in this study. This email provided information about the 6th follow-up period study procedures and contained a link to an electronic (n = 191) consent form. After agreeing to participate, the individuals were directed to a website that contained the baseline Figure 1. Flow of participants through the study. survey. The first 200 runners who agreed to participate in the study, met the inclusion criteria, and fully completed The follow-up survey contained information about training, the baseline survey were included. This survey contained the presence of any RRI during the period, motivation questions regarding personal characteristics, running to run, and any running races that the participant had routines, and previous RRI. Also a specific question was competed in over the preceding two weeks. These questions included to confirm that runners were injury-free before elicited information about the following variables: number starting the follow-ups. All questions and details about the of times that the participant had trained; the total distance baseline survey are described in Appendix 1 (see eAddenda run (in kilometres); average time for each running session; for Appendix 1) and were published elsewhere (Hespanhol predominant type of training surface (asphalt, cement, Junior et al 2012). grass, dirt, sand, gravel); predominant type of terrain (flat course, uphill, downhill, or mixed); amount of speed Follow-up survey and outcome measures Data collection consisted of six follow-up surveys (Appendix 2, see eAddenda for Appendix 2) sent to the runners by email every 14 days throughout the 12-week study period. Messages were sent by email every two weeks to remind the participants to complete the online survey for the previous fortnight. A reminder email was sent if the survey was not completed in three days. If runners had not completed the survey eight days after the initial email, they were then contacted by phone to remind them to complete the survey either online or over the phone. A reminder letter was sent by regular mail with a pre-paid return envelope if none of the previous reminder attempts was successful. Participants who received a reminder by regular mail could complete a printed survey that had the same questions as the online version. In order to minimise the recall bias in the information collected in these follow-up surveys, we sent all runners a running log by regular mail to help them to record each running session. We requested that participants complete the running log with all relevant information and transfer these data while completing the fortnightly follow- up survey. 264 Journal of Physiotherapy 2013 Vol. 59 – © Australian Physiotherapy Association 2013. Open access under CC BY-NC-ND license.
Hespanhol Junior et al: Predictive factors and incidence of running injuries 5BCMF. Personal and training characteristics of participants. Variable All Injured Uninjured p (n = 191) (n = 60) (n = 131) Age (yr), mean (SD) 42.8 (10.5) 41.8 (10.2) 42.9 (10.5) 0.249 Gender, n female (%) 50 (26) 11 (18) 39 (30) 0.095 Height (cm), mean (SD) 0.196 Weight (kg), mean (SD) 171.1 (9.4) 172.4 (8.8) 170.5 (9.7) 0.449 BMI, mean (SD) 72.0 (14.0) 73.1 (11.8) 71.4 (14.9) 0.825 Education, n (%) 24.4 (3.1) 24.5 (2.7) 24.4 (3.3) 0.525 Elementary school 3 (2) 0 (0) 3 (2) 0.066 High school 25 (13) 6 (10) 19 (15) < 0.001 University degree 82 (43) 27 (45) 55 (42) < 0.001 Postgraduate degree 81 (42) 27 (45) 54 (41) < 0.001 Running experience (yr), median (IQR) 5.0 (3.0 to 9.0) 4.0 (2.3 to 6.8) 5.0 (3.0 to 9.5) < 0.001 Frequency (sessions/wk), mean (SD) 3.0 (1.6) 1.9 (1.5) 3.2 (1.5) < 0.001 Distance (km/wk), median (IQR) 28.5 (15.0 to 41.0) 15 (2.5 to 26.3) 30 (18.0 to 42.5) 0.071 Duration (min/session), median (IQR) 60 (50 to 80) 50 (15 to 60) 60 (50 to 90) 0.087 Surface < 0.001 Hard (times/wk)a, median (IQR) 1.5 (0.5 to 3.0) 0.5 (0.0 to 2.0) 1.5 (0.5 to 3.0) 0.001 Soft (times/wk)b, median (IQR) 0.0 (0.0 to 0.5) 0.0 (0.0 to 0.0) 0.0 (0.0 to 1.0) 0.186 Treadmill (times/wk), median (IQR) 0.0 (0.0 to 0.5) 0.0 (0.0 to 1.0) 0.0 (0.0 to 1.5) < 0.001 Other (times/wk)c, median (IQR) 0.0 (0.0 to 0.0) 0.0 (0.0 to 0.0) 0.0 (0.0 to 0.0) 0.213 Terrain <0.001 Flat (times/wk), median (IQR) 2.0 (0.5 to 3.0) 1.0 (0.0 to 2.0) 2.0 (1.0 to 3.0) 0.366 Uphill (times/wk), median (IQR) 0.0 (0.0 to 0.0) 0.0 (0.0 to 0.0) 0.0 (0.0 to 0.5) 0.308 Downhill (times/wk), median (IQR) 0.0 (0.0 to 0.0) 0.0 (0.0 to 0.5) 0.0 (0.0 to 0.0) 0.185 Mixed (times/wk)d, median (IQR) 0.5 (0.0 to 1.0) 0.0 (0.0 to 0.5) 0.5 (0.0 to 1.5) Training 0.171 Speed (times/wk), median (IQR) 0.0 (0.0 to 1.0) 0.0 (0.0 to 0.5) 0.0 (0.0 to 1.0) Interval (times/wk), median (IQR) 0.0 (0.0 to 1.0) 0.0 (0.0 to 0.0) 0.0 (0.0 to 1.0) 0.667 Missed training Personal reasons (n/wk), median (IQR) 0.0 (0.0 to 1.0) 0.0 (0.0 to 1.5) 0.0 (0.0 to 1.0) 0.718 Lack of motivation (n/wk), median (IQR) 0.0 (0.0 to 0.0) 0.0 (0.0 to 0.0) 0.0 (0.0 to 0.0) Unfavourable weather (n/wk), median (IQR) 0.0 (0.0 to 0.0) 0.0 (0.0 to 0.0) 0.0 (0.0 to 0.0) 0.549 How do you feel? n (%) Motivated (majority of 6 follow-ups) 138 (72) 37 (62) 101 (77) Neutral (majority of 6 follow-ups) 21 (11) 9 (15) 12 (9) Poorly motivated (majority of 6 follow-ups) 15 (8) 7 (12) 8 (6) Draw between any category (6 follow-ups) 17 (9) 7 (12) 10 (8) Training monitoring, n (%) Coaches 79 (41) 23 (38) 56 (43) Web spreadsheets 19 (10) 5 (8) 14 (11) No training plan 93 (49) 61 (47) Participated in a race during the study, n (%) 32 (53) Yes (at least 1 race) 174 (91) 120 (92) No (no participation during all the study) 17 (9) 54 (90) 11 (8) Previous running-related injury, n (%) 6 (10) None 90 (47) 64 (49) 1 53 (28) 26 (43) 38 (29) 2 39 (20) 15 (25) 24 (18) 3 15 (25) 9 (5) 5 (4) 4 (7) SD = standard deviation, IQR = interquartile range, BMI = body mass index. aHard surface = asphalt and cement. bSoft surface = dirt, grass and gravel. cOther surface = sand and synthetic. dMixed terrain = uphill and downhill Journal of Physiotherapy 2013 Vol. 59 – © Australian Physiotherapy Association 2013. Open access under CC BY-NC-ND license. 265
Research 5BCMF. The 84 running-related injuries by type and anatomical location. Characteristic of RRI n (%) Duration of RRI Lost training Pain in wks TFTTJPOTXL intensity Type 25 (30) mean (SD) mean (SD) .VTDMFTUSBJOSVQUVSFUFBS 12 (14) mean (SD) 10 (12) 4.3 (2.9) 5.8 (2.0) Low back pain 7 (8) 3.6 (2.7) 1.6 (1.0) 5.2 (2.5) Tendinopathy 6 (7) 2.4 (0.8) 2.8 (2.0) 6.0 (2.0) Plantar fasciitis 4 (5) 4.0 (2.1) 5.7 (5.5) 5.8 (2.5) Meniscal or cartilage damage 3 (4) 4.7 (3.5) 4.0 (5.0) 3.6 (2.4) $POUVTJPOIBFNBUPNBFDDIZNPTJT 2 (2) 3.2 (1.8) 4.6 (2.3) 6.4 (1.8) Intense spasm or severe cramp 2 (2) 2.5 (1.0) 3.8 (4.2) 4.8 (2.2) 4QSBJO JOKVSZPGUIFKPJOUBOEPSMJHBNFOUT 2.0 (0.0) 3.3 (2.3) 3.7 (0.6) Stress fracture (overload) 1 (1) 3.0 (1.4) 3.8 (4.2) 7.8 (1.5) \"SUISJUJTTZOPWJUJTCVSTJUJT 1 (1) 4.0 (0.0) Dislocation, subluxation 1 (1) 2.0a 9.0a Patellar chondromalacia 10 (12) 2.0a 3.0a 3.0a Not identified 2.0a 3.7 (1.4) 8.7 (0.8) Anatomical location 16 (19) 12.0a 4.4 (3.3) 3.9 (2.2) Knee 14 (17) 3.3 (1.8) 'PPUUPFT 12 (14) 4.2 (3.3) 4.9 (2.7) Leg 12 (14) 4.3 (3.0) 4.5 (4.4) 5.6 (2.4) Lumbar spine 12 (14) 3.7 (2.7) 3.9 (2.1) 5.5 (2.2) Thigh 6 (7) 4.0 (3.1) 1.8 (1.0) 5.6 (2.4) Ankle 5 (6) 2.5 (0.9) 3.4 (3.1) 5.9 (1.9) )JQHSPJO 2.5 (1.2) 2.4 (1.8) 5.3 (2.6) Achilles tendon (calcaneal) 3 (4) 2.7 (1.0) 6.8 (4.0) 7.3 (1.4) Cervical spine 2 (2) 4.0 (3.5) 1.7 (1.7) 6.1 (2.1) 1FMWJTTBDSVNCVUUPDLT 2 (2) 4.7 (1.2) 4.3 (4.2) 4.7 (3.1) RRI = running-related injury. aAbsolute numbers 3.0 (1.4) 4.6 (1.1) 5.4 (1.5) 7.0 (4.2) training (ie, training sessions that include some bouts of injury in this study was classified as the first RRI developed high speed running during a very short period); number of by the runners during the 12-week follow-up. interval training sessions as different running intensities (ie, Fartlek); motivation during training (motivated, neutral, Data analysis or poorly motivated); amount and type of running races performed; and absence of training due to personal reasons, Our sample size was estimated using an anticipated RRI motivation, or unfavourable weather conditions (eg, rain). incidence of 26% in the population based upon a previous Participants were also asked whether they failed to train for study (Buist et al 2010), with an estimation accuracy of 25% at least one session due to the presence of any RRI during and a significance level of 5%. This analysis suggested a the period (see Question 12 in Appendix 2 on the eAddenda sample of at least 175 runners. Expecting a loss of follow up for details). In this case, the participant was asked to report of approximately 10–15%, we decided to recruit a sample the symptoms/diagnosis and the anatomical region that was of 200 runners. Descriptive statistics were used to present injured, as well as to rate the pain intensity using a 10-point the characteristics of the participants. Chi-square, Mann- (1–10) pain numerical rating scale. Whitney, and Student’s t-tests were performed to check differences between those who developed RRI during the The primary outcome of this study was the incidence 12-week follow-up and those who did not. The distribution of RRI. The definition of RRI used was ‘any pain of of the data was checked by visual inspection of histograms. musculoskeletal origin attributed to running by the runners themselves and severe enough to prevent the runner from The incidence of RRI was calculated as the percentage of performing at least one training session’ (Bovens et al injured runners and as RRIs per 1000 hours of exposure to 1989, Macera et al 1989, van Middelkoop et al 2007, Van running. The exposure to running was calculated using the Middelkoop et al 2008b). Recurrent RRI during the 12- exposure time from the beginning of the study until the end week follow-up period was defined, based on previous of follow-up (12 weeks). To determine possible associations studies, as an RRI of the same type and at the same site between training characteristics and RRI, we initially as the index injury and which occurred after the runner performed a univariate analysis using the generalised returned to full participation in running sessions after the estimating equations (GEE) for each independent variable index injury (Fuller et al 2006, Fuller et al 2007). The index with RRI as the dependent variable. The variables that had significant associations with p < 0.20 in the univariate 266 Journal of Physiotherapy 2013 Vol. 59 – © Australian Physiotherapy Association 2013. Open access under CC BY-NC-ND license.
Hespanhol Junior et al: Predictive factors and incidence of running injuries 5BCMF. Univariate binary logistic analysis using 5BCMF. Multivariate binary logistic analysis using generalised estimating equations. generalised estimating equation. Variable Odds ratio p Variable Odds ratio p (95% CI) (95% CI) Running 0.99 (0.94 to 1.03) 0.601 %VSBUJPO NJOTFTTJPO a 1.01 (1.00 to 1.02) 0.008 experience (yr) 1.01 (0.87 to 1.18) 0.856 Type of surface (times/ Frequency 1.00 (0.99 to 1.01) 0.920 week) 1.06 (0.86 to 1.31) 0.588 TFTTJPOTXFFL 1.01 (1.00 to 1.02) 0.017 0.25 (0.05 to 1.25) 0.092 Distance (km/week) Hardb 1.14 (0.99 to 1.32) 0.074 Otherc 0.65 (0.38 to 1.13) 0.126 %VSBUJPO NJO 0.89 (0.71 to 1.11) 0.287 Type of terrain 0.12 (0.01 to 1.75) 0.122 session)a 1.03 (0.87 to 1.21) 0.745 (times/week) Type of surface 0.23 (0.04 to 1.25) 0.088 Uphill 1.46 (1.02 to 2.10) 0.039 (times/week) Downhill 0.61 (0.43 to 0.88) 0.008 0.97 (0.81 to 1.17) 0.773 Type of training Hardb 0.53 (0.26 to 1.08) 0.081 (times/week) 1– Softc 0.09 (0.004 to 2.08) 0.133 Speed training 1.88 (1.01 to 3.51) 0.046 Treadmill 1.00 (0.82 to 1.22) 1.000 Interval training Otherd Previous RRI Type of terrain 1.25 (0.93 to 1.67) 0.134 No (times/week) 0.71 (0.48 to 1.03) 0.061 Yes Flat Uphill 1 – CI = confidence interval, RRI = running-related injury. aThe odds Downhill 1.22 (0.64 to 2.32) 0.554 Mixede 0.89 (0.35 to 2.25) 0.808 ratio indicates the change in odds for a 10-units increase. Type of training bAsphalt and cement. cSand and synthetic (times/week) 1 – Speed training 0.79 (0.49 to 1.28) 0.331 Results Interval training How do you feel? 1 – Flow of participants through the study Motivated 2.21 (1.22 to 4.01) 0.009 Neutral Of the 200 runners who were enrolled in the study, 191 Poorly motivated answered all six questionnaires corresponding to the 12 Participated in a weeks of follow-up (96%) as presented in Figure 1. The race during the characteristics of the recreational runners are presented in study Table 1. No Yes Incidence of RRI Previous RRI No During the 12-week follow-up, 84 RRIs were registered Yes by 60 (31%) of the 191 recreational runners analysed. The incidence of RRI in this 12-week follow-up was 10 RRIs CI = confidence interval, RRI = runing-related injury. aThe odds per 1000 hours of running exposure. Of the injured runners, 70% (42/60) developed one RRI, 22% (13/60) developed ratio indicates the change in odds for a 10-units increase. two injuries, 7% (4/60) developed three injuries, and 2% bAsphalt and cement. cDirt, grass and gravel. dSand and (1/60) developed four injuries. Of the runners that presented synthetic. eUphill and downhill two or more RRIs in this study, 28% (5/18) represented recurrences. The mean duration of the RRIs registered in analysis were selected for inclusion in the multivariate this study was 3.4 weeks (SD 2.3), an average of 3.9 running binary logistic analysis to control for confounders using sessions per runner (SD 3.3) were missed due to RRIs, and GEE. The GEE was described as an appropriate method the mean pain intensity of these injuries was 5.6 points (SD to analyse longitudinal data with recurrent events (Twisk et 2.3) on a 10-point scale. The type of RRI and anatomic al 2005). As we collected the RRI information fortnightly, region results are fully described in Table 2. we used predictors from the preceding 14 days to predict RRI occurring in a given fortnight to be sure that the Predictive factors for RRI predictors were related to period before the RRI occurred. The results were expressed as odds ratios (OR) and 95% Table 3 describes the results of the univariate GEE analysis. CI. For continuous variables the ORs indicate the change in The variables with a p < 0.20 in this analysis were included odds for a one-unit increase, except for duration of training, in the multivariate GEE analysis, which is presented in which indicates the change in odds for a 10-unit increase. Table 4. The training characteristics that were identified as Predictive factors were classified as follows: risk factors for risk factors for RRI in the final model were: previous RRI RRI if the 95% CI around the OR was greater than 1.0, or (OR 1.88, 95% CI 1.01 to 3.51), duration of training session protective factors for RRI if the 95% CI around the OR was (OR 1.01, 95% CI 1.00 to 1.02), and speed training (OR 1.46, lower than 1.0. 95% CI 1.02 to 2.10). Interval training was identified as the protective factor against the development of RRIs (OR 0.61, 95% CI 0.43 to 0.88). Journal of Physiotherapy 2013 Vol. 59 – © Australian Physiotherapy Association 2013. Open access under CC BY-NC-ND license. 267
Research Discussion our results add important information about the association between training variables with RRI in recreational runners. The results of this study are relevant because they provide Second, we performed a statistical analysis to determine new information about the incidence of RRIs and modifiable the predictive factors of RRI that take into account the predictive factors for RRI in recreational runners. The recurrent events and the variation of the time-dependent identification of the RRI incidence in recreational runners variables during the study. To our knowledge, no studies is important to monitor interventions that can influence the with the purpose of identifying predictive factors of RRI rate of RRI in this population. In addition, the identification have used this longitudinal statistical technique. of modifiable risk factors is important because this may lead to modifications in the injury risk profile and the There are some limitations to this study. First, the information can be used in the development of preventive recreational runners who participated in this study were interventions. recruited from the same database, which may limit the generalisability of our results. Second, self-report injuries The incidence of RRI found in this study (31%) was lower were used in the study. The logistics of this study did not than those previously reported: 79% at six months follow-up allow for confirmation of diagnosis by a health professional. (Lun et al 2004) and 51% at 12 months follow-up (Macera et Therefore, to facilitate injury reporting participants were al 1989) in recreational runners not enrolled or training to required to select options from drop-down boxes with the participate in races. This may be explained by these previous additional option of entering a response to an empty box if studies using longer follow-up and different RRI definitions. there was no suitable option in the drop-down boxes. Third, While these previous studies considered a reduction of this study had a relatively short follow-up period (ie, 12 the running volume due to injury enough to define a RRI weeks). We suggest conducting further prospective studies (Lun et al 2004, Macera et al 1989), our study used a more with longer follow-up periods and with more accurate rigorous criterion (ie, missing at least one training session diagnosis. due to RRI). Despite this, these results are worrying because the incidence of RRI in recreational runners may increase In conclusion, this prospective cohort study demonstrated from 31% in three months (as we found in this study) to 51% that the incidence of RRI in recreational runners was 31% in one year (Macera et al 1989). These high RRI rates are or 10 RRIs per 1000 hours of running exposure. The most likely to decrease running adherence leading to a negative frequent type of injury was muscle injury and the most influence in an active lifestyle and increasing the costs of affected anatomical region was the knee. The relevant risk health care. factors for RRI in recreational runners were identified in this study as previous RRI and speed training, while the We found that previous RRI was associated with higher risk protective factor identified was interval training. Q of RRI in recreational runners. A systematic review on this topic concluded that this variable had strong evidence to be eAddenda: Appendix 1 and 2 available at jop.physiotherapy. a risk factor of RRI (van Gent et al 2007). Two possible asn.au explanations for these findings are: the ‘new’ injury is an exacerbation of an earlier injury that was not completely Ethics: The Ethics Committee of the Universidade Cidade recovered (Taunton et al 2003, Wen et al 1998); and injured de São Paulo approved this study (number 13506607). All runners may adopt a different biomechanical pattern in participants gave written informed consent before data order to protect the injured anatomical region and this could collection began. predispose them to a new injury. Competing interests: None declared. Duration of training, speed training, and interval training were also associated with higher RRI. Despite statistical Support: None. significance, the OR of duration of training was very small indicating an irrelevant effect in real life. This means that Acknowledgements: Luiz Carlos Hespanhol Junior is in our study and in recreational runners generally, other a PhD student supported by CAPES (Coordenação de training characteristics can be more important to predict Aperfeiçoamento de Pessoal de Nível Superior), process RRI. Speed training was associated with higher RRI. This number 0763-12-8, Ministry of Education of Brazil. can be explained by an increase in the running intensity We thank CORPORE Brasil for their assistance in the overloading the musculoskeletal structures, predisposing recruitment of the study participants, as well as Aline Carla recreational runners to injury. The fact that interval training Araújo Carvalho, Bruno Tirotti Saragiotto and Tiê Parma was associated with lower RRI in this study also supports Yamato for their help in the data collection, and Professor this hypothesis. Most of the recreational runners who Jos Twisk for statistical advice. perform interval training switch from normal or slightly higher intensity intervals to lower or much lower intensity Correspondence: Luiz Carlos Hespanhol Junior, intervals (eg, walking), resulting in a lower total training Department of Public & Occupational Health, VU intensity in a given running session, decreasing the odds University Medical Center, The Netherlands. Email: luca_ of injury. [email protected] We consider that the strengths of this study are two-fold. First, we measured some training variables (duration of training session, speed training, interval training, and the level of motivation to run) that were not measured in previous observational prospective studies with recreational runners not enrolled or training to participate in races. Therefore, 268 Journal of Physiotherapy 2013 Vol. 59 – © Australian Physiotherapy Association 2013. Open access under CC BY-NC-ND license.
Hespanhol Junior et al: Predictive factors and incidence of running injuries References high school cross-country runners. American Journal of Epidemiology 163: 151–159. Bovens AM, Janssen GM, Vermeer HG, Hoeberigs JH, Janssen MP, Verstappen FT (1989) Occurrence of running Satterthwaite P, Norton R, Larmer P, Robinson E (1999) Risk injuries in adults following a supervised training program. factors for injuries and other health problems sustained in International Journal of Sports Medicine 10: S186–S190. a marathon. British Journal of Sports Medicine 33: 22–26. Buist I, Bredeweg SW, Bessem B, van Mechelen W, Lemmink Scheer BV, Murray A (2011) Al Andalus Ultra Trail: an KA, Diercks RL (2010) Incidence and risk factors of running- observation of medical interventions during a 219-km, related injuries during preparation for a 4-mile recreational 5-day ultramarathon stage race. Clinical Journal of Sports running event. British Journal of Sports Medicine 44: 598– Medicine 21: 444–446. 604. Taunton JE, Ryan MB, Clement DB, McKenzie DC, Lloyd- Fuller CW, Ekstrand J, Junge A, Andersen TE, Bahr R, Dvorak Smith DR, Zumbo BD (2003) A prospective study of running J, et al (2006) Consensus statement on injury definitions and injuries: the Vancouver Sun Run ‘In Training’ clinics. British data collection procedures in studies of football (soccer) Journal of Sports Medicine 37: 239–244. injuries. British Journal of Sports Medicine 40: 193–201. Twisk JW, Smidt N, de Vente W (2005) Applied analysis Fuller CW, Molloy MG, Bagate C, Bahr R, Brooks JH, Donson of recurrent events: a practical overview. Journal of H, et al (2007) Consensus statement on injury definitions Epidemioogy and Community Health 59: 706–710. and data collection procedures for studies of injuries in rugby union. British Journal of Sports Medicine 41: 328–331. van Gent RN, Siem D, van Middelkoop M, van Os AG, Bierma- Zeinstra SM, Koes BW (2007) Incidence and determinants Hespanhol Junior LC, Costa LO, Carvalho AC, Lopes AD (2012) of lower extremity running injuries in long distance runners: A description of training characteristics and its association a systematic review. British Journal of Sports Medicine 41: with previous musculoskeletal injuries in recreational 469–480. runners: a cross-sectional study. Revista Brasileira de Fisioterapia 16: 46–53. van Middelkoop M, Kolkman J, van Ochten J, Bierma-Zeinstra SM, Koes BW (2007) Course and predicting factors of lower- Jacobs SJ, Berson BL (1986) Injuries to runners: a study of extremity injuries after running a marathon. Clinical Journal entrants to a 10,000 meter race. American Journal of Sports of Sports Medicine 17: 25–30. Medicine 14: 151–155. van Middelkoop M, Kolkman J, van Ochten J, Bierma-Zeinstra Lun V, Meeuwisse WH, Stergiou P, Stefanyshyn D (2004) SM, Koes BW (2008a) Prevalence and incidence of lower Relation between running injury and static lower limb extremity injuries in male marathon runners. Scandinavian alignment in recreational runners. British Journal of Sports Journal of Medicine and Science in Sports 18: 140–144. Medicine 38: 576–580. van Middelkoop M, Kolkman J, van Ochten J, Bierma-Zeinstra Lysholm J, Wiklander J (1987) Injuries in runners. American SM, Koes BW (2008b) Risk factors for lower extremity Journal of Sports Medicine 15: 168–171. injuries among male marathon runners. Scandinavian Journal of Medicine and Science in Sports 18: 691–697. Macera CA, Pate RR, Powell KE, Jackson KL, Kendrick JS, Craven TE (1989) Predicting lower-extremity injuries among Wen DY, Puffer JC, Schmalzried TP (1998) Injuries in runners: habitual runners. Archives of Internal Medicine 149: 2565– a prospective study of alignment. Clinical Journal of Sports 2568. Medicine 8: 187–194. Marti B (1991) Health effects of recreational running in women. Williams PT (1997) Relationship of distance run per week to Some epidemiological and preventive aspects. Sports coronary heart disease risk factors in 8283 male runners. Medicine 11: 20–51. The National Runners’ Health Study. Archives of Internal Medicine 157: 191–198. Pileggi P, Gualano B, Souza M, Caparbo VF, Pereira RMR, Pinto ALS, et al (2010) Incidência e fatores de risco de lesões Williams PT (2007) Maintaining vigorous activity attenuates osteomioarticulares em corredores: um estudo de coorte 7-yr weight gain in 8340 runners. Medicine & Science in prospectivo. [Incidence and risk factors of musculoskeletal Sports & Exercise 39: 801–809. injuries in runners: a prospective cohort study.] Revista Brasileira de Educacao Fisica e Esporte 24: 453–462. Williams PT (2008) Relationship of running intensity to hypertension, hypercholesterolemia, and diabetes. Rauh MJ, Koepsell TD, Rivara FP, Margherita AJ, Rice SG Medicine & Science in Sports & Exercise 40: 1740–1748. (2006) Epidemiology of musculoskeletal injuries among Journal of Physiotherapy 2013 Vol. 59 – © Australian Physiotherapy Association 2013. Open access under CC BY-NC-ND license. 269
Appraisal Critically Appraised Papers Sun-style T’ai Chi improves walking endurance and health-related quality of life in people with COPD Synopsis achieve a moderate dyspnoea by encouraging participants to imagine pushing against a resistance, perform deeper Summary of: Leung RWM, et al (2013) Short-form Sun- squats or with the use of wrist weights. Participants were style t’ai chi as an exercise training modality in people with also encouraged to complete 30 minutes of SSTC each COPD. Eur Respir J 41: 1051–1057. [Prepared by Kylie Hill, day that they did not attend a supervised session and were CAP Editor.] provided with a training booklet and DVD to facilitate home training. Those in the control group continued with Question: Does short-form Sun-style T’ai Chi (SSTC) usual medical care. Outcome measures: The primary change exercise capacity, balance, physical performance, outcome measure was time walked during the endurance quadriceps strength, health-related quality of life (HRQoL) shuttle walk test (ESWT) at 12 weeks. Results: A total of or mood in patients with chronic obstructive pulmonary 38 participants completed the study. On completion of the disease (COPD)? Design: Randomised controlled trial with training period, greater gains were seen in the intervention concealed allocation and blinding of outcome assessors. group compared with the control group in time walked Setting: The out-patient department of a hospital in Sydney, during the ESWT (348 sec; 95% confidence interval [CI], Australia. Participants: Adults with stable COPD were 186 to 510 sec). Significant between group differences, included if they had not participated in exercise training in in favour of the intervention group, were also seen in the previous 12 months, had no significant co-morbidities measures of balance, physical performance, quadriceps that precluded participation in SSTC and did not require force, HRQoL and anxiety. Conclusion: The use of SSTC supplemental oxygen during exercise. Randomisation is an effective intervention to improve several outcomes in allocated 22 to the intervention group and 20 to the control people with COPD, including walking endurance, HRQoL group. Interventions: Participants in the intervention group and quadriceps force. attended two supervised 1-hour sessions of SSTC training each week, for 12 weeks. Training intensity was titrated to western exercise training. The focus on rhythmic breathing during SSTC may have expedited better diaphragmatic Commentary control during exercise in people with COPD. Chronic obstructive pulmonary disease (COPD) is a A key goal of pulmonary rehabilitation is to encourage the condition characterised by impaired pulmonary function patient to maintain an active lifestyle. If the practice of T’ai and reduced exercise capacity and health-related quality of Chi arouses sufficient interest in people with COPD for them life (HRQoL). There is compelling evidence that pulmonary to adopt this as a daily exercise of moderate intensity, future rehabilitation is effective at improving exercise capacity, work should focus on how SSTC could be incorporated in physical function and HRQoL in people with stable COPD pulmonary rehabilitation. (Lacasse et al 2006) and following exacerbations (Puhan et al 2011). Appropriate exercise reconditioning is essential for Alice Jones successful pulmonary rehabilitation but exercise outcomes Department of Physiotherapy, The University of Sydney, vary depending on the mechanism of exercise limitation (Plankeel et al 2005). The mode of exercise training should Australia be versatile and tailored to an individual’s needs. The study School of Rehabilitation Sciences, Griffith University, by Leung et al (2013) describes a Sun-Style T’ai Chi (SSTC) exercise program which resulted in a clinically important Australia increase in endurance, balance and HRQoL scores. The Incremental Shuttle Walk Test (ISWT) was used to References establish SSTC exercise intensity, which was assessed as moderate. The commendable aspect of the study was adding Lacasse Y et al (2006) Cochrane Database Syst Rev 4: resistance loading to the wrists during T’ai Chi practice to CD003793. ensure standardisation of exercise intensity. Most T’ai Chi studies allow their subjects to practice traditional, relaxed, Plankeel J et al (2005) Chest 127:110–116. smooth and rhythmical movements. The SSTC combined a ‘hard and soft’ form of Chinese martial arts, which ensured Puhan M et al (2011) Cochrane Database Syst Rev 10: a training intensity high enough to yield physiological CD003793. training benefit, which satisfies the essential principle of Troosters T et al (2005) Am J Respir Crit Care Med 172: 19–38. Journal of Physiotherapy 2013 Vol. 59 – © Australian Physiotherapy Association 2013 273
Appraisal Clinimetrics The Work Limitations Questionnaire (WLQ-25) Description can also be obtained (Lerner et al, 2001). Productivity loss can be evaluated using an algorithm provided by the The Work Limitations Questionnaire (WLQ-25) was developers in their manual. developed by Lerner and colleagues (Lerner et al 2001). It is one of the most commonly used questionnaires to evaluate Clinical measurement properties: A systematic review of at-work disability and productivity loss. It contains 25 items the psychometric properties of the WLQ-25 revealed that arranged under four subscales addressing four dimensions the scales have been assessed in various populations and of job demands namely: time demands, physical demands, have demonstrated acceptable levels of validity, reliability mental/interpersonal demands, and output demands. The and responsiveness. (William et al 2007). The WLQ-25 has time demands subscale contains five items on punctuality, demonstrated low to moderate correlation with other pain pacing, and productivity. The physical demands subscale and disability measures (r = 0.28 to 0.67) with the physical has six items covering static positioning, moving around, demand subscale showing the lowest correlation (Roy et al lifting, repetitive movements, posture, and use of tools. 2011). In a study involving 836 patients with rheumatoid The mental or interpersonal demands subscale contains arthritis, Walker and colleagues found the WLQ-25 to nine items that assess concentration and on-the-job social correlate moderately with pain, quality of life and fatigue. interactions. The output demands subscale contains five (r = 0.46 to-0.60) (Walker et al 2005). The internal items determining the volume and quality of work (Lerner consistency of the subscales ranges from 0.77 to 0.97 et al 2001). (Lerner et al 2002, Walker et al 2005; Beaton et al 2010). Test-retest validity ranges from 0.69-0.80 for the four Instructions to client and scoring: For each of the sub scales (Lerner 2001). An exploratory factor analysis questions the clients are asked to rate their level of difficulty revealed one main factor explaining around 77% of the in handling job specific demands in the past two weeks in variance (Walker et al 2005). view of their current health status. Of the four sub scales, except for the physical demands subscale, the clients are The WLQ-25 has been shown to be sensitive to change with asked how much of time they experience difficulty. The a standardized responsive mean of 0.65 for the summed physical demands subscale inquires into how much of the score and 0.63 for the index score. The clinically important time they are able to do items without difficulty. A five difference for the WLQ has been reported as 13/100 points point ordinal response scale ranging from zero-four with for the summed score, 5/28.6 points for the index score an additional sixth option’does not apply to my job’ is used. (Roy et al 2011). The total scores range from 0 – 100 % and an index score References Commentary Beaton DE et al (2010) Arthritis Care and Research 62: 28–37. At-work disability is a vital construct that is of interest for rehabilitation professionals, employers and policy makers. Lerner D et al (2002) Journal of Clinical Epidemiology 55: It is responsible for a major loss in productivity resulting in 197–208. a huge but not obviously seen economic burden to society. WLQ-25 is one of the most commonly used tools for Lerner D et al (2001) Medical Care 39: 72–85. accessing this construct. Further evaluation of the WLQ- 25 using modern clinical measurement methods would Roy JS et al (2011) Physical Therapy 91: 254–266. enhance its validity. Moreover the different versions of the WLQ-25 have to be compared to prescribe a standard Williams RM et al (2007) Journal of Occupational Rehabilitation version for use by clinicians. 17: 504–521. Vanitha Arumugam and Joy C MacDermid Walker N et al (2005) Journal of Rheumatology 32: 1006–1012. Hand and Upper Limb Centre, St. Joseph’s Health Centre University of Western Ontario London, ON, Canada 276 Journal of Physiotherapy 2013 Vol. 59 – © Australian Physiotherapy Association 2013
Appraisal Critically Appraised Papers Wearing pedometers in conjunction with daily step goals and incentives can increase physical activity among children Synopsis 8000 steps per day for at least half of the days in a month received a voucher worth SGD$30. Prizes of SGD$120 Summary of: Finkelstein EA, et al (2013) A cluster were awarded via lotteries held monthly for children who randomized controlled trial of an incentive-based outdoor met their monthly pedometer step goal and attended at least physical activity program. J Pediaitr DOI 10.1016/j. 2 outdoor sessions per month. The control group continued jpeds.2013.01.009. [Prepared by Nora Shields, CAP Editor.] with their usual daily activities. Outcome measures: The primary outcome was the mean number of steps per day Question: Does an incentive-based physical activity in the last week of the 9-month program. The pedometers intervention increase physical activity and fitness in worn by children in the intervention group were unsealed children aged 6–12 years? Design: Cluster randomised, but children in the control group wore sealed pedometers. controlled trial with concealed allocation and blinded Secondary outcome measures were 6-minute walk test outcome assessment. Setting: Hospital and community (6MWT) distance, Pediatric Quality of Life Inventory, settings in Singapore. Participants: Healthy children aged and body mass index. Results: 251 children (mean age 6–12 years. Children with any severe chronic medical 8.2 years, SD 1.5) completed the study. At the end of the condition, such as Type 1 diabetes, were excluded. 9-month program, the mean number of steps per day was Randomisation of 212 families (285 participants) allocated significantly more in the intervention group, by 893 steps 106 families (138 participants) to the incentive-based (95% CI 759 to 1027), with 24% of the intervention group physical activity program group and 106 families (147 and 2% of the control group reaching the target of 8000 participants) to a control group. Interventions: Both groups steps per day. The groups did not differ significantly on received pamphlets presenting information on the benefits any of the secondary outcomes. Conclusion: Pedometers of physical activity. In addition, the intervention group and incentives increased the mean daily number of steps participated in a 9-month incentive-based physical activity performed by children but did not result in improved program. Participants in the intervention group received health outcomes at follow-up. These results are consistent information on structured weekend outdoor activities with behavioural change theory that incentives motivate including 2–3 hour hikes at nature reserves and parks, and sustained behaviour change through feedback and tangible families were encouraged to attend sessions at least twice a reinforcements. month. The children in this group also received a pedometer to track daily steps taken and were offered incentives to [95% CIs calculated by the CAP Editor.] meet a goal of 8000 steps per day. Each child who logged accurate measure of fitness as the heart rates obtained at Commentary the end of the test would suggest a less than maximal effort (O’Donovan et al 2013). Due to concerns of decreasing levels of physical activity in children, there is a need for interventions aimed at increasing Improvement in physical activity as measured by steps taken physical activity. Higher levels of physical activity in was evident in this study indicating it is possible to increase children may decrease risk factors for cardiovascular physical activity in young children. It adds to the growing disease and cancer in adulthood and prevent overweight evidence for including incentives as part of interventions and obesity. to change physical activity behaviour. Pedometers are low cost and widely available. When faced with the challenge A strong aspect of the methods implemented by Finkelstein of increasing activity levels in children, therapists should and colleagues was the use of pedometers. Pedometers consider using pedometers and other incentives to motivate capture the number of steps taken. Determining if an behaviour change along with family involvement. Initial exercise intervention leads to physical activity behaviour costs in terms of pedometers, vouchers, and prizes may seem change can be difficult. Between group differences can be high but are low in terms of preventing the considerable due to the control group increasing their overall activity or healthcare costs due to chronic disease in later life. the intervention group decreasing their background physical activity levels due to engaging in the prescribed exercise Juliette Hussey intervention (known as the ‘activitystat’ hypothesis) Discipline of Physiotherapy, Trinity College Dublin, (Rowlands 1998). Measuring overall physical activity continuously throughout the duration of the study helps Ireland account for these unexpected changes. References No changes in secondary outcomes (BMI, 6MWT, and quality of life) were found. However the levels of BMI and O’Donovan C et al (2013) Pediatr ObesEPJK quality of life appeared to be within normal ranges. It is 6310.2013.00172.x. possible that the relatively moderate target of 8000 steps may not have been sufficient to result in secondary changes Rowlands T (1998) Med Sci Sports Exerc 30: 392. such as BMI. Also, the 6MWT may not have been an 272 Journal of Physiotherapy 2013 Vol. 59 – © Australian Physiotherapy Association 2013
Search
