Research 15 Study SMD (95% CI) Study SMD (95% CI) Random Random Christofoletti 2008 Arcoverde 2014 Eggermont 2009a Bossers 2015 Eggermont 2009b Eggermont 2009a Hoffmann 2016 Eggermont 2009b Lam 2015 Hong 2018 Lam 2011 Lam 2011 Maki 2012 Lü 2015 Nagamatsu 2013 Yoon 2013 Scherder 2005 Sungkarat 2017 Total Zhu 2018 –2 –1 0 1 2 print & web 4C=FPO Total Favours con Favours exp print & web 4C=FPO –2 –1 0 1 2 Figure 8. Standardised mean difference (95% CI) of the effect of exercise on working memory by pooling data from eight trials (n = 646). Favours con Favours exp Figure 4. Standardised mean difference (95% CI) of the effect of exercise on delayed in MMSE are related to language and working memory (10 from memory by pooling data from 11 trials (n = 1294). Orientation, five from Attention and Calculation and four from Lan- guage).97 Improvement in MMSE may be induced predominantly by additional 12 weeks or more (ie, 12 and 48 weeks).46,60,86 It is worth enhanced language ability and working memory. The conclusion that noting that all trials that reported sustained improvement in global exercise improves global cognition has to be interpreted carefully as cognition included people with MCI as participants and had longer exercise may not benefit other cognitive functions, as shown in the total training durations ( 21 hours). On the contrary, in those trials present review. Also, most of the included studies recruited consid- that reported no sustained improvement, the participants had mild to erably more women than men. Finally, since not all of the included moderate dementia and the total training duration was shorter ( 18 trials explicitly disclosed the type of dementia (eg, Alzheimer’s, hours). However, the current results should be interpreted with vascular) of the participants, the current review was unable to esti- caution, as there is insufficient data to estimate the effect of different mate the effect of exercise on various types of dementia. trial characteristics (eg, participant population, training protocol) on the sustainability of improvement. Also, no study has examined the This systematic review had some limitations. The mean baseline sustainability of the benefits of exercise on behavioural problems. MMSE scores were used to classify the selected studies into different patient sub-groups. One potential problem for this classification is Most trials had acceptable attendance rates. The only trial69 that that it cannot take the heterogeneity of the sample within a study had poor attendance (33%) included participants with a markedly low into account. The comparison of exercise effectiveness on people with baseline MMSE score (8.8), while the participants of all other trials different levels of cognitive impairment across studies is therefore had baseline MMSE scores . 15. Severe cognitive impairment could compromised. Also, the influence of specific parameters of exercise be a potential cause of a low attendance rate. protocols on training effect remains largely uninvestigated. There were large inter-study variations in training protocol. This study tried Regarding safety, the adverse events that were considered to be to reduce the heterogeneity by selecting trials that shared a particular potentially intervention-related were observed in few cases (n = 54) parameter for sensitivity analysis (eg, total training duration). How- compared with the total number of participants exposed to the ex- ever, other dimensions of an exercise program might have also ercise intervention (n = 1850). Exercise intervention appears to be contributed to the difference in effect size obtained across analyses. In largely safe in people with MCI or dementia. However, it is recom- addition, the large inter-study variations in training protocol made it mended that exercise be administered with supervision. difficult to identify an ‘optimal’ training protocol in terms of training The included studies had some common limitations. Most of the trials examined global cognition using MMSE. Nineteen of 30 scores SMD (95% CI) Study SMD (95% CI) Random Random Arcoverde 2014 Study Bossers 2015 Hoffmann 2016 Bossers 2015 Hong 2018 Christofoletti 2008 Liu-Ambrose 2016 Eggermont 2009a Nagamatsu 2012 Eggermont 2009b van Uffelen (men) 2008 Hong 2018 van Uffelen (women) 2008 Scherder 2005 Total Total print & web 4C=FPO –4 –3 –2 –1 0 1 2 3 4 –2 –1 0 1 print & web 4C=FPO Favours con Favours exp Favours con Favours exp Figure 6. Standardised mean difference (95% CI) of the effect of exercise on recognition Figure 10. Standardised mean difference (95% CI) of the effect of exercise on cognitive by pooling data from six trials (n = 291). flexibility by pooling data from seven trials (n = 544).
16 Law et al: Exercise for cognition and behaviour in dementia Study SMD (95% CI) Study SMD (95% CI) Random Random Arcoverde 2014 Cancela 2016 Barnes 2013 Fleiner 2017 Bossers 2015 Hoffmann 2016 Cancela 2016 Holthoff 2015 Christofoletti 2008 Lam 2011 Eggermont 2009a Lowery 2014 Eggermont 2009b Rolland 2007 Hoffmann 2016 Telenius 2015a Hong 2018 Yang 2015 Lam 2011 Total Lü 2015 Maki 2012 –2 –1 0 1 2 print & web 4C=FPO Scherder 2005 Yoon 2013 Favours con Favours exp Zhu 2018 Figure 16. Standardised mean difference (95% CI) of the effect of exercise on behav- Total ioural problem by pooling data from 9 trials (n = 1172). print & web 4C=FPO –2 –1 0 1 2 Favours con Favours exp In the future, more studies should be conducted to evaluate the effect of exercise on specific cognitive functions. The heterogeneity in Figure 12. Standardised mean difference (95% CI) of the effect of exercise on attention most areas of cognitive functions remains large, while determinants by pooling data from 15 trials (n = 1264). of effective training remain unclear. More effort should be put into future studies to scrutinise the pivotal factors (eg, type of training, frequency and duration. Therefore, analyses could only be conducted target population) that contribute to the favourable treatment out- with total training duration, which can also shed some light on the comes. Research on people with a more severe grade of dementia is proposed time of an ‘optimal’ training protocol. Finally, publication needed, especially considering that this review showed that people bias was present in most of the meta-analyses; this was taken into with moderate dementia tend to demonstrate greater improvement. consideration in the rating of the quality of the evidence according to Whether the effect of exercise on cognitive function and behavioural the GRADE system. changes could last after discontinuing the exercise training also re- quires further investigation. Study SMD (95% CI) Random In conclusion, exercise appears to be safe in people with MCI and dementia when under supervision. There was low quality of evidence Barnes 2013 to support using non-specific physical exercise to reduce the decline Bossers 2015 in global cognition in people with MCI or dementia. The quality of Christofoletti 2008 evidence improved to moderate when aerobic exercise, particularly at Eggermont 2009a moderate to high intensity, was adopted. The effect appeared to be Eggermont 2009b more pronounced in those with moderate-grade dementia relative to Hoffmann 2016 those at an earlier stage. Total training duration of . 24 hours could Holthoff 2015 lead to a greater training effect. For individual cognitive functions, Hong 2018 there is low quality of evidence showing that exercise attenuates the Lam 2015 decline in working memory. Moderate evidence was found that ex- Lam 2011 ercise in general has no substantial effect on attention. Aerobic ex- Maki 2012 ercise appears to have no substantial effect on cognitive flexibility in Scherder 2005 people with MCI or mild dementia. The effect of exercise on other Toots 2017 cognitive functions was largely uncertain. It remains controversial van Uffelen (men) 2008 whether the effect of exercise on cognitive function could last after van Uffelen (women) 2008 the termination of exercise. There was also moderate quality evidence to support the use of exercise in alleviating behavioural problems. Total What was already known on this topic: People with mild print & web 4C=FPO –2 –1 0 1 2 3 cognitive impairment and dementia exhibit deterioration in cognitive function and may have behavioural problems. Exercise Favours con Favours exp improves cognition, but the evidence for this benefit focuses on global measures of cognitive function, which combine specific Figure 14. Standardised mean difference (95% CI) of the effect of exercise on language aspects such as attention, executive function, memory, language by pooling data from 14 trials (n = 1593). and reasoning. What this study adds: Exercise reduces global cognitive decline and lessens behavioural problems in people with mild cognitive impairment or dementia. Its benefits on cognitive function are mainly due to its effects on working memory. The most pronounced effect on global cognition appears to arise from aerobic exercise at moderate or greater intensity with at least 24 hours of total training time.
Research 17 Footnotes: a Comprehensive Meta-analysis Version 2, Biostat, 26. Wilson RS, Leurgans SE, Boyle PA, Bennett DA. Cognitive decline in prodromal Englewood, NJ, 2005. b Review Manager Version 5.3, Copenhagen: Alzheimer disease and mild cognitive impairment. Arch Neurol. 2011;68:351–356. The Nordic Cochrane Centre, The Cochrane Collaboration, 2014. 27. Rovio S, Spulber G, Nieminen LJ, Niskanen E, Winblad B, Tuomilehto J, et al. The effect of midlife physical activity on structural brain changes in the elderly. Neu- eAddenda: Figures 3, 5, 7, 9, 11, 13, 15, 17 and Appendices 1, 2, 3 can robiol Aging. 2010;31:1927–1936. be found online at https://doi.org/10.1016/j.jphys.2019.11.014. 28. Tan ZS, Spartano NL, Beiser AS, DeCarli C, Auerbach SH, Vasan RS, et al. Physical activity, brain volume, and dementia risk: The Framingham Study. J Gerontol A Biol Ethics approval: Nil. Sci Med Sci. 2017;72:789–795. Competing interests: Nil. 29. Northey JM, Cherbuin N, Pumpa KL, Smee DJ, Rattray B. Exercise interventions for cognitive function in adults older than 50: a systematic review with meta-analysis. Source(s) of support: This research did not receive any specific Br J Sports Med. 2018;52:154–160. grant from funding agencies in the public, commercial, or not-for- 30. Song D, Yu DSF, Li PWC, Lei Y. The effectiveness of physical exercise on cognitive profit sectors. and psychological outcomes in individuals with mild cognitive impairment: A Acknowledgements: We would like to thank the researchers of systematic review and meta-analysis. Int J Nurs Stud. 2018;79:155–164. the original studies that provided us their data, and Dr. Sam Chan for 31. Young J, Angevaren M, Rusted J, Tabet N. Aerobic exercise to improve cognitive function in older people without known cognitive impairment. Cochrane Database his advice on categorisation of cognitive functions. Syst Rev. 2015;4:CD005381. Provenance: Not invited. Peer reviewed. 32. Burgener SC, Yang Y, Gilbert R, Marsh-Yant S. The effects of a multimodal inter- vention on outcomes of persons with early-stage dementia. Am J Alzheimers Dis Correspondence: Marco YC Pang, Department of Rehabilitation Other Demen. 2008;23:382–394. Sciences, The Hong Kong Polytechnic University, Hong Kong. Email: 33. Yaguez L, Shaw KN, Morris R, Matthews D. The effects on cognitive functions of a movement-based intervention in patients with Alzheimer’s type dementia: a pilot [email protected] study. Int J Geriatr Psychiatry. 2011;26:173–181. References 34. Ahlskog JE, Geda YE, Graff-Radford NR, Petersen RC. Physical exercise as a pre- ventive or disease-modifying treatment of dementia and brain aging. Mayo Clin 1. Alzheimer’s Disease International. Policy Brief for Heads of Government: The Global Proc. 2011;86:876–884. Impact of Dementia. London, UK: Alzheimer’s Disease International; 2013. 35. Lautenschlager NT, Cox K, Cyarto EV. The influence of exercise on brain aging and 2. Carson S, McDonagh MS, Peterson K. A systematic review of the efficacy and safety dementia. Biochim Biophys Acta. 2012;1822:474–481. of atypical antipsychotics in patients with psychological and behavioral symptoms of dementia. J Am Geriatr Soc. 2006;54:354–361. 36. Goldet G, Howick J. Understanding GRADE: an introduction. J Evid Based Med. 2013;6:50–54. 3. Lyketsos CG, Lopez O, Jones B, Fitzpatrick AL, Breitner J, DeKosky S. Prevalence of neuropsychiatric symptoms in dementia and mild cognitive impairment: results 37. Aguiar P, Monteiro L, Feres A, Gomes I, Melo A. Rivastigmine transdermal patch and from the cardiovascular health study. JAMA. 2002;288:1475–1483. physical exercises for Alzheimer’s disease: a randomized clinical trial. Curr Alz- heimer Res. 2014;11:532–537. 4. Seitz D, Purandare N, Conn D. Prevalence of psychiatric disorders among older adults in long-term care homes: A systematic review. Int Psychogeriatr. 38. Arcoverde C, Deslandes A, Moraes H, Pompeu FA, Silveira H, Mouta R, et al. 2010;22:1025–1039. Treadmill training as an augmentation treatment for Alzheimer’s disease: a pilot randomized controlled study. Arq Neuropsiquiatr. 2014;72:190–196. 5. Hersch EC, Falzgraf S. Management of the behavioral and psychological symptoms of dementia. Clin Interv Aging. 2007;2:611–621. 39. Baker LD, Frank LL, Foster-Schubert K, Green PS, Wilkinson CW, McTiernan A, et al. Effects of aerobic exercise on mild cognitive impairment: a controlled trial. Arch 6. Alzheimer’s Association. Alzheimer’s disease facts and figures. Alzheimers Dement. Neurol. 2010;67:71–79. 2014;10:2. 40. Barnes DE, Santos-Modesitt W, Poelke G, Kramer AF, Castro C, Middleton LE, et al. 7. Cuijpers P. Depressive disorders in caregivers of dementia patients: a systematic The mental activity and exercise (max) trial: A randomized controlled trial to review. Aging Ment Health. 2005;9:325–330. enhance cognitive function in older adults. JAMA Intern Med. 2013;173:797–804. 8. Donaldson C, Tarrier N, Burns A. Determinants of carer stress in Alzheimer’s dis- 41. Bossers WJR, van der Woude LHV, Boersma F, Hortobágyi T, Scherder EJA, van ease. Int J Geriatr Psychiatry. 1998;13:248–256. Heuvelen MJG. A 9-week aerobic and strength training program improves cogni- tive and motor function in patients with dementia: a randomized, controlled trial. 9. Fauth EB, Gibbons A. Which behavioral and psychological symptoms of dementia Am J Geriatr Psychiatry. 2015;23:1106–1116. are the most problematic? Variability by prevalence, intensity, distress ratings, and associations with caregiver depressive symptoms. Int J Geriatr Psychiatry. 42. Cancela JM, Ayán C, Varela S, Seijo M. Effects of a long-term aerobic exercise 2013;29:263–271. intervention on institutionalized patients with dementia. J Sci Med Sport. 2016;19:293–298. 10. Ferri CP, Prince M, Brayne C, Brodaty H, Fratiglioni L, Ganguli M, et al. Global prevalence of dementia: a Delphi consensus study. Lancet. 2005;366:2112–2117. 43. Christofoletti G, Oliani MM, Gobbi S, Stella F, Bucken Gobbi LT, Renato Canineu P. A controlled clinical trial on the effects of motor intervention on balance and 11. Lyketsos CG, Carrillo MC, Ryan JM, Khachaturian AS, Trzepacz P, Amatniek J, et al. cognition in institutionalized elderly patients with dementia. Clin Rehabil. Neuropsychiatric symptoms in Alzheimer’s disease. Alzheimers Dement. 2008;22:618–626. 2011;7:532–539. 44. Eggermont LHP, Knol DL, Hol EM, Swaab DF, Scherder EJA. Hand motor activity, 12. Ory M, Yee JL, Tennstedt SL, Schulz R. The extent and impact of dementia care: cognition, mood, and the rest-activity rhythm in dementia: a clustered RCT. Behav unique challenges experienced by family caregivers. In: Schulz R, ed. Handbook on Brain Res. 2009;196:271–278. dementia caregiving: evidence-based interventions for family caregivers. New York: Springer; 2000:1–32. 45. Eggermont LHP, Swaab DF, Hol EM, Scherder EJA. Walking the line: a randomised trial on the effects of a short term walking programme on cognition in dementia. 13. Russ TC, Hamer M, Stamatakis E, Starr JM, Batty G. Psychological distress as a risk J Neurol Neurosurg Psychiatry. 2009;80:802–804. factor for dementia death. Arch Intern Med. 2011;171:1859. 46. Fiatarone Singh MA, Gates N, Saigal N, Wilson GC, Meiklejohn J, Brodaty H, et al. 14. Wimo A, Jönsson L, Bond J, Prince M, Winblad B. The worldwide economic impact The Study of Mental and Resistance Training (SMART) Study—resistance training of dementia 2010. Alzheimers Dement. 2013;9:1–11. and/or cognitive training in mild cognitive impairment: a randomized, double- blind, double-sham controlled trial. J Am Med Dir Assoc. 2014;15:873–880. 15. Farina N, Rusted J, Tabet N. The effect of exercise interventions on cognitive outcome in Alzheimer’s disease: a systematic review. Int Psychogeriatr. 2014;26:9–18. 47. Fleiner T, Dauth H, Gersie M, Zijlstra W, Haussermann P. Structured physical ex- ercise improves neuropsychiatric symptoms in acute dementia care: a hospital- 16. Forbes D, Forbes SC, Blake CM, Thiessen EJ, Forbes S. Exercise programs for people based RCT. Alzheimers Res Ther. 2017;9:68. with dementia. Cochrane Database Syst Rev. 2015;4:CD006489. 48. Hoffmann K, Sobol NA, Frederiksen KS, Vogel A, Vestergaard K, Brændgaard H, 17. Forbes D, Thiessen EJ, Blake CM, Forbes SC, Forbes S. Exercise programs for people et al. Moderate-to-high intensity physical exercise in patients with Alzheimer’s with dementia. Cochrane Database Syst Rev. 2013;12:CD006489. disease: a randomized controlled trial. J Alzheimers Dis. 2016;50:443–453. 18. Groot C, Hooghiemstra AM, Raijmakers PGHM, Van Berckel BN, Scheltens P, 49. Hokkanen L, Rantala L, Remes AM, Härkönen B, Viramo P, Winblad I. Dance and Scherder EJ, et al. The effect of physical activity on cognitive function in patients movement therapeutic methods in management of dementia: a randomized, with dementia: A meta-analysis of randomized control trials. Ageing Res Rev. controlled study. J Am Geriatr Soc. 2008;56:771–772. 2016;25:13–23. 50. Holthoff VA, Marschner K, Scharf M, Steding J, Meyer S, Koch R, et al. Effects of 19. Öhman H, Savikko N, Strandberg TE, Pitkälä KH. Effect of physical exercise on physical activity training in patients with Alzheimer’s dementia: results of a pilot cognitive performance in older adults with mild cognitive impairment or de- RCT study. PLoS One. 2015;10:e0121478. mentia: a systematic review. Dement Geriatr Cogn Disord. 2014;38:347–365. 51. Hong SG, Kim JH, Jun TW. Effects of 12-week resistance exercise on electroen- 20. Zheng G, Xia R, Zhou W, Tao J, Chen L. Aerobic exercise ameliorates cognitive cephalogram patterns and cognitive function in the elderly with mild cognitive function in older adults with mild cognitive impairment: a systematic review and impairment: a randomized controlled trial. Clin J Sport Med. 2018;28:500–508. meta-analysis of randomised controlled trials. Br J Sports Med. 2016. 52. Hu JP, Guo YH, Wang F, Zhao XP, Zhang QH, Song QH. Exercise improves cognitive 21. Du Z, Li Y, Li J, Zhou C, Li F, Yang X. Physical activity can improve cognition in function in aging patients. Int J Clin Exp Med. 2014;7:3144–3149. patients with Alzheimer’s disease: a systematic review and meta-analysis of ran- domized controlled trials. Clin Interv Aging. 2018;13:1593–1603. 53. Kemoun G, Thibaud M, Roumagne N, Carette P, Albinet C, Toussaint L, et al. Effects of a physical training programme on cognitive function and walking efficiency in 22. Dunham J. Cognition. In: Darity WA, ed. International Encyclopedia of the Social elderly persons with dementia. Dement Geriatr Cogn Disord. 2010;29:109–114. Sciences. Detroit: Macmillan Reference USA; 2008. 54. Kim MJ, Han CW, Min KY, Cho CY, Lee CW, Ogawa Y, et al. Physical exercise with 23. Horning S. Aging and Cognition. In: Ramachandran V, ed. Encyclopedia of Human multicomponent cognitive intervention for older adults with Alzheimer’s Disease: Behavior. 12th ed. London: Academic Press; 2012. a 6-month randomized controlled trial. Dement Geriatr Cogn Dis Extra. 2016;6:222– 232. 24. Aggarwal NT, Wilson RS, Beck TL, Bienias JL, Bennett DA. Mild cognitive impair- ment in different functional domains and incident Alzheimer’s disease. J Neurol 55. Kwak YS, Um SY, Son TG, Kim DJ. Effect of regular exercise on senile dementia Neurosurg Psychiatry. 2005;76:1479–1484. patients. Int J Sports Med. 2008;29:471–474. 25. Cloutier S, Chertkow H, Kergoat MJ, Gauthier S, Belleville S. Patterns of cognitive 56. Lam LCW, Chan WC, Leung T, Fung AW, Leung EM. Would older adults with mild decline prior to dementia in persons with mild cognitive Impairment. J Alzheimers cognitive impairment adhere to and benefit from a structured lifestyle activity Dis. 2015;47:901–913.
18 Law et al: Exercise for cognition and behaviour in dementia intervention to enhance cognition?: A cluster randomized controlled trial. PLoS home residents with dementia: a single blinded randomized controlled trial. BMC One. 2015;10:e0118173. Geriatr. 2015;15:158. 57. Lam LCW, Chau RCM, Wong BML, Fung AW, Lui VW, Tam CC, et al. Interim follow- 77. Toots A, Littbrand H, Bostrom G, Hörnsten C, Holmberg H, Lundin-Olsson L, et al. up of a randomized controlled trial comparing Chinese-style mind body (Tai Chi) Effects of exercise on cognitive function in older people with dementia: a ran- and stretching exercises on cognitive function in subjects at risk of progressive domized controlled trial. J Alzheimers Dis. 2017;60:323–332. cognitive decline. Int J Geriatr Psychiatry. 2011;26:733–740. 78. Toulotte C, Fabre C, Dangremont B, Lensel G, Thévenon A. Effects of physical 58. Lam LCW, Chau RCM, Wong BML, Fung AW, Tam CW, Leung GT, et al. A 1-year training on the physical capacity of frail, demented patients with a history of randomized controlled trial comparing mind body exercise (Tai Chi) with falling: a randomised controlled trial. Age Ageing. 2003;32:67–73. stretching and toning exercise on cognitive function in older chinese adults at risk 79. van Uffelen JGZ, Chinapaw MJM, van Mechelen W, Hopman-Rock M. Walking or of cognitive decline. J Am Med Dir Assoc. 2012;13:568.e15–568.e20. vitamin B for cognition in older adults with mild cognitive impairment? A rand- 59. Landi F, Russo A, Bernabei R. Physical activity and behavior in the elderly: a pilot omised controlled trial. Br J Sports Med. 2008;42:344–351. study. Arch Gerontol Geriatr Suppl. 2004;9:235–241. 80. Varela S, Ayán C, Cancela JM, Martín V. Effects of two different intensities of aerobic 60. Lautenschlager NT, Cox KL, Flicker L, Foster JK, van Bockxmeer FM, Xiao J, et al. exercise on elderly people with mild cognitive impairment: a randomized pilot Effect of physical activity on cognitive function in older adults at risk for Alzheimer study. Clin Rehabil. 2012;26:442–450. disease: A randomized trial. JAMA. 2008;300:1027–1037. 81. Venturelli M, Scarsini R, Schena F. Six-month walking program changes cognitive 61. Lee HJ, Don Kim K. Effect of physical activity on cognition and daily living activities and ADL performance in patients with Alzheimer. Am J Alzheimers Dis Other Demen. of the elderly with mild dementia. J Phys Ther Sci. 2018;30:428–433. 2011;26:381–388. 62. Liu-Ambrose T, Best JR, Davis JC, Eng JJ, Lee PE, Jacova C, et al. Aerobic exercise and 82. Vreugdenhil A, Cannell J, Davies A, Razay G. A community-based exercise pro- vascular cognitive impairment: A randomized controlled trial. Neurology. gramme to improve functional ability in people with Alzheimer’s disease: a ran- 2016;87:2082–2090. domized controlled trial. Scand J Caring Sci. 2012;26:12–19. 63. Lowery D, Cerga-Pashoja A, Iliffe S, Thuné-Boyle I, Griffin M, Lee J, et al. The effect of 83. Wei XH, Ji LL. Effect of handball training on cognitive ability in elderly with mild exercise on behavioural and psychological symptoms of dementia: the EVIDEM-E cognitive impairment. Neurosci Lett. 2014;566:98–101. randomised controlled clinical trial. Int J Geriatr Psychiatry. 2014;29:819–827. 84. Yang SY, Shan CL, Qing H, Wang W, Zhu Y, Yin MM, et al. The effects of aerobic 64. Lü J, Sun M, Liang L, Feng Y, Pan X, Liu Y. Effects of momentum-based dumbbell exercise on cognitive function of Alzheimer’s disease patients. CNS Neurol Disord training on cognitive function in older adults with mild cognitive impairment: a Drug Targets. 2015;14:1292–1297. pilot randomized controlled trial. Clin Interv Aging. 2015;11:9–16. 85. Yoon JE, Lee SM, Lim HS, Kim TH, Jeon JK, Mun MH. The effects of cognitive activity 65. Maki Y, Ura C, Yamaguchi T, Murai T, Isahai M, Kaiho A, et al. Effects of intervention combined with active extremity exercise on balance, walking activity, memory using a community-based walking program for prevention of mental decline: a level and quality of life of an older adult sample with dementia. J Phys Ther Sci. randomized controlled trial. J Am Geriatr Soc. 2012;60:505–510. 2013;25:1601–1604. 66. Miu DKY, Szeto SL, Mak YF. A randomised controlled trial on the effect of exercise 86. Zhu Y, Wu H, Qi M, Zhang Q, Zhou L, Wang S, et al. Effects of a specially designed on physical, cognitive and affective function in dementia subjects. Asian J Gerontol aerobic dance routine on mild cognitive impairment. Clin Interv Aging. Geriatr. 2008;3:8–16. 2018;13:1691–1700. 67. Nagamatsu LS, Chan A, Davis JC, Beattie BL, Graf P, Voss MW, et al. Physical activity 87. Gomez-Pinilla F, Hillman C. The influence of exercise on cognitive abilities. Compr improves verbal and spatial memory in older adults with probable mild cognitive Physiol. 2013;3:403–428. impairment: a 6-month randomized controlled trial. J Aging Res. 88. Kirk-Sanchez N, McGough E. Physical exercise and cognitive performance in the 2013;2013:861893. elderly: current perspectives. Clin Interv Aging. 2013:51. 68. Nagamatsu LS, Handy TC, Hsu CL, Voss M, Liu-Ambrose T. Resistance training 89. Colcombe SJ, Erickson KI, Scalf PE, Kim JS, Prakash R, McAuley E, et al. Aerobic promotes cognitive and functional brain plasticity in seniors with probable mild exercise training increases brain volume in aging humans. J Gerontol A Biol Sci Med cognitive impairment: A 6-month randomized controlled trial. Arch Intern Med. Sci. 2006;61:1166–1170. 2012;172:666–668. 90. Erickson KI, Miller DL, Weinstein AM, Akl SL, Banducci SE. Physical activity and 69. Rolland Y, Pillard F, Klapouszczak A, Reynish E, Thomas D, Andrieu S, et al. Exercise brain plasticity in late adulthood: a conceptual review. Ageing Res. 2012;3:34–47. program for nursing home residents with Alzheimer’s disease: a 1-year random- 91. Erickson KI, Voss MW, Prakash RS, Basak C, Szabo A, Chaddock L, et al. Exercise ized, controlled trial. J Am Geriatr Soc. 2007;55:158–165. training increases size of hippocampus and improves memory. Proc Natl Acad Sci 70. Scherder EJA, Van Paasschen J, Deijen JB, Deijen JB, Van Der Knokke S, Orlebeke JF, U S A. 2011;108:3017–3022. et al. Physical activity and executive functions in the elderly with mild cognitive 92. Paterson DH, Warburton DER. Physical activity and functional limitations in older impairment. Aging Ment Health. 2005;9:272–280. adults: a systematic review related to Canada’s Physical Activity Guidelines. Int J 71. Sobol NA, Dall CH, Høgh P, Hoffmann K, Steen Frederiksen K, et al. Change in fitness Behav Nutr Phys Act. 2010;7:38. and the relation to change in cognition and neuropsychiatric symptoms after 93. Gligoroska J, Manchevska S. The effect of physical activity on cognition physio- aerobic exercise in patients with mild Alzheimer’s Disease. J Alzheimers Dis. logical mechanisms. Mater Sociomed. 2012;24:198. 2018;65:137–145. 94. Cummings J, Mega M, Gray K, Rosenberg-Thompson S, Carusi DA, Gornbein J. The 72. Steinberg M, Leoutsakos J-MS, Podewils LJ, Lyketsos CG. Evaluation of a home- Neuropsychiatric Inventory: Comprehensive assessment of psychopathology in based exercise program in the treatment of Alzheimer’s disease: The Maximizing dementia. Neurology. 1994;44:2308–2314. Independence in Dementia (MIND) study. Int J Geriatr Psychiatry. 2009;24:680– 95. Nowrangi MA, Lyketsos CG, Rosenberg PB. Principles and management of neuro- 685. psychiatric symptoms in Alzheimer’s dementia. Alzheimers Res Ther. 2015;7:12. 73. Stevens J, Killeen M. A randomised controlled trial testing the impact of exercise on 96. Kennedy RE, Cutter GR, Wang G, Schneider LS. Using baseline cognitive severity for cognitive symptoms and disability of residents with dementia. Contemp Nurse. enriching Alzheimer’s disease clinical trials: How does Mini-Mental State Exami- 2006;21:32–40. nation predict rate of change? Alzheimers Dement. 2015;1:46–52. 74. Sungkarat S, Boripuntakul S, Chattipakorn N, Watcharasaksilp K, Lord SR. Effects of 97. Folstein MF, Folstein SE, McHugh PR. “Mini-mental state”. A practical method for Tai Chi on Cognition and Fall Risk in Older Adults with Mild Cognitive Impairment: grading the cognitive state of patients for the clinician. J Psychiatr Res. A Randomized Controlled Trial. J Am Geriatr Soc. 2017;65:721–727. 1975;12:189–198. 75. Telenius EW, Engedal K, Bergland A. Effect of a high-intensity exercise program on physical function and mental health in nursing home residents with dementia: an Websites assessor blinded randomized controlled trial. PLoS One. 2015;10:e0126102. 76. Telenius EW, Engedal K, Bergland A. Long-term effects of a 12 weeks high-intensity www.pedro.org.au functional exercise program on physical function and mental health in nursing
Journal of Physiotherapy 66 (2020) 61–63 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: Blinding: what, why, when and how? What is blinding? Unique to RCTs, blinding is also used in an attempt to prevent controls obtaining effective non-study interventions (co-in- Blinding involves manipulating or concealing information from terventions) or from obtaining the intervention outside the trial people involved in clinical research. The amount and nature of the (contamination).5 For instance, in situations where trial participants concealed information varies depending upon the study design and know that they have been allocated to the placebo group, they may the role that the person plays within the research. This ‘information’ seek effective treatments. Co-intervention and contamination will might be the hypothesis being evaluated, treatment or exposure generally result in an underestimation of the measure of effect of the under investigation, participants’ group allocation, outcome being intervention, biasing the observed measure of effect towards the null. assessed, tests that are evaluated, or names of the investigators conducting the research. The evidence of a placebo effect (change in outcome due to the perception that something is being done) in healthcare is well Blinding is a critical component in many study designs. The established. Any placebo effect is usually perceived as being an effect concept of blinding is mostly associated with randomised controlled that improves participant outcomes rather than one that is a harmful trials (RCTs) but is also important in many other study types, effect. The use of blinding in placebo-controlled RCTs attempts to including cohort studies, diagnostic test accuracy studies, case- minimise the impact the placebo effect has on the trial results. The control studies and systematic reviews. It may be possible to with- measured effect of an intervention consists of both the effect of the hold information from any of several groups involved in clinical intervention itself and the placebo effect due to receiving healthcare. research: participants, healthcare providers, data collectors, assessors By including a placebo arm of the trial (as opposed to a control group of outcomes (or exposures), data analysts, data safety and monitoring that is left untreated) and ensuring that all participants are blinded, teams, and manuscript writers.1 any placebo effect is presumed to be the same in both groups. Therefore, the difference in outcomes between the control and Why is blinding important? intervention group can be assumed to be due to the intervention and not to any placebo effect. The ultimate aim of blinding is to avoid bias. Once those involved in a research project are unblinded (eg, once they are aware of which In case-control studies participants are blinded to the outcome, to group the participant belongs to, which outcome is important, etc) reduce differential error in the reporting or measurement of the bias may be introduced, which threatens the validity of the study exposure by cases compared with controls. This can be challenging as results. participants are often aware of their outcome status. As a result, participants in case-control studies may be deliberately blinded to the Empirical evidence suggests that the absence of blinding can research hypothesis or to the specific exposure of interest in an impact study results.2–4 RCTs that do not employ blinding demon- attempt to minimise this differential error in the measurement of strate, on average, larger effects of intervention.2,3 For example, odds exposure. ratios have been demonstrated to be on average 17% greater in un- blinded RCTs than in blinded ones.3 In any study design, it may be possible, and ideal, to blind par- ticipants to the research hypothesis. This may be particularly useful in The impact of blinding varies slightly, depending upon the study RCTs and cohort studies when participants are unable to be blinded to design, the information that is being concealed and who is being the treatment/exposure under investigation, or in case-control blinded. studies when participants are aware of their outcome status. Blind- ing may prevent differential errors in the reporting of outcomes or Blinding participants/patients exposures due to the knowledge of treatment or outcome. In RCTs and cohort studies, blinding of participants to the treat- Blinding of healthcare professionals ment/exposure group is undertaken in an attempt to reduce differ- ential error in the reporting or measurement of the outcome due to During an RCT, blinding of healthcare professionals and those the behaviour of the participants. This differential error is known as caring for participants is important because knowledge of the ascertainment bias. Ascertainment bias is likely to occur when exposure can play an important role in the decision to adjust the blinding is not implemented and the patients assess the main out- intervention or control intervention, and could result in a decrease comes themselves (ie, for unblinded Patient Reported Outcome of contrast between interventions. The decision to withdraw a Measurement). Ascertainment bias can bias the observed measure of participant from a trial may be influenced by the knowledge of effect either towards or away from the null. The size and the direction which treatment/exposure group the participant belongs to.6 For of the bias will depend upon the participants’ knowledge or beliefs example, if the healthcare provider knows that their patient is in about which treatment/exposure group they belong to, as well as the placebo group, they may withdraw the patient from the trial if their beliefs about the impact of the treatment/exposure upon the their symptoms become worse so that an active treatment can be outcome. prescribed. https://doi.org/10.1016/j.jphys.2019.11.017 1836-9553/© 2019 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/).
62 Appraisal Research Note Blinding research team members compared with surgery, neither the trial participant nor the health- care providers can be blinded. Some ethical committees agree to keep In RCTs and cohort studies, as with participants, outcome asses- participants naïve, meaning that they are unaware of the research sors and data analysts can be blinded to the treatment/exposure hypothesis and therefore the comparator treatment. This can only be group of the participants in order to minimise ascertainment bias. A performed when both interventions (eg, surgery and physiotherapy) blinded outcome assessment is particularly important when the are frequently provided treatments and both are effective. The po- outcome being assessed is subjective. tential for bias caused by unblinded healthcare providers can be minimised by having different healthcare providers give the experi- In some case-control studies, depending upon the design of the mental and control interventions. study and nature of the outcome, it may be possible to blind re- searchers who are responsible for gathering data regarding exposure Blinding data collectors, assessors, analysists, data monitoring teams and to the case/control status of the participants. However, in many case- manuscript writers in RCTs control studies, this may be difficult or impossible to achieve. In this case, it may be possible to blind these researchers to the exposure of It may be possible to withhold information about allocation from interest and hypothesis under investigation. data collectors and assessors by ensuring that, at the time data are collected and assessed, the data collectors and assessors have had no In diagnostic accuracy studies, investigators who are conducting prior contact with participants and have not been informed about the index test should be blinded to the results of the reference test allocation. and investigators conducting the reference test should be blinded to the results of the index test. If blinding is not implemented, agree- When outcomes are not self-reported, it is usually possible to ment between the index and reference tests may be inflated, thus blind the outcome assessor to the participants’ treatment group. This artificially inflating the accuracy (sensitivity and specificity) of the reduces the potential for ascertainment bias. In order to maintain test under investigation. However, there is a limited amount of con- blinding, it is important that the outcome assessor has not been flicting evidence that blinding causes bias in diagnostic accuracy involved in delivering the intervention. It is necessary to hide any studies.7 clues about allocation. For instance, in an RCT comparing exercise with an injection in patients with shoulder pain, all participants – In systematic reviews, blinding of reviewers to the authors and those who did and did not receive the injection – should wear a band journals of the original studies is often attempted. However, given the aid over the injection site while their outcomes are being assessed. resources that are required and the uncertain benefit of minimising bias, this may not be required.8,9 However, to minimise bias in this It may also be possible to withhold information about allocation process it is recommended that multiple researchers independently from data analysts, data safety and monitoring teams, and manuscript (blindly) review the full text of publications for inclusion, extract writers by providing them with coded data. However, perfect blind- data, and assesses the risk of bias.9,10 ing about allocation requires strict adherence to trial protocols; in some circumstances, protocol violations or obvious effects of in- When and how do we blind? terventions, or adverse effects of interventions may mean that blinding is imperfect. It may be difficult to blind these individuals Despite its importance, blinding is often difficult and sometimes because patterns in the data may be strongly (rightly or wrongly) impossible to implement. The manner in which it is implemented suggestive of allocation. varies depending upon the study design. While there are differing points of view as to the ideal way to implement or measure the Cohort studies success of blinding,11 it is well accepted that this feature is crucial to the rigour of clinical research. Depending upon the design of a cohort or observational study, participants may or may not be able to be blinded to their exposure Randomised controlled trials status. Some cohort studies allow participants a choice as to their exposure group and therefore blinding of participants is impossible. Allocation concealment or blinding? As a result of this limitation, greater emphasis should be placed on Blinding and allocation concealment are terms that are often rigorous, unbiased assessment of outcomes.6 However, in other observational designs, where cohorts of individuals are followed up confused. The term blinding in an RCT refers to keeping those over time, and multiple exposures and/or multiple outcomes are involved with the trial unaware of the assigned intervention. Allo- measured (such as The Framingham study) participants may be un- cation concealment refers to processes that prevent those who are aware of their exposure levels and, therefore, participant blinding recruiting participants from knowing the upcoming treatment as- may be possible. signments. The decision to accept or reject a participant is therefore made without the knowledge of which treatment group this partic- When outcome data are collected, as with RCTs, outcome asses- ipant will be assigned to. Allocation concealment is a methodological sors usually can and should be blind to the participants’ exposure technique seen mainly in RCTs, and can always be implemented.12 group, except when the patient assesses their own outcome, like pain, disability and quality of life. This is particularly important if the In contrast to allocation concealment, blinding occurs after allo- outcomes are subjective, which lends greater scope for individual cation and cannot always be implemented. Allocation concealment biases to enter.15 aims to reduce selection bias,12 whereas blinding aims to reduce measurement bias. Despite the best efforts in blinding outcome assessors, circum- stances beyond the control of the researchers or assessors may result Blinding participants, investigators and healthcare providers in RCTs in the assessors becoming unblinded. Given that participants may not A range of methodological devices can be used to withhold in- be blinded in cohort studies, they may divulge some information specifically about their exposure status to the outcome assessors. formation about allocation. For example, participants or healthcare providers can be blinded by providing a sham intervention. Credible Case-control studies sham interventions may assist in the conduct of the trial because few participants want to receive a placebo treatment and few healthcare The design of case-control studies makes implementing blinding providers want to administer one. Placebos are commonly used in challenging. As participants are selected based on the presence or explanatory trials.13 absence of the outcome, it can be impossible to blind participants to their outcome. Participants may be fully aware as to whether they are Techniques used to blind participants in pragmatic trials will vary a ‘case’ or a ‘control’ and are therefore unblinded. Therefore, the recall and are dependent upon the intervention being evaluated. In trials of of exposures is fraught with differential measurement error. Cases complex interventions such as surgery or physiotherapy, blinding of may be likely to recall exposures differently to controls.16,17 participants, investigators and/or doctors is difficult, and may be impossible.4,14 For example, when community-based healthcare is
Appraisal Research Note 63 Although ideal, blinding of assessors is often difficult to maintain however, been some recognition of the problems in methods used to throughout the entire process of exposure ascertainment, and needs a assess blinding. Consequently, a recommendation to report the thorough logistic organisation to make sure that different people in adequacy of blinding that had previously been made in the CONSORT the study do different things. For example, if the outcome assessors statement has been retracted.28,29 are also involved in interviewing the participants, the outcome status of the participants may be visually apparent, or become apparent Summary during the interview process.16 Nonetheless, blinding of outcome assessors in case-control studies is possible but is dependent upon Blinding is an important methodological technique for minimising the outcome and exposures being evaluated.18 bias in many types of clinical research. Although blinding may be difficult or impossible to implement, researchers designing such It may be possible to blind participants and investigators to the studies should make every reasonable attempt to mask the partici- research hypothesis or objective of the study, so that they are un- pants’ exposure or treatment from key groups involved in the study. aware of the specific exposures the research is investigating. This may Readers of clinical research should critically evaluate the attempts at be achieved and maintained throughout the data collection process blinding and consider the amount and extent of measurement bias with detailed surveys, interviews or assessments, including a range of that may have entered the study due to imperfect blinding. potential exposures. It sounds plausible that such strategies may minimise the differential errors associated with recall bias on behalf Competing interests: Nil. of the participant as well as interviewer bias; however, there does not Sources of support: Nil. appear to be any evidence supporting this hypothesis. Acknowledgements: None. Provenance: Invited. Peer reviewed. Issues to consider when evaluating the success of blinding Correspondence: Erin Mathieu, Sydney School of Public Health, Faculty of Medicine and Health, The University of Sydney, Australia. Even when the investigators can withhold information about Email: [email protected] allocation from trial participants and others involved in the design, assessment and analysis of the study, these people may develop be- Erin Mathieu liefs about allocation. Their beliefs may be based on experiences of Sydney School of Public Health, Faculty of Medicine and Health, the intervention such as side effects or improvement in outcomes, or some other characteristic of the intervention such as taste or after- The University of Sydney, Australia taste of the intervention.19 Beliefs about allocation may be based on cues that are misinterpreted as evidence of receiving a particular References intervention. For example, a participant in a placebo-controlled trial of a non-steroidal anti-inflammatory drug who develops a gastric 1. Devereaux PJ, et al. Evid Based Ment Health. 2002;5:36–37. ulcer may perceive that they are in the intervention arm, even if they 2. Colditz GA, et al. Stat Med. 1989;8:441–454. are actually in the control arm. Or outcome assessors may perceive a 3. Schulz KF, et al. JAMA. 1995;273:408–412. participant to be in the intervention arm due to an improvement in 4. Miller JN, et al. Stat Med. 1989;8:455–466. their outcome, even if it is later established that the intervention was 5. Sackett DL. Int J Epidemiol. 2007;36:664–665. ineffective. Typically, people will have varying degrees of strength of 6. Day SJ, et al. BMJ. 2000;321:504. belief (some may be absolutely convinced, whereas others may have 7. Whiting PF, et al. J Clin Epidemiol. 2013;66:1093–1104. weak suspicions) and some beliefs are correct and others are incor- 8. Berlin JA. Lancet. 1997;350:185–186. rect.19 Based on evidence from placebo studies, the impact that these 9. Hoiggins J, Green S, eds. The Cochrane Collaboration. Cochrane handbook for sys- beliefs have on the study, whether correct or incorrect, can introduce bias in the estimates of effect.20 tematic reviews of interventions, 2011. 10. Buscemi N, et al. J Clin Epidemiol. 2006;59:697–703. There is considerable variability in researchers’ understandings 11. Schulz KF. Equine Vet J. 2005;37:394–395. about what defines the success or failure of blinding.21,22 It has been 12. Elkins M. J Physiother. 2013;59:134–136. claimed that blinding has been maintained when all or most partic- 13. Schulz KF, et al. Ann Intern Med. 2002;136:254–259. ipants (or assessors) believe that they received the experimental 14. Deyo RA, et al. Am J Phys Med Rehabil. 1990;69:6–10. intervention, or when all or most participants (or assessors) are un- 15. Chang A, et al. Arthritis Rheum. 2005;52:3515–3519. certain about which group they were allocated to, or when the beliefs 16. Kopec JA, et al. J Epidemiol Community Health. 1990;44:179–186. about allocation are no more accurate than chance.23,24 17. Herbert R. J Physiother. 2017;63:264–266. 18. Plinsinga ML, et al. Br J Sports Med. 2018;52:284–291. Despite the demonstrated importance of blinding, there is little 19. Mathieu E, et al. Australas Epidemiol. 2012;19:26–32. consensus regarding when, how or even if the adequacy of blinding 20. Kaptchuk TJ, et al. New Eng J Med. 2015;373:8–9. should be assessed. Some authorities recommend that the adequacy 21. Boutron I, et al. J Clin Epidemiol. 2005;58:1220–1226. of blinding should be explicitly assessed.21,25,26 Many methods have 22. Rees JR, et al. Contemp Clin Trials. 2005;26:25–37. been used to assess blinding and all involve asking about beliefs of 23. Bang H, et al. Control Clin Trials. 2004;25:143–156. allocation, and conclusions made regarding the success of blinding 24. James KE, et al. Stat Med. 1996;15:1421–1434. can vary depending upon the method that was chosen.27 There has, 25. Hrobjartsson A, et al. Int J Epidemiol. 2007;36:654–663. 26. Fergusson D, et al. BMJ. 2004;328:432. 27. Mathieu E. In: Internet Randomised Controlled Trials: An Evaluation of their uses, limitations and biases. Sydney: University of Sydney; 2012. 28. Schulz KF, et al. Ann Intern Med. 2010;152:726–732. 29. Schulz KF, et al. Lancet. 2010;375:1144–1146.
Journal of Physiotherapy 66 (2020) 39–44 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 State anxiety improves prediction of pain and pain-related disability after 12 weeks in patients with acute low back pain: a cohort study Joannes M Hallegraeff a, Ronald Kan a, Emiel van Trijffel a,b,c, Michiel F Reneman d a SOMT University of Physiotherapy, Amersfoort; b Experimental Anatomy Research Department, Department of Physiotherapy, Human Physiology and Anatomy, Faculty of Physical Education and Physiotherapy, Vrije Universiteit Brussel; c Department of Manual Therapy, Faculty of Medicine and Pharmacy, Vrije Universiteit Brussel, Brussels, Belgium; d University of Groningen, University Medical Center Groningen, Department of Rehabilitation Medicine, Groningen, The Netherlands KEY WORDS ABSTRACT Spinal low back pain Question: Do measures of state anxiety and trait anxiety in people with acute low back pain (ALBP) improve Anxiety prediction of chronic low back pain (CLBP), defined as pain or pain-related disability at 12 weeks? Design: Prognostic factor Observational multi-centre prospective cohort study in primary physiotherapy care with measurements at Risk baseline and at 12 weeks of state and trait anxiety, as well as other established prognostic factors for Prediction CLBP. Participants: People with nonspecific ALBP, aged 18 to 60 years, who had been pain free for 3 Psychological months before their current ALBP, and who were being treated according the Dutch clinical guidelines. Outcome measures and analysis: CLBP was defined as a pain score 3/10 on the Numerical Pain Rating Scale (primary outcome), and as a pain-related disability score 19/70 on the Pain Disability Inventory. Univariate and multivariate logistic regression analyses estimated how the risk of CLBP differed with state and trait anxiety and other established prognostic factors. Results: Most (204 of 225) participants completed both assessments. State anxiety was an independent predictor of CLBP, whether defined as pain or pain- related disability at 12 weeks, in contrast to trait anxiety. State anxiety improved the predictive perfor- mance of the model, with area under the curve (AUC) increasing from 0.64 (95% CI 0.56 to 0.71) to 0.75 (95% CI 0.68 to 0.82) and Nagelkerke’s R2 increasing from 0.08 to 0.24 for the primary outcome measure, pain. For the secondary outcome measure, pain-related disability: AUC 0.63 (95% CI 0.54 to 0.72) improved to 0.73 (95% CI 0.65 to 0.82) and Nagelkerke’s R2 increased from 0.05 to 0.16. Adding trait anxiety to the prognostic model for pain improved the AUC from 0.64 (95% CI 0.56 to 0.71) to 0.70 (95% CI 0.62 to 0.77) and Nagel- kerke’s R2 from 0.08 to 0.15. Conclusion: State anxiety in patients with ALBP improved prediction of CLBP, defined as pain and pain-related disability at 12 weeks. [Hallegraeff JM, Kan R, van Trijffel E, Reneman MF (2020) State anxiety improves prediction of pain and pain-related disability after 12 weeks in patients with acute low back pain: a cohort study. Journal of Physiotherapy 66:39–44] © 2019 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 is a temporary subjective condition that reflects feelings of fear and tension due to physical or perceived threats.6,7 In primary care physiotherapy, the natural course of acute low back pain (ALBP) is favourable, whereby a reduction in pain and Pain-related anxiety is related to perception of pain and thoughts disability can be expected within 4 to 6 weeks.1 However, recurrence about the consequences of pain8 and is associated with cognitive and rates within a year are high and the risk of chronic low back pain somatic anxiety symptoms and pain avoidance behaviour.9 Negative (CLBP) is substantial.1,2 Psychological factors such as depression, distress, negative expectations and somatisation can negatively in- emotions such as anxiety and fear increase pain when fear is an active fluence the course of low back pain and treatment outcomes.3 In addition, anxiety may also play a role in the transition of ALBP to CLBP response to a present threat and when anxiety is more defensive and and in the persistence or recurrence of pain.4 Anxiety is the antici- passively oriented.10 State anxiety or trait anxiety might facilitate the pation of a future threat, which leads to nervous behaviour, concen- transition from ALBP to CLBP.4,8,9 Higher anxiety-related sensations tration problems, inner agitation and somatic symptoms such as fatigue, worry, restlessness and muscular tension.5 ‘Trait anxiety’ is are associated with a decrease in perceived pain tolerance and more ever-present and reflects a person’s general and long-standing char- severe pain perceptions.11 In patients with chronic pain, trait anxiety acteristics regarding how they to respond to perceived threats; in- has also been found to be associated with pain intensity.12,13 Besides dividuals high in trait anxiety react with high ‘state anxiety’,5 which anxiety, fear of pain plays a role in fear-avoidance behaviour and pain intensity in people with ALBP.14 Anxiety and fear of pain are assumed to be related to avoidance behaviour and reduced physical performance.13,15 There is evidence that targeting anxiety can reduce pain.16 https://doi.org/10.1016/j.jphys.2019.11.011 1836-9553/© 2019 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/).
40 Hallegraeff et al: State anxiety in acute low back pain Few studies have investigated the role of state anxiety and trait obtained at baseline and at 12 weeks in a standardised sequence: anxiety in the transition from ALBP to CLBP. State anxiety has been Numeric Pain Rating Scale; State and Trait Anxiety Inventory (STAI); found to be similar between patients with ALBP and those with CLBP, and Pain Disability Inventory.21–23 but trait anxiety was significantly higher in CLBP.15 However, that study included patients in tertiary care and did not specify whether State and trait anxiety back pain was of specific or non-specific origin. In the current study, we hypothesised that state anxiety and trait anxiety could be pre- Levels of state and trait anxiety were measured using the revised dictive of poor outcomes in patients with ALBP. STAI scales in the authorised Dutch translation (STAI-Y).6,24,25 The STAI-Y is a self-reported questionnaire comprising two subscales of Therefore, the research question for this observational multi- 20 items each. The State Anxiety scale (STAI-S) evaluates the current centre prospective cohort study was: state of anxiety, asking how respondents feel ‘at this moment’ using items that measure subjective feelings of tension, nervousness, worry Do measures of state anxiety and trait anxiety in people with and arousal. The Trait Anxiety scale (STAI-T) evaluates relatively acute low back pain improve prediction of chronic low back pain, stable aspects of ‘anxiety disposition’, including general states of defined as pain or pain-related disability at 12 weeks? tranquillity, confidence and safety. The response items for the STAI-S include: ‘not at all’, ‘somewhat’, ‘moderately so’ and ‘very much so’. Methods The STAI-T response options are: ‘almost never’, ‘sometimes’, ‘often’ and ‘almost always’. Ten items of each subscale should be reversed Design and the subsequent summing of the 20 items results in scores ranging from 20 to 80, with higher scores reflecting higher levels of anxiety. The STROBE statement was used as a guide to report this study, Internal consistency (Cronbach’s alpha) was measured among Dutch and the study methods and results are reported according to the students and reached scores of 0.87 to 0.95 for the STAI-S and 0.85 to REMARK checklist.17 A longitudinal multi-centre prospective cohort 0.94 for the STAI-T. Test-retest reliability in a healthy population study design with two measurement occasions was conducted. showed Pearson correlation coefficients (rp) for the STAI-T 0.86 and Recruitment occurred at 26 primary care physiotherapy practices for STAI-S 0.52.25 Construct validity of the STAI is good with rp 0.73 throughout the Netherlands. The study was a Phase II confirmatory and rp 0.85 with the Cattell and Scheier’s Anxiety Scale Questionnaire, prognostic study to establish the independent prognostic value of respectively.26 state anxiety and trait anxiety in patients with ALBP.18,19 Known factors adjusted in the model Physiotherapists working in primary care and who were studying a Master of Science degree in Manual Therapy were requested to The following factors were considered as ‘known’ prognostic participate in the study by collecting data on consecutive patients factors in the model because there is evidence27–31 that they increase with ALBP. Participation was based on personal commitment; as a the risk of CLBP: pain intensity at onset, assessed with the Numeric consequence, geographically, the centres and practices were spread Pain Rating Scale, with scores ranging from 0 to 10; pain-related randomly throughout the Netherlands. All patients with ALBP disability measured by the Pain Disability Inventory; duration of attended a baseline assessment in the acute stage (, 6 weeks), within low back pain measured as number of days having low back pain; 1 week after initial contact, and all self-reported data were collected physical workload (classified in four categories: office worker, driver, before history taking and physical examination. Patients were reas- physically heavy work and healthcare worker); and widespread pain sessed 12 weeks after baseline assessment for pain and disability. (classified in three categories: lumbosacral region, radiation proximal Confidential records were stored in a secure area with limited access. to the knee and radiation distal to the knee). Anonymity and confidential input of data were guaranteed to protect the health status information of the participants. Encoded data were Outcome measures entered in an encrypted electronic database. At 12 weeks, pain and pain-related disability were assessed using Participants the Numeric Pain Rating Scale and the Pain Disability Inventory, respectively. The Numerical Pain Rating Scale is a one-dimensional People with ALBP were consecutively included if they were aged measure of pain intensity in adults. The Pain Disability Inventory is 18 to 60 years, reported low back pain of , 6 weeks duration with or a seven-item ordinal scale to measure the impact of pain on seven without radiating pain, and had been pain-free for at least 3 months domains of functioning on a scale ranging from 0 to 70, where 70 is before the onset of their current back pain. They were also required to the most severe score. Test-retest reliability of the Pain Disability be able to read and understand the Dutch language. People were Inventory is adequate (ICC 0.76) and the minimum clinically impor- excluded if they had undergone previous lumbar spinal surgery, tant change is between 8.5 and 9.5 points.22,23 To classify CLBP pa- experienced any neurological signs in their lower limbs, or were tients as having the outcome pain at 12 weeks, a cut-off for pain was diagnosed with a specific cause of low back pain such as lumbar defined as greater than ‘very mild’ pain that resulted in a pain score spinal stenosis, current malignancy, spondyloarthropathy, osteopo- 3/10 on the Numerical Pain Rating Scale and with the outcome rosis, spondylolisthesis, major trauma, infection or systemic disease. pain-related disability at 12 weeks, a score 19/70 on the Pain During follow-up, standard care physiotherapy according to the Disability Inventory.31,32 Dutch clinical practice guideline for low back pain was carried out.20 Prior to participation, the study protocol was explained verbally and Data analysis in writing to patients, after which patients signed informed consent. It was expected to find a prevalence of anxiety of 30% in the ALBP Data collection population.24 Using this estimate, the a priori sample size calculation revealed that to ensure the width of the 95% confidence interval Data collection was carried out from January 2016 to March 2016 would be 6 6.9% around a proportion of 30% having anxiety as and January 2017 to March 2017. All self-reported measures were determined using the STAI scale, at least 170 participants would be obtained in a separate room prior to physical examination and required. standard care. Two days before baseline assessment, each participant was contacted by email or telephone as a reminder. Participants were Univariate logistic regression analyses were performed to estimate asked to complete blank items in the event of non-completion of the independent, uncorrected association of state anxiety and trait questionnaire items. Data from the following questionnaires were anxiety with pain and pain-related disability. Subsequently, separate multivariate logistic regression models were produced to estimate
Research 41 People presenting to primary Table 2 care physiotherapy with acute Univariate regression and multivariate regression analyses of state anxiety for pain at 12 weeks, classified as a Numerical Pain Rating Scale score 3/10, adjusted for pain, low back pain (n = 301) pain-related disability, duration, physical workload and widespread pain. Excluded (n = 69) Model Variable Pain . specific low back pain Beta Odds Ratio (95% CI) p-value . older age Unadjusted State anxiety 0.08 1.1 (1.0 to 1.1) 0.00 Adjusted State anxiety 0.08 1.1 (1.1 to 1.1) 0.00 Pain 0.29 1.3 (1.1 to 1.7) 0.01 Pain-related disability –0.02 1.0 (1.0 to 1.0) 0.16 Duration 0.14 1.2 (0.99 to 1.3) 0.06 Physical workload 0.01 1.0 (0.9 to 1.2) 0.89 Widespread pain 0.17 1.2 (0.7 to 2.1) 0.58 Eligible (n = 232) anxiety and trait anxiety and outcomes were calculated as b co- Declined (n = 7) efficients, odds ratios (ORs) with their 95% confidence interval (CI), and p-values. To prevent over-fitting, the total number of variables included in the model was limited to five variables to ensure 10 events per variable.33 Included (n = 225) Model fit statistics .Lost to follow-up (n = 21) To quantify how close predictions were to the actual outcome, the did not attend second Brier test was used, ranging from 0 to 0.25, whereby a score closer to assessment 0 indicates good performance.18,34 Explained variance was calculated applying the standard procedure of Nagelkerke’s R2. Whether the Followed up 12 weeks later prediction of the model aligned with the observed probability of (n = 204) developing CLBP was assessed by the Hosmer-Lemeshow goodness- of-fit test, where a larger p-value indicates a better fit. Figure 1. Flow diagram of participants through the study. It was hypothesised that an increase of 10% chance would be the independent contribution of state anxiety and trait anxiety considered clinically relevant for predicting CLBP by adding state or adjusted for the effects of currently known prognostic factors for trait anxiety in the prognostic model, adjusted with the five ‘known’ predicting low back pain status at 12 weeks. These five baseline factors as shown by the area under the Receiver Operating Charac- variables, the ‘known’ factors (pain intensity, pain-related disability, teristic (ROC) curve (AUC). Moreover, an AUC 0.70 would be pain duration, physical work load and widespread pain), were forced acceptable.35 It was also hypothesised that explained variance must into the prognostic model as adjusted irrespective of their univariate show improvement . 0.20 when state or trait anxiety is added. association with the outcomes.27–30 Associations between state Significance level for the multivariate analysis was set at a = 0.05. All analyses were performed using IBM SPSS Statistics v. 25.0. Table 1 Results Baseline demographic and clinical characteristics of the study completers, and values 12 weeks after enrolment in the study. The flow of participants through the study is shown in Figure 1. Baseline demographic and clinical characteristics of the study par- Characteristic Baseline 12 weeks ticipants with complete datasets (n = 204) are presented in Table 1. (n = 204) (n = 204) The normally distributed data showed no floor or ceiling effects. Univariate regression and multivariate regression analyses were Age (y), mean (SD) 41 (12) performed on state anxiety and trait anxiety for both pain and pain- Gender, n female (%) 103 (51) related disability at 12 weeks, and adjusted for pain, pain-related Recurrent, n (%) 148 (73) disability, duration, physical workload and widespread pain. Localisation, n (%) 152 (75) The adjusted multivariate regression of state anxiety showed an lumbosacral region 42 (21) increase of 10% in the probability of developing CLBP with every unit radiation to proximal knee 10 (5) increase in the score on the STAI-S. In addition, pain intensity at onset radiation to distal knee was also an independent predictor for CLBP with OR 1.3 (95% CI 1.1 to Workload, n (%) 88 (43) 1.7) and p = 0.01 (Table 2). office worker 6 (3) driver Table 3 heavy physical work 71 (35) Univariate regression and multivariate regression analyses of state anxiety for pain- healthcare worker 35 (17) related disability at 12 weeks, measured as a Pain Disability Inventory score 19/70, Physically active, n (%) 141 (69) adjusted for pain, pain-related disability, duration, physical workload and Body mass index (kg/cm2), mean (SD) 24.9 (6.9) widespread pain. State anxiety (STAI-S), mean (SD) 35.0 (11.0) Trait anxiety (STAI-T), mean (SD) 35.2 (10.5) Model Variable Pain Pain intensity (NPRS 0/10), mean (SD) 6.4 (1.7) Greater than ‘very mild’ pain 3/10 NPRS, n (%) 25.9 (14.2) Beta Odds Ratio (95% CI) p-value Disability (PDI 0/70), mean (SD) 24.7 (14.7) 27.1 (15.6) Greater than ‘mild’ disability 19/70 PDI, n (%) 2.2 (2.3) Unadjusted State anxiety 0.07 1.1 (1.0 to 1.1) 0.00 Adjusted State anxiety 0.07 1.1 (1.0 to 1.2) 0.00 73 (36) Pain 0.03 1.0 (0.8 to 1.3) 0.80 8.3 (11.0) Pain-related disability 0.02 1.0 (1.0 to 1.1) 0.14 35 (17) Duration 0.03 1.0 (0.9 to 1.2) 0.76 Physical workload –0.09 0.9 (0.8 to 1.1) 0.34 NPRS = Numerical Pain Rating Scale, PDI = Pain Disability Inventory, STAI-S = State and Widespread pain –0.15 0.9 (0.4 to 1.7) 0.66 Trait Anxiety Inventory-State, STAI-T = State and Trait Anxiety Inventory-Trait.
42 Hallegraeff et al: State anxiety in acute low back pain Table 4 1,0 Univariate regression and multivariate regression analyses of trait anxiety for pain at 12 weeks, classified as a Numerical Pain Rating Scale score 3/10, adjusted for pain, pain-related disability, duration, physical workload and widespread pain. Model Variable Pain 0,8 Beta Odds Ratio (95% CI) p-value Unadjusted Trait anxiety 0.05 1.1 (1.0 to 1.1) 0.01 Adjusted Trait anxiety 0.05 1.1 (1.0 to 1.1) 0.01 Pain 0.29 1.3 (1.1 to 1.6) 0.01 Pain-related disability –0.01 1.0 (0.9 to 1.0) 0.24 Sensitivity 0,6 Duration 0.15 1.2 (1.0 to 1.3) 0.049 Physical workload 0.02 1.0 (0.9 to 1.2) 0.83 0,4 Widespread pain 0.17 1.2 (0.7 to 2.0) 0.55 The adjusted multivariate logistic regression of state anxiety 0,2 showed an increase of 10% in the probability of developing CLBP with every unit increase in the score on the STAI-S. No other independent 0,0 0,2 0,4 0,6 0,8 1,0 predictors were identified (Table 3). 0,0 1 – Specificity Reference line State anxiety added The adjusted multivariate logistic regression of trait anxiety Five known factors showed an increase of 10% in the probability of developing CLBP with every unit increase in the score on the STAI-T. Pain intensity and duration of pain were independent predictors for CLBP (Table 4.) The adjusted multivariate logistic regression of trait anxiety showed no additional predictive value. No other independent pre- dictors were identified (Table 5). Model fit statistics Figure 2. Receiver Operating Characteristic (ROC) curve of state anxiety (red line) to visualise the discriminatory power. The outcome is the primary outcome: pain at 12 The predictive performance of the model was quantified in terms weeks. of overall fit and discrimination. Discrimination with state anxiety added improved into acceptable accuracy for the primary outcome, anxiety, not trait anxiety, is a prognostic factor and may be of addi- pain: AUC improved from 0.64 (95% CI 0.56 to 0.71) to 0.75 (95% CI tional value in the clinical setting. 0.68 to 0.82) and Nagelkerke’s R2 improved from 0.08 to 0.24. Discrimination with state anxiety added improved into acceptable Current guidelines for low back pain emphasise that psychological accuracy for the secondary outcome, pain-related disability: AUC factors, of which anxiety is one, are risk factors for poor prognosis at improved from 0.63 (95% CI 0.54 to 0.72) to 0.73 (95% CI 0.65 to 0.82) the acute stage.19 Although trait anxiety reflects a person’s long- and Nagelkerke’s R2 improved from 0.05 to 0.16. Discrimination with standing characteristic, it is remarkable that trait anxiety showed a trait anxiety added improved for the primary outcome, pain: AUC substantial decrease during the 12-week period in the transition from improved from 0.64 (95% CI 0.56 to 0.71) to 0.70 (95% CI 0.62 to 0.77) ALBP to CLBP. This implies that the general characteristics of trait and Nagelkerke’s R2 improved from 0.08 to 0.15. On the secondary anxiety may be debatable, although in a nonclinical population both outcome, no improvement was detectable. Hosmer and Lemeshow state anxiety and trait anxiety scores differed, which was supported (p-values) varied between 0.22 and 0.94 and Brier scores varied be- by neuroanatomical correlations.36 In addition, mean anxiety scores tween 0.13 and 0.22. at baseline were different from values found in other studies and it is The ROC curves illustrate the discrimination of the primary and 1,0 secondary outcomes due to the addition of state anxiety to the model (Figures 2 and 3). 0,8 Discussion This prospective prognostic cohort study found evidence that Sensitivity 0,6 state anxiety in people with ALBP is an independent predictor of the risk of developing CLBP at 12 weeks, in contrast with trait anxiety. In 0,4 the multivariate adjusted model, the addition of state anxiety improved discrimination between people at risk or not of developing CLBP in contrast with adding trait anxiety. This suggests that state Table 5 0,2 Univariate regression and multivariate regression analyses of trait anxiety for pain- related disability at 12 weeks, measured as a Pain Disability Inventory score 19/70, adjusted for pain, pain-related disability, duration, physical workload and widespread pain. Model Variable Pain 0,0 Odds Ratio (95% CI) 0,0 Beta p-value 0,2 0,4 0,6 0,8 1,0 1 – Specificity Reference line Unadjusted Trait anxiety 0.03 1.0 (0.9 to 1.1) 0.11 State anxiety added Adjusted Trait anxiety 0.02 1.0 (0.9 to 1.1) 0.26 Pain 0.06 1.0 (0.8 to 1.4) 0.61 Pain-related disability 0.02 1.0 (0.9 to 1.1) 0.09 Five known factors Duration 0.02 1.0 (0.9 to 1.2) 0.80 Physical workload –0.07 0.9 (0.8 to 1.1) 0.49 Figure 3. Receiver Operating Characteristic (ROC) curve of state anxiety (red line) to Widespread pain –0.09 0.9 (0.5 to 1.7) 0.78 visualise the discriminatory power. The outcome is the secondary outcome: pain- related disability at 12 weeks.
Research 43 unknown what represents valid anxiety scores in CLBP. Newcomer that, within the Dutch regulations, formal ethical approval was not et al recruited participants from a tertiary care clinic in which trait anxiety was more strongly present than state anxiety.15 In contrast to needed. All participants gave written informed consent before data the current sample of patients with ALBP, Bean et al found no sig- nificant role for state anxiety to predict disability and pain in patients collection. with CLBP.16 Their study did not distinguish between state anxiety and trait anxiety, and did not measure anxiety at the acute stage.16 In Competing interests: Nil. addition, a systematic review found limited evidence that anxiety is predictive of disability in patients with subacute, non-malignant pain Source(s) of support: Nil. conditions.37 Acknowledgements: Thanks to all patients and physiotherapists Strengths of this study were: the use of a thoroughly validated instrument for measuring state and trait anxiety, with good who participated in this study. discriminant properties and ease of implementation in clinical prac- tice; and the recruitment in primary care with acceptable loss to Provenance: Not invited. Peer reviewed. follow-up. In addition, the sample size and event rate appeared suf- ficient to allow a maximum of five variables in the model and, Correspondence: Joannes M Hallegraeff, SOMT University consequently, provide stable predictors. However, analysing missing values at baseline showed that pain intensity and disability were 10% of Physiotherapy, Amersfoort, The Netherlands. Email: and 17% below the mean scores, respectively. In contrast to what was expected, the adjusted prognostic factor effects of state anxiety and [email protected] trait anxiety were approximately equal to the unadjusted ones. The predictive performance of the prognostic model is quantified in terms References of overall measures of fit and discrimination, whereby the increase in the explained variance may be regarded as a relevant scientific and 1. Maher C, Underwood M, Buchbinder R. Non-specific low back pain. Lancet. clinical contribution. When focused on state anxiety as a prognostic 2017;389:736–737. factor, the explained variance tripled on both primary and secondary outcome measures. In addition, on both outcomes the AUC showed 2. Kent PM, Keating JL. Can we predict poor recovery from recent-onset nonspecific substantial improvement in discriminant capacity by adding state low back pain? A systematic review. Man Ther. 2008;13:12–28. anxiety, which was in contrast to trait anxiety. Nevertheless, a large part of the variance remains unexplained; this should be investigated 3. Ramond A, Bouton C, Richard I, Roquelaure Y, Baufreton C, Legrand E, et al. Psy- in further research. chosocial risk factors for chronic low back pain in primary care–a systematic re- view. Fam Pract. 2011;28:12–21. A limitation of the study results is that only complete cases were used and imputation was not used. However, it is debatable whether 4. Kessler RC. The epidemiology of pure and comorbid generalized anxiety disorder: a the bias of imputation would be less than the selection bias that may review and evaluation of recent research. Acta Psychiatr Scand Suppl. 2000;406:7–13. arise from using only complete cases.18 Selection bias may have been present, when selection was based on incorrect identification of pa- 5. American Psychiatric Association. Diagnostic and Statistical Manual of Mental Dis- tients at baseline and during follow-up in patients who had with- orders. (5th ed). Arlington: American Psychiatric Publishing; 2013. drawn. In our study we assumed that these cases were a random selection of the data, although analysing these cases would have been 6. Spielberger CD, Gorsuch RL, Lushene R, Vagg PR, Jacobs GA. Manual for the State- preferable. Trait Anxiety Inventory. Palo Alto, CA: Consulting Psychologists Press; 1983. The results of this study are in line with recommendations 7. van der Ploeg HM, Defares PB, Spielberger CD. Handleiding bij de Zelf-Beoordelings emphasising the need for attention to psychological factors in phys- Vragenlijst ZBV. Een Nederlandse bewerking van de Spielberger State-Trait Anxiety iotherapy care.24 Although it requires specific professional skills, ev- Inventory STAI-DY. Amsterdam: Pearson Assessment and Information BV; 2000. idence exists that patient education can provide reassurance.38 Replication of this study’s findings is needed in other multiple in- 8. Campbell P, Bishop A, Dunn KM, Main CJ, Thomas E, Foster NE. Conceptual overlap dependent confirmatory studies to assess the best way of classifying of psychological constructs in low back pain. Pain. 2013;154:1783–1791. state anxiety and also to investigate the mechanisms that lie behind the construct of state anxiety in nonspecific low back pain, how it is 9. Lumley MA, Cohen JL, Borszcz GS, Cano A, Alison M, Porter LS, et al. Pain and affected, and the role of physiotherapists in the management of these emotion: a biopsychosocial review of recent research. J Clin Psychol. 2011;67:942– patients in clinical practice. It is recommended that clinicians 968. consider state anxiety as a potential predictor of poor outcome in patients with ALBP. Further research on anxiety as a psychosocial 10. Thibodeau MA, Welch PG, Katz J, Asmundson GJ. Pain-related anxiety influences factor and as part of screening instruments in ALBP patients is pain perception differently in men and women: a quantitative sensory test across warranted.39 thermal pain modalities. Pain. 2013;154:419–426. What is already known on this topic: Several factors 11. Esteve R, Ramírez-Maestre C, López-Martínez AE. Experiential avoidance and contribute to the transition from acute to chronic low back pain. anxiety sensitivity as dispositional variables and their relationship to the adjust- However, the influence of anxiety is unknown. ment to chronic pain. Eur J Pain. 2012;16:718–726. What this study adds: Patients with acute low back pain presenting in primary care with higher levels of state anxiety 12. Leeuw M, Goossens ME, Linton SJ, Crombez G, Boersma K, Vlaeyen JW. The fear- have higher risk of developing chronic low back pain and pain- avoidance model of musculoskeletal pain: current state of scientific evidence. related disability at 12 weeks. Further validation of these find- J Behav Med. 2007;30:77–94. ings is warranted. Meanwhile, physiotherapists could consider identifying those patients at-risk, in order to intervene and 13. McNally RJ. Anxiety sensitivity and panic disorder. Biol Psychiatry. 2002;52:938– decrease the risk of chronicity. 946. Ethics approval: The Medical Ethics Committee (METC) of the 14. George SZ, Fritz JM, McNeil DW. Fear-avoidance beliefs as measured by the fear- University of Groningen reviewed the study procedures and decided avoidance beliefs questionnaire: change in fear-avoidance beliefs questionnaire is predictive of change in self-report of disability and pain intensity for patients with acute low back pain. Clin J Pain. 2006;22:197–203. 15. Newcomer KL, Shelerud RA, Vickers Douglas KS, Larson DR, Crawford BJ. Anxiety levels, fear-avoidance beliefs, and disability levels at baseline and at 1 year among subjects with acute and chronic low back pain. PM R. 2010;2:514–520. 16. Bean DJ, Johnson MH, Kydd RR. Relationships between psychological factors, pain, and disability in complex regional pain syndrome and low back pain. Clin J Pain. 2014;30:647–653. 17. Vandenbroucke JP, von Elm E, Altman DG, Gøtzsche PC, Mulrow CD, Pocock SJ, et al. Strengthening the Reporting of Observational Studies in Epidemiology (STROBE): explanation and elaboration. Int J Surg. 2014;12:1500–1524. 18. Riley RD, van der Windt DA, Croft P, Moons KGM. Prognosis research in health care, Concepts, Methods, and Impact. 1st ed. Oxford, UK: Oxford University Press; 2019. 19. Hayden JA, Côté P, Steenstra IA, Bombardier C. Identifying phases of investigation helps planning, appraising, and applying the results of explanatory prognosis studies. J Clin Epidemiol. 2008;61:552–560. 20. Staal JB, Hendriks EJ, Heijmans M, Kiers H, Lutgers-Boomsma AM, Rutten G, et al. KNGF-richtlijn Lage rugpijn [KNGF low back pain guideline]. Amersfoort, Netherlands: Koninklijk Nederlands Genootschap voor Fysiotherapie; 2013. 21. Van der Ploeg HM. De zelf-beoordelings vragenlijst (STAI-DY). Tijdschr Psychiatr. 1982;24:576–588. 22. Soer R, Köke AJ, Speijer BL, Vroomen PC, Smeets RJ, Coppes MH, et al. Reference values of the pain disability index in patients with painful musculoskeletal and spinal disorders: a cross-national study. Spine. 2015;40:E545–E551. 23. Soer R, Reneman MF, Vroomen PC, Stegeman P, Coppes MH. Responsiveness and minimal clinically important change of the Pain Disability Index in patients with chronic back pain. Spine. 2012;37:711–715. 24. Bener A, Verjee M, Dafeeah EE, Falah O, Al-Juhaishi T, Schlogl J, et al. Psychological factors: anxiety, depression, and somatization symptoms in low back pain patients. J Pain Res. 2013;6:95–101. 25. Knight RG, Waal-Manning HJ, Spears GF. Some norms and reliability data for the State-Trait Anxiety Inventory and the Zung Self-Rating Depression scale. Br J Clin Psychol. 1983;22:245–249. 26. Cattell RB, Scheier IH. Handbook for the IPAT Anxiety Scale Questionnaire: Self Analysis Form. (2nd ed). Champaign: Institute for Personality and Ability Testing; 1963. 27. Hayden JA, Dunn KM, van der Windt DA, Shaw WS. What is the prognosis of back pain? Best Pract Res Clin Rheumatol. 2010;24:167–179.
44 Hallegraeff et al: State anxiety in acute low back pain 28. Campbell P, Foster NE, Thomas E, Dunn KM. Prognostic indicators of low back pain 35. Mandrekar JN. Receiver operating characteristic curve in diagnostic test assess- in primary care: five-year prospective study. J Pain. 2013;14:873–883. ment. J Thorac Oncol. 2010;5:1315–1316. 29. Artus M, Campbell P, Mallen CD, Dunn KM, van der Windt DA. Generic prognostic 36. Donzuso G, Cerasa A, Gioia MC, Caracciolo M, Quattrone A. The neuroanatomical factors for musculoskeletal pain in primary care: a systematic review. BMJ Open. correlates of anxiety in a healthy population: Differences between the state-trait 2017;7:e012901. anxiety inventory and the Hamilton anxiety rating scale. Brain Behav. 2014;4:504–514. 30. Chou R, Shekelle P. Will this patient develop persistent disabling low back pain? JAMA. 2010;303:1295–1302. 37. Valentin GH, Pilegaard MS, Vaegter HB, Rosendal M, Ørtenblad L, Væggemose U, et al. Prognostic factors for disability and sick leave in patients with subacute 31. Traeger AC, Hubscher M, Henschke N, Moseley GL, Lee H, McAuley JH. Effect of non-malignant pain: a systematic review of cohort studies. BMJ Open. primary care-based education on reassurance in patients with acute low back pain: 2016;6:e007616. systematic review and meta-analysis. JAMA Intern Med. 2015;175:733–743. 38. Steffens D, Maher CG, Pereira LS, Stevens ML, Oliveira VC, Chapple M, et al. Pre- 32. Violante FS, Mattioli S, Bonfiglioli R. Low-back pain. Handb Clin Neurol. vention of low back pain: a systematic review and meta-analysis. JAMA Intern Med. 2015;131:397–410. 2016;176:199–208. 33. Peduzzi P, Concato J, Kemper E, Holford TR, Feinstein AR. A simulation study of the 39. Karran EL, McAuley JH, Traeger AC, Hillier SL, Grabherr L, Russek LN, et al. Can number of events per variable in logistic regression analysis. J Clin Epidemiol. screening instruments accurately determine poor outcome risk in adults with 1996;49:1373–1379. recent onset low back pain? A systematic review and meta-analysis. BMC Medicine. 2017;15:1–15. 34. Steyerberg EW. Clinical Prediction Models A Practical Approach to Development, Validation and Updating. (4th ed). New York: Springer-Verlag; 2009.
Journal of Physiotherapy 66 (2020) 45–53 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 Therapeutic alliance facilitates adherence to physiotherapy-led exercise and physical activity for older adults with knee pain: a longitudinal qualitative study Andrew J Moore a,b, Melanie A Holden a, Nadine E Foster a, Clare Jinks a a Primary Care Centre Versus Arthritis, School of Primary, Community and Social Care, Keele University, Keele; b Musculoskeletal Research Unit, Bristol Medical School, University of Bristol, Bristol, United Kingdom KEY WORDS ABSTRACT Physiotherapy Questions: What are people’s experiences and perceived impact of physiotherapist-led exercise Therapeutic alliance interventions for knee pain attributable to osteoarthritis? What barriers and facilitators to change in Osteoarthritis exercise and physical activity behaviour exist over time? Design: A longitudinal qualitative study was un- Knee dertaken; it involved face-to-face, semi-structured and longitudinal interviews. Participants: Interviews Exercise were undertaken with older adults with knee pain and who had been randomised to one of three Physical activity physiotherapist-led exercise intervention arms in the Benefits of Effective Exercise for knee Pain (BEEP) trial. Thirty participants were enrolled in this qualitative study, with interviews scheduled at the end of the trial intervention period and 12 months later. Data analysis: A ‘layered approach’ to thematic analysis was used, including open coding (using constant comparison), deductive coding and within-case and cross-case longitudinal analysis of change. Results: Different levels of exercise supervision, progression and individu- alisation emerged, matching the content of the intervention protocols. Barriers to exercise and general physical activity were similar across intervention arms (lack of motivation, time, physical environment, lack of supervision and/or monitoring). Despite individualising exercise programs and specifically targeting ex- ercise, some barriers to adherence remained at 12 months. Factors facilitating longer-term exercise adherence included change in or retained knowledge about the role of exercise for knee pain and the presence and quality of a therapeutic alliance, which was also reflective of the participants’ experience of the intervention, regardless of the trial arm. Conclusion: Despite a focus on individualisation and exercise adherence, barriers remained in the longer term. Strong therapeutic alliance during treatment appeared to facilitate adherence to exercise and general physical activity. The findings highlight ongoing physiotherapy support and therapeutic alliance as targets for future adherence-enhancing interventions for exercise in older adults with knee pain. [Moore AJ, Holden MA, Foster NE, Jinks C (2020) Therapeutic alliance facilitates adherence to physiotherapy-led exercise and physical activity for older adults with knee pain: a longitudinal qualitative study. Journal of Physiotherapy 66:45–53] © 2019 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 pain, physical limitations, low self-efficacy, depression, anxiety, time constraints, poor social support, non-positive experiences of physical Musculoskeletal conditions, after mental health, are the leading cause of disability worldwide.1,2 Osteoarthritis (OA) is associated with activity and beliefs about OA, exercise and pain also impact on ex- more limitations in activities of daily living among older people than ercise adherence.8–12 External factors include physiotherapists’ pro- any other disease.3 Exercise, including local muscle strengthening fessional care and encouragement towards patients, while and general physical activity (general aerobic fitness), is recom- mended as a core treatment.4 Although there is evidence of small-to- environmental factors include weather and suitable/accessible exer- moderate benefits from exercise and physical activity for knee OA, cise environments.7,9,13 Clinical recommendations for promoting ex- these decline over time, which is potentially explained by reduced adherence.5 Previous research has identified a wide array of factors ercise adherence include: improving the nature of the relationship that influence adherence to exercise and physical interventions for knee OA, including: the person’s physical capacity to exercise; pre- between patient and provider; individually tailoring exercise pro- vious exercise history; symptom relief and improvement; positive exercise experiences and beliefs; healthcare professional and social grams; and better supervision to facilitate the prescription of pro- support; and expectations of improvement.6–12 Lack of motivation, gressive exercise programs tailored to suit the changing needs of patients.7,11,13 However, whether the influence of these factors changes over time, regarding the uptake and maintenance of exercise and physical activity behaviour, remains unexplored.12 Understanding whether and how these factors change over time can potentially provide https://doi.org/10.1016/j.jphys.2019.11.004 1836-9553/© 2019 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/).
46 Moore et al: Physiotherapy-led intervention for knee pain evidence that will enable the development and optimisation of ex- intervention period; and follow-up interviews approximately 12 ercise interventions for OA, to provide maximum benefit for patients months after the end of the intervention period. Interviews were over the longer term. Additionally, the majority of studies have undertaken by AM, MH and DR, all of whom are experienced in focused on prescribed exercise rather than general physical activity.12 qualitative research methods. The post-intervention interview topic guide included open questions to explore participants’ views on: The Benefits of Effective Exercise for knee Pain (BEEP) trial14 aimed what happened during physiotherapy sessions; the impact of treat- to test whether enhanced physiotherapist-led exercise interventions ment on exercise and general physical activity behaviour; and beliefs could provide larger and longer-lasting benefits for people with knee about changes in knee symptoms and exercise behaviour over the OA. It was hypothesised that: greater individualisation, supervision and course of treatment. Follow-up interviews explored participants’ ex- progression of a lower limb exercise program would be superior to periences of: exercise and treatment over the longer term; percep- usual National Health Service (NHS) physiotherapy exercise; and a tions of adherence to the prescribed exercises and changes in physical program that targets exercise adherence in the longer term, supporting activity behaviour over time; and impact on their knee (the topic the transition from lower-limb exercise to general physical activity, guides are presented in Appendices 1 and 2 on the eAddenda). All would be superior to usual NHS physiotherapy exercise. interviews were conducted in the participants’ homes. Field notes were written up as soon as possible after the interviews to record the Within the trial, participants were randomised to one of three interviewers’ immediate impressions. Interviews were audio- groups. The usual physiotherapy care (UC) control group received up recorded (with the participant’s consent), transcribed by an to four treatment sessions of advice and exercise over 12 weeks (n = independent university-approved transcription company and ano- 176 participants). The individually tailored exercise (ITE) group nymised. Commercial software was used for data management.25 received six to eight sessions of individualised, supervised and pro- gressed lower limb exercises over 12 weeks (n = 178 participants). Data analysis The targeted exercise adherence (TEA) group were transitioned from lower limb exercise to general physical activity, with eight to ten A ‘layered approach’ to thematic analysis was undertaken.26 Open contacts over 6 months (n = 172 participants). The interventions are coding of all transcripts identified: experiences of interventions; described in full elsewhere and this paper contains some information views on the trial exercises and general physical activity; and barriers included in the report summarising the much wider research pro- and facilitators to exercise. Adopting constant comparison, transcripts gramme of which this was part.14–16 The trial results showed that were read and re-read, an initial coding framework was devised and participants receiving UC, ITE and TEA, on average, experienced codes were grouped into themes by AM.27,28 CJ, MH and NF inde- moderate improvements in pain and function, as measured by the pendently coded a sample of transcripts. Codes and initial themes Western Ontario and McMaster Universities Arthritis Index were discussed, agreed, and then applied to the dataset, with ongoing (WOMAC).17 However, there were no significant differences between refinement. Data collection and analysis were carried out iteratively arms at 6 months (primary outcome), or at other time points (3, 6, 9 so that emerging themes (post-intervention) could be explored dur- and 18 months) and adherence declined over time in all arms.16 ing follow-up interviews. Sampling continued until no new themes emerged, indicating data saturation.29 Next, deductive coding This paper reports findings from a program evaluation utilising included pre-determined codes of individualisation, supervision and qualitative interviews alongside the trial to understand the experi- progression because these were core characteristics of the trial ence of trial participants. Qualitative methods are important in trials interventions.14 as they: help to explain how complex interventions are implemented within a particular social context; facilitate interpretation of trial Finally, longitudinal qualitative analysis was used, which is outcomes; and highlight to researchers any future trials in which particularly useful when studying changes in health behaviour and interventions are most likely to be effective.18–22 Furthermore, a within program evaluations to further an understanding of not just longitudinal qualitative approach was used to understand adherence whether an intervention is perceived to work, but why it works and to exercise and physical activity across the course of the trial.23,24 This whether it is feasible and acceptable over time.24,30 Within-case and study aimed to investigate participants’ experiences of treatment, and cross-case longitudinal analyses using descriptive and interpretive barriers and facilitators to exercise and general physical activity questions were used to determine elements of behaviour change behaviour in the longer term. within the data.23 Summaries of each participant’s data were used to identify perceptions of changes in the trial exercises and general Therefore, the research questions for this longitudinal qualitative physical activity, and influences of change during the period from study were: post-intervention to follow-up (within case) and in each intervention arm (cross-case).23,26 1. What are people’s experiences and perceived impact of physiotherapist-led exercise interventions for knee pain attribut- Results able to osteoarthritis? Participants 2. What barriers and facilitators to change in exercise and physical activity behaviour exist over time? Thirty participants consented to and completed a post- intervention interview. There was a spread of participants across Methods the three intervention arms and according to age, gender, baseline and 18-month follow-up physical activity levels and adherence to Participants exercise at 3, 6 and 18 months (measured in trial questionnaires). Twenty-two of the participants agreed to be interviewed at follow-up The BEEP trial participants were adults aged 45 years, with knee 12 months later (See Table 1). pain and/or stiffness in one or both knees and who met criteria for a clinical diagnosis of OA.4 None of the participants were previously Themes known to the interviewers. Purposive sampling was undertaken (using trial questionnaire data) to ensure a diverse range of charac- The results are presented in four parts. The first part outlines teristics, including: age, gender, trial intervention arm, baseline pain participant perceptions of what happened during the trial interven- severity and changes in knee pain and function according to WOMAC tion period. The second and third parts focus on perceptions of scores after intervention.16 change over time, and barriers and facilitators to exercise and phys- ical activity in the longer term. The fourth part focuses on Data collection Face-to-face, semi-structured interviews with a subsample of participants from each intervention arm were undertaken at two time points: post-intervention interviews shortly following the BEEP
Research 47 Table 1 Characteristics of participants. Group Gender Age WOMACa PASEb (mean) I have been doing my exercises as often as advisedc identifier range Pain Function Baseline 18 months 3 months 6 months 18 months (y) Usual care 42 M 55 to 64 Worse Worse 149 170 Disagree Disagree Strongly disagree Worse Worse 32 364 F 75 Worse 205 64 Not sure Strongly disagree Strongly disagree NR Improved 69 1135 F 55 to 64 Improved Improved 212 188 Strongly agree Agree Strongly agree Improved Improved NR 6153 M 75 Improved Improved 160 NR Agree Agree Missing No change Improved 200 6937 M 55 to 64 Improved Worse 187 518 Disagree Disagree Disagree Worse 31 7058 M 65 to 74 Worse 117 Disagree Disagree Disagree Worse No change 207 7082 F 65 to 74 Improved 157 152 Agree Not sure Not sure No change Improved 266 7283 F 65 to 74 Improved Improved 97 188 Agree Agree Disagree Improved Improved NR 8460 F 75 Improved 292 157 Strongly agree Disagree Disagree Improved Worse 295 11 702 M 45 to 54 Worse 299 NR Agree Disagree Strongly agree Worse Improved Individually tailored exercise Worse 25 Improved Improved 175 11 M 55 to 64 Improved 136 245 Strongly agree Strongly agree Strongly agree Improved 196 26 F 55 to 64 Improved Worse 147 193 Agree Agree Disagree No change Improved 312 36 M 65 to 74 Improved 139 NR Strongly agree Not sure Disagree Worse Improved 226 61 F 75 Improved Improved 124 178 Strongly agree Agree Not sure Improved 143 1171 M 75 Improved Worse 116 243 Not sure Agree Strongly agree Worse 30 1481 M 45 to 54 Worse Improved 219 Agree Agree Strongly disagree Improved Worse 6517 F 45 to 54 Improved Improved 261 Strongly agree Agree Strongly disagree 6878 F 55 to 64 NR 189 Strongly agree Strongly agree Agree Improved Targeted exercise adherence 3 F 55 to 64 90 Strongly agree Strongly agree Agree 30 M 65 to 74 129 Agree Agree Agree 181 F 75 NR Agree Missing Missing 1736 M 55 to 64 110 Disagree Disagree Disagree 3657 M 65 to 74 196 Agree Disagree Disagree 4677 M 65 to 74 167 Agree Agree Agree 4876 M 75 55 Agree Agree Missing 5210 F 75 65 Strongly agree Agree Missing 6436 F 55 to 64 NR Agree Strongly agree Disagree 6544 F 45 to 54 209 Agree Disagree Not sure 7880 M 45 to 54 168 Strongly disagree Agree Disagree 8141 F 65 to 74 43 Strongly agree Agree Strongly agree F = female, M = male, NR = not recorded, WOMAC = Western Ontario and McMaster Universities Arthritis Index. See main text for descriptions of the interventions given to each of the three trial groups. a The Western Ontario and McMaster Universities Arthritis Index (WOMAC) consists of 24 items divided into three sub-scales: pain, physical function and stiffness. The pain subscale ranges from 0 (no pain) to 20 (maximum pain) and the function subscale ranges from 0 (no disability) to 68 (maximum disability). The psychometric properties of the WOMAC16 have been extensively studied in knee pain populations in clinical trials of different interventions including exercise.48 WOMAC Pain and Physical Function Scores were taken from self-completed questionnaires returned at 3 months follow-up. Improved: any reduction in pain or disability subscale scores from baseline to 3-month questionnaire. Worse: any increase in pain or disability subscale score from baseline to 3-month questionnaire. b The Physical Activity Scale for the Elderly (PASE) assesses physical activity levels over a 1-week period combining physical activity from several domains including household, occupational and leisure.49 Scores range from 0 to 793, with higher scores indicating greater physical activity. c Single item adherence question. participants’ suggested improvements to physiotherapy in- introducing new exercises and maintaining improvements in exer- terventions. Illustrative data excerpts are available in Table 2 on the cise. Higher levels of progression were described by participants in eAddenda. the ITE and TEA arms, such as moving beyond specific lower limb exercises to more general physical activity, which was often linked to Participant perception of the trial interventions their own individual exercise goals, and valued activities (eg, cycling, attending a gym). Physiotherapy-led exercises were described by participants across all arms of the trial, and the core exercise characteristics of supervi- Individualisation sion, progression and individualisation were experienced at different Patient interviews revealed evidence of a basic level of exercise levels (basic and higher) and varied across each arm. individualisation in all three intervention arms. Exercises were Supervision adapted based on the individual’s ability to perform them, reducing A basic level of supervision was evident in the participants’ nar- or increasing exercise difficulty, or providing alternatives if a partic- ipant found them too difficult (eg, due to pain). Individualisation in ratives. They described how the treating physiotherapist demon- the UC arm did not extend beyond this basic level. However, a higher strated and explained the purpose of the exercises before observing level of individualisation was evident in both the ITE and TEA arms, and assessing them as they performed the exercises, and monitored with participants reporting that the physiotherapists adjusted their for any symptom response to the exercise program. Some physio- exercise programs to suit their individual needs (eg, co-morbidities) therapists performed the exercises with the participants, creating a and personal exercise goals. Providing explanations for adjustments sense of partnership. Some participants noted particularly high levels and exercise replacements, monitoring participants’ progress and of attentiveness from the physiotherapists, suggesting a higher level moving beyond the specific lower limb exercises of the BEEP trial of supervision. Participants across all arms of the trial reported being characterised a higher level of individualisation. motivated when performing the exercises under the supervision of the physiotherapists; however, their motivation faded over time Overall, in this sample, it was found that the core intervention without supervision. constructs (supervision, progression and individualisation) were delivered as expected. Therefore, the findings of no substantial Progression between-group differences in the trial may not have been due to poor A basic level of exercise progression was evident in all arms and delivery of these key intervention components. The trial paper re- ports on other aspects of fidelity (eg, delivery of cognitive behavioural included increasing the number of repetitions of exercises,
48 Moore et al: Physiotherapy-led intervention for knee pain Table 3 Barriers to physiotherapy-led exercises and general physical activity. Exercise Usual care Individually targeted exercise Targeted exercise adherence Post-intervention Follow-up Post-intervention Follow-up Post-intervention Follow-up Exercise Patient: ongoing Patient: pain Patient: lack of Patient: pain Patient: pain, Patient: no prescribed pain, attitudes to free, negative enjoyment, reduced worsening, reduced injury, lack of more pain so in the trial exercise and pain, attitudes towards self-efficacy exercise self-efficacy, motivation, no not motivated physical status exercise and pain fear of falling, change in Replaced with concerned about knowledge, No change in Replaced with other activities doing the wrong comorbidity, injury, knowledge other activities thing, lack of negative beliefs Time and place enjoyment about exercise Severity of limitations for pain exercise Lack of monitoring Burden of Replaced with Time and place Time and place exercise limitations: limitations other activity no time, poor Conflicting advice access to facilities Lack of supervision from HCPs Physiotherapist: Time and place poor explanation Physiotherapist: and delivery poor communication, limitations of intervention negative attitude Weather General Patient: lack of Patient: ongoing Patient: lack of Patient: pain physical motivation or pain motivation free, worsening activity enjoyment and self-efficacy, pain, lack of comorbidities, motivation and Time limitations already active so exercise self-efficacy Replaced no perceived need Time and place exercises with Weather limitations other activities Weather Lack of support Weather Cost HCPs = healthcare providers. See main text for descriptions of the interventions given to each of the three trial groups. and educational strategies) and so the findings are complementary in return of pain or future surgery), observing/feeling benefits, and helping to interpret the trial results. knowledge (ie, change in knowledge or retained knowledge) about the role of exercise for pain. Participants who felt that they were Longer-term barriers to exercise and general physical activity ‘naturally active’ and exercised regularly prior to the trial generally maintained their physical activity behaviour into the longer term, A wide range of barriers to exercise and general physical activity regardless of intervention arm. A strong therapeutic alliance between were identified (Table 3), with some being common across the physiotherapist and participant was also a powerful facilitator of intervention arms of the trial. Participant factors (existence of pain or exercise and physical activity. other physical symptoms, low exercise self-efficacy, lack of motiva- tion, or incentive to exercise during pain-free periods), time and Change in knowledge place, weather, lack of supervision and monitoring remained as bar- Knowledge about exercise specific to individual needs facilitated riers to exercise or uptake of general physical activity in the longer term. The excerpts in Table 2 on the eAddenda are illustrative of these exercise and physical activity in the longer term and was linked to themes. Barriers relating to a lack of motivation, unsuitable physical participants’ and physiotherapists’ characteristics. In general, partic- environment and a lack of support or incentives during pain-free ipants reported learning about the benefits of exercising for joint pain periods remained at follow-up. Participants also had ongoing from the physiotherapist and information booklets, and experien- concerns about exercising when in pain and a lack of exercise self- tially during supervision of their exercise program in the trial inter- efficacy. At follow-up, continuing knee pain, instability, fear of fall- vention sessions. Physiotherapists’ explanations of the value and ing or ‘doing the wrong thing’ also deterred some participants from rationale of exercise for knee pain, how to perform the exercises and continuing with exercise in the longer term. Whilst articulated as what should feel right and wrong changed participants’ perceptions barriers to exercise in post-intervention interviews, the burden of of exercise. A participant in the ITE intervention arm retained exercises and physiotherapist-related factors (eg, perceptions that the knowledge about the role of exercise for pain and how this was physiotherapist was disinterested) did not appear to be barriers in the supported by reassurance from the physiotherapist, increasing the longer term. New barriers at 12 months were lack of support to ex- participant’s exercise self-efficacy and ability to exercise despite pain. ercise and receipt of conflicting advice from healthcare practitioners. A model explaining longer-term barriers to exercise and physical Therapeutic alliance activity is presented in Figure 1 on the eAddenda. It illustrates how Therapeutic alliance — defined as a sense of collaboration, warmth the barriers are linked and how there are often multiple influences on adherence to an exercise and physical activity intervention. and support between physiotherapists and participants — emerged as a core facilitator of exercise in the short and longer term regardless of Longer-term facilitators to exercise and general physical activity intervention arm.31–34 A therapeutic alliance facilitated agreement on the exercise intervention and generated an affective bond. Table 5 A wide range of facilitators to exercise and general physical ac- outlines features of a therapeutic alliance between the participants tivity were identified (Table 4), with some being common across the and physiotherapists in all three intervention arms alongside infor- intervention arms of the trial. A model explaining longer-term facil- mation about adherence to exercise and physical activity. Participants itators to exercise and physical activity is presented in Figure 2 on the spoke positively about physiotherapists who took time to get to know eAddenda. The most prominent facilitators remaining at 12 months them and valued physiotherapists’ ability to understand and empa- were participant-related factors, specifically an individual’s ‘naturally thise. Participants appreciated physiotherapists taking time to explain active’ identity, motivation regarding prevention (ie, avoiding the exercises and show how and why they worked, and positively encouraging and empowering them to contribute to their own treat- ment and to ask questions in return. Continuity over treatment sessions
Research 49 Table 4 Facilitators to physiotherapy-led exercises and general physical activity. Exercise Usual care Individually targeted exercise Targeted exercise adherence Post-intervention Follow-up Post-intervention Follow-up Post-intervention Follow-up Exercise Patient: enjoyment, Patient: pain free, Patient: feeling benefit, Time and place: Patient: self-efficacy, Patient: naturally prescribed increased self-efficacy, attitudes towards naturally active, can fit into daily naturally active active, feeling the in the trial feeling benefit exercise and pain, increased self-efficacy, life and work benefit feeling the benefit enjoyment Change in knowledge General Time and place: Time and place: Change in knowledge physical can fit in to daily Prevention: avoiding Time and place: can fit Supported by family activity life or work future surgery into daily life and work fit into daily life and work Physiotherapist: Change in knowledge Change in knowledge Physiotherapist: positive, instilled Replaced with other Physiotherapist: instilled confidence, confidence, reassuring, reassuring activities reassuring, respectful, respectful, listened open, trusting Supervision Supervision Prevention: avoiding Prevention: avoiding surgery surgery Change in knowledge Patient: self-efficacy, Patient: naturally Patient: feeling the Patient: naturally active naturally active active benefit, naturally Support from others active, enjoyment Change in knowledge Change in knowledge Change in Physiotherapist: knowledge Change in knowledge Time and place: fits inspiring, realistic Time and place: into daily life and work fit into daily life Physiotherapist: and work encouragement Prevention: avoiding surgery See main text for descriptions of the interventions given to each of the three trial groups. with the same physiotherapist helped to build a sense of collaboration, a longer period and different modes of delivery (eg, email or tele- and participants spoke about developing rapport, openness and phone sessions and reminders, physiotherapist home visits and trusting relationships where they felt valued, respected and supported. community-based activities). Figure 2 on the eAddenda presents a model of facilitating factors Discussion to explain the interplay between participant and physiotherapy characteristics, therapeutic alliance, knowledge and longer-term ex- This qualitative study embedded within a randomised trial aimed ercise or physical activity. Participants from all intervention arms to investigate participants’ experiences and perceptions of treatment, reported anxiety about exercising with pain but physiotherapists’ particularly supervision, individualisation, progression and barriers reassurance during supervision (linked to therapeutic alliance above) and facilitators to exercise and general physical activity behaviour in helped them to understand when pain was an indication to stop and the longer term. This helped to shed light on the quantitative clinical when it was safe to work through pain. Participants described how results of the trial and provided a novel understanding of changes in changes in their knowledge and confidence to exercise (fostered barriers and facilitators to exercise and physical activity over time. during interactions in treatment sessions) impacted on exercise and This understanding can inform how to optimise future exercise in- physical activity over 12 months later. terventions for OA and to provide maximum benefit for patients over the longer term. Time and place Time and place remained as an exercise facilitator 12 months after The qualitative findings showed that participants experienced different levels of supervision, individualisation and progression the trial intervention period, as the nature of the exercises meant that within each arm. This provides context for the other qualitative participants could fit them into their daily lives at home and work (eg, findings and for the interpretation of the quantitative clinical results sitting at work, waiting for the kettle to boil). because the physiotherapists generally appeared to deliver core intervention components well. The lack of significant differences in Support and supervision clinical outcomes between the intervention arms may therefore be Regular contact and support from others (family and friends) also attributed to other factors. For example, regardless of intervention arm, the barriers and facilitators to exercise and physical activity emerged as a factor that facilitated exercise and physical activity in were common and similar, both at the end of the trial intervention the longer term. For those who had moved on to more general period and at longer-term follow-up. This has added to previous physical activity, staff at local gyms were an important source of research on barriers and facilitators to exercise by identifying the motivation, especially when they took an interest in their knee longer-term barriers as predominantly patient factors, but which also problem and progress. Participants also reported wanting supervision include time and place, the weather and ongoing supervision and from staff at local gyms, like they had experienced with the physio- monitoring.7,9,12,35,36 All three interventions appeared to facilitate therapists in the trial. those who saw themselves as ‘naturally active’ to maintain levels of physical activity despite their knee pain. Suggested improvements to physiotherapy interventions The importance of a therapeutic alliance during treatment in Participants from all intervention arms talked of a need for more facilitating both short-term and longer-term adherence to exercise regular physiotherapist reviews (eg, 6, 12 or 24 months after the end and general physical activity behaviour has also been highlighted of the initial intervention) to enable changes in exercises to be made within this study, suggesting that it may be an important target in appropriately and remind participants of the importance of future exercise interventions for knee OA. Previous studies have continuing with exercise and physical activity. Participants in the UC recognised the importance of therapeutic alliance (or patient- and ITE arms felt that regular reviews with physiotherapists rather provider relationship) as a central component of the therapeutic than primary care physicians (general practitioners) or hospital con- process and a determinant of treatment outcome in psychotherapy, sultants (specialists) were more appropriate, as they provided the but the impact on musculoskeletal outcomes has not been studied ‘right kind of advice’ and support and were ‘easier to talk to’. Other suggestions included more treatment sessions, treatment spread over
Table 5 Characteristics of a therapeutic alliance across the three trial intervention arms and participants’ perceptions of adherence. Therapeutic Example Illustrative quotation alliance characteristic Mutual Equity in the The thing is you do the exercise ‘cause you feel that you Investment work/reciprocity person down. You know you do them ‘cause in the first in going to do me good, it’s going to yeah’, but also there’s think, ‘Oh he’s gone out of his way to explain these things t do it’s only fair that I do them so at least I can tell him wha next time I meet him’, you know. (UC 7058 FU) Appreciative of other So I, I think, and I think it’s because of [physiotherapist] in though I wanted to let him down ‘cause he’d been so good I also think that because they see you trying, it motivates Personal Joint motivations Watched me doing all the exercise, then she’d write on t interactions/ Negotiation/agreement wanted me to do, you know. But I said to her, I said, ‘Well affective bond Feeling at ease/ day.’ She said, ‘Well as long as you do thoroughly, but do t relaxed/valued you know.’ (UC 6153 PI) Communication Perception of therapist (eg, good, nice) I mean, he always had time to talk to you, and say, you Getting to know each anything?’ He didn’t rush you in and rush you out or, lik other/making connections sometimes but, with people, but, no, he was very good. ( Attentiveness to other She was a nice young lass, you know, bedside manner, th Openness, honesty and sort of suspected what it was, do you know what I mean trust We were talking about gardening whatever as I was doin Listening she sort of mentioned things about her life and what she co pieces and it just made it a much more enjoyable experie made a difference. It made me feel I wanted to do the ex She’d had a car crash. She herself, her leg’s badly damage into sort of what it was all about, you know. (TEA 30 FU I think she was far more realistic and she seemed more i and what I did. I know the other girl talked about it but course, not out of . made you feel particularly valuable I do think it was probably seeing that (trial) physio that r think this is a shared thing, he is trying to help me. I wan on such a level that I could share things that I’d maybe f share. (UC 1135 FU) Of course she said, ‘Three times a day’ and I’d look at he that?’ I said ‘Once.’ [..] Oh yes, I was honest with her, y thorough.’ [yes], you know. (UC 6153 PI) And then you start asking more... you develop a bit of a r the study I’m sure the physio gets a little bit more out of it on, on what you’re saying. (ITE 1481 PI) So I think, I, I was quite impressed with the physio in that what I was saying with regard to both the pain in the kn problems and the hernia. (TEA 7880 PI)
don’t want to let the other Perceptions of adherencea to exercise and 50 Moore et al: Physiotherapy-led intervention for knee pain nstance you think, ‘Oh that’s physical activity levels at end of treatment a secondary thing there you to me and shown me what to and 12-month follow-up at sort of effect its having the At end of treatment was partially adherent to exercises and was an active hill walker. At follow-up no longer did the exercises from the trial but was an active n a way, I didn’t, don’t feel as hill walker and had joined a gym. (UC 7058) d, got me so far. (TEA 3657 PI) s them as well. (ITE 1481 FU) At end of treatment was adherent to exercise and joined a gym. At follow-up was partially adherent, joined different gym and cycled. (TEA 3657) the list how many times she At end of treatment was adherent to exercises, used exercise bike. At follow-up l I’m doing them once, once a was not adherent to exercises from the trial but did other sitting exercises, cycled the ten times of each exercise, and walked. (ITE 1481) At end of treatment was adherent to exercises, withdrawn from study at follow-up. know, ‘Any questions or (UC 6153) ke, you know, it does happen (ITE 61 FU) At end of treatment adherent to exercises from the trial and active through he fact that she straightaway dancing. At follow-up continued to dance but not doing exercises from the trial as n? (TEA 30 FU) knee worsened and awaiting knee replacement. (ITE 61) At end of treatment was partially adherent to exercises from the trial, cycled, ng my exercises as well. And walked dog and did morning stretches. At follow-up continued to cycle, and walk ould do with various bits and and did exercises from the trial 2 or 3 times a week. (TEA 30) ence I think. [.] I think it At end of treatment partially adherent to exercises from the trial. At follow-up xercises more. (ITE 26 FU) partially adherent as tried to do some exercises but others were too painful (due to ed. So she had got an insight Baker’s cyst). (ITE 26) U) At end of treatment partially adherent to exercises from the trial, cycled, walked interested in me as a person dog and did morning stretches. At follow-up continued to cycle, and walk and did it was more as a matter of exercises from the trial 2 or 3 times a week. (TEA 30) At end of treatment partially adherent to exercises from the trial. At follow-up e I suppose. (ITE 26 PI) partially adherent as tried to do some exercises but others were too painful (due to really made me open up and Baker’s cyst). (ITE 26) Adherent both at end of the trial intervention and at follow-up. Continued with the nted to be helped. But he was exercises because felt left knee was deteriorating. (UC 1135) found in the past difficult to At end of treatment was adherent to exercises, withdrawn from study at follow-up. er. ‘How many times you do (UC 6153) yes. I said, ‘Once but very At end of treatment was adherent to exercises, used exercise bike. At follow-up rapport and I think that from was not adherent to exercises from the trial but did other sitting exercises, cycled because you start expanding and walked. (ITE 1481) At end of treatment partially adherent and did mountain biking and walking. At t she listened and understood follow-up partially adherent, did exercises from the trial but irregularly. (TEA nee and my mental health 7880)
Research 51 At end of treatment partially adherent and did mountain biking and walking. At At end of treatment was adherent to exercises, used exercise bike. At follow-up Adherent both at end of the trial intervention and at follow-up. Continued with the At end of treatment was adherent to exercises from the trial, declined interview at ITE = individually tailored exercise group (six to eight sessions of individualised, supervised and progressed lower limb exercises over 12 weeks), FU = follow-up, PI = post-intervention, TEA = targeted exercise adherence group (transitioned from lower much.37–39 Ferreira et al reported that positive therapeutic follow-up partially adherent, did exercises from the trial but irregularly. (7880 was not adherent to exercises from the trial but did other sitting exercises, cycled exercises because felt left knee was deteriorating. (UC 1135) follow-up. (ITE 6878) limb exercise to general physical activity, with eight to ten contacts over 6 months), UC = usual care group (up to four treatment sessions of advice and exercise over 12 weeks). alliance ratings between physiotherapists and patients were associ- TEA) and walked. (ITE 1481) ated with improved outcomes in low back pain.40 A need for further a Adherence was categorised as: adherent (participants described still doing their allocated exercises at both time points); partially adherent (participants described doing some of their allocated exercises or general physical activity, or had started exploration of the role of therapeutic alliance in relation to She explained even though the exercises might cause pain, as said, she sort of By going seeing someone every week for a period of time, I think you, you develop Reassure me. Reassure me. Give me the right exercises to do, if it wasn’t going to do any I’ll be a lot more confident, I think I’ll be a bit, say, worried, but at the moment, I’ve got replacement or new types of physical activity); and non-adherent (participants described not doing their allocated exercises at either time point). rehabilitation outcomes has been recognised.39,41 A systematic re- suggested that the, the problem amongst other things was the lack of strength in the some trust, some openness comes from the... from my part, comes from that as well, further damage, if it was arthritis. I think the cartilage problem is caused by, or could sort of motivation that I’m being seen sort of every month. Um, just reassurance really, view by Hall et al reported significant positive associations between muscle. So she said by building the muscles up that will support the knee better [yeah] some understanding. (ITE 1481 PI) have been caused by the arthritis. I needed reassurance that it was okay to actually do I suppose isn’t it. Um, I’m doing my exercises and I’m going and um, the physio’s going therapeutic alliance and global perceived effect of treatment, changes in the long run. (TEA 7880 PI) the exercises and I wasn’t going to cause further damage. Confidence I think because through the exercises with me and telling me this is – well I know it’s improved, but it’s in pain, physical function, patient satisfaction with treatment, sometimes trying to do these things on your own is a bit scary if you get stuck, because just a bit of reassurance and – and motivation to keep going a bit longer [mmm] with depression and general health status.34 The current study adds to this my leg does lock. And I think probably him encouraging me to do the right exercises, them. (ITE 6878 PI) previous research in two ways. First, it suggests that therapeutic and do them every day, which I did do. (UC 1135 PI) alliance during the treatment phase may have a longer-term influ- ence on patients’ exercise and general physical activity behaviour. Explanations offered and Reassurance/confidence This supports recent evidence suggesting that therapeutic alliance understood (exercise (exercise self-efficacy) may be the best predictor of adherence to exercise.42 Second, it self-efficacy) identifies key features of therapeutic alliance that appear to be fa- cilitators to exercise adherence, which include: mutual investment, the quality of personal interactions and communication, and an af- fective bond. Various conceptual frameworks have been used to explain ther- apeutic alliance in relation to physiotherapy, but the heterogeneity of these models limits their application in the musculoskeletal pain context.39 Identifying an optimal framework for use in musculo- skeletal pain rehabilitation settings may be informed by the current analysis. Quantitative measures of therapeutic alliance are available. Babatunde et al identified 26 measures; among these, the Working Alliance Inventory is the most used.39 However, Hall et al found that measures developed from psychotherapy exhibit a ceiling effect and require re-contextualisation for a musculoskeletal pain context.43 Despite specifically targeting exercise adherence in the longer term in the TEA intervention in the BEEP trial, similar barriers to exercise were identified in participants in each of the three intervention arms and remained over time. When asked about potential improvements to the interventions used in the BEEP trial, participants suggested regular reviews (with a physiotherapist or primary care physician) over a longer period of time and different modes of delivery. The best method for providing ongoing support for exercise to this patient population remains un- clear. The addition of two 30-minute booster sessions to a 12-week supervised physiotherapy exercise program did not increase adher- ence or improve pain or function in a trial by Bennell et al.44 Although regular reviews are a component of a chronic disease management model, there are barriers to the use of OA chronic disease service management models.45 It is possible that harnessing the potential of eHealth to provide ongoing reminders and monitoring may help to maintain adherence over time.46,47 A strength of this work is that the analysis was not constrained to the core constructs of the exercise interventions and did not attempt to either confirm or reject the clinical trial results, but to explain them.26 The qualitative data analysis was conducted while the au- thors were unaware of the quantitative clinical trial results, in order to maintain an interpretive approach. By sampling for a diverse range of characteristics and applying a within-case and cross-case longi- tudinal approach, it went beyond a cross-sectional analysis of the qualitative data set, ensuring that data saturation was achieved in accordance with the aims of the study. The use of descriptive and analytical questions about change in behaviour also provided insights into longitudinal data analysis, which is seldom reported.23 A limi- tation is that it did not also interview physiotherapists to explore their experiences of delivering the trial interventions or to gain un- derstanding of their perceptions of therapeutic alliance. Future research is required to establish if it is possible to harness the po- tential of a greater therapeutic alliance between patients and phys- iotherapists, and whether regular reviews and ongoing support of patients lead to sustained changes in exercise and physical activity behaviour in the longer term. There were also eight participants who declined to a follow-up interview. The purposive sampling frame- work ensured a good range of patient characteristics and a state of thematic saturation was reached within the analysis; the authors are confident in the themes that were identified. However, they cannot be sure that the addition of data from the eight participants who
52 Moore et al: Physiotherapy-led intervention for knee pain declined a follow-up interview would have changed the results, and 2. Murray CJ, Vos T, Lozano R, Naghavi M, Flaxman AD, Michaud C, et al. Disability- accept this as a limitation of longitudinal qualitative studies. adjusted life years (DALYs) for 291 diseases and injuries in 21 regions, 1990-2010: a systematic analysis for the Global Burden of Disease Study 2010. Lancet. The interview participants experienced different levels of super- 2012;380:2197–2223. vision, individualisation and progression, yet similar barriers and facilitators to exercise and physical activity were found irrespective of 3. Jagger C, Matthews R, Spiers N, Brayne C, Comas-Herrera A, Robinson T, et al. which of the three trial intervention arms participants were allocated Compression or expansion of disability?: forecasting future disability levels under to. Despite a focus on supervision, individualisation and progression changing patterns of diseases. London: King’s Fund; 2006. to improve exercise adherence, barriers remained 12 months after contact with the physiotherapists ended, including lack of support 4. National Institute for Health and Care Excellence (2014). Osteoarthritis: Care and and regular reviews, continuing knee pain, fear of doing the wrong Management in adults. Clinical guidelines CG177. https://www.ncbi.nlm.nih.gov/ thing, and the weather. However, the presence and quality of a pubmedhealth/PMH0068962. Accessed 31 July, 2018. therapeutic alliance during treatment appeared to facilitate adher- ence to exercise and general physical activity. These findings help to 5. Fransen M, McConnell S, Harmer AR, Van der Esch M, Simic M, Bennell KL. Exercise explain the quantitative trial results.17 They also highlight that for osteoarthritis of the knee: a Cochrane systematic review. Br J Sports Med. models of ongoing support and enhanced therapeutic alliance should 2015;49:1554–1557. be targets for future trials of exercise in older adults with knee pain. 6. Hendry M, Williams NH, Markland D, Wilkinson C, Maddison P. Why should we What was already known on this topic: Osteoarthritis is exercise when our knees hurt? A qualitative study of primary care patients with common and can impair function. Exercise, including local osteoarthritis of the knee. BMC Fam Pract. 2006;23:558–567. muscle strengthening and general physical activity, is recom- mended in clinical guidelines. The benefits obtained after 7. Marks R. Knee osteoarthritis and exercise adherence: a review. Curr Aging Sci. commencing exercise typically decline over time, which is 2012;5:72–83. probably due to progressive reductions in adherence to the exercise. 8. Jack K, McLean SM, Moffet JK, Gardiner E. Barriers to treatment adherence in What this study adds: Patients with knee osteoarthritis re- physiotherapy outpatient clinics: A systematic review. Man Ther. 2010;15:220– ported that lack of motivation, time, physical environment and 228. lack of monitoring were barriers to exercise and general physical activity. These barriers seemed similar, regardless of the levels of 9. Petursdottir U, Arnadottir SA, Halldorsdottir S. Facilitators and barriers to exer- supervision, individualisation and progression that were provided cising among people with osteoarthritis: a phenomenological study. Phys Ther. when exercise was commenced. The presence and quality of a 2010;90:1014–1025. therapeutic alliance with a physiotherapist facilitated adherence to exercise and general physical activity. 10. Holden MA, Nicholls EE, Young J, Hay EM, Foster NE. Role of exercise for knee pain: what do older adults in the community think? Arthritis Care Res. 2012;64:1554– eAddenda: Table 2, Figures 1 and 2, and Appendices 1 and 2 can be 1564. found online at https://doi.org/10.1016/j.jphys.2019.11.004. 11. Hurley MV, Walsh N, Bhavnani V, Britten N, Stevenson F. Health beliefs before and Ethics approval: The North West 1 Research Ethics Committee, after participation on an exercised based rehabilitation programme for chronic Cheshire, UK (REC reference: 10/H1017/45) approved this study. All knee pain: doing is believing. BMC Musculoskelet Disord. 2010;11:31. participants gave written informed consent before data collection began. 12. Kanavaki AM, Rushton A, Efstathiou N, Alrushud A, Klocke R, Abhishek A, et al. Barriers and facilitators of physical activity in knee and hip osteoarthritis: a sys- Competing interests: Nil. tematic review of qualitative evidence. BMJ Open. 2017;7:e017042. Source(s) of support: This independent research was funded by the National Institute for Health Research (RP-PG-0407-10386) and 13. Bennell K, Dobson F, Hinman RS. Exercise in osteoarthritis: Moving from pre- the Arthritis Research UK Centre in Primary Care grant (18139). CJ is scription to adherence. Best Pract Res Clin Rheumatol. 2014;28:93–117. part-funded by the National Institute for Health Research (NIHR) Applied Research Centre (ARC) West Midlands. NF is a NIHR Senior 14. Foster NE, Healey EL, Holden MA, Nicholls E, Whitehurst DG, Jowett S, et al. Investigator and was supported through an NIHR Research Profes- A multicentre, pragmatic, parallel group, randomised controlled trial to compare sorship (NIHR-RP-011-015). The views expressed are those of the the clinical and cost-effectiveness of three physiotherapy-led exercise in- authors and not necessarily those of the NHS, the NIHR or the terventions for knee osteoarthritis in older adults: the BEEP trial protocol (ISRCTN: Department of Health and Social Care. 93634563). BMC Musculoskelet Disord. 2014;15:254. Acknowledgements: We thank all the patients who participated in the interview study. We thank Dr Diane Roberts (DR) for her 15. Holden MA, Case R, Healey EL, Hill S, Mullis R, Roddy E, et al. Content and eval- contribution to the BEEP pilot study interviews and for conducting uation of the benefits of effective exercise for older adults with knee pain trial early interviews for the main qualitative study linked to the BEEP physiotherapist training program. Arch Phys Med Rehabil. 2017;98:866–873. trial. We thank the administrative staff and others who supported the qualitative research study within the BEEP trial from Keele Clinical 16. Hay E, Dziedzic K, Foster N, Peat G, van der Windt D, Bartlam B, et al. Optimal Trials Unit, including research program support from Liz Mason. We primary care management of clinical osteoarthritis and joint pain in older people: thank Professor Peter Croft who led the NIHR program grant within a mixed-methods programme of systematic reviews, observational and qualitative which the BEEP trial was funded, and all the members of the TSC and studies, and randomised controlled trials. Programme Grants Appl Res. 2018;6:4. DMC, including the PPIE members. Provenance: Not invited. Peer reviewed. 17. Bellamy N, Buchanan WW, Goldsmith CH, Campbell J, Stitt LW. Validation study of Correspondence: Dr Andrew J Moore, Musculoskeletal Research WOMAC: a health status instrument for measuring clinically important patient Unit, Bristol Medical School, University of Bristol, Bristol, United relevant outcomes to antirheumatic drug therapy in patients with osteoarthritis of Kingdom. Email: [email protected] the hip or knee. J Rheumatol. 1988;15:1833–1840. References 18. Oakley A, Strange V, Bonell C, Allen E, Stephenson J. Process evaluation in rando- mised controlled trials of complex interventions. BMJ. 2006;332:413–416. 1. March L, Smith EU, Hoy DG, Cross MJ, Sanchez-Riera L, Blyth F, et al. Burden of disability due to musculoskeletal (MSK) disorders. Best Pract Res Clin Rheumatol. 19. Moore GF, Audrey S, Barker M, Bond L, Bonell C, Hardeman W, et al. Process 2014;28:353–366. evaluation of complex interventions: Medical Research Council guidance. BMJ. 2015;350:h1258. 20. Lewin S, Glenton C, Oxman A. Use of qualitative methods alongside randomised controlled trials of complex healthcare interventions: methodological study. BMJ. 2009;339:b3496. 21. O’Cathain A, Thomas KJ, Drabble SJ, Rudolph A, Hewison J. What can qualitative research do for randomised controlled trials? A systematic mapping review. BMJ Open. 2013;3:e002889. 22. Hoddinott P, Britten J, Pill R. Why do interventions work in some places and not others: a breastfeeding support group trial. Soc Sci Med. 2010;70:769–778. 23. Saldana J. Longitudinal qualitative research. Analyzing change through time. Califor- nia: AltaMira Press; 2003. 24. Calman L, Brunton L, Molassiotis A. Developing longitudinal qualitative designs: lessons learned and recommendations for health services research. BMC Med Res Methodol. 2013;13:14. 25. NVivo qualitative data analysis software; QSR International Pty Ltd. Version 10, 2012. 26. Plano Clark VL, Schumacher K, West CM, Edrington J, Dunn LB, Harzstark A, et al. Practices for embedding an interpretive qualitative approach within a randomized clinical trial. J Mix Methods Res. 2013;7:219–242. 27. Charmaz K. Constructing grounded theory: A practical guide through qualitative analysis. London: Sage; 2006. 28. Boyatzis RE. Transforming Qualitative Information: Thematic Analysis and Code Development. London: Sage; 1998. 29. Guest G, Bunce A, Johnson L. How many interviews are enough? Field Meth. 2006;18:59–82. 30. Morris RL, Sanders C, Kennedy AP, Rogers A. Shifting priorities in multimorbidity: a longitudinal qualitative study of patient’s prioritization of multiple conditions. Chronic Illn. 2011;7:147–161. 31. Freud S. The Dynamics of Transference. London: Hogarth Press; 1958. 32. Greenson RR. Technique and Practice of Psychoanalysis. New York: International Universities Press; 1967. 33. Bordin ES. The generalizability of the psychoanalytic concept of the working alli- ance. Psychotherapy: Theory, Research, and Practice. 1979;16:252–260. 34. Hall AM, Ferreira PH, Maher CG, Latimer J, Ferreira ML. The influence of the therapist-patient relationship on treatment outcome in physical rehabilitation: a systematic review. Phys Ther. 2010;90:1099–1110.
Research 53 35. Veenhof C, van Hasselt TJ, Koke AJ, Dekker J, Bijlsma JW, van den Ende CH. Active 42. Wright BJ, Galtieri NJ, Fell M. Non-adherence to prescribed home rehabilitation involvement and long-term goals influence long-term adherence to behavioural exercises for musculoskeletal injuries: the role of the patient-practitioner rela- graded activity in patients with osteoarthritis: a qualitative study. Aust J Physiother. tionship. J Rehabil Med. 2014;46:153–158. 2006;52:273–278. 43. Hall AM, Ferreira ML, Clemson L, Ferreira P, Latimer J, Maher CG. Assessment of the 36. Dobson F, Bennell KL, French SD, Nicolson PJ, Klaasman RN, Holden MA, et al. therapeutic alliance in physical rehabilitation: a RASCH analysis. Disabil Rehabil. Barriers and facilitators to exercise participation in people with hip and/or knee 2012;34:257–266. osteoarthritis: synthesis of the literature using behavior change theory. Am J Phys Med Rehabil. 2016;95:372–389. 44. Bennell KL, Kyriakides M, Hodges PW, Hinman RS. Effects of two physiotherapy booster sessions on outcomes with home exercise in people with knee osteoar- 37. Ardito RB, Rabellino D. Therapeutic alliance and outcome of psychotherapy: historical thritis: a randomized controlled trial. Arthritis Care Res. 2014;66:1680–1687. excursus, measurements, and prospects for research. Front Psychol. 2011;2:270. 45. Brand CA, Ackerman IN, Tropea J. Chronic disease management: improving care for 38. Martin DJ, Garske JP, Davis MK. Relation of the therapeutic alliance with people with osteoarthritis. Best Pract Res Clin Rheumatol. 2014;28:119–142. outcome and other variables: A meta-analytic review. J Consult Clin Psychol. 2000;68:438. 46. Cottrell MA, Hill AJ, O’Leary SP, Raymer ME, Russell TG. Patients are willing to use telehealth for the multidisciplinary management of chronic musculoskeletal con- 39. Babatunde F, MacDermid J, MacIntyre N. Characteristics of therapeutic alliance in ditions: a cross-sectional survey. J Telemed Telecare. 2018;24:445–452. musculoskeletal physiotherapy and occupational therapy practice: a scoping re- view of the literature. BMC Health Serv Res. 2017;17:375. 47. Bossen D, Veenhof C, Van Beek KE, Spreeuwenberg PM, Dekker J, De Bakker DH. Effectiveness of a web-based physical activity intervention in patients with knee and/ 40. Ferreira PH, Ferreira ML, Maher CG, Refshauge KM, Latimer J, Adams RD. The or hip osteoarthritis: randomized controlled trial. J Med Internet Res. 2013;15:e257. therapeutic alliance between clinicians and patients predicts outcome in chronic low back pain. Phys Ther. 2013;93:470–478. 48. O’Reilly SC, Muir KR, Doherty M. Screening for pain in knee osteoarthritis: which question? Ann Rheum Dis. 1996;55:931–933. 41. Kayes NM, McPherson KM. Human technologies in rehabilitation: ‘Who’ and ‘How’ we are with our clients. Disabil Rehabil. 2012;34:1907–1911. 49. Washburn RA, Smith KW, Jette AM, Janney CA. The physical activity scale for the elderly (PASE): development and evaluation. J Clin Epi. 1993;46:153–162.
Search
