Journal of Physiotherapy 67 (2021) 12–26 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 Multidisciplinary, exercise-based oncology rehabilitation programs improve patient outcomes but their effects on healthcare service-level outcomes remain uncertain: a systematic review Amy M Dennett a,b, Mitchell Sarkies c,d, Nora Shields a, Casey L Peiris a, Cylie Williams d,e, Nicholas F Taylor a,b a School of Allied Health, Human Services and Sport, La Trobe University, Melbourne, Australia; b Allied Health Clinical Research Office, Eastern Health, Melbourne, Australia; c School of Public Health, Curtin University, Perth, Australia; d School of Primary and Allied Health Care, Monash University, Melbourne, Australia; e Department of Allied Health, Peninsula Health, Melbourne, Australia KEY WORDS ABSTRACT Cancer Question: What is the effect of multidisciplinary, exercise-based, group oncology rehabilitation programs on Physical therapy healthcare service outcomes and patient-level outcomes, including quality of life and physical and Exercise psychosocial function? Design: Systematic review with meta-analysis of randomised controlled trials. Multidisciplinary Participants: Adults diagnosed with cancer. Intervention: Multidisciplinary, group-based rehabilitation Rehabilitation that includes exercise for cancer survivors. Outcome measures: Primary outcomes related to health service delivery, including costs, hospitalisations and healthcare service utilisation. Secondary outcomes were patient-level measures, including: the European Organisation for Research and Treatment of Cancer Quality- of-life Questionnaire, 30-second timed sit to stand and the Hospital Anxiety and Depression Scale. The ev- idence was evaluated using the PEDro Scale and the Grades of Research, Assessment, Development and Evaluation (GRADE) approach. Results: Seventeen trials (1,962 participants) were included. There was un- certainty about the effect of multidisciplinary, exercise-based rehabilitation on healthcare service outcomes, as only one trial reported length of stay and reported wide confidence intervals (MD 2.4 days, 95% CI 23.1 to 7.8). Multidisciplinary, exercise-based rehabilitation improved muscle strength (1RM chest press MD 3.6 kg, 95% CI 0.4 to 6.8; 1RM leg press MD 19.5 kg, 95% CI 12.3 to 26.8), functional strength (30-second sit to stand MD 6 repetitions, 95% CI 3 to 9) and reduced depression (MD 20.7 points, 95% CI 21.2 to 20.1) compared to usual care. There was uncertainty whether multidisciplinary rehabilitation programs are more effective when delivered early versus late or more effective than exercise alone. Adherence was typically high (mean weighted average 76% sessions attended) with no major and few minor adverse events reported. Conclusion: Multidisciplinary, exercise-based oncology rehabilitation programs improve some patient-level outcomes compared with usual care. Further evidence from randomised trials to determine their effect at a healthcare service level are required if these programs are to become part of standard care. Trial registration: PROS- PERO CRD42019130593. [Dennett AM, Sarkies M, Shields N, Peiris CL, Williams C, Taylor NF (2021) Multidisciplinary, exercise-based oncology rehabilitation programs improve patient outcomes but their effects on healthcare service-level outcomes remain uncertain: a systematic review. Journal of Phys- iotherapy 67:12–26] © 2020 Published by Elsevier B.V. on behalf of Australian Physiotherapy Association. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). Introduction Oncology rehabilitation aims to address physical and psychological impairments to maintain or restore function, reduce symptom Rehabilitation is an essential element of cancer survivorship care.1 burden, maximise independence and improve quality of life.7 With growing numbers of cancer survivors,2 there is a need to pro- Oncology rehabilitation programs may help manage cancer as a vide interventions that mitigate the adverse effects of cancer treat- chronic disease, which in turn may improve patient and healthcare ments and prevent future disease and cancer recurrence. The most service outcomes. commonly reported issues for cancer survivors are physical problems (such as pain and fatigue) and psychosocial problems (such as fear of Exercise is an important component of oncology rehabilitation. As recurrence).3–5 Cancer survivors also have an increased risk of a structured form of physical activity, exercise improves a variety of developing secondary comorbidities such as cardiovascular disease.6 outcomes for cancer survivors, including fatigue, depression, cardio- respiratory fitness and quality of life.8–10 Supervised exercise https://doi.org/10.1016/j.jphys.2020.12.008 1836-9553/© 2020 Published by Elsevier B.V. on behalf of Australian Physiotherapy Association. This is an open access article under the CC BY-NC-ND license (http:// creativecommons.org/licenses/by-nc-nd/4.0/).
Research 13 programs yield the greatest improvements in these outcomes.8 Method Higher doses of exercise are also associated with greater reductions in cancer-related mortality, cancer recurrence and cardiovascular This systematic review was reported in accordance with the disease risk.11,12 Given these benefits, guidelines support the inte- Preferred Reporting Items for Systematic Reviews and Meta-Analysis gration of exercise into cancer supportive care.9,13–15 (PRISMA) statement.27 Multidisciplinary, oncology rehabilitation programs have been Identification and selection of studies implemented for people with cancer.16 These programs are based on exercise and supplemented by multidisciplinary education and life- Studies were identified through electronic database searches of style counselling. Similar to a cardiac rehabilitation model,17 these MEDLINE, CINAHL, Embase, Sport Discus, PsycINFO and PEDro from programs reduce mortality and readmissions and improve quality of the earliest available time until April 2019. The search strategy used life in people with cardiovascular disease.18 Exercise-based oncology synonyms and MeSH terms focusing on the three key concepts of rehabilitation shares similar goals to cardiac rehabilitation, including: cancer, multidisciplinary rehabilitation and exercise (Appendix 1). reducing symptoms, managing risk factors and improving functional Filters were applied to the search strategy to limit study design to capacity and psychosocial well-being.16,17 Oncology rehabilitation is only include randomised controlled trials.28 Citation tracking of also commonly delivered in groups to provide peer support to par- included trials was performed using Google Scholar and reference ticipants and offer greater access, efficiency and cost savings.19,20 lists of all included papers were checked to supplement database Delivering exercise-based oncology rehabilitation programs for can- searches. cer survivors is feasible,16 and participants describe programs as helping them ‘return to normal’ after cancer diagnosis.19 Two reviewers independently screened the titles and abstracts of each study against predefined inclusion criteria. Any disagreement was Despite the similarities, there are important differences between discussed between the two reviewers until consensus was reached. Full- oncology rehabilitation and cardiac rehabilitation. Unlike cardiac text articles were obtained and screened by two independent reviewers disease, the effects of cancer and its associated treatment may not be to determine final eligibility for study inclusion. Agreement between localised to one physiological system and patients may present with reviewers was assessed using the kappa statistic (k). multiple complex impairments, suggesting that the role of the multidisciplinary team may be even more pertinent.7 Two previous To be included, randomised controlled trials needed to evaluate a reviews have evaluated the effect of oncology rehabilitation programs multidisciplinary, group exercise-based, oncology rehabilitation pro- generally.21,22 Four of the 15 trials described across the two reviews gram for adults with cancer. Comparison groups needed to be usual were delivered in a multidisciplinary, exercise-based group model. care or an alternative intervention (eg, exercise only). Trials could These reviews concluded that multidimensional rehabilitation re- include any healthcare service or patient-level outcome. For the duces fatigue among breast cancer survivors21 and improves quality purpose of this review, healthcare service outcomes included of life among a mixed cohort of cancer survivors.22 No review has healthcare-related costs, hospital length of stay, hospital read- specifically evaluated the effect of multidisciplinary, exercise-based missions, emergency department presentations and healthcare ser- group oncology rehabilitation. vice utilisation such as outpatient visits. Trials were excluded if they were single-discipline programs (eg, support groups with no exercise There are currently no standardised guidelines for the imple- component), unstructured exercise, alternative exercise/psychosocial mentation of multidisciplinary, exercise-based oncology rehabilita- groups (mindfulness meditation, yoga, Pilates, Qi-Gong), single- tion programs, and access to programs is poor.20,23 For wider component programs (eg, exercise only supervised by a physiother- implementation of multi-disciplinary, exercise-based oncology reha- apist and exercise physiologist), individual (1:1) rehabilitation bilitation programs into clinical practice, there is a need to evaluate programs, tele-rehabilitation (unless broadcast simultaneously with a programs at a healthcare service level. Healthcare service outcomes supervised centre-based group), home-based therapy, survivorship may include healthcare-related costs, hospital length of stay, hospital clinics or single assessment only with no group follow-up (Box 1). readmissions, emergency department presentations and healthcare service utilisation such as outpatient visits. Multidisciplinary, Box 1. Inclusion criteria. exercise-based oncology rehabilitation is a resource-intensive and complex intervention. Despite promising results about feasibility of Design exercise-based oncology rehabilitation from pre-to-post and retro- Randomised controlled trial spective studies,16 there is limited evidence from randomised Participants controlled trials on their effect on outcomes such as patient mortality, Adults with cancer readmissions and costs to healthcare services. Only one systematic Intervention review has evaluated the costs of multidimensional, oncology reha- Multidisciplinary ( 2 disciplines) and multicomponent group- bilitation broadly. This review of six studies (of which only one non- randomised study included a multidisciplinary, exercise-based based rehabilitation that includes supervised exercise oncology program) found preliminary evidence of potential cost Exercise intervention with aerobic and/or resistance exercise savings for healthcare services (incremental cost-effectiveness ratios ranged from 2V16,976 to V11,057 per quality-adjusted life year).24 as a core component Costs were up to V793 per patient for a group-based exercise and The rehabilitation program is time-limited with a minimum of 2 psychosocial intervention.25 Another review found conflicting evi- dence of cost-effectiveness of physiotherapy-only interventions weeks’ duration (including exercise) for people with breast cancer.26 Data relating to Outcome measures healthcare service costs of multidisciplinary exercise-based group Healthcare service outcomes (costs, hospital re-admissions, oncology rehabilitation will help policy-makers and healthcare ser- vice providers make important decisions about implementing length of stay, emergency department presentations, health- exercise-based oncology rehabilitation into practice. care service utilisation) Patient-level outcomes (health-related quality of life; physical Therefore, the research questions for this systematic review were: and psychosocial function; symptom and impairment measures; participation measures including physical activity levels) 1. What is the effect of multidisciplinary, exercise-based, group Process measures (adherence, adverse events) oncology rehabilitation programs on healthcare service outcomes? Comparison Oncology rehabilitation compared to usual care 2. What is the effect of multidisciplinary, exercise-based group Oncology rehabilitation compared to equivalent alternative oncology rehabilitation programs on patient-level outcomes, intervention (eg, telehealth intervention, written/digital including quality of life and physical and psychosocial function? information materials, exercise only) Comparing two different models of cancer rehabilitation (eg, at two times or at two intensities)
14 Dennett et al: Multidisciplinary, exercise-based oncology rehabilitation programs Articles found via search of databases Articles found via reference lists or citation tracking (n = 1,882) (n = 3) Records after duplicates removed (n = 1,510) Titles and abstracts screened (n = 1,510) Excluded based on title and abstract (n = 1,445) Full-text articles assessed for eligibility (n = 65) Excluded after evaluation of full text (n = 39) • not a randomised trial (n = 5) • exercise/physiotherapy only (n = 9) • 1:1 intervention (n = 14) • not supervised exercise-based (n = 5) • wrong type of exercise (n = 1) • evaluates component of rehabilitation (n = 3) • inadequate control (n = 2) Duplicate trial data (n = 9) Trials included in review (n = 17) Inadequate data for meta-analysis (n = 7) Trials included in meta-analysis (n = 10) Figure 1. Flow of trials through the review. Assessment of methodological quality Meta-analysis was performed using a random-effects model with clinically homogenous data using RevMan software.32 If more than Two reviewers independently used the 11-item Physiotherapy low levels of statistical heterogeneity were observed (I2 . 25%), Evidence Database (PEDro) scale to assess methodological quality. sensitivity analyses were completed excluding trials hypothesised to The PEDro scale comprises one item evaluating external validity and be contributing to heterogeneity. The Grades of Research, Assess- 10 items evaluating internal validity. The PEDro score demonstrates ment, Development and Evaluation (GRADE) approach was used to convergent and construct validity29 and has moderate levels of inter- determine the overall quality of evidence (between ‘high’ and ‘very rater reliability (ICC 0.68, 95% CI 0.57 to 0.76).30 A score of 8 on the low’) presented in each meta-analysis. Evidence was downgraded PEDro scale is considered the highest possible score, as participants based on predetermined criteria by one level if: PEDro score was , 6 and therapists in exercise trials are unable to be blinded. for the majority of trials used in the meta-analysis;33 there was greater than low levels of statistical heterogeneity between trials Data analysis (I2 . 25%);34 and there were large confidence intervals (exceeding the minimum clinically important difference). Results were summarised To compare groups, mean differences (MD) were calculated from descriptively when data could not be pooled for meta-analysis. post-intervention means and standard deviations (SD) for homoge- nous outcomes. If change scores were presented, post-intervention Results means were calculated in reference to the baseline mean and the post-intervention SD was imputed using baseline data.31 Where there Study selection were insufficient data for analysis or inconsistency in data, authors were contacted to seek further information. Database searches identified 1,882 articles for screening. An additional two articles were identified by citation tracking and one
Research 15 Table 1 PEDro scores of the included trials. Trial Random Concealed Groups similar Participant Therapist Assessor , 15% Intention-to- Between-group Point estimates Total comparisons reported and variability (0 to allocation allocation at baseline blinding blinding binding dropout treat analysis 10) Y Y Adamsen Y Y Y N N NY Y Y 7 N 2009 42 Y N Bergland Y N Y N N NY N Y Y5 Y 1994 43 N Y Bergland Y N Y N N NN N Y Y3 Y 2007 40 Y Y Bourke Y Y Y N N YY Y Y Y8 Y 2011 41 Y Y Cho YN Y N N NN N Y3 2006 47 Clark YN Y N N NY N Y5 2013 44 Ghavarmi Y Y Y N N NY N Y6 2017 48 Hubbard Y N N N N NN N N1 2016 52 Korstjens Y N Y N N NY Y Y6 2008 45 Midtgaard Y N Y N N NN Y Y5 2013 36 O’Neill Y Y N N N YY N Y6 2018 37 Quist YY Y N N YN Y Y7 2018 38 Rummans Y N Y N N NY N Y5 2006 46 Sandmael Y N Y N N NN N Y4 2017 49 Sheppard Y N Y N N NN N Y4 2016 50 Spahn Y Y N N N NY Y Y6 2013 51 Uster YY Y N N YY Y Y8 2017 39 Y = Yes, N = No. article through checking reference lists. The total yield was reduced to Intervention 1,510 after duplicates were removed. Sixty-five full-text articles were Rehabilitation was delivered in outpatient hospital settings. Group screened and reduced to 26 articles after assessment against inclu- sion criteria. The inter-rater agreement between two reviewers for sizes ranged from 3 to 25 participants and were often led by full text was ‘very good’ (k = 0.835, 95% CI 0.697 to 0.972).35 The 26 physiotherapists36–40,42,44–46,49,52 (11 trials) and nurses38,40–47,52 (10 included articles reported data from 17 randomised controlled trials trials). Dietitians (six trials)37,39,43,47,49,50 also contributed to pro- (Figure 1). Where duplicate trial data were presented, the first article grams. Programs were supervised twice-weekly for 4 weeks to 1 year, published with participant outcome data was considered the primary with most being 10 to 12 weeks in duration. Nine trials included both trial throughout (Appendix 2). Four authors responded to requests for aerobic and resistance training.36–39,41–43,45,52 Exercise was described missing data.36–39 as individually tailored in 12 trials.36–39,41–43,45,47–49,52 Exercise was supplemented by multidisciplinary information sessions (12 Methodological quality trials),37,38,40,41,43,44,46,47,49–52 relaxation (seven trials)40,42–44,46,51,52 and individual nutrition counselling (six trials) (Table 3).37–39,48,49,51 The quality of trials included in this review was moderate, with a Four trials included standardisation processes within the rehabilita- mean score of 5 (range 1 to 8) on the PEDro scale (Table 1). Inter-rater tion program comprising staff training44–46 and written mate- agreement on quality criteria was ‘very good’ (k = 0.887, 95% CI 0.832 rials.44,46,50 Two trials compared multidisciplinary rehabilitation with to 0.941).35 Seven trials had concealed allocation (41%) and four trials supervised, group-based exercise45,51 and two trials compared early (24%) had blinded assessors. rehabilitation with late rehabilitation.38,49 The other 13 trials compared multidisciplinary exercise-based cancer rehabilitation with Study characteristics usual care (Table 2). Participants Adverse events Trials included 1,962 participants. The mean age of participants Six trials reported on adverse events.36–39,45,51 No serious adverse ranged from 4636 to 69 years40,41 and most were women (Table 2). events were attributed to the rehabilitation interventions in these six Seven trials included participants with a mix of cancer types36,39,42–46 trials. In one trial,36 six participants in the rehabilitation group (of 214 and nine trials included people with early-stage cancer total) developed lymphoedema. This did not limit group participa- only.37,38,43,45,47–51 Four trials evaluated rehabilitation exclusively in tion. Exacerbation of pre-existing musculoskeletal discomfort was an advanced cancer cohort.36,39,44,46 Trials were usually conducted reported by seven of 107 participants across two trials.37,51 No other after cancer treatment completion, with six trials including partici- adverse events attributed to the intervention were reported in the pants during treatment39,42,44,46,49,52 and one trial including partici- remaining three trials. One trial comparing prehabilitation plus early pants pre-surgery.38 Only four trials reported baseline physical rehabilitation with prehabilitation plus late rehabilitation found an performance status,39,44,46,49 which was described as fully active or overall postoperative pulmonary complication rate of 23% within 30 restricted, in only strenuous activities for the majority of participants. days after surgery, with complications highest in the early
16 Dennett et al: Multidisciplinary, exercise-based oncology rehabilitation programs Table 2 Characteristics of the included trials. Trial Participants Intervention Outcome measures Country Adamsen n = 269 Exp = Oncology rehab, 4/wk, 6 wks Muscle strength = 1RM 2009 42 Fitness = VO2 max (stationary cycle) Age (yr) = 48 (11) Con = Usual care: allowed to freely increase activity Fatigue = FACT An Denmark Physical Activity = leisure time PA Gender = 196 F, 73 M self-report Cancer type = mixed QoL = EORTC QLQ-C30, MOS SF-36 Follow-up = 0, 6 wks Cancer stage = mixed Time since diagnosis = 12 wks Cancer treatment = during chemotherapy Baseline performance status = NR Bergland n = 199 Exp = Oncology rehab, 11 sessions, 7 wks Strength and activity = self-reported 1994 43 Age (yr) = Exp 53 (NR), Con 54 (NR) Con = Usual care: n = 65 had no intervention, n = 36 offered a session with Psychological = Modified HADS, Gender = 192 F, 7 M dietitian and oncologist Sweden Cancer type = mixed Mental Adjustment to Cancer Scale Cancer stage = early Symptoms = Breast cancer Time since diagnosis = within 8 wks Cancer treatment = post-treatment symptoms Baseline performance status = NR Activity = Activities of daily living QoL = Global Quality of life Follow-up = 0, 3, 6, 12 months Bergland n = 211 Exp = Oncology rehab, 7 sessions, 7 wks Psychological = HADS 2007 40 Age (yr) = 69 (12) Gender = 0 F, 211 M Con = Usual care: included two information leaflets and opportunity to phone QoL = EORTC QLQ-C30 Sweden Cancer type = prostate Cancer stage = mixed nurse with questions Follow-up = 0, 6, 12 months Time since diagnosis = within 6 months Cancer treatment = watchful waiting, during and post-treatmenta Baseline performance status = NR Bourke n = 18 Exp = Oncology rehab, 1 to 2/wk, 12 wks Fatigue = FACT-F 2011 41 Age (yr) = 69 (8) Con = Usual care: included nurse-led follow-up Fitness = VO2 peak (Bruce protocol) Gender = 6 F, 12 M Muscle strength = knee extensor UK Cancer type = colon Cancer stage = mixed strength, isometric muscle fatigue Time since diagnosis = 6 to 24 months Functional activity = 30-s STS Cancer treatment = post-treatment Physical Activity = Godin Leisure- Baseline performance status = NR time Index Cho 2006 47 n = 65 Exp = Oncology rehab, 2/wk, 10 wks QoL = FACT-C South Korea Age (yr) = 49 (NR) Con = Usual care: no group rehabilitation Follow-up = 0, 12 wks Gender = 65 F, 0 M ROM = Shoulder Cancer type = breast Psychological = Psychosocial Cancer stage = early Time since diagnosis = 15 months Adjustment Score Cancer treatment = post-treatment QoL = Quality of Life Baseline performance status = NR Follow-up = 0, 10 wks Clark n = 131 Exp = Oncology rehab, up to 4 wks Psychological = POMS 2013 44 Sleep = PSQI Age (yr) = 59 (NR) Con = Usual care: including medical appointments and referrals to specialists Physical Activity = Exercise USA Gender = 45 F, 86 M when needed behaviour questionnaire QoL = FACIT-Spiritual Wellbeing, Cancer type = mixed FACT-G, Caregiver QoL Cancer stage = advanced questionnaire Follow-up = 0, 4, 27 wks Time since diagnosis = within 12 months Cancer treatment = during radiotherapy Baseline performance status = ECOG 0 (n = 64), 1 (n = 67) Ghavarmi n = 80 Exp = Oncology rehab, 5/wk, 24 wks Body Composition = BMI 2017 48 Age (yr) = 49 (9) Con = Usual care Fatigue = CFS Gender = 80 F, 0 M Sleep = PSQI Iran Cancer type = breast QOL = EORTC QLQ-C30, EORTC-BR23 Cancer stage = early Follow-up = 0, 24 wks Time since diagnosis = 3 to 18 months post-treatment Cancer treatment = post-treatment Baseline performance status = NR Hubbard n = 41 Exp = Cardiac rehab, 1 to 2/wk, 6 to 12 wks Adverse events 2016 52 Age (yr) = NR Con = Usual care: given booklet Staying healthy after bowel cancer Intervention adherence and Gender = NR UK Cancer type = colorectal attendance Cancer stage = NR Staff and patient experience Time since diagnosis = NR Follow-up = post-intervention, time Cancer treatment = post-surgery Baseline performance status = NR period not stated
Research 17 Table 2 (Continued) Trial Participants Intervention Outcome measures Country Korstjens n = 147 Exp= Oncology rehab (physiotherapy, CBT), 2/wk, 12 wks Cognition = SPSI-R Con = Active: physiotherapy only Psychological = HADS 2008 45 Age (yr) = 49 (11) Fitness = VO2 max (Cycle ergometer) Strength = MVC (Elbow and knee) Netherlands Gender = 123 F, 24 M hand held dynamometer Cancer type = mixed Fatigue = MFI Physical Activity = PASE Cancer stage = early QoL = MOS SF-36, EORTC QLQ-C30 Follow-up = 0, 3, 6, 12 months Time since diagnosis = 1-year post- treatment Cancer treatment = post-treatment Baseline performance status = NR Midtgaard n = 214 Exp = Oncology rehab, 1/wk, 52 wks Fitness = VO2 max (cycle ergometer) 2013 36 Age (yr) = Exp 48 (10), Con 46 (12) Con = Usual care: feedback and general exercise advice after fitness testing at Muscle strength = 1RM chest and Gender = 178 F, 36 M each assessment Denmark Cancer type = mixed leg press Cancer stage = advanced Fatigue = EORTC-QLQ C30 Time since diagnosis = 80 d post- Psychological = HADS treatment Physical activity = Saltin and Grimby Cancer treatment = post chemotherapy Baseline performance status = NR Questionnaire QOL = MOS SF-36, EORTC QLQ-C30 Follow-up 0, 6, 12 months O’Neill n = 43 Exp = Oncology rehab, 2/wk, 12 wks Fitness: VO2 max (cycle ergometer) 2018 37 Body composition = Anthropometric Age (yr) = 66 (9) Con = Usual care Ireland measurements Gender = 8 F, 35 M Physical activity = Actigraph GT3X1 QoL = EORTC QLQ-C30 Cancer type = upper gastrointestinal Follow-up = 0, 3, 6 months Cancer stage = early Time since diagnosis = 45 d post-surgery Cancer treatment = post-treatment Baseline performance status = NR Quist n = 235 Exp = Early rehabilitation: unsupervised training 2 wks preop and supervised Fitness = VO2 max (cycle ergometer) 2018 38 Psychological = HADS, Distress Age (yr) = early 66 (NR), late 65 (NR) training 2 wks postop, 2/wk, 12 wks Denmark thermometer Gender = 117 F, 118 M Con = Active: late rehabilitation: started 6 wks post-op after 2 wks pre-op Lung capacity = FEV1 Muscle strength = 1RM chest and Cancer type = lung unsupervised training leg press Cancer stage = early Functional activity = 6MWT Physical Activity = Physical Activity Time since diagnosis = 2 to 14 wks post- Scale surgery QoL = MOS SF-36, FACT-L, EORTC Cancer treatment = post-surgery QLQ-C30, MSPSS Follow-up= 0, 14, 26, 52 wks Baseline performance status = NR Rummans n = 115 Exp = Oncology rehab, 8 sessions, 4 wks Psychological = POMS-SF 2006 46 Age (yr) = 60 (11) Gender = 37 F, 78 M Con = Usual care: usual outpatient appointments and any other support sought QoL = Spitzer QoL, LASA QoL, FACIT- USA Cancer type = mixed Cancer stage = advanced by the patient Spiritual wellbeing Time since diagnosis = NR Cancer treatment = during radiotherapy Follow-up = 0, 4, 8, 27 wks Baseline performance status = ECOG: 0 (n = 33), 1 (n = 66), 2 (n = 4) Sandmael n = 41 Exp = Oncology rehab during first wk of radiotherapy, 2/wk, 6 wks Body composition = CT 2017 49 Age (yr) = 63 (9) Con = Active: oncology rehab, 2 to 4 wks at end of radiotherapy, 3/wk and 2 Physical Activity = HUNT PA-Q Gender = 16 F, 25 M voluntary sessions, 3 wk Follow-up = 0, 6, 14 wks Norway Cancer type = head and neck Cancer stage = early Time since diagnosis = 2 wks Cancer treatment = during radiotherapy Baseline performance status = KPS 90 (n = 33) Sheppard n = 31 Exp = Oncology rehab, 1/wk, 12 wks Fitness = VO2 peak (treadmill) 2016 50 Age (yr) = 55 (10) Con = Usual care: patients given the NCI booklet Facing Forward Life After Cancer Body composition USA Gender = 31 F, 0 M Treatment Dietary intake Cancer type = breast Physical activity = IPAQ-SF Cancer stage = early Follow-up = 0, 12 wks Time since diagnosis 6 months post but , 5 years Cancer treatment = post-treatment Baseline performance status = NR Spahn n = 64 Exp = Oncology rehab, 1/wk, supervised walking 3/wk, 10 wks Fatigue = VAS 0-100 mm scale, MFI 2013 51 Age (yr) = Exp 58 (9), Con 55 (11) Con = Active: supervised walking training only, 30 min, 3/wk, 10 wks Psychological = HADS Gender = 64 F, 0 M Symptoms = MRS Germany Cancer type = breast QoL = EORTC QLQ-C30 Cancer stage = early Follow-up = 0, 10, 22 wks Time since diagnosis = 50 months Cancer treatment = post-treatment Baseline performance status = NR
18 Dennett et al: Multidisciplinary, exercise-based oncology rehabilitation programs Table 2 (Continued) Trial Participants Intervention Outcome measures Country Uster n = 58 Exp = Oncology rehab, 2/wk, 12 wks Length of stay of unplanned 2017 39 Age (yr) = 63 (10) Con = Usual care: no exercise intervention offered to patients. Request to admissions Switzerland Gender = 18 F, 40 M maintain physical activity level. Nutritional support by a dietician as needed. 3-month survival Dietary intake Cancer type = mixed Strength = Grip, 1RM leg press Body composition Cancer stage = advanced Functional activity = 6MWT, 30-s Time since diagnosis = 12 months STS QoL = EORTC QLQ-C30 Cancer treatment = during and post- Follow-up = 0, 3, 6 months treatmenta Baseline performance status = ECOG: 0 (n = 4), 1 (n = 35), 2 (n = 16), 3 (n = 3) Age is mean (SD). BMI = body mass index, CFS = Cancer Fatigue Scale, Con = control group, CT = computed tomography, ECOG = Eastern Cooperative Oncology Group Performance Status, EORTC BR23 = European Organisation for Research and Treatment of Cancer Quality-of-life Questionnaire Breast, EORTC QLQ-C30 = European Organisation for Research and Treatment of Cancer Quality-of-life Questionnaire Core 30, Exp = experimental group, FACT An = Functional Assessment of Cancer Therapy - Anaemia, FACT-C = Functional Assessment of Cancer Therapy - Colorectal Cancer, FACT-F = Functional Assessment of Cancer Therapy - Fatigue, FACT-G = Functional Assessment of Cancer Therapy - General, FACT-L = Functional Assessment of Cancer Therapy - Lung, FACIT = Functional Assessment of Chronic Illness Therapy, FEV = forced expiratory volume, HADS = Hospital Anxiety and Depression Scale, HUNT PAQ = Nord Trondelag Health Study Questionnaire, IPAQ-SF = International Physical Activity Questionnaire Short Form, KPS = Karnofsky Performance Status, LASA = Linear Analogue Scales of Assessment, MFI = Multidimensional Fatigue Inventory, MOS-SF 36 = Medical Outcomes Study Short Form 36, MRS = Menopausal Rating Scale, MSPSS = Multidimensional Scale Of Perceived Social Support, MVC = maximum voluntary contraction, NR = not reported, PA = physical activity, PASE = Physical Activity Scale for the Elderly, POMS = Profile of Mood States, PSQI = Pittsburgh Sleep Quality Index, QOL = quality of life, RM = repetition maximum, SPSI-R = Social Problem Solving Inventory, STS = sit to stand, VAS = visual analogue scale, 6MWT = 6-minute walk test. a Studies included participants who were receiving treatment or had completed treatment at the time of trial. In the trial by Uster et al 2017, two participants received radiotherapy, 22 received chemotherapy. rehabilitation group (33% compared with 14%).38 The prevalence of cancer were removed from analysis,36,39,42 exercise-based oncology cardiac complications was 13%, with no difference between groups.38 rehabilitation compared with usual care improved global quality of Adherence life (MD 26, 95% CI 9 to 43, I2 = 86%) and insomnia (MD 239 points, 95% CI 246 to 232, I2 = 0%). When two trials implementing long-term Thirteen trials reported adherence to the rehabilitation programs. interventions36,48 (6 and 12 months, respectively) were removed Adherence varied between 67%36 and 94%,37,49 with a mean weighted average of 76% of sessions attended. In one trial that included pre- from analysis, global quality of life may have improved compared habilitation for lung cancer surgery, six participants allocated to with usual care (MD 9 points, 95% CI 1 to 16, I2 = 51%). prehabilitation did not receive the intervention due to ‘lack of time’ (mean 8, SD 3 days to surgery) and one participant chose not to Two short-term interventions (4 weeks) both completed in attend due to low motivation.38 Only 25% of participants accom- plished daily preoperative exercise and there was high dropout advanced cancer populations that were not included in meta-analysis among both early (39%) and late (42%) rehabilitation groups.38 due to heterogeneous outcomes found short-term improvements in Effects of multidisciplinary exercise-based cancer rehabilitation quality of life that were not sustained at 6 months.44,46 Another trial compared to usual care not included in meta-analysis comprising early and advanced colo- Healthcare service outcomes One trial evaluated the effect of multidisciplinary, exercise-based rectal cancer survivors found a clinically significant effect in favour of rehabilitation for improved quality of life using the FACT-C (MD 14 oncology rehabilitation on hospital length of stay for unplanned ad- points, 95% CI 2 to 27) (Appendix 3).41 missions.39 The mean between-group difference was 2.4 days in favour of the intervention group (95% CI 23.1 to 7.8). There were no significant differences between multidisciplinary, exercise-based oncology rehabilitation and usual care for other Patient-level outcomes Quality of life: Meta-analysis of five trials,36,37,39,42,48 including 557 quality of life outcomes measured by the EORTC QLQ-C30 (Table 4, participants, found low-quality evidence of a medium-sized, clinically significant observed mean effect for exercise-based oncology reha- Appendix 4). bilitation improving global quality of life measured by the European Physical outcomes: Meta-analysis of two trials36,42 including 388 Organisation for Research and Treatment of Cancer Quality-of-life Questionnaire Core 30 (EORTC QLQ-C30) (MD 14 points, 95% CI 21 participants showed high-quality evidence of improvement in upper to 28; I2 = 95%) when compared with usual care; however, the lower bound of the confidence interval indicated uncertainty (Table 4, body muscle strength measured by 1RM chest press (MD 3.6 kg, 95% Figure 2). For a more detailed forest plot, see Figure 3 on the eAd- CI 0.4 to 6.8, I2 = 18%) and moderate-quality evidence of improvement denda. Meta-analysis of two trials with longer term follow-up also demonstrated uncertainty of a clinically significant between-group in lower body muscle strength measured by 1RM leg press (MD 19.5 effect at the 3-month follow-up (Table 4). kg, 95% CI 12.3 to 26.8, I2 = 0%) when comparing exercise-based Sensitivity analysis was completed for quality of life due to high I2 oncology rehabilitation with usual care (Figure 4). Meta-analysis of values. Removal of trials including any participants receiving cancer two trials39,41 including 55 participants found low-quality evidence of treatment from analysis39,42 indicated uncertainty as to whether there were improvements in any quality of life-related outcome for improvement in functional strength measured with the 30-second sit the effect of exercise-based oncology rehabilitation compared with to stand test (MD 6 repetitions, 95% CI 3 to 9, I2 = 0%) when comparing usual care (global quality of life subscale MD 18 points, 95% CI 24 to 40, I2 = 97%). When trials including any participants with advanced exercise-based oncology rehabilitation with usual care (Figure 4). For a more detailed forest plot, see Figure 5 on the eAddenda. No dif- ferences between multidisciplinary, exercise-based oncology reha- bilitation and usual care were found for fatigue, body composition or cardiorespiratory fitness (Table 4). Psychological outcomes: Meta-analysis of two trials36,43 including 465 participants found moderate-quality evidence of small reductions in depression (MD 20.7 points, 95% CI 21.2 to 20.1, I2 = 0%) but not anxiety as measured by the Hospital Anxiety and Depression Scale after exercise-based oncology rehabilitation when compared with usual care (Figure 6). For a more detailed forest plot, see Figure 7 on the eAddenda. Additional outcomes relating to quality of life, physical activity levels, physical impairments and psychological status were not included in meta-analysis due to heterogeneity and/or insuffi- cient data (Appendix 3).
Table 3 Characteristics of the multidisciplinary oncology rehabilitation programs. Trial Supervised exercise FITT Non-exercise comp Country core Relaxation (30 min, 4/wk), massage (30 Adamsen 2009 42 F: 3/wk min, 2/wk), body awareness and Hospital I: high restoration (90 min, 1/wk) T: RT, AT, balance Information session (1/wk, 4 wks), Bergland 1994 43 T: 90 min coping training (1/wk, 3 wks) NR F: 1/wk (4 wks) Information session (60 min/session) I: NR T: AT, RT, mobility Information sessions (15 to 30 mins, 1/ T: 120 min fortnight) Bergland 2007 40 F: 1/wk Information session (90 min, 1/wk), Hospital I: NR support group (60 min, 1/wk) T: Pelvic floor, fitness training, relaxation, Bourke 2011 41 breathing Hospital T: 60 min Cho 2006 47 F: 2/wk 3 2 wks, 1/wk 3 6 wks Cancer centre (tertiary I: moderate to high hospital) T: AT, RT T: 30 mins F: 2/wk I: low to moderate T: AT, flexibility T: 90 min Clark 2013 44 F: NR Information session (60 min/session), Cancer centre (tertiary I: NR relaxation (15 min/session) hospital) T: conditioning T: 20 min Ghavarmi 2017 48 F: 3 to 5/wk Individual diet advice (1/wk) Cancer centre I: high Relaxation, information session (1/wk) T: AT Hubbard 2016 52 T: 60 min Hospital or sports centre F: 1 to 2/wk I: moderate to high T: AT, RT, flexibility T: 75 to 90 min
ponents Disciplines involved Group Program Information topics size additional Nurse specialist duration sessions Physiotherapist 7 to 10 6 wk 4 d/wk Take-home info on Nurse specialist 7 wk 11 sessions Side effects progressive muscle Physical trainer 7 wk Crisis and emotional reactions relaxation Oncologist Diet Psychologist Alternative treatments Dietitian Return to work Anxiety Handling situations and rights Booster session 2 months Physiotherapist 3 to 10 7 sessions Prostate cancer after conclusion of program Nurse Side effects Urologist Crisis and emotional reactions Incontinence and sexual function HEP 2/wk in last 6 wks, Nurse 12 wk 1 to 2 d/wk Healthy eating nutrition info pack Exercise physiologist HEP booklets (wk 5 to 10) Nurse 15 10 wk 2 d/wk Understanding breast cancer Research 19 with education content and Surgeon Treatment and complications 2 elastic balls Dietitian up to 4 Nutrition and diet Image consultant wk Sexual and daily life Physiatrist Prevention and management Exercise prescription 24 wk Lymphoedema Fitness instructor Breast reconstruction Image management Relationships/communication Latest cancer treatment Telephone counselling after Psychologist 4 6 sessions Health behaviour change structured intervention (10 Physical Radiation and chemotherapy calls), manual Therapist Mood and coping Physiatrist Spirituality Nurse Problem solving Social worker Community resources Chaplain Advanced care planning Record keeping Physical activity Defining quality of life Communication strategies Diet diaries, written Researcher 15 5 d/wk Healthy eating information Exercise coach HEP written info Physiotherapist 15 to 25 6 to 12 wk 1 or 2 d/wk Diet Nurse Physical activity Physiotherapy Relaxation/stress management assistant Cardiac-specific information sessions
Table 3 (Continued) Supervised exercise FITT Non-exercise comp core Trial F: 2/wk CBT (120 min, 1/wk) Country I: moderate to high T: AT, RT, sports Korstjens 2008 45 T: 120 min University medical centre, hospital, rehab centre Midtgaard 2013 36 F: 1/wk Counselling (120 min, 3 individual Hospital I: high sessions, tri-monthly; 6 group sessions, T: AT, RT bi-monthly) O’Neill 2018 37 T: 90 min Cancer centre Individual diet counselling, information F: 1 to 2/wk sessions (7 sessions) Quist 2018 38 I: low to moderate Hospital T: AT, RT Group education (3 sessions), individual T: 20 to 35 min counselling (3 sessions) Rummans 2006 46 Cancer centre F: 2/wk Information sessions (60 min/session), I: moderate to high relaxation (10 to 20 min/session) Sandmael 2017 49 T: AT, RT Hospital T: 60 min Nutrition counselling, (during group, 1 session, post group 1/wk), information Sheppard 2016 50 F: NR session (1/wk) Hospital I: NR T: conditioning Information session (60 min, 1/fortnight), Spahn 2013 51 T: 20 min MI phone coaching, (on wks in between, Hospital 15 min) F: 2/wk I: moderate Information sessions, guided discussion, T: RT, AT stress, nutrition, relaxation, T: 30 to 35 min hydrotherapy, mindfulness (1 3/wk, 6 hrs) F: 1/fortnight I: moderate T: NR T: 30 min F: 3 in 10 wks I: high T: AT T: 30 min Uster 2017 39 F: 2/wk Nutrition counselling (3 sessions) Hospital I: moderate to high T: AT, RT, balance T: 60 min AT = aerobic training, CBT = Cognitive Behavioural Therapy, FITT = frequency, intensity, type, time (per session), HEP = home
ponents Disciplines involved Group Program Information topics 20 Dennett et al: Multidisciplinary, exercise-based oncology rehabilitation programs size additional Psychologist duration sessions Experience with cancer Nurse 8 to 12 Stress, relaxation and fatigue CBT homework (30 min) Physiotherapist 12 wk 2 d/wk Exercise physiology Social worker 20 Illness perceptions Optimism Self-management Goal setting Decision-making Psychologist 52 wk 1 d/wk Physiotherapist Exercise physiologist HEP (1/ wk in wks 1 to 2, Surgeon 12 wk 2 d/wk Mindfulness then 5/wk in wks 10 to 12) Dietitian Physiotherapist 12 wk 2 d/wk Health promoting behaviour Preop HEP, smoking Occupational cessation, diet advice as therapist needed Psychotherapist Physiotherapist Nurse Psychologist 4 wk 8 sessions Symptom management Physical therapist Spiritual guidance Physiatrist 6 wk Financial resources Nurse 12 wk Advanced care planning Social worker CBT for coping Chaplain Open discussion and support Oral supplements Physiotherapist 2 d/wk Cancer treatment side effects Dietitian Physical activity Mental health Pedometers and step goal, Exercise Fortnightly Nutrition education food diary, manual to store Physiologist resources Nutritionist Walking HEP (30 min, 3/wk) Sports therapist 10 wk 1 d/wk Psyche and pain Other disciplines NR 12 wk 2 d/wk Relaxation response Stress management Cognitive restructuring Yoga Nutrition Social competence Coping with disease Oral supplements (post Physiotherapist exercise) Dietitian e exercise program, MI = motivational interviewing; NR = not reported, RT = resistance training.
Research 21 Effects of multidisciplinary exercise-based cancer rehabilitation rehabilitation than for other chronic diseases. Unlike cardiac or pul- compared with supervised group exercise programs monary disease, cancer affects multiple systems and patients may not present as an inpatient at initial diagnosis or at any stage throughout Meta-analysis of two trials45,51 including 191 participants found their treatment continuum. Therefore, in evaluating cost- low-to-moderate-quality evidence of no difference between multi- effectiveness, other healthcare service outcomes relating to eco- disciplinary, exercise-based oncology rehabilitation and supervised nomic benefits – such as return to work, medication adherence and group exercise for improving quality of life (MD 2.1 points, 95% CI 28.8 outpatient healthcare service utilisation – also need to be considered to 13.0) or reducing fatigue (Multidimensional Fatigue Inventory - in addition to hospitalisations. General Fatigue MD 20.2 points, 95% CI 21.2 to 0.7) or depression (MD 20.6 points, 95% CI 22.9 to 1.6) and anxiety (MD 20.4 points, 95% There was uncertainty about whether multidisciplinary, exercise- CI 23.3 to 2.4) immediately post-intervention (Appendix 4). based oncology rehabilitation has a positive impact on cancer survivors’ quality of life. Improvements in global quality of life Effect of timing of rehabilitation demonstrated a mean 14-point improvement on the EORTC QLQ-C30, indicating a clinically meaningful improvement, but this estimate One trial38 compared early and delayed rehabilitation after lung came with substantial uncertainty.54–56 However, there was more cancer surgery. A 12-week supervised exercise program was initiated confidence in the results when only short-term interventions and 2 or 6 weeks after surgery. There was less reduction in FEV1 in the people with early-stage cancer were included. Additionally, studies early rehabilitation group compared with late rehabilitation at week included in the review but not included in the meta-analyses also 26 (MD 0.15 L, 95% CI 0.05 to 0.25) and week 52 (MD 0.1 L, 95% CI 0 to reported positive effects on quality of life. This may be explained by 0.25). Due to wide confidence intervals there was uncertainty about the nature of living with cancer as a chronic disease and the transient whether there were any other between-group differences at week 26 effect that interventions such as rehabilitation may have on outcomes. (primary outcome VO2 peak MD 23 ml/minute, 95% CI 288 to 82).38 Participants also had reduced levels of depression and improved their In another trial49 comparing rehabilitation during and after radio- muscle strength, which may have contributed to improved overall therapy in people with early head and neck cancer, there was un- quality of life by enhancing their ability to participate in functional certainty due to the relatively wide confidence intervals about the activity. Cancer treatment negatively affects quality of life among magnitude of the between-group differences with respect to body people with cancer and they report high levels of unmet physical and composition at week 14 (muscle mass MD 2.9 cm2/m2, 95% CI 24.5 to psychosocial need.55,57 Oncology rehabilitation may mitigate these 10.3; body weight MD 8 kg, 95% CI 25 to 21). negative effects. Our results provide further explanation as to the ef- fect of multidisciplinary rehabilitation on quality of life compared with Discussion previous reviews 21,22 and the findings are consistent with a qualitative analysis that exercise-based oncology rehabilitation helps people ‘re- The main findings of this review of multidisciplinary, exercise- turn to normal’ after cancer diagnosis.19 A key difference between this based oncology rehabilitation programs were: there was uncer- review and those previously published is that all trials included in this tainty about the impact of such programs on healthcare service review implemented supervised group-based exercise. Supervision of outcomes from a single trial on length of stay, due to wide confidence exercise increases the likelihood of benefits related to quality of life.8 intervals; there was high-quality evidence of improved muscle This may be due to the additional benefits of attention from thera- strength, improved functional strength and reduced depression in a pists, including guided exercise progression, and the potential for mixed cohort of cancer survivors compared with usual care but un- increased adherence to the exercise protocol resulting in the certainty about improvements in short-term global quality of life completion of a greater volume of exercise.8 compared with usual care; and typically, programs were well atten- ded and safe. In addition, evidence from two trials demonstrated Surprisingly, multidisciplinary, exercise-based oncology rehabili- uncertainty about whether multidisciplinary, exercise-based tation did not improve cardiorespiratory fitness or reduce cancer- oncology rehabilitation may be more or less effective than super- related fatigue. These are two important endpoints of oncology vised group exercise rehabilitation alone in cancer survivors who rehabilitation known to improve with exercise alone.10,58 Our results have completed treatment. There is also uncertainty whether early may be related to issues with exercise prescription. In order for rehabilitation after surgery for lung cancer or during radiotherapy for physiological adaptations to occur to induce change in cardiopul- head and neck cancer is more or less effective than delayed monary systems, exercise must be prescribed at the appropriate rehabilitation. dosage in accordance with FITT (frequency, intensity, type and timing) training principles, with attention paid to exercise specificity This review highlights the lack of data relating to outcomes that and progressive overload. Moderate-intensity aerobic exercise im- are meaningful to healthcare services. While recommendations have proves cardiorespiratory fitness and fatigue levels of cancer survi- been made to integrate exercise into standard cancer care,14 this vors.10,33,58 There was poor reporting of exercise interventions in this cannot be achieved without evidence of the impact of exercise-based review’s included trials, which was consistent with other exercise rehabilitation programs at healthcare system level, particularly in trials on cancer.59 Moreover, supervised exercise was mostly public healthcare systems where resources are limited. The multi- completed once or twice a week, with only half of the included trials disciplinary, exercise-based oncology rehabilitation model is a com- including specific details related to aerobic exercise intensity, varying plex and resource-intensive intervention. Cardiac and pulmonary from low-to-moderate to high-intensity aerobic exercise. To be able rehabilitation interventions have extensive evidence relating to ser- to translate interventions into practice, adequate reporting of trials is vice outcomes, which has assisted the integration of such programs important for clinicians who practise in exercise-based oncology into standard care pathways. A review of 14,486 cardiac rehabilitation rehabilitation. Clinicians currently have difficulty prescribing exercise participants demonstrated reduced risk of hospital admission and in oncology rehabilitation programs to the appropriate intensity reduced cardiovascular mortality, with four trials demonstrating cost required to improve outcomes.20 Therefore, exercise-based oncology savings.18 Similarly, in a review of pulmonary rehabilitation, re- rehabilitation interventions used in trials need to be reported using ductions in hospital admissions and mortality were also estab- recognised checklists such as TIDieR60 and described appropriately lished.53 In the present review, only one trial evaluated a healthcare using FITT principles, so that clinicians can provide effective exercise service outcome. Only two reviews have previously included prescription in practice. healthcare service outcomes by evaluating costs of oncology reha- bilitation more broadly; data specifically relating to exercise-based Another possible explanation for the absence of between-group interventions are inconclusive.24,26 It should also be considered that differences in fatigue may be related to timing and baseline status demonstrating cost-effectiveness may be more difficult for oncology of participants. Participants in this review had low-to-moderate levels of fatigue prior to commencing rehabilitation. The majority of participants in this review had completed treatment and had early stage cancer, which may explain why their fatigue did not
Table 4 Meta-analysis, effect of multidisciplinary, exercise-based oncology rehabilitation programs on patient-related outcomes vers Outcome Time point Trials (n) Quality of Life (0 worst to 100 best) Post-intervention 536,37,39,42,48 EORTC QLQ-C30 Global Follow-up 237,39 EORTC QLQ C30 Physical function Post-intervention EORTC QLQ-C30 Role function Follow-up 536,37,39,42,48 EORTC QLQ-C30 Emotional function Post-intervention 237,39 EORTC QLQ-C30 Cognitive function Follow-up EORTC QLQ-C30 Social function Post-intervention 536,37,39,42,48 Follow-up 237,39 Post-intervention Follow-up 536,37,39,42,48 Post-intervention 237,39 Follow-up 536,37,39,42,48 237,39 536,37,39,42,48 237,39 Symptoms (0 best to 100 worst) Post-intervention 536,37,39,42,48 EORTC QLQ-C30 Fatigue Follow-up 237,39 Post-intervention EORTC QLQ-C30 Nausea/vomiting Post-intervention 536,37,39,42,48 EORTC QLQ-C30 Pain Post-intervention 536,37,39,42,48 EORTC QLQ-C30 Dyspnoea Post-intervention 536,37,39,42,48 EORTC QLQ-C30 Insomnia Post-intervention 536,37,39,42,48 EORTC QLQ-C30 Appetite Post-intervention 536,37,39,42,48 EORTC QLQ-C30 Constipation Post-intervention 536,37,39,42,48 EORTC QLQ-C30 Diarrhoea Post-intervention 536,37,39,42,48 EORTC QLQ-C30 Financial difficulty 536,37,39,42,48 Body composition Post-intervention 437,41,48,50 BMI (kg/m2) Post-intervention 237,50 Waist circumference (cm) Physical capacity Post-intervention 336,37,50 VO2 peak (ml/kg/min) Post-intervention 239,41 30-sec Sit to Stand (repetitions) Post-intervention 236,42 Strength: 1RM chest (kg) Post-intervention 236,42 Strength: 1RM leg (kg)
sus usual care. MD (95% CI) I2 MCID GRADE 22 Dennett et al: Multidisciplinary, exercise-based oncology rehabilitation programs Participants Exp minus Con (n) 5 to 8 Lowa 14 (21 to 28) 95% Moderateb 557 8 (21 to 18) 0% 2 to 7 Lowa 77 Lowa 10 (22 to 22) 97% 6 to 12 Lowa 551 3 (29 to 15) 41% Moderateb 77 7 (22 to 16) 82% 6 to 9 Lowa 7 (27 to 20) 0% Lowa 551 8 (22 to 17) 86% 3 to 7 Lowa 76 Lowa 6 (229 to 41) 89% 3 to 8 Lowa 550 4 (25 to 13) 88% Lowa 77 24 to 29 27 (225 to 11) 60% 23 to 29 Lowa 551 7 (23 to 17) 88% 25 to 29 Lowa 76 22 to 29 Lowa 28 (256 to 40) 90% 25 to 29 Lowa 551 27 to 213 Lowa 77 29 (223 to 6) 94% 24 to 210 Lowa 23 (214 to 9) 90% 23 to 211 High 674 27 (214 to 0) 91% Lowa 77 210 (223 to 3) 93% 23 Moderatec 546 25 (213 to 2) 82% Lowa 551 215 (233 to 4) 94% NA 551 NA Lowa 551 0 (21 to 0) 0% Moderateb 550 26 (214 to 3) 70% NA 551 21 (28 to 7) 83% 2 High 551 27 (215 to 1) 78% Lowa 551 NA High 22.2 (25.2 to 0.9) 67% NA Moderateb 159 20.7 (26.9 to 5.6) 0% 62 1.08 (20.52 to 2.68) 0% 215 6 (3 to 9) 0% 55 388 3.6 (0.4 to 6.8) 18% 388 19.5 (12.3 to 26.8) 0%
Research 23 GRADE Moderated Moderated GRADE working group grades of evidence (see reason for downgrade). PEDro score , 6 was considered lower quality. MCID taken from pulmonary rehabilitation cohort for HADS70 and 30-sec sit to stand71 because no MCID was available for oncology Study MD (95% CI) rehabilitation. MCID for EORTC from Cocks et al 2012.56 Range of scores available representing the smallest clinically relevant difference. Adamsen 2009 Random BMI = body mass index, EORTC QLQ-C30 = European Organisation for Research and Treatment of Cancer Quality-of-life Questionnaire Core 30, GRADE = Grades of Research, Assessment, Development and Evaluation, HADS = Hospital Anxiety and Ghavami 2017 21.8 to 21.3 21.7 to –1.5 Depression Scale, MCID = minimum clinically important difference, NA = not available, NR = not reported, 1RM = one repetition maximum. Midtgaard 2013 O’Neill 2018 a Reason for downgrade: heterogeneity, wide confidence intervals. Uster 2018 b Reason for downgrade: wide confidence intervals. c Reason for downgrade: heterogeneity. Pooled d Both trials rated lower quality. MCID –50 –25 0 25 50 (kg) Favours Favours control experimental 20.7 (21.2 to 20.1) 0% Figure 2. Mean difference (95% CI) in effect of multidisciplinary, exercise-based oncology rehabilitation compared with usual care on global quality of life measured with the European Organisation for Research and Treatment of Cancer Quality-of-life Questionnaire Core 30 (EORTC-QLQ C30) by pooling data from five trials. MD (95% CI) I2 20.7 (21.4 to 0.1) 0% significantly improve, given that the prevalence of cancer-related Exp minus Con fatigue is higher among patients completing treatment and typi- cally resolves within a year of treatment completion.61,62 In addition, Participants baseline physical performance of participants in relation to their (n) disease status was largely unknown. Given that the greatest gains are made by participants with the highest needs,63 it is possible that there was a ceiling effect of rehabilitation interventions with low baseline fatigue and high physical performance. The results from this review provide preliminary evidence from two trials that multidisciplinary education in addition to exercise may 338 338 A Study MD (95% CI) Random Adamsen 2009 Midtgaard 2013 Pooled Trials 236,43 236,43 –8 –4 0 48 (n) (kg) Favours Favours control experimental B Study MD (95% CI) Random Adamsen 2009 Time point Post-intervention Post-intervention Midtgaard 2013 Pooled –20 –10 0 10 20 (kg) Favours Favours control experimental Psychological well-being (0 best to 21 worst) C Study MD (95% CI) Random Bourke 2011 Uster 2018 Pooled –10 –5 0 5 10 Table 4 (Continued) HADS Anxiety HADS Depression Favours (repetitions) Favours Outcome control experimental Figure 4. Mean difference (95% CI) in effect of multidisciplinary, exercise-based oncology rehabilitation compared with usual care on strength measured by the (A) 1 repetition maximum (1RM) chest press, (B) 1RM leg press and (C) 30-second sit to stand.
24 Dennett et al: Multidisciplinary, exercise-based oncology rehabilitation programs Study MD (95% CI) service cost savings.69 There were also high levels of unexplained Bergland 1994 Random heterogeneity in the meta-analysis, which may limit the confidence Midtgaard 2013 in the estimate of the pooled effect. To account for this, sensitivity Pooled analysis was completed considering cancer type, treatment phase and program duration; however, further subgroup analysis was limited by –1 –0.5 0 0.5 1 lack of trials. Favours Favours Only one randomised controlled trial reported the effect of experimental control multidisciplinary, exercise-based oncology rehabilitation programs on healthcare service outcomes. If exercise-based rehabilitation pro- Figure 6. Mean difference (95% CI) in effect of multidisciplinary, exercise-based grams are to become part of standard care, more research is required oncology rehabilitation compared with usual care on depression measured with the to evaluate oncology rehabilitation at a healthcare service level. This Hospital Anxiety and Depression Scale (HADS) by pooling data from two trials. review of 17 trials demonstrated moderate-to-high-quality evidence that multidisciplinary, exercise-based oncology rehabilitation im- not provide additional benefit compared with exercise alone. This is proves strength and reduces depression, and low-quality evidence consistent with previous findings related to oncology,64 pulmonary65 that it may improve quality of life compared with usual care. Strong and cardiac rehabilitation.18 It may be expected that a complex recommendations relating to optimal content, timing and duration of multidisciplinary rehabilitation intervention would be more effective programs were unable to be made. However, preliminary evidence than rehabilitation alone, as it can address multiple issues.66 suggests that exercise-based oncology rehabilitation that includes However, there is strong evidence that exercise alone improves added multidisciplinary education may not provide additional benefit cancer-related outcomes.9,11 Therefore, routine addition of multidis- compared with supervised, group-based exercise alone, nor does ciplinary education to exercise may need to be reconsidered. In the providing early compared to late rehabilitation in head and neck and two trials that compared multidisciplinary rehabilitation and super- lung cancer survivors. vised group exercise, one multidisciplinary intervention was a 6-hour mind-body intervention, including guided discussion and lectures, What was already known on this topic: Among people with and one comprised a 2-hour cognitive behavioural training session cancer, exercise helps to reduce fatigue and depression, while including education, discussion and problem solving. Adherence to improving cardiorespiratory fitness and quality of life. Supervised specific elements of the rehabilitation program may be reduced due exercise programs with higher doses of exercise yield the to participants needing to direct focus on multiple areas. Education greatest improvements. Multidisciplinary oncology rehabilitation may also not be delivered in a way that is consistent with adult programs supplement the exercise with lifestyle counselling and learning needs, reducing its potential effectiveness.65 As up to 80% of educational sessions from various healthcare professionals. existing oncology rehabilitation programs include education ses- What this study adds: Multidisciplinary, exercise-based sions,20 more research is required to justify their inclusion as a core oncology rehabilitation programs offered in a group format in- element of cancer rehabilitation. crease limb muscle strength, improve performance on functional strength tasks and reduce depression. Current evidence does not Timing of exercise-based oncology rehabilitation along the cancer confirm whether they improve outcomes at the healthcare ser- continuum may not be as important as having access to services. Two vice level, such as costs, hospitalisations and length of stay. trials comparing early and delayed treatment found no between- group difference in most outcomes. Moreover, in one trial offering eAddenda: Figures 3, 5 and 7 and Appendices 1 to 4 can be found early rehabilitation for lung cancer surgery there was high dropout, online at https://doi.org/10.1016/j.jphys.2020.12.008. low adherence and higher complications in the earlier rehabilitation group.38 These findings support qualitative data from participants of Ethics approval: Not applicable. exercise-based cancer rehabilitation, suggesting that patients prefer Competing interests: Nil. flexible programs they can access when they feel ready to participate, Source(s) of support: This work was supported by the Evidence especially when undergoing cancer treatment.19,67 However, our Translation in Allied Health (EVITAH) group, an initiative funded by findings are based on limited data. Recent guidelines have been the National Health and Medical Research Council. released supporting the implementation of prehabilitation for people Acknowledgements: We thank the Evidence Translation in Allied with cancer, where it may be more appropriate to provide individual Health (EVITAH) group for their contribution in initiating and guiding multidisciplinary therapy than group-based therapies.68 the completion of this systematic review. Provenance: Not invited. Peer reviewed. To our knowledge, this review is the first to evaluate the effect of Correspondence: Amy M Dennett, Allied Health Clinical multidisciplinary, exercise-based oncology rehabilitation group pro- Research Office, Eastern Health, Melbourne, Australia. Email: grams. This review was reported in line with PRISMA guidelines and [email protected] prospectively registered. It comprised randomised controlled trials and trials were not restricted to countries where English was the References primary language. This inclusion reduced the risk of selection and publication bias, increasing confidence in the results. This review also 1. Institute of Medicine, National Research Council of the National Academies. included trials with participants with a variety of cancers, across the From cancer patient to cancer survivor - lost in transition. Washington, DC: An treatment continuum including people with advanced cancer, which American Society of Clinical Oncology and Institute of Medicine symposium; is often underrepresented in exercise and cancer trials. 2006. The methodological quality of most of the included trials was low, 2. Stewart BW. World Cancer Report 2014. In: Wild CP, ed. Lyon: International Agency largely due to lack of allocation concealment and blinded assessment. for Research on Cancer/World Health Organization; 2014. We were unable to make strong recommendations relating to the duration and timing of exercise-based oncology rehabilitation pro- 3. Burg MA, Adorno G, Lopez EDS, Loerzel V, Stein K, Wallace C, et al. Current unmet grams due to the limited number and diversity of studies. Most needs of cancer survivors: analysis of open-ended responses to the American included trials were short-term and completed after treatment in Cancer Society Study of Cancer Survivors II. Cancer. 2015;121:623–630. participants with early-stage cancer. There were insufficient trials to complete a moderator analysis. Furthermore, no trials examined 4. Lisy K, Langdon L, Piper A, Jefford M. Identifying the most prevalent unmet needs multi-disciplinary, exercise-based prehabilitation, which is an area of of cancer survivors in Australia: a systematic review. Asia Pac J Clin Oncol. increasing interest in cancer, particularly due to potential healthcare 2019;15:e68–e78. 5. Reilly CM, Bruner DW, Mitchell SA, Minasian LM, Basch E, Dueck AC, et al. A literature synthesis of symptom prevalence and severity in persons receiving active cancer treatment. Support Care Cancer. 2013;21:1525–1550. 6. Armenian SH, Xu L, Ky B, Sun C, Farol LT, Pal SK, et al. Cardiovascular disease among survivors of adult-onset cancer: a community-based retrospective cohort study. J Clin Oncol. 2016;34:1122.
Research 25 7. Silver JK, Raj VS, Fu JB, Wisotzky EM, Smith SR, Kirch RA. Cancer rehabilitation and fitness, and patient-reported outcomes in cancer survivors: a randomized, palliative care: critical components in the delivery of high-quality oncology ser- controlled trial. Ann Oncol. 2013;24:2267–2273. vices. Support Care Cancer. 2015;23:3633–3643. 37. O’Neill LM, Guinan E, Doyle SL, Bennett AE, Murphy C, Elliott JA, et al. The RESTORE randomized controlled trial: impact of a multidisciplinary rehabilitative program 8. Buffart LM, Kalter J, Sweegers MG, Courneya KS, Newton RU, Aaronson NK, et al. on cardiorespiratory fitness in esophagogastric cancer survivorship. Ann Surg. Effects and moderators of exercise on quality of life and physical function in pa- 2018;268:747–755. tients with cancer: an individual patient data meta-analysis of 34 RCTs. Cancer 38. Quist M, Sommer MS, Vibe-Petersen J, Stærkind MB, Langer SW, Larsen KR, et al. Treat Rev. 2017;52:91–104. Early initiated postoperative rehabilitation reduces fatigue in patients with oper- able lung cancer: a randomized trial. Lung Cancer. 2018;126:125–132. 9. Campbell KL, Winters-Stone KM, Wiskemann J, May AM, Schwartz AL, 39. Uster A, Ruehlin M, Mey S, Gisi D, Knols R, Imoberdorf R, et al. Effects of nutrition Courneya KS, et al. Exercise guidelines for cancer survivors: consensus statement and physical exercise intervention in palliative cancer patients: A randomized from international multidisciplinary roundtable. Med Sci Sports Exerc. controlled trial. Clin Nutr. 2018;37:1202–1209. 2019;51:2375–2390. 40. Berglund G, Petersson LM, Eriksson KC, Wallenius I, Roshanai A, Nordin KM, et al. “Between Men”: a psychosocial rehabilitation programme for men with prostate 10. Mustian KM, Alfano CM, Heckler C, Kleckner AS, Kleckner IR, Leach CR, et al. cancer. Acta Oncol. 2007;46:83–89. Comparison of pharmaceutical, psychological, and exercise treatments for cancer- 41. Bourke L, Thompson G, Gibson DJ, Daley A, Crank H, Adam I, et al. Pragmatic related fatigue: a meta-analysis. JAMA Oncol. 2017;3:961–968. lifestyle intervention in patients recovering from colon cancer: a randomized controlled pilot study. Arch Phys Med Rehabil. 2011;92:749–755. 11. Cormie P, Zopf EM, Zhang X, Schmitz KH. The impact of exercise on cancer mor- 42. Adamsen L, Quist M, Andersen C, Møller T, Herrstedt J, Kronborg D, et al. Effect of a tality, recurrence, and treatment-related adverse effects. Epidemiol Rev. multimodal high intensity exercise intervention in cancer patients undergoing 2017;39:71–92. chemotherapy: randomised controlled trial. BMJ. 2009;339:b3410. 43. Berglund G, Bolund C, Gustafsson UL, Sjoden PO. A randomized study of a reha- 12. Jones LW, Habel LA, Weltzien E, Castillo A, Gupta D, Kroenke CH, et al. Exercise and bilitation program for cancer patients: the ‘starting again’ group. Psycho Oncol. risk of cardiovascular events in women with nonmetastatic breast cancer. J Clin 1994;3:109–120. Oncol. 2016;34:2743–2749. 44. Clark MM, Rummans TA, Atherton PJ, Cheville AL, Johnson ME, Frost MH, et al. Randomized controlled trial of maintaining quality of life during radiotherapy for 13. Armenian SH, Lacchetti C, Barac A, Carver J, Constine LS, Denduluri N, et al. Pre- advanced cancer. Cancer. 2013;119:880–887. vention and monitoring of cardiac dysfunction in survivors of adult cancers: 45. Korstjens I, May AM, van Weert E, Post WJ, van der Schans CP, van den Borne B, American Society of Clinical Oncology clinical practice guideline. J Clin Oncol. et al. Quality of life after self-management cancer rehabilitation: a randomized 2017;35:893–911. controlled trial comparing physical and cognitive-behavioral training versus physical training. Psychosom Med. 2008;70:422–429. 14. Cormie P, Atkinson M, Bucci L, Cust A, Eakin E, Hayes S, et al. Clinical Oncology 46. Rummans TA, Clark MM, Sloan JA, Frost MH, Bostwick JM, Atherton PJ, et al. Society of Australia position statement on exercise in cancer care. Med J Aust. Impacting quality of life for patients with advanced cancer with a structured 2018;209:184–187. multidisciplinary intervention: a randomized controlled trial. J Clin Oncol. 2006;24:635–642. 15. Runowicz CD, Leach CR, Henry NL, Henry KS, Mackey HT, Cowens-Alvarado RL, 47. Cho OH, Yoo YS, Kim NC. Efficacy of comprehensive group rehabilitation for et al. American Cancer Society/American Society of Clinical Oncology breast cancer women with early breast cancer in South Korea. Nurs Health Sci. 2006;8:140–146. survivorship care guideline. J Clin Oncol. 2016;34:611–635. 48. Ghavami H, Akyolcu N. Effects of a lifestyle interventions program on quality of life in breast cancer survivors. UHOD. 2017;27:91–99. 16. Gilchrist SC, Barac A, Ades PA, Alfano CM, Franklin BA, Jones LW, et al. Cardio- 49. Sandmael JA, Bye A, Solheim TS, Stene GB, Thorsen L, Kaasa S, et al. Feasibility and oncology rehabilitation to manage cardiovascular outcomes in cancer patients and preliminary effects of resistance training and nutritional supplements during survivors: a scientific statement from the American Heart Association. Circulation. versus after radiotherapy in patients with head and neck cancer: a pilot ran- 2019;139:e997–e1012. domized trial. Cancer. 2017;123:4440–4448. 50. Sheppard VB, Hicks J, Makambi K, Hurtado-de-Mendoza A, Demark-Wahnefried W, 17. Dittus KL, Lakoski SG, Savage PD, Kokinda N, Toth M, Stevens D, et al. Exercise- Adams-Campbell L. The feasibility and acceptability of a diet and exercise trial in based oncology rehabilitation: leveraging the cardiac rehabilitation model. overweight and obese black breast cancer survivors: The Stepping STONE study. J Cardiopulm Rehabil Prev. 2015;35:130–139. Contemp Clin Trials. 2016;46:106–113. 51. Spahn G, Choi KE, Kennemann C, Lüdtke R, Franken U, Langhorst J, et al. Can a 18. Anderson L, Thompson DR, Oldridge N, Zwisler AD, Rees K, Martin N, et al. Exer- multimodal mind-body program enhance the treatment effects of physical activity cise-based cardiac rehabilitation for coronary heart disease. Cochrane Database Syst in breast cancer survivors with chronic tumor-associated fatigue? A randomized Rev. 2016;1:CD001800. controlled trial. Integ Cancer Ther. 2013;12:291–300. 52. Hubbard G, Adams R, Campbell A, Kidd L, Leslie SJ, Munro J, et al. Is referral of 19. Dennett AM, Peiris CL, Taylor NF, Reed MS, Shields N. ‘A good stepping stone postsurgical colorectal cancer survivors to cardiac rehabilitation feasible and to normality’: a qualitative study of cancer survivors’ experiences of an acceptable? A pragmatic pilot randomised controlled trial with embedded quali- exercise-based rehabilitation program. Support Care Cancer. 2019;27:1729– tative study. BMJ Open. 2016;6, e009284. 1736. 53. Puhan MA, Gimeno-Santos E, Cates CJ, Troosters T. Pulmonary rehabilitation following exacerbations of chronic obstructive pulmonary disease. Cochrane 20. Dennett AM, Peiris CL, Shields N, Morgan D, Taylor NF. Exercise therapy in oncology Database Syst Rev. 2016;12:CD005305. rehabilitation in Australia: a mixed-methods study. Asia Pac J Clin Oncol. 54. Cocks K, King MT, Velikova G, Martyn St-James M, Fayers PM, Brown JM. Evidence- 2017;13:e515–e527. based guidelines for determination of sample size and interpretation of the Eu- ropean Organisation for the Research and Treatment of Cancer Quality of Life 21. Myrhaug HT, Mbalilaki JA, Lie N-EK, Hansen T, Nordvik JE. The effects of multi- Questionnaire Core 30. J Clin Oncol. 2011;29:89. disciplinary psychosocial interventions on adult cancer patients: a systematic re- 55. Hong F, Bosco JLF, Bush N, Berry DL. Patient self-appraisal of change and minimal view and meta-analysis. Disabil Rehabil. 2020;42:1062–1070. clinically important difference on the European Organization for the Research and Treatment of Cancer Quality of Life Questionnaire Core 30 before and during 22. Scott DA, Mills M, Black A, Cantwell M, Campbell A, Cardwell CR, et al. Multidi- cancer therapy. BMC Cancer. 2013;13:165. mensional rehabilitation programmes for adult cancer survivors. Cochrane Data- 56. Cocks K, King MT, Velikova G, Martyn St-James M, Fayers PM, et al. Evidence-based base Syst Rev. 2013;28:CD007730. guidelines for interpreting change scores for the European Organisation for the Research and Treatment of Cancer Quality of Life Questionnaire Core 30. Eur J 23. Canestraro A, Nakhle A, Stack M, Strong K, Wright A, Beauchamp M, et al. Oncology Cancer. 2012;48:1713–1721. rehabilitation provision and practice patterns across Canada. Physiother Can. 57. Veloso AG, Sperling C, Holm LV, Nicolaisen A, Rottmann N, Thayssen S, et al. Unmet 2013;65:94–102. needs in cancer rehabilitation during the early cancer trajectory – a nationwide patient survey. Acta Oncol. 2013;52:372–381. 24. Mewes JC, Steuten LM, Ijzerman MJ, van Harten WH. Effectiveness of multidi- 58. Scott JM, Zabor EC, Schwitzer E, Koelwyn GJ, Adams SC, Nilsen TS, et al. Efficacy of mensional cancer survivor rehabilitation and cost-effectiveness of cancer reha- exercise therapy on cardiorespiratory fitness in patients with cancer: a systematic bilitation in general: a systematic review. Oncologist. 2012;17:1581–1593. review and meta-analysis. J Clin Oncol. 2018;36:2297–2305. 59. Neil-Sztramko SE, Winters-Stone KM, Bland KA, Campbell KL. Updated systematic 25. Gordon LG, Scuffham P, Battistutta D, Graves N, Tweeddale M, Newman B. A cost- review of exercise studies in breast cancer survivors: attention to the principles of effectiveness analysis of two rehabilitation support services for women with breast exercise training. Br J Sports Med. 2019;53:504–512. cancer. Breast Cancer Res Treat. 2005;94:123–133. 60. Hoffmann TC, Glasziou PP, Boutron I, Milne R, Perera R, Moher D, et al. Better reporting of interventions: template for intervention description and replication 26. Khan KA, Mazuquin B, Canaway A, Petrou S, Bruce J. Systematic review of economic (TIDieR) checklist and guide. BMJ. 2014;348:g1687. evaluations of exercise and physiotherapy for patients treated for breast cancer. 61. Bower JE. Cancer-related fatigue–mechanisms, risk factors, and treatments. Nat Rev Breast Cancer Res Treat. 2019;176:37–52. Clin Oncol. 2014;11:597–609. 62. Servaes P, Verhagen C, Bleijenberg G. Fatigue in cancer patients during and after 27. Moher D, Liberati A, Tetzlaff J, Altman DG, PRISMA Group. Preferred reporting treatment: prevalence, correlates and interventions. Eur J Cancer. 2002;38:27– items for systematic reviews and meta-analyses: the PRISMA statement. PLoS Med. 43. 2009;6, e1000097. 63. Cormie P, Lamb S, Newton R, Valentine L, McKiernan S, Spry N, et al. One size doesn’t fit all: responders and nonresponders to exercise interventions for cancer 28. Scottish Intercollegiate Guidelines Network (SIGN). Search Filters; 2019. https:// patients and survivors. Oral abstracts. Asia Pac J Clin Oncol. 2017;13:85–86. www.sign.ac.uk/search-filters.html. Accessed 20 March, 2019. 29. Macedo LG, Elkins MR, Maher CG, Moseley AM, Herbert RD, Sherrington C. There was evidence of convergent and construct validity of Physiotherapy Evidence Database quality scale for physiotherapy trials. J Clin Epidemiol. 2010;63:920–925. 30. Maher CG, Sherrington C, Herbert RD, Moseley AM, Elkins M. Reliability of the PEDro scale for rating quality of randomized controlled trials. Phys Ther. 2003;83:713–721. 31. Wiebe N, Vandermeer B, Platt RW, Klassen TP, Moher D, Barrowman NJ. A systematic review identifies a lack of standardization in methods for handling missing variance data. J Clin Epidemiol. 2006;59:342–353. 32. Review Manager (RevMan) [computer program]. Version 5.3. Copenhagen: The Nordic Cochrane Centre, The Cochrane Collaboration; 2014. 33. Dennett AM, Peiris CL, Shields N, Prendergast LA, Taylor NF. Moderate-intensity exercise reduces fatigue and improves mobility in cancer survivors: a systematic review and meta-regression. J Physiother. 2016;62:68–82. 34. Higgins JPT, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta- analyses. BMJ. 2003;327:557–560. 35. GraphPad. Quick Calcs. Quantify agreement with kappa; 2018. https://www. graphpad.com/quickcalcs/kappa1/. Accessed 10 April, 2019. 36. Midtgaard J, Christensen JF, Tolver A, Jones LW, Uth J, Rasmussen B, et al. Efficacy of multimodal exercise-based rehabilitation on physical activity, cardiorespiratory
26 Dennett et al: Multidisciplinary, exercise-based oncology rehabilitation programs 64. Dennett AM, Shields N, Peiris CL, Reed MS, O’Halloran PD, Taylor NF. Does 68. Macmillan Cancer Support. Prehabilitation for people with cancer: Principles and psychoeducation added to oncology rehabilitation improve physical activity guidance for prehabilitation; 2019. https://www.macmillan.org.uk/assets/ and other health outcomes? A systematic review. Rehabil Oncol. 2017;35: prehabilitation-guidance-for-people-with-cancer.pdf. Accessed 9 July, 2019. 61–71. 69. Cavalheri V, Jenkins S, Hill K. Physiotherapy practice patterns for patients under- 65. Blackstock FC, Evans RA. Rehabilitation in lung diseases: ‘Education’ component of going surgery for lung cancer: a survey of hospitals in Australia and New Zealand. pulmonary rehabilitation. Respirology. 2019;24:863–870. Intern Med J. 2013;43:394–401. 66. Squires JE, Sullivan K, Eccles MP, Worswick J, Grimshaw JM. Are multifaceted in- 70. Smid DE, Franssen FME, Houben-Wilke S, Vanfleteren LE, Janssen DJ, Wouters EF, terventions more effective than single-component interventions in changing et al. Responsiveness and MCID estimates for CAT, CCQ, and HADS in patients with health-care professionals’ behaviours? An overview of systematic reviews. Imple- COPD undergoing pulmonary rehabilitation: a prospective analysis. J Am Med Dir ment Sci. 2014;9:152. Assoc. 2017;18:53–58. 67. Dennett AM, Harding KE, Reed MS. The challenge of timing: a qualitative study on 71. Zanini A, Crisafulli E, D’Andria M, Gregorini C, Cherubino F, Zampogna E, clinician and patient perspectives about implementing exercise-based rehabilita- et al. Minimal clinically important difference in 30 second sit-to-stand test tion in an acute cancer treatment setting. Support Care Cancer. 2020;28: after pulmonary rehabilitation in patients with COPD. Eur Respir J. 6035–6043. 2018;52:OA5199.
4 Cumulative number of records clinical practice guidelines Editorial systematic reviews 1000 randomised trials therapeutic exercises are the best for rehabilitation of elbow, wrist 800 and hand disorders. Also, we need to determine what role and rele- 600 vance central mechanisms and effects on the sensorimotor system 400 have in these disorders, and how to approach these mechanisms to 200 improve outcomes for patients. Competing interest: Nil. Source of support: Nil. Acknowledgement: Nil. Provenance: Invited. Not peer reviewed. Correspondence: Mark Elkins, Centre for Education & Workforce Development, Sydney Local Health District, Sydney, Australia. Email: [email protected] 0 References 1960 1970 1980 1990 2000 2010 2020 Year 1. Bonnevie T, et al. J Physiother. 2020;66:3–4. 2. Dennett A, et al. J Physiother. 2020;66:70–72. Figure 1. Cumulative evidence on the Physiotherapy Evidence Database (PEDro) about 3. Hwang R, et al. J Physiother. 2020;66:193–195. the effects of physiotherapy interventions on musculoskeletal disorders of the elbow, 4. Dorsch S, et al. J Physiother. 2020;66:211–212. wrist and hand, based on the October 2020 update of the database. 5. De Putter CE, et al. J Bone Jt Surg - Ser A. 2012;94:e56. 6. Larsen CF, et al. Eur J Epidemiol. 2004;19:323–327. Bisset and Prof Bill Vicenzino expertly summarises the available ev- 7. Taylor NF, et al. Aust J Physiother. 2007;53:7–16. idence about the burden associated with the condition, its manage- 8. Ziebart C, et al. Hand Ther. 2019;24:69–81. ment (with a particular focus on physiotherapy interventions), and 9. Bruder AM, et al. J Physiother. 2017;63:205–220. future directions for research and clinical practice. 10. Lewis KJ, et al. J Physiother. 2020;66:97–104. 11. Kay S, et al. Aust J Physiother. 2008;54:253–259. In summary, this online article collection includes a range of 12. Reid SA, et al. J Physiother. 2020;66:105–112. important developments in the physiotherapy management of 13. Jongs RA, et al. J Physiother. 2012;58:173–180. elbow, wrist and hand disorders, mainly related to exercises. It also 14. Bruder AM, et al. J Physiother. 2016;62:145–152. highlights an important unanswered question: while exercises seem 15. Blanquero J, et al. J Physiother. 2020;66:236–242. to be effective in the rehabilitation of these disorders, are the 16. Blanquero J, et al. J Physiother. 2019;65:81–87. exercises that we currently use ideal or do we need to include other 17. Villafañe JH, et al. J Physiother. 2013;59:25–30. exercises, interventions or dosages? Future research should address 18. Bisset LM, et al. J Physiother. 2015;61:174–181. two important issues. First, we need to determine which types of Websites PEDro www.pedro.org.au Paper of the Year 2020 The Editorial Board of Journal of Physiotherapy is pleased to announce the 2020 Paper of the Year Award. The winning paper is judged by a panel of members of the International Advisory Board who do not have a conflict of interest with any of the papers under consideration. They vote for the paper published in the 2020 calendar year that, in their opinion, has the best combination of scientific merit and application to the clinical practice of physiotherapy. The winning paper is ‘Preoperative physiotherapy is cost-effective for preventing pulmonary complications after major abdominal surgery: a health economic analysis of a multicentre randomised trial.’1 The authors are Ianthe Boden, PhD, from Launceston General Hospital and The University of Melbourne, and her colleagues in Australia and New Zealand.1 The paper addressed uncertainty around the cost-effectiveness of preoperative physiotherapy for patients undergoing major abdominal surgery. Before major abdominal surgery, a single physiotherapy session involving education and training markedly reduces the incidence of postoperative pulmonary complications. However, uncertainty about the cost-effectiveness of preoperative physiotherapy may be making some hospitals reluctant to institute this intervention. The winning paper examined whether preoperative physiotherapy aimed at preventing postoperative pulmonary complications is cost-effective from the hospital’s perspective. Cost-effectiveness and quality-adjusted life year gains were most evident when experienced physiotherapists delivered the intervention. Nevertheless, even across a range of physiotherapists who participated in the study, preoperative physiotherapy aimed at preventing postoperative pulmonary complications was highly likely to be cost-effective. This is the second cardiorespiratory study that joins recent winners in the geriatrics, sports and neurology subdisciplines.2–5 The members of the Editorial Board congratulate Dr Boden and her co-authors on their success. References 1. Boden I, Robertson IK, Neil A, Reeve J, Palmer AJ, Skinner EH, Browning L, Anderson L, Hill C, Story D, Denehy L. Preoperative physiotherapy is cost-effective for preventing pulmonary complications after major abdominal surgery: a health economic analysis of a multicentre randomised trial. J Physiother. 2020;66:180–187. 2. Moreno NA, de Aquino BG, Garcia IF, Tavares LS, Costa LF, Giacomassi IWS, Lunardi AC. Physiotherapist advice to older inpatients about the importance of staying physically active during hospitalisation reduces sedentary time, increases daily steps and preserves mobility: a randomised trial. J Physiother. 2019;65:208–214. 3. McKeough Z, Cheng SWM, Alison J, Jenkins C, Hamer M, Stamatakis E. Low leisure-based sitting time and being physically active were associated with reduced odds of death and diabetes in people with chronic obstructive pulmonary disease: a cohort study. J Physiother. 2018;64:114–120. 4. Al Attar WSA, Soomro N, Pappas E, Sinclair PJ, Sanders RH. Adding a post-training FIFA 111 exercise program to the pre-training FIFA 111 injury prevention program reduces injury rates among male amateur soccer players: a cluster-randomised trial. J Physiother. 2017;63:235–242. 5. van den Berg M, Sherrington C, Killington M, Smith S, Bongers B, Hassett L, Crotty M. Video and computer-based interactive exercises are safe and improve task-specific balance in geriatric and neurological rehabilitation: a randomised trial. J Physiother. 2016;62:20–28. https://doi.org/10.1016/j.jphys.2020.12.003 1836-9553/
Journal of Physiotherapy 67 (2021) 49–55 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 People with Parkinson’s disease are more willing to do additional exercise if the exercise program has specific attributes: a discrete choice experiment Serene S Paul a, Colleen G Canning a, Niklas Löfgren b,c, Cathie Sherrington d,e, Deborah C Lee a, Julie Bampton a, Kirsten Howard e a Discipline of Physiotherapy, Faculty of Medicine and Health, University of Sydney, Sydney, Australia; b Function Area Occupational Therapy and Physiotherapy, Allied Health Professionals Function, Karolinska University Hospital, Stockholm, Sweden; c Department of Neurobiology, Care Sciences and Society, Division of Physiotherapy, Karolinska Institutet, Stockholm, Sweden; d Institute for Musculoskeletal Health, Faculty of Medicine and Health, University of Sydney, Sydney, Australia; e School of Public Health, Faculty of Medicine and Health, University of Sydney, Sydney, Australia KEY WORDS: ABSTRACT Parkinson disease Question: What specific attributes of exercise programs influence the preferences of people with Parkinson’s Patient preference disease for additional exercise compared with their current practice? What trade-offs are participants willing Exercise to make between exercise program attributes? Design: Discrete choice experiment. Participants: Five Physical therapy hundred and forty people with Parkinson’s disease. Intervention: Participants decided whether they would Discrete choice experiment adopt a hypothetical program in addition to their current exercise routine. Outcome measures: Exercise program attributes included: type, number of sessions/week, location, travel time/session, delivery mode, supervisor’s expertise, extent of supervision, benefits for physical and psychological function and out-of- pocket cost/session. Results: Participants preferred additional exercise when programs: provided physical (OR 1.85, 95% CI 1.61 to 2.13) or psychological (OR 1.45, 95% CI 1.26 to 1.67) benefit, involved less travel time (ORs 1.50 to 2.02) and were supervised by qualified professionals with Parkinson’s disease expertise (ORs 1.51 to 1.91). Participants were most willing to add multimodal exercise to their exercise routine (ORs 2.01 to 2.19). Participants were less likely to prefer higher cost programs (OR 0.65, 95% CI 0.60 to 0.71, per AU$10 cost increase) or group sessions compared to individual sessions (OR 0.72, 95% CI 0.54 to 0.96). Men preferred adding strengthening exercises (OR 2.00, 95% CI 1.23 to 3.26) and women had a preference against adding aerobic exercise (OR 0.33, 95% CI 0.15 to 0.73). Participants not currently exercising were more likely to prefer adding exercise compared with those already exercising 300 minutes weekly (OR 1.74, 95% CI 1.15 to 2.63). Conclusion: People with Parkinson’s disease were more willing to participate in exercise programs that cost less, involve less travel, provide physical or psychological benefits and are supervised by qualified pro- fessionals. To enable more people with Parkinson’s disease to exercise, health services should provide pro- grams addressing these factors and account for sex differences. [Paul SS, Canning CG, Löfgren N, Sherrington C, Lee DC, Bampton J, Howard K (2021) People with Parkinson’s disease are more willing to do additional exercise if the exercise program has specific attributes: a discrete choice experiment. Journal of Physiotherapy 67:49–55] © 2020 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 Emerging evidence indicates that physical activity at moderate to Parkinson’s disease is a common neurodegenerative disorder with high intensities may reduce the risk of developing Parkinson’s disease an incidence of 8 to 18 per 100 person-years.1 People present with and can attenuate motor symptoms in this population.6,7 Recent ev- motor symptoms such as impaired gait, balance and freezing of gait, as well as with a variety of non-motor symptoms, most commonly idence suggests that people with more severe Parkinson’s disease cognitive decline and depression.1 Strong evidence supports the have most to gain by participating in formal exercise.8 This highlights benefits of exercise for people with Parkinson’s disease by decreasing both motor and non-motor symptoms, improving quality of life and the importance of providing adequate exercise opportunities for reducing falls.2,3 Preliminary evidence indicates that aerobic exercise can induce neuroplasticity, delay disease progression, reduce people at different stages of the disease spectrum. depressive symptoms and improve cognition.4,5 Strength and balance training can improve gait, increase quality of life and reduce falls.2 Despite the many benefits of exercise, people with Parkinson’s disease are less active than the general population, even at earlier stages of the disease, and their physical activity levels progressively decline.9 This is problematic for people with Parkinson’s disease and healthcare systems, given that exercise interventions are likely to be cost-effective and can be cost-saving in people with milder disease https://doi.org/10.1016/j.jphys.2020.12.007 1836-9553/© 2020 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/).
50 Paul et al: Exercise preferences in Parkinson’s disease severity,10 particularly if delivered by physiotherapists with expertise was made to improve clarity; data from these 17 pilot participants in Parkinson’s disease.11 Therefore, there is an urgent need to inves- were included in the final sample. tigate ways to encourage this population to become more engaged in exercise. Based on the prior parameter estimates, a final Bayesian d-efficient discrete choice experiment design with 48 questions A core principle of patient-centred healthcare for people with (ie, four blocks of 12 questions) was generated using NGENE design Parkinson’s disease is to offer interventions that are effective and softwarea (d-error 0.058; s-estimate [minimum sample size] 227 have high likelihood of being adopted.12 However, there is a lack of participants). This approach maximised the precision of estimates of research to determine the features of exercise programs that would the value of attributes by presenting participant choice sets that enhance the adoption of evidence-based programs. For example, it is minimised the elements of the asymptotic variance-covariance unknown whether the willingness to undertake an exercise program matrix of the statistical methods used to analyse the data. is influenced by attributes such as its benefits, its costs, the avail- ability of transport, and the type, duration and frequency of exercise. Participants It is also unknown what trade-offs between these factors people with Parkinson’s disease would make in deciding whether to undertake an Community-dwelling people with Parkinson’s disease were recruited exercise program. Personal characteristics in people with Parkinson’s through advertisements in private neurology clinics and newspapers, disease influence their exercise behaviour.13–15 Greater exercise self- and through Parkinson’s New South Wales and other consumer networks efficacy, higher education and older age have been found to induce for people with Parkinson’s disease between May 2016 and January 2018. positive exercise behaviours,13 while barriers include low expecta- People were eligible to participate if they had been diagnosed with Par- tions of benefit, more severe motor symptoms, lack of time, fear of kinson’s disease and resided in Australia. People with a diagnosis of de- falling and non-motor symptoms such as apathy and fatigue.13–15 mentia or cognitive impairment, or with insufficient English language However, it is unclear how these characteristics might modify pref- ability to understand and complete the experiment were excluded. erences for attributes of exercise programs. Yet, such information is required to guide provision of exercise programs for this population. Data collection Additionally, given the neurodegenerative nature of Parkinson’s dis- ease with its changing impact on physical function, understanding All data for this study were collected via an anonymous survey. the exercise preferences of people with different disease stages is Participants were provided with a link to complete the survey online critical for promoting long-term maintenance of exercise and physical or they completed a paper copy, which was returned via mail; hence, activity. informed consent was implied upon submission or return of a completed survey. The survey took approximately 30 minutes to Therefore, the research questions for this discrete choice experi- complete and comprised three sections: background information, ment were: including demographics and Parkinson’s disease history; exercise history; and the discrete choice experiment. 1. What specific attributes of exercise programs influence the pref- erences of people with Parkinson’s disease for additional exercise The demographic information that was collected included age, compared with their current practice? sex, living in metropolitan or regional areas, domestic arrangement, preferred language, level of education, employment, annual income 2. What trade-offs are participants willing to make between exercise and health insurance. Parkinson’s disease-related information program attributes? included number of years since diagnosis, whether participants were taking anti-Parkinsonian medication, and self-rating of disease Method severity using the Hoehn and Yahr scale. Data were also collected on performance of activities of daily living from Part 2 of the Movement Design Disorders Society sponsored version of the Unified Parkinson’s Disease Rating Scale.19 Preferences for additional exercise were evaluated using a discrete choice experiment, a quantitative technique underpinned by random Participants also reported whether they undertook regular exer- utility16 and consumer choice17 theories, and based upon the premise cise. Exercisers reported the types of exercise undertaken, the num- that a healthcare service or intervention can be described by its ber of sessions per week and minutes per session, but no measure of characteristics or attributes. In a discrete choice experiment, the exercise intensity. All exercise reported by participants, except for levels of each attribute (eg, exercise type) that are presented to par- flexibility exercises (eg, stretches), had the potential to be of mod- ticipants are varied in a series of questions, and respondents choose erate intensity, although a proportion likely involved light-intensity the option that they prefer for each question. Discrete choice exper- exercise.20,21 Thus, exercisers who reported undertaking at least iments determine which attributes are driving preferences of con- 300 minutes of exercise (excluding flexibility exercises) per week sumers and the trade-offs between attributes that people are willing were classified as such,22 while exercisers who did not meet this to accept. threshold were classified as performing some exercise. This discrete choice experiment presented participants with a After reporting their current exercise, participants were presented series of questions, asking them to choose whether they would add with 12 choice sets describing a hypothetical new program where the presented exercise program to their current exercise regimen. The levels of each attribute were systematically varied. Attributes and primary outcome measure was participant preference for additional levels (Table 1) were based upon interviews of people with Parkin- exercise (over and above what they were currently doing). The design son’s disease,15 the literature on adherence to exercise in this popu- and analysis of the discrete choice experiment followed good lation13–15 and pilot testing the questionnaire. Participants were research practice guidelines.18 asked to decide whether or not they would adopt each hypothetical program in addition to their current exercise routine. Participants had An initial discrete choice experiment design was pilot tested to to complete a minimum of four choice sets to be included in the assess understanding and comprehension of attributes, using a con- analysis. An example of a choice task is presented in Figure 1. venience sample. Six participants were asked to complete the experiment to provide lay feedback in terms of understanding, Data analysis formatting or text changes. Following language modifications, a sec- ond pilot was conducted with 17 respondents to collect response data Analyses were conducted using a mixed logit modelb using 20,000 to calculate prior parameter estimates with which to inform the Halton draws with a panel specification to account for correlated design of the main study. From this second pilot, only a slight choices within an individual.23 All attributes were initially specified modification to the wording of one attribute description (travel time) as random with normal distributions. Respondent demographic characteristics were specified as categorical variables and were effects
Research 51 Table 1 Attributes and levels within the discrete choice experiment. Attribute Description of the levels presented to participants Exercise type Dance program (eg, tango or waltz/foxtrot or Irish set dancing) Balance exercise program (exercises to challenge balance, performed while standing, may include Tai-chi) Muscle strength exercise program (exercises aimed at increasing muscle strength, eg, resistance exercises using free weights, exercises using your body weight, weight machines, weighted vests, TheraBand) Aerobic exercise program (exercises to challenge the heart and lungs and make you puff, eg, fast walking, jogging, running, cycling) Walking exercise program (eg, walking indoors and/or outdoors, treadmill walking, walking using visual and/or auditory cues) Multimodal exercise program (including at least 2 of the following exercises): balance exercises, strength exercises, aerobic exercises, walking exercises Number of 45-minute exercise sessions per week 1 3 45-minute session per week 2 3 45-minute sessions per week 3 3 45-minute sessions per week 4 3 45-minute sessions per week Exercise location At home In the local neighbourhood (eg, local parks, streets or outdoor public spaces) At a hospital or health centre/practice At a community centre or facility (eg, gym, community hall) At multiple locations including home At multiple locations excluding home Travel time per exercise session Less than 5 minutes each way travel time 5 minutes each way travel time 10 minutes each way travel time 15 minutes each way travel time 30 minutes each way travel time 60 minutes or more each way travel time Exercise delivery mode: individual and/or in a group Individual session/s Group session/s of two or more people Individual plus group sessions Supervisor’s expertise There is no one supervising the exercise Supervised by a family member, friend or carer Supervised by a fitness, exercise or dance instructor without specific expertise in Parkinson’s disease Supervised by a fitness, exercise or dance instructor with specific expertise in Parkinson’s disease Supervised by a physiotherapist without specific expertise in Parkinson’s disease Supervised by a physiotherapist with specific expertise in Parkinson’s disease Amount of supervision All of the exercise is supervised Some of the exercise is supervised None of the exercise is supervised The effect on your Parkinson’s disease motor You will experience no improvement in Parkinson’s disease motor symptoms and physical function symptoms and physical function You will experience small improvement in Parkinson’s disease motor symptoms and physical function You will experience moderate improvement in Parkinson’s disease motor symptoms and physical function You will experience large improvement in Parkinson’s disease motor symptoms and physical function Effect on your overall feeling of wellbeing You will experience no improvement in overall feeling of wellbeing You will experience small improvement in overall feeling of wellbeing You will experience moderate improvement in overall feeling of wellbeing You will experience large improvement in overall feeling of wellbeing Out of pocket cost (in AU$) including travel Out of pocket cost = $0 per session (including travel costs) costs per session Out of pocket cost = $10 per session (including travel costs) Out of pocket cost = $25 per session (including travel costs) Out of pocket cost = $50 per session (including travel costs) Out of pocket cost = $100 per session (including travel costs) Out of pocket cost = $150 per session (including travel costs) coded. We examined interactions between attributes themselves, and attribute level increased; a negative coefficient indicated that an between attributes and population characteristics (eg, age, sex, in- alternative was less preferred as the attribute level increased. Results come, Parkinson’s disease severity, current exercise status) before were also reported as the odds (95% CI) of adding the new program to deciding on a final model specification. In the final model specifica- participants’ existing exercise routine. Goodness-of-fit differences tion, exercise type, travel time and exercise delivery mode were between models were examined using log-likelihood function, specified as random, categorical, effects-coded variables with a pseudo R2 and Akaike Information Criteria. All analyses used com- normal distribution; the extent of benefit for physical function and mercial softwarea. Trade-offs between attributes were calculated for psychological function were random, linear variables with a from the ratio of individual specific parameters and presented as normal distribution; and cost was specified as a random, linear var- mean willingness to pay for each level of improvement in physical iable with a triangular distribution. The number of exercise sessions/ and psychological benefit, and for exercise delivered by a professional week, location of exercise and exercise supervisor’s expertise were with expertise in Parkinson’s disease. non-random, categorical, effects-coded variables; and the extent of supervision was a non-random, linear variable. Interactions between Results exercise type and participant sex were also included. A total of 551 participants with Parkinson’s disease returned Results were reported as b coefficients and 95% CI. A positive b surveys; however, 11 participants completed fewer than four of the 12 coefficient indicated that an alternative was more preferred as the
52 Paul et al: Exercise preferences in Parkinson’s disease Dance program (eg, tango, waltz/foxtrot or Irish set dancing) were exercising 300 minutes weekly22 (OR 1.74, 95% CI 1.15 to 2.63, p = 0.009). Other sociodemographic characteristics 2 including age, 2 x 45-minute sessions per week disease severity and annual income 2 did not significantly influence the participants’ preferences for additional exercise (Table 5). At a hospital or health centre/practice Research question 2 15 minutes each way travel time Trade-offs were estimated between cost and other attributes to Group sessions calculate willingness to pay for specific program characteristics. We found that participants’ willingness to pay for each additional level of Supervised by a physiotherapist with specific expertise in PD improvement in physical and psychological benefit was AU$16.71 (95% CI 15.80 to 17.61) and AU$9.77 (95% CI 9.40 to 10.14), respec- Some of the exercise is supervised tively. Participants were willing to pay more for a program delivered by a professional with experience in Parkinson’s disease (exercise You will experience large improvement in your PD motor symptoms and physical function instructor: AU$17.16, 95% CI 16.49 to 17.82; physiotherapist: AU$10.94, 95% CI 10.51 to 11.37). Participants were also willing to pay You will experience moderate improvement in your overall feeling of wellbeing AU$8.74 (95% CI 8.40 to 9.07) for an exercise program that included some individual sessions. Out of pocket cost = $25 per session (including travel costs) Discussion Would you be willing to add this exercise program to your current weekly routine? This discrete choice experiment conducted among a large sample Yes No is the first of its type addressing this question and provides insight into the preferences of people with Parkinson’s disease to add addi- Figure 1. Example of a choice scenario. tional exercise to their existing weekly routine. As 86% of the sample reported some degree of exercising, it was unsurprising that the choice sets and were excluded from analysis, based on a priori criteria participants mostly preferred not to add exercise to their current for completion, giving a total of 6,346 choice sets for analysis. Among routines. Nevertheless, certain program characteristics increased the the 540 participants who were analysed, 395 completed online sur- likelihood that people with Parkinson’s disease would be willing to veys and 145 completed paper surveys. The sociodemographic char- add exercise, including: lower costs, less travel time, physical and acteristics of the analysed participants are summarised in Table 2. psychological benefit and supervision by physiotherapists or exercise Participants had a mean age of 69 years (range 41 to 95). The clinical instructors with Parkinson’s disease expertise. characteristics of the participants are summarised in Table 3. The mean disease duration was 7 years (range 0 to 37) and the severity of The willingness of people with Parkinson’s disease to perform Parkinson’s disease was primarily mild to moderate (84% self- additional exercise to obtain physical and psychological benefits identified as being in Hoehn and Yahr stages 1 to 3). The exercise highlights the known efficacy of exercise for improving physical history of the participants is summarised in Table 4. This sample function such as increasing strength, balance and mobility;24 and consisted primarily of exercisers, with 86% reporting undertaking reducing motor24 and non-motor3 symptoms. A recent report found some exercise and 43% already exercising 300 minutes weekly.22 that only 25% of Australians aged . 65 years met recommended levels of physical activity.22 Hence, 43% of the sample (ie, half those Research question 1 who reported currently exercising) already exercising 300 minutes weekly indicates that Australians with Parkinson’s disease are The final model (Table 5) had a good model fit (adjusted Akaike becoming increasingly aware of the benefits of exercise for health and Information Criteria of 0.888, log likelihood 22,327.5 and pseudo r2 disease management. Nevertheless, more than half the sample likely 0.376), approximately equivalent to an r2 of 0.80 in a linear regression did not exercise sufficiently, reiterating the difficulty that most people model.23 Overall, participants expressed a strong underlying prefer- with Parkinson’s disease have in achieving adequate exercise doses.25 ence, all else being equal, for not adding extra exercise to their cur- rent exercise regimens (as suggested by the significant constant, p , Despite people with Parkinson’s disease demonstrating some 0.0001). However, this underlying preference could be mitigated by willingness to add exercise to their existing routine, including the characteristics of the exercise programs. minority of participants who were not currently exercising, prag- matically a threshold was noted. Unsurprisingly, there was an aver- The participants were more willing to add additional exercise to sion to adding four sessions of additional exercise, although adding their existing exercise routine when programs lead to physical (OR two or three sessions was not significantly worse than adding one 1.85, 95% CI 1.61 to 2.13, p , 0.0001) or psychological benefits (OR session; there was also aversion to participating in exercise that 1.45, 95% CI 1.26 to 1.67, p , 0.0001). They also preferred programs incurred high costs or large amounts of travel. Given the multisystem delivered by physiotherapists with expertise in Parkinson’s disease nature of Parkinson’s disease and possible comorbidities,1 it is para- (OR 1.51, 95% CI 1.08 to 2.11, p = 0.02) or qualified exercise instructors mount that management plans for people with Parkinson’s disease, with expertise in Parkinson’s disease (OR 1.91, 95% CI 1.18 to 3.09, p = including exercise, are not unduly burdensome and that all facets can 0.009) compared with no supervision. Preferences for programs be adhered to by the individual. As the benefits of exercise dissipate varied between men and women: men with Parkinson’s disease upon cessation of exercise,26 prescription of tailored exercise pro- preferred adding additional strengthening exercises (OR 2.00, 95% CI grams maximising adherence is required. Thus, health professionals 1.23 to 3.26, p = 0.005) and women were against adding aerobic ex- should consider each person’s circumstances, including using ercises (OR 0.33, 95% CI 0.15 to 0.73, p = 0.006) to their existing ex- community-based exercise options close to home, although it may ercise routine. Both men and women with Parkinson’s disease present pragmatic challenges to access exercise providers with Par- showed a preference for adding multimodal exercise to their existing kinson’s disease expertise. exercise routine (ORs 2.01 to 2.19, p = 0.001 to 0.009). The finding that people with Parkinson’s disease prefer exercise to Unsurprisingly, the participants were more likely to prefer exer- be supervised by health professionals with expertise in Parkinson’s cise that did not incur much travel (ie, 10 minutes travel each way, disease highlights the importance of specialised care delivery.11 Given ORs 1.50 to 2.02, p , 0.001 to 0.03) but were less likely to prefer known cost savings when interventions have been delivered by programs with higher costs (OR 0.65 per AU$10 increase in cost, 95% physiotherapists with expertise in Parkinson’s disease,11 the current CI 0.60 to 0.71, p , 0.001) or a high frequency of sessions (ie, four results and those of previous studies indicate that in order to increase additional sessions: OR 0.41, 95% CI 0.30 to 0.55, p , 0.001). The participants also had a lower preference for group exercise compared with individual exercise (OR 0.72, 95% CI 0.54 to 0.96, p = 0.03). The extent of supervision (none, some sessions or all sessions) did not significantly influence preferences for additional exercise. Current exercise status influenced preferences for adding exercise. Those who were not currently exercising were more likely to prefer adding exercise to their current practice compared with those who
Research 53 Table 2 Table 3 Sociodemographic characteristics of the analysed participants. Clinical characteristics of the analysed participants. Characteristic Participants Characteristic Participants (n = 540) (n = 540) Age (yr), mean (SD) Parkinson’s disease duration (yr), mean (SD) a Age 65 yr, n % 69 (9) Taking antiparkinsonian medication, n % 7 (6) Sex, n (%) 398 (74) Hoehn and Yahr stage, n (%) b 511 (95) male 320 (59) 0 58 (11) female 220 (41) 1 199 (37) Domestic location, n (%) 2 108 (20) metropolitan 414 (77) 3 144 (27) regional 126 (23) 4 Domestic arrangement, n (%) 5 17 (3) lives alone 67 (12) MDS-UPDRS-II (0 to 56), mean (SD) c 5 (1) lives with spouse, carer or family 457 (85) 11.6 (7.7) lives in aged care facility other 12 (2) MDS-UPDRS-II = Movement Disorder Society-sponsored revision of the Unified Main language spoken at home, n (%) 4 (1) English Parkinson’s Disease Rating Scale. other 530 (98) a n = 14 participated in the study the year that they were diagnosed with Parkinson’s Highest level of education, n (%) a 10 (2) completed primary education disease. completed secondary education 74 (14) b n = 531. completed higher education or vocational training 51 (9) c n = 538. Employment, n (%) 414 (77) full time modulate messages about exercise benefits for people with Parkin- part time 48 (9) son’s disease to highlight the importance of specific types of exercise retired/pensioner 44 (8) (eg, aerobic5 and balance27 exercises) or to highlight particular home duties 426 (79) components of exercise (eg, the strength or balance components unemployed 17 (3) within multimodal exercise2) within a program. Such exercise types Annual income (AU$), n (%) b 5 (1) may have a higher likelihood of adoption if each exercise type is $35,000 thoroughly presented with different options, adapted to individual $35,001 to $65,000 138 (26) capabilities and integrated into multimodal exercise for people with $65,001 to $95,000 134 (25) Parkinson’s disease. $95,001 to $125,000 78 (14) $125,001 to $150,000 The vast majority of the participants in this study had mild-to- . $150,001 50 (9) moderate Parkinson’s disease and were already exercising, which Private health insurance, n yes (%) 16 (3) was similar to other studies investigating exercise in this popula- Health insurance covers physiotherapy, n yes (%) c 34 (6) tion,14,25 although it was unknown for how long participants had 465 (86) been exercising. While this sample allowed us to investigate uptake a n = 539. 389 (85) of additional exercise, which is particularly relevant for people who b n = 450. engage in some but inadequate amounts of exercise, it was difficult to c n = 460 (ie, five missing). derive specific recommendations for people who currently do not engage in exercise. Encouragingly, people with Parkinson’s disease adherence to exercise and deliver cost-effective care, exercise pre- not currently exercising were willing to begin exercising, regardless of scription for this population should be made by health professionals disease severity. Therefore, it is vital that future studies specifically with expertise in exercise prescription and Parkinson’s disease investigate this group of people in order to obtain greater under- management. Educating local gym instructors about Parkinson’s standing of initiatives with potential to be effective in increasing disease may be a cost-effective way of increasing options for people uptake of exercise in non-exercisers. Future studies should also re- with Parkinson’s disease to exercise with adequate support. In cruit people with greater severity of Parkinson’s disease who may contrast, this study found that people with Parkinson’s disease were averse to adding group exercise sessions. This finding should be Table 4 interpreted in the context of current exercise habits, where 60% of the Exercise history of the analysed participants. sample already attended group classes and 14% did not exercise. It is therefore likely that offering a combined model (group and/or indi- Characteristic Participants vidual sessions) with some Parkinson’s disease-specific supervision (n = 540) will be most attractive to a broad range of people with Parkinson’s disease. Performs exercise at least once per week, n (%) 467 (86) Already exercising 300 minutes/week, n (%) a 233 (43) This study also provided new insights into the preferences of Exercise type, n (%) b,c people with Parkinson’s disease for participating in different types of 55 (12) exercise, with results highlighting some discrepancies between what dance 179 (38) people with Parkinson’s disease prefer and what health professionals balance exercises 150 (32) may prescribe. Men showed a clear preference for adding strength strengthening exercises 158 (34) training and women were averse to participating in aerobic exercise. aerobic exercises 330 (71) People with Parkinson’s disease of both sexes did not express any walking 129 (28) preference for participating in balance exercises, yet this type of ex- flexibility exercises 262 (56) ercise is crucial for falls prevention in people with Parkinson’s dis- multimodal exercise d ease.27 The reasons for this indifference regarding balance exercises Exercise delivery mode, n (%) e 139 (26) were not revealed in this study but may include previous experiences individual 85 (16) of overly repetitive balance training that was not sufficiently chal- group 238 (44) lenging, particularly for those at earlier disease stages. Recent evi- individual and group dence demonstrates that people with Parkinson’s disease can successfully participate in challenging and varied balance training.8 a All exercise types except for flexibility exercises were assumed to have the Additionally, people with Parkinson’s disease reported an interest in undertaking multimodal exercise, which often includes balance and potential to be of moderate intensity. other functional exercises. Therefore, healthcare providers could b n = 467 (ie, exercised at least once per week). c Participants reported 1 session of an exercise type per week. Percentages were calculated using the denominator of n = 467 participants who reported exercising at least once per week, so the total exceeds 100% as many participants reported doing multiple types of exercises. d Combines 2 types of exercise. e n = 462 (ie, five missing out of n = 467 who report exercising).
54 Paul et al: Exercise preferences in Parkinson’s disease Table 5 Preferences of people with Parkinson’s disease for adding extra exercise to their current exercise routine. Attributes b (95% CI) OR (95% CI) p Random parameters 1.12 (0.70 to 1.80) , 0.0001 1.01 (0.67 to 1.52) Constant (no additional exercise) 3.13 (2.16 to 4.11) 1.34 (0.80 to 2.26) 0.63 Exercise type (versus dance) 2.00 (1.23 to 3.26) 0.96 0.12 (20.36 to 0.59) 0.33 (0.15 to 0.73) 0.27 balance 2 female 0.01 (20.40 to 0.42) 0.81 (0.44 to 1.49) 0.005 balance 2 male 0.29 (20.22 to 0.82) 1.22 (0.72 to 2.07) 0.006 strengthening 2 female 1.35 (0.88 to 2.07) 0.49 strengthening 2 male 0.70 (0.21 to 1.18) 2.01 (1.19 to 3.38) 0.47 aerobic 2 female 21.10 (21.89 to 20.31) 2.19 (1.37 to 3.50) 0.17 aerobic 2 male 20.21 (20.83 to 0.40) 0.009 walking 2 female 1.84 (1.12 to 3.02) 0.001 walking 2 male 0.20 (20.34 to 0.73) 2.02 (1.36 to 3.02) multimodal 2 female 0.30 (20.13 to 0.73) 1.50 (1.04 to 2.17) 0.02 multimodal 2 male 0.70 (0.17 to 1.22) 0.97 (0.62 to 1.52) 0.0005 Travel time per session (min) each way (versus 60 min) 0.79 (0.32 to 1.25) 0.70 (0.48 to 1.01) ,5 0.03 5 0.61 (0.11 to 1.10) 0.72 (0.54 to 0.96) 0.90 10 0.71 (0.31 to 1.10) 1.02 (0.75 to 1.39) 0.06 15 0.41 (0.04 to 0.78) 30 20.03 (20.48 to 0.42) 1.85 (1.61 to 2.13) 0.03 Delivery (versus individual sessions only) 20.36 (20.73 to 0.01) 0.88 group sessions 1.45 (1.26 to 1.67) group and individual sessions 20.33 (20.62 to 20.04) , 0.0001 Effect on motor symptoms and physical function 0.02 (20.28 to 0.33) 0.65 (0.60 to 0.71) per level of improvement a , 0.0001 Effect on overall feeling of wellbeing 0.62 (0.48 to 0.76) 1.14 (0.80 to 1.62) per level of improvement a 1.00 (0.75 to 1.35) , 0.0001 Out-of-pocket costs 0.37 (0.23 to 0.51) 0.41 (0.30 to 0.55) per AU$10 increase 0.47 20.43 (20.15 to 20.34) 0.59 (0.34 to 1.00) 0.98 0.76 (0.51 to 1.12) , 0.0001 Non-random parameters 0.88 (0.60 to 1.31) 1.32 (0.74 to 2.35) 0.05 45-min sessions (n/wk) (versus 1 session) 0.13 (20.22 to 0.48) 1.51 (1.04 to 2.20) 0.17 2 0.00 (20.29 to 0.30) 0.53 3 20.90 (21.19 to 20.60) 0.92 (0.60 to 1.40) 0.34 4 0.77 (0.47 to 1.25) 0.03 20.53 (21.06 to 20.00) 1.91 (1.18 to 3.09) Location of exercise (versus at home) 20.28 (20.66 to 0.11) 1.14 (0.87 to 1.51) 0.68 local neighbourhood 20.13 (20.52 to 0.27) 1.51 (1.08 to 2.11) 0.29 hospital or health centre/practice 0.28 (20.30 to 0.85) 0.009 community centre or facility 1.12 (0.74 to 1.71) 0.35 multiple locations including home 0.41 (0.04 to 0.79) 0.02 multiple locations excluding home 20.09 (20.51 to 0.33) 0.58 Supervisor’s expertise (versus no supervision) 20.26 (20.75 to 0.22) family member, friend or carer 0.009 qualified fitness, exercise or dance instructor without PD expertise 0.65 (0.16 to 1.13) 0.11 qualified fitness, exercise or dance instructor with PD expertise 0.13 (20.14 to 0.41) physiotherapist without PD expertise 0.41 (0.08 to 0.75) 0.24 physiotherapist with PD expertise 0.54 0.12 (20.30 to 0.53) 0.39 Extent of supervision per level increase in supervision b Sociodemographic characteristics (random parameter) 0.55 (0.14 to 0.97) 1.74 (1.15 to 2.63) 20.26 (20.57 to 0.06) 0.77 (0.57 to 1.06) Current exercise (versus base level 300 min/wk) c not currently exercising currently exercising but , 300 min/wk Sociodemographic characteristics (non-random parameters) Age . 65 years (vs , 65) 20.15 (20.40 to 0.10) 0.86 (0.67 to 1.11) 0.07 (20.16 to 0.30) 1.07 (0.85 to 1.35) Annual income . AU$65,000 per year (vs , $65,000) 20.01 (20.23 to 0.21) 0.99 (0.79 to 1.24) PD severity high (vs low) d Model fit: Akaike Information Criterion = 0.888, log likelihood 22327.5, pseudo r2 = 0.376. MDS-UPDRS-II = Movement Disorder Society-sponsored revision of the Unified Parkinson’s Disease Rating Scale, PD = Parkinson’s disease. a Gain was quantified as: none, small, moderate or large. b Extent of supervision was quantified as: none, some or complete. c Flexibility exercises were excluded as this type of exercise is unlikely to be of at least moderate intensity. d Disease severity was categorised at the mean value of the MDS-UPDRS-II (ie, low disease severity was scored as 10 and high disease severity as 11). require more support to exercise safely and to probe exercise habits in structured exercise yields the most benefit for improving physical function.30 Future studies should investigate the benefits and relation to fall frequency. adherence of people with Parkinson’s disease for incorporating incidental physical activity in addition to, or as partial substitution The highly educated sample is likely a limitation of this study of, structured exercise programs, particularly for people with more because high education levels are known to influence adherence to severe disease31 who may struggle to meet physical activity exercise,13 probably through better health literacy.28 Additionally, recommendations.22 since data were self-reported, we were unable to accurately deter- In conclusion, this study was the first to examine, from the perspective of people with Parkinson’s disease, which features of mine the intensity of reported exercises. Older adults and people exercise programs were associated with willingness to perform more with greater levels of disability may struggle to meet the recom- exercise. While the majority of this large sample of people with mended moderate-to-vigorous levels of physical activity and exer- cise.22 Although lower intensity and incidental physical activity in older adults is associated with improved health outcomes,29
Research 55 Parkinson’s disease exercised regularly, more than half did not meet Parkinson disease: a phase 2 randomized clinical trial. JAMA Neurol. 2018;75:219– exercise recommendations. This suggests a need to account for the 226. exercise preferences of people with Parkinson’s disease to increase 6. Ellis T, Rochester L. Mobilizing Parkinson’s Disease: The Future of Exercise. exercise uptake (eg, by providing multimodal exercise programs J Parkinsons Dis. 2018;8:S95–S100. designed by exercise providers with Parkinson’s disease expertise to 7. Fang X, Han D, Cheng Q, Zhang P, Zhao C, Min J, et al. Association of levels of meet the preferences of the individual), while at the same time tar- physical activity with risk of Parkinson Disease: a systematic review and meta- geting vital impairments. Additionally, future studies should investi- analysis. JAMA Netw Open. 2018;1:e182421. gate whether measures could be taken to provide local exercise 8. Lofgren N, Conradsson D, Joseph C, Leavy B, Hagstromer M, Franzen E. Factors facilities with personnel knowledgeable in Parkinson’s disease to associated with responsiveness to gait and balance training in people with Par- reduce travel time, cost, and thereby improve the availability of ex- kinson Disease. J Neurol Phys Ther. 2019;43:42–49. ercise for people with Parkinson’s disease. 9. Cavanaugh JT, Ellis TD, Earhart GM, Ford MP, Foreman KB, Dibble LE. Capturing ambulatory activity decline in Parkinson’s disease. J Neurol Phys Ther. 2012;36:51– What was already know on this topic: Although exercise is 57. beneficial and cost-effective for people with Parkinson’s disease, 10. Farag I, Sherrington C, Hayes A, Canning CG, Lord SR, Close JC, et al. Economic most are less active than the general population (even at earlier evaluation of a falls prevention exercise program among people with Parkinson’s stages of the disease) and their physical activity levels progres- disease. Mov Disord. 2016;31:53–61. sively decline. 11. Ypinga JHL, de Vries NM, Boonen L, Koolman X, Munneke M, Zwinderman AH, et al. What this study adds: People with Parkinson’s disease are Effectiveness and costs of specialised physiotherapy given via ParkinsonNet: a more willing to participate in exercise programs that cost less, retrospective analysis of medical claims data. Lancet Neurol. 2018;17:153–161. involve less travel, are supervised by qualified professionals with 12. van der Eijk M, Nijhuis FA, Faber MJ, Bloem BR. Moving from physician-centered Parkinson's disease expertise, and provide physical or psycho- care towards patient-centered care for Parkinson’s disease patients. Parkinsonism logical benefits. Men were more likely to consider adopting Relat Disord. 2013;19:923–927. strengthening exercises, whereas women were averse to 13. Ellis T, Cavanaugh JT, Earhart GM, Ford MP, Foreman KB, Fredman L, et al. Factors adopting aerobic exercise. To facilitate more exercise among associated with exercise behavior in people with Parkinson disease. Phys Ther. people with Parkinson’s disease, healthcare services should 2011;91:1838–1848. provide programs aligned with these factors. 14. Ellis T, Boudreau JK, DeAngelis TR, Brown LE, Cavanaugh JT, Earhart GM, et al. Barriers to exercise in people with Parkinson disease. Phys Ther. 2013;93:628–636. Footnotes: a NGENE design software, ChoiceMetrics, Sydney, 15. O’Brien C, Clemson L, Canning CG. Multiple factors, including non-motor impair- Australia. b NLOGIT Version 5.0, Econometric Software, Sydney, ments, influence decision making with regard to exercise participation in Par- Australia. kinson’s disease: a qualitative enquiry. Disabil Rehabil. 2016;38:472–481. 16. McFadden DL. Conditional Logit Analysis of Qualitative Choice Behavior. In: Ethics approval: This study was approved by the University of Zarembka P, ed. Frontiers in Econometrics. NY: Wiley; 1973:105–142. Sydney Human Research Ethics Committee (project number 2016/ 17. Thaler R. Toward a positive theory of consumer choice. J Econ Behav Organ. 199). As data were collected via an anonymous survey, informed 1980;1:39–60. consent was implied upon submission or return of a completed 18. Bridges JF, Hauber AB, Marshall D, Lloyd A, Prosser LA, Regier DA, et al. Conjoint survey. analysis applications in health–a checklist: a report of the ISPOR Good Research Practices for Conjoint Analysis Task Force. Value Health. 2011;14:403–413. Competing interests: Nil. 19. Goetz CG, Tilley BC, Shaftman SR, Stebbins GT, Fahn S, Martinez-Martin P, et al. Source(s) of support: This study was funded by a Parkinson’s NSW Movement Disorder Society-sponsored revision of the Unified Parkinson’s Disease grant (2015-16). Rating Scale (MDS-UPDRS): scale presentation and clinimetric testing results. Mov Acknowledgements: The authors wish to thank the people with Disord. 2008;23:2129–2170. Parkinson’s disease who participated in this research. 20. Wallen MB, Franzen E, Nero H, Hagstromer M. Levels and patterns of physical Provenance: Not invited. Peer reviewed. activity and sedentary behavior in elderly people with mild to moderate Parkinson Correspondence: Serene S Paul, Discipline of Physiotherapy, Disease. Phys Ther. 2015;95:1135–1141. School of Health Sciences, Faculty of Medicine and Health, University 21. Conradsson D, Nero H, Lofgren N, Hagstromer M, Franzen E. Monitoring training of Sydney, Sydney, Australia. Email: [email protected] activity during gait-related balance exercise in individuals with Parkinson’s dis- ease: a proof-of-concept-study. BMC Neurol. 2017;17:19. References 22. Australian Institute of Health and Welfare. Physical activity across the life stages. Cat. no. PHE 225. Canberra, Australia. 2018. 1. McLean G, Hindle JV, Guthrie B, Mercer SW. Co-morbidity and polypharmacy in 23. Hensher DA, Rose JM, Greene WH. Applied Choice Analysis. 2nd ed. Cambridge: Parkinson’s disease: insights from a large Scottish primary care database. BMC Cambridge University Press; 2015. Neurol. 2017;17:126. 24. Tomlinson CL, Patel S, Meek C, Herd CP, Clarke CE, Stowe R, et al. Physiotherapy intervention in Parkinson’s disease: systematic review and meta-analysis. Br J Med. 2. Mak MK, Wong-Yu IS, Shen X, Chung CL. Long-term effects of exercise and physical 2012;345:e5004. therapy in people with Parkinson disease. Nat Rev Neurol. 2017;13:689–703. 25. Oguh O, Eisenstein A, Kwasny M, Simuni T. Back to the basics: Regular exercise matters in Parkinson’s disease: Results from the National Parkinson Foundation QII 3. Wu PL, Lee M, Huang TT. Effectiveness of physical activity on patients with Registry study. Parkinsonism Relat Disord. 2014;20:1221–1225. depression and Parkinson’s disease: a systematic review. PLoS One. 2017;12: 26. Wallen MB, Hagstromer M, Conradsson D, Sorjonen K, Franzen E. Long-term effects e0181515. of highly challenging balance training in Parkinson’s disease-a randomized controlled trial. Clin Rehabil. 2018;32:1520–1529. 4. Silveira CRA, Roy EA, Intzandt BN, Almeida QJ. Aerobic exercise is more effective 27. Shen X, Wong-Yu IS, Mak MK. Effects of exercise on falls, balance, and gait ability in than goal-based exercise for the treatment of cognition in Parkinson’s disease. Parkinson’s disease: a meta-analysis. Neurorehabil Neural Repair. 2016;30:512–527. Brain Cogn. 2018;122:1–8. 28. Liu YB, Liu L, Li YF, Chen YL. Relationship between health literacy, health-related behaviors and health status: a survey of elderly Chinese. Int J Environ Res Public 5. Schenkman M, Moore CG, Kohrt WM, Hall DA, Delitto A, Comella CL, et al. Effect of Health. 2015;12:9714–9725. high-intensity treadmill exercise on motor symptoms in patients with de novo 29. Buman MP, Hekler EB, Haskell WL, Pruitt L, Conway TL, Cain KL, et al. Objective light-intensity physical activity associations with rated health in older adults. Am J Epidemiol. 2010;172:1155–1165. 30. Brach JS, Simonsick EM, Kritchevsky S, Yaffe K, Newman AB, Health. Aging and Body Composition Study Research Group. The association between physical function and lifestyle activity and exercise in the health, aging and body compo- sition study. J Am Geriatr Soc. 2004;52:502–509. 31. Snider J, Müller MLTM, Kotagal V, Koeppe RA, Scott PJ, Frey KA, et al. Non- exercise physical activity attenuates motor symptoms in Parkinson disease independent from nigrostriatal degeneration. Parkinsonism Relat Disord. 2015;21:1227–1231.
Journal of Physiotherapy 67 (2021) 5–11 KEY WORDS 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 Neck pain Invited Topical Review Physical therapy Literature review Physiotherapy management of neck pain Arianne P Verhagen Discipline of Physiotherapy, Graduate School of Health, University of Technology Sydney, Sydney, Australia [Verhagen AP (2021) Physiotherapy management of neck pain. Journal of Physiotherapy 67:5–11] © 2020 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 In 2017, the Global Burden of Disease study calculated that neck pain had: an age-standardised point prevalence of 3,551/100,000 Neck pain is defined as pain in the neck with or without pain people, with a 95% uncertainty interval (UI) from 3,140 to 3,978; and referred into one or both upper limbs that lasts for at least 1 day.1,2 In an annual incidence of 807/100,000 people (95% UI 714 to 913).2 Both 2008, the Task Force on Neck Pain defined the anatomical region of the incidence and prevalence of neck pain increased with age and neck pain (Figure 1).3 This anatomical definition was developed based were greater among females than males. The prevalence of neck pain on consensus because the Task Force found more than 300 case did not change substantially between 1990 and 2010.2 Up to 70% of definitions for neck pain. People with neck pain may also have people can expect to experience some neck pain in their lifetime, accompanying headache or shoulder pain, but neck pain is the pri- although in most cases neck pain will not seriously interfere with mary complaint. daily activities and participation.2,3,5 Categorisation The incidence of serious pathology (Grade IV) is low, up to 2% in referred patients,1 while the incidence of cervical radiculopathy In 2008, the Task Force on Neck Pain proposed a classification of (Grade III) ranges from 6.3 to 21 per 10,000 people.6 This wide people with neck pain into four categories.3 This classification is range is due to variation in the definitions of ‘radiating or radicular based on the Quebec Task Force classification of whiplash.4 The only symptoms’ that are used in practice and research.6,7 Often the difference between both classifications is that the Quebec Task Force definition is not limited to ‘the presence of neurological signs or also defined a Grade 0, which means that there was a trauma present sensory deficits’ but includes only radiating symptoms. According but no pain. In the Task Force on Neck Pain classification, Grade I to III to the Task Force on Neck Pain, these patients cannot be regarded neck pain is regarded as non-specific neck pain (Table 1).3 Grade I and as having a Grade III neck pain. The vast majority of patients have II neck pain are distinguished by the amount of interference with Grade I or II neck pain, often estimated to be . 90% of patients.1 activities of daily living. A person with Grade III neck pain (also called cervical radiculopathy) also has objective neurologic signs (such as There are several factors that indicate an increased risk of devel- decreased deep tendon reflexes, weakness or sensory deficits) and oping neck pain. The most important of these prognostic factors are: positive findings on provocation and reduction tests. People with trauma, work-related factors (low job satisfaction, poor perceived Grade IV neck pain suffer from major pathologies, and this grade work support, high work stress levels), psychological factors (self- corresponds with specific neck pain. perceived depression, poor psychological health) and smoking.8–10 Degeneration of the cervical disc does not appear to be a risk fac- Incidence and prevalence tor.8 The economic burden of neck pain has not been evaluated extensively.1,2 In the Global Burden of Disease study, out of the 291 conditions studied, neck pain was found to rank 21st in terms of overall burden Diagnosis and assessment and fourth in terms of overall disability; therefore, neck pain is a serious public health problem in the general population.1,2 Among all The diagnostic process within physiotherapy practice consists of musculoskeletal disorders, low back pain (ranked first) and neck pain history taking, physical examination and, if deemed necessary, (ranked fourth) are the most common worldwide. Nevertheless, the (referral for) diagnostic imaging. The aim of history taking is to find amount of research involving people with low back pain greatly information that informs the patient’s prognosis and whether the outweighs that involving people with neck pain.2 patient belongs to a subgroup that warrants a different management strategy. History taking leads to an initial hypothesis, which can be https://doi.org/10.1016/j.jphys.2020.12.005 1836-9553/© 2020 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/).
6 Verhagen: Physiotherapy management of neck pain Figure 1. Posterior and lateral views of the anatomic region of the neck used in the Task Force on Neck Pain definition of neck pain.3 confirmed or excluded via physical examination (or diagnostic im- Historically patients with trauma-related neck pain (previously aging). A flow chart of the diagnostic process is presented in Figure 2. called a whiplash or a whiplash-associated disorder) are regarded as a subgroup of patients with neck pain. The distinguishing characteristic Red flags is that they have experienced a trauma (often a car accident). Greater severity of pain at baseline or at consultation is associated with a First, a physiotherapist has to rule out serious pathology or red poorer prognosis.17,18 flags. Red flags are patterns of signs or symptoms (warning signals) that may indicate serious pathology requiring further medical diag- Patients with work-related neck pain (ie, neck pain due to work that nostic procedures. decreases on the weekend or during periods off work) are also regarded as a subgroup of patients with neck pain, as they seem to have a poorer The most well-known screening methods for a fracture among prognosis.17,19 There is a wide variety of work-related prognostic factors patients with neck pain after trauma are the Canadian cervical spine that have shown to be related to this poorer prognosis. rule (C-Spine) and the National Emergency X-Radiography Utilization Study (NEXUS).11,12 According to a systematic review, the sensitivity of To date, it is unclear whether patients with cervicogenic headache both methods is high; therefore, for patients with a negative result for (ie, headache that typically develops after neck pain and is often either screening method, the possibility of fracture could reliably be exacerbated by neck movements) are a subgroup of headache pa- excluded (high Sensitivity and a Negative test rules Out the diagnosis; tients or neck pain patients.20–22 Unfortunately, data on prognosis SnNOut).11 No red flags for a malignancy have been evaluated. and prognostic factors for this subgroup are lacking. Other known screening tests are tests for higher cervical (liga- In the literature, there is no consensus regarding the classification of ment) instability or arteria vertebralis insufficiency. The aim of these patients with or without cervical radiculopathy on the basis of symp- tests is to identify patients at high risk of a serious complication when toms and neurological investigation, except that the patients have pain receiving cervical spinal manipulation. However, these screening radiating to the arm, often following a radicular pattern.6,7 According to methods have been poorly researched and to date have not been the definition of the Task Force on Neck Pain, these patients have validated.13,14 Despite this, most guidelines for manual therapists and neurological symptoms or sensory deficits (such as sensory loss or chiropractors recommend performing these screening tests.15,16 altered reflexes).3 One small study showed that a loss of sensation and pain radiating to the elbow both have a high specificity and can History taking therefore be used to diagnose (or rule in) a cervical radiculopathy (high Specificity and a Positive test rules In the diagnosis; SpPIn).23 Having The next step in the diagnostic process is to look at prognostic and radiating symptoms without these neurological symptoms and sen- differentiating factors in order to make an estimation of the patient’s sory deficits is regarded as Grade II neck pain. The prognosis of patients prognosis or subgroup, relevant to the management strategy with cervical radiculopathy is favourable; the majority of patients (Figure 2). recover within 4 to 6 months.24 The categories discussed above (eg, trauma-related neck pain, cervicogenic headache, neck pain with Table 1 radiculopathy and work-related neck pain) are similar to the categories Grades of neck pain defined by the Task Force on Neck Pain.3 in the clinical practice guideline by Blanpied et al (neck pain with movement control disorders (including whiplash-associated disor- Grade Explanation ders), neck pain with headaches, neck pain with radiating pain, and neck pain with mobility deficit).22 I Neck pain and associated disorders with no signs or symptoms suggestive of major structural pathology and no or minor interference with activities of Physical examination daily living Physical examination may consist of inspection of the posture, II No signs or symptoms of major structural pathology, but major interference palpation, measuring the range of motion, measuring muscle with activities of daily living strength, testing reflexes, testing sensation and specific tests. The choice of which of these physical examination procedures will be III No signs or symptoms of major structural pathology, but presence of performed depends on the findings from the history taking and neurologic signs such as decreased deep tendon reflexes, weakness, or whatever diagnosis or diagnoses those findings suggest. The aim of sensory deficits in the upper extremity physical examination is to confirm or rule out that initial diagnosis. IV Signs or symptoms of major structural pathology, which include (but are Unfortunately, little is known regarding the diagnostic value of not limited to) fracture, vertebral dislocation, injury to the spinal cord, general physical examination for patients with neck pain. There is infection, neoplasm, or systemic disease including inflammatory arthropathies
Invited Topical Review 7 Neck pain Red flags? Yes Referral No Prognostic/diagnostic factors Trauma Work Headache Sensory loss Radiating below elbow Trauma- Work-related Neck pain Cervicogenic Cervical related headache radiculopathy Figure 2. Flow chart of the diagnostic process. also no information on the diagnostic value of specific tests to However, imaging is usually discouraged unless there is severe differentiate between neck pain patients and patients with trauma- trauma,36 mainly because diagnostic imaging also produces a high related neck pain, work-related neck pain or cervicogenic headache. number of false positives. In a study with 1,211 relatively healthy and To confirm or rule out the initial diagnosis of cervical radiculopathy, asymptomatic participants who received diagnostic imaging using guidelines advise specific tests.25 The most well-known specific tests MRI, over 87% of the participants presented with a ‘bulging disc’ and are: Spurling’s test, the traction test, the Upper Limb Tension Test and 5.3% with a spinal cord compression.37 the shoulder abduction test. Several studies show that the Spurling’s test and the traction test both have a relatively high specificity. Prognosis and course Specificity varies from 89 to 100% for Spurling’s test and from 90 to 97% for the traction test.26,27 Therefore, both tests seem useful to In 2008, the Task Force on Neck Pain estimated that 50 to 85% of confirm the initial hypothesis (SpPIn). On the other hand, the Upper people with neck pain do not make a full recovery, indicating that Limb Tension Test can be used to rule out a cervical radiculopathy due neck pain has an episodic and recurrent character.8 In addition, a to a high sensitivity varying from 87 to 93% (SnNOut).26 The repro- systematic review found that, in people with acute neck pain, the ducibility of the specific tests (reported as a kappa value) ranges from pooled mean pain score decreased by 45% during the first 6.5 weeks, 13 to 93%.27 Although neurological testing of dermatomes and myo- but after that no further reduction in pain was found.38 In this study tomes is recommended, its diagnostic validity has not been assessed. the prognosis was calculated based on the recovery rates from cohort studies and from participants randomised to a control arm that did Clinical prediction rules not receive any treatment. The prognosis for patients with cervical radiculopathy is more favourable than for patients with neck pain Many clinical prediction rules exist, although most of them have without radiculopathy.24 been developed using unsatisfactory methods or have not been validated.28 A systematic review found a total of 99 prediction models In general, several factors have been identified in the literature for neck pain or trauma-related neck pain, of which three were that are likely related to a poorer prognosis: previous episodes of promising enough for use in physiotherapy and other primary care neck pain, concurrent low back pain, concurrent headaches, poor settings.28 One of the promising models was developed for people health, psychological factors (such as anxiety, worry, frustration and with neck pain, and two specifically for people with trauma-related depression) and work-related symptoms (such as low job satisfaction, neck pain.29–31 A consistent factor related to high likelihood of re- high physical job demands and little influence on work situa- covery included in all three models was age (, 35 years). Low initial tion).17,19,39 In contrast, younger age, an active coping style and disability score (, 32%, assessed with the Neck Disability Index) optimistic outlook appear to be related to a favourable prognosis.17 seemed relevant for the trauma-related neck pain models only.28 Physiotherapy treatment The Keele Subgroup Targeted Treatment (STarT) Back Screening Tool was initially developed for people with acute low back pain, but The majority of neck pain guidelines on diagnosis and treatment was recently modified and validated for people with neck pain.32 It of patients with neck pain recommend a combination of manual aims to stratify people with neck pain into low, medium and high risk therapy, exercise and education as the preferred evidence-based for chronic complaints combined with a targeted treatment for each physiotherapy treatments.25,40 Massage might be beneficial (incon- category, but the predictive validity is low. sistent evidence) and psychological (behavioural) treatment and multidisciplinary treatment are effective in some subgroups of pa- Diagnostic imaging tients. All other interventions lack a clear evidence base. Various guidelines recommend not to refer people with neck pain Education to imaging. Despite this, diagnostic imaging is sometimes used to confirm or rule out a specific pathology – most often a cervical rad- Education is defined as a process of enabling individuals to make iculopathy (cervical disc herniation). The sensitivity and specificity of informed decisions about their personal health-related behaviour.41 various imaging techniques varies from 27 to 96%.33 Ruling out a According to a Cochrane review, patient education (or the provision fracture can best be done using a computed tomography (CT) scan, of information) is regarded as an essential part of communication which has a sensitivity of 96 to 99%.34 Specific magnetic resonance between the physiotherapist and the patient.42 Unfortunately, that imaging (MRI) techniques seem to be valid for diagnosing a cervical review failed to show evidence that education is beneficial in the disc herniation, with sensitivity and specificity between 95 and 97%.35 treatment of neck pain patients. A more recent systematic review
8 Verhagen: Physiotherapy management of neck pain concluded that structured patient education alone is equally benefi- Mobilisation, manipulation, advice and exercise cial compared with other conservative interventions for patients with neck pain with or without traumatic origin.41 The patient educational Mobilisations and manipulations are rarely used as a unimodal interventions that are evaluated and recommended by the guidelines intervention; more often they are administered in combination with are: reassuring patients that the pain is not a serious condition; advice and/or exercises.51–53 The combined treatment of exercise and providing information on pain and prognosis, including information manipulations seem to be more effective (moderate-quality evidence) that imaging is not recommended; advising to stay active; and than exercises alone for immediate pain, but not on all other out- educating about self-care, exercises and (stress) coping skills.36,41,42 comes for people with neck pain (Figure 4, with a detailed forest plot available in Appendix 2 on the eAddenda).52 Unfortunately, the effect Exercise size is small (SMD 0.15, 95% CI 0.00 to 0.30) and there is probably not a clinically relevant benefit of adding mobilisations or manipulations Physical exercises vary widely from general land-based or aquatic to exercises. exercise to neck-specific endurance, strength, stretching or McKenzie exercises. The most recent Cochrane review on exercises for me- Massage chanical neck disorders found that a wide variety of exercises had been evaluated, varying from breathing exercises to strength and Massage therapy is one of the oldest treatment strategies for endurance exercises.43,44 In this review, the quality of the evidence musculoskeletal pain. It involves mobilisation and manipulation of was categorised as very low, low, moderate or good, according to the the soft tissues of the body through touch.54 There is a wide spectrum Grading of Recommendations, Assessment, Development and Evalu- of techniques that fall under the umbrella term of massage therapy. ations (GRADE) system.45 The review concluded that when exercise The different techniques vary in the manner in which touch is was compared with no treatment or placebo, or evaluated as an applied, as well as the amount of pressure that is applied.54 Massage additional treatment: strength, endurance and stabilising exercises techniques commonly used by physiotherapist are known as con- were beneficial in chronic neck pain (moderate-quality evidence); ventional western massage and were found to be beneficial (in one only strength and endurance exercises were beneficial in chronic small study) in the treatment of patients with neck pain compared cervicogenic headaches (moderate-quality evidence); and there was a with no treatment or placebo.49 small benefit of stretching, strengthening and stabilisation exercises in acute cervical radiculopathy (low-quality evidence). The stand- Non-physiotherapy management ardised effect sizes varied from 0.3 to 0.7 (95% CI 0.1 to 1.3), which can be regarded as small to moderate effects. There were no studies that Medication evaluated exercises in patients with acute neck pain. A recent network meta-analysis showed that no specific exercise was found to People with neck pain might take over-the-counter medication, be superior in people with chronic non-specific neck pain.46 such as paracetamol or non-steroidal anti-inflammatory drugs (NSAIDs). Although physiotherapists cannot prescribe pain medica- Several researchers have assumed that changes in motor control tion, it is important for them to know the evidence about relevant in the deep cervical muscles contribute to the origin or persistence of medications in order to help patients with their questions. neck pain.47 A recent systematic review aimed to investigate this hypothesis and evaluated whether motor control exercises (ie, cranio- A recent systematic review evaluated the effectiveness of para- cervical flexion exercises) are more effective than no intervention for cetamol in patients with musculoskeletal disorders but did not people with chronic neck pain. The authors found clinically relevant identify any trials evaluating paracetamol in patients with neck benefits (standardised effect sizes between 0.33 and 0.58) on pain pain.55 A few randomised trials evaluating NSAIDs for patients with and disability.47 neck pain exist; they showed NSAIDs to be better than placebo,56 equally effective as muscle relaxants or acupuncture,57,58 but less Mobilisation and manipulation beneficial than spinal manipulation and exercises.59 The only high- quality study on NSAIDs in (sub)acute neck pain patients (72 pa- Physiotherapists often offer ‘manual therapy’, aiming to improve tients) found diclofenac gel (a topical NSAID) to be more effective spinal joint motion and restore range of motion. Manual therapy than placebo in reducing pain.60 This evidence is supported by an consists of various techniques, including mobilisations and manipu- overview of Cochrane reviews in patients with chronic pain on topical lations. Mobilisations are defined as using low-grade/velocity, small- NSAIDs.60 This overview found that topical diclofenac was effective amplitude or large-amplitude passive movement techniques within with a number needed to treat of 9.8, based on data from six trials the patient’s range of motion and within the patient’s control. with 2,343 participants (moderate quality evidence). Oral NSAIDs also Manipulation is defined as a localised high-velocity and low- seem to be effective in patients with spinal pain compared with amplitude force directed at specific cervical or thoracic spinal seg- placebo (MD 16 mm on a 100-mm visual analogue scale, 95% CI 12 to ments near the end of the patient’s range of motion and without their 21), which was above the a priori defined 10-mm threshold for control. clinical relevance.61 A Cochrane review and another systematic review both found that Surgery cervical mobilisations and manipulations were equally beneficial (moderate-quality evidence) in patients with non-specific neck Patients with ongoing neck pain that is not responsive to con- pain.48,49 According to the Cochrane review, cervical manipulations servative care are frequently referred to secondary care for further show a small beneficial effect (low-quality evidence), but thoracic assessment with a chance to receive corticosteroid injections or manipulations show a larger beneficial effect when compared to an surgery. inactive treatment (moderate-quality evidence), indicating that thoracic manipulations were more beneficial than cervical manipu- No systematic review has evaluated corticosteroid injections for lations.48 A more recent systematic review evaluating the effective- neck pain, but some randomised trials exist.62–65 All of these trials ness of thoracic manipulations could not confirm this finding based evaluated corticosteroid injections in patients with cervical radicul- on two studies that directly compared cervical with thoracic ma- opathy. Only one of the studies compared injection with physio- nipulations.50 That review, on the other hand, found that thoracic therapy interventions (education, electrophysical agents, massage manipulations were more beneficial than mobilisations and standard and exercise).64 In this three-arm trial, there were no important dif- care (very low-quality evidence) with mean differences in pain on a ferences in the primary outcome (arm pain) between injections 100-mm visual analogue scale of 14 mm (95% CI 6 to 22) and 13 mm alone, physiotherapy interventions alone, or combined injections and (95% CI 4 to 22), respectively (Figure 3, with a detailed forest plot physiotherapy. available in Appendix 1 on the eAddenda).
Invited Topical Review 9 Comparison WMD (95% CI) Study Random Thoracic manipulation versus any mobilisation Cleland 2007 Masaracchio 2013 Salom-Moreno 2014 Suvarnnato 2013 Pooled Thoracic manipulation versus thoracic mobilisation Cleland 2007 Salom-Moreno 2014 Suvarnnato 2013 Pooled –50 –25 0 25 50 (0 to 100 mm) Favours thoracic Favours manipulation mobilisation Figure 3. Weighed mean difference in immediate/short-term effect of thoracic spine manipulation versus mobilisation on neck pain severity, measured on a 100-mm visual analogue scale. The upper comparison pools all studies where thoracic manipulation was compared with any mobilisation, and the lower comparison is the subgroup of three studies where the thoracic manipulation was compared with thoracic mobilisation. Modified from Masaracchio et al.50 Study SMD (95% CI) A systematic review including nine controlled studies found Akhter 2014 Random overall no important differences between surgery or conservative Celenay 2015 care in neck pain patients (very low quality evidence).66 In addition, Celenay 2016 very small differences in benefits and harms between the various Dziedzic 2005 surgical techniques were found, and no additional benefit was found Evans 2012 by adding fusion to anterior decompression techniques.67,68 Ganesh 2015a Ganesh 2015b Future research Yang 2015 Most systematic reviews discussed above found: a limited number Pooled of studies on the target intervention, studies with overall (very) small sample sizes, a high proportion of studies with high risk of bias, and –2 –1 0 1 2 marked clinical heterogeneity between the studies. These findings hamper firm conclusions from being drawn and indicate that future Favours exercise and Favours exercise research will likely change current conclusions and recommendations. manual therapy therapy Compared to low back pain, with a more or less similar burden of Figure 4. Standardised mean difference in immediate effect of combined exercise disease, neck pain is a relatively understudied condition and more research is therefore warranted.2 A recent Delphi consensus study on therapy and manual therapy versus exercise therapy alone on neck pain severity. research priorities in neck pain research concluded that the main Modified from Fredlin et al.52 research priority was to evaluate the effectiveness and cost- effectiveness of all major interventions.69 The second most impor- tant research priority was to evaluate how to best translate the research findings into clinical practice. Research into diagnostic as- sessments was priority 11 out of 15 priorities. Research on risk stratification could be conducted using clinical prediction models/rules including evaluating the impact of such rules in risk-stratified neck pain trials. These studies increase the knowl- edge on the validity of diagnostic assessment and, for example, focus on which patients might benefit from which treatment strategies.70
10 Verhagen: Physiotherapy management of neck pain Further research could evaluate the optimal characteristics and 13. Hutting N, Verhagen AP, Vijverman V, Keesenberg MD, Dixon G, Scholten- dose of the most frequently used interventions. This may help to Peeters GG. Diagnostic accuracy of premanipulative vertebrobasilar insufficiency reduce the heterogeneity between studies.71 tests: a systematic review. Man Ther. 2013;18:177–182. The Global Burden of Disease study on neck pain concludes that: 14. Hutting N, Scholten-Peeters GG, Vijverman V, Keesenberg MD, Verhagen AP. Diagnostic accuracy of upper cervical spine instability tests: a systematic review. increasing population awareness about neck pain and its risk factors Phys Ther. 2013;93:1686–1695. as well as the importance of early detection and management is warranted to reduce the future burden of this condition.2 15. International Federation of Orthopaedic Manipulative Physical Therapists (IFOMPT). Cervical Framework. https://www.ifompt.org/site/ifompt/IFOMPT%20Cervical% This indicates a global patient/population education or mass 20Framework%20final%20September%202020.pdf. Accessed 1 November, 2020. media campaign. A recent systematic review on mass media cam- paigns in low back pain suggested that these campaigns might be 16. Australian Physiotherapy Association. Clinical guide to safe manual therapy prac- effective for changing health beliefs.72 tice in the cervical spine. www.physiotherapy.asn.au/cervicalspine. Accessed 1 November, 2020. Conclusion 17. Walton DM, Macdermid JC, Giorgianni AA, Mascarenhas JC, West SC, Zammit CA. Physiotherapists frequently see patients with neck pain in clinical Risk factors for persistent problems following acute whiplash injury: update of a practice; it is one of the four musculoskeletal disorders that have a systematic review and meta-analysis. J Orthop Sports Phys Ther. 2013;43:31–43. major burden on society. Stratifying patients in either subgroups or based on their prognosis (prediction of recovery) might be useful in 18. Williamson E, Williams MA, Gates S, Lamb SE. Risk factors for chronic disability in a guiding physiotherapy management decisions. Manual therapy, ex- cohort of patients with acute whiplash associated disorders seeking physiotherapy ercise and education – usually in combination – seem to be the treatment for persisting symptoms. Physiotherapy. 2015;101:34–43. preferred evidence-based physiotherapy treatments for most patients with neck pain. Nevertheless, most interventions and management 19. Verwoerd M, Wittink H, Maissan F, de Raaij E, Smeets RJEM. Prognostic factors for strategies are not based on firm evidence and effect sizes are small. persistent pain after a first episode of nonspecific idiopathic, non-traumatic neck Clinicians need to be aware of this and keep abreast of new findings pain: a systematic review. Musculoskelet Sci Pract. 2019;42:13–37. in the many avenues of research into the management of neck pain. 20. Becker WJ. Cervicogenic headache: evidence that the neck is a pain generator. eAddenda: Appendices 1 and 2 can be found online at https://doi. Headache. 2010;50:699–705. org/10.1016/j.jphys.2020.12.005. 21. Vincent MB. Cervicogenic headache: the neck is a generator: con. Headache. Ethics approval: Nil. 2010;50:706–709. Competing interests: Nil. Source(s) of support: Nil. 22. Blanpied PR, Gross AR, Elliott JM, Devaney LL, Clewley D, Walton DM, et al. Neck Acknowledgements: Nil. pain: revision 2017: clinical practice guidelines linked to the international classi- Provenance: Invited. Peer reviewed. fication of functioning, disability and health from the orthopaedic section of the Correspondence: Arianne P Verhagen, Discipline of Physiotherapy, American Physical Therapy Association. J Orthop Sports Phys Ther. 2017;47:A1-A83. Graduate School of Health, University of Technology Sydney, Sydney, Australia. Email: [email protected] 23. Wainner RS, Fritz JM, Irrgang JJ, Boninger ML, Delitto A, Allison S. Reliability and diagnostic accuracy of the clinical examination and patient self-report measures References for cervical radiculopathy. Spine. 2003;28:52–62. 1. Hoy D, March L, Woolf A, Blyth F, Brooks P, Smith E, et al. The global burden of neck 24. Wong JJ, Côté P, Quesnele JJ, Stern PJ, Mior SA. The course and prognostic factors of pain: estimates from the global burden of disease 2010 study. Ann Rheum Dis. symptomatic cervical disc herniation with radiculopathy: a systematic review of 2014;73:1309–1315. the literature. Spine J. 2014;14:1781–1789. 2. Safiri S, Kolahi AA, Hoy D, Buchbinder R, Mansournia MA, Bettampadi D, et al. 25. Parikh P, Santaguida P, Macdermid J, Gross A, Eshtiaghi A. Comparison of CPG’s for Global, regional, and national burden of neck pain in the general population, 1990- the diagnosis, prognosis and management of non-specific neck pain: a systematic 2017: systematic analysis of the Global Burden of Disease Study 2017. BMJ. review. BMC Musculoskelet Disord. 2019;20:81. 2020;368:m791. 26. Thoomes EJ, van Geest S, van der Windt DA, Falla D, Verhagen AP, Koes BW, et al. 3. Haldeman S, Carroll L, Cassidy JD, Schubert J, Nygren A. The Bone and Joint Decade Value of physical tests in diagnosing cervical radiculopathy: a systematic review. 2000-2010 Task Force on Neck Pain and Its Associated Disorders: executive sum- Spine J. 2018;18:179–189. mary. Spine. 2008;33:S5–S7. 27. Lemeunier N, da Silva-Oolup S, Chow N, Southerst D, Carroll L, Wong JJ, et al. 4. Spitzer WO. Scientific monograph of the Quebec Task Force on Whiplash-Associ- Reliability and validity of clinical tests to assess the anatomical integrity of the ated Disorders: Redefining ‘whiplash’ and its management. Spine. 1995;20:1–73. cervical spine in adults with neck pain and its associated disorders: Part 1-A systematic review from the Cervical Assessment and Diagnosis Research Evalua- 5. Cohen SP, Hooten WM. Advances in the diagnosis and management of neck pain. tion (CADRE) Collaboration. Eur Spine J. 2017;26:2225–2241. BMJ. 2017;358:j3221. 28. Wingbermühle RW, van Trijffel E, Nelissen PM, Koes B, Verhagen AP. Few prom- 6. Thoomes EJ, Scholten-Peeters GG, de Boer AJ, Olsthoorn RA, Verkerk K, Lin C, et al. ising multivariable prognostic models exist for recovery of people with non- Lack of uniform diagnostic criteria for cervical radiculopathy in conservative specific neck pain in musculoskeletal primary care: a systematic review. intervention studies: a systematic review. Eur Spine J. 2012;21:1459–1470. J Physiother. 2018;64:16–23. 7. Iyer S, Kim HJ. Cervical radiculopathy. Curr Rev Musculoskelet Med. 2016;9:272– 29. Schellingerhout JM, Heymans MW, Verhagen AP, Lewis M, de Vet HCW, Koes BW. 280. Prognosis of patients with nonspecific neck pain: development and external validation of a prediction rule for persistence of complaints. Spine. 2010;35: 8. Hogg-Johnson S, van der Velde G, Carroll LJ, Holm LW, Cassidy JD, Guzman J, et al. E827–E835. The burden and determinants of neck pain in the general population: results of the Bone and Joint Decade 2000-2010 Task Force on Neck Pain and Its Associated 30. Sterling M, Hendrikz J, Kenardy J, Kristjansson E, Dumas JP, Niere K, et al. Assess- Disorders. Spine. 2008;33:S39–S51. ment and validation of prognostic models for poor functional recovery 12 months after whiplash injury: A multicentre inception cohort study. Pain. 2012;153:1727– 9. Kim R, Wiest C, Clark K, Cook C, Horn M. Identifying risk factors for first-episode 1734. neck pain: a systematic review. Musculoskelet Sci Pract. 2018;33:77–83. 31. Ritchie C, Hendrikz J, Jull G, Elliott J, Sterling M. External validation of a clinical 10. Jun D, Zoe M, Johnston V, O’Leary S. Physical risk factors for developing non- prediction rule to predict full recovery and ongoing moderate/severe disability specific neck pain in office workers: a systematic review and meta-analysis. Int following acute whiplash injury. J Orthop Sports Phys Ther. 2015;45:242–250. Arch Occup Environ Health. 2017;90:373–410. 32. Bier JD, Ostelo RWJG, Koes BW, Verhagen AP. Validity and reproducibility of the 11. Michaleff ZA, Maher CG, Verhagen AP, Rebbeck T, Lin CW. Accuracy of the Canadian modified STarT Back Tool (Dutch version) for patients with neck pain in primary C-spine rule and NEXUS to screen for clinically important cervical spine injury in care. Musculoskelet Sci Pract. 2017;31:22–29. patients following blunt trauma: a systematic review. CMAJ. 2012;184:E867–E876. 33. Nordin M, Carragee EJ, Hogg-Johnson S, Weiner SS, Hurwitz EL, Peloso PM, et al. 12. Nordin M, Randhawa K, Torres P, Yu H, Haldeman S, Brady O, et al. The Global Spine Assessment of neck pain and its associated disorders: results of the Bone and Joint Care Initiative: a systematic review for the assessment of spine-related complaints Decade 2000-2010 Task Force on Neck Pain and Its Associated Disorders. Spine. in populations with limited resources and in low- and middle-income commu- 2008;33:S101–S122. nities. Eur Spine J. 2018;27:816–827. 34. Holmes JF, Akkinepalli R. Computed tomography versus plain radiography to screen for cervical spine injury: a meta-analysis. J Trauma. 2005;58:902–905. 35. Shim JH, Park CK, Lee JH, Choi JW, Lee DC, Kim DH, et al. A comparison of angled sagittal MRI and conventional MRI in the diagnosis of herniated disc and stenosis in the cervical foramen. Eur Spine J. 2009;18:1109–1116. 36. Lin I, Wiles L, Waller R, Goucke R, Nagree Y, Gibberd M, et al. What does best practice care for musculoskeletal pain look like? Eleven consistent recommenda- tions from high-quality clinical practice guidelines: systematic review. Br J Sports Med. 2020;54:79–86. 37. Nakashima H, Yukawa Y, Suda K, Yamagata M, Ueta T, Kato F. Narrow cervical canal in 1211 asymptomatic healthy subjects: the relationship with spinal cord compression on MRI. Eur Spine J. 2016;25:2149–2154. 38. Hush JM, Lin CC, Michaleff ZA, Verhagen A, Refshauge KM. Prognosis of acute idiopathic neck pain is poor: a systematic review and meta-analysis. Arch Phys Med Rehabil. 2011;92:824–829. 39. Wirth B, Humphreys BK, Peterson C. Importance of psychological factors for the recovery from a first episode of acute non-specific neck pain; a longitudinal observational study. Chiropr Man Therap. 2016;24:9. 40. Corp N, Mansell G, Stynes S, Wynne-Jones G, Morsø L, Hill JC, et al. Evidence-based treatment recommendations for neck and low back pain across Europe: a sys- tematic review of guidelines. Eur J Pain. 2020. https://doi.org/10.1002/ejp.1679. 41. Yu H, Côté P, Southerst D, Wong JJ, Varatharajan S, Shearer HM, et al. Does struc- tured patient education improve the recovery and clinical outcomes of patients
Invited Topical Review 11 with neck pain? A systematic review from the Ontario Protocol for Traffic Injury 57. Khwaja SM, Minnerop M, Singer AJ. Comparison of ibuprofen, cyclobenzaprine or Management (OPTIMa) Collaboration. Spine J. 2016;16:1524–1540. both in patients with acute cervical strain: a randomized controlled trial. CJEM. 42. Gross A, Forget M, St George K, Fraser MM, Graham N, Perry L, et al. Patient ed- 2010;12:39–44. ucation for neck pain. Cochrane Database Syst Rev. 2012;3:CD005106. 43. Gross A, Kay TM, Paquin JP, Blanchette S, Lalonde P, Christie T, et al. Exercises for 58. Cho JH, Nam DH, Kim KT, Lee JH. Acupuncture with non-steroidal anti-inflam- mechanical neck disorders. Cochrane Database Syst Rev. 2015;1:CD004250. matory drugs (NSAIDs) versus acupuncture or NSAIDs alone for the treatment of 44. Gross AR, Paquin JP, Dupont G, Blanchette S, Lalonde P, Cristie T, et al. Exercises chronic neck pain: an assessor-blinded randomised controlled pilot study. Acu- for mechanical neck disorders: a Cochrane review update. Man Ther. 2016;24: punct Med. 2014;32:17–23. 25–45. 45. GRADE Working Group. Grading quality of evidence and strength of recommen- 59. Bronfort G, Evans R, Anderson AV, Svendsen KH, Bracha Y, Grimm RH. Spinal dations. BMJ. 2004;328(7454):1490. manipulation, medication, or home exercise with advice for acute and subacute 46. de Zoete RM, Armfield NR, McAuley JH, Chen K, Sterling M. Comparative effec- neck pain: a randomized trial. Ann Intern Med. 2012;156:1–10. tiveness of physical exercise interventions for chronic non-specific neck pain: a systematic review with network meta-analysis of 40 randomised controlled trials. 60. Derry S, Wiffen PJ, Kalso EA, Bell RF, Aldington D, Phillips T, et al. Topical analgesics Br J Sports Med. 2020. bjsports-2020-102664. for acute and chronic pain in adults - an overview of Cochrane Reviews. Cochrane 47. Martin-Gomez C, Sestelo-Diaz R, Carrillo-Sanjuan V, Navarro-Santana MJ, Bardon- Database Syst Rev. 2017;5:CD008609. Romero J, Plaza-Manzano G. Motor control using cranio-cervical flexion exercises versus other treatments for non-specific chronic neck pain: a systematic review 61. Machado GC, Maher CG, Ferreira PH, Day RO, Pinheiro MB, Ferreira ML. Non-ste- and meta-analysis. Musculoskelet Sci Pract. 2019;42:52–59. roidal anti-inflammatory drugs for spinal pain: a systematic review and meta- 48. Gross A, Langevin P, Burnie SJ, Bédard-Brochu MS, Empey B, Dugas E, et al. analysis. Ann Rheum Dis. 2017;76:1269–1278. Manipulation and mobilisation for neck pain contrasted against an inactive control or another active treatment. Cochrane Database Syst Rev. 2015;9:CD004249. 62. Manchikanti L, Cash KA, Pampati V, Malla Y. Two-year follow-up results of fluo- 49. Wong JJ, Shearer HM, Mior S, Jacobs C, Côté P, Randhawa K, et al. Are manual roscopic cervical epidural injections in chronic axial or discogenic neck pain: a therapies, passive physical modalities, or acupuncture effective for the manage- randomized, double-blind, controlled trial. Int J Med Sci. 2014;11:309–320. ment of patients with whiplash-associated disorders or neck pain and associated disorders? An update of the Bone and Joint Decade Task Force on Neck Pain 63. Woo JH, Park HS. Cervical transforaminal epidural block using low-dose local and Its Associated Disorders by the OPTIMa collaboration. Spine J. 2016;16: anesthetic: a prospective, randomized, double-blind study. Pain Med. 2015;16:61– 1598–1630. 67. 50. Masaracchio M, Kirker K, States R, Hanney WJ, Liu X, Kolber M. Thoracic spine manipulation for the management of mechanical neck pain: a systematic review 64. Cohen SP, Hayek S, Semenov Y, Pasquina PF, White RL, Veizi E, et al. Epidural steroid and meta-analysis. PLoS One. 2019;14:e0211877. injections, conservative treatment, or combination treatment for cervical radicular 51. Miller J, Gross A, D’Sylva J, Burnie SJ, Goldsmith CH, Graham N, et al. Manual pain: a multicenter, randomized, comparative-effectiveness study. Anesthesiology. therapy and exercise for neck pain: a systematic review. Man Ther. 2010;15: 2014;121:1045–1055. 334–354. 52. Fredin K, Lorås H. Manual therapy, exercise therapy or combined treatment in the 65. Bureau NJ, Moser T, Dagher JH, Shedid D, Li M, Brassard P, et al. Transforaminal management of adult neck pain – a systematic review and meta-analysis. Mus- versus intra-articular facet corticosteroid injections for the treatment of cervical culoskelet Sci Pract. 2017;31:62–71. radiculopathy: a randomized, double-blind, controlled study. Am J Neuroradiol. 53. Coulter ID, Crawford C, Vernon H, Hurwitz EL, Khorsan R, Booth MS, et al. 2014;35:1467–1474. Manipulation and mobilization for treating chronic nonspecific neck pain: a sys- tematic review and meta-analysis for an appropriateness panel. Pain Physician. 66. van Middelkoop M, Rubinstein SM, Ostelo R, van Tulder MW, Peul W, Koes BW, 2019;22:E55–E70. et al. Surgery versus conservative care for neck pain: a systematic review. Eur Spine 54. Patel KC, Gross A, Graham N, Goldsmith CH, Ezzo J, Morien A, et al. Massage J. 2013;22:87–95. for mechanical neck disorders. Cochrane Database Syst Rev. 2012;9: CD004871. 67. van Middelkoop M, Rubinstein SM, Ostelo R, van Tulder MW, Peul W, Koes BW, 55. Machado GC, Maher CG, Ferreira PH, Pinheiro MB, Lin CW, Day RO, et al. Efficacy et al. No additional value of fusion techniques on anterior discectomy for neck and safety of paracetamol for spinal pain and osteoarthritis: systematic pain: a systematic review. Pain. 2012;153:2167–2173. review and meta-analysis of randomised placebo controlled trials. BMJ. 2015;350: h1225. 68. Verhagen AP, van Middelkoop M, Rubinstein SM, Ostelo R, Jacobs W, Peul W, et al. 56. Predel HG, Giannetti B, Pabst H, Schaefer A, Hug AM, Burnett I. Efficacy and safety Effect of various kinds of cervical spinal surgery on clinical outcomes: a systematic of diclofenac diethylamine 1.16% gel in acute neck pain: a randomized, double- review and meta-analysis. Pain. 2013;154:2388–2396. blind, placebo-controlled study. BMC Musculoskelet Disord. 2013;14:250. 69. Silva PV, Costa LOP, Maher CG, Kamper SJ, Costa LDCM. The new agenda for neck pain research: a modified Delphi study. J Orthop Sports Phys Ther. 2019;49:666–674. 70. Schellingerhout JM, Verhagen AP, Heymans MW, Pool JJ, Vonk F, Koes BW, et al. Which subgroups of patients with non-specific neck pain are more likely to benefit from spinal manipulation therapy, physiotherapy, or usual care? Pain. 2008;139: 670–680. 71. Price J, Rushton A, Tyros V, Heneghan NR. Consensus on the exercise and dosage variables of an exercise training programme for chronic non-specific neck pain: protocol for an international e-Delphi study. BMJ Open. 2020;10:e037656. 72. Suman A, Armijo-Olivo S, Deshpande S, Marietta-Vasquez J, Dennett L, Miciak M, et al. A systematic review of the effectiveness of mass media campaigns for the management of low back pain. Disabil Rehabil. 2020:1–29.
Journal of Physiotherapy 67 (2021) 69–71 Appraisal 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 Note: Diagnostic test accuracy studies Introduction accuracy study can also investigate combinations of findings, which may be more informative in clinical practice. In clinical physiotherapy practice, making a diagnosis is an essential part of patient management. The diagnostic process typi- This Research Note describes how physiotherapists can interpret cally consists of history taking, physical examination and any addi- classic diagnostic research and explain the difference between classic tional investigations deemed necessary by the physiotherapist. A diagnostic research and diagnostic models. This will help physio- diagnosis enables the physiotherapist to: apply the appropriate evi- therapists to consider which tests to use in clinical practice and why. dence to decisions about possible treatments, address the patient’s desire for information about their current condition and provide any Classic diagnostic test accuracy studies evidence about the prognosis of people with that condition. A diag- nosis therefore facilitates shared decision making by the physio- Scientific research into the value of a diagnostic process typically therapist and the patient about management.1 It is therefore crucial consists of evaluating the accuracy of individual diagnostic tests. for physiotherapists to be able to understand research into diagnostic These studies are called diagnostic test accuracy studies. The validity tests and interpret the results of their diagnostic tests on patients. of the index test is assessed by comparing the outcomes of that test with those of a recognised and valid reference test. Research describing the validity or accuracy of diagnostic methods gained prominence in the 1950s, particularly regarding the validity of In order to quantify the validity of an index test, the index test as psychological tests.2,3 ‘Validity’ in this context refers to whether a test well as the reference test commonly divide patients into two cate- is able to measure what the clinician aims to measure (ie, the gories: test positive and test negative. A 2 3 2 table (also known as a construct of interest). For a number of tests, the validity is obvious: if four-field table or cross classification table) can then be created to the physiotherapist wishes to find out how tall someone is, a depict how the study participants have been scored on the two tests measuring tape is clearly a valid instrument for this. However, (Figure 1). Based on such a table, a range of diagnostic statistics can determining the validity of many diagnostic tests relevant to phys- be calculated, including sensitivity and specificity. iotherapy is more complex (eg, tests to confirm whether a patient who presents with knee pain has a meniscal injury or not). Some index tests or reference tests may be measured as contin- uous data (eg, age). Therefore, to calculate sensitivity and specificity Diagnostic research aims to assess the validity of index tests by the researcher needs to dichotomise the test results, which inevitably comparing them with a reference test. A reference test (previously limits the interpretation of the test performance. The choice of a cut called a ‘gold standard’ test) is considered the best available test for point (eg, age dichotomised at 40 years) is quite often arbitrary. To the condition of interest (target condition) but it may be difficult to avoid this, the accuracy of the index test can be evaluated for multiple administer, expensive or even invasive. Index tests are generally cut points or included as a continuous variable (with regression easier, less expensive and/or safer to administer in clinical practice. analysis). Most physiotherapy diagnostic research focuses on physical exami- nation tests as index tests, such as McMurray’s test for those with Critical appraisal knee pain.4 However, any element of the history (eg, presence of a cough) or a questionnaire (eg, Cumberland Ankle Instability Tool) can Diagnostic test accuracy studies, like other studies, can be of high also be regarded as a diagnostic test and evaluated for its accuracy.5 In or low methodological quality. Tools such as the QUADAS-2 tool addition, while most research focuses on single tests, a diagnostic (Quality Assessment of Diagnostic Accuracy Studies 2)6 can be used to rate the quality of classic diagnostic test accuracy studies. This tool a Reference test b positive negative Sensitivity = true positives / all participants with the condition = 75 / (75 + 50) = 60% Index test 75 175 Specificity = true negatives / all participants without the condition negative positive = 200 / 375 = 53% 50 200 Figure 1. Elements of a classic diagnostic test accuracy study, including (a) a 2 3 2 table and (b) calculation of sensitivity and specificity. https://doi.org/10.1016/j.jphys.2020.12.004 1836-9553/© 2020 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/).
70 Appraisal Research Note Box 1. Design elements that increase the believability of a Diagnostic models classic diagnostic test accuracy study. In daily clinical practice, the diagnostic strategy rarely, if ever, A consecutive or random sample of patients was enrolled relies on a single clinical variable or test.9 In principle, each diagnostic The study avoids inappropriate exclusions variable or test contributes towards the eventual likelihood that a The index test and reference test are assessed independently patient has a particular disease or condition; in other words, the diagnostic process is a multivariable process.8 (ie, without knowledge of the outcome of the other test) The threshold for a positive test is clear Diagnostic research is increasingly being conducted with the aim All participants received both tests of ascertaining which combination of clinical variables or diagnostic The timing between the performance of the index and tests will enable the clinician to make a diagnosis with a reasonable degree of certainty. Within that combination, how much value a reference tests is clearly appropriate (ie, the result on either particular test adds to the (final) diagnosis can also be investigated. test is unlikely to change during that time) For example, what is the additional value of a physical examination test (eg, Lachman’s test) on top of history taking (eg, mechanism of comprises four domains: patient selection, index test, reference injury, rapid effusion) for the diagnosis of an anterior cruciate liga- standard, and flow and timing. For a busy clinician it is helpful to ment (ACL) rupture? To determine this, an analysis firstly focused on consider that a diagnostic accuracy study will probably be trust- the likelihood of the diagnosis based on the combination of clinical worthy if it meets the criteria shown in Box 1. variables from the history, and then added the results of physical examination tests to the analysis, to evaluate whether the likelihood Interpretation of a diagnosis changed substantially.10 The combination of three history items (a person has all three positive), each with an inde- Many people struggle to interpret diagnostic accuracy measures pendent sensitivity and specificity varying from 43 to 73%, decreases such as sensitivity and specificity. For example, what does a high the sensitivity to 18% and increases the specificity to 99%. Adding the degree of sensitivity, such as 92%, mean in practice? Correct inter- anterior drawer test, the only physical examination test with prom- pretation involves both understanding what the terms (eg, speci- ising results in this study, decreased the sensitivity to 16% and did not ficity) and the numbers (eg, 80%) mean, and just as importantly change the specificity (SpPIn).10 understanding how the test is intended to be used in a clinical setting. Some tests are used primarily as a screening test, while others are Diagnostic models are multivariable models, meaning that mul- used to make a diagnosis that directs treatment decisions. Screening tiple variables (index tests) are evaluated for their collective associ- tests such as the C-Spine rule (to identify possibility of cervical spine ation with the outcome of a reference standard (eg, ACL injury versus fracture) are used to rule out serious pathologies and determine if no ACL injury). further testing is needed. Therefore, it is most important that screening tests do not miss people with the diagnosis (ie, screening Critical appraisal tests should have a high sensitivity). However, it is less critical that when a screening test is positive the person actually has the diagnosis For diagnostic models it is also worthwhile to think whether the (ie, it is not crucial that screening tests also have high specificity). In study could have introduced bias concerning: the selection of pa- an ideal world, tests would have high sensitivity and specificity. tients; the conduct or interpretation of the index test and the refer- However, this is rarely the case, so the utility of a test for a specific ence test; the flow of participants; and the order and timing of the purpose may depend on whether that particular usage requires high index and reference tests. The same criteria of trustworthiness sensitivity or high specificity. mentioned for single diagnostic tests hold for diagnostic models, although the analysis methods are much more important. The terms SpPIn and SnNOut were introduced in the 1990s to help with the interpretation of sensitivity and specificity.2,7 SpPIn stands Interpretation for Specificity high and test Positive rules disease In; and SnNOut for Sensitivity high and test Negative rules disease Out. For instance, in Diagnostic models commonly use a regression analysis, and the people with low back pain with a suspected disc herniation the contribution of each individual variable or test is often expressed as a overall sensitivity of the straight leg raising test (SLR) is 92% (95% CI regression coefficient (beta) for a continuous measure, or an odds 87 to 95).8 This means that almost all patients with low back pain and ratio (OR) for a dichotomous measure. For example, if the index test is a disc herniation will have a positive SLR and very few patients with a dichotomous measure (eg, sex) with an OR of 1.2 for males, this herniation will have a false negative result (approximately 8%). It can means that the likelihood of a disorder is slightly higher in males therefore be concluded with a high degree of sensitivity (eg, . 90%) (someone with a positive test result for a given sex), compared with that patients with a negative outcome on the SLR test most likely do females (negative test result). If the index test is continuous (eg, age) not have a herniation (SnNOut). Such an index test is therefore good with an OR of 1.2, this means that the likelihood of the disorder rises at ruling out the disease or condition. Because tests rarely have both slightly with each year increase in the patient’s age. high sensitivity and high specificity, when a test has high sensitivity it often means that the test also produces many false positive results. To evaluate the diagnostic value of a diagnostic model, a receiver The reverse applies to the crossed SLR test that has a high specificity operator characteristic (ROC) curve is presented. This is a graph (SpPIn): this test is good at ruling in the disease or condition,8 but is depicting the sensitivity of the model against the specificity (tech- not good at ruling out the condition (disc herniation). nically, the false positivity rate, ie, 1 – specificity) at different cut-off points. The area under the curve (AUC) indicates how effectively the It is not possible or sensible to provide exact sensitivity and whole model can discriminate between people with and without the specificity values that are considered ‘high’ (or at least high enough to particular condition. A model with an AUC of 1 is a perfect model (ie, comply with the SpPIn and SnNOut rules), since these values depend it can identify all patients with the condition without producing any on the clinical consequences. In a person with low back pain the false positives or false negatives). In contrast, a model with an AUC of consequence of missing a diagnosis of acute cauda equina is very 0.5 has no value whatsoever (ie, it cannot discriminate at all between different to the consequence of missing a disc herniation, so higher people with or without the condition). With an AUC between 0.6 and sensitivity would be required in a test for cauda equina. For muscu- 0.7 the model is considered ‘reasonable’ and with an AUC 0.7 the loskeletal disorders, values 90% with reasonably narrow confidence model is considered ‘good’;11 however, as discussed above, the clin- intervals are often considered sufficiently high for clinical utility.7 ical use and consequences of the diagnostic model need to be considered.
Appraisal Research Note 71 The interpretation of beta coefficients, ORs or an AUC is rather the model. Ideally, models developed in one study are then validated in impractical for clinicians. Therefore, a multivariable diagnostic model a new population, but examples of this are relatively rare. is ideally converted into a score chart or nomogram, which the clinician can easily use to calculate the likelihood of a certain diag- Summary nosis for a particular patient. For example, in patients with shoulder pain, a nomogram was developed that includes one history item This Research Note summarised and contrasted classic diagnostic (male sex) combined with three physical examination tests (positive test accuracy studies and diagnostic models. When interpreting re- lift-off test, Jobe test, and external rotation strength ratio between the sults from a diagnostic study, it is important to clearly understand the affected and unaffected shoulder), which can predict a rotator cuff purpose of the test(s) under investigation.5 The role of a diagnostic tear with 83% accuracy.12 test can be: to screen people to exclude the need to undergo further investigation (in this case a high sensitivity (SnNOut) is important); to Diagnostic models versus classic diagnostic test accuracy studies make a diagnosis to guide prognostication or further treatment de- cisions; or just to monitor the condition.18 Classic diagnostic test accuracy studies (ie, the evaluation of the accuracy of an individual test) often have limited direct applicability to Currently, most diagnostic research focuses on the diagnostic ac- daily practice. In daily practice, the diagnostic process almost always curacy of single tests, but diagnostic accuracy of a model based on consists of a cluster of tests, and the two most important things for a multiple variables or tests better reflects clinical practice and may clinician are: the likelihood of a diagnosis based on all tests; and what produce better results. A growing number of diagnostic models are is the additional contribution of each test (ie, to determine whether it is being developed. Diagnostic models that are found to be adequately worth adding that test to the battery of previous tests).9 A further accurate can be translated into nomograms to enhance their clinical limitation of classic diagnostic research, compared with diagnostic usefulness. Ultimately the value of any clinical diagnosis, based on models, is that it is more often carried out in a select patient group that single tests or a model including multiple tests, depends on whether is not representative of the patients seen in daily practice (selection the diagnosis improves prediction of important patient outcomes or bias) and is therefore likely to overestimate the accuracy (eg, sensitivity helps direct treatment decisions. and specificity) of the index test.13,14 This overestimation resulting from patient selection is most common when the reference test is costly or Competing interests: Nil. invasive (eg, a surgical procedure or magnetic resonance imaging). Sources of support: Nil. Acknowledgements: Nil. There are two situations in which classic diagnostic test accuracy Provenance: Invited. Peer reviewed. studies have a clear role. These are: the use of screening tests among Correspondence: Arianne Verhagen, Discipline of Physiotherapy, healthy subjects (eg, breast cancer screening using a mammogram) Graduate School of Health, University of Technology Sydney, Sydney, and the evaluation of new tests. A screening test is often performed Australia. Email: [email protected] without the usual history taking and physical examination, with the primary aim of excluding people without the disease or condition Arianne Verhagena and Mark Hancockb from undergoing costly and intensive follow-up investigations (ie, aDiscipline of Physiotherapy, Graduate School of Health, University of ruling a condition out (SnNOut)). Technology Sydney, Sydney, Australia With a new test, a stand-alone test should be reasonably able to bFaculty of Medicine and Health Science, Macquarie University, Sydney, discriminate between patients with and without the condition. If a test is unable to do so, further research into the use of the index test Australia and its value as part of a battery of tests in a more realistic clinical situation is unlikely to be appropriate. References Diagnostic models also have some shortcomings. For instance, 1. Davidson M. Aust J Physiother. 2002;48:227–233. inadequate sample size can lead to an overestimation of the likelihood 2. Cronbach LJ, Gleser GC. Psychol Bull. 1953;50:456–473. of a particular diagnosis.15 Furthermore, it is only possible to include in 3. Cronbach LJ, Meehl PE. Psychol Bull. 1955;52:281–302. the model those variables that have been measured. There are often 4. Kaizik MA, et al. Physiother Res Int. 2020;25, e1871. considerable differences between studies as to which variables (index 5. Cohen JF, et al. BMJ Open. 2016;6, e012799. tests) were deemed relevant and included, so comparison between 6. Whiting PF, et al. Ann Intern Med. 2011;155:529–536. studies is difficult. For instance, a systematic review of diagnostic 7. Hegedus EJ, Stern B. J Manip Physiol Ther. 2009;17:E1–E5. models identified four models for the diagnosis of symptomatic sacro- 8. Van der Windt DA, et al. Cochrane Database Syst Rev. 2010;2:CD007431. iliac joints in people with low back pain, only evaluating the combi- 9. Moons KGM, et al. Clin Chem. 2004;50:473–476. nation of physical examination tests.16 Of these four clusters, the 10. Wagemakers HP, et al. Arch Phys Med Rehabil. 2010;91:1452–1459. ‘cluster of Laslett’ is most well-known.17 All these models indicated that 11. Hosmer DW, Lemeshow S. Applied Logistic Regression. 2nd ed. New York: Wiley; three out of five specific tests must be positive, but the tests that were included in the analyses differed somewhat from one another; there- 2000. fore, the clusters differed. The sensitivity of the clusters varied from 45 12. Jain NB, et al. Orthop J Sports Med. 2018;6:2325967118784897. to 91%, and the specificity from 57 to 89%. These models used a more 13. Mower WR. Ann Emerg Med. 1999;33:85–91. classic approach (using sensitivity and specificity) and looked at the 14. Sackett DL, Haynes RB. BMJ. 2002;324:539–541. number of tests instead of weighting each test for its contribution to 15. Retel Helmrich IRA, et al. J Physiother. 2019;4:243–245. 16. Petersen T, et al. Musculoskelet Disord. 2017;18:188. 17. Laslett M. J Man Manip Ther. 2008;16:142–152. 18. Bossuyt PM, et al. BMJ. 2006;332:1089–1092.
Journal of Physiotherapy 67 (2021) 56–61 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 Students achieve comparable performance scores for clinical placements in public and private sectors: a longitudinal observational study Vidya Lawton , Taryn M Jones , Catherine M Dean Department of Health Professions, Faculty of Medicine, Health and Human Sciences, Macquarie University, Sydney, Australia KEY WORDS ABSTRACT Clinical education Questions: Does student clinical performance differ according to healthcare sector? Does student clinical Clinical competence performance at Macquarie University differ from other Australian graduate entry-level programs? Design: A Physiotherapy education longitudinal observational study with comparison to national data. Participants: A total of 284 physio- Healthcare sector therapy students from Macquarie University. Outcome measures: Each student’s clinical performance was Physical therapy evaluated by a clinical educator using the Assessment of Physiotherapy Practice (APP) tool at the end of four 5-week clinical placements. Four measures of clinical performance were analysed: Total APP score, Employability Skills, Clinical Skills and a global rating of performance. A between-group difference in the APP results of 5% was nominated a priori as large enough to be considered important. Results: Of the 1,136 placements, 533 (47%) were undertaken in the private sector. Among their four placements, 99% of students had at least one private sector placement and 70% had two or more private sector placements. There were negligible differences between private and public sector placements in Total APP scores (MD 0%, 95% CI 21 to 1), Employability Skills scores (MD 2% higher in the public sector, 95% CI 1 to 3) and Clinical Skills scores (MD 1% higher in the private sector, 95% CI 21 to 3). On the global rating of performance, 88% of placements in each sector were rated as being either good or excellent. Students in the private sector were 9% (95% CI 3 to 14) more likely to be rated as excellent compared with the public sector. There were negligible differences in clinical performance between the Macquarie University and other Australian graduate-entry students. Conclusion: Macquarie University’s practice of increasing private sector participation in clinical education had no adverse effects on student clinical performance, and it is likely to be beneficial in better preparing students for work in the private sector. [Lawton V, Jones TM, Dean CM (2021) Students achieve comparable performance scores for clinical placements in public and private sectors: a longitudinal observational study. Journal of Physiotherapy 67:56–61] © 2020 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 Australia, compared with the current 44 programs offered by 21 universities across the country.5 This represents a 125% increase in The landscape of the Australian physiotherapy workforce has the number of programs since 2005. Student numbers have doubled changed significantly over the past two decades. The profession has from 4,605 in 2011 to 8,472 in 2018.4,6 Whilst this supports the trebled in size from 13,399 registered physiotherapists in 1998 to healthcare needs of a growing and ageing population, it has inten- 33,792 in 2019.1,2 Changes in the healthcare needs of the Australian sified pressure on an already stretched academic and teaching population have resulted in an increased demand for access to workforce, as well as creating an increased challenge to secure high- physiotherapy from the community and private providers. Australia’s quality clinical education. growing private healthcare sector is now managing individuals who were once solely treated within the public sector: for example, 69% of Despite the private healthcare sector now being the predominant individuals undergoing elective surgical procedures, such as total working environment for Australian physiotherapists, the clinical knee replacements, are now undertaking these procedures within the education programs of most universities have remained largely private sector.3 These changes to healthcare service delivery are also within the public sector. The private sector accounted for only 14% of reflected in the physiotherapy workforce, with 72% of Australian all clinical placements in 2009 for one large university with two physiotherapists now working within the private sector.4 entry-level physiotherapy programs.7 This finding is consistent with later data published by Health Workforce Australia in 2014, which Similarly, physiotherapy education is intimately linked with the reported that only 6% of clinical placements occurred in private evolution and development of the profession and has also grown practice settings.8 Across the allied health professions, clinical significantly. In 1990, there were only five physiotherapy programs in placements in the private sector have the potential to not only ease https://doi.org/10.1016/j.jphys.2020.12.001 1836-9553/© 2020 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 57 the burden of clinical education but also prepare students to work Four variables were collected: Total APP scores (all 20 items), effectively within the private sector. However, there has been some Employability Skills scores (the first six items of the APP), Clinical reluctance from the private sector to participate in clinical education, Skills scores (the final 14 items of the APP), all of which were con- with concerns about patient satisfaction, quality of care, reduced verted into a percentage, and the global rating of each student, as revenue and perceived liability issues.9–12 measured by the global rating scale. Detailed information about the APP and resources to support clinical educators are available on the When its Doctor of Physiotherapy (DPT) commenced in 2012, ClinEdAus website.19 Macquarie University became the 19th university to offer a physio- therapy entry-level degree in Australia. One of the aims of this degree Data analysis was to better align clinical education with current and future phys- iotherapy workforce requirements. One component of the clinical Data collected from five cohorts were combined and quantita- education program with the Macquarie University DPT involves 5- tively analysed using commercial softwarea. Descriptive statistics week clinical placement blocks in diverse areas of physiotherapy were used to describe the demographic characteristics of the par- practice, comparable to those undertaken by other Australian phys- ticipants, and the profile of the clinical education sites, placements iotherapy education providers. Within these 5-week clinical place- and educators. All available de-identified clinical performance data ments Macquarie University has had a deliberate intention to secure for other graduate entry-level programs from the last placement in placements in the private sector. each program from 2018 were extracted from the APPlinkup database with permission.20 This extraction resulted in data for 517 students. Therefore, the aim of this study was to comprehensively describe Total APP scores, Employability Skills scores and Clinical Skills scores the profile of students and clinical performance in 5-week clinical were analysed using independent t-tests to determine the mean placements and to answer two specific research questions: difference between public and private sectors, and between Mac- quarie University and other Australian entry-level programs. These 1. Does student clinical performance differ according to healthcare mean differences were presented with 95% confidence intervals (95% sector? CI). The percentages of students rated as inadequate, adequate, good and excellent were analysed using the Newcombe-Wilson method 2. Does student clinical performance at Macquarie University differ without continuity correction to determine the difference between from other Australian graduate entry-level programs? sectors, and between Macquarie University and other Australian graduate entry-level programs and presented as risk difference (95% Method CI).21,22 A between-group difference in the APP results of 5% was nominated a priori as large enough to be considered important. Design Results A longitudinal, retrospective analysis was conducted on five co- horts of students across four clinical placements in the DPT at Mac- Flow of participants through the study quarie University. The Macquarie University DPT is a 3-year Masters (extended) entry-level physiotherapy degree. Students complete A total of 311 students successfully met entry requirements and three 5-week clinical placements during the first half of the second enrolled into the DPT at Macquarie University. The flow of partici- year, and the third-year students complete another 5-week advanced pants is seen in Figure 1, and demographic data are presented in clinical placement in the final semester. The clinical placements are Table 1. Among the 1,136 placements completed over the 7-year data located within both the private and public sectors in hospital, com- collection period, 99% of students had at least one private sector munity or mixed settings. Students who had commenced the phys- placement and 70% had completed two or more placements in the iotherapy program between 2012 and 2016 were eligible for the private sector. The public sector accounted for the majority of study and were included in the final analysis if they had successfully placements in Rehabilitation (80%) and Acute settings (69%), whereas graduated by June 2019. the private sector accounted for the majority of Primary Care (79%) and Elective (57%) placements. All APP evaluations had the global Outcome measures rating score completed; however, 177 (16%) completed APP evalua- tions had one or more individual items scored as N/A. In the public The primary outcome was the clinical performance of students measured by the Assessment of Physiotherapy Practice (APP), which Students enrolled in is a standardised, reliable and valid assessment tool used by all the course (n = 311) physiotherapy degrees nationally.13–15 The APP is a 20-item instru- ment, covering seven domains of physiotherapy practice. The first Excluded by leaving the two domains (professional behaviour and communication) were course (n = 26) operationally defined as Employability Skills. Although these skills are also known as soft skills or generic skills, they enable graduates to withdrew from the be more likely to have success in their chosen occupations.16,17 The course (n = 20) other five domains (assessment, analysis/planning, intervention, excluded for academic evidence-based practice and risk management) were operationally reasons (n = 6) defined as Clinical Skills, which are the discipline-specific skills that are required by the profession to practise competently. Our opera- Excluded from the tional division of the APP was broadly in line with concurrent work analysis (n = 1) that used factor analysis to validate the division of the APP into Professional (items 1 to 4) and Clinical (items 5 to 20) subscores.18 yet to graduate (n = 1) Each item is assessed on a scale from 0 to 4, with 2 defined as the minimum standard for entry-level physiotherapists. Every item in the Analysed (n = 284) APP must be scored or, if students have not had an opportunity to demonstrate any skills or behaviours listed under a particular item, Figure 1. Flow of Macquarie University participants through the study. marked as not applicable (N/A). In addition, a global rating as a ho- listic representation of a student’s performance was evaluated using a global rating scale, defined by four distinct categories of a student’s overall performance: inadequate, adequate, good and excellent.13
58 Lawton et al: Clinical placements in private and public sectors Table 1 to 14) more likely to be rated as excellent compared with the public Demographics of Macquarie University students and completed clinical placements. sector (Figure 3). Characteristic (n = 284) Comparison of clinical performance with other Australian graduate entry-level programs Age, mean (SD) 24 (3) Sex, n male (%) 161 (57) There were no important differences in Total APP scores (MD 0%, Previous degree, n (%) 95% CI 22 to 1), Employability Skills scores (MD 21%, 95% CI 22 to 1) 111 (39) or Clinical Skills scores (MD 21%, 95% CI 23 to 1) between the clinical Exercise and Sports Science 88 (31) performance of Macquarie University students and students from Health Science 85 (30) other graduate entry-level programs in Australia undertaking their other 3.0 (0.3) final placement, as shown in Table 4. There was minimal difference in Grade point average (0 worst to 4 best), mean (SD) the percentage of students in each category of the global rating scale Student exposure to private sector, n placements (%) 3 (1) (Figure 4). Compared with other graduate entry programs, students at 0 83 (29) Macquarie University had a similar likelihood of being rated excellent 1 146 (51) (RD 21%, 95% CI 26 to 9) or good (RD 4%, 95% CI 23 to 11). The 2 50 (18) numerical data used to generate Figures 2 to 4 are available in 3 Appendix 1 on the eAddenda. 4 2 (1) sector, 95% of APP evaluations were completed without N/As compared with 72% in the private sector. There were no other missing data. Clinical performance over time Discussion Clinical performance for each placement is outlined in Table 2 and There were no important differences in student performance be- Figure 2. There were gradual but small increases in APP scores over tween the public and private sector clinical placements, or more time. broadly when comparing Macquarie University students with stu- dents completing placements within other Australian graduate entry- Clinical performance between sectors level physiotherapy programs. Given that almost all of Macquarie University students had some exposure to the private sector in at There were negligible differences between private and public least one 5-week clinical placement prior to graduation, the findings sector placements in Total APP scores (MD 0%, 95% CI 21 to 1), of this study suggest that students are able to gain the knowledge, Employability Skills scores (MD 2% higher in the public sector, 95% CI skills and attributes required for registration, regardless of the sector 1 to 3) and Clinical Skills scores (MD 1% higher in the private sector, in which they complete their placements. 95% CI 21 to 3), as shown in Table 3. Although the percentage of students rated as good or excellent was the same (88%) for both Student performance was similar between public and private sectors, there was a difference in the distribution of global ratings sector placements within the Macquarie University DPT program. between the sectors. Students in the private sector were 9% (95% CI 3 Furthermore, the performance of Macquarie University students was on a par with other graduate entry-level programs, both published23 and unpublished.20 Importantly, this was achieved with three times Table 2 Mean (SD) clinical performance scores of Macquarie University students by placement. APP scores (%) Placement Total (Items 1 to 20) 1 2 3 4 Employability Skills (Items 1 to 6) Clinical Skills (Items 7 to 20) 74 (12) 76 (12) 76 (12) 81 (12) 83 (13) 83 (13) 84 (12) 87 (12) 70 (12) 73 (13) 73 (12) 78 (13) APP = Assessment of Physiotherapy Practice. Percentage of students 0% 20% 40% 60% 80% 100% Placement 1 (n = 284) Placement 2 (n = 284) Placement 3 (n = 284) Placement 4 (n = 284) Inadequate Adequate Good Excellent Figure 2. Percentage of Macquarie University students rated within each category of the global rating scale across Placements 1 to 4.
Research 59 Table 3 Difference between Mean (SD) clinical performance scores of Macquarie University students by sector, and mean difference (95% CI) between sectors. sectors (95% CI) APP scores (%) All Sector Private minus public (n = 1,136) 0 (21 to 1) Private Public 22 (23 to 21) (n = 533) (n = 603) 1 (21 to 3) Total (Items 1 to 20) 77 (12) 77 (13) 77 (12) 100% Employability Skills (Items 1 to 6) 84 (13) 83 (13) 85 (12) Clinical Skills (Items 7 to 20) 74 (13) 74 (14) 73 (12) APP = Assessment of Physiotherapy Practice. Percentage of students 0% 20% 40% 60% 80% Private (n = 533) Public (n = 603) Inadequate Adequate Good Excellent Figure 3. Percentage of Macquarie University students rated within each category of the global rating scale in private and public sector placements. Table 4 Mean (SD) clinical performance scores of Macquarie University students on Placement 4 and of students on their final placement in other Australian graduate entry-level programs, and mean difference (95% CI) between programs. APP scores (%) Programs Difference between programs (95% CI) Macquarie University Other programs Macquarie University minus (n = 284) (n = 517) other programs Total (Items 1 to 20) 81 (12) 81 (11) 0 (22 to 1) Employability Skills (Items 1 to 6) 87 (12) 88 (11) 21 (22 to 1) Clinical Skills (Items 7 to 20) 78 (13) 79 (12) 21 (23 to 1) APP = Assessment of Physiotherapy Practice. Percentage of students 0% 20% 40% 60% 80% 100% Macquarie (n = 284) Other (n = 517) Inadequate Adequate Good Excellent Figure 4. Percentage of students rated within each category of the global rating scale for Macquarie University placement 4 and other Australian graduate-entry physiotherapy programs final placement. more private sector placements than previously reported,7 with 533 sector.24 Although three (1%) students did not complete a 5-week (47%) of Macquarie University placements in the private sector and placement in the private sector, all students were exposed to this 364 (32%) in the private practice setting, which is five times more sector in integrated sports, and work health and safety placements than the national data for this segment of the private sector.8 Private undertaken in a different component of the program. The complete sector placements varied according to type, ranging from 20% for exposure to the private sector seen within the Macquarie University rehabilitation placements to 79% for primary care. This range broadly DPT is a large increase compared with 56% of students, which has reflects the distribution of where those physiotherapy services are been reported previously.7 Private sector placements are important delivered. For example, only 14% of the rehabilitation services for for all students, given the current physiotherapy workforce where stroke participating in the national audit were from the private almost 73% of physiotherapists work within the private sector.25
60 Lawton et al: Clinical placements in private and public sectors Private sector placements have the added benefit of building unique approved this study. All participants gave written informed consent skills necessary for a sector where students are more likely to work upon graduation,26 as well as allowing for comprehensive exposure before data collection began. to contemporary healthcare environments with a greater breadth of healthcare service delivery models. With an increase in physio- Competing interests: Nil. therapy programs nationally and an increase in pressure to provide high-quality clinical experiences to students within these programs, Source(s) of support: Macquarie University Doctor of Physio- private sector placements contribute to the sustainability of clinical education within physiotherapy programs. therapy Clinical Education Component was supported by Health Communication skills, eagerness to learn, empathy and having a Workforce Australia’s Clinical Training Fund in 2012 until 2015. good knowledge base have been identified by both clinical educators Funding: This research did not receive any specific grant from and university educators as being desirable attributes prior to clinical placement.27,28 The implication is that these employability skills need funding agencies in the public, commercial, or not-for-profit sectors. to be developed transparently and deliberately within academic curricula.29,30 Provenance: Not invited. Peer reviewed. Australian graduate entry-level students, regardless of university, Correspondence: Vidya Lawton, Department of Health Professions, consistently performed better in their employability skills than clin- ical skills, suggesting that these skills are being successfully devel- Faculty of Medicine, Health and Human Sciences, Macquarie Univer- oped. The Macquarie University program is a graduate entry-level program and the older, more mature student cohort may have sity, Sydney, Australia. Email: [email protected] developed their employability skills in their previous degree and/or had greater employment exposure. Furthermore, all academic units References within the first year of the DPT, prior to the commencement of the 5- week block clinical placements, include integrated clinical experi- 1. Physiotherapy Labour Force 1998. National Health Labour Force Series Number 22. ences within the community with a focus on the development of Canberra: Australian institute of Health and Welfare; 1998. communication skills, self-reflection and characteristics of profes- sionalism, which may have contributed to the higher scores for 2. 2018/2019 Annual Report. Physiotherapy Board of Australia web site. https://www. employability skills. physiotherapyboard.gov.au/. Accessed 18 October, 2019. One potential limitation of this research is the difference in usage 3. Knee replacement hospitalisations 18 years and over. Australian Commission on of the APP between sectors. There were more items scored as N/A on Safety and Quality in Healthcare web site; 2017. https://www.safetyandquality.gov. the APP evaluations completed in the private sector (28%) compared au/sites/default/files/migrated/4.1-Knee-replacement.pdf. Accessed 19 October, with the public sector (5%), particularly regarding Item 18, in which 2019. discharge planning is assessed. Previous research has found that Item 18 had high N/A counts.13,31 It may also partly reflect that only six 4. 2017/2018 Annual Report. Australian Health Practitioner Regulations Agency web private sector sites out of 68 were included in the development of the site. www.ahpra.gov.au/annualreport/2018. Accessed 10 October, 2019. APP.32 The concept of discharge planning is likely to be considered differently between sectors and settings. Within the private sector, 5. Approved Programs of Study. Australian Health Practitioner Regulations Agency intervention and management often tend to focus on support and the web site; 2019. https://www.ahpra.gov.au/Education/Approved-Programs-of- development of self-management strategies rather than on discharge Study.aspx?ref=Physiotherapist. Accessed 27 September, 2019. preparation. 6. 2010/2011 Annual Report. Australian Health Practitioner Regulations Agency web In conclusion, Macquarie University’s practice of increasing pri- site. www.ahpra.gov.au/annualreport/2011. Accessed 10 October, 2019. vate sector participation in clinical education had no adverse effects on student clinical performance and provides the opportunity to 7. Dean C, Stark A, Gates C, Czerniec SA, Hobbs C, Bullock LL, et al. A profile of prepare students for work in the private sector, where the majority physiotherapy clinical education. Aust Health Rev. 2009;33:38–46. will seek future employment. Furthermore, private sector placements also alleviate the clinical education burden in the public sector, 8. Health Workforce Australia. Clinical Training Profile: Physiotherapy; 2014. thereby contributing to the sustainability of clinical education com- https://acdhs.edu.au/wp-content/uploads/2018/05/Clinical-Training-Profiles-Project- ponents of physiotherapy programs. Physiotherapy.pdf. Accessed 19 November, 2020. What was already known on this topic: In Australia, 9. Doubt L. Clinical education in private practice: an interdisciplinary project. J Allied although 72% of physiotherapists work in the private sector, Health. 2004;33:47–50. many university physiotherapy courses offer clinical placements for students that are primarily in the public sector. 10. Sokkar C, McAllister L, Raymond J, Penman M. Supervisors’ perceptions of student What this study adds: The clinical performance of physio- placements in speech-language pathology private practice: benefits, challenges therapy students was similar on clinical placements regardless of and strategies. Speech Lang Hear. 2019;22:122–133. whether the placement occurred in the private or public sector. The clinical performance of physiotherapy students at a univer- 11. Bridle M, Hawkes B. A survey of Canadian occupational therapy private practice. sity that provided a substantial proportion of clinical placements Can J Occup Ther. 1990;57:160–166. in the private sector was similar to the clinical performance of physiotherapy students nationally. 12. Sloggett K. Private practice: benefits, barriers and strategies of providing fieldwork placements. Can J Occup Ther. 2003;70:42–50. Footnotes: a IBM-SPSS Statistics version 24, SPSS Inc., Chicago, USA. 13. Dalton M. Development of the assessment of physiotherapy practice - a standardised and validated approach to assessment of professional competence in physiotherapy eAddenda: Appendix 1 can be found online at DOI: https://doi. [PhD Thesis]. Queensland, Australia: Griffith University; 2011. org/10.1016/j.jphys.2020.12.001. 14. Dalton M, Davidson M, Keating J. The Assessment of Physiotherapy Practice (APP) is Ethics approval: The Macquarie University Human Research a valid measure of professional competence of physiotherapy students: a cross- Ethics Committee (Medical Sciences; Reference No: 5201600425) sectional study with Rasch analysis. J Physiother. 2011;57:239–246. 15. Dalton M, Davidson M, Keating J. The Assessment of Physiotherapy Practice (APP) is a reliable measure of professional competence of physiotherapy students: a reli- ability study. J Physiother. 2012;58:49–56. 16. Barrie S. Understanding what we mean by the generic attributes of graduates. High Educ. 2006;51:215–241. 17. Osmani M, Weerakkody V, Hindi N, Eldabi T. Graduates employability skills: A review of literature against market demand. J Educ Bus. 2019;94: 423–432. 18. Reubenson A, Ng L, Gucciardi DF. The Assessment of Physiotherapy Practice tool provides informative assessments of clinical and professional dimensions of stu- dent performance in undergraduate placements: a longitudinal validity and reli- ability study. J Physiother. 2020;66:113–119. 19. Clinical Education Australia. Enabling Clinical Education Skills. Clinical Education Australia website. https://clinedaus.org.au/. Accessed 29 January, 2020. 20. APPLinkup. APPLinkUp website. https://www.applinkup.com. Accessed 20 September, 2019. 21. Herbert R. Research Note: Significance testing and hypothesis testing: meaning- less, misleading and mostly unnecessary. J Physiother. 2019;65:178–181. 22. Physiotherapy Evidence Database Confidence Interval Calculator. Physiotherapy Evidence Database web site. https://www.pedro.org.au/english/downloads/ confidence-interval-calculator/. Accessed 20 September, 2019. 23. Brooks SM, Milne N, Orr RM, Terry R. Is a students’ ability to critically self-reflect, related to their performance on physiotherapy clinical placements? Aust J Clin Educ. 2016;1:6. 24. National Stroke Audit: Acute Services Report 2019. National Stroke Foundation web site. https://informme.org.au/stroke-data/Acute-audits. Accessed 1 February, 2020. 25. Physiotherapists 2017 Fact sheet. Health Workforce Australia web site. https://hwd. health.gov.au/. Accessed 30 September, 2019. 26. Bacopanos E, Edgar S. Employment patterns of Notre Dame graduate physiother- apists 2006-12: targeting areas of workforce need. Aust Health Rev. 2015;40:188– 193. 27. Cross V. Begging to Differ? Clinicians’ and academics’ views on desirable attributes for physiotherapy students on clinical placement. Assess Eval High Educ. 1998;23:295–311.
Research 61 28. Cross V. Approaching consensus in clinical competence assessment: third round of 32. Dalton M, Davidson M. Development of the Assessment of Physiotherapy Practice a delphi study of academics’ and clinicians’ perceptions of physiotherapy un- (APP): A standardised and valid approach to assessment of clinical competence in dergraduates. Physiotherapy. 2001;87:341–350. physiotherapy. Australian Learning and Teaching Council (ALTC) Final Reports web site; 2009. www.altc.edu.au. Accessed 10 October, 2019. 29. Chipchase LS, Galley P, Jull G, McMeeken JM, Refshauge K, Naylor M, et al. Char- acteristics of student preparedness for clinical learning: clinical educator per- Websites spectives using the Delphi approach. BMC Med Educ. 2012;12:112. APPLinkUp www.applinkup.com 30. Trede F. Role of work-integrated learning in developing professionalism and pro- ClinEdAus www.clinedaus.org.au fessional identity. Asia-Pac J Coop Educ. 2012;13:159–167. 31. Judd BK, Scanlan JN, Alsion JA, Waters D, Gordon CJ. The validity of a professional competence tool for physiotherapy students in simulation-based clinical educa- tion: a Rasch analysis. BMC Med Educ. 2016;16:196.
Journal of Physiotherapy 67 (2021) 1–2 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 The role of the physiotherapist in treating survivors of sexual assault Janine Stirling a, K Jane Chalmers b,c, Lucy Chipchase c,d a Department of Counselling and Psychotherapy, Australian College of Applied Psychology, Sydney, Australia; b IIMPACT in Health Research, Allied Health and Human Performance, University of South Australia, Adelaide, Australia; c School of Health Sciences, Western Sydney University, Sydney, Australia; d College of Nursing and Health Sciences, Flinders University, Adelaide, Australia In memory of India Eve Chipchase history of sexual assault.7 Similarly, women who have been sexually assaulted have been found to perceive symptoms of incontinence or The Me Too Movement has facilitated an international conversation constipation as more severe and life impacting than those without a on sexual harassment and sexual assault, paving the way for change.1 sexual assault history.8 While suitably qualified physiotherapists in This topic – once shrouded in secrecy, silence and shame – is the Women’s, Men’s and Pelvic Health subdiscipline play a funda- currently under the spotlight, challenging society in a new and dy- mental role in the management of these pelvic floor conditions, there namic way. With open dialogue and readily accessible information, are two aspects that require further consideration. the need for healthcare professionals to know more and do more is compelling. Thus, it is time to reflect on the role that physiotherapists First, given that one in four women have experienced sexual as- may play, either explicitly or implicitly, in the management of people sault, we must consider how this knowledge can be used in practice who have undergone sexual assault. The intent of this editorial is to to achieve the best possible outcome for women seeking treatment. raise awareness and demonstrate a need for specific skills and This may require physiotherapists, in the process of taking a history, training to meet the complex needs of women who are survivors of to use screening tools that address aspects of sexual wellbeing. Sexual sexual assault trauma. The primary focus is on women’s experiences; trauma impacts a person’s psychological and physical health, and men, however, may be similarly impacted. treatment effectiveness relies on treating the underlying issues as well as focusing on presenting symptoms.9 Thus, one could argue that In Australia, a population survey conducted in 2016 found that one it is an ethical duty to screen for sexual trauma and to obtain as much in every two women experienced sexual harassment.2 Sexual clinically relevant information as possible to support assessment, harassment includes a range of unwanted behaviours such as treatment and referral.9 There are different ways to obtain this in- touching, kissing, fondling and showing or sending sexually offensive formation, either through subjective questioning, or through the use material via text, email or social media.2,3 Sexual assault, on the other of more structured assessment tools. For example, the Adverse hand, is an act of a sexual nature that involves threat, intimidation Childhood Experience Questionnaire for Adults10 is a 10-question tick and physical force, carried out against a person’s will and includes box tool that can identify exposure to sexual, emotional and physical rape, attempted rape and/or indecent assault.3 One in six women adversity in childhood (Box 1). Alternatively, the American College of have experienced a sexual assault in Australia, although if childhood Obstetricians and Gynecologists suggests five screening questions sexual abuse is included, this ratio becomes one in four women.2,3 that can be asked to screen women for sexual assault (Box 2).11 The The terms sexual assault and sexual abuse (often associated with American College of Obstetricians and Gynecologists suggests that behaviour toward children, not adults) are often used interchange- healthcare practitioners screen those presenting with pelvic pain, ably in the literature. For the purposes of this paper, the term sexual sexual dysfunction or dysmenorrhoea.11 The numerous health im- assault is used and encompasses rape, sexual abuse and assault. plications associated with sexual assault means that it is highly probable that all physiotherapists will encounter patients with a Sexual assault impacts on a person’s physical, social, emotional history of sexual assault in their practice, whether explicitly identified and psychological health. A systematic review of international papers, or not. Training in conducting a subjective assessment and use of the including over three million participants, reported that sexual assault aforementioned assessment tools is important to ensure that phys- had a significant association with a lifetime diagnosis of depression, iotherapists are competent in dealing with the possible responses to post-traumatic stress disorder, anxiety, eating disorders, sleep disor- the screening questions, while remaining sensitive to the needs of the ders and suicide attempts.4 Similarly, Paras et al5 found a statistically patient throughout the process.12 significant association between women with a history of rape and a subsequent diagnosis of fibromyalgia, chronic pelvic pain and A second aspect for physiotherapists to consider when treating a gastrointestinal disorders. Indeed, any exposure to trauma – be it patient presenting with a history of sexual assault is the survival psychological, emotional, physical or sexual in nature – results in an response and sequelae. Approximately 70% of women report tonic individual being 2.7 times more likely to experience a somatic syn- immobility during a sexual assault.13 Tonic immobility is a profound, drome than if they had no exposure to trauma.6 These last two global motor inhibition where the skeletal muscles tense rigidly and studies were systematic reviews and meta-analyses of studies con- are unable to be moved voluntarily.13,14 Not only is voluntary move- ducted internationally.5,6 ment affected, but vocal capacity is also diminished.14,15 These motor inhibitions mean that key survival-based actions of screaming and Physiotherapists who work in the Women’s, Men’s and Pelvic movements of fighting or running away, which ordinarily protect the Health subdiscipline of the profession have a role to play with women body during a sympathetic response to threat, are not available.15 who are survivors of sexual assault because they are more likely to Women who experience tonic immobility during sexual assault are have multiple pelvic floor complaints compared with those without a https://doi.org/10.1016/j.jphys.2020.11.008 1836-9553/© 2020 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/).
2 Editorial Box 1. Adverse Childhood Experience Questionnaire for willingness to access physiotherapy services? If a patient does access Adults.10 physiotherapy and vestiges of the survival response arise, what conditions are necessary to promote a positive treatment outcome? Did you feel that you didn’t have enough to eat, had to wear dirty clothes, or had no one to protect or take care of you? A trauma-informed approach is recommended when treating a patient who presents with a history of any type of trauma.17 The five Did you lose a parent through divorce, abandonment, death, or core principles of trauma-informed care include: providing emotional other reason? and physical safety by attuning to a patient’s needs; being trustworthy; offering choices to patients; collaborating with them; and empowering Did you live with anyone who was depressed, mentally ill, or them.10,17 In a qualitative study involving 27 survivors of childhood attempted suicide? sexual abuse who received or were referred to physiotherapy, a need for safety was considered the prevailing theme when discussing how Did you live with anyone who had a problem with drinking or health professionals can practise with more sensitivity to their needs.18 using drugs, including prescription drugs? Safety was impacted if patients perceived a lack of control. They valued accepting environments where the physiotherapist respected them, Did your parents or adults in your home ever hit, punch, beat or was informed about how trauma impacts the body and was attentive to threaten to harm each other? their personal boundaries. Survivors also expressed a need to work in partnership with a healthcare team consisting of a physiotherapist, Did you live with anyone who went to jail or prison? psychotherapist and general practitioner.18 A similar approach was Did a parent or adult in your home ever swear at you, insult documented by Dunleavy and Slowik in the physiotherapy manage- ment of a patient with low back pain and a history of sexual assault 40 you, or put you down? years prior.16 With collaborative identification of stress responses and Did a parent or adult in your home ever hit, beat, kick or triggers, management of hyperarousal, and a slow, graded exposure to triggering stimuli, the patient reported positive outcomes in her back physically hurt you in any way? pain after 2 years and an improvement in post-traumatic stress disor- Did you feel that no one in your family loved you or thought you der symptoms over 4 years.16 were special? With very little available research to address how physiotherapists Did you experience unwanted sexual contact (such as fondling can work with women who have been sexually assaulted, we must consider how to build capacity within the profession to ensure that or oral/anal/vaginal intercourse/penetration)? women who are survivors of sexual assault receive the care and support they require. This editorial is a first step that has aimed to highlight gaps Box 2. American College of Obstetricians and Gynecologists in research and in clinical practice. There is a strong need for: good sexual assault screening questions.11 screening of sexual assault history and subsequent psychological illness in women; a requirement for future research to provide more detailed Has anyone ever touched you against your will or without your approaches to delivering trauma-informed pelvic healthcare to women consent? who are survivors of sexual assault; and competency-based training for physiotherapists so that they can deal with the issues that may arise Have you ever been forced or pressured to engage in sexual from screening. This is a call to action to further our profession’s capacity activities when you did not want to? to help the one in four women who have experienced sexual assault. Have you ever had unwanted sex while under the influence of Ethics approval: Nil. alcohol or drugs? Competing interests: Nil. Source(s) of support: Nil. Do you feel that you have control over your sexual relationships Acknowledgements: Nil. and will be listened to if you say ‘no’ to sexual activities? Provenance: Not invited. Peer reviewed. Correspondence: Lucy Chipchase, Flinders University, Adelaide, Is your visit today because of a sexual experience you did not Australia. Email: lucy.chipchase@flinders.edu.au want to happen? References Floyd S, Anderson J. American College of Obstetricians and Gynecologists Committee Opinion Number 777: Sexual Assault. 1. Lore G, et al. Policy Options Politiques. 2020. https://policyoptions.irpp.org/ Obstet Gynecol. 2019;133:e296– e301. magazines/february-2020/support-services-still-lacking-for-survivors-of-sexual- assault/. Accessed 9 September, 2020. twice as likely to develop post-traumatic stress disorder and are three times more likely to have severe depression 6 months after sexual 2. Australian Bureau of Statistics. Personal Safety. Canberra: ABS; 2017. https:// assault.13 Women describe the experience of tonic immobility as www.abs.gov.au/ausstats/[email protected]/Lookup/by%20Subject/4906.0w2016wMain% leaving them vulnerable to the feeling that immobility may occur in 20FeatureswAbout%20the%20Personal%20Safety%20Survey%20w2. Accessed 9 other stressful life situations that involve sexual contact, or when September, 2020. they feel afraid, out of control, angry or disregarded.15 It is plausible that when physiotherapists treat women with a history of sexual 3. Australian Bureau of Statistics. Personal Safety. Canberra: ABS; 2012. https:// assault that involuntary reflexive responses relating to the assault www.abs.gov.au/ausstats/[email protected]/Lookup/4906.0Chapter5002012. Accessed 9 may be elicited during the course of treatment. This is particularly September, 2020. likely if a woman is seeking help for a distressing pelvic condition that requires an internal examination and related treatment. 4. Chen LP, et al. Mayo Clin Proc. 2010;85:618–629. 5. Paras ML, et al. JAMA. 2009;302:550–561. To date, one case study has been published addressing how a 6. Afari N, et al. Psychosom Med. 2014;76:2–11. patient who received physiotherapy management of lower back pain 7. Beck JJ, et al. J Sexual Med. 2009;6:193–198. later developed symptoms associated with a prior history of sexual 8. Imhoff LR, et al. Arch Surg. 2012;147:1123–1129. assault.16 In this case study, a number of physiological signs and 9. Probst DR, et al. J Aggress Maltreatment Trauma. 2011;20:199–226. symptoms were observed in the patient after 10 weeks of treatment, 10. Provider Toolkit. Screening and Responding to the Impact of ACE’s and Toxic Stress. including sweating, shaking, suppressed breathing, freezing, hyper- vigilance, an inability to focus, and outbursts of anger.16 The patient May 2020. https://www.acesaware.org/wp-content/uploads/2020/05/ACEs-Aware- then disclosed experiencing flashbacks of her sexual assault 40 years Provider-Toolkit-5.21.20.pdf. Accessed 9 September, 2020. prior and was subsequently diagnosed with post-traumatic stress 11. Floyd S, Anderson J. Obstet Gynecol. 2019;133:e296–e301. disorder. This report of delayed-onset post-traumatic stress disorder 12. Frawley HC, et al. Physiother Theory Pract. 2019;35:1117–1130. during physiotherapy sessions draws into focus a number of key 13. Moller A, et al. Acta Obstet Gynecol Scand. 2017;96:932–938. questions that require investigation. For example, how many women 14. de Kleine RA, et al. Psychiatry Res. 2018;270:1105–1109. experience similar somatic symptoms during visits to physiothera- 15. TeBockhorst SF, et al. Psychol Trauma. 2015;7:171–178. pists? Are physiotherapists equipped to detect and manage these 16. Dunleavy K, Slowik AK. Phys Ther. 2012;92:339–351. uncomfortable somatic experiences? To what degree does a fear of 17. Kezelman C, Stavropoulos P. Blue Knot Foundation. 2018;1–147. uncomfortable somatic experiences interfere with a woman’s 18. Schachter CL, et al. Can Fam Physician. 2004;50:405–412.
Search
