["Research 179 Percentage of participants 0% 20% 40% 60% 80% 100% Benefits Students bring enthusiasm and readiness to learn Clinical educators develop additional knowledge\/skills in supervision\/teaching Private practitioners\/clinical educators can gauge student suitability for employment Graduates are orientated to private practice and may be more work ready Barriers Limited caseload available for students Clients do not agree to student consultations Third-party funders will not reimburse for student consultations Financial costs associated with supervision Extra time and effort associated with supervision Inadequate financial contribution from university Insufficient clinic space or equipment Inadequate clinical educator training and support Enablers Clinical educators do not lose income by supervising students Clinical educators gain income from supervising students Clinical educators receive training and support from the university University contributes financially towards costs of supervision University recognises clinical educator and private practice contributions Strongly disagree Disagree Neutral Agree Strongly agree Figure 1. Percentage of participants reporting each category of agreement with the survey statement. caseload and was identi\ufb01ed as a barrier in this study and other As suggested by university clinical education managers,4 a common research.8\u201310 In addition, this study highlights the speci\ufb01c barrier barrier to hosting placements reported by participants in this study related to clients whose treatment is funded by a third party (eg, was the cost and time associated with supervising and supporting students; these costs related to the earning potential of clinical edu- Medicare, motor vehicle accident, workers compensation). Current cators (ie, the number of clients able to be seen, and\/or the ability to charge for consultations). These perspectives appear to vary: a recent legislative and funding restrictions were perceived to prevent or limit retrospective study found no signi\ufb01cant differences in occasions of service and business income in 12 private practices when they were student involvement in care if the client was to be billed for services hosting students compared with a control period when they were not hosting students.30 Moreover, when excluding the \ufb01rst week of provided. Participants in this study perceived care to carry minimal placement and comparing the later weeks of hosting a student with not having a student, there was a slightly positive effect on practice risk with student input due to the clinical educators\u2019 close supervi- income when hosting students.30 Future research should therefore sion. Research already suggests that outcomes of care are similar.23 explore whether practitioner perceptions match with educator and practice incomes and explore how and why costs of hosting students Veri\ufb01cation of the safety and effectiveness of student care in clin- may differ within different placement contexts and models. ical trials may assist efforts to advocate for a change in legislation and Participants in this research wanted \ufb01nancial reimbursement, funding limitations. In the meantime, educators may be able to in- training and support from universities to host students on placement. Interestingly, a third of participants reported that no one at their crease student exposure to different caseloads by sharing supervision practice had attended any clinical educator training and , 50% received \ufb01nancial remuneration from universities. Wanting more with other practitioners, using peer-assisted learning and organising support from universities, speci\ufb01cally frequent on-site visits, is active observation opportunities.24,25 Students may also bene\ufb01t from consistent with previous research,9 but the capacity to provide undertaking relevant business administration and client service ac- tivities (eg, clinical documentation and reports, reception duties) and re\ufb02ective practice tasks (ie, self-evaluation, journaling, research), as these non-clinical activities are important to prepare them for private practice employment.5 Other novel approaches to clinical placement models that universities and private practitioners might like to explore include telehealth,26 simulation,27 peer simulation28 and clinical research placements.29","180 Peiris et al: Hosting students in private practice physiotherapy on-site support may vary by institution. Where it is not possible for a Correspondence: Casey L Peiris, Discipline of Physiotherapy, School university to provide this type of support because of staf\ufb01ng, funding or geographical limitations, remote video support might be a suitable of Allied Health, Human Services and Sport, La Trobe University, alternative, as well as better preparing students (eg, through early observation experience) and educators (eg, clinical educator training, Melbourne, Australia. Email: [email protected] having a clear model to follow) and providing a range of independent learning activities for students to complete on placement to give References private practitioners some breathing space. These strategies, com- bined with close communication with the universities, as suggested 1. Australian Government. Department of Health and Aged Care. Factsheet, Allied in previous research,4 would also enhance bene\ufb01ts and help manage Health, 2019: Physiotherapists. https:\/\/hwd.health.gov.au\/resources\/publications\/ poorly performing students. factsheet-alld-physiotherapists-2019.pdf. published February 2020; accessed 27 July 2022. This study focused on Australian private practice physiotherapists who had recently hosted students; therefore, the results may not be 2. Lawton V, Jones TM, Dean CM. Students achieve comparable performance scores representative of all physiotherapists in all locations and times. for clinical placements in public and private sectors: a longitudinal observa- However, participants were from six states and territories in tional study. J Physiother. 2021;67:56\u201361. https:\/\/doi.org\/10.1016\/j.jphys.2020. Australia, representing a broad national perspective. No response rate 12.001 to the surveys was reported due to the use of snowballing recruit- ment. However, considering that 643 students completed a private 3. Health Workforce Australia. Clinical training pro\ufb01le: Physiotherapy. Adelaide: practice placement in Australia in 2017,4 the sample (who hosted 208 Health Workforce Australia. https:\/\/www.google.com\/url?sa=t&rct=j&q=& students per year) can be assumed to host one-third of all placements esrc=s&source=web&cd=&ved=2ahUKEwiR0eHFueXsAhX8wjgGHTQSD5kQFjAAeg in Australia and are therefore likely representative of clinical educa- QIAxAC&url=https%3A%2F%2Facdhs.edu.au%2Fwp-content%2Fuploads%2F2018% tors across Australia. 2F05%2FClinical-Training-Pro\ufb01les-Project-Physiotherapy.pdf&usg=AOvVaw1vEZS fADvXdUuocDhRWrXY. published March 2014; accessed 03 November 2020. Private practitioners perceived that hosting students had wide- ranging bene\ufb01ts; however, they did report them to be costly and 4. Peiris CL, Reubenson A, Dunwoodie R, Lawton V, Francis-Cracknell A, Wells C. time-consuming. Universities can further support clinical educators Clinical placements in private practice for physiotherapy students are perceived as by providing tailored training, speci\ufb01c guidelines and communication safe and bene\ufb01cial for students, private practices and universities: a national related to private practice placements, and by providing consistent mixed-methods study. J Physiother. 2022;68:61\u201368. https:\/\/doi.org\/10.1016\/j.jphys. levels of remuneration for hosting students. Universities and private 2021.12.007 practices can work together, along with professional bodies, govern- ment and third-party funders, to address the barriers to private 5. Wells C, Olson R, Bialocerkowski A, Carroll S, Chipchase L, Reubenson A, practice placements for students. This is necessary to ensure that et al. Work readiness of new graduate physical therapists for private private practice placement experiences continue to be available so practice in Australia: Academic faculty, employer, and graduate perspectives. that physiotherapy students can gain contextual skills suited to pri- Phys Ther. 2021;101:pzab078. https:\/\/doi.org\/10.1093\/ptj\/pzab078. PMID: vate practice and there is a smoother transition to the workforce. 33686439. What was already known on this topic: About half of all 6. Atkinson R, McElroy T. Preparedness for physical therapy in private practice: physiotherapists in Australia work in private practice but , 10% Novices identify key factors in an interpretive description study. Man Ther. of student placements occur in private practice. It is important 2016;22:116\u2013121. for students to gain exposure to private practice prior to gradu- ation to improve their work readiness for this setting. 7. Davies JM, Edgar S, Debenham JA. A qualitative exploration of the factors in\ufb02u- What this study adds: Private practitioners perceived that encing the job satisfaction and career development of physiotherapists in private hosting students was beneficial for practitioners, clients and practice. Man Ther. 2016;25:56, 51. future recruitment but reported that they were costly and time- consuming. Universities can reduce barriers by providing 8. Kent FM, Richards KL, Haines TP, Morgan PE, Maloney SR, Keating JL. Patient and tailored training and support, and need to work with practi- practitioner perceptions of student participation in private practice consultations: tioners, professional bodies, government and third-party funders A mixed-methods study. Focus Health Prof Educ. 2015;16:42\u201354. to address policy barriers. 9. Doubt L, Paterson M, O\u2019Riordan A. Clinical education in private practice: An Footnotes: a SurveyMonkey software, SurveyMonkey, San Mateo, interdisciplinary project. J Allied Health. 2004;33:47\u201350. USA. 10. Maloney P Stagnitti K, Schoo A. Barriers and enablers to clinical \ufb01eldwork edu- b SPSS Statistics for Windows Version 26, IBM Corp, Armonk, USA. cation in rural public and private allied health practice. Higher Edu Res Dev. c NVivo 12 software, QSR International Pty Ltd, Denver, USA. 2013;32:420\u2013436. eAddenda: Appendix 1 can be found online at https:\/\/doi.org\/10. 1016\/j.jphys.2023.05.009 11. Hall M, Poth C, Manns P, Beaupre L. An exploration of Canadian physiotherapists\u2019 Ethics approval: The study was approved by ethics committees at decisions about whether to supervise physiotherapy students: Results form a na- The University of Queensland, La Trobe University, Curtin University, tional survey. Physiother Can. 2016;68:141\u2013148. https:\/\/doi.org\/10.3138\/ptc.2014- Macquarie University, Monash University and Charles Sturt Univer- 88E sity. All participants gave written informed consent before data collection began. 12. Forbes R, Dinsdale A, Dunwoodie R, Birch S, Brauer S. Weighing up the bene\ufb01ts and Competing interests: The authors have no con\ufb02icts of interest to challenges of hosting physiotherapy student placements in private practice; a declare. qualitative exploration. Physiother Theory Pract. 2020;27:1\u201311. Source(s) of support: Nil. Acknowledgements: We acknowledge the La Trobe University 13. Forbes R, Dinsdale A, Dunwoodie R, Birch S, Brauer S. Exploring strategies used by Social Research Assistance Platform for providing research support physiotherapy private practices in hosting student clinical placements. Aust J Clin for interview transcription. We would also like to acknowledge Educ. 2020;7:1\u201313. https:\/\/doi.org\/10.53300\/001c.17206 Cambridge McCormick and Alexis Nicholson for their assistance in organising focus groups and transcribing interviews. 14. Smith PM, Corso LM, Cobb N. The perennial struggle to \ufb01nd clinical place- ment opportunities: A Canadian national survey. Nurs Educ Today. 2010;30:798\u2013803. 15. Rodger S, Webb G, Devitt L, Gilbert J, Wrightson P, McMeeken J. A clinical education and practice placements in the allied health professions: an international perspective. J Allied Health. 2008;37:53\u201362. 16. McMeeken J, Grant R, Webb G, Krause KL, Garnett R. Australian physiotherapy student intake is increasing and attrition remains lower than the university average: a demographic study. Aust J Physiother. 2008;54:65\u201371. 17. Shields N, Bruder A, Taylor NF, Angelo T. Getting \ufb01t for practice: An innovative paediatric clinical placement provided physiotherapy students opportunities for skill development. Physiotherapy. 2013;99:159\u2013164. https:\/\/doi.org\/10.1016\/j. physio.2012.02.001 18. Dean CM, Stark AM, Gates CA, Czerniec SA, Hobbs CL, Bullock LD, et al. A pro\ufb01le of physiotherapy clinical education. Aust Health Rev. 2009;33:38\u201346. https:\/\/doi.org\/ 10.1071\/ah090038 19. Portney L, Watkins MP. Foundations of clinical research applications to practice. 3rd ed. Upper Saddle River, NJ: Pearson Education; 2009. 20. Saunders B, Sim J, Kingstone T, Baker S, Water\ufb01eld J, Bartlam B, et al. Saturation in qualitative research: exploring its conceptualization and operationalization. Qual Quant. 2018;52:1893\u20131907. 21. Liamputtong P. Qualitative research methods. 3rd ed. Hong Kong: Oxford Univer- sity Press; 2009. 22. Thompson Burdine J, Thorne S, Sandhu G. Interpretive description: A \ufb02exible qualitative methodology for medical education research. Med Educ. 2021;55(3):336\u2013343. 23. Chia RX, Gomersall SR, Fooken J, Birch S, Dinsdale A, Dunwoodie R, et al. Physio- therapy student clinical placements in Australian private practice: Patient-reported outcomes with supervised student care. Physiother Res Int. 2022;27:e1929. https:\/\/ doi.org\/10.1002\/pri.1929 24. Sevenhuysen S, Nickson W, Farlie MK, Raitman L, Keating JL. The development of a peer assisted learning model for the clinical education of physiotherapy students. J Peer Learning. 2013;6:30\u201345.","Research 181 25. Sevenhuysen S, Farlie MK, Keating JL, Haines TP, Molloy E. Physiotherapy students 28. Dalwood N, Bowles K-A, Williams C, Morgan P, Pritchard S, Blackstock F. Students and clinical educators perceive several ways in which incorporating peer-assisted as patients: A systematic review of peer simulation in health care professional learning could improve clinical placements: a qualitative study. J Physiother. education. Med Educ. 2020;54:387\u2013399. 2015;61:87\u201392. 29. Dario A, Simic M. Innovative physiotherapy clinical education in response to the 26. Bacon R, Hopkins S, Kellett J, Millar C, Smillie L, Sutherland R. The bene\ufb01ts, chal- COVID-19 pandemic with a clinical research placement model. J Physiother. lenges and impacts of telehealth student clinical placements for accredited health 2021;67:235\u2013237. programs during the COVID-19 pandemic. Front Med. 2022;30:842685. 30. Forbes R, Dinsdale A, Brauer SG, Dunwoodie R, Fooken J, Khanna D, et al. Hosting 27. Chipchase L, Blackstock F, Patman S, Barnett-Harris A. Keep the momentum going: pre-registration physiotherapy students in Australian private practices does not pushing the boundaries of clinical learning and assessment. J Physiother. change service and economic outcomes; an economic analysis. Musc Sci Pract. 2018;64:205\u2013207. 2021;52:102318. https:\/\/doi.org\/10.1016\/j.msksp.2021.102318","Journal of Physiotherapy 69 (2023) 182\u2013188 j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m \/ l o c a t e \/ j p hy s Research Physiotherapists should apply health coaching techniques and incorporate accountability to foster adherence to a walking program for low back pain: a qualitative study Natasha C Pocovi a, Julie Ayre b, Simon D French c, Chung-Wei Christine Lin d, Anne Tiedemann d, Christopher G Maher d, Dafna Merom e, Kirsten McCaffrey f, Mark J Hancock a a Department of Health Professions, Faculty of Medicine, Health and Human Sciences, Macquarie University, Sydney, Australia; b Sydney Health Literacy Lab, School of Public Health, Faculty of Medicine and Health, The University of Sydney, Sydney, Australia; c Department of Chiropractic, Faculty of Medicine, Health and Human Sciences, Macquarie University, Sydney, Australia; d Institute for Musculoskeletal Health, The University of Sydney and Sydney Local Health District, Sydney, Australia; e School of Science and Health, Western Sydney University, Sydney, Australia; f Sydney Health Literacy Lab, School of Public Health, Faculty of Medicine and Health, The University of Sydney, Sydney, Australia KEY WORDS ABSTRACT Exercise Questions: What motivates individuals to start a walking program for the prevention of low back pain? What Adherence strategies optimise short-term and long-term adherence to a walking program? What strategies can Low back pain physiotherapists incorporate into clinical practice to facilitate commencement of and adherence to a Patient preference walking program? Design: Qualitative study. Participants: Twenty-two adults recently recovered from an Prevention episode of non-speci\ufb01c low back pain who participated in a 6-month, progressive and individualised walking Qualitative research program that was prescribed by a physiotherapist trained in health coaching. Methods: Semi-structured focus groups conducted online following completion of the walking program. Interview questions explored: primary motivations for starting a walking program, identi\ufb01cation of which elements were useful in opti- mising adherence to the program, and identi\ufb01cation of the barriers to and facilitators of engagement with the program. Audio recordings were transcribed and thematic analysis was conducted. Results: Three major themes were identi\ufb01ed. Theme one identi\ufb01ed that strong motivators to start a walking program were anticipated improvements in low back pain management and the added general health bene\ufb01ts of a more active lifestyle. Theme two identi\ufb01ed that fear of high-impact exercises led to avoidance; however, walking was considered a safe exercise option. Theme three identi\ufb01ed accountability, enjoyment of exercise and health bene\ufb01ts were critical to adherence. Conclusion: Participants recently recovered from low back pain re\ufb02ected positively on a physiotherapist-prescribed walking program. Participants described what elements of the program were crucial to starting exercise and optimising adherence. These \ufb01ndings have informed a list of practical recommendations for physiotherapists to improve patient commencement and adherence to exercise. [Pocovi NC, Ayre J, French SD, Lin C-WC, Tiedemann A, Maher CG, Merom D, McCaffrey K, Hancock MJ (2023) Physiotherapists should apply health coaching techniques and incorporate accountability to foster adherence to a walking program for low back pain: a qualitative study. Journal of Physiotherapy 69:182\u2013188] \u00a9 2023 Australian Physiotherapy Association. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http:\/\/creativecommons.org\/licenses\/by-nc-nd\/4.0\/). Introduction enhanced cognitive function,6,7 promotion of mental wellbeing8,9 and improved physical function.10\u201312 Despite these wide-ranging Low back pain is a musculoskeletal condition that is commonly bene\ufb01ts, global estimates indicate that more than one-quarter (28%) treated by physiotherapists and is a leading cause of disability of adults do not meet the World Health Organization\u2019s Guidelines on worldwide.1 The burden of low back pain is largely the result of the Physical Activity, presenting a major public health problem requiring recurrent nature of the condition, with approximately 70% of in- urgent action.13,14 Physiotherapists are well placed to address this dividuals experiencing a recurrence within 12 months following re- public health concern by encouraging their patients to have a more covery.2 Current evidence supports the use of exercise for the physically active lifestyle. prevention of low back pain recurrences, but no speci\ufb01c exercise is shown to provide superior bene\ufb01t.3,4 Identifying an exercise strategy that patients are compliant with ideally provides general health bene\ufb01ts and prevents future re- Exercise is associated with signi\ufb01cant health bene\ufb01ts, including currences of low back pain, which would be a signi\ufb01cant advance- the prevention and management of non-communicable disease,5 ment in public health. Non-adherence to prescribed home-based https:\/\/doi.org\/10.1016\/j.jphys.2023.05.010 1836-9553\/\u00a9 2023 Australian Physiotherapy Association. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http:\/\/creativecommons.org\/ licenses\/by-nc-nd\/4.0\/).","Research 183 Table 1 Intervention description using the Template for Intervention Description and Replication (TIDieR) checklist. 1. Brief name The WalkBack Trial 2. Why Low back pain (LBP) is recognised globally as a prevalent, costly and disabling condition. Recurrences are common and contribute to much of the burden of LBP. Current evidence favours exercise and education for prevention of LBP, but an optimal intervention has not yet been established. Walking is a simple, widely accessible, low-cost intervention that has yet to be evaluated. 3. What materials Participants allocated to the walking\/education intervention received: \u000f sessions with a physiotherapist with the primary aim of designing a progressive and individually tailored walking program. \u000f education focused on a modern understanding of LBP that reduces the threat and fear associated with pain and advice on strategies to reduce the risk of a recurrence of LBP. \u000f a wearable physical activity tracker to measure daily steps. \u000f a walking diary to act as a motivator in completing the program and provide a degree of accountability. 4. What procedures \u000f The \ufb01rst of three face-to-face or video conference calls with the clinician was used to collaboratively design a walking program and provide education and advice related to LBP and the rationale for undertaking the program. \u000f Three telephone calls used health coaching principles to identify barriers to and facilitators of engagement in the walking program, and to provide support to assist participants achieve the walking goals or modify the program as required. \u000f All sessions focused on progression of the walking program by increasing frequency, duration and intensity throughout the program. 5. Who provided Clinicians with a tertiary quali\ufb01cation in physiotherapy who have received training through Wellness Coaching Australia on the topic of behaviour change and coaching delivered the intervention. 6. How The tailored walking program was re-assessed and progressed during follow-up consults with the clinician. 7. Where The intervention was delivered at approximately 25 private physiotherapy clinics across greater Sydney and south-east Queensland, Australia. 8. When and how much \u000f Following randomisation, those in the walking\/education intervention received six sessions with a physiotherapist. \u000f Participants were booked in for an initial consult (week 0) with a physiotherapist lasting approximately 45 minutes. \u000f The telephone-based coaching occurred at three time points (week 2, week 8 and week 26), taking approximately 15 minutes based on each participant\u2019s requirement. \u000f Two follow-up sessions (face-to-face or video conferencing depending on proximity to available clinics) took place in week 4 and week 12 and lasted approximately 30 minutes. 9. Tailoring The walking program was tailored to participant goals, preferences and current walking capacity. exercise is reported to be between 50 and 70% in patients with low Methods back pain.15 A previous pilot trial for low back pain prevention found that centre-based exercise programs, which require a large time and Design travel commitment, were unacceptable to consumers.16 Similarly, a discrete choice experiment identi\ufb01ed a preference for exercise pro- Online focus groups explored participants\u2019 perspectives and ex- grams of shorter duration, lower cost and that take place outside of a periences with the intervention arm of the WalkBack trial. Reporting gym setting.17 Addressing these barriers, walking is a simple, popular of the current study conformed with the consolidated criteria for and low-cost exercise that can be performed at a convenient time and reporting qualitative research (COREQ) guideline.20 Ethical approval place for individuals. A recent systematic review investigating the was obtained from the Macquarie University Human Research Ethics effectiveness of walking\/running for the treatment and prevention of Committee (HREC Approval Number: 52021963424976). non-speci\ufb01c low back pain found high-certainty evidence that walking was more effective in reducing short-term pain (SMD \u20130.23, Overview of the WalkBack Trial 95% CI \u20130.35 to \u20130.10) and disability (SMD \u20130.19, 95% CI \u20130.33 to \u20130.06) compared with minimal or no intervention.18 There were no A detailed overview of the WalkBack trial can be found in the trials investigating the effectiveness of walking\/running for low back published protocol.19 In this two-armed randomised trial, a walking pain prevention; this evidence gap is being addressed by the current and education (intervention) aiming to prevent low back pain WalkBack trial. recurrence were compared with a control group. Included partici- pants were adults who had recovered from a recent episode of low The randomised controlled WalkBack trial19 developed an indi- back pain and at trial commencement: were not engaged in regular vidualised and progressive walking program speci\ufb01cally for the pre- walking for exercise, were not meeting physical activity guidelines, vention of low back pain recurrences. The intervention incorporated and did not have comorbidities preventing safe participation in a elements intentionally designed to optimise adherence, including walking program. health coaching, support from trained physiotherapists and activity tracking. This qualitative study explored the perspectives and expe- Participants allocated to the intervention received six sessions riences of participants in the intervention arm of the trial. Speci\ufb01cally, (three face-to-face or telehealth sessions based on geographical the current study aimed to explore participants\u2019 primary motivation proximity to trial-trained clinicians, and three telephone calls) with a for starting a walking program, identify which elements were useful physiotherapist to facilitate a progressive and individualised walking in optimising adherence to the program and identify the barriers or program over a 6-month period. The aim was to reach a minimum facilitators to engagement with the program. It was anticipated that dosage of walking \ufb01ve times per week, for at least 30 minutes by the this qualitative study may highlight strategies that could be used by end of the program, and for participants to then continue indepen- clinicians to increase the physical activity levels of their patients and dently. The physiotherapist underwent training in health coaching improve adherence to prescribed exercise programs. that comprised motivational interviewing, goal setting and building self-ef\ufb01cacy. This training aimed to optimise participant adherence Therefore, the research questions for this qualitative study were: with the intervention. A summary of the intervention following the Template for Intervention Description and Replication (TIDieR) 1. What motivates individuals to start a walking program for the checklist21 is presented in Table 1. prevention of low back pain? Participants allocated to the control group did not receive any 2. What strategies optimise short-term and long-term adherence to a treatment as part of their involvement in the trial. However, they walking program? could engage in strategies to prevent or treat episodes of low back pain as required. 3. What strategies can physiotherapists incorporate into clinical practice to facilitate commencement of and adherence to a walking program?","184 Pocovi et al: Optimising adherence to a walking program Table 2 Data analysis Characteristics of participants. The analysis team comprised a mix of musculoskeletal researchers Characteristic Participants (n = 22) (with physiotherapy, chiropractic and exercise science backgrounds), behavioural scientists with qualitative research expertise, and two Gender, n (%) 13 (59) consumers with experience of back pain and involvement in the female 9 (41) WalkBack trial. These consumers did not contribute data to the focus male 56 (10) groups, but instead advised on interpretation and analysis of collected data. The moderator of the focus groups and primary ana- Age (yr), mean (SD) 3 (14) lyst (NCP) has background training in physiotherapy but was not an Relative status of economic advantage a, n (%) 2 (9) intervention provider in the WalkBack trial. 4 (18) quintile 1 4 (18) Framework analysis24 was used to analyse data, which involved quintile 2 9 (41) the following steps: familiarisation with the transcripts; assigning quintile 3 codes to sections of transcripts; sorting codes under similar themes\/ quintile 4 5 (23) sub-themes; charting quotes under each theme; and summary and quintile 5 4 (18) interpretation of the data. Previous number of low back pain episodes b, n (%) 3 (14) 1 to 5 episodes 8 (36) Two transcripts were independently coded by JA and NCP to 6 to 10 episodes 2 (9) observe similarities and differences in coding. Any discrepancies were 11 to 25 episodes discussed, and these principles guided coding of the remaining three 26 to 100 episodes 11 (50) transcripts. Preliminary themes with supporting quotes were then . 100 episodes 11 (50) presented to the consumers and the research team. Through discus- Intervention delivery, n (%) sions, the preliminary themes were critically reviewed, re\ufb02ected on face-to-face 7.95 (1.71) and revised. The revised themes were continuously compared with telehealth 7.14 (2.14) the transcripts to ensure that updated themes were supported by the Self-rated adherence (0 to 10) c, mean (SD) data. at 3 months at 6 months Results a Economic advantage is based on participants\u2019 postcodes and evaluated using the Flow of participants through the study Socio-Economic Indexes for Areas (SEIFA) interactive maps produced by the During recruitment, 47 participants allocated to the intervention Australian Bureau of Statistics. Quintile 1 = 20% least advantaged population. arm of the WalkBack trial were candidates for the current study (ie, had completed their 6-month follow-up questionnaire and were Quintile 5 = 20% most advantaged population. invited to participate). Of these, 25 agreed to participate, of which 22 b An episode of low back pain was de\ufb01ned as pain in the area between the 12th rib provided data for the focus groups (three participants were unable to attend the scheduled times). Five focus groups with a mean duration and buttock crease not attributed to a speci\ufb01c diagnosis such as vertebral fracture or of 56 minutes were conducted. cancer, lasting . 24 hours with a pain intensity . 2 (0 to 10 Numeric Pain Rating Scale). Participant characteristics c 0 is \u2018not at all compliant\u2019 and 10 is \u2018very compliant\u2019. The percentage of male and female participants closely re\ufb02ected the numbers enrolled in the main trial, with the majority being fe- Participants males (59%) and a mean age of 56 years. Most participants lived in metropolitan areas across Australia, and more than half of the par- Consecutive participants in the intervention arm of the WalkBack ticipants (59%) lived in areas indexed to be in the two highest quin- trial were invited (between May and August 2021) to participate in tiles of relative socioeconomic advantage, using the Socio-Economic the current study, shortly after completing the 6-month intervention. Indexes for Areas (SEIFA) 2016 interactive classi\ufb01cation maps.25 The The questionnaire at the 6-month assessment concluded with a participants who agreed to participate in the focus groups, by chance, description of the qualitative study, followed by a question asking had characteristics that re\ufb02ected a wide range of age, sex, socioeco- whether participants were interested in \ufb01nding out more. Those nomic status, history of low back pain and self-reported adherence indicating \u2018Yes\u2019 were sent an email with the Participant Information (Table 2). Therefore, although all consenting participants took part, and Consent Form and then contacted via telephone to discuss po- which is not typical of purposive sampling, the goal of purposive tential involvement. sampling was achieved. Sampling was planned to be purposive, to ensure a balanced representation based on participants\u2019 age, sex, number of previous low back pain episodes, mode of intervention delivery (ie, face-to- face or telehealth) and self-rated adherence with the intervention. Self-rated adherence was collected using a modi\ufb01ed version of the Brief Adherence Rating Scale (collected at 3 and 6 months) where 0 is \u2018not at all compliant\u2019 and 10 is \u2018very compliant\u2019.22 Recruitment ceased when ongoing analysis indicated that suf\ufb01cient data were collected to address the study aims and no new themes were identi\ufb01ed from the \ufb01nal interviews.23 Focus groups Main \ufb01ndings An experienced qualitative researcher (JA) trained a fellow Three main themes were formed from the data. The \ufb01rst theme research team member (NCP) who then conducted the focus groups. detailed the various motivators for involvement in an exercise pro- Questions were piloted between members of the research team prior gram to prevent low back pain. The second theme comprised beliefs to data collection. The focus groups were semi-structured and ques- around exercise and low back pain. The third theme discussed the tions focused on: motivations for participation in the WalkBack trial; process of commencing a walking program and explored what re\ufb02ections of experience with the program, detailing speci\ufb01c ele- assisted initial adherence and long-term engagement. Illustrative ments of the program that they liked or disliked; barriers and facil- quotes from the focus groups are used throughout the results section itators to engagement with the program; and the opportunity to to support the key \ufb01ndings. provide general feedback. The interview guide used to facilitate the discussion is presented in Appendix 1 on the eAddenda. Theme 1: Identifying motivations for engaging in low back pain prevention The focus groups were conducted via Zoom. Sessions were recorded and transcribed verbatim by an external third party. The Low back pain is characterised by a largely unpredictable and script was compared with the focus group recording and checked for \ufb02uctuating pattern of recurring episodes, impacting individuals errors by the interviewer (NCP).","Research 185 differently. As a result, participants\u2019 motivations to then engage in Theme 2: Positive views on exercise, but cautious of which exercise exercise to manage their low back pain varied. All participants shared is \u2018right\u2019 for low back pain a desire to use a preventative approach to evade a long-term course of low back pain, increase their self-management over the condition, Participants were aware of the health bene\ufb01ts associated with a and reduce the impacts of low back pain on participation in daily more active lifestyle but were not engaging in exercise regularly at activities of importance to them. commencement of the trial. One reservation for engaging in exercise was identifying a safe and appropriate exercise for low back pain that Subtheme 1.1: I\u2019ve had a history of low back pain and don\u2019t want to re- would not exacerbate their symptoms. Due to this uncertainty, par- experience it ticipants would avoid high-impact exercises, but there was consensus amongst participants that walking was a generally safe exercise un- Participants described varied experiences ranging from a long- likely to trigger low back pain. standing history of repeated episodes to a single distinct and severe episode. Participants\u2019 interest in the trial had stemmed from a \u2018fear\u2019 of Subtheme 2.1: I know increasing exercise and reducing sedentary future exacerbations and wanting to identify a strategy to mitigate behaviour is bene\ufb01cial to general health the frequency, severity and impact of future recurrences. The walking program was seen as an opportunity to tap into the . having had a couple of years of intermittent pain, I didn\u2019t want to various bene\ufb01ts achieved by a more active lifestyle. Some participants go through it again. It was really debilitating and I was just fright- re\ufb02ected positively on impacts of exercise to their general health, ened to do anything . because it was so unpleasant. (Female, 68 while others believed they could use physical activity to address years) health-related ailments. . when my lower back twinged, I literally collapsed on the \ufb02oor, I I truly believe that movement helps you . but I wasn\u2019t necessarily couldn\u2019t move, managed to crawl to the bed and then spent the next doing it regularly. (Male, 42 years) week lying in bed . particularly being so young, that prevention aspect for me was really critical. It was terrifying. (Female, If I committed to daily walking, the hope was that I would also lose 32 years) weight. I have obstructive sleep apnoea and I had tried to lose weight for a few years. I struggled with a bit of a bad hip too, so needed to Subtheme 1.2: I want to help myself and decrease my reliance on lose weight . It [motivation to exercise] was multi-faceted for me. healthcare to manage my back pain (Female, 67 years) Most participants had not previously considered a preventative Some participants noted that periods of lockdown during the approach to managing their low back pain. The intervention\u2019s focus COVID-19 pandemic ampli\ufb01ed sedentary behaviour. Participants on a preventative self-management strategy was appealing because it described a link between exacerbations of their low back pain and was an opportunity to independently self-manage their condition. these periods of increased sedentary behaviour. I was interested in what could be done, especially from a remote area. You\u2019re a lot more sedentary [working from home] because you\u2019re not It\u2019s easier, I\u2019m sure, if you\u2019re in the city and have access to medical walking to and from train stations or bus stations or going out at help but I\u2019m remote so I needed something that I could do . to help lunch. I noticed that sitting all day is not great . from psychological prevent back pain. (Female, 57 years) bene\ufb01ts to general health and \ufb01tness (Female, 54 years) Participants described not wanting to rely on the healthcare sys- Subtheme 2.2: I\u2019m concerned that engaging in high-impact exercise will tem. The time and money commitment for treating low back pain was worsen my low back pain frustrating. There was also a desire to prevent future and potentially more severe episodes, which may require extensive medical man- Many participants believed that exercises of high intensity or agement such as medication or considerations of surgery. impact contributed to low back pain. This led to avoidance of speci\ufb01c exercises perceived to increase the chance of injury or recurrence. . going to a physio all the time, it\u2019s a lot of money too; whereas, now we can help ourselves [with a walking program], that\u2019s the best part. Because I surf, I know surfers have problems with their backs in (Female, 58 years) general. So I thought, perhaps I have to stop sur\ufb01ng if I want to not have pain . (Male, 47 years) Subtheme 1.3: I\u2019ve previously dealt with pain, but it\u2019s now impacting There was one participant who spoke extensively about advice aspects of daily life important to me provided by his healthcare provider. The advice deterred the partic- ipant from continuing with his previously active lifestyle and high- Some participants indicated that they would have previously lights the potential impact of fearful messaging from healthcare endured low back pain and rarely taken action to address it. The professionals. desire to reduce the impact of future episodes on participation in activities perceived as critical to daily life (eg, occupation, housework) I saw a physio and he said \u2018With your degenerative back, you should or commitments important to them (eg, leisure, family activities, not be running. You should be very careful about how much exercise hobbies) was a motivator for engaging in prevention. you do and make sure it\u2019s low impact\u2019 and that was a bit of a shock for me . He said, \u2018You\u2019ve got to be really careful or you\u2019ll end up in a I\u2019ve dealt with back pain for many years. I\u2019m that stupid, arrogant, wheelchair.\u2019 He gave me this framework of we\u2019ve got a serious health \u2018Yeah - I\u2019ll be right\u2019 just push through, deal with everyone\u2019s problems problem here \u2013 don\u2019t overdo it. (Male, 50 years) and ignore my own. But when you see the disappointment on your son\u2019s face when all he wants to do is kick a footy some days and you Subtheme 2.3: I perceive walking as an easy to do exercise, unlikely to can\u2019t do it, then that was it [enough motivation] for me. (Male, 42 trigger my back pain years) Contrasting with the avoidance of high-impact exercises, partici- I\u2019m on a farm, so there\u2019s a need to be \ufb01t and healthy to help on the pants reported that walking was one of the few exercises they felt farm. I just would \ufb01nd myself putting my back out just with ordinary they could manage, even during an acute episode of low back pain. daily tasks and then I can\u2019t work for a couple of weeks. (Female, 57 While participants reported a fear of engaging in certain exercise, the years) same beliefs were not held about walking.","186 Pocovi et al: Optimising adherence to a walking program Walking is one of those activities I can still do, even when I\u2019m going addressed by discussing with their physiotherapist and adjusting through periods where there\u2019s things like bending and lifting that I aspects of the program to suit their needs and lifestyle. can\u2019t do . (Male, 42 years) Four months of the year, we have incredibly high temperatures. It\u2019s Additionally, walking was reported by most participants to be an not conducive to walking unless you get up at \ufb01ve in the morning. accessible type of exercise that could be more easily tailored to their I mean, some people can manage that, but it\u2019s not a good start for me. needs and embedded within their daily routine. The ease of walking I ended up going to the gym and walking on the treadmill during that added to the appeal of the program. heat instead. (Female, 60 years) It was practical and something that you can do as part of your I like my mornings. I prefer sleeping, so we knew that right off the bat, normal routine. (Male, 49 years) it\u2019s not going to be a morning walk. So that\u2019s why I felt the program was tailored around my lifestyle, it was incorporated into my day, It can be tailored to all ages, from young people with back problems walking to and from meetings. (Female, 57 years) to elderly . And whether you work full-time, part-time and all that, you can always manage to \ufb01nd half an hour to walk. (Female, 58 Many participants described setting unrealistic goals for their years) walking early in the program. These goals were discussed in consultation with the physiotherapist, who provided guidance in Theme 3: Structures facilitating the starting of and ongoing setting realistic and achievable walking targets. This expertise and adherence to an exercise program guidance were perceived as important to a participant\u2019s success and maximising chances of sustained engagement. Participants were able to identify key elements of the program that assisted their transition from a sedentary lifestyle to a more I thought the idea was to get faster and longer through the program. active one. Strategies used to start exercising related to being held The physiotherapist kept pulling me back and saying, \u2018Let\u2019s make it accountable to their exercise program, and shared planning of the achievable\u2019. I\u2019m glad he did, otherwise I would have been writing program between the participant and physiotherapist. The use of a down that I\u2019ll walk for two hours daily, which is not achievable at all. \ufb02exible and adaptive program, support from the physiotherapist, (Male, 47 years) enjoyment of exercise and observed health bene\ufb01ts appeared critical to optimising long-term adherence to the walking program. Many participants expressed that daily life could impede their ability to remain consistent with an exercise program and emphas- Subtheme 3.1: When starting to exercise, being held accountable ised that the \ufb02exibility of the program, alongside positive support improved my adherence provided by the physiotherapist, allowed setbacks to not be seen as a failure, but instead a normal process in the journey to making exer- Participants believed that involvement in a structured and cise a habit. monitored walking program would force commitment to more reg- ular exercise. Participants described low motivation as a major hurdle I have a son who was in and out of hospital, so sometimes I couldn\u2019t to commencing exercise independently. walk. The physiotherapist said, \u2018Look, just put in some more intense sessions when you can and don\u2019t worry about when you can\u2019t\u2019. You I would have to agree that accountability was really big for me ... it\u2019s just go with ebb and \ufb02ow, \ufb01t it in when you can and just relax if you great when you have that structure around you, because I don\u2019t know can\u2019t do it. (Female, 67 years) if I can create it for myself to be honest. (Female, 38 years) Participants described various strategies useful for starting the Subtheme 3.3: The bene\ufb01ts I saw from walking enhanced my long-term walking program. Most crucial to increasing adherence were regular adherence check-ins with the physiotherapist and diarising of exercise. Participants described bene\ufb01ts that ranged from reduced severity I\u2019ve got to put these steps in, put the step count down, because [the of episodes through to complete resolution of low back pain re- physiotherapist] is going to want to see it. So it\u2019s like, \u2018Okay, I gotta do currences with the program. No participants described walking to it. Don\u2019t miss it\u2019. (Female, 49 years) have a deleterious impact on their back pain. Participants noted that once they had become less adherent to walking, the bene\ufb01ts were not The diary was a fantastic motivator for me. I wanted to follow the sustained in the long term. This was a motivator to continue regular rules and I knew [the physiotherapist] was going to be checking, so I exercise. wanted it to look the best. (Female, 32 years) If I did have some time without walking, I would start to feel a Although most participants continued to maintain some level of twinge, and as soon as I started walking again, it just went away. It walking beyond the intervention period, many indicated that it was was like magic . Who would\u2019ve thought you could prevent your more dif\ufb01cult to maintain the program when no longer being back pain just by walking? (Female, 68 years) monitored by the treating clinician. I need to get back to [walking] because I felt the difference once I Having to be accountable was important for me. Gradually life stopped. I had low back pain a few weeks ago for the \ufb01rst time in intervened, and the appealing thing about it initially was how easy it months, and it coincided with stopping the routine walks. (Male, 49 [walking] was. You didn\u2019t need equipment. You didn\u2019t need anything. years) I was annoyed and disappointed that when I wasn\u2019t held so accountable, it did drop off. (Female, 68 years) Aside from the bene\ufb01ts related to low back pain, participants were motivated to continue with walking because various additional Subtheme 3.2: A \ufb02exible and supportive program encouraged my bene\ufb01ts were noted. These bene\ufb01ts included improved sleep, mood, ongoing engagement mental health and improved general health and \ufb01tness. Although accountability was \ufb02agged by participants as an I\u2019ll de\ufb01nitely be doing more walking than I was before the program. important driver to start exercising, lapses in adherence to the pro- Similar to others [participants], the weight loss, \ufb01tness and mental gram and disengagement with the program still occurred. Primarily, health bene\ufb01ts have been really good. I\u2019m now using it as a strategy participants described that barriers such as poor weather and lack for a whole bunch of my wellness as well as my back pain. (Female, of time in\ufb02uenced their adherence; however, these were often 32 years)","Research 187 Table 3 Recommendations to aid clinicians in promoting a walking-based exercise program and assist patient commencement and adherence to exercise. Incorporate prescription of general exercise in routine clinical practice. \u000f Ask patients about current levels of physical activity and if sedentary, assess willingness to change this behaviour. \u000f If the patient is willing to increase physical activity, discuss the health bene\ufb01ts afforded by a more active lifestyle (eg, weight loss, disease management, mental health bene\ufb01ts).5 \u000f Understand patient perceptions of physical activity and address misconception or fears related to exercise.29 Use health coaching techniques during the collaborative planning of exercise goals. \u000f Share planning of exercise programs with the patient, considering exercise preferences, functional ability and lifestyle.28 \u000f Provide guidance in setting realistic and achievable goals.30 \u000f Discuss potential barriers to exercise and how the patient may mitigate these. \u000f Modify (increase or decrease) exercise targets as required and provide positive support even when exercise targets are unmet.30 Use accountability as a strategy to aid early adherence to exercise.26 \u000f Once the program is prescribed, ensure scheduled follow-ups to speci\ufb01cally review and discuss exercise progress. \u000f Monitor a patient\u2019s compliance with exercise using exercise logs, pedometer data, etc. \u000f Provide feedback on patient performance and adherence to the program at each follow-up session.31 Support patient in long-term adherence to exercise. \u000f Discuss progress and ask patient to re\ufb02ect on additional health bene\ufb01ts attained (eg, better sleep, reduced stress, weight loss). \u000f Review with patient whether health bene\ufb01ts from walking are sustained during periods when walking has ceased. \u000f Work with patient to ensure that the ongoing program is enjoyable (eg, exercising in new environments, forming social exercise networks).31 Note: These recommendations have been formulated from the key \ufb01ndings of the current qualitative study, with the addition of supporting literature. I found mentally that I was feeling calmer . I just felt more relaxed provides general health bene\ufb01ts, which may motivate a patient to during the [walking] program. (Male, 60 years) start a more physically active lifestyle. Almost all participants described a genuine enjoyment for walking Participants described accountability to the treating physiothera- following the program. It was a means of exercising, allowing pist as critical in the early stages of adherence to exercise, a notion exploration of local geography, whilst also establishing new hobbies, supported by previous literature.26 While all participants valued the interests and expanding social networks. importance of exercise, many re\ufb02ected on how a key challenge pre- viously faced when trying to create regular exercise habits was to I\u2019ve joined a hiking group and meet new people and it has been great convert positive intentions into action. A recent qualitative study fun. I\u2019ve really enjoyed it because I\u2019m travelling a bit more, going to with physiotherapists identi\ufb01ed that supporting patients who are not places that I hadn\u2019t been before . (Male, 60 years) in pain is dif\ufb01cult (ie, patients working towards prevention and self- management). Implementing passive support strategies such as an Going to scenic places and trying to go to a nice park, or somewhere open-door policy or encouraging patients to email or call if needed beautiful just made it [walking] a lot more enjoyable . I learnt was largely unsuccessful, with few patients acting on these oppor- where all the great parks were and always had shoes in my car and a tunities.27 In contrast to this, the WalkBack trial took a more active pedometer in my bag. I really enjoyed learning to be a \u2018walker\u2019, it was approach, where accountability was implemented with regularly pretty, pretty cool. (Female, 38 years) scheduled follow-up sessions, ongoing monitoring of participant progress, and encouraging participants to diarise their exercise (ie, Discussion use of pedometer and walking diary). Overall, participants re\ufb02ected positively on the ability of these strategies to improve adherence in This qualitative study identi\ufb01ed three major themes related to the early stages of a walking program. initiation and adherence to a physiotherapist-prescribed walking program for the prevention of low back pain recurrence. Theme one Participants re\ufb02ected positively on clinicians\u2019 use of health described motivations to initiate exercise, which varied between coaching and its impact on adherence in the early stages of the participants but mostly included anticipated improvements to low program. The key aspects of health coaching were the shared plan- back pain management and the added general health bene\ufb01ts of a ning and ongoing adjustment of realistic exercise goals. The exercise more active lifestyle. Theme two described participants\u2019 attitudes program in WalkBack was designed collaboratively with the partici- towards exercise and fears of exacerbating pain, leading to avoidance pant and accommodated their preferences, functional ability, avail- of high-impact and intensive exercise, whereas walking was widely able time and lifestyle. Despite participants valuing this approach, considered a safe exercise option. In the third theme, participants and the positive effects of shared decision-making in improving identi\ufb01ed accountability as a crucial strategy in the early stages of outcomes and satisfaction, the literature indicates that physiothera- their exercise program, while enjoyment of exercise and experiencing pists struggle to provide shared decision-making and it is an area improved health aided long-term adherence. requiring improvement.27,28 Participants in this study were motivated to start exercise for This study highlighted that enjoyment of the program and satis- two main reasons: to potentially prevent low back pain and to faction with outcomes (ie, feeling better both mentally and physi- improve their general health. Exercise is recommended for the cally) were critical to long-term exercise adherence. The participants prevention of low back pain,3,4 but it is unclear whether a particular used walking to explore their local area, build social networks and exercise type offers superior bene\ufb01t. Physiotherapists aim to pro- establish new hobbies (eg, hiking). These strategies led to greater vide targeted treatment to address highly speci\ufb01c pathology and enjoyment of the program and appeared to improve long-term impairments, and may neglect incorporating prescription of general adherence. Participants also described that the bene\ufb01ts felt from exercise in routine clinical practice. The effectiveness of the Walk- walking (eg, improved low back pain, reduced stress, better sleep) Back intervention (individually prescribed, progressive walking and were short-lasting. If participants had a period of non-adherence, the education program) for preventing low back pain is yet to be loss of these health bene\ufb01ts was a strong in\ufb02uence in recommencing determined, as the trial is not yet complete. Regardless of its effects exercise. Clinicians should prompt patient\u2019s re\ufb02ection on the addi- on low back pain prevention, clinicians may consider the prescrip- tional health bene\ufb01ts attained by a more physically active lifestyle, to tion of walking as an adjunct to targeted exercise for patients.18 aid long-term adherence to exercise. Walking has the advantage of being accessible, low-cost, safe and Based on these study \ufb01ndings and supporting literature, we have developed a list of practical recommendations to aid clinicians\u2019 promotion of a walking-based exercise program and to assist pa- tient commencement and adherence to exercise (Table 3). These","188 Pocovi et al: Optimising adherence to a walking program recommendations are relevant to promoting walking programs for 4. de Campos TF, Maher CG, Fuller JT, Steffens D, Attwell S, Hancock MJ. Prevention reasons other than prevention of low back pain, including achieving strategies to reduce future impact of low back pain: a systematic review and meta- physical activity recommendations. analysis. Brit J Sports Med. 2021;55:468\u2013476. A strength of this study was the re\ufb01nement of themes by a 5. Warburton DE, Bredin SS. Health bene\ufb01ts of physical activity: a systematic review multidisciplinary authorship team and two consumers who provided of current systematic reviews. Curr Opin Cardiol. 2017;32:541\u2013556. their perspectives on the interpretation of themes. This study re\ufb02ects the perspectives of participants with an underlying interest in 6. Nuzum H, Stickel A, Corona M, Zeller M, Melrose RJ, Wilkins SS. Potential bene\ufb01ts walking, mostly living in metropolitan areas across Australia, posing a of physical activity in MCI and dementia. Behav Neurol. 2020:7807856. https:\/\/doi. potential limitation to the generalisability of the \ufb01ndings. The par- org\/10.1155\/2020\/7807856 ticipants may have provided different views on exercise motivation, adherence and enjoyment compared with those not interested in 7. Erickson KI, Hillman C, Stillman CM, Ballard RM, Bloodgood B, Conroy DE, et al. exercise, those already engaged in high levels of exercise and those in Physical activity, cognition, and brain outcomes: a review of the 2018 physical rural or remote areas with different priorities. activity guidelines. Med Sci Sports Exerc. 2019;51:1242. To conclude, this study found that sedentary people who had 8. Kandola A, Stubbs B. Exercise and anxiety. Phys Exerc Human Health. 2020:345\u2013 recently recovered from low back pain were interested in a walking 352. program because they thought that walking was a safe and accessible way of potentially managing low back pain, with the potential for 9. Kelly P, Williamson C, Niven AG, Hunter R, Mutrie N, Richards J. Walking on sun- broader health bene\ufb01ts. It also identi\ufb01ed that elements of account- shine: scoping review of the evidence for walking and mental health. Brit J Sports ability, health coaching, enjoyment and the bene\ufb01ts of exercise, Med. 2018;52:800\u2013806. which improved participant adherence to walking in the short-term and long-term. 10. World Health Organisation. Physical Activity - Fact Sheet. https:\/\/www.who.int\/ news-room\/fact-sheets\/detail\/physical-activity. Accessed May 16, 2022. What was already known on this topic: In people with low back pain, walking reduces short-term pain and disability; how- 11. Lin Y-H, Chen Y-C, Tseng Y-C, Tsai S-T, Tseng Y-H. Physical activity and successful ever, adherence to walking or other general exercise is low to aging among middle-aged and older adults: a systematic review and meta-analysis moderate in this population. of cohort studies. Aging. 2020;12:7704. What this study adds: Physiotherapists should use health coaching techniques to collaboratively plan walking or other 12. Stessman J, Hammerman-Rozenberg R, Cohen A, Ein-Mor E, Jacobs JM. Physical general exercise goals as part of routine care for people with low activity, function, and longevity among the very old. Arch Intern Med. back pain. Arranging for patients to be accountable at scheduled 2009;169:1476\u20131483. follow-ups or via exercise logging aids early adherence to the general exercise. Reflecting on health benefits obtained from the 13. Guthold R, Stevens GA, Riley LM, Bull FC. Worldwide trends in insuf\ufb01cient physical general exercise supports longer term adherence. activity from 2001 to 2016: a pooled analysis of 358 population-based surveys with 1.9 million participants. Lancet Global Health. 2018;6:e1077\u2013e1086. Ethics approval: The Macquarie University Human Research Ethics Committee approved this study. All participants gave written 14. World Health Organization. WHO guidelines on physical activity and sedentary informed consent before data collection began. behaviour: at a glance. https:\/\/www.who.int\/publications\/i\/item\/9789240014886. Accessed May 5, 2023. Competing interests: Nil. Source(s) of support: Natasha C Pocovi received a three-year 15. Beinart NA, Goodchild CE, Weinman JA, Ayis S, Godfrey EL. Individual and scholarship from Macquarie University (Macquarie University intervention-related factors associated with adherence to home exercise in chronic Research Excellence Scholarship) and a one-year NHMRC Low Back low back pain: a systematic review. Spine J. 2013;13:1940\u20131950. Pain Centre of Research Excellence \u2013 ANZBACK PhD scholarship. Acknowledgements: The investigators acknowledge the National 16. Stevens ML, Lin CC, Hancock MJ, Wisby-Roth T, Latimer J, Maher CG. A physiotherapist- Health and Medical Research Council (NHMRC) Australia for funding led exercise and education program for preventing recurrence of low back pain: a the WalkBack trial from which participants were recruited - randomised controlled pilot trial. Physiotherapy. 2018;104:217\u2013223. APP1161889. The trial has been endorsed by the ANZMUSC Clinical Trial Network indicating its high priority and quality, importance to 17. Ferreira GE, Howard K, Zadro JR, O\u2019Keeffe M, Lin C-WC, Maher CG. People consumers\/patients, clinicians and policy makers, and its potential to considering exercise to prevent low back pain recurrence prefer exercise programs improve patient outcomes. The authors thank all participants that differ from programs known to be effective: a discrete choice experiment. recruited into the study. The authors acknowledge the time and effort J Physiother. 2020;66:249\u2013255. contributed to analysis by our consumers with lived experience of low back pain, Rebecca Bates and Douglas Parke. 18. Pocovi NC, de Campos TF, Lin C-WC, Merom D, Tiedemann A, Hancock MJ. Walking, Correspondence: Natasha C Pocovi, Department of Health Pro- cycling, and swimming for nonspeci\ufb01c low back pain: a systematic review with fessions, Faculty of Medicine, Health and Human Sciences, Macquarie meta-analysis. J Orthop Sports Phys Ther. 2021;52:85\u201399. University, Sydney, Australia. Email: [email protected] 19. Pocovi NC, Lin C-WC, Latimer J, Merom D, Tiedemann A, Maher C, et al. Effec- References tiveness and cost-effectiveness of a progressive, individualised walking and edu- cation programme for prevention of low back pain recurrence in adults: study 1. James SL, Abate D, Abate KH, Abay SM, Abbafati C, Abbasi N, et al. Global, regional, protocol for the WalkBack randomised controlled trial. BMJ Open. 2020;10: and national incidence, prevalence, and years lived with disability for 354 diseases e037149. and injuries for 195 countries and territories, 1990\u20132017: a systematic analysis for the Global Burden of Disease Study 2017. Lancet. 2018;392(10159):1789\u20131858. 20. Tong A, Sainsbury P, Craig J. Consolidated criteria for reporting qualitative research https:\/\/doi.org\/10.1016\/S0140-6736(18)32279-7 (COREQ): a 32-item checklist for interviews and focus groups. Int J Qual Health Care. 2007;19:349\u2013357. 2. da Silva T, Mills K, Brown BT, Pocovi N, de Campos T, Maher C, et al. Recurrence of low back pain is common: a prospective inception cohort study. J Physiother. 21. Hoffmann TC, Glasziou PP, Boutron I, Milne R, Perera R, Moher D, et al. Better 2019;65:159\u2013165. https:\/\/doi.org\/10.1016\/j.jphys.2019.04.010 reporting of interventions: template for intervention description and replication (TIDieR) checklist and guide. BMJ. 2014;348:g1687. https:\/\/doi.org\/10.1136\/bmj. 3. Steffens D, Maher CG, Pereira LS, Stevens ML, Oliveira VC, Chapple M, et al. Pre- g1687 vention of low back pain: a systematic review and meta-analysis. JAMA Intern Med. 2016;176:199\u2013208. 22. Byerly MJ, Nakonezny PA, Rush AJ. The Brief Adherence Rating Scale (BARS) vali- dated against electronic monitoring in assessing the antipsychotic medication adherence of outpatients with schizophrenia and schizoaffective disorder. Schiz- ophr Res. 2008;100:60\u201369. https:\/\/doi.org\/10.1016\/j.schres.2007.12.470 23. Fusch PI, Ness LR. Are we there yet? Data saturation in qualitative research. Qual Rep. 2015;20:1408. 24. Ritchie J, Lewis J, Nicholls CM, Ormston R. Qualitative research practice: A guide for social science students and researchers. Thousand Oaks, CA, USA: Sage; 2013. 25. Statistics ABo. Census of Population and Housing: Socio-Economic Indexes for Areas (SEIFA), Australia, 2016 [Internet]. Socio-economic Indexes for Areas (SEIFA) 2016 https:\/\/www.abs.gov.au\/ausstats\/[email protected]\/Lookup\/by%20Subject\/2033.0.55. 001w2016wMain%20FeatureswIRSAD%20Interactive%20Mapw16. Accessed June 7, 2022. 26. Oussedik E, Foy CG, Masicampo E, Kammrath LK, Anderson RE, Feldman SR. Accountability: a missing construct in models of adherence behavior and in clinical practice. Patient Prefer Adher. 2017;11:1285. 27. Ayre J, Jenkins H, McCaffrey KJ, Maher CJ, Hancock MJ. Physiotherapists have some hesitations and unmet needs regarding delivery of exercise programs for low back pain prevention in adults: A qualitative interview study. Musculoskel Sci Pract. 2022;62:102630. 28. Dierckx K, Deveugele M, Roosen P, Devisch I. Implementation of shared decision making in physical therapy: observed level of involvement and patient preference. Phys Ther. 2013;93:1321\u20131330. 29. Taulaniemi A, Kankaanp\u00e4\u00e4 M, Rinne M, Tokola K, Parkkari J, Suni JH. Fear-avoidance beliefs are associated with exercise adherence: secondary analysis of a randomised controlled trial (RCT) among female healthcare workers with recurrent low back pain. BMC Sports Sci, Med Rehabil. 2020;12:1\u201313. 30. Meade LB, Bearne LM, Sweeney LH, Alageel SH, Godfrey EL. Behaviour change techniques associated with adherence to prescribed exercise in patients with persistent musculoskeletal pain: systematic review. Brit J Health Psychol. 2019;24:10\u201330. 31. Edmunds J, Ntoumanis N, Duda JL. Helping your clients and patients take owner- ship over their exercise: Fostering exercise adoption, adherence, and associated well-being. ACSM\u2019s Health & Fitness J. 2009;13:20\u201325.","Journal of Physiotherapy 69 (2023) 141\u2013147 j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m \/ l o c a t e \/ j p hy s Invited Topical Review Physiotherapy management of moderate-to-severe traumatic brain injury Leanne Hassett a,b a Faculty of Medicine and Health, University of Sydney, Sydney, Australia; b Implementation Science Academy, Sydney Health Partners, Sydney, Australia KEY WORDS [Hassett L (2023) Physiotherapy management of moderate-to-severe traumatic brain injury. Journal of Physiotherapy 69:141\u2013147] Brain injury \u00a9 2023 Australian Physiotherapy Association. Published by Elsevier B.V. This is an open access article under Rehabilitation Physical therapy the CC BY-NC-ND license (http:\/\/creativecommons.org\/licenses\/by-nc-nd\/4.0\/). Introduction reported in mild-to-moderate injuries) and may be due to damage to the central vestibular system due to the upper motor neuron lesion or What is traumatic brain injury? the peripheral vestibular system if there is injury around the ear (eg, temporal bone fracture).11 Traumatic brain injury (TBI) is de\ufb01ned as an alteration in brain function caused by an external force to the head such as from road Burden of traumatic brain injury traf\ufb01c accidents (car, bike, pedestrian), falls, blast injuries, acts of violence and sporting injuries.1 It is a leading cause of long-term The reported incidence of TBI varies considerably between studies. disability and can occur at any time across the lifespan.2 This injury Internationally, an annual TBI incidence proportion of 295\/100,000 used to be most common in young male adults; however, with an was found in a pooled analysis of 82 studies considering all ages and ageing population, the incidence is growing in older adults (higher in severities.7 Moderate and severe injuries are estimated to account for females) who sustain a fall.3 TBIs are typically categorised as mild, most health-related costs,7 although they make up a small proportion moderate or severe, with severity based on the Glasgow Coma Scale,4 of the total incidence (23 moderate TBI\/100,000 and 13 severe TBI\/ length of loss of consciousness and\/or duration of post-traumatic 100,000). While other health conditions such as stroke are more amnesia (PTA)5 (see Table 1).6 Although the majority of TBIs are common, TBI primarily affects people during their most economically mild (approximately 80%), there is still a signi\ufb01cant number of people productive years and the effects are lifelong.12 Consequently, the who sustain a moderate or severe TBI and survive.7 economic and social costs are very high, for example: the estimated overall healthcare cost attributable to nonfatal TBI in USA was The presentation of an individual with moderate-to-severe TBI US$40.6 billion in 201613 and in Australia, the lifetime cost of new varies considerably depending on the site and severity of the brain cases of TBI was AU$10.5 billion in 2008.14 injury and other injuries sustained at the same time (eg, fractures, amputation, peripheral nerve damage). The person with TBI may Physiotherapy management of moderate-to-severe traumatic present with cognitive and language impairments, behavioural dis- brain injury turbances, and primary and secondary physical impairments.8 Cognitive and language impairments include slowed processing Adults and children who sustain a moderate-to-severe TBI are speed, reduced short-term memory, de\ufb01cits in executive functioning likely to spend days to weeks in the acute care setting,15,16 including and aphasia.9 Behavioural disturbances include mood disorders (eg, an intensive care unit where they may be intubated and ventilated for depression), personality changes (eg, passivity) and aggressive be- more severe injuries. During this acute phase, the role of the phys- haviours.10 Primary physical impairments include neuromotor im- iotherapist is to work as part of the treating team to improve respi- pairments resulting directly from the upper motor neuron lesion (eg, ratory function and prevent respiratory complications and secondary weakness, loss of coordination, reduced sensation and spasticity) and brain damage. If the individual with TBI is in a coma or con\ufb01ned to impairments resulting from orthopaedic injuries sustained in the bed and restricted in movement, particularly if spasticity is present, accident (eg, pain).8 Secondary physical impairments result from they are at risk of developing muscle contractures resting in a prolonged bed rest and immobility caused by a combination of the shortened position,8 as well as complications such as heterotopic above-mentioned primary impairments of TBI (eg, apathy, weakness) ossi\ufb01cation (ectopic bone growth in soft tissues near large joints such and include contracture and cardiorespiratory deconditioning.8 as hip and knee)17 and pressure injuries. A key role of physiotherapy Damage to the vestibular system (with symptoms of dizziness, loss in the acute care setting is to prevent or minimise the development of of balance and nausea) is also common after TBI (although more often https:\/\/doi.org\/10.1016\/j.jphys.2023.05.015 1836-9553\/\u00a9 2023 Australian Physiotherapy Association. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http:\/\/creativecommons.org\/ licenses\/by-nc-nd\/4.0\/).","142 Hassett: Physiotherapy management of traumatic brain injury Table 1 range of motion) or both. It is likely that early severe spasticity will Classi\ufb01cation of traumatic brain injury severity.6 contribute to the development of contracture, as do muscle weakness and environmental factors such as positioning in bed. A more detailed Injury severity Post-traumatic Glasgow Coma Loss of explanation of the mechanisms and relationship between spasticity amnesia Scale consciousness and contracture is beyond the scope of this review; see Boyd et al25 for further reading. Mild 0 to 24 h 13 to 15 0 to 30 min Moderate . 24 h to 7 d 9 to 12 . 30 min to 24 h The modi\ufb01ed Tardieu scale26,27 is recommended for assessing Severe: 3 to 8 hypertonus in the clinical setting and, importantly, to determine \u000f very severe .7d . 24 h whether spasticity and\/or contracture is causing hypertonus. In this \u000f extremely 7 to 28 d scale the muscle to be tested is moved passively at two speeds, very . 28 d slow (V1) and very fast (V3), and the quality and range of motion is severe assessed to determine the presence of spasticity and\/or contracture. Many clinical settings use the Modi\ufb01ed Ashworth Scale;28 however, it contracture (see below for more detail) and to monitor, report and is not as accurate as the Tardieu scale at detecting spasticity (as it manage complications with the medical and nursing teams. does not test movements at speed to elicit the velocity-dependent hyperactive stretch re\ufb02ex, which is the key feature of spasticity).29 From the acute care setting, some individuals with moderate-to- severe TBI will require admission to inpatient rehabilitation, where It is less clear how best to manage contracture and spasticity. A the length of stay often extends from weeks to months.15,16 Physio- commonly applied intervention is low load prolonged stretch, typi- therapists play an essential role as part of a multidisciplinary team that cally applied for at least 30 minutes up to 24 hours a day if serial provides inpatient rehabilitation care. On discharge, physiotherapy casting is used. A Cochrane review with meta-analysis evaluated the management is often still required as part of multidisciplinary care that effectiveness of stretch compared with no stretch in individuals with can extend for more than a year to improve functional outcomes and neurological conditions with or at risk of contracture.30 The review support the goal of community re-integration.18 found that stretch was of little clinical bene\ufb01t on joint mobility (MD 2 deg, 95% CI 0 to 3, I2 = 37%, 18 studies, 549 participants, high-certainty This section provides a summary of physiotherapy interventions evidence) and uncertain short-term effects on pain and activity lim- provided to individuals with moderate-to-severe TBI as part of itations. Sub-group analysis for acquired brain injury showed greater rehabilitation, along with the current evidence that underpins it. It is effect on joint range of motion (MD 9 deg, 95% CI 1 to 16, I2 = 63%, not possible to cover all interventions delivered by physiotherapists, three studies, 35 participants); however, the estimate was less precise but it is intended that the key interventions are discussed. and limited by a small number of studies and participants. Physiotherapy during post-traumatic amnesia Serial casting is often used in the management of signi\ufb01cant contracture in individuals with TBI. The limb is positioned so that the Individuals with moderate-to-severe TBI who require inpatient shortened muscle is in a stretched position and the cast is applied. rehabilitation may be admitted to a specialist brain injury unit, Every 3 to 7 days, the cast is removed and reapplied with the aim of particularly due to behavioural and cognitive impairments that can gradually stretching the muscle and regaining full range of motion. A be challenging to manage on a standard rehabilitation ward. When an systematic review including randomised controlled trials (RCTs) and individual is admitted to inpatient rehabilitation, they are ideally uncontrolled studies found improvements in joint mobility but not ready to actively engage in rehabilitation. This may be challenging for function.31 Since that review was published, a high-quality RCT was those with extremely severe injuries (such as those with disorder of conducted in which 2 weeks of serial casting was compared with 30 consciousness)19 who will be limited in their participation. Most in- minutes of stretch per day to reduce elbow \ufb02exor contracture in 26 dividuals with TBI who have had a period of coma will likely emerge adults with TBI.32 Although there was a large and signi\ufb01cant from coma into PTA,5 which is de\ufb01ned as a period of confusion and improvement in elbow extension range of motion immediately after disorientation, with retrograde and anterograde amnesia, poor intervention (MD 22 deg, 95% CI 12 to 31), this effect was reduced 1 attention and, frequently, agitation.20 In severe injuries, the length of day later (MD 11 deg, 95% CI 0 to 21) and there was an uncertain PTA is a good indicator of long-term outcomes.21 Length of PTA is between-group difference 4 weeks later (MD 2 deg, 95% CI \u201313 to 17). ideally assessed daily using a standardised scale such as the West- mead PTA scale.20 This scale includes 12 questions that test orienta- Box 1. Strategies for working with an individual in post- tion and memory, and the person is considered out of PTA when they traumatic amnesia. score 12\/12, 3 days in a row. The length of PTA is calculated as the number of days from injury to the \ufb01rst day they score 12\/12. \u000f Practice of whole tasks that are relevant to the patient (rather than part practice) Physiotherapists may be uncertain about commencing retraining of motor tasks when a person is still in PTA. Although PTA includes \u000f Consistency of environment (eg, exercise set up in same way de\ufb01cits in short-term memory, procedural memory (ie, memory of each time) and structuring the environment to make the task the steps, actions or sequences to perform common activities) is easier to perform retained.22 The newly updated International Cognitive (INCOG) guidelines for cognitive rehabilitation after TBI provide a speci\ufb01c \u000f The use of concrete goals and specific numbers of repetitions recommendation for physiotherapy during PTA based on level B ev- \u000f Using demonstration and performing the exercise with the idence (cohort study23). The guidelines recommend that \u2018physical therapists should make efforts to provide therapy to patients in PTA, patient while \ufb02exibly adapting session length, intensity and location based \u000f Keep instruction brief, allow time to process instruction and on the degree of agitation, cognitive impairment and fatigue of the person with TBI\u2019.20 See Box 1 for some strategies for providing minimise manual guidance physiotherapy for a person in PTA. \u000f Provide short, frequent sessions if physical and\/or cognitive Management of hypertonus fatigue is apparent, consider time of day that the patient is most alert Hypertonus (an increase in the resistance that is felt when moving \u000f Remove distractions and triggers for outbursts of aggressive a relaxed limb) is commonly seen in people with TBI along the con- behaviour (eg, noise, other individuals) from the training tinuum of care. Hypertonus may be due to spasticity (\u2018a motor dis- environment order characterised by a velocity-dependent increase in tonic stretch \u000f Ensure your safety by considering safe distance where re\ufb02exes (\u2018muscle tone\u2019) with exaggerated tendon jerks resulting from possible, safety of objects used for training, and have an hyperexcitability of the stretch re\ufb02ex as one component of the upper additional staff member if you are at all concerned motor neuron syndrome\u201924) or contracture (a loss of passive joint \u000f Time physiotherapy with pain medication if needed Based on expert opinion and recommendations from Moseley et al8 and Spiteri et al.23","Invited Topical Review 143 Study MD (95% CI) Study MD (95% CI) Brown 2005 Random Brown 2005 Random Tefertiller 2022 Wilson 2006 Pooled Pooled \u20132 \u20131 0 1 2 \u20134 \u20132 0 2 4 Favours Favours body Favours Favours body overground walking weight-supported overground walking weight-supported treadmill training treadmill training Figure 1. Mobility training using bodyweight-supported treadmill training versus Figure 2. Mobility training using bodyweight-supported treadmill training versus overground walking on 10-m walk test (m\/s). overground walking on Functional Ambulation Category. Participants in this study ranged from paralysed to very weak in their electromyographic biofeedback, working in mid-range and reducing elbow extensor muscles and this changed very little over the study friction or removing gravity for active movement.36 period. This result suggests that serial casting induces only transient increases in range of motion that are unlikely to be maintained if Once muscles are strong enough to work against gravity, it is muscles around the joint remain very weak. recommended that they are strengthened with task-speci\ufb01city in mind. A systematic review conducted in neurological populations Management of spasticity can draw upon pharmacological options found that although muscles can be made stronger with strength (eg, oral or intrathecally administered baclofen, botulinum toxin-A) and training, this does not necessarily carry over to improvements at the non-pharmacological options (eg, casting, splinting).33 Similar to the activity level if task-speci\ufb01city is not considered.37 A recent high- evidence of stretch for contracture, the effectiveness of these in- quality RCT tested the effectiveness of ballistic resistance training terventions is uncertain. A Cochrane review on the topic in TBI included compared with usual care on improving high-level mobility in in- nine studies with 134 participants and was unable to synthesise the dividuals with TBI aged 15 to 65 years.38 The targeted muscle groups data due to poor reporting and study quality.33 With the limited were the muscles shown to be important for walking (ankle plan- available evidence on the management of spasticity in neurological tar\ufb02exors, hip \ufb02exors and hip extensors for power generation, and populations, practice guidelines and consensus statements have been knee extensors for power absorption).39 The trial found that replacing developed to guide clinical practice. A recent study synthesised these three sessions\/week of non-ballistic exercises with ballistic resistance documents and provided three recommendations: focal spasticity exercises resulted in similar or better mobility measured using the management should be multidisciplinary, patient-centred and goal High-Level Mobility Assessment Tool (MD 3 points, 95% CI 0 to 6);38 directed; routine measurement of impairment and activity are strongly this was largely maintained at 6 months. Sub-group analysis of par- endorsed; and botulinum toxin-A injection should only be provided as ticipants scoring , 27 on the High-Level Mobility Assessment Tool at part of an integrated approach to focal spasticity management.34 baseline showed greater bene\ufb01t (MD 6 points, 95% CI 1 to 10). The authors of this RCT concluded that ballistic resistance training has a Providing evidence-based management for contracture and similar or better effect on mobility than non-ballistic training and spasticity in individuals with TBI remains challenging for clinicians. It may have greater bene\ufb01t for individuals with more severe mobility is unlikely that 30 to 60 minutes of routine stretch over a short period limitations. Because of a transient decrement in balance at the end of of time will prevent or reverse contracture. Provision of daily stretch the intervention period, the authors also recommended that ballistic in the long term has not been evaluated and therefore positive effects training should not be delivered at the expense of balance training. on important outcomes such as pain and pressure injury prevention cannot be ruled out. The effect on carer burden also needs to be Retraining functional motor tasks discerned. Serial casting may be an effective way of reversing contracture, but the effect is likely to be transient if muscle activity Activity limitations are common after moderate-to-severe TBI, does not return and highlights the importance of strength training in particularly on admission to inpatient rehabilitation. For example, an very weak muscles. When moderate-to-severe spasticity is present, Australian adult cohort study using prospectively collected clinical the risk of contracture developing may be reduced by positioning at- data over a 13-year period (2000 to 2013; n = 613) found that, on risk muscles in lengthened positions combined with more neutral admission, 27% could stand up equal weightbearing, 26% could walk positions through the day and commencing active training that en- at \u0015 1 m\/s independently, and 45% had upper limb dysfunction. On courages movement through range. Pharmacological interventions discharge (median length of stay 52 days, IQR 28 to 129) this such as botulinum toxin-A could be trialled where moderate-to- improved considerably: 65% could stand up equal weightbearing, 70% severe focal spasticity is present and interferes with function or could walk at \u0015 1 m\/s and only 20% had upper limb dysfunction.40 personal care. In this instance, patient-centred goals should be Improvements in physical function have also been shown in older identi\ufb01ed and a multidisciplinary team involved. adults with TBI undertaking inpatient rehabilitation.41 Strength training Mobility training is often a focus of physiotherapy management and patient goals in rehabilitation, with a certain level of mobility Limited research has been conducted in TBI to guide strength required for the individual to be discharged home from hospital and training, despite muscle weakness being a very common physical to be safe walking in the community. At the time of discharge from impairment. Individuals with TBI experience muscle weakness due to hospital, most individuals with moderate-to-severe TBI are inde- the upper motor neuron lesion causing a disruption to the motor pendent in their mobility; however, high-level balance and mobility neurons normally activating muscles. Muscle weakness is also due to limitations may persist longer-term.42 Mobility training typically in- disuse of muscles from prolonged inactivity, which causes the mus- volves high repetitions of part-task and whole-task practice, for cles to atrophy. Muscle weakness can be more signi\ufb01cant in those example: to improve walking the individual may need to perform with severe TBIs, due to hormonal disturbances from the brain injury part-task practice such as stepping practice to improve shifting their and acute care management that causes hypercatabolism.35 For par- weight onto their affected leg during stance phase, and whole-task alysed or very weak muscles, strength training needs to be set-up to practice of walking overground or on a treadmill. make it as easy as possible to elicit muscle activity. Strategies may include using electrical stimulation (with the individual engaged in Mobility training is the area of TBI physical rehabilitation with the trying to work with the machine to turn their muscles on), most RCTs conducted. This research has not been synthesised in a","144 Hassett: Physiotherapy management of traumatic brain injury Study MD (95% CI) Study SMD (95% CI) Tefertiller 2022 Random Cuthbert 2014 Random Gil-Gomez 2011 Pooled Straudi 2017 Tefertiller 2019 \u20134 \u20132 0 2 4 Pooled Favours Favours treadmill \u20132 \u20131 0 1 2 overground walking training (\u00b1 BWS) with virtual reality Figure 3. Mobility training using treadmill training (with or without bodyweight Favours balance Favours balance support, BWS) with virtual reality versus overground walking on Functional Ambu- training without VR training with VR lation Category. Figure 9. Virtual reality balance training versus other balance training on systematic review, which would be bene\ufb01cial in the future. Many of the RCTs compare the same dose of different types of mobility performance-based measures of balance. Data presented as SMD and 95% CIs. Two studies47,48 used Berg Balance Scale (score 0 to 56; higher score indicates better training including treadmill training compared with overground balance) and two studies49,50 used Community Balance and Mobility Scale (scale 0 to walking,43-46 and virtual reality balance training compared with other 96; higher score better balance). Gil-Gomez study48 included participants with an types of balance training.47-50 Three of the four studies that compared acquired brain injury, of which three of 17 participants had sustained a traumatic treadmill training with overground walking included participants who would not be independent walkers,43,45,46 and who were . 1 year after brain injury. injury;43,44,46 interventions were prescribed to be performed for 30 Cardiorespiratory \ufb01tness training and promotion of physical to 60 minutes, two to three times per week for 4 to 14 weeks. Overall, activity no difference was found between interventions (see Figures 1 to 4; for Reduced cardiorespiratory \ufb01tness is a common secondary impairment after severe TBI. Data from 11 studies with 234 partici- detailed forest plots, see Figures 5 to 8 on the eAddenda), although pants measuring peak oxygen uptake in individuals with TBI showed most trials reported improvements in both groups from baseline to an average value of 27.2 mL\/kg\/min, which is below the 10th end of intervention. Similarly for the virtual reality balance training percentile \ufb01tness level in normative data matched for age and sex.35 Reduced cardiorespiratory \ufb01tness in individuals with TBI is caused compared with other balance interventions, three of the four studies from a combination of injury factors and prolonged inactivity. Phys- included participants . 1 year post-injury48-50 and with high-level mobility problems.47,48,50 Interventions were prescribed to be per- ical inactivity is common during hospitalisation, is often extreme in the acute care setting and continues during inpatient rehabilitation.59 formed 15 to 60 minutes, three to \ufb01ve times per week for 4 to 12 weeks. Overall, there was uncertainty as to whether virtual reality was Physical inactivity continues after discharge, despite improvements in superior to other balance training interventions (see Figure 9; for a mobility due to personal (eg, fatigue) and environmental (eg, cost, transport) factors.60 A recently published USA-based cohort study detailed forest plot, see Figure 10 on the eAddenda), although most with 472 participants found that the majority of adults with trials reported improvements in both groups from baseline to end of moderate-to-severe TBI did not meet physical activity guidelines intervention. Taken together these studies indicate that clinicians (55%), and this was worse for adults aged \u0015 45 years (68%).61 For a detailed description of the aetiology of reduced cardiorespiratory could use these interventions to improve mobility, although further \ufb01tness and the relationship with physical inactivity see Hassett trials would help to con\ufb01rm this recommendation, particularly et al 2015.35 including participants in the \ufb01rst 6 months after injury. Fitness training should be commenced in inpatient rehabilitation Some RCTs have also evaluated providing an additional dose of when the patient is medically stable and able to exercise using large mobility training in additional to usual rehabilitation or usual activ- muscle groups. Assessment of \ufb01tness and physical activity levels ities. These interventions include training of a speci\ufb01c mobility task (including pre-injury levels) should be conducted prior to (eg, sit to stand,51 walking with rhythmic cues52 or additional cognitive tasks53) as well as combined mobility training.54-58 These studies commencing \ufb01tness training. Standardised \ufb01tness testing protocols may not be suitable for many individuals with TBI; however, modi\ufb01ed varied in quality (PEDro score ranging from 5 to 8\/10) and included protocols have been developed and tested in adults62,63 and chil- children and adolescents with acquired brain injuries52,54,55 and adults dren64 with TBI. Fitness tests are not always conducted by clinicians with TBI at varying times post-injury. Effectiveness of these in- prior to prescribing \ufb01tness training; however, this limits the ability to set suitable training parameters and assess an individual\u2019s response terventions varied, with some studies demonstrating between-group differences on some outcomes (eg, sit to stand ability,51,55 Timed up Study SMD (95% CI) and Go Test54 and some variables for walking52,57), but not others. Batemen 2001 Random Ding 2021 Overall, these studies provide some evidence for mobility training in Driver 2004 Laskin 2001 individuals with TBI, but the variety of interventions and outcomes Tomoto 2022 make it dif\ufb01cult to be certain of these effects. Pooled Study MD (95% CI) Freivogel 2009 Random Pooled \u20134 \u20132 0 2 4 \u20132 \u20131 0 1 2 3 Favours BWSTT or Favours robotic Favours con Favours exp overground walking treadmill training Figure 11. Updated \ufb01gure from Cochrane review of \ufb01tness training after TBI, which Figure 4. Mobility training using robotic treadmill training versus bodyweight- supported treadmill training or overground walking on Functional Ambulation Cate- evaluated effect of cardiorespiratory \ufb01tness training on \ufb01tness measured by power gory. Data are change scores. output (Watts).67 An additional two studies68,69 measured \ufb01tness using peak oxygen uptake (VO2peak). Forest plot is presented as SMD and 95% CI. Rated as low-certainty evidence.","Invited Topical Review 145 to higher intensity exercise. A recent qualitative study conducted TBI.76 Despite the lack of direct evidence supporting the guidelines with clinicians working in stroke rehabilitation identi\ufb01ed several for individuals with TBI, on balance it would be suitable to recom- barriers to implementing \ufb01tness testing and training in rehabilita- mend and promote these guidelines for individuals who have phys- tion.65 Barriers included clinicians\u2019 beliefs around how hard their ical, cognitive and behavioural capacity to achieve these levels. The patients can exercise and concerns of causing a cardiac event. WHO guidelines also provide good practice points, which promote Cochrane reviews on this topic in stroke66 and TBI67 populations have that some activity is better than none, and these points may be more demonstrated safety of \ufb01tness training in these populations, although suitable for those who are particularly sedentary or would be chal- proper pre-exercise screening of individuals is recommended as part lenged to meet the full guidelines. of the assessment process.62 Future directions for research and practice The evidence to support the effectiveness of \ufb01tness training in TBI has been synthesised in a Cochrane review that includes eight RCTs The current evidence to inform clinical practice in TBI rehabili- with 399 adult participants.67 The review found low-certainty evi- tation is limited. In all areas of physiotherapy management there is dence of improvement in cardiorespiratory \ufb01tness (MD 35 Watts, 95% low or very low certainty evidence, mostly due to studies with small CI 3 to 68), reduction in depression for studies conducted post- sample sizes, high risk of bias, and measuring a range of outcomes hospital (SMD \u20130.61, 95% CI \u20131.10 to \u20130.11, low-certainty evidence) using different outcome measures. These challenges limit the ability and no clear \ufb01ndings for other outcomes, mostly due to limited data. to synthesise studies and make \ufb01rm recommendations for clinical The addition of two small subsequently published community-based practice. The implication for clinicians is that they are forced to rely \ufb01tness trials68,69 does not change the certainty of evidence or direc- on expert opinion and clinical experience, which leads to inconsis- tion of effect of \ufb01tness training on cardiorespiratory \ufb01tness (SMD tency in care for those living with TBI. One strategy to address this 0.53, 95% CI 0.11 to 0.95) (Figure 11; for a detailed forest plot, see inconsistency is the development of clinical practice guidelines, Figure 12 on the eAddenda). where indirect evidence from similar populations (eg, stroke, cere- bral palsy) can be combined with the current limited evidence in TBI Fitness training can be prescribed according to the American as well as expert opinion to develop consensus recommendations for College of Sports Medicine guidelines for individuals with stroke and care. This process can also help to set research priorities and col- brain injury.70 Guidelines recommend that \ufb01tness training is pre- laborations between specialist brain injury services both nationally scribed three to \ufb01ve times\/week, 20 to 60 minutes in duration and at and internationally, which is essential to enhance the collective ca- an intensity of 40 to 85% heart rate reserve or 13\/20 on the Borg pacity to recruit suf\ufb01cient sample sizes to answer key questions of rating scale, or with an energy expenditure of 300 kcal per session. interest. Achieving suf\ufb01cient intensity of training can be challenging for in- dividuals with a severe injury, particularly early in rehabilitation.71 Regarding priority gaps in the evidence, there is a clear need for The use of a circuit class where patients rotate around a circuit of research across the lifespan. Trials summarised in this review are exercise stations is one strategy to achieve suf\ufb01cient dosage of predominantly trials including working age adults. Studies including \ufb01tness training. This was demonstrated in an observational study children and adolescents and older adults are desperately needed to (with embedded RCT) including 53 individuals with severe TBI un- \ufb01ll these evidence gaps. Broader inclusion criteria may be needed in dertaking inpatient rehabilitation.72 The circuit class provided a low- children, where numbers are smaller (eg, acquired brain injury); intensity, long-duration exercise session that met the caloric \ufb01tness however, participant demographic and injury data should be criteria of . 300 kcal per session for 62% (95% CI 49 to 74) of collected and individual participant data accessible so that data can participants.72 be synthesised between studies. As mentioned above, physical inactivity is common after TBI, All areas of physiotherapy management require further research, which increases the risk of developing chronic health conditions such given the low or very low certainty evidence that is currently avail- as diabetes and heart disease. People living with TBI have been shown able. Deciding on priorities of these areas should be conducted with to be at greater risk of developing chronic health conditions key stakeholders, including people living with TBI, to ensure that the compared with those without TBI, and comorbidities occurring after most important questions are addressed \ufb01rst. Collaborations between TBI are associated with higher mortality.73 Given the health risks of consumer organisations, academics and clinical services and creating physical inactivity, physiotherapists should promote physical activity learning healthcare systems77 where data can be collected and used to their patients, and support them to identify and participate in for clinical and research purposes will also assist. Data registries, such preferred physical activity. A controlled trial tested a 12-week phys- as those that exist in the USA and are currently being designed ical activity intervention (stage-matched behaviour change activities, in Australia,78 may also help to promote collaboration and research exercise prescription, community access facilitation and relapse efforts. To aid this process, consensus on a core set of physical prevention strategies) compared with control in 43 adults with ac- outcome measures to be collected will be important. Similar core sets quired brain injury.74 The intervention effectively increased adoption already exist for psychosocial function in adults79 and children80 of physical activity (time in moderate-to-vigorous physical activity: with TBI and would improve our ability to compare results MD 13 min\/day, 95% CI 1 to 25), but it was not maintained at follow- across studies and pool data for meta-analysis. Such steps will up. Further rigorous studies are needed to determine how best to consolidate current knowledge, identify evidence gaps, and facilitate support individuals with TBI to be active. optimised, evidence-based care for individuals living with moderate- to-severe TBI. In 2020, the World Health Organization (WHO) released the \ufb01rst global physical activity and sedentary behaviour guidelines for chil- eAddenda: Figures 1, 2, 3, 4, 9 and 11 can be found online at dren and adults living with disability.75 The guidelines specify that https:\/\/doi.org\/10.1016\/j.jphys.2023.05.015 people living with disability should, where possible, undertake moderate to vigorous physical activity (150 to 300 minutes\/week for Ethics approval: Nil. adults, 60 minutes\/day for children), muscle strengthening activities Competing interests: Nil. (two or more times\/week for adults, three or more times\/week for Source(s) of support: NHMRC 2020 MRFF Traumatic Brain Injury children), multicomponent exercises including balance and strength Mission Stream 2 2021, Grant number: APP2009099. activities for older adults living with disability, and to limit the Acknowledgements: I would like to acknowledge the research amount of time being sedentary. Whilst the guidelines are informed team working on BRIDGES (BRain Injury: Developing Guidelines for by evidence from neurological conditions such as multiple sclerosis, physical activitiES) project who conducted some of the searches and Parkinson\u2019s disease and stroke, they do not include direct evidence of extracted data that I have used in this review: Liam Johnson, Belinda the effects of physical activity for people living with TBI. A recently Wang, Pien Alferink, Kerry West, Sakina Chagpar, Ella Bracone. conducted rapid review on the topic did not \ufb01nd any evidence to Provenance: Commissioned. Peer reviewed. con\ufb01rm or refute the suitability of the guidelines for individuals with","146 Hassett: Physiotherapy management of traumatic brain injury Correspondence: Leanne Hassett, Faculty of Medicine and 31. Mortenson PA, Eng JJ. The use of casts in the management of joint mobility and hypertonia following brain injury in adults: a systematic review. Phys Ther. Health, University of Sydney, Sydney, Australia. Email: 2003;83:648\u2013658. [email protected] 32. Moseley AM, Hassett LM, Leung J, Clare JS, Herbert RD, Harvey LA. Serial casting versus positioning for the treatment of elbow contractures in adults with References traumatic brain injury: a randomized controlled trial. Clin Rehabil. 2008;22: 406\u2013417. 1. Menon DK, Schwab K, Wright DW, Maas AI. Position statement: de\ufb01nition of traumatic brain injury. Arch Phys Med Rehabil. 2010;91:1637\u20131640. 33. Synnot A, Chau M, Pitt V, O\u2019Connor D, Gruen RL, Wasiak J, et al. Interventions for managing skeletal muscle spasticity following traumatic brain injury. Cochrane 2. Pozzato I, Tate RL, Rosenkoetter U, Cameron ID. Epidemiology of hospitalised Database Syst Rev. 2017;11:CD008929. traumatic brain injury in the state of New South Wales, Australia: a population- based study. Aust N Z J Public Health. 2019;43:382\u2013388. 34. Williams G, Singer BJ, Ashford S, Hoare B, Hastings-Ison T, Fheodoroff K, et al. A synthesis and appraisal of clinical practice guidelines, consensus statements and 3. Gardner RC, Dams-O\u2019Connor K, Morrissey MR, Manley GT. Geriatric Traumatic Cochrane systematic reviews for the management of focal spasticity in adults and Brain Injury: Epidemiology, Outcomes, Knowledge Gaps, and Future Directions. children. Disabil Rehabil. 2022;44:509\u2013519. J Neurotrauma. 2018;35:889\u2013906. 35. Hassett L, Moseley A, Harmer A. The aetiology of reduced cardiorespiratory \ufb01tness 4. Teasdale G, Jennett B. Assessment of coma and impaired consciousness. A practical among adults with severe traumatic brain injury and the relationship with physical scale. Lancet. 1974;2:81\u201384. activity: A narrative review. Brain Impair. 2015;17:43\u201354. 5. Russell WR, Smith A. Post-traumatic amnesia in closed head injury. Arch Neurol. 36. Ada L, Canning CG. Changing the way we view the contribution of motor impair- 1961;5:4\u201317. ments to physical disability after stroke. In: Ada L, Ellis E, Refshauge KM, eds. Science-based rehabilitation: theories into practice. Edinburgh, New York: Butter- 6. Bradshaw B. Report to the Surgeon General Traumatic Brain Injury Task Force. May 15, worth-Heinemann; 2005:87\u2013106. 2007. 2008. 37. Williams G, Kahn M, Randall A. Strength training for walking in neurologic reha- 7. Nguyen R, Fiest KM, McChesney J, Kwon CS, Jette N, Frolkis AD, et al. The Inter- bilitation is not task speci\ufb01c: a focused review. Am J Phys Med Rehabil. national Incidence of Traumatic Brain Injury: A Systematic Review and Meta- 2014;93:511\u2013522. Analysis. Can J Neurol Sci. 2016;43:774\u2013785. 38. Williams G, Hassett L, Clark R, Bryant AL, Morris ME, Olver J, et al. Ballistic resis- 8. Moseley AM, Hassett LM. Traumatic Brain Injury. In: Carr JH, Shepherd RB, eds. tance training has a similar or better effect on mobility than non-ballistic exercise Neurological rehabilitation. Optimizing motor performance. 2nd edition. Philadephia: rehabilitation in people with a traumatic brain injury: a randomised trial. Elsevier; 2010. J Physiother. 2022;68:262\u2013268. 9. Novack TA, Alderson AL, Bush BA, Meythaler JM, Canupp K. Cognitive and func- 39. Williams G, Hassett L, Clark R, Bryant A, Olver J, Morris ME, et al. Improving tional recovery at 6 and 12 months post-TBI. Brain Inj. 2000;14:987\u2013996. walking ability in people with neurologic conditions: a theoretical framework for biomechanics-driven exercise prescription. Arch Phys Med Rehabil. 2019;100:1184\u2013 10. Kant R, Duffy JD, Pivovarnik A. Prevalence of apathy following head injury. Brain Inj. 1190. 1998;12:87\u201392. 40. Wong S, Hassett L, Liu J, Simpson G, Hodgkinson A, Sherrington C. Physical out- 11. Crampton A, Garat A, Shepherd HA, Chevignard M, Schneider KJ, Katz-Leurer M, comes for people admitted to an adult Brain Injury Rehabilitation Unit: a cohort et al. Evaluating the vestibulo-ocular re\ufb02ex following traumatic brain injury: a study. In: ASSBI\/NZRA Conference. New Zealand; 2019. scoping review. Brain Inj. 2021;35:1496\u20131509. 41. No\u00ebl F, Gagnon MP, Lajoie J, C\u00f4t\u00e9 M, Caron SM, Martin A, et al. Inpatient physical 12. Ma VY, Chan L, Carruthers KJ. Incidence, prevalence, costs, and impact on therapy in moderate to severe traumatic brain injury in in older adults: a scoping disability of common conditions requiring rehabilitation in the United States: review. Int J Environ Res Public Health. 2023;20:3367. stroke, spinal cord injury, traumatic brain injury, multiple sclerosis, osteoar- thritis, rheumatoid arthritis, limb loss, and back pain. Arch Phys Med Rehabil. 42. Ponsford JL, Downing MG, Olver J, Ponsford M, Acher R, Carty M, et al. Longitudinal 2014;95:986\u2013995. follow-up of patients with traumatic brain injury: outcome at two, \ufb01ve, and ten years post-injury. J Neurotrauma. 2014;31:64\u201377. 13. Miller GF, DePadilla L, Xu L. Costs of nonfatal traumatic brain injury in the United States, 2016. Med Care. 2021;59:451\u2013455. 43. Brown TH, Mount J, Rouland BL, Kautz KA, Barnes RM, Kim J. Body weight- supported treadmill training versus conventional gait training for people with 14. Access Economics. The economic cost of spinal cord injury and traumatic brain injury chronic traumatic brain injury. J Head Trauma Rehabil. 2005;20:402\u2013415. in Australia. The Victorian Neurotrauma Initiative; 2009. 44. Tefertiller C, Ketchum JM, Bartelt P, Peckham M, Hays K. Feasibility of virtual reality 15. Baguley IJ, Nott MT, Howle AA, Simpson GK, Browne S, King AC, et al. Late mortality and treadmill training in traumatic brain injury: a randomized controlled pilot after severe traumatic brain injury in New South Wales: a multicentre study. Med J trial. Brain Inj. 2022;36:898\u2013908. Aust. 2012;196:40\u201345. 45. Wilson DJ, Powell M, Gorham JL, Childers MK. Ambulation training with and 16. Dahdah MN, Barisa MT, Schmidt K, Barnes SA, Dubiel R, Dunklin C, et al. Comparative without partial weightbearing after traumatic brain injury: results of a random- effectiveness of traumatic brain injury rehabilitation: differential outcomes ized, controlled trial. Am J Phys Med Rehabil. 2006;85:68\u201374. across TBI model systems centers. J Head Trauma Rehabil. 2014;29:451\u2013459. 46. Freivogel S, Schmalohr D, Mehrholz J. Improved walking ability and reduced 17. Dizdar D, Tiftik T, Kara M, Tun\u00e7 H, Ers\u00f6z M, Akkus\u00b8 S. Risk factors for developing therapeutic stress with an electromechanical gait device. J Rehabil Med. heterotopic ossi\ufb01cation in patients with traumatic brain injury. Brain Inj. 2009;41:734\u2013739. 2013;27:807\u2013811. 47. Cuthbert JP, Staniszewski K, Hays K, Gerber D, Natale A, O\u2019dell D. Virtual reality- 18. Ponsford J, Harrison-Felix C, Ketchum JM, Spitz G, Miller AC, Corrigan JD. Outcomes based therapy for the treatment of balance de\ufb01cits in patients receiving inpa- 1 and 2 years after moderate to severe traumatic brain injury: an international tient rehabilitation for traumatic brain injury. Brain Inj. 2014;28:181\u2013188. comparative study. Arch Phys Med Rehabil. 2021;102:371\u2013377. 48. Gil-G\u00f3mez JA, Llor\u00e9ns R, Alca\u00f1iz M, Colomer C. Effectiveness of a Wii balance 19. Pistarini C, Maggioni G. Disorders of Consciousness. In: Platz T, ed. Clinical Path- board-based system (eBaViR) for balance rehabilitation: a pilot randomized clinical ways in Stroke Rehabilitation: Evidence-based Clinical Practice Recommendations. trial in patients with acquired brain injury. J Neuroeng Rehabil. 2011;8:30. Cham (CH): Springer; 2021:57\u201370. 49. Straudi S, Severini G, Sabbagh Charabati A, Pavarelli C, Gamberini G, Scotti A, et al. 20. Bayley M, Ponsford J, Trevena-Peters J, Janzen S, Harnett A, Marshall S, et al. The effects of video game therapy on balance and attention in chronic ambulatory INCOG 2.0 Guidelines for Cognitive Rehabilitation Following Traumatic Brain traumatic brain injury: an exploratory study. BMC Neurol. 2017;17:86. Injury, Part I: Posttraumatic Amnesia. J Head Trauma Rehabil. 2023;38:24\u201337. 50. Tefertiller C, Hays K, Natale A, O\u2019Dell D, Ketchum J, Sevigny M, et al. Results from a 21. Walker WC, Stromberg KA, Marwitz JH, Sima AP, Agyemang AA, Graham KM, et al. randomized controlled trial to address balance de\ufb01cits after traumatic brain injury. Predicting long-term global outcome after traumatic brain injury: development of Arch Phys Med Rehabil. 2019;100:1409\u20131416. a practical prognostic tool using the traumatic brain injury model systems national database. J Neurotrauma. 2018;35:1587\u20131595. 51. Canning CG, Shepherd RB, Carr JH, Alison JA, Wade L, White A. A randomized controlled trial of the effects of intensive sit-to-stand training after recent trau- 22. Ewert J, Levin HS, Watson MG, Kalisky Z. Procedural memory during posttraumatic matic brain injury on sit-to-stand performance. Clin Rehabil. 2003;17:355\u2013362. amnesia in survivors of severe closed head injury. Implications for rehabilitation. Arch Neurol. 1989;46:911\u2013916. 52. Kim SJ, Shin YK, Yoo GE, Chong HJ, Cho SR. Changes in gait patterns induced by rhythmic auditory stimulation for adolescents with acquired brain injury. Ann N Y 23. Spiteri C, Ponsford J, Williams G, Kahn M, McKay A. Factors affecting participation Acad Sci. 2016;1385:53\u201362. in physical therapy during posttraumatic amnesia. Arch Phys Med Rehabil. 2021;102:378\u2013385. 53. Evans JJ, Green\ufb01eld E, Wilson BA, Bateman A. Walking and talking therapy: improving cognitive-motor dual-tasking in neurological illness. J Int Neuropsychol 24. Lance JW. Symposium synopsis. In: Feldman RG, Young RR, Koella WP, eds. Spas- Soc. 2009;15:112\u2013120. ticity: disordered motor control. Chicago: Year Book Medical Publishers; 1980:485\u2013 494. 54. Katz-Leurer M, Rotem H, Keren O, Meyer S. The effects of a \u2018home-based\u2019 task- oriented exercise programme on motor and balance performance in children 25. Boyd R, Ada L. Physiotherapy management of spasticity. In: Barnes MP, Johnson GR, with spastic cerebral palsy and severe traumatic brain injury. Clin Rehabil. eds. Upper Motor Neurone Syndrome & Spasticity. 2nd edition. Cambridge University 2009;23:714\u2013724. Press; 2008:79\u201398. 55. Baque E, Barber L, Sakzewski L, Boyd RN. Randomized controlled trial of web-based 26. Tardieu G, Shentoub S, Delarue T. [Research on a technic for measurement of multimodal therapy for children with acquired brain injury to improve gross motor spasticity]. Rev Neurol (Paris). 1954;91:143\u2013144. capacity and performance. Clin Rehabil. 2017;31:722\u2013732. 27. Held J, Pierrot-Deseilligny E. Reeducation Motrice des Affections Neurologiques. Paris: 56. Peirone E, Goria PF, Anselmino A. A dual-task home-based rehabilitation Bailliere; 1969. programme for improving balance control in patients with acquired brain injury: a single-blind, randomized controlled pilot study. Clin Rehabil. 28. Bohannon RW, Smith MB. Interrater reliability of a modi\ufb01ed Ashworth scale of 2014;28:329\u2013338. muscle spasticity. Phys Ther. 1987;67:206\u2013207. 57. McClanachan NJ, Gesch J, Wuthapanich N, Fleming J, Kuys SS. Feasibility of gaming 29. Patrick E, Ada L. The Tardieu Scale differentiates contracture from spasticity console exercise and its effect on endurance, gait and balance in people with an whereas the Ashworth Scale is confounded by it. Clin Rehabil. 2006;20:173\u2013182. acquired brain injury. Brain Inj. 2013;27:1402\u20131408. 30. Harvey LA, Katalinic OM, Herbert RD, Moseley AM, Lannin NA, Schurr K. Stretch for 58. S\u00e4rk\u00e4m\u00f6 T, Huttula L, Leppelmeier J, Molander K, Forsbom MB, S\u00e4ynevirta K, et al. the treatment and prevention of contractures. Cochrane Database Syst Rev. DARE to move: feasibility study of a novel dance-based rehabilitation method in 2017;1:CD007455. severe traumatic brain injury. Brain Inj. 2021;35:335\u2013344.","Invited Topical Review 147 59. Hassett L, Wong S, Sheaves E, Daher M, Grady A, Egan C, et al. Time use and 71. Bateman A, Culpan FJ, Pickering AD, Powell JH, Scott OM, et al. The effect of aerobic physical activity in a specialised brain injury rehabilitation unit: an observational training on rehabilitation outcomes after recent severe brain injury: a randomized study. Brain Inj. 2018;32:850\u2013857. controlled evaluation. Arch Phys Med Rehabil. 2001;82:174\u2013182. 60. Lorenz LS, Charrette AL, O\u2019Neil-Pirozzi TM, Doucett JM, Fong J. Healthy body, 72. Hassett LM, Moseley AM, Whiteside B, Barry S, Jones T. Circuit class therapy can healthy mind: A mixed methods study of outcomes, barriers and supports for provide a \ufb01tness training stimulus for adults with severe traumatic brain injury: a exercise by people who have chronic moderate-to-severe acquired brain injury. randomised trial within an observational study. J Physiother. 2012;58:105\u2013112. Disabil Health J. 2018;11:70\u201378. 73. Izzy S, Chen PM, Tahir Z, Grashow R, Radmanesh F, Cote DJ, et al. Association of 61. Pham T, Green R, Neaves S, Hynan LS, Bell KR, Juengst SB, et al. Physical activity and traumatic brain injury with the risk of developing chronic cardiovascular, perceived barriers in individuals with moderate-to-severe traumatic brain injury. endocrine, neurological, and psychiatric disorders. JAMA Netw Open. 2022;5: PM&R. 2022. e229478. 62. Hassett LM, Harmer AR, Moseley AM, Mackey MG. Validity of the modi\ufb01ed 20- 74. Clanchy KM, Tweedy SM, Trost SG. Evaluation of a physical activity intervention for metre shuttle test: assessment of cardiorespiratory \ufb01tness in people who have adults with brain impairment: a controlled clinical trial. Neurorehabil Neural Repair. sustained a traumatic brain injury. Brain Inj. 2007;21:1069\u20131077. 2016;30:854\u2013865. 63. Vitale AE, Jankowski LW, Sullivan SJ. Reliability for a walk\/run test to estimate 75. Carty C, van der Ploeg HP, Biddle SJ, Bull F, Willumsen J, Lee L, et al. The \ufb01rst global aerobic capacity in a brain-injured population. Brain Inj. 1997;11:67\u201376. physical activity and sedentary behavior guidelines for people living with disability. J Phys Act Health. 2021;18:86\u201393. 64. Rossi C, Sullivan SJ. Motor \ufb01tness in children and adolescents with traumatic brain injury. Arch Phys Med Rehabil. 1996;77:1062\u20131065. 76. Johnson L, Williams G, Sherrington C, Pilli K, Chagpar S, Auchettl A, et al. The effect of physical activity on health outcomes in people with moderate-to-severe trau- 65. Connell LA, Klassen TK, Janssen J, Thetford C, Eng JJ. Delivering intensive rehabil- matic brain injury: a rapid systematic review with meta-analysis. BMC Public itation in stroke: factors in\ufb02uencing implementation. Phys Ther. 2018;98:243\u2013250. Health. 2023;23:63. 66. Saunders DH, Sanderson M, Hayes S, Johnson L, Kramer S, Carter D, et al. Physical 77. Etheredge LM. A rapid-learning health system. Health Aff. 2007;26:w107\u2013w118. \ufb01tness training for patients with stroke. Stroke. 2020;51:e299\u2013e300. 78. Fitzgerald M, Ponsford J, Lannin NA, O\u2019Brien TJ, Cameron P, Cooper DJ, et al. AUS- 67. Hassett L, Moseley AM, Harmer AR. Fitness training for cardiorespiratory condi- TBI: The Australian health informatics approach to predict outcomes and monitor tioning after traumatic brain injury. Cochrane Database Syst Rev. intervention ef\ufb01cacy after moderate-to-severe traumatic brain injury. Neurotrauma 2017;12:CD006123. Rep. 2022;3:217\u2013223. 79. Honan CA, McDonald S, Tate R, Ownsworth T, Togher L, Fleming J, et al. Outcome 68. Ding K, Tarumi T, Tomoto T, Bell KR, Madden C, Dieppa M, et al. A proof-of-concept instruments in moderate-to-severe adult traumatic brain injury: recommen- trial of a community-based aerobic exercise program for individuals with trau- dations for use in psychosocial research. Neuropsychol Rehabil. 2019;29: matic brain injury. Brain Inj. 2021;35:233\u2013240. 896\u2013916. 80. Wearne T, Anderson V, Catroppa C, Morgan A, Ponsford J, Tate R, et al. Psychosocial 69. Tomoto T, Le T, Tarumi T, Dieppa M, Bell K, Madden C, et al. Carotid arterial functioning following moderate-to-severe pediatric traumatic brain injury: rec- compliance and aerobic exercise training in chronic traumatic brain injury: a pilot ommended outcome instruments for research and remediation studies. Neuro- study. J Head Trauma Rehabil. 2022;37:263\u2013271. psychol Rehabil. 2020;30:973\u2013987. 70. Palmer-McLean K, Harbst KB. Stroke and brain injury. In: Durstine J, et al. eds. ACSM\u2019s Exercise Management for Persons with Chronic Diseases and Disabilities. Champaign (IL): Human Kinetics; 2009:287\u2013297.","Journal of Physiotherapy 69 (2023) 139 j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m \/ l o c a t e \/ j p hy s Readers\u2019 Choice Award The Editorial Board is pleased to announce the annual Readers\u2019 Choice Award, which recognises the paper published in Journal of Physio- therapy that generates the most interest by readers. The winning paper from a given year is the paper that is downloaded the greatest number of times in the 6 months after its day of publication. The Readers\u2019 Choice Award for 2022 is therefore announced in mid-2023. The winning paper from among those published in 2022 is the Invited Topical Review \u2018Physiotherapy management of Achilles tendinopathy\u2019 by Professor Peter Malliaras.1 Professor Malliaras is the Director of Higher Degree Research in the School of Primary and Allied Health Care, Monash University, Melbourne, and the Director of the Monash Musculoskeletal Research Unit. He is also a practising clinical physiotherapist. The winning paper expertly summarises the results of a large amount of research into the epidemiology, pathophysiology, diagnosis, natural history, assessment and treatment of Achilles tendinopathy. Future directions for research are also discussed. The winning paper is one of the journal\u2019s popular Invited Topical Reviews series. Consistent with previous years, this and the other Invited Topical Reviews from 2022 all received a high number of downloads.1\u20134 In fact, about half papers that have so far won the Journal of Physio- therapy Readers\u2019 Choice award have been from this series.1,5\u20137 The entire series of the Invited Topical Reviews, including the winning paper by Professor Malliaras, are freely available in full text as a collection on the Journal\u2019s website: https:\/\/www.sciencedirect.com\/journal\/journal-of- physiotherapy\/special-issue\/10XGNCFN4S9. These reviews cover clinical topics as diverse as whiplash-associated disorders, lung cancer, falls and Down syndrome. The Editorial Board of Journal of Physiotherapy congratulates Professor Malliaras on his success. References 1. Malliaras P. Physiotherapy management of Achilles tendinopathy. J Physiother. 2022;68:221\u2013237. 2. Thomas P, Baldwin C, Beach L, Bissett B, Boden I, Cruz SM, et al. Physiotherapy management for COVID-19 in the acute hospital setting and beyond: an update to clinical practice recommendations. J Physiother. 2022;68:8\u201325. 3. Holland AE. Physiotherapy management of interstitial lung disease. J Physiother. 2022;68:158\u2013164. 4. Dylke E. Measurement of breast cancer-related lymphoedema. J Physiother. 2022;68:238\u2013243. 5. Hodgson CL, Tipping CJ. Physiotherapy management of intensive care unit-acquired weakness. J Physiother. 2017;63:4\u201310. 6. Harvey LA. Physiotherapy rehabilitation for people with spinal cord injuries. J Physiother. 2016;62:4\u201311. 7. Bisset LM, Vicenzino B. Physiotherapy management of lateral epicondylalgia. J Physiother. 2015;61:174\u2013181. https:\/\/doi.org\/10.1016\/j.jphys.2023.05.003 1836-9553\/","Journal of Physiotherapy 69 (2023) 198\u2013202 j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m \/ l o c a t e \/ j p hy s Appraisal Research Note: Interpreting \ufb01ndings of a systematic review using GRADE methods Introduction What results are GRADE applied to? The interpretation of \ufb01ndings from a systematic review is a critical GRADE is used to assess the certainty of the synthesised result for last step in the review process that can in\ufb02uence decisions about each critical and important outcome. When planning a review, au- whether the bene\ufb01ts of a therapy outweigh the harms. Although thors identify and prioritise patient-important outcomes, addressing researchers are accustomed to interpreting \ufb01ndings, there can be both desirable and undesirable effects. Cochrane and GRADE guid- considerable variability in the approach taken, and it is not uncom- ance suggests limiting evidence summaries to the seven outcomes of mon for there to be no mention of the factors that review authors greatest importance for decision-making (a discussion of methods for have considered when interpreting their \ufb01ndings. Grading of Rec- identifying and prioritising outcomes is beyond the scope of this ommendations Assessment, Development and Evaluation (GRADE) research note; for guidance, see Cochrane Handbook for Intervention methods were developed to provide an agreed framework for sys- Reviews2). tematic and transparent interpretation of results. The methods pro- vide an approach for authors to consider and integrate information What are we rating our certainty in? about a body of evidence (eg, study characteristics, risk of bias, con- sistency of results across studies) when communicating their cer- Interpreting a result from a synthesis requires consideration of the tainty (or con\ufb01dence) in the effects of interventions based on results size effect (of importance to patients or not) and how certain (or from a synthesis. con\ufb01dent) we are that this re\ufb02ects the true intervention effect. To make this interpretation, authors must decide and communicate This research note aimed to provide an overview and examples what they are rating their certainty in (referred to as \u2018the target of of how the certainty of evidence is assessed using GRADE in sys- certainty\u2019 in GRADE).3 Two approaches are common in systematic tematic reviews of interventions. Readers are referred to the GRADE reviews, and both involve interpreting where the point estimate lies guidance and concept papers for in-depth coverage of GRADE in relation to a threshold. In the \ufb01rst approach, authors rate their methods.1,a certainty that there is evidence of an important effect or not (ie, an effect large enough to be of importance to patients)b using a Overview, terminology and scope threshold for a minimally important difference (MID; illustrated in Figure 1).3 In the second approach, authors rate their certainty that A GRADE assessment involves the consideration of factors (do- there is evidence of \u2018an effect\u2019 (ie, a non-null effect), in which case the mains) that in\ufb02uence certainty that the synthesised result from a line of \u2018no effect\u2019 is used as the threshold. An alternative is to interpret body of evidence re\ufb02ects the true effect of an intervention. Judge- results using multiple thresholds that de\ufb01ne a range of effect sizes ments for each GRADE domain are based on the information gener- (eg, trivial, small, moderate, large effects; see the approach used by ated and synthesised in a systematic review (eg, assessments of bias the American College of Physicians5 as implemented in their review in individual studies and from missing results in a synthesis, sensi- of corticosteroids for low back pain6). tivity analyses, characteristics of included studies). GRADE does not provide or prescribe methods for synthesis or generating other in- In this research note, we focus on the use of GRADE for rating cer- formation required for assessing certainty; however, an understand- tainty that there is evidence of an important effect. The effects of physical ing of these methods and terminology is important for understanding therapies are commonly interpreted in relation to established values GRADE and is assumed in this note. Box 1 gives an overview of the for a minimum important difference on a particular scale (eg, 10 points GRADE process and Table 1 provides de\ufb01nitions of key terms and on a 100-point scale for function or disability), so the concept of notes about scope. interpreting whether there is evidence of an important effect will be familiar to many readers and is used to illustrate the GRADE approach in this research note. GRADE guidance suggests that using a threshold for an important effect in systematic reviews is useful for decision- a This note draws on extensive GRADE guidance (for a complete list, see https:\/\/ b In some circumstances, setting a threshold for an important effect involves training.cochrane.org\/grade-approach) and the summary of that guidance pre- considering all critical and important outcomes and their relative importance sented in the Cochrane Handbook for Systematic Reviews of Interventions (value) to patients. Societal and economic outcomes may be included among these (referred to as \u2018the Cochrane Handbook) (https:\/\/training.cochrane.org\/handbook). outcomes. In such circumstances, an important effect might be de\ufb01ned as \u2018a The paper is not of\ufb01cial GRADE guidance or a GRADE concept paper, both of which bene\ufb01t large enough that patients consider that it outweighs the costs, risks and are produced and approved by the GRADE Working Group (https:\/\/www. other inconveniences of using the intervention\u2019. If such an approach was used, a gradeworkinggroup.org\/). GRADE and Cochrane guidance are complemented by higher threshold for important bene\ufb01t would be set if the intervention has practical resources including software for producing GRADE summary of \ufb01ndings appreciable side effects and costs. GRADE typically reserves this approach for tables (https:\/\/www.gradepro.org\/) and training materials to help authors to guidelines. In systematic reviews, GRADE recommends setting thresholds inde- implement GRADE (https:\/\/training.cochrane.org\/grade-approach). pendently for each outcome considering the value patients place on the outcome.4 https:\/\/doi.org\/10.1016\/j.jphys.2023.05.012 1836-9553\/\u00a9 2023 Australian Physiotherapy Association. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http:\/\/creativecommons.org\/ licenses\/by-nc-nd\/4.0\/).","Appraisal Research Note 199 makers.3 It is important to recognise that changing the threshold(s) for Box 1. Overview of GRADE process to rate and communicate rating certainty can change the interpretation of the point estimate and certainty of evidence. The steps listed should be completed in certainty in that estimate. For this reason, the threshold for interpreting the order shown for each outcome and comparison. intervention effects should be considered when planning a review; methods for determining the thresholds are beyond the scope of this \u000f Consider each GRADE domain separately and decide if you research note (see GRADE guidelines 323 and the Cochrane Handbook have \u2018no concerns\u2019 (no downgrade), \u2018serious concerns\u2019 for Reviews of Intervention for guidance7). (downgrade \u20131) or very serious concerns (downgrade \u20132) such that your certainty in the effect estimate is reduced Assessment of GRADE domains (see Table 1 for list of domains and text for what to consider when rating each domain). Concerns about imprecision Imprecision may be \u2018extremely serious\u2019 warranting a downgrade of \u20133 (eg, interpretation of the upper and lower bound of 95% CI In a GRADE assessment, a synthesis result is considered imprecise indicates that the estimate is compatible with large benefit when the 95% CI crosses the threshold of interest (eg, the threshold and large harm).8 for an important difference as shown in Figure 1).8 In Figure 1, only those estimates that cross the threshold for important bene\ufb01t or \u000f Decide on your overall certainty. For randomised trials, the important harm are rated down for imprecision (the extent of rating body of evidence begins at \u2018high\u2019 certainty (4 points). Each down depends on whether one or both thresholds are crossed). downgrade by one point (\u20131) reduces certainty one level to Contrary to what authors accustomed to interpreting \ufb01ndings based moderate, low or very low. on statistical signi\ufb01cance might imagine, the \ufb01rst estimate in Figure 1 would not be rated down for imprecision. Despite crossing the null \u000f Explain reasons for rating down. Concise, informative, (ie, 0), the 95% CI does not cross either threshold. In all but one of the accurate and easy to understand explanations for GRADE examples of imprecise results in Figure 1, only a single threshold is judgements should be reported.16 crossed (leading to a rating of serious imprecision). Where two thresholds are crossed, further downgrades are considered, with the \u000f Present a concise summary of the synthesised result and possibility that imprecision is so extreme that the result is too un- GRADE assessment for each outcome in a summary of certain to be interpreted. The preferred approach in GRADE is to rate findings table (one table per comparison; one row per imprecision based on absolute effects (see Figure 1, where dichoto- outcome; report results in relative and absolute terms; report mous outcomes are presented as an absolute risk difference, and overall certainty, judgements for each domain and example below for continuous outcomes). explanations for rating down).13 Consider the example of arthroscopic surgery for knee osteoar- \u000f Write a plain language interpretation of each finding. thritis.9 The authors concluded that arthroscopic surgery leads to GRADE has guidance on writing informative, plain language little or no difference in pain, as the improvement seen excluded any statements to communicate the effect and certainty in the clinically important threshold, based on a prede\ufb01ned important dif- effect for each outcome.18 ference of 12 points on a 0 to 100 point pain scale: evidence from four trials (309 participants) indicated that compared with placebo, mean \u000f Communicate certainty wherever results are reported. improvement with knee arthroscopy was 4.6 points better on a 0 to Different formats are appropriate for different sections of 100 point scale (where 0 is no pain). As the 95% CI was smaller than the systematic review (see PRISMA 2020 for details17). the 12-point threshold (0.02 to 9 points better), the result was not downgraded for imprecision. for investigating heterogeneity (such as meta-regression or sub- group analyses) fail to identify a credible explanation (eg, differ- Inconsistency ences in PICO or methodology), then authors should rate down.11 In systematic reviews, authors must consider whether the results Indirectness of the individual studies contributing to a meta-analysis are consis- tent. Important unexplained differences in the direction or size of the While all studies included in a review must meet the review PICO effects (ie, heterogenous results), reduces certainty in the combined criteria, typically only a subset contributes to each synthesis (com- effect estimate and can make the estimate uninterpretable in extreme parison). This subset of studies must have examined the intervention, cases. Identifying inconsistency in a GRADE assessment begins by comparator and outcome of interest; however, they may not directly considering: whether the con\ufb01dence intervals of the individual address all components of the PICO for the synthesis (comparison).12 studies in the meta-analysis overlap (suggesting compatible in- For example, the studies contributing to a synthesis may involve only terpretations of effect); the I2 statistic for quantifying inconsistency part of the population of interest (eg, only higher resource settings (ie, variation in the effects due to heterogeneity rather than leading to uncertainty about the effects in lower resource settings), chance10); and the strength of evidence for heterogeneity (eg a p- implement the intervention in a speci\ufb01c way (eg a one-off exercise value from a chi-square test).11 Cochrane guidance advises against intervention when usual practice is multiple sessions), or measure using these approaches individually to judge inconsistency; for surrogate outcomes (eg reporting general medical deterioration example, rather than using cut-offs for I2 to judge inconsistency among older hospital patients during unplanned hospital stays rather alone, I2 values must be considered in relation to the direction and than a direct measure of deterioration such as new incidence of magnitude of study results and the strength of evidence for hetero- delirium). Differences between the study PICO characteristics (iden- geneity.10 None of the above methods indicate whether the observed ti\ufb01ed through a synthesis of study characteristics) and the synthesis inconsistency in effect estimates is important. As with imprecision, PICO should not automatically be assumed to be important. If authors such a judgement should be made by considering where the point have empirical evidence that a factor that differs between the study estimates of individual studies lie in relation to the threshold for an characteristics and synthesis PICO is likely to modify intervention important difference.11 Returning to Figure 1, the point estimates for effects importantly, they may consider rating down for indirectness, individual studies may vary in size (ranging from small but important especially if all or the majority of data come from such studies. An bene\ufb01t to large bene\ufb01t), but if all lie to the right of the threshold for alternative to rating down is to limit the interpretation based on important bene\ufb01t, then authors need not be concerned about characteristics of the body of evidence (eg, we have high certainty inconsistency. Where point estimates lie either side of one or both about effects on mild disease, but are uncertain about effects on thresholds, then inconsistency is a concern. If pre-speci\ufb01ed analyses moderate and severe disease). Risk of bias Systematic review authors are accustomed to reporting an assessment of the risk of bias in each included study in their review.","200 Appraisal Research Note Table 1 Terms used in this paper and notes about scope. Term Notes about scope Certainty (or con\ufb01dence) The certainty that the true intervention effect \u2018lies on one side of a speci\ufb01ed threshold or within a chosen range\u2019. 4 Authors can choose as the in the evidence threshold, the null (the line of null effect) or an important effect (the minimum important difference between interventions). Or authors can rate their certainty that the true effect lies within a chosen range (ie, within the range de\ufb01ned for trivial, small, moderate or large effects). Authors interpret where the point estimate of the synthesis result lies in relation to this threshold (eg, important bene\ufb01t) and then rate their certainty that this point estimate re\ufb02ects the true effect. GRADE domains The factors considered when rating certainty (or con\ufb01dence) in the synthesis result. Five domains are considered when rating the certainty of Imprecision a body of evidence from randomised trials (RCTs), non-randomised studies of interventions (NRSI) or both: Inconsistency Does interpreting the con\ufb01dence interval limits of the combined estimate of effect lead to importantly different interpretations of the bene\ufb01ts Indirectness (or harms) of the intervention? Risk of bias Does the effect size vary importantly from study to study without a credible explanation for the variation? Publication bias Are there differences between the questions the studies address and our synthesis question such that the results may not apply to our question? Do the studies have important methodological limitations such that we might be over (or under) estimating the size of the intervention effect? Are we missing studies or results that we suspect were not reported because they were considered unfavourable? For NRSIs, \u2018upgrading\u2019 factors may apply: large effect size, dose response, opposing residual confounding.1 Consideration of these factors is outside the scope this paper (see GRADE guidelines 1819). In principle, these domains could apply to any body of evidence, but compelling examples of reasons to rate up evidence from randomised trials are lacking. PICO questions Each systematic review, synthesis within a review, and included study addresses a question that can be de\ufb01ned in terms of the Population(s), Intervention(s), Comparator(s), and Outcome(s) (PICOs) of interest. The PICO framework is widely used to de\ufb01ne review and synthesis questions. An example of a review question speci\ufb01ed in PICO terms is \u2018For people with osteoarthritis of the knee [population], what is the effect of exercise [intervention] compared to no exercise or a different type of exercise [comparator] on pain, physical and psychosocial functioning, and quality of life [outcomes]?\u2019 PICO criteria Are speci\ufb01c decision rules, used to operationalise the PICO questions, that de\ufb01ne which studies are eligible for the review and, ideally, each synthesis. Study characteristics Includes the PICO questions addressed in a study, the number and characteristics of participants, and the study methodology (design features and risk of bias). Comparison The intervention groups (or contrast) compared in a synthesis. In this research note, we focus on comparison of two intervention groups, which is referred to as a pairwise comparison. A newer method \u2013 network meta-analysis \u2013 enables multiple interventions to be compared in the same analysis. GRADE methods exist for these analyses, but are beyond the scope of this note.20 Synthesis (of results) The statistical combination of results from a set of studies addressing the same (or similar) PICO question. In principle, any method of quantitative synthesis could be used; however, in this research note, the synthesis method is assumed to be meta-analysis because this provides the most complete information for a GRADE assessment. In this note, we refer to the result from the statistical combination of results from a set of studies as the \u2018synthesis result\u2019, the \u2018meta-analytic result\u2019 or a \u2018combined estimate\u2019 (below). Synthesis (of study This refers to the synthesis of information about study characteristics that is needed for a GRADE assessment, such as the study PICO characteristics) characteristics, risk of bias and so on. Combined estimate of Also known as \u2018combined estimate\u2019 or \u2018meta-analysis estimate\u2019. The estimate of the effect from the synthesis of results from two or more intervention effect studies using meta-analysis. The analysis yields an overall estimate (which we refer to as the \u2018point estimate\u2019) and a con\ufb01dence interval (generally a 95% CI, a loose interpretation of which is that it indicates the range of values in which we can be 95% con\ufb01dent that the true effect lies).7 Body of evidence The set of studies that are eligible for inclusion in a given synthesis (ie, comparison and outcome) and to which GRADE is applied. Includes all studies that measure the outcome irrespective of whether results are available for inclusion in the synthesis or not. Figure 1. Interpretation of effect estimates using a threshold for an important difference of 5% (5 more people per 100 experiencing the outcome is important bene\ufb01t, 5 fewer is important harm). Point estimates that lie in the greyed area are interpreted as unimportant or trivial (\u2018little or no difference\u2019 between treatment and control). Imprecision is judged according to whether or not the 95% CI crosses one or both thresholds. Where the 95% CI is extremely wide (bottom estimate) the result is considered too imprecise to interpret (extremely serious imprecision). Note, for ef\ufb01ciency of presentation we use a horizontal line to depict a \u2018combined estimate\u2019 (discussed under Imprecision) and results from single studies (discussed under Inconsistency), whereas a diamond would generally be used for combined estimates. (Figure adapted from Nikolakopoulou et al.21) RD = risk difference.","Appraisal Research Note 201 In a GRADE assessment, authors consider the potential for the com- which results were unavailable, the conclusion was unlikely to bined effect estimate from a meta-analysis to be over-estimated or change substantially if results from this trial were included. under-estimated (biased) because of methodological limitations in the studies contributing to an analysis (eg arising from the random- Summarising and communicating \ufb01ndings isation process, deviations from the intended interventions, missing outcome data, measurement of the outcome, and selection of the Once each GRADE domain has been considered, an overall rating reported result).13 Where studies at high risk of bias in\ufb02uence the of certainty is made and communicated using one of four levels (high, estimate of effect importantly, then authors should consider rating moderate, low and very low certainty) (see Box 1). GRADE summary down for risk of bias. Decisions about whether studies at risk of bias of \ufb01ndings tables are widely used to present the results assessed (the in\ufb02uence the estimate are informed by results from whatever strat- synthesised result for each outcome reported in relative and absolute egy authors used to address risk of bias in their review. For example, terms), the overall GRADE judgement for each outcome, and expla- authors may undertake a sensitivity analysis wherein only the subset nations for downgrades and more borderline decisions. These ex- of studies judged to be at low risk of bias are included in the meta- planations are an important feature of the GRADE approach; they analysis. Alternatively, the meta-analysis may be strati\ufb01ed by risk of enable readers to con\ufb01rm decisions, apply the results appropriately, bias judgements and the stratum and overall effect estimates re- and reuse or reassess the evidence for different contexts.16,17 Finally, ported. In either case, there is no need to rate down if interpretation authors are encouraged to communicate the size of effect and cer- of the point estimate for all studies and each subset is similar. tainty of evidence in text and summary tables using GRADE plain However, there is reason for concern if the point estimate from low language statements.18 risk of bias studies differs importantly from an estimate that includes studies at high risk of bias (ie, leading to a different interpretation of Conclusion the intervention effect, especially if the latter is more favourable). Authors should plan which estimate to report from strati\ufb01ed analyses GRADE methods provide a widely used approach for systematic \u2014 a decision that has implications for the GRADE judgement. One review authors to interpret synthesis results and communicate their option is to report the estimate from the subset of lower risk of bias certainty in those results. This research note has outlined how GRADE studies and not downgrade for risk of bias. The alternative is to report methods are used to systematically and transparently consider factors the estimate from all studies and rate down for risk of bias. Both that are widely agreed to in\ufb02uence certainty in evidence about the options involve trade-offs, for example the \ufb01rst approach may in- effects of an intervention on outcomes. crease imprecision if a large amount of data are excluded. If neither strati\ufb01ed nor sensitivity analyses are possible, the weight of studies at Competing interests: Sue Brennan is the Director of the Mel- high risk of bias in the analysis is considered. Where studies at high bourne GRADE centre, a member of the GRADE Guidance Group for risk of bias contribute more weight, and so are likely to in\ufb02uence the the international GRADE Working Group, and a Senior Research combined effect estimate importantly, authors should consider rating Fellow at Cochrane Australia. Renea Johnston is a member of the down. Melbourne GRADE centre, and a Managing Editor of Cochrane Musculoskeletal. Publication (reporting) bias Acknowledgements: Sue Brennan\u2019s position at Cochrane Australia Reporting bias occurs when study results are not reported because is funded by the Australian Government through the National Health the results are considered unfavourable to the intervention (eg, and Medical Research Council. We thank Joanne McKenzie (Methods magnitude or direction of effect suggests harm or no bene\ufb01t). In sys- in Evidence Synthesis Unit, School of Public Health and Preventive tematic reviews, authors need to consider two main forms of reporting Medicine, Monash University) and Melissa Murano (Cochrane bias. The \ufb01rst is when a study result is available for inclusion in a Australia, School of Public Health and Preventive Medicine, Monash synthesis but the result has been selectively reported (eg, the most University) for critical review of the manuscript and suggestions for favourable result from multiple measures of the outcome; the most improvement. favourable result from multiple analyses). This form of reporting bias is considered in the assessment of the risk of bias in an individual study Provenance: Invited. Peer reviewed. and the GRADE risk of bias domain. The second form of reporting bias occurs when a study measures an outcome but the result is unavailable Sue E Brennan and Renea V Johnston for the synthesis because it is considered unfavourable to the inter- School of Public Health and Preventive Medicine, Monash University, vention and either the outcome, or the entire study, is unavailable. Cochrane provides a framework for assessing bias due to missing re- Melbourne, Australia sults in a synthesis,14 and a tool is being developed to guide the assessment (www.riskofbias.info\/welcome\/rob-me-tool). These References methods note limitations of graphical and statistical methods for detecting publication bias (funnel plots and tests of asymmetry), 1. Cochrane. Cochrane training - GRADE approach. Viewed 13 April 2023 from instead emphasising assessment of selective non-reporting of results https:\/\/training.cochrane.org\/grade-approach from studies identi\ufb01ed as eligible for the review. GRADE judgments of publication bias should be based on the results from these assessments. 2. McKenzie J, et al. Chapter 3: De\ufb01ning the criteria for including studies and how Similar to judgements for the GRADE risk of bias domain, authors need they will be grouped for the synthesis. In: Higgins JPTTJ, Chandler J, Cumpston M, to consider the likely impact of missing results on the estimate of Li T, Page MJ, Welch VA, ed. Cochrane Handbook for Systematic Reviews of In- intervention effect. If the amount of missing data is trivial and unlikely terventions version 63. www.trainingcochraneorg\/handbook. 2002. to change the point estimate and its interpretation importantly, there is no need for concern. Unlike other GRADE domains, publication bias is 3. Zeng L, et al. J Clin Epidemiol. 2021;137:163\u2013175. either \u2018strongly suspected\u2019 or \u2018not detected\u2019. 4. Hultcrantz M, et al. J Clin Epidemiol. 2017;87:4\u201313. 5. Chou R, et al. Ann Intern Med. 2017;166:480\u2013492. Consider a review of vertebroplasty for osteoporotic vertebral 6. Chou R, et al. Cochrane Database Syst Rev. 2022;10:CD012450. compression fracture that found high-quality evidence based upon 7. Sch\u00fcnemann HJ, Vist GE, Higgins JPT, Santesso N, Deeks JJ, Glasziou P, et al. Chapter \ufb01ve trials (541 randomised participants, one trial with incomplete data reported) that vertebroplasty provides no important reduction 15: Interpreting results and drawing conclusions. In: Higgins JPT, Thomas J, in pain or disability compared with placebo (sham vertebroplasty).15 Chandler J, Cumpston M, Li T, Page MJ, Welch VA, ed. Cochrane Handbook for The authors decided not to downgrade certainty for publication bias, Systematic Reviews of Interventions version 63. www.trainingcochraneorg\/ as although there was one additional placebo-controlled trial for handbook. 2022. 8. Zeng L, et al. J Clin Epidemiol. 2022;150:216\u2013224. 9. O\u2019Connor D, et al. Cochrane Database Syst Rev. 2022;3:CD014328. 10. Deeks JJ, Higgins JPT, Altman DG. Chapter 10: Analysing data and undertaking meta-analyses. In: Higgins JPT, Thomas J, Chandler J, Cumpston M, Li T, Page MJ, Welch VA, ed. Cochrane Handbook for Systematic Reviews of Interventions version 63. www.trainingcochraneorg\/handbook. 2022.","202 Appraisal Research Note 11. Guyatt G, et al. J Clin Epidemiol. 2023;158:70\u201383. Page MJ, Welch VA, (ed). Cochrane Handbook for Systematic Reviews of Interventions 12. Guyatt GH, et al. J Clin Epidemiol. 2011;64:1303\u20131310. version 63. www.trainingcochraneorg\/handbook. 2022. 13. Sch\u00fcnemann HJ, Higgins JPT, Vist GE, Glasziou P, Akl EA, Skoetz N, et al. Chapter 14: 15. Buchbinder R, et al. Cochrane Database Syst Rev. 2018;11:CD006349. 16. Santesso N, et al. J Clin Epidemiol. 2016;74:28\u201339. Completing \u2018Summary of \ufb01ndings\u2019 tables and grading the certainty of the evidence. 17. Page MJ, et al. BMJ. 2021;372:n71. In: Higgins JPT, Thomads J, Chandler J, Cumpston M, Li T, Page MJ, Welch VA, ed. 18. Santesso N, et al. J Clin Epidemiol. 2020;119:126\u2013135. Cochrane Handbook for Systematic Reviews of Interventions version 63. www. 19. Sch\u00fcnemann HJ, et al. J Clin Epidemiol. 2019;111:105\u2013114. trainingcochraneorg\/handbook. 2022. 20. Brignardello-Petersen R, et al. BMJ. 2020;371:m3900. 14. Page MJ, Higgins JPT, Sterne JAC. Chapter 13: Assessing risk of bias due to missing 21. Nikolakopoulou A, et al. PLoS Med. 2020;17:e1003082. results in a synthesis. In: Higgins JPT, Thomas J, Chandler J, Cumpston M, Li T,","Journal of Physiotherapy 69 (2023) 133\u2013135 j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m \/ l o c a t e \/ j p hy s Editorial Who should judge treatment effects as unimportant? Christina Abdel Shaheed a,b, Stephanie Mathieson a,b, Ross Wilson c, Ann-Mason Furmage d, Christopher G Maher a,b aSydney School of Public Health, Faculty of Medicine and Health, The University of Sydney, Sydney, Australia; bInstitute for Musculoskeletal Health, Sydney, Australia; cDunedin School of Medicine, University of Otago, Dunedin, New Zealand; dRoyal Prince Alfred Hospital Consumer Advisory Group, Sydney Local Health District, Sydney, Australia Introduction and the clinical context in which the decision is being made, these estimates need to be derived independently for every decision It is common to encounter authors of trials, systematic reviews problem. This evidence does not exist in most clinical scenarios and guidelines dismissing treatments on the grounds that their ef- investigated in research, so an approach is required for when a well- fects are not large enough to be worthwhile. The basis is often as derived estimate of the smallest worthwhile effect is unavailable. simple as the treatment effect not surpassing a standard benchmark, such as 15 units on a 0 to 100 pain scale. We have been guilty of this Why can\u2019t there be just one generic smallest worthwhile effect practice ourselves, but now realise that things are more complex. This benchmark? Editorial begins with a brief introduction to the smallest worthwhile effect, also referred to as the minimum clinically important differ- While it may initially seem appealing to adopt a generic smallest ence. It then explains the limitations of the standard benchmark worthwhile effect that applies across the board, this convention approach and proposes a more nuanced approach that considers the causes problems for both science and the patient. In the scienti\ufb01c complexity of the clinical encounter and the wishes of the patient. An community, an obvious problem is that researchers cannot agree on example from pain research is used. the benchmark, so decisions on treatment effects that are large enough to be worthwhile are quite inconsistent across studies. What is the smallest worthwhile effect? Although it may be useful to know the average smallest worthwhile effect in a sample of the population (eg, from a planning perspective), The smallest worthwhile effect is an attribute of the clinical sce- a desire for standardisation has inadvertently created disorder. For nario, not of the outcome measure.1 What constitutes a clinically decades we have been trying to \ufb01nd a single value for a concept that important effect is speci\ufb01c to the individual patient encounter. should not have a single value. Consequently, there is a wide range of Context-speci\ufb01c factors \u2014 such as the patient\u2019s baseline health status, estimates and much confusion over the utility of these estimates. cost, risk of harms, accessibility of the treatment being offered and impact on other aspects of health-related quality of life \u2014 and idio- The approach is also inef\ufb01cient as it is dichotomising the trial\/ syncratic patient preferences are likely to in\ufb02uence the perceived review results (as being worthwhile or not worthwhile). The practice importance of a treatment effect. This challenges subsequent thinking resembles using p-values to label treatments as effective or ineffec- about the smallest worthwhile effect, where it has typically been tive. Applying the same smallest worthwhile effect cut-off to a considered as a \ufb01xed attribute of the measure (eg, 15 points on a 0 (no placebo-controlled trial and a comparative effectiveness trial seems pain) to 100 (worst pain) scale). unwise because all non-zero differences are likely to be worth noting when comparing two treatments head-to-head (eg, when required Despite the focus on estimating the single smallest worthwhile (ie, PRN) versus time-contingent dosing of drug X). effect for an outcome measure, the range of obtained estimates varies widely.1 In the pain \ufb01eld, thresholds for (within-group) pain reduc- The practice also ignores the reality that treatments vary enor- tion ranging from 8 to 40 points on a 0 to 100 pain scale have been mously in costs, ease of access, complexity and potential for harm. It suggested to de\ufb01ne the smallest worthwhile effect,2\u20135 and in sys- makes little sense to apply the same smallest worthwhile effect to, for tematic reviews of drug interventions for musculoskeletal pain, the example: a 3-day course of paracetamol, a 6-week residential pain smallest worthwhile effect (between-group difference) typically management program, a spinal cord stimulator, spinal fusion surgery ranges from 9 points6,7 to 20 points8 (on a 0 to 100 pain scale). There and restructuring the health system, as in the STarT Back trial is also ongoing debate around whether the (between-group) smallest (strati\ufb01ed primary care management for low back pain).10 Yet in the worthwhile effect should be 10 or 15 points (or higher) on a 0 to 100 pain discipline, particularly the back pain discipline, we have ended pain scale in systematic reviews of back pain interventions. However, up in the untenable position where a complex intervention that re- in a clinician-patient interaction, the smallest worthwhile effect value quires restructuring of the health system can be judged against the is best arrived at by discussion with the particular patient for whom same benchmark as a comparative effectiveness trial of two dosing the therapy is being considered,9 alongside a consideration of cost, regimens for an over-the-counter medicine. harms and convenience of the treatment. As the average smallest worthwhile effect in a given scenario will depend on things like the However, the more important problems lie within the realm of costs, risk of harms, ease of implementation, convenience and patients. The primary problem is that it takes the patient out of accessibility of the intervention (and the relevant comparator\/alter- decision-making and harks back to a more paternalistic stage in native), the population for whom the treatment is being considered, medicine. When policymakers or guideline committees dismiss effective therapies because their mean effects are less than a generic smallest worthwhile effect benchmark imposed by researchers, https:\/\/doi.org\/10.1016\/j.jphys.2023.04.001 1836-9553\/\u00a9 2023 Australian Physiotherapy Association. Published by Elsevier B.V. This is an open access article under the CC BY license (http:\/\/creativecommons.org\/licenses\/by\/ 4.0\/).","134 Editorial patients may lose access to therapies that they would deem worth- including clear information on the likely improvement in outcome while in their particular circumstances. Conversely, patients may be with the intervention, relative to baseline (eg, baseline pain score), and offered therapies because the treatment effects are larger than the the likely outcome without the treatment. Here, researchers should assumed smallest worthwhile effect, without suf\ufb01cient regard to the attempt to describe the following: What would be the pre- and post- individual circumstances and preferences of the patient. treatment outcomes (eg, average pain score) for the typical partici- pant if they received the active or the control treatment (as might be The generic benchmarks that have been set in the pain \ufb01eld also seen in a shared decision-making tool)?13 What would be the risk of seem based upon an expectation of large treatment effects. The re- adverse events with active or control treatments (as might be seen in ality is that most treatments in medicine have modest effects11,12 and a shared decision-making tool)? Are there alternative treatments? If so a more constructive approach would be to guide patients in so, what is known about them? choosing between them (or no treatment). It is also important to acknowledge that the treatment effect What should researchers do? represents an average response in trial participants, and this cannot be used to reliably predict individual patient response. However, the Treatments vary enormously in costs, ease of access, complexity mean treatment effect estimate can serve as an important guide in and potential for harm, and the controls similarly vary: another active discussions. treatment, placebo, no treatment or usual care. Instead of adopting a generic smallest worthwhile effect benchmark, the following strate- Help readers to contextualise the treatment contrast by providing gies should be considered to support a shared decision-making information such as treatment cost, complexity, ease of access and process. safety Do not dichotomise as \u2018use treatment\u2019 versus \u2018don\u2019t use treatment\u2019; Divorced of a clinical context, the effect size alone is insuf\ufb01cient to instead outline the magnitude and precision of the treatment help a patient make an informed decision. Researchers should pro- estimate vide clear information on the nature of the trial intervention (eg, cost, program duration, availability and potential for harm), supplemented Rather than dichotomising results and making recommendations with trial participants\u2019 views on the treatment program. This infor- \u2018to use\u2019 or \u2018not use\u2019 the intervention based on an arbitrary smallest mation should also be provided for the control intervention. The worthwhile effect threshold, researchers should convey the magni- varying nature of interventions demands more nuanced interpreta- tude of the effect\u2019s estimate, precision and certainty of evidence. The tion of study \ufb01ndings, which considers and openly discusses the mean change and 95% con\ufb01dence interval should be clearly described broader clinical context. together with a description of the cost, ease of access, complexity and potential for harm of the treatment. A more nuanced approach moving forward It is important to avoid conclusions that explicitly dismiss the There are more nuanced approaches that can avoid imposing a use of an intervention (eg, \u2018our \ufb01ndings do not support the use of X\u2019 decision based on a single value, without regard to the speci\ufb01c or \u2018the effects of X were not considered clinically worthwhile\u2019) as clinical circumstances. For example, the American College of Phy- this removes the patient from the decision-making. Instead, con- sicians recommends that effect estimates be categorised as \u2018small\u2019, clusions can be framed as \u2018there is [high\/moderate\/low\/very low] \u2018moderate\u2019 or \u2018large\u2019.14 Categorising effect estimates in this way (or certainty evidence that X provides [large\/moderate\/small\/very small] using a similar, adapted approach depending on the treatment effects for [outcome].\u2019. As previously described, judging whether a intervention) overcomes the labels that imply a treatment is only given treatment effect is meaningful or worthwhile for a patient can worthwhile if it meets a pre-speci\ufb01ed threshold value. In a recent be understood with regard to the preferences of the individual pa- review of paracetamol for pain, a mean change score (between- tient, the clinical context in which the treatment decision is being group difference) , 10 points was considered as very small, 10 to 19 made, and other relevant aspects of the treatment under consider- points as small, 20 to 29 point as moderate and . 30 points as large ation. It is possible for a patient to favour a treatment with a small on a 0 to 100 pain scale,15 an approach that resembles recommen- effect that is cheap, easily accessible and safe over a treatment with dations from the American College of Physicians. These approaches a larger effect but which carries greater risk of harm, is costly or is avoid the pitfalls associated with imposing a single value and burdensome to the patient. Therefore, in shared decision-making, a empower the reader to make their own decision about the clinical decision around the clinical utility of an intervention should not be utility of the treatment. imposed. Finally, engaging consumers in the design and reporting of studies Don\u2019t throw away data by only focusing on mean between-group is increasingly being recognised as important. This engagement difference process should include clear input on the outcomes that should be collected, reporting of those outcomes and general interpretation of Basing decisions around clinical utility of an intervention using results. A consumer perspective may just reveal that many treatments the mean between-group difference in pain scores says little about previously dismissed for not meeting a pre-speci\ufb01ed threshold are patient satisfaction with the treatment or the proportion of patients worthwhile in the individual\u2019s particular circumstance, and supports who achieved minimal or no pain as a result of treatment compared the case for considering treatments without regard to a pre- with those outcomes in the control intervention. This information is determined smallest worthwhile effect. useful in helping to guide patient decision-making and re\ufb02ects the considerations that patients see as important.1 When presenting in- Conclusion formation about treatment response, researchers should provide commentary around these additional outcomes and how they Researchers are moving away from a smallest worthwhile effect differed between the study arms. threshold and characterising the magnitude of effect as (albeit arbi- trary) small, moderate or large, which then leaves the patient to Explain to readers what the mean effect would represent for the decide whether a small, moderate or large effect is worthwhile for trial participant them. In the absence of a well-derived estimate of the smallest worthwhile effect, this thinking ought to be applied in the reporting Researchers can assist readers (patients, clinicians, policymakers, of trial or systematic review results. A consideration of the above guideline committees, etc) to better understand the mean effect by points within a shared decision-making model can support clinicians","Editorial 135 and patients to make informed decisions about treatment, without References imposing an arbitrary smallest worthwhile effect cut-off in the setting of an individual patient making a shared decision with their 1. Ferreira M. J Physiother. 2018;64:272\u2013274. clinician. 2. Olsen MF, et al. J Clin Epidemiol. 2018;101:87\u2013106.e102. 3. Grilo RM, et al. Joint Bone Spine. 2007;74:358\u2013361. Footnotes: Nil. 4. Kelly AM. Acad Emerg Med. 1998;5:1086\u20131090. Ethics approval: Not applicable. 5. Olsen MF, et al. BMC Med. 2017;15:35. Competing interests: Nil. 6. Machado GC, et al. BMJ. 2015;350:h1225. Source(s) of support: Nil. 7. Wandel S, et al. BMJ. 2010;341:c4675. Acknowledgements: We would like to thank Professor Steve 8. Abdel Shaheed C, et al. JAMA Intern Med. 2016;176:958\u2013968. Kamper and Professor Robert Herbert for their input. 9. Hush JM, et al. J Physiother. 2011;57:143\u2013144. Provenance: Not invited. Peer reviewed. 10. Hill JC. et al. Lancet. 2011;378(9802):1560\u20131571. Correspondence: Christina Abdel Shaheed, Institute for Musculo- 11. Pereira TV, et al. JAMA. 2012;308(16):1676\u20131684. skeletal Health, Sydney, Australia. Email: Christina.AbdelShaheed@ 12. Hayes MJ, et al. CMAJ Open. 2018;6:E31\u2013E38. sydney.edu.au 13. Jansen J, et al. BMJ. 2016;353:i2893. 14. Chou R, et al. Ann Intern Med. 2017;166:480\u2013492. 15. Abdel Shaheed C, et al. Med J Aust. 2021;214:324\u2013331."]
Search
