["34 Thompson et al: Mobile apps supporting exercise for musculoskeletal pain 24. McGuinness L, Higgins JPT. Risk-of-bias VISualization (robvis): An R package and 37. Blanquero J, Cortes-Vega MD, Rodriguez-Sanchez-Laulhe P, Corrales-Serra BP, Shiny web app for visualizing risk-of-bias assessments. Res Syn Meth. 2020;1\u20137. G\u00f3mez-Patricio E, D\u00edaz-Matas N, et al. Feedback-guided exercises performed on a tablet touchscreen improve return to work, function, strength and healthcare 25. Sch\u00fcnemann HBJ, Guyatt G, Oxman A. GRADE Handbook. 2013. usage more than an exercise program prescribed on paper for people with 26. Page MJ HJ, Sterne JAC. Cochrane Handbook for Systematic Reviews of Intervention. wrist, hand or \ufb01nger injuries: a randomised trial. J Physiother. 2020;66:236\u2013 242. Chapter 13: Assessing risk of bias due to missing results in a synthesis. 2021. 27. Lee J, Lee M, Lim T, Kim T, Kim S, Suh D, et al. Effectiveness of an application-based 38. Goldsmith ES, Taylor BC, Greer N, Murdoch M, MacDonald R, McKenzie L, et al. Focused Evidence Review: Psychometric Properties of Patient-Reported Outcome neck exercise as a pain management tool for of\ufb01ce workers with chronic neck pain Measures for Chronic Musculoskeletal Pain. J Gen Int Med. 2018;33:61\u201370. and functional disability: a pilot randomized trial. Eur J Integrative Med. 2017;12:87\u201392. 39. Schafer AGM, Zalpour C, von Piekartz H, Hall TM, Paelke V. The Ef\ufb01cacy of 28. van Reijen M, Vriend I, Zuidema V, van Mechelen W, Verhagen EA. Increasing Electronic Health-Supported Home Exercise Interventions for Patients With compliance with neuromuscular training to prevent ankle sprain in sport: does the Osteoarthritis of the Knee: Systematic Review. J Med Internet Res. 2018;20: \u2019strengthen your ankle\u2019 mobile App make a difference? A randomised controlled e152. trial [with consumer summary]. Br J Sports Med. 2016;50:1200\u20131205. 29. Van Reijen M, Vriend I, Zuidema V, van Mechelen W, Verhagen EA. The \u201cStrengthen 40. Devji T, Guyatt GH, Lytvyn L, Brignardello-Petersen R, Foroutan F, Sadeghirad B, your ankle\u201d program to prevent recurrent injuries: A randomized controlled trial et al. Application of minimal important differences in degenerative knee disease aimed at long-term effectiveness. J Sci Med Sport. 2017;20:549\u2013554. outcomes: A systematic review and case study to inform BMJ Rapid Recommen- 30. Thiengwittayaporn S, Wattanapreechanon P, Sakon P, Peethong A, Ratisoontorn N, dations. BMJ Open. 2017;7. Charoenphandhu N, et al. Development of a mobile application to improve exercise accuracy and quality of life in knee osteoarthritis patients: a randomized 41. Karayannis NV, Sturgeon JA, Chih-Kao M, Cooley C, Mackey SC. Pain interference controlled trial. Arch Orthop Trauma Surg. 2021. and physical function demonstrate poor longitudinal association in people living 31. Alasfour M, Almarwani M. The effect of innovative smartphone application on with pain: A PROMIS investigation. Pain. 2017;158:1063\u20131068. adherence to a home-based exercise programs for female older adults with knee osteoarthritis in Saudi Arabia: a randomized controlled trial. Disabil Rehabil. 42. Booth J, Moseley GL, Schiltenwolf M, Cashin A, Davies M, H\u00fcbscher M. Exercise for 2020:1\u20138. chronic musculoskeletal pain: A biopsychosocial approach. Musculoskelet Care. 32. Abadiyan F, Hadadnezhad M, Khosrokiani Z, Letafatkar A, Akhshik H. Adding a 2017;15:413\u2013421. smartphone app to global postural re-education to improve neck pain, posture, quality of life, and endurance in people with nonspeci\ufb01c neck pain: a randomized 43. Hunt MA, Birmingham TB, Skarakis-Doyle E, Vandervoort AA. Towards a controlled trial. Trials. 2021;22. biopsychosocial framework of osteoarthritis of the knee. Disabil Rehabil. 30:54\u2013 33. Chitkar SSA, Hakkak HRM, Saadati H, Hosseini SH, Jafari Y, Ganji R. The effect of 61. mobile-app-based instruction on the physical function of female patients with knee osteoarthritis: a parallel randomized controlled trial. BMC Women\u2019s Health. 44. Burgess R, Mansell G, Bishop A, Lewis M, Hill J. Predictors of functional outcome in 2021:21. musculoskeletal healthcare: An umbrella review. Eur J Pain. 2020;24:51\u201370. 34. Chhabra HS, Sharma S, Verma S. Smartphone app in self-management of chronic low back pain: a randomized controlled trial. Eur Spine J. 2018;27:2862\u20132874. 45. Bachmann C, Oesch P, Bachmann S. Recommendations for Improving Adherence to 35. Anan T, Kajiki S, Oka H, et al. Effects of an arti\ufb01cial intelligence-assisted health Home-Based Exercise: A Systematic Review. Phys Med Rehabil Kurortmed. program on workers with neck\/shoulder pain\/stiffness and low back pain: Ran- 2018;28:20\u201331. domized controlled trial. JMIR mHealth uHealth. 2021;9. 36. Choi Y, Nam J, Yang D, Jung W, Lee HR, Kim SH. Effect of smartphone application- 46. Michie S, Atkins L, West R. The Behaviour Change Wheel: A Guide to Designing supported self-rehabilitation for frozen shoulder: a prospective randomized con- Interventions. In: Silverback Publishing; 2014. www.silverbackpublishing.org. trol study. Clin Rehabil. 2019;33:653\u2013660. Accessed June 25, 2021. 47. Minozzi S, Cinquini M, Gianola S, Gonzalez-Lorenzo M, Banzi R. The revised Cochrane risk of bias tool for randomized trials (RoB 2) showed low interrater reliability and challenges in its application. J Clin Epidemiol. 2020;126:37\u201344. 48. Minozzi S, Dwan K, Borrelli F, Filippini G. Reliability of the revised Cochrane risk-of- bias tool for randomised trials (RoB2) improved with the use of implementation instruction. J Clin Epidemiol. 2022;141:99\u2013105.","Journal of Physiotherapy 69 (2023) 47\u201352 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 Normal gait speed varies by age and sex but not by geographical region: a systematic review Addison Williams Andrews a, Srikant Vallabhajosula a, Sarah Boise a, Richard W Bohannon b aDepartment of Physical Therapy Education, Elon University, Elon, USA; bPhysical Therapy Consultants, Fuquay-Varina, USA KEYWORDS ABSTRACT Gait Questions: What are comfortable gait speed values for apparently healthy adults? How do these differ by age Walking speed group, sex and geographical region? Design: Systematic review of observational studies with meta-analysis. Reference value Participants: Apparently healthy, community-dwelling adults who have undergone measurement of Adult comfortable gait speed. Search method: Potentially relevant studies were identi\ufb01ed in four databases. Geography Extracted data from studies that satis\ufb01ed the eligibility criteria were added to a database containing the same information from a meta-analysis published a decade ago. Outcome measures: The weighted mean comfortable gait speed was calculated along with the 95% con\ufb01dence interval for each stratum of age\/sex using a random-effects model. Mean gait speeds were further strati\ufb01ed by the continent where the study took place. Tests of homogeneity included I2 and prediction intervals. Results: Meta-analysis of data from 51,248 apparently healthy adults was strati\ufb01ed by age (in decades) and sex. Male gait speed slowed beyond age 50 years whereas female gait speed slowed beyond age 30 years. The weighted mean gait speed ranged from 97 cm\/s (females aged \u0015 80 years) to 140 cm\/s (males aged 40 to 49 years). The I2 values ranged from 0 to 34.07; prediction interval ranges varied from a low of 30 (125 to 155 cm\/s; males aged 40 to 49 years) to a high of 77 (83 to 160 cm\/s; females aged 60 to 69 years). There was considerable overlap in con\ufb01dence intervals between continents for each sex\/age group. Conclusions: Comfortable gait speed slowed through the adult years, but males maintained a faster walking speed than females. Further strati\ufb01cation of comfortable gait speed by geographical region is not warranted. [Andrews AW, Vallabhajosula S, Boise S, Bohannon RW (2023) Normal gait speed varies by age and sex but not by geographical region: a sys- tematic review. Journal of Physiotherapy 69:47\u201352] \u00a9 2022 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 periodically from a population-based sample.10,11 Consequently, the present analysis, completed in the last year, used meta-analysis to Gait is the primary means by which independently mobile in- update comfortable gait speed reference values for apparently dividuals move within their life space (eg, home or neighbourhood).1 healthy adults, particularly as it might differ between populations in The speed at which they walk, classi\ufb01ed by some as a \u2018vital sign\u2019,2,3 different regions. has been shown to be related to other important variables such as survival2 and the capacity for community ambulation.4 Given these Therefore, the research questions for this systematic review were: facts, a sense of \u2018normal gait speed\u2019 is appropriate for interpreting the performance or health status of individuals. These individuals include 1. What are comfortable gait speed values for apparently healthy both those who are healthy and those with musculoskeletal, neuro- adults? logical or cardiovascular health-related problems. 2. How do these differ by age group, sex and geographical region? Covariates that have been shown to affect normal gait speed include sex and age.5\u20138 Speci\ufb01cally, normal gait speed is greater in Method males than females,5,6 while normal gait speed is greater in younger adults than older adults.7,8 Identi\ufb01cation and selection of studies Several individual studies have attempted to provide summaries This review is reported according to the guidelines found in the of normal gait speed; potentially more representative indicators have Preferred Reporting Items for Systematic Reviews and Meta-Analysis been derived via systematic reviews and meta-analyses.9 The last (PRISMA) statement.12 Furthermore, to ensure integrity of the study, such analysis of which we are aware was published a decade ago.9 the investigators adhered to a Quality Checklist for a Systematic The literature suggests that reference values be updated https:\/\/doi.org\/10.1016\/j.jphys.2022.11.005 1836-9553\/\u00a9 2022 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\/).","48 Andrews et al: Normal gait speed by geographical region Box 1. Inclusion criteria. Articles retrieved by search \u2022 PubMed (n = 408) Design \u2022 CINAHL (n = 345) \u2022 Web of Science (n = 869) \u000f Observational studies \u2022 ResearchGate (n = 260) Participants Duplicates excluded (n = 438) \u000f Community-dwelling \u000f \u001518 years of age Articles evaluated (n = 1,444) Outcome measures \u000f Walking distance between 2 m and 30 m \u000f Instructed to walk at comfortable, normal or natural gait speed \u000f Specified that acceleration and deceleration phases were not included in the timed assessment of gait speed Review or Meta-Analysis.13 This study is the culmination of a 4-year Articles excluded after evaluation of text (n = project entailing three rounds of updated reviews. Given the 1,406)a ongoing nature of the work, the investigators were not granted \u2022 gait speed not collected (n = 269) registration by an outside agency. Electronic identi\ufb01cation of poten- \u2022 non-healthy participants (n = 263) tially relevant studies began with the search of three databases: \u2022 < 20 participants (n = 211) PubMed, the Cumulative Index of Nursing and Allied Health (CINAHL) \u2022 systematic review (n = 137) and Web of Science. Additional articles that cited the original \u2022 not usual gait speed (n = 118) research9 were reviewed on ResearchGate. The search terms using \u2022 no acceleration phase (n = 76) Boolean mode were (Speed OR Velocity) AND (gait OR walking) AND \u2022 children (n = 72) (Norms OR normative OR reference). Each of these databases was \u2022 data not stratified by decade or sex (n = 63) searched from 2008 to 2019. Reviewed research articles were limited \u2022 timed distance < 2 m or > 30 m (n = 63) to those published in English. One researcher independently \u2022 treadmill (n = 63) reviewed the search results to determine potential eligibility for in- \u2022 non-human participants (n = 35) clusion. Only articles that speci\ufb01cally mentioned untimed accelera- \u2022 not in English (n = 27) tion and deceleration phases were included because the nature of the \u2022 not published in a peer-reviewed journal (n = start (ie, standing start versus dynamic start) affected gait speed in previous studies.14,15 Other criteria for inclusion are included in Box 1. 9) Duplicate articles identi\ufb01ed in other reviewed databases were Articles added to previous meta- discarded. Two investigators reviewed the methods and results sec- analysis (n = 38) tions to determine whether each potentially relevant article met the eligibility criteria. Articles were excluded if they: measured gait speed Figure 1. Flow of recent studies through the updated review. in animals; included participants with a gait-affecting pathology (eg, a Initial reason identi\ufb01ed for exclusion. Several articles failed to meet two or more stroke); did not include original data (eg, systematic review, previ- inclusion\/exclusion criteria. ously published data); did not collect or present gait speed values; only analysed gait on a treadmill; or had a gait path that included Data analysis turns or stopping. Authors were contacted regarding articles that did not specify suf\ufb01cient information to determine whether it met the All data from both the previous meta-analysis9 and this review of eligibility criteria or did not present their data in the requisite format articles published within the past decade were entered into a (ie, gait speed means and SDs, strati\ufb01ed by sex and decade of life). spreadsheet speci\ufb01cally designed for meta-analysis.16 The weighted Those articles that provided the requested data were included in the mean comfortable gait speed was calculated along with the 95% CI for analyses. To minimise the risk of investigator bias, two investigators each stratum of age\/sex using a random-effects model. Weighting had to agree on whether each study met all the eligibility criteria. All refers to the number of participants within a study who \ufb01t within the data included in the meta-analysis were veri\ufb01ed for accuracy of input sex\/age stratum in question; mean gait speed values from studies by two investigators working independently. Two of the investigators with more participants were given more weight in the calculations. identi\ufb01ed studies whose mean gait speed for one or more groups of Mean gait speeds were further strati\ufb01ed by the continent on which participants were three or more SDs above or below the overall mean the study took place. for that group; studies deemed outliers were excluded from further analysis. Tests of homogeneity included I2 and prediction intervals.17\u201319 Tests of homogeneity were only conducted if the sex\/age stratum Assessment of characteristics of trials had \ufb01ve or more source articles20 and \u0015 25 participants. Participants: The eligibility criteria and country in which the data Results were collected were extracted for each trial, along with whether the participants were enrolled as a population-based sample, a conve- Search results yielded 1,882 articles published over the past nience sample or some other cohort. We also extracted the sample decade. After removal of duplicates, 1,444 abstracts or full-texts were size, the sex of the participants and the age range by decade (ie, 18 to reviewed for adherence to inclusion and exclusion criteria. Articles 29, 30 to 39, 40 to 49, 50 to 59, 60 to 69, 70 to 79, 80 to 99) of the were excluded from the analysis if they did not meet one required participants. inclusion or exclusion criterion; several articles failed to meet more than one criterion (see Figure 1). Thirty-eight articles met all eligi- Outcome measures: We extracted the timed distance (in metres), bility criteria and were incorporated into the analysis.6,8,21\u201356 Table 1 untimed distance for acceleration and deceleration (in metres), provides a summary of the included articles.6,8,21\u201356 testing conditions (eg, footwear, presence of markers for motion analysis, number of trials), instructions to participants, method of timing gait and mean (SD) gait speed in cm\/s for each sex\/age stra- tum. This information was added to a database containing the same information from a previous meta-analysis published a decade ago.9","Research 49 Table 1 Table 1 Continued Test distance and Participants Speed (cm\/s) Summary of recent studies added to original meta-analysis.9 Study equipment Sampling Study Test distance and Participants Speed (cm\/s) Goutier, 201032 7.5 m Sex (n) Age (y) Mean (SD) Sampling equipment Convenience Accelerometer Adams, 201621 Sex (n) Age (y) Mean (SD) Switzerland Population-based Gunasekaran, 201633 M (10) 20 to 29 130 (16) Netherlands 4.9 m M (475) 50 to 59 131 (16) Convenience F (10) 20 to 29 130 (9) Gait mat M (729) 60 to 69 127 (16) India Alcock, 201522 M (564) 70 to 79 120 (19) 4.0 m M (330) 60 to 65 83 (31) Convenience 6.0 m M (247) 80 to 89 103 (21) Hollman, 201134 Stopwatch M (95) 65 to 70 73 (19) UK Motion analysis M (14) 90 to 99 92 (16) Strati\ufb01ed sample M (69) 70 to 79 72 (17) Bailey, 201623 F (608) 50 to 59 125 (17) USA 5.6 m F (185) 60 to 65 75 (16) Convenience system F (885) 60 to 69 123 (16) Gait mat F (34) 65 to 70 70 (15) USA 4.9 m F (685) 70 to 79 144 (19) Kawai, 20198 F (10) 70 to 79 65 (14) Gait mat F (300) 80 to 89 99 (21) Population-based 5.0 m M (27) 70 to 74 117 (16) Bruening, 201524 90 to 99 77 (24) Japan Plantar pressure M (30) 75 to 79 122 (15) Convenience 15.0 m F (18) 60 to 69 140 (14) M (37) 80 to 84 112 (17) USA Motion analysis F (15) 70 to 79 120 (20) Kemmler, 201535 platform M (14) 85 to 89 101 (22) F (16) 80 to 89 110 (26) Population-based F (33) 70 to 74 116 (20) Canney, 201725 system F (8) 60 to 69 123 (16) Germany 10.0 m F (77) 75 to 79 112 (17) Population-based M (8) 70 to 79 119 (21) Kirkwood, 201636 Photocell sensors F (43) 80 to 84 101 (15) Ireland 4.9 m M (8) 80 to 89 120 (16) Convenience F (33) 85 to 89 98 (20) Gait mat M (5) 50 to 59 97 (12) Brazil M (188) 70 to 74 132 (21) Chen, 201626 F (2) 60 to 69 112 (33) Kirkwood, 201837 M (152) 75 to 79 126 (25) Convenience 4.6 m F (16) 70 to 79 125 (26) Cross-sectional M (107) 80 to 84 123 (22) Taiwan Stopwatch F (19) 80 to 89 104 (20) Brazil M (44) 85 to 96 108 (23) F (5) 18 to 20 129 (13) F (238) 70 to 74 133 (20) Chui, 201027 3.7 m M (6) 20 to 29 133 (12) Lee, 201438 F (275) 75 to 79 130 (20) Convenience Gait mat M (27) 30 to 39 133 (20) Convenience F (160) 80 to 84 121 (21) USA M (6) 40 to 49 136 (9) USA F (48) 85 to 96 106 (22) 10.0 m M (14) 50 to 59 151 (31) F (1,022) 70 to 79 131 (22) Crockett, 201828 Stopwatch M (3) 18 to 19 126 (19) L\u00f6fgren, 201339 F (303) 80 to 95 111 (26) Convenience F (2) 20 to 29 133 (15) Convenience Canada 6.0 m F (22) 30 to 39 137 (20) Sweden 5.6 m F (95) 70 to 74 121 (18) Photocell sensors F (7) 40 to 49 150 (10) Lythgo, 201140 Gait mat F (77) 75 to 79 115 (18) Elbaz, 201829 F (4) 40 to 49 151 (19) Convenience F (30) 80 to 89 105 (23) Population-based 10.0 m M (29) 50 to 59 144 (17) Australia 5.6 m F (42) 18 to 20 137 (16) France Stopwatch M (1,000) 60 to 69 136 (19) Malatesta, 201741 Gait mat F (112) 20 to 29 140 (15) M (792) 70 to 79 127 (20) Convenience F (42) 30 to 39 134 (15) Freund, 201930 5.0 m M (407) 80 to 89 115 (19) Switzerland 4.6 m F (83) 40 to 49 135 (15) Convenience Motion analysis M (68) 40 to 49 146 (18) Gait mat F (31) 50 to 59 126 (18) USA F (208) 50 to 59 141 (19) McKay, 201742 F (162) 60 to 69 125 (18) Fryzowicz, 201831 system F (1,239) 60 to 69 133 (18) Population-based 7.3 m M (21) 65 to 69 102 (21) Convenience 23.0 m F (938) 70 to 79 120 (20) Australia Gait mat M (39) 70 to 75 108 (20) Poland Motion analysis F (423) 80 to 89 105 (23) M (28) 76 to 79 100 (21) Gillain, 20176 system F (79) 60 to 69 134 (25) Middleton, 201643 4.6 m M (41) 80 to 89 94 (20) Convenience M (5) 70 to 79 105 (22) Convenience Gait mat F (40) 65 to 69 110 (25) Belgium M (9) 80 to 89 95 (14) USA 15.0 m F (36) 71 to 75 103 (21) M (9) 50 to 59 111 (16) Inertial sensor F (36) 76 to 79 91 (22) F (5) 60 to 69 119 (19) system F (43) 80 to 89 84 (21) F (20) 70 to 79 102 (21) F (13) 60 to 69 123 (16) F (12) 80 to 89 94 (12) 6.1 m F (43) 70 to 79 121 (23) F (4) 70 to 79 155 (58) Gait mat F (23) 80 to 89 108 (21) M (4) 80 to 89 130 (15) M (17) 18 to 19 152 (16) M (26) 90 to 99 109 (38) 10.0 m M (9) 20 to 27 155 (11) M (5) 70 to 79 134 (26) Stopwatch F (43) 18 to 19 158 (19) F (15) 80 to 89 105 (12) F (12) 20 to 22 143 (17) F (51) 90 to 99 80 (17) M (10) 20 to 29 130 (19) F (17) 50 to 59 140 (30) M (3) 60 to 69 160 (35) F (18) 50 to 59 130 (30) M (10) 70 to 79 144 (15) F (14) 60 to 69 150 (20) F (10) 20 to 29 144 (10) F (16) 60 to 69 140 (20) F (4) 60 to 69 145 (9) F (10) 70 to 79 130 (10) F (7) 70 to 79 143 (13) F (10) 70 to 79 130 (19) M (50) 20 to 29 133 (14) F (6) 65 to 69 123 (19) M (50) 30 to 39 131 (13) M (419) 70 to 74 115 (20) M (50) 40 to 49 133 (17) M (489) 75 to 79 110 (19) M (50) 50 to 59 125 (15) M (391) 80 to 84 102 (21) M (50) 60 to 69 124 (13) M (124) 85 to 89 103 (18) M (50) 70 to 79 125 (17) M (19) 65 to 69 113 (18) M (50) 80 to 99 110 (21) F (663) 70 to 74 106 (18) F (50) 20 to 29 137 (14) F (764) 75 to 79 98 (19) F (50) 30 to 39 131 (14) F (707) 80 to 84 92 (19) F (50) 40 to 49 134 (15) F (195) 85 to 89 91 (17) F (50) 50 to 59 130 (14) F (16) 50 to 59 144 (19) F (50) 60 to 69 124 (17) F (20) 60 to 69 142 (14) F (50) 70 to 79 117 (23) F (20) 70 to 79 133 (15) F (50) 80 to 99 109 (23) F (20) 20 to 29 137 (11) M (5) 70 to 79 114 (17) F (28) M (8) 80 to 89 94 (20) 60 to 69 143 (17) M (3) 90 to 93 96 (15) M (9) 70 to 79 142 (9) F (3) 60 to 69 113 (10) M (13) 80 to 89 123 (3) F (22) 70 to 79 116 (19) M (3) 60 to 69 128 (21) F (20) 80 to 89 107 (22) F (25) 70 to 79 111 (18) F (2) 90 to 90 120 (3) F (7)","50 Andrews et al: Normal gait speed by geographical region Table 1 Continued Table 1 Continued Study Test distance and Participants Speed (cm\/s) Study Test distance and Participants Speed (cm\/s) Sampling equipment Sampling equipment Sex (n) Age (y) Mean (SD) Sex (n) Age (y) Mean (SD) Nilsson, 201744 6.0 m Waanders, 201656 5.0 m Population-based Stopwatch F (954) 75 to 79 130 (20) Convenience Motion analysis M (13) 20 to 29 141 (13) Sweden Netherlands F (11) 20 to 29 145 (16) Novaes, 201145 10.0 m system Convenience Stopwatch Brazil M (9) 40 to 49 135 (11) M (13) 50 to 59 134 (22) Oh-Park, 201046 4.6 m M (11) 60 to 69 126 (15) Including studies from the previously published meta-analysis,9 Convenience Gait mat M (10) 70 to 79 109 (18) data from a total of 79 articles were included in this updated meta- USA F (11) 40 to 49 127 (20) analysis. The results of the meta-analysis of 51,248 apparently F (8) 50 to 59 127 (15) healthy adults strati\ufb01ed by age\/decade and sex are presented in Ruggero, 201347 4.6 m F (9) 60 to 69 107 (17) Table 2. The weighted mean of gait speed for females ranged from 97 Population-based Stopwatch F (8) 70 to 79 102 (10) cm\/s (females aged \u0015 80 years) to 138 cm\/s (females aged 18 to 29 Brazil M (26) 70 to 74 112 (20) years). The weighted mean of gait speed for males ranged from 105 10.0 m M (45) 75 to 79 112 (18) cm\/s (males aged \u0015 80 years) to 140 cm\/s (males aged 40 to 49 years). Sampaio, 201748 Stopwatch M (28) 80 to 84 108 (17) Convenience M (17) 85 to 89 102 (18) The I2 values ranged from 0 to 34; prediction interval ranges Brazil F (52) 70 to 74 110 (15) varied from a low of 30 (125 to 155 cm\/s) to a high of 77 (83 to 160 F (65) 75 to 79 102 (19) cm\/s). Thus, with 95% certainty, the average observed gait speed for Seino, 201449 5.0 m F (49) 80 to 84 100 (16) females aged between 18 to 29 years falls between 134 and 143 cm\/s; Cohort Stopwatch F (22) 85 to 89 101 (15) 15% of the variance in observed gait speed re\ufb02ects variance in true Japan M (59) 65 to 69 123 (31) gait speed rather than sampling error. With 95% certainty, the true M (61) 70 to 79 118 (27) gait speed for females aged between 18 and 29 years falls between Singh, 201850 6.0 m M (16) 80 to 89 94 (19) 117 and 159 cm\/s (prediction interval range: 42 cm\/s). Other rows of Cross-sectional Stopwatch F (125) 65 to 69 112 (20) results presented for other groups in Table 2 may be interpreted Malaysia F (100) 70 to 79 106 (22) similarly. F (20) 80 to 89 85 (34) Southard, 201051 7.0 m F (3) 90 to 99 55 (14) The results of the meta-analysis for Asia, Australia, Europe, North Convenience Stopwatch M (25) 60 to 64 140 (20) America and South America strati\ufb01ed by age and sex are presented in USA M (35) 65 to 69 140 (20) Tables 3 to 7, respectively (see eAddenda for Tables 3 to 7). Differ- 2.0 m M (27) 70 to 74 130 (30) ences in recorded gait speeds between geographical regions (ie, Svoboda, 201752 Photocell M (24) 75 to 79 120 (30) continents) could not be analysed with an analysis of variance, since Convenience sensors M (11) 80 to 89 100 (30) assumptions of homoscedasticity were violated. Several groups Czech Republic F (127) 60 to 64 130 (20) further strati\ufb01ed by continent had too few source articles to develop F (123) 65 to 69 130 (20) con\ufb01dence interval values. There were overlaps in con\ufb01dence interval Thaweewannakij, 201353 3.0 m F (102) 70 to 74 120 (20) values for each sex\/age group between continents, based on obser- Population-based Stopwatch F (68) 75 to 79 120 (30) vation. The only exception to this overlap involved Asian females and Thailand F (35) 80 to 89 100 (30) males aged between 18 and 29 years, whose gait speeds were lower M (479) 65 to 69 139 (22) than the gait speeds of their age-matched and sex-matched coun- Turcato, 201554 20.0 m M (722) 70 to 74 133 (23) terparts residing on other continents. Convenience Photocell M (612) 75 to 79 126 (24) Italy sensors M (293) 80 to 84 116 (25) Discussion Vallabhajosula, 201955 M (48) 85 to 99 111 (28) Convenience 4.9 m F (588) 65 to 69 139 (22) The investigation methods were consistent with those employed USA Gait mat F (845) 70 to 74 131 (23) in the original meta-analysis of normative gait speed values.9 The F (653) 75 to 79 118 (25) results of this current study led to the addition of comfortable gait F (352) 80 to 84 105 (28) speed data from 28,137 healthy adults who participated in studies F (68) 85 to 99 92 (25) published over the past decade. This resulted in gait speed data for M (88) 60 to 69 110 (27) 51,248 healthy adults strati\ufb01ed by sex and decade of life. M (49) 70 to 79 100 (29) M (8) 80 to 89 93 (19) In comparison with the previous study, all new strati\ufb01ed and F (131) 60 to 69 107 (23) weighted means fell within the previously reported 95% CIs;9 thus, the F (47) 70 to 79 96 (24) updated weighted means are unsurprising, given the previous work. F (2) 80 to 89 74 (28) However, the 95% CIs in the current meta-analysis are narrower than M (2) 40 to 49 109 (48) they were in the previous study. Increasing the number of analysed M (5) 50 to 59 119 (36) participants by 128% helped to narrow the con\ufb01dence intervals. F (4) 40 to 49 105 (37) F (7) 50 to 59 125 (18) Except for males aged 18 to 29 years, prediction interval ranges F (2) 60 to 69 150 (0) were lower for both males and female participants aged , 50 years. M (10) 60 to 69 118 (21) This indicates better homogeneity among the younger groups than M (11) 70 to 79 103 (13) the older groups. Even though all investigators required all their M (2) 80 to 89 115 (4) participants to be independently ambulatory in the community, some F (51) 60 to 69 119 (14) participants with comorbidities that may affect gait speed (eg, dia- F (39) 70 to 79 107 (16) betes mellitus)44 could have been included in some studies in this F (9) 80 to 89 105 (13) meta-analysis but not in others. Since older individuals are more M (96) 60 to 69 116 (21) likely to have comorbidities that may affect gait, this variance in M (180) 70 to 79 109 (20) exclusion criteria may have contributed to less homogeneity in the M (44) 80 to 90 97 (20) older participants. Another viable explanation for this observation of F (320) 60 to 69 108 (15) less homogeneity among older adults is that health status may be F (329) 70 to 79 99 (15) more variable among otherwise independent older adults, leading to F (61) 80 to 94 88 (18) M (10) 20 to 29 151 (10) F (13) 20 to 29 155 (15) M (2) 60 to 69 131 (11) M (4) 70 to 79 128 (7) M (9) 80 to 96 116 (35) F (6) 60 to 69 134 (19) F (18) 70 to 79 109 (27) F (8) 80 to 92 94 (32)","Research 51 Table 2 Participants (n) Gait speed (cm\/s) Homogeneity Meta-analysis results strati\ufb01ed by age and sex. 508 Prediction interval Age Sex Source articles (n) 182 (min to max) 456 2,417 Weighted mean 95% CI I2 (%) 42 (117 to 159) 9,448 (min to max) 31 (122 to 153) 15,361 15 38 (115 to 153) 18 to 29 F 23 3,588 138 (108 to 158) 134 to 143 34 63 (93 to 156) 30 to 39 F 10 318 138 (128 to 163) 133 to 143 4 77 (83 to 160) 40 to 49 F 12 134 134 (105 to 150) 129 to 139 15 65 (81 to 146) 50 to 59 F 23 163 125 (90 to 155) 118 to 131 0a 58 (68 to 126) 60 to 69 F 55 1,947 121 (70 to 150) 116 to 126 0a 74 (99 to 173) 70 to 79 F 78 4,074 113 (65 to 150) 110 to 117 6 34 (122 to 156) 801 F 55 6,632 93 to 101 0a 30 (125 to 155) 18 to 29 M 25 1,784 97 (55 to 142) 129 to 143 0a 60 (104 to 164) 30 to 39 M 11 136 (108 to 155) 134 to 144 4 74 (89 to 163) 40 to 49 M 9 139 (120 to 154) 135 to 145 0a 50 (95 to 145) 50 to 59 M 13 140 (109 to 151) 126 to 142 0a 69 (71 to 140) 60 to 69 M 35 134 (112 to 151) 120 to 132 15 70 to 79 M 54 126 (73 to 160) 117 to 123 0a 801 M 42 120 (72 to 155) 100 to 111 105 (83 to 130) F = female, M = male, max = maximum, min = minimum. less homogeneous gait speeds. In support of this latter explanation, prediction intervals only if there were more than 25 participants in Oh-Park et al found slower comfortable gait speeds in older adults that group. Nevertheless, the calculation of prediction intervals with gait abnormalities compared with older adults without gait may not have been appropriate for groups with less than 10 source abnormalities, even though all participants were community- articles.17 dwelling and independently ambulatory.46 This study provides the results of a meta-analysis of the In concurrence with previous studies, gait speed decreased as age comfortable gait speed of 51,248 healthy adults strati\ufb01ed by age\/ increased.5,57 There was a steady decline in the weighted mean of gait decade and sex. Tests of homogeneity were applied for groups with speed for females with increasing age. The weighted mean of gait more than \ufb01ve source articles and more than 25 subjects. Based on speed for males remained stable from 18 to 29 years to 40 to 49 years prediction interval ranges, younger adults may be more homoge- and thereafter decreased with increasing age. neous than older adults. Results were also presented for each stratum by continent. Based on this analysis, further strati\ufb01cation of There was considerable overlap in con\ufb01dence interval values comfortable gait speed by continent is not warranted. for each sex\/age group between continents. The only exception to this overlap involved Asian females and males aged between 18 What is already known on this topic: Comfortable gait and 29 years. The gait speed con\ufb01dence intervals for both groups speed is a useful indicator of the mobility and health status of were lower and did not overlap at all with the gait speed con\ufb01- individuals. Previous studies have suggested possible dence intervals for those of the same sex and age who resided on geographical differences in comfortable gait speed. other continents. However, this geographical differentiation is What this study adds: Meta-analysis of data from 51,248 tenuous, given that gait speeds were obtained from 21 Asian fe- healthy adults showed that comfortable gait speed progressively males and 26 Asian males in the 18 to 29 age group. Thus, the slows by decade among males aged . 50 years and females results of this study do not support the differentiation of aged . 30 years. Stratification of comfortable gait speed by comfortable gait speeds by geographical region. Considerable geographical region is not warranted, given considerable overlap physical and cultural variation, even within the same continent, in confidence intervals between continents. may explain some of the consistent overlap in gait speed con\ufb01- dence intervals between continents. eAddenda: Tables 3 to 7 can be found online at https:\/\/doi.org\/10. 1016\/j.jphys.2022.11.005 There were some limitations to the current study. Ironically, the rigid application of the eligibility criteria may have limited Ethics approval: The Elon University Institutional Review Board the number of suitable participants who were included in the (Ethics Committee) approved this study. analysis, for example: since gait speed may be in\ufb02uenced by whether or not the acceleration and decelerations were timed,14 Competing interests: Nil. documentation of untimed acceleration and deceleration phases Source(s) of support: Nil. or con\ufb01rmation from the investigators themselves was required Acknowledgements: The authors acknowledge the diligent con- prior to inclusion in the database. When a de\ufb01nite response to tributions of Riley Beck, Hannah Buckingham and Rory Cullen. the inquiries could not be obtained, those participants were not Providence: Not invited. Peer reviewed. included; thus, participants whose gait speeds were actually Correspondence: Bill Andrews, Department of Physical Therapy measured with untimed acceleration and deceleration phases may Education, Elon University, Elon, USA. Email: [email protected] have been excluded. References Also, bias may not have been adequately accounted for, although this was a meta-analysis of a clinical measurement, not an inter- 1. Taylor JK, Buchan IE, van der Veer SN. Assessing life-space mobility for a more vention. In order to mitigate the risk of bias, the eligibility criteria holistic view on wellbeing in geriatric research and clinical practice. Aging Clin Exp were strictly applied and agreement from two or more investigators Res. 2019;31:439\u2013445. was required before including a study in the analysis. Furthermore, other confounding variables that affect gait speed besides age and sex 2. Hardy SE, Perera S, Roumani YF, Chandler JM, Studenski SA. Improvement in usual were not accounted for (eg, height),29,58 although they were also gait speed predicts better survival in older adults. J Am Geriatr Soc. 2007;55:1727\u2013 infrequently reported. 1734. Another limitation of this study lay in the calculation of pre- 3. Middleton A, Fritz SL, Lusardi M. Walking speed: the functional vital sign. J Aging diction intervals, particularly when the groups were further Phys Act. 2015;23:314\u2013322. strati\ufb01ed by continent. We chose to calculate prediction intervals only for those groups that contained participants from \ufb01ve or 4. Salbach NM, O\u2019Brien KK, Brooks D, et al. Reference values for standardized more source articles. Since this is a lower cut-off for calculation of tests of walking speed and distance: a systematic review. Gait Posture. a prediction interval than others suggest,17 we also calculated 2015;41:341\u2013360.","52 Andrews et al: Normal gait speed by geographical region 5. Bohannon RW, Wang YC. Four-meter gait speed: normative values and reliability 33. Gunasekaran V, Banerjee J, Dwivedi SN, Upadhyay AD, Chatterjee P, Dey AB. determined for adults participating in the NIH Toolbox Study. Arch Phys Med Normal gait speed, grip strength and thirty seconds chair stand test among older Rehabil. 2019;100:509\u2013513. Indians. Arch Gerontol Geriatr. 2016;67:171\u2013178. 6. Gillain S, Boutaayamou M, Dardenne N, et al. Data set of healthy old people 34. Hollman JH, McDade EM, Petersen RC. Normative spatiotemporal gait parameters assessed for three walking conditions using accelerometric and opto-electronic in older adults. Gait Posture. 2011;34:111\u2013118. methods. Aging Clin Exp Res. 2017;29:1201\u20131209. 35. Kemmler W, Teschler M, Goisser S, et al. Prevalence of sarcopenia in Germany and 7. Herssens N, Verbecque E, Hallemans A, Vereeck L, Van Rompaey V, Saeys W. Do the corresponding effect of osteoarthritis in females 70 years and older living in spatiotemporal parameters and gait variability differ across the lifespan of healthy the community: results of the FORMoSA study. Clin Interv Aging. 2015;10:1565\u2013 adults? A systematic review. Gait Posture. 2018;64:181\u2013190. 1573. 8. Kawai H, Taniguchi Y, Seino S, et al. Reference values of gait parameters measured 36. Kirkwood RN, Gomes HA, Sampaio RF, Furtado SR, Moreira BS. Spatiotemporal and with a plantar pressure platform in community-dwelling older Japanese adults. variability gait data in community-dwelling elderly women from Brazil. Braz J Phys Clin Interv Aging. 2019;14:1265\u20131276. Ther. 2016;20:258\u2013266. 9. Bohannon RW, Andrews AW. Normal walking speed: a descriptive meta-analysis. 37. Kirkwood RN, Moreira B, Mingoti SA, Faria BF, Sampaoi RF, Resende RA. The Physiotherapy. 2011;97:182\u2013189. slowing down phenomenon: What is the age of major gait velocity decline? Maturitas. 2018;115:31\u201336. 10. Ritchie RF, Palomaki G. Selecting clinically relevant populations for reference in- tervals. Clin Chem Lab Med. 2004;42:702\u2013709. 38. Lee SW, Verghese J, Holtzer R, Mahoney JR, Oh-Park M. Trunk sway during walking among older adults: norms and correlation with gait velocity. Gait Posture. 11. Strauss E, Sherman EM, Spreen O. A Compendium of Neuropsychological Tests. 2014;40:676\u2013681. Administration, Norms, and Commentary. New York: Oxford University Press; 2006. 39. L\u00f6fgren N, Halvarsson A, St\u00e5hle A, Franz\u00e9n E. Gait characteristics in older women 12. Moher D, Shamseer L, Clarke M, et al. Preferred reporting items for systematic with osteoporosis and fear of falling. Eur J Physiother. 2013;15:139\u2013145. review and meta-analysis protocols (PRISMA-P) 2015 statement. Syst Rev. 2015;4:1. 40. Lythgo N, Wilson C, Galea M. Basic gait and symmetry measures for primary 13. Strang J, Pilling S, Albert E, et al. Drug Misuse: Psychosocial Interventions. Leicester school-aged children and young adults. II: walking at slow, free and fast speed. Gait (UK): British Psychological Society; 2008. Posture. 2011;33:29\u201335. 14. Pasma JH, Stijntjes M, Ou SS, Blauw GJ, Meskers CG, Maier AB. Walking speed in 41. Malatesta D, Canepa M, Menendez Fernandez A. The effect of treadmill and elderly outpatients depends on the assessment method. Age (Dordr). overground walking on preferred walking speed and gait kinematics in healthy, 2014;36:9736. physically active older adults. Eur J Appl Physiol. 2017;117:1833\u20131843. 15. Phan-Ba R, Calay P, Grodent P, et al. A corrected version of the Timed-25 Foot Walk 42. McKay MJ, Baldwin JN, Ferreira P, et al. Spatiotemporal and plantar pressure pat- Test with a dynamic start to capture the maximum ambulation speed in multiple terns of 1000 healthy individuals aged 3-101 years. Gait Posture. 2017;58:78\u201387. sclerosis patients. NeuroRehabilitation. 2012;30:261\u2013266. 43. Middleton A, Fulk GD, Beets MW, Herter TM, Fritz SL. Self-selected walking speed is 16. Neyeloff JL, Fuchs SC, Moreira LB. Meta-analyses and Forest plots using a microsoft predictive of daily ambulatory activity in older adults. J Aging Phys Act. excel spreadsheet: step-by-step guide focusing on descriptive data analysis. BMC 2016;24:214\u2013222. Res Notes. 2012;5:52. 44. Nilsson AG, Sundh D, Johansson L, et al. Type 2 diabetes mellitus is associated with 17. Borenstein M. Research Note: In a meta-analysis, the I2 index does not tell us how better bone microarchitecture but lower bone material strength and poorer much the effect size varies across studies. J Physiother. 2020;66:135\u2013139. physical function in elderly women: a population-based study. J Bone Miner Res. 2017;32:1062\u20131071. 18. Higgins JP, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta- analyses. BMJ. 2003;327:557\u2013560. 45. Novaes RD, Miranda AS, Dourado VZ. Usual gait speed assessment in middle-aged and elderly Brazilian subjects. Rev Bras Fisioter. 2011;15:117\u2013122. 19. Borenstein M, Hedges LV, Higgins JPT, HR R. Introduction to Meta-analysis. West Sussex, UK: John Wiley & Sons, Ltd; 2009. 46. Oh-Park M, Holtzer R, Xue X, Verghese J. Conventional and robust quantitative gait norms in community-dwelling older adults. J Am Geriatr Soc. 2010;58:1512\u20131518. 20. Jackson D, Turner R. Power analysis for random-effects meta-analysis. Res Synth Methods. 2017;8:290\u2013302. 47. Ruggero CR, Bilton TL, Teixeira LF, et al. Gait speed correlates in a multiracial population of community-dwelling older adults living in Brazil: a cross-sectional 21. Adams HHH, Verlinden VJA, Callisaya ML, et al. Heritability and genome-wide population-based study. BMC Public Health. 2013;13:182. association analyses of human gait suggest contribution of common variants. J Gerontol A Biol Sci Med Sci. 2016;71:740\u2013746. 48. Sampaio RAC, Sampaio PYS, Castano LAA, et al. Cutoff values for appendicular skeletal muscle mass and strength in relation to fear of falling among Brazilian 22. Alcock L, O\u2019Brien TD, Vanicek N. Age-related changes in physical functioning: older adults: cross-sectional study. Sao Paulo Med J. 2017;135:434\u2013443. correlates between objective and self-reported outcomes. Physiotherapy. 2015;101:204\u2013213. 49. Seino S, Shinkai S, Fujiwara Y, et al. Reference values and age and sex differences in physical performance measures for community-dwelling older Japanese: a pooled 23. Bailey S, Bailey E. Predictors of Health Related Quality of Life in Adults 50 Years and analysis of six cohort studies. PLoS One. 2014;9:e99487. Older. J Nov Physiother. 2016;6:281. 50. Singh DKA, Murukesu R, Shahar S, et al. Discriminative and predictive ability of 24. Bruening DA, Frimenko RE, Goodyear CD, Bowden DR, Fullenkamp AM. Sex dif- physical performance measures in identifying fall risk among older adults. Sains ferences in whole body gait kinematics at preferred speeds. Gait Posture. Malaysiana. 2018;47:2769\u20132776. 2015;41:540\u2013545. 51. Southard V, Dave A, Douris P. Exploring the role of body mass index on balance 25. Canney M, Sexton DJ, O\u2019Connell MD, Kenny RA, Little MA, O\u2019Seaghdha CM. Kidney reactions and gait in overweight sedentary middle-aged adults: a pilot study. J Prim function estimated from cystatin C, but not creatinine, is related to objective tsts of Care Community Health. 2010;1:178\u2013183. physical performance in community-dwelling older adults. J Gerontol A Biol Sci Med Sci. 2017;72:1554\u20131560. 52. Svoboda Z, Bizovska L, Janura M, Kubonova E, Janurova K, Vuillerme N. Variability of spatial temporal gait parameters and center of pressure displacements during 26. Chen HY, Tang PF. Factors contributing to single- and dual-task Timed \u201cUp & Go\u201d gait in elderly fallers and nonfallers: a 6-month prospective study. PLoS One. test performance in middle-aged and older adults who are active and dwell in the 2017;12:e0171997. community. Phys Ther. 2016;96:284\u2013292. 53. Thaweewannakij T, Wilaichit S, Chuchot R, et al. Reference values of physical 27. Chui KK, Lusardi MM. Spatial and temporal parameters of self-selected and fast performance in Thai elderly people who are functioning well and dwelling in the walking speeds in healthy community-living adults aged 72-98 years. J Geriatr Phys community. Phys Ther. 2013;93:1312\u20131320. Ther. 2010;33:173\u2013183. 54. Turcato AM, Godi M, Giordano A, Schieppati M, Nardone A. The generation of 28. Crockett K, Kontulainen SA, Farthing JP, et al. Differences in function and fracture centripetal force when walking in a circle: insight from the distribution of ground risk in postmenopausal women with and without a recent distal radius fracture. reaction forces recorded by plantar insoles. J Neuroeng Rehabil. 2015;12:4. J Aging Phys Act. 2018;26:136\u2013145. 55. Vallabhajosula S, Humphrey SK, Cook AJ, Freund JE. Concurrent validity of the Zeno 29. Elbaz A, Artaud F, Dugravot A, Tzourio C, Singh-Manoux A. The gait speed Walkway for measuring spatiotemporal gait parameters in older adults. J Geriatr advantage of taller stature is lost with age. Sci Rep. 2018;8:1485. Phys Ther. 2019;42:E42\u2013E50. 30. Freund JE, Stetts DM, Oostindie A, Shepherd J, Vallabhajosula S. Lower 56. Waanders J, Beijersbergen C, Murgia A, Hortobagyi T. Functional relevance of Quarter Y-Balance Test in healthy women 50-79 years old. J Women Aging. relative maintenance of maximal eccentric quadriceps torque in healthy old adults. 2019;31:475\u2013491. Gerontology. 2016;62:588\u2013596. 31. Fryzowicz A, Murawa M, Kabacinski J, Rzepnicka A, Dworak LB. Reference values of 57. Beauchet O, Allali G, Sekhon H, et al. Guidelines for assessment of gait and refer- spatiotemporal parameters, joint angles, ground reaction forces, and plantar ence values for spatiotemporal gait parameters in older adults: the Biomathics and pressure distribution during normal gait in young women. Acta Bioeng Biomech. Canadian Gait Consortiums Initiative. Front Hum Neurosci. 2017;11:353. 2018;20:49\u201357. 58. Bendall MJ, Bassey EJ, Pearson MB. Factors affecting walking speed of elderly 32. Goutier KM, Jansen SL, Horlings CG, Kung UM, Allum JH. The in\ufb02uence of walking people. Age Ageing. 1989;18:327\u2013332. speed and gender on trunk sway for the healthy young and older adults. Age Ageing. 2010;39:647\u2013650.","Journal of Physiotherapy 69 (2023) 6 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 Paper of the Year 2022 The Editorial Board of Journal of Physiotherapy is pleased to announce the 2022 Paper of the Year Award. The winning paper is judged by a panel of members of the International Advisory Board who do not have a con\ufb02ict of interest with any of the papers under consideration. They vote for the paper published in the 2022 calendar year that, in their opinion, has the best combination of scienti\ufb01c merit and application to the clinical practice of physiotherapy. The winning paper is Some conservative interventions are more effective than others for people with chronic non-speci\ufb01c neck pain: a systematic review and network meta-analysis.1 The authors are Greta Castellini, Paolo Pillastrini, Carla Vanti, Silvia Bargeri, Silvia Giagio, Elena Bordignon, Francesco Fasciani, Francesco Marzioni, Tiziano Innocenti, Alessandro Chiarotto, Silvia Gianola and Lucia Bertozzi. The winning paper expertly summarises the results of an enormous amount of research: 119 randomised trials with over 12,000 participants. The review found evidence that a combination of active and\/or passive multimodal non-pharmacological inventions (eg, exercise and manual therapy) were effective for pain on a 0-to-10 scale at 1 month (MD range 0.84 to 3.74) and at 3 to 6 months (MD range 1.06 to 1.49), and effective on disability on a 0-to-100 scale at 1 month (MD range 10.26 to 14.09) and 3 to 6 months (MD range 5.60 to 16.46) in people with chronic non- speci\ufb01c neck pain. Compared with inert treatment, anti-in\ufb02ammatory drugs alone or in combination with another non-pharmacological treatment did not reduce pain at 1 month, at 3 to 6 months, or at 12 months. Most adverse events were experienced following acupunc- ture\/dry needling intervention. The winning paper is one of several network meta-analyses1\u20133 published by the Journal. It gave due consideration to assumptions about transitivity, inconsistency, heterogeneity and coherence. It is a wonderful complement to the Journal\u2019s recent Invited Topical Review on physiotherapy management of neck pain.4 The members of the Editorial Board congratulate Greta Castellini, Paolo Pillastrini and their co-authors on their success. References 1. Castellini G, Pillastrini P, Vanti C, Bargeri S, Giagio S, Bordignon E, et al. Some conservative interventions are more effective than others for people with chronic non-speci\ufb01c neck pain: a systematic review and network meta-analysis. J Physiother. 2022;68:244\u2013254. 2. Hayden JA, Ellis J, Ogilvie R, Stewart SA, Bagg MK, Stanojevic S, et al. Some types of exercise are more effective than others in people with chronic low back pain: a network meta- analysis. J Physiother. 2021;67:252\u2013262. 3. Scapini KB, Bohlke M, Moraes OA, Rodrigues CG, In\u00e1cio JF, Sbruzzi G, et al. Combined training is the most effective training modality to improve aerobic capacity and blood pressure control in people requiring haemodialysis for end-stage renal disease: systematic review and network meta-analysis. J Physiother. 2019;65:4\u201315. 4. Verhagen AP. Physiotherapy management of neck pain. J Physiother. 2021;67:5\u201311. https:\/\/doi.org\/10.1016\/j.jphys.2022.12.001 1836-9553\/","Journal of Physiotherapy 69 (2023) 7\u201314 j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m \/ l o c a t e \/ j p hy s Invited Topical Review Physiotherapy management of bronchiectasis in adults Annemarie L Lee a,b a Department of Physiotherapy, Monash University, Melbourne, Australia; b Institute for Breathing and Sleep, Melbourne, Australia KEY WORDS [Lee AL (2023) Physiotherapy management of bronchiectasis in adults. Journal of Physiotherapy 69:7\u201314] \u00a9 2022 Australian Physiotherapy Association. Published by Elsevier B.V. This is an open access article under the Bronchiectasis Physical therapy CC BY-NC-ND license (http:\/\/creativecommons.org\/licenses\/by-nc-nd\/4.0\/). Pulmonary rehabilitation Education Introduction studied.4 There has been a consistent increase in recent years across a range of countries and in those of an older age.5,6 Given the increasing Bronchiectasis is characterised by chronic cough with sputum age of adults with bronchiectasis and that another condition is production, dyspnoea and reduced exercise tolerance, all of which considered to be the underlying cause in approximately 30% of peo- contribute to poorer health-related quality of life (HRQOL). Physio- ple, it is unsurprising that individuals with this syndrome have an therapy forms an integral part of management, with the goal of average of four comorbidities.7 Common comorbidities include addressing symptoms to improve health status and minimise the rate coexisting respiratory diseases or conditions that arise from the of decline in respiratory function. This review summarises the aeti- systemic in\ufb02ammation inherent in bronchiectasis such as vascular ologies and clinical presentation of bronchiectasis; the corresponding diseases, metastatic malignancy, gastro-oesophageal re\ufb02ux disease, burden on individuals and the healthcare system; key components of and musculoskeletal dysfunction with osteoporosis and peripheral bronchiectasis care, focusing on physiotherapy interventions; and muscle weakness. future directions for research and practice. The extent and clinical course of bronchiectasis is variable; some What is bronchiectasis individuals with mild symptoms at diagnosis or localised disease may report a stable trajectory, while those with severe symptoms at the Bronchiectasis, a chronic respiratory condition, is a clinical syn- outset may present with diffuse disease and experience a rapid drome characterised by radiologically abnormal and permanent progressive decline. Within the natural history of bronchiectasis, dilatation of the bronchi, persistent cough, airway in\ufb02ammation and acute exacerbations, de\ufb01ned as worsening of usual respiratory infection.1 Diagnosis is con\ufb01rmed via radiological features evident on symptoms and requiring a change in treatment,8 are signi\ufb01cant high-resolution computed tomography chest scans, which are un- events that in\ufb02uence an individual\u2019s clinical presentation, condition dertaken when an individual is clinically stable.2 Bronchiectasis is trajectory and overall prognosis. caused by a heterogeneous range of clinical disorders; common aetiologies are post-infection (bacterial, viral and mycobacterial), Burden of bronchiectasis genetic disorders (primary ciliary dyskinesia), humoral immunode- \ufb01ciencies, autoimmune diseases, in\ufb02ammatory conditions or Chronic cough is the most frequently experienced symptom, idiopathic. The pathophysiology of bronchiectasis is still best under- affecting up to 96% of adults with bronchiectasis. Sputum expecto- stood by the vicious cycle that was \ufb01rst proposed by Cole:3 airway ration, shortness of breath, haemoptysis, chest pain and, to a lesser in\ufb02ammation, primarily dominated by neutrophils, contributes to extent, fatigue have also been reported.9 Other clinical characteristics impaired mucociliary clearance; this impairment leads to mucus include extrapulmonary manifestations of peripheral muscle weak- hypersecretion and airway obstruction, with an increase in microbial ness, presenting as reduced functional exercise capacity and physical colonisation, which contributes to infection and structural lung activity,10 and anxiety and depression.11 The severity of symptoms is damage. Colonisation can arise from a mix of organisms, among often synonymous with the severity of disease and co-existing con- which Pseudomonas aeruginosa, Haemophilus in\ufb02uenzae and Strepto- ditions. Collectively, respiratory and systemic features, including coccus pneumoniae are the most frequently isolated pathogens. those derived from comorbid conditions, contribute to a heightened treatment burden, poorer physical\/social functioning, poorer role While the worldwide prevalence of bronchiectasis is not precisely performance and reduced perceptions of health.12 From the patient known, international data currently estimate rates between 53 and perspective, HRQOL is further impaired by social embarrassment and 566 cases per 100,000 inhabitants, depending on the population stigma due to chronic cough and sputum expectoration, limitations https:\/\/doi.org\/10.1016\/j.jphys.2022.11.013 1836-9553\/\u00a9 2022 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\/).","8 Lee: Physiotherapy management of bronchiectasis in adults on daily activities and psychological symptoms.13 Although not well Airway clearance techniques understood, intangible costs related to work productivity impairment due to absenteeism and presenteeism of those with bronchiectasis Airway clearance techniques are recommended as part of man- and the subsequent burden on caregivers and family as a result of agement of people with bronchiectasis, with techniques customised these indirect effects have been observed.14 and modi\ufb01ed as required, particularly during an acute exacerbation.19 Broad categories of techniques include positional options (gravity- Acute exacerbations are responsible for disease progression and assisted drainage), those which modulate expiratory \ufb02ow (active deterioration in lung function. The number of exacerbations per year cycle of breathing technique, autogenic drainage and forced expira- independently predicts future mortality, with the risk of mortality tory technique) and positive pressure devices (both oscillatory and doubled in frequent exacerbators.15 Those of the frequent exacerbator non-oscillatory devices). Techniques are applied across the disease phenotype have a greater likelihood of hospitalisation and poorer spectrum, with therapists modifying approaches to accommodate HRQOL.16 Although a causal relationship has not been demonstrated, comorbidities as necessary.21 The physiological rationale is that these the co-existence of chronic obstructive pulmonary disease (COPD) or techniques augment sputum clearance via a number of mechanisms: asthma in 20 to 60% of individuals with bronchiectasis is linked to a improvement in collateral ventilation and independence; increase in higher 5-year mortality rate (55% compared with 20% among those expiratory air\ufb02ow velocity; reduction in airway resistance; use of without COPD or asthma).17 gravity; changes in airway pressure; and generation of airway oscil- lations. A recent review of Australian practice found that up to 58% of Within the primary and secondary healthcare settings, direct costs adults with bronchiectasis regularly used airway clearance tech- to the health system are signi\ufb01cant. A systematic review found that niques,22 and the proportion of clinicians prescribing techniques those with bronchiectasis have a longer length of hospital stay (4.5 increased to 89% during an acute exacerbation.21 versus 2.5 days), a greater number of admissions (0.6 versus 0.4 per year), more outpatient consultations (20 versus 14), more emergency When compared with no treatment, a Cochrane review showed room visits and more days of medication use compared with those that airway clearance techniques were associated with short-term without bronchiectasis.14 Costs in the primary and secondary care improvement in sputum expectoration (MD 8.4 ml, 95% CI 3.4 to settings in Spain have been reported at V4,672, which increased to 13.4) and health status (MD \u201314.8 points, 95% CI \u201318.0 to \u201311.6), but V9,999 in those with a high mortality risk, while annual costs ranged had no clear effect on acute exacerbation rate (RR 0.71, 95% CI 0.23 to from US$13,244 to US$67,764 in the USA.14 This is largely driven by 2.25).23 Acknowledging the different mechanisms of action between higher annual hospitalisation rates and medication (respiratory and techniques, a more recent Cochrane review showed that positive those prescribed to manage comorbidities). Those colonised with expiratory pressure (PEP) therapy had similar effects on HRQOL, Pseudomonas aeruginosa have a 6.5-fold increase in hospital admis- symptoms, sputum expectoration and lung volumes compared with sions.18 Costs have not been quanti\ufb01ed for the time required for other techniques for clinically stable patients.24 In those with an acute treatment (including physiotherapy), diagnostic tests, monitoring, exacerbation of bronchiectasis, six studies of 120 patients highlighted nutritional interventions, and requirements for social care, including the safety of a range of airway clearance techniques, with suggestions adaptations to the home for those with severe disease, but such costs that the active cycle of breathing technique may offer more clinical are likely to be considerable. bene\ufb01t compared with gravity-assisted drainage and manual tech- niques.25 The results of these reviews were heavily in\ufb02uenced by Management of bronchiectasis studies of a short (single sessions to maximum of 1 week) or medium duration (4 weeks to 3 months) for selected outcomes, and a mix of The main goals of treatment for bronchiectasis are to control and techniques and outcomes, which limited the pooling of data, irre- alleviate symptoms, reduce the incidence of acute exacerbations and spective of clinical state. prevent disease progression. Following the game-changing studies of macrolides with reduced frequency of acute exacerbations, robust Since these reviews, additional short-term, randomised controlled evidence supports their use.19,20 Other medical approaches comprise trials comparing techniques have emerged.26\u201328 The effects of anti-in\ufb02ammatory agents, inhaled antibiotics and bronchodilators, therapist-made Bottle PEP therapy found that a single treatment corticosteroid and immune-modulating therapies.2,19,20 Physio- session led to equivalent sputum weight during treatment compared therapy is equally important in bronchiectasis management; a sum- with the active cycle of breathing technique.27 These \ufb01ndings are mary of interventions together with the level of evidence that encouraging, given that this is a simple and inexpensive method of underpins their use is provided in Figure 1. applying oscillating PEP therapy. The longer-term effect of techniques modulating expiratory \ufb02ow is outlined by the pooled analysis Several high-quality Benefit from the intervention No effect from the intervention randomised trials Airway clearance techniques Pulmonary rehabilitation during Pulmonary rehabilitation when an acute exacerbation clinically stable Single high-quality Domiciliary humidified oxygen randomised trials or therapy several low-quality randomised trials Respiratory muscle training Uncontrolled trials or Mucoactive agents (hypertonic expert opinion and isotonic saline) Non-invasive ventilation Rehabilitation pre- and post- transplantation Education and self- management Figure 1. Interventions for the management of bronchiectasis with associated levels of evidence.","Invited Topical Review 9 Study OR (95% CI) for those with recurrent exacerbations.19 This form of pharmacological Livnat 2021 Random therapy aims to improve mucociliary clearance and promote sputum Uzmezoglu 2018 expectoration by reducing the viscosity of secretions and improving sputum hydration. Mucoactive agents prescribed in people with bron- Pooled chiectasis include nebulised hypertonic (concentrations ranging from 3 to 10%) and isotonic saline (concentration of 0.9%), with recent data 0.05 0.2 1 5 20 outlining that 14% and 4% of Australians with bronchiectasis use hyper- tonic or isotonic saline, respectively, on a regular basis.35 Favours ACBT or AD Favours oscillating PEP A previous systematic review comparing hypertonic with isotonic Figure 2. Increase in self-reported sputum expectoration comparing oscillating PEP saline in adults with bronchiectasis reported similar improvement in therapy and techniques modulating expiratory \ufb02ow for 4 weeks. Data are odds ratios. dynamic lung function measurements (forced expiratory volume in 1 ACBT = active cycle of breathing technique, AD = autogenic drainage, PEP = positive second (SMD 0.34, 95% CI \u20130.06 to 0.75) and forced vital capacity expiratory pressure. (SMD 0.40, 95% CI \u20130.01 to 0.80) and a mixed impact on HRQOL following 3 months of treatment, or no difference between tech- demonstrating similar proportions of individuals reporting an in- niques after 12 months of treatment.36 Similar effects on selected crease in self-reported sputum expectoration, irrespective of tech- HRQOL tools were noted in a recent study of 22 people with primary nique, following 4 weeks of daily oscillating PEP therapy versus ciliary dyskinesia comparing twice daily hypertonic saline (7%) with autogenic drainage26 or twice daily oscillating PEP therapy versus isotonic saline (0.9%) over 4 months.37 The \ufb01ndings on HRQOL be- active cycle of breathing technique and gravity-assisted drainage28 tween studies highlight the importance of suf\ufb01cient power to detect (Figure 2; see Figure 3 on the eAddenda for a detailed forest plot). changes in selected outcomes and the potential bene\ufb01t in using a The impact on generic and disease-speci\ufb01c HRQOL measures were bronchiectasis-speci\ufb01c QOL tool to assess clinical effect. also equivalent between techniques. Together with previous studies, these trial \ufb01ndings emphasise that no single airway clearance tech- Practically, physiotherapists prescribe isotonic saline or propose a nique results in superior physiological or patient-reported outcomes recommendation for hypertonic saline as adjuncts to airway clear- in the short term. ance techniques in this population and within these prescriptions, guidance for treatment order is included. While it is common for There is a large, long-term trial of airway clearance techniques hypertonic saline to be administered prior to or during airway that has implications for practice.29 A randomised controlled trial clearance therapy (practice informed by studies in cystic \ufb01brosis38), with 44 participants with stable bronchiectasis, whose disease isotonic saline is more frequently delivered in conjunction with severity ranged from mild to severe, found a reduction in the number airway clearance techniques; a combined approach reduces treat- of acute exacerbations, improvement in disease-speci\ufb01c and cough- ment burden.39 Physiotherapists also provide guidance to ensure related QOL and increase in sputum expectoration over a 12-month optimal medication delivery, including the type of nebuliser and period for those receiving slow expiration with an open glottis in a breathing technique. lateral posture twice daily compared with upper limb stretches. As the only study to date that has examined the effects of airway Pulmonary rehabilitation and physical activity clearance techniques over a period of . 3 months, this provides con\ufb01dence that undertaking a form of airway clearance therapy in the Pulmonary rehabilitation guidelines and bronchiectasis manage- long term leads to health bene\ufb01ts, including those linked to ment guidelines recommend the inclusion of adults with bronchiec- morbidity and mortality compared with no treatment. tasis within these programs as part of overall care.19,20,40 Pulmonary rehabilitation programs for adults with bronchiectasis are conducted To ensure effective implementation in the long term, adherence to in a mix of settings, including hospital or community outpatient the airway clearance techniques is critical. Adults with bronchiectasis programs, home-based or telerehabilitation, with the exercise struc- have an adherence rate of 41%; those with better physical function, ture focused on moderate-to-high intensity endurance and resistance with a greater burden of treatment and more severe respiratory exercise prescription, for a typical duration of 6 to 8 weeks. Co- symptoms are more likely to be adherent.30 Predictors of compliance interventions range across education, self-management, airway are older age and a strong belief in treatment necessity.31 Capturing clearance techniques and respiratory muscle training. A recent the patient\u2019s and clinician\u2019s voice, recent qualitative studies have Cochrane review of six studies demonstrated that in clinically stable shown that barriers to airway clearance techniques include lack of individuals, rehabilitation with or without co-interventions improved perceived bene\ufb01t, poor motivation, time commitment and lack of functional exercise tolerance re\ufb02ected in the incremental shuttle resources, while enablers are use of a tailored, personalised approach, walk distance (MD 87 m, 95% CI 43 to 132) and disease-speci\ufb01c perceived health and QOL bene\ufb01t, inclusion of self-management HRQOL (MD \u20139.6 points, \u201311.7 to \u20133.6) compared with usual care.41 strategies and working with the interdisciplinary team.32,33 While the magnitude of improvements were clinically signi\ufb01cant, they were not sustained in the long term. This review con\ufb01rms that In summary, current evidence for airway clearance techniques in this intervention, which was originally designed for and applied to bronchiectasis supports that undertaking a form of therapy yields those with COPD, offers a similar degree of bene\ufb01t in bronchiectasis. some clinical bene\ufb01t in the short and long term, but there is not This observation has been con\ufb01rmed in a propensity-matched necessarily a single, optimal or superior technique. There is signi\ufb01- observational study, with similar completion rates and improve- cant heterogeneity in disease severity, technique prescription, treat- ment in clinical outcomes observed in 426 patients (213 with bron- ment duration and outcomes between reported studies. Data for chiectasis, 213 with COPD).42 exacerbations and HRQOL are scant, with few long-term studies and a lack of consensus of key outcomes for airway clearance trials.34 De- The emergence of alternative rehabilitation models examines mographic and clinical factors appear to serve as enablers or barriers whether home-based programs offer similar results to those con- to adherence in the adult population; this information may be ducted within the hospital or community environment. Two recent applied in the clinical setting to guide practice. trials studied home-based pulmonary rehabilitation for people with bronchiectasis, with training consisting of walking or cycling or step Mucoactive agents training combined with resistance exercise.43,44 The pooled effects demonstrated similar improvement in functional exercise capacity for Mucoactive therapy has been recommended for people with bron- both home-based and hospital outpatient-based models of care chiectasis in cases where airway clearance techniques are ineffective20 or (Figure 4; see Figure 5 on the eAddenda for a detailed forest plot). Following the escalation in telerehabilitation, a randomised controlled equivalence trial demonstrated a similar degree of improvement in functional exercise capacity and HRQOL domains in","10 Lee: Physiotherapy management of bronchiectasis in adults Subgroup SMD (95% CI) While all programs provide adjunct management, consisting of Study Fixed education and\/or initial instruction or review of airway clearance therapy, there is no consistency between studies. It is therefore Hospital-based pulmonary rehabilitation dif\ufb01cult to determine the overall contribution of these co-treatments to the clinical outcomes. Ideally, prior to commencing a rehabilitation Lee 2014 program, airway clearance techniques are already prescribed. To date, commonly included education topics have varied from airway clear- Mandal 2012 ance therapy, bene\ufb01ts of exercise, breathing techniques, underlying lung pathology, inhaled medication, pacing and energy conservation, Newall 2005 nutrition, self-management, and exercise maintenance options.41 Recent updates of topics desired by patients in this setting Subtotal (Box 1)48\u201350 could be considered in future iterations of the education component of pulmonary rehabilitation. Home-based pulmonary rehabilitation De Jesus 2022 Despite the bene\ufb01ts of pulmonary rehabilitation, attendance rates Jose 2021 are low (22%), although it is indicated in 67% of the population.35 For Subtotal some adults with an adequate, individually-tailored physical activity routine with which they are engaged, a structured program may not be Pooled required. While physical activity levels are lower compared with healthy aged-matched peers,10 physical activity interventions in this population \u20134 \u20132 0 24 beyond pulmonary rehabilitation have not been examined. Patient- identi\ufb01ed barriers to exercise are embarrassment about symptoms, Favours usual care Favours rehabilitation breathlessness, fear of exacerbating symptoms or causing an exacerba- tion, and insuf\ufb01cient time.51 Not unexpectedly, enjoyable and motivating Figure 4. Differences in walking distance following pulmonary rehabilitation versus activities that are undertaken with a support network promote greater usual care in hospital-outpatient pulmonary rehabilitation and home-based pulmo- engagement. These factors warrant consideration to improve uptake and nary rehabilitation. Data are standardised mean difference (either 6-minute walk completion of pulmonary rehabilitation or engagement with physical distance or incremental shuttle walk distance). activity, with the necessary adaptations to accommodate the needs of individuals and comorbidities central to success. Oxygen therapy those undertaking telerehabilitation compared with centre-based Hypoxaemia may be present in people with bronchiectasis, rehabilitation in a subgroup of individuals with bronchiectasis.45 particularly during an acute exacerbation with reduced ventilation or Although trials are few, the \ufb01ndings to date outline that alternative air\ufb02ow obstruction and in those with severe disease and chronic gas ways of delivering pulmonary rehabilitation can be safely delivered, exchange abnormalities. In the absence of speci\ufb01c guidelines for appear clinically bene\ufb01cial and may overcome access issues. bronchiectasis, the recommendations for treating hypoxaemia are derived from position statements and guidelines for acute oxygen These positive effects were less evident when applied to in- therapy and domiciliary oxygen therapy.52,53 These recommendations dividuals following an acute exacerbation of bronchiectasis. guide empirical and pragmatic approaches in clinical practice to treat Following 14 days of antibiotic treatment, 48 patients were rando- acute and chronic hypoxaemia. Like other chronic lung diseases, mised to either outpatient exercise rehabilitation and education or physiotherapists optimise oxygen therapy in the acute or rehabilita- usual care for 6 weeks,46 with no clear between-group differences in tion environment to ensure adequate dosage and the most suitable functional exercise capacity, patient-reported outcomes or time to mode of oxygen delivery. next exacerbation. The \ufb01ndings contrast with studies of people with COPD, in whom pulmonary rehabilitation following discharge for an One of the more recent systems used is high-\ufb02ow nasal therapy acute exacerbation led to selected clinical improvement.47 The lack of (HFNT), a gas delivery system capable of providing heated and hu- bene\ufb01t may be attributable to natural recovery in this population. midi\ufb01ed air or supplemental oxygen via a heated pass-over humidi- Further examination related to the timing of prescribing this inter- \ufb01er and nasal cannulae. Options for air\ufb02ow commonly range from 20 vention in this clinical state is required. to 60 l\/min, while oxygen can be titrated to 60 to 100%. The rationale is that an increase in airway surface liquid volume under conditions Box 1. Education topics for people with bronchiectasis. From of 37\u000eC and 100% relative humidity enhances mucociliary clearance. Lee et al,50 Burge et al48 and Hester et al.69 Other physiological effects are a reduction in anatomical dead space, enhanced carbon dioxide washout, reduced work of breathing and \u000f Understanding the aetiology of the condition and prognosis, generation of positive end-expiratory pressure, which reduces dysp- including uncertainty noea and improves oxygenation.54 A short-term study highlighted that 3 hours of treatment per day with HFNT over 7 days improved \u000f Knowledge about self-management, including strategies for mucociliary transport in adults with bronchiectasis,55 suggesting an early recognition of exacerbations and developing support and adjunct role of humidi\ufb01ed oxygen therapy in improving sputum coping mechanisms clearance. A longer term, randomised controlled study of 45 people with bronchiectasis who produced daily sputum (. 5 ml) received \u000f Becoming an active partner in disease management domiciliary humidi\ufb01ed air at 20 to 25 l\/min via HFNT for a period of \u0015 \u000f Understanding of the physical, psychological and social 2 hours\/day for a 12-month period;54 fewer exacerbations and a better HRQOL were noted in those receiving this therapy compared impacts of living with the condition with usual care.56 These preliminary \ufb01ndings offer encouragement \u000f Medication and rescue packs towards the prescription of humidi\ufb01ed oxygen therapy, although \u000f Lifestyle factors of diet and sleep further studies across a broader range of disease severity are required \u000f Knowledge for carers to clarify the bene\ufb01ts. \u000f Airway clearance therapy - alternative options for techniques, Evidence is very limited regarding HFNT in those with acute ex- frequency of review and advice on cleaning of devices acerbations. An observational study of 15 patients with acute respi- \u000f Physical activity beyond pulmonary rehabilitation ratory failure secondary to an exacerbation of COPD with concurrent \u000f Treatment for extra-pulmonary conditions bronchiectasis applied HFNT and supplemental oxygen at 50 l\/min, \u000f Management anxiety, depression and symptom stigma \u000f End-of-life care and advance directives","Invited Topical Review 11 with the fraction of inspired oxygen titrated to maintain percuta- information within these settings, patients desire credible and neous saturations of 92%.57 Improvements in respiratory rate, gas trustworthy sources of information that are accessible outside the exchange parameters, symptoms and ease of expectoration were re- clinical environment,69 which include digital websites. When ported after 24 hours of treatment. With the contribution of bron- designed in partnership with consumers, such websites have a use- chiectasis to the underlying exacerbation being unclear, further trials fulness rating of 93% and offer an increased understanding of their are necessary to determine the clinical effect of this practice. disease and treatment options\u2014lending support for a \u2018patient as partner\u2019 role in developing such resources.69 Non-invasive ventilation The basis of self-management in bronchiectasis is derived from Non-invasive ventilation is applied to treat acute or chronic type II other respiratory conditions and is designed to assist patients in respiratory failure in those with bronchiectasis. Physiologically, non- understanding their condition, treatments and their bene\ufb01ts. A sys- invasive ventilation corrects acid-base balance, unloads respiratory tematic review and realist synthesis of self-management in adults muscles and improves symptoms. For those with acute respiratory with bronchiectasis included data from six studies consisting of 288 failure secondary to an acute exacerbation of bronchiectasis, feasi- patients,70 and highlighted three types of self-management pro- bility studies have demonstrated improvement in selected gas ex- grams: education and action planning; education and airway clear- change parameters and respiratory symptoms, with suggestions for a ance therapy; and education, exercise and airway clearance therapy. reduced need for intubation and duration of stay in intensive Components within each program are outlined in Table 1. These ap- care.58,59 Within their scope of practice, physiotherapists play a role proaches to self-management applied in a mix of study designs have in the provision of this treatment, including the practicalities of led to short-term improvements in self-ef\ufb01cacy and HRQOL, while identifying suitable candidates, establishing treatment and adjusting long-term outcomes are unknown.70 The success of self-management parameters to ensure patient comfort and clinical ef\ufb01cacy.60 This role comes from recognising that there is no \u2018one size \ufb01ts all\u2019, with indi- is likely to extend to the provision of airway clearance therapy to vidualised, \ufb02exible and responsive self-management plans needed by maximise secretion clearance. those living with bronchiectasis.49 The priorities of self-management program should be guided by an individual\u2019s current status and For those with chronic respiratory failure, domiciliary nasal non- knowledge for maximum gain, for example: education about the invasive ventilation has been associated with stabilisation of blood disease is less useful for a well-informed patient, but a focus on gas measurements and reduction in hospitalisation duration over a dealing with the psychosocial factors and stigma of bronchiectasis 12-month period and improvement in QOL.61,62 The role of the may be a higher priority. physiotherapist in domiciliary non-invasive ventilation services is less clear and may be dependent on the physiotherapists\u2019 clinical skill Management of comorbidities and con\ufb01dence in this environment. For people with bronchiectasis, comorbid conditions (including Respiratory muscle training anxiety and depression, incontinence, musculoskeletal pain and rhi- nosinusitis) are potentially amenable to physiotherapy. While exer- Respiratory muscle weakness has been observed in people with cise training alone had no effect on anxiety or depression,71 an bronchiectasis;63 this feature has been linked to dyspnoea and observational study showed that those with at least mild anxiety or impaired ef\ufb01ciency of coughing, which may limit secretion clearance. depression improved in their psychological symptoms following In the event of respiratory muscle weakness contributing to respi- pulmonary rehabilitation beyond the minimum important differ- ratory symptoms in a spontaneously breathing patient with bron- ence.72 The inclusion of an education session on psychological well- chiectasis, inspiratory muscle training is recommended.64 A mix of being in addition to endurance and strength exercise training is a approaches for this mode of intervention, including pressure necessity for improvement. threshold training of 30 to 70% maximal inspiratory pressures and airway clearance techniques (PEP therapy) were recently examined in Urinary incontinence affects 45 to 55% of adults with bronchiec- a systematic review.65 Pooled results revealed an improvement in tasis, with some reporting symptoms for a duration of 5 to 10 years, maximal inspiratory pressures (MD 6.08, 95% CI 138 to 10.77), but no with detrimental effects on HRQOL.73,74 While the ef\ufb01cacy of treat- change in maximal expiratory pressures (MD 2.04, 95% CI \u20133.33 to ment in bronchiectasis is unproven, various treatment options con- 7.31). Importantly, the effect of respiratory muscle training on sisting of urinary incontinence management plans with pelvic \ufb02oor patient-centred outcomes is mixed, with greater bene\ufb01ts evident in muscle strengthening, urge suppression, bladder retraining and toilet functional exercise tolerance and HRQOL with high-intensity scheduling have been prescribed. compared with low- to medium-intensity training protocols.66 A more recent study found similar improvements in maximal inspira- While musculoskeletal complications have been reported in those tory pressures, while interviews with individuals revealed a with bronchiectasis, only one study has examined the presence of perceived improvement in physical ability and health due to inspi- pain.75 Pain locations were chest, middle and low back pain (due to ratory muscle training.67 When combined with exercise training, coughing) and neck pain. Poor posture was evident, with an increased while there is no initial superiority in adding respiratory muscle frequency of postural malalignment (scoliosis and kyphosis) in those training immediately following treatment, the bene\ufb01ts of an exercise with bronchiectasis compared with healthy controls. At a minimum, program on exercise capacity were sustained at 3 months after the postural review and prescription of musculoskeletal interventions program\u2019s conclusion.68 While these studies hint at physiological designed to address the contributing factors to poor posture and pain bene\ufb01ts for respiratory muscle training, there is a lack of translational by physiotherapists are worthy of further study. effect illustrating the impact on activities limited by symptoms or synonymous with exercise intolerance. The bene\ufb01t in using this mode In some individuals with bronchiectasis, a united airways\u2019 disease of training may depend on the degree of hyperin\ufb02ation, degree of syndrome is apparent with a concurrent diagnosis of rhinosinusitis; it weakness and the optimal prescription, which has not yet been can affect approximately 62% of adult cases and has been linked to identi\ufb01ed. poorer HRQOL.76 There is paucity of evidence about treatment, which has included nasal saline irrigation, antibiotics, topical or systemic Education and self-management steroids, mucolytics, decongestants, antihistamines and surgery.77 It is recommended that physiotherapists managing individuals with these symptoms liaise closely with respiratory physicians and ear, nose and throat specialists to optimise treatment where possible. Patient education is a fundamental part of management, with Surgery and lung transplantation information and knowledge delivered in a variety of formats, including pulmonary rehabilitation, clinical consultations, patient Surgical management for those with bronchiectasis is preserved webinars\/seminars or in the inpatient setting. In addition to receiving for those with localised disease, with cohort studies demonstrating","12 Lee: Physiotherapy management of bronchiectasis in adults Table 1 Self-management program content and outcomes applied in bronchiectasis. Adapted from Tsang et al.70 Self-management Program components and content Outcomes programs \u000f Improved patient knowledge Education and action \u000f Interactive sessions by healthcare professionals with visual and auditory learning resources (including about symptoms and ability to planning handouts, posters and videos). recognise exacerbations Education and airway \u000f Education on provision of airway clearance therapy \u000f Improved self-ef\ufb01cacy and clearance techniques \u000f Educational topics: signs and symptoms, treatment goals, problem-solving and general health promotion health-related quality of life Education, exercise (nutrition) \u000f Improvement in self-ef\ufb01cacy and airway clearance \u000f Participants completed worksheets outlining usual symptoms, daily management of symptoms and and health-related quality of life techniques triggers for exacerbation \u000f Improved health-related quality \u000f Weekly action planning and goal setting of life and exercise capacity \u000f Action plans served as treatment seeking support tool to set weekly goals \u000f Awareness of psychosocial factors (relationships with people and levels of anxiety and depression) \u000f Provision of educational materials to encourage independence in further aspects of their care \u000f Knowledge regarding how to deal with the social and psychological aspects of bronchiectasis \u000f Education regarding differentiation in symptoms between stable and exacerbation phase \u000f Role of family members in providing emotional support and advice \u000f Provision of structured program of exercise, education and psychosocial support, instructing in problem- solving skills, behavioural management and decision-making \u000f Individual tailoring of exercise program based on patient tolerance to accommodate disease severity \u000f Instruction in breathing techniques (pursed lips breathing, exhalation on effort and forward leaning position) and examples of airway clearance techniques improvement in HRQOL following surgery.78 The focus of physio- In the meantime, use of recently developed bronchiectasis-speci\ufb01c therapy interventions in this clinical scenario involves education tools (Quality of Life-Bronchiectasis and Bronchiectasis Health regarding postoperative management, respiratory care and early Questionnaire)84 are applicable. Clarity on the effects of hypertonic mobilisation to facilitate discharge planning.79 and isotonic saline inhalation and humidi\ufb01ed HFNT alone or in conjunction with airway clearance therapy would further inform While surgery involving removal of a portion of the lung is rarely clinical practice. A growth in studies of longer duration (12 months) performed, lung transplantation is undertaken for life-threatening, will enable more robust interpretation of the effects of airway advanced disease. While only making up a small portion of lung clearance techniques on exacerbation rates, hospitalisations, HRQOL, transplants, physiotherapists form a pivotal role in optimising man- and direct and indirect measures of cost-effectiveness, which are all agement of those with bronchiectasis awaiting a lung transplant and outcomes re\ufb02ecting morbidity and impacting on mortality and the following the procedure. Physiotherapists prescribe pre-transplant healthcare system. exercise programs with a mix of endurance and resistance training, while maintaining optimal airway clearance therapy. Although based While pulmonary rehabilitation in a stable clinical state causes on uncontrolled studies, a systematic review demonstrated that this short-term bene\ufb01t, it is unclear how to retain these improvements. intervention package improved exercise capacity, while effects on Greater examination of the role of exercise training and mobilisation HRQOL were mixed.80 In addition, physiotherapists are involved as in the inpatient setting and of pulmonary rehabilitation following part of the multidisciplinary team in providing education and support hospitalisation for an acute exacerbation is needed to guide clinical to address the stress and expectations of the waiting period for practice. Home-based rehabilitation and telerehabilitation offer im- transplant, address energy conservation and manage infection control provements, but how to accurately assess exercise capacity and pe- for some individuals.81 Post-transplant rehabilitation programs have ripheral muscle strength within these settings to maximise access a similar focus to pre-transplantation, with education and exercise requires attention. Identi\ufb01cation of common physical activity in- training across a variable duration.80 Despite the absence of terventions undertaken by this population is the \ufb01rst step towards comparator groups or speci\ufb01c data for those with bronchiectasis, establishing their impact, as not all individuals will engage with bene\ufb01ts from clinical trials of mixed diagnoses have demonstrated pulmonary rehabilitation. Understanding patient factors in\ufb02uencing improvements in clinical outcomes.80 To maintain improvements, adherence to and engagement with physical activity and exercise recommendations for ongoing exercise, with identi\ufb01cation of enjoy- interventions will enable greater personalised tailoring of therapies. able physical activities is important in long-term management. Clarity on the role of education and self-management in bron- Future directions for research and practice chiectasis is required, including content and availability of resources, modes of delivery, evaluation of impact and translation into daily Bronchiectasis is no longer considered an orphan disease and the practice to facilitate behavioural change. Given the mix of comor- growth of international registries in recent years has energised op- bidities in bronchiectasis that may be amenable to physiotherapy portunities for future research. Most trials of airway clearance ther- interventions, research should gain a greater understanding of the apy have focused on single treatment techniques and short-term breadth and clinical impact of these co-existing conditions, establish outcomes; given combinations of techniques are frequently applied in reliable and valid measurement tools, and explore treatment options clinical practice,39 there is a need for pragmatic studies of individuals within the scope of physiotherapy practice. Finally, continuation of diagnosed with comorbidities examining the clinical effect of com- translation of \ufb01ndings into practice is critical to implementing bined approaches in a stable clinical state and during an acute evidence-based physiotherapy care for adults with bronchiectasis. exacerbation. To date, all studies have applied models of physio- therapy treatment that were delivered in person. Given the chal- eAddenda: Figures 3 and 5 can be found online at https:\/\/doi.org\/ lenges to accessing physiotherapy,82 exploration of tele- 10.1016\/j.jphys.2022.11.013 physiotherapy that includes delivery of airway clearance therapy should be explored. Of importance in all studies is consumer input Ethics approval: Not applicable. into study design and outcomes for evaluation. The recent recogni- Competing interests: Nil. tion of the need for a core set of outcomes in physiotherapy trials and Source(s) of support: Nil. their subsequent identi\ufb01cation83 will further inform which patient- Acknowledgements: Nil. reported and patient-experience outcomes should be incorporated. Provenance: Invited. Peer reviewed. Correspondence: Annemarie L Lee, Department of Physiotherapy, Monash University, Melbourne, Australia. Email: Annemarie.Lee@ monash.edu","Invited Topical Review 13 References 31. McCullough AR, Tunney MM, Elborn JS, Bradley JM, Hughes CM. Predictors of adherence to treatment in bronchiectasis. Respir Med. 2015;109:838\u2013845. 1. Chang AB, Fortescue R, Grimwood K, Alexopoulou E, Bell L, Boyd J, et al. European Respiratory Society guidelines for the management of children and adolescents 32. Franks LJ, Walsh JR, Hall K, Adsett JA, Morris NR. Patient perspectives of airway with bronchiectasis. Eur Respir J. 2021;58:2002990. clearance techniques in bronchiectasis. Physiother Theory Pract. Published online ahead of print, Sept 20 2022; https:\/\/doi.org\/10.1080\/09593985.2022.2126741. 2. Hill AT, Sullivan AL, Chalmers JD, De Soyza A, Elborn JS, Floto RA, et al. British Thoracic Society guideline for bronchiectasis in adults. Thorax. 2019;74(Suppl 1):1\u2013 33. Lawton K, Royals K, Carson-Chahhoud K, Smith B, Veale A. Management: a health 69. professional perspective. Respirology. 2019;24(Suppl 1):101. 3. Cole PJ. In\ufb02ammation: a two-edged sword-the model of bronchiectasis. Eur J Respir 34. Welsh EJ, Evans DJ, Fowler SJ, Spencer S. Interventions for bronchiectasis: an Dis Suppl. 1986;147:6\u201315. overview of Cochrane systematic reviews. Cochrane Database Syst Rev. 2015;7:CD010337. 4. Contarini M, Finch S, Chalmers JC. Bronchiectasis: a case-based approach to investigation and management. Eur Respir Rev. 2018;27:180016. 35. Visser SK, Bye PTP, Fox GJ, Burr LD, Chang AB, Holmes-Liew C-L, et al. Management of Australian adults with bronchiectasis in tertiary care: evidence-based or access- 5. Seitz AE, Olivier KN, Adjemian J, Holland SM, Prevots DR. Trends in bronchiectasis driven? Lung. 2019;197:803\u2013810. among medicare bene\ufb01ciaries in the United States, 2000 to 2007. Chest. 2012;142:432\u2013439. 36. Tarrant BJ, Le Maitre C, Romero L, Steward R, Button BM, Thompson BR, et al. Mucoactive agents for chronic, non-cystic \ufb01brosis lung disease: a systematic re- 6. Quint JK, Millett ERC, Joshi M, Navaratnam V, Thomas SL, Hurst JR, et al. Changes in view and meta-analysis. Respirology. 2017;22:1084\u20131092. the incidence, prevalence and mortality of bronchiectasis in the UK from 2004 to 2013: a population-based cohort study. Eur Respir J. 2016;47:186\u2013193. 37. Paff T, Daniels JMA, Weerskink EJ, Lutter R, Noordegraaf AV, Haarman EG. A randomised controlled trial on the effect of inhaled hypertonic saline on quality 7. McDonnell MJ, Aliberti S, Goeminne PC, Restrepo MI, Finch S, Pesci A, et al. of life in primary ciliary dyskinesia. Eur Respir J. 2017;49:1601770. Comorbidities and the risk of mortality in patients with bronchiectasis: an inter- national multicentre cohort study. Lancet Respir Med. 2016;4:969\u2013979. 38. Elkins M, Dentice R. Timing of hypertonic saline inhalation for cystic \ufb01brosis. Cochrane Database Syst Rev. 2020;2:CD008816. 8. Hill AT, Haworth CS, Aliberti S, Barker A, Blasi F, Boersma W, et al. Pulmonary exacerbation in adults with bronchiectasis: a consensus de\ufb01nition for clinical 39. Lee AL, Baenziger S, Louey A, Jennings S, Solin P, Hoy R. A review of physiotherapy research. Eur Respir J. 2017;49:1700051. practice for people with bronchiectasis. Eur Respir J Open Res. 2020;7. https:\/\/doi. org\/10.1183\/23120541.00569-2020. 9. Quint JK, Smith MP. Paediatric and adult bronchiectasis: diagnosis, disease burden and prognosis. Respirology. 2019;24:413\u2013422. 40. Alison JA, McKeough ZJ, Johnston K, McNamara RJ, Spencer LM, Jenkins SC, et al. Australian and New Zealand pulmonary rehabiliation guidelines. Respirology. 10. De Camargo AA, Boldorini JC, Holland AE, de Castro RA, de Cordoba Lanza F, 2017;22:800\u2013819. Athanazio RA, et al. Determinants of peripheral muscle strength and activity in daily life in people with bronchiectasis. Phys Ther. 2018;98:153\u2013161. 41. Lee AL, Gordon C, Osadnik CR. Exercise training for people with bronchiectasis. Cochrane Database Syst Rev. 2021;3:CD 013110. 11. Oliveira C, Olveira G, Gaspar I, Dorado A, Cruz I, Soriguer F, et al. Depression and anxiety symptoms in bronchiectasis: associations with health-related quality of 42. Patel S, Cole AD, Nolan CM, Barker RE, Jones SE, Kon S, et al. Pulmonary rehabili- life. Qual Life Res. 2013;22:597\u2013605. tation in bronchiectasis: a propensity-matched study. Eur Respir J. 2019;53: 1801264. 12. Ringshausen FC, Sokol P, Barten G, De Roux A, Diel R, Konwert S, et al. Quality of life is associated with increased symptom burden and markers of disease severity: 43. de Jes\u00fas S, Pacheco VA, Morales AV, Rodriguez AMM, Garcia RR, Arnedillo- data from the German Bronchiectasis Registry PROGNOSIS. Eur Respir J. Mu\u00f1oz A. Exercise capacity and physical activity in non-cystic \ufb01brosis bronchiec- 2018;52:PA2670. tasis after a pulmonary rehabilitation home-based programme: a randomised controlled trial. Int J Environ Res Public Health. 2022;19:11039. 13. Dudgeon EK, Crichton M, Chalmers JD. \u201cThe missing ingredient\u201d: the patient perspective of health related quality of life in bronchiectasis: a qualitative study. 44. Jos\u00e9 A, Holland AE, Selman JPR, de Camargo CO, Fonseca DS, Athanazio RA, et al. BMC Pulm Med. 2018;18:81. Home-based pulmonary rehabilitation in people with bronchiectasis: a rando- mised controlled trial. Eur Respir J Open Res. 2021;7:00021\u20132021. 14. Goeminne PC, Hernandez F, Diel R, Filonenko A, Hughes R, Juelich F, et al. The economic burden of bronchiectasis: known and unknown - a systematic review. 45. Cox NS, McDonald CF, Mahal A, Alison JA, Wootton R, Hill CJ, et al. Tele- BMC Pulm Med. 2019;19:3388. rehabilitation for chronic respiratory disease: a randomised controlled equivalence trial. Thorax. 2022;77:643\u2013651. 15. Chalmers JD, Goeminne PC, Aliberti S, McDonnell MJ, Lonni S, Davidson J, et al. The bronchiectasis severity index. An International derivation and validation study. Am 46. Chalmers JD, Crichton ML, Brady G, Finch S, Lonergan M, Fardon TC. Pulmonary J Respir Crit Care Med. 2014;189:576\u2013585. rehabilitation after exacerbation of bronchiectasis: a pilot randomised controlled trial. BMC Pulm Med. 2019;17:85. 16. Chalmers JD, Aliberti S, Filonenko A, Shteinberg M, Goeminne PC, Hill AT, et al. Characterisation of the \u201cFrequent Exacerbator Phenotype\u201d in bronchiectasis. Am J 47. Puhan M, Scharplatz M, Troosters T, Haydn Walters E, Steurer J. Pulmonary reha- Respir Crit Care Med. 2018;197:1410\u20131420. bilitation following exacerbations of chronic obstructive pulmonary disease. Cochrane Database Syst Rev. 2016;2016:CD0053005. 17. Polverino E, Dimakou K, Hurst J, Martinez-Garcia MA, Miravitlles M, Paggiaro P, et al. The overlap between bronchiectasis and chronic airway disease: state of the 48. Burge AT, Lee A, Nicholes M, Purcell S, Miller B, Norris N, et al. Advance care art and future directions. Eur Respir J. 2018;52:1800328. planning education in pulmonary rehabilitation: a qualitative study exploring participant perspectives. Palliat Med. 2013;27:508\u2013515. 18. Finch S, McDonnell MJ, Abo-Leyah H, Aliberti S, Chalmers JD. A comprehensive analysis of the impact of Pseudomonas aeruginosa conolisation on prognosis in 49. Kelly CA, Tsang A, Lynes D, Spencer S. \u2018It\u2019s not one size \ufb01ts all\u2019: a qualitative study of adult bronchiectasis. Ann Am Thorac Soc. 2015;12:1602\u20131611. patients\u2019 and healthcare professionals\u2019 views of self-management for bronchiec- tasis. BMJ Open Respir Res. 2021;8:e000862. 19. Chang AB, Bell SC, Torzillo PJ, King PT, Maguire G, Byrnes CA, et al. Chronic sup- purative lung disease and bronchiectasis in children and adults in Australia and 50. Lee AL, Smith R, Burr L, Chang AB, Holmes-Liew C-L, King P, et al. \\\"Teach me how to New Zealand. Thoracic Society of Australia and New Zealand guidelines. Med J Aust. look after myself\u2019: What people with bronchiectasis want from education in a 2015;202:21\u201323. pulmonary rehabilitation setting. Clin Respir J. Published online ahead of print, Nov 7 2022; https:\/\/doi:10.1111\/crj.13563. 20. Polverino E, Goeminne PC, O\u2019Donnell MJ, Aliberti S, Marshall SE, Loebinger MR, et al. European Respiratory Society guidelines for the management of adult 51. Royle H, Kelly C. \\\"The likes of me running and walking? No change\u2019: Exploring the bronchiectasis. Eur Respir J. 2017;50:1700629. perceptions of adult patients with bronchiectasis towards exercise. Chron Illn. Published online ahead of print, May 23 2022: https:\/\/doi: 10.1177\/174239532211 21. Phillips J, Lee AL, Pope R, Hing W. Physiotherapists\u2019 use of airway clearance tech- 08223 17423953221108223. niques during an acute exacerbation of bronchiectasis: a survey study. Arch Physiother. 2021;11:3. 52. Barrett A, Beasley R, Buchan C, Chien J, Farah CS, King G, et al. Thoracic Society of Australia and New Zealand position statement on acute oxygen use in adults: 22. Webb EM, Holland AE, Chang AB, Burr L, Holmes-Liew C-L, Jayaram L, et al. \u2019Swimming between the \ufb02ags\u2019. Respirology. 2022;27:262\u2013276. Physiotherapy practice for adults with bronchiectasis: Australian bronchiectasis registry data. Respirology. 2022;27:TO034. 53. McDonald CF, Whyte K, Jenkins S, Serginson J, Frith P. Clinical practice guideline on adult domiciliary oxygen therapy: executive summary from the Thoracic Society of 23. Lee AL, Burge AT, Holland AE. Airway clearance techniques in bronchiectasis. Australia and New Zealand. Respirology. 2016;21:76\u201378. Cochrane Database Syst Rev. 2015;11:CD008351. 54. Rea H, McAuley S, Jayaram L, Garrett J, Hockey H, Storey L, et al. The clinical utility 24. Lee AL, Burge AT, Holland AE. Positive expiratory pressure therapy versus other of long-term humidi\ufb01cation therapy in chronic airway disease. Respir Med. airway clearance techniques in bronchiectasis. Cochrane Database Syst Rev. 2010;104:525\u2013533. 2017;9:CD011699. 55. Hasani A, Chapman TH, McCool D, Smith RE, Dilworth JP, Agnew JE. Domiciliary 25. Phillips J, Lee AL, Pope R, Hing W. Effect of airway clearance techniques in patients humidi\ufb01cation improves lung mucociliary clearance in patients with bronchiec- experiencing an acute exacerbation of bronchiectasis: a systematic review. Physi- tasis. Chron Respir Dis. 2008;5:81\u201386. other Theory Pract. 2019;36:1300\u20131315. 56. Good WR, Garrett J, Hockey HUP, Jayaram L, Wong C, Rea H. The role of high-\ufb02ow 26. Livnat G, Yaari N, Stein N, Bentur L, Hanna M, Harel M, et al. 4-week daily airway nasal therapy in bronchiectasis: a post hoc analysis. Eur Respir J Open Res. clearance using oscillating positive-end expiratory pressure versus autogenic 2021;7:00711\u20132020. drainage in bronchiectasis patients: a randomised controlled trial. Eur Respir J Open Res. 2021;7:00426\u20132021. 57. Crimi C, Noto A, Cortegiani A, Campisi R, Hef\ufb02er E, Gregoretti C, et al. High \ufb02ow nasal therapy use in patients with acute exacerbation of COPD and bronchiectasis: 27. Santos MD, Milross MA, McKenzie DK, Alison JA. Bubble-positive expiratory a feasibility study. J COPD. 2020;17:184\u2013190. pressure device and sputum clearance in bronchiectasis: a randomised cross-over study. Physiother Res Int. 2019;25:e1836. 58. Phua J, Ang YL, See KC, Mukhopadhyay A, Santiago EA, Dela Pena EG, et al. Noninvasive and invasive ventilation in acute respiratory failure associated with 28. zmezog\u0015lu B, Altiay G, O} zdemir L, Tuna H, Su}t N. The ef\ufb01cacy of \ufb02utter and active bronchiectasis. Intensive Care Med. 2010;36:638\u2013647. cycle of breathing techniques in patients with bronchiectasis: a prospective, randomised, comparative study. Turk Thorac J. 2018;19:103\u2013109. 59. Sadigov A, Bakhshaliyeva K, Ibrahimov T, Ismailzadeh J. Noninvasive ventilation in acute respiratory failure caused by non-cystic \ufb01brosis bronchiectasis and chronic 29. Mu\u00f1oz G, de Gracia J, Bux\u00f3 M, Alvarex A, Vendrell M. Long-term bene\ufb01ts of airway obstructive pulmonary disease. Eur Respir J. 2015;46:PA4808. clearance in bronchiectasis: a randomised placebo-controlled trial. Eur Respir J. 2018;51:1701926. 60. Piper AJ, Moran FM. Non-invasve ventilation and the physiotherapist: current state and future trends. Phys Ther Rev. 2006;11:37\u201343. 30. McCullough AR, Tunney MM, Quittner AL, Elborn JS, Bradley JM, Hughes CM, et al. Treatment adherence and health outcomes in patients with bronchiectsis. BMC 61. Simonds AK, Elliott MW. Outcome of domiciliary nasal intermittent positive Pulm Med. 2014;14:107. pressure ventilation in restrictive and obstructive disorders. Thorax. 1995;50:604\u2013 609.","14 Lee: Physiotherapy management of bronchiectasis in adults 62. Gacouin A, Desrues B, L\u00e9na H, Quinquenel ML, Dassonville J, Dalaval P. Long-term 73. Duignan N, McDonnell MJ, Mokoka MC, Rutherford RM. High prevalence of stress nasal intermitten positive pressure ventilation (NIPPV) in sixteen consecutive urinary incontinence in adult patients with bronchiectasis. Ir Med J. patients with bronchiectasis: a retrospective study. Eur Respir J. 1996;9:1246\u20131250. 2016;109:440. 63. Moran FM, Piper AJ, Elborn JS, Bradley JM. Respiratory muscle pressures in non-CF 74. Prys-Pricard CO, Niven R. Urinary incontinence in patients with bronchiectasis. Eur bronchiectasis: repeatability and reliability. Chron Respir Dis. 2010;7:165\u2013171. Respir J. 2006;27:866\u2013867. 64. Bott J, Blumenthal S, Buxont M, Ellum S, Falconer C, Garrod R, et al. Guidelines for 75. Vardar-Yagli N, Seglam M, Calik-Kutukcu E, Inal-Ince D, Arikan H, Coplu L. the physiotherapy management of the adult, medical, spontaneously breathing Increased pain sensitivity, postural abnormalities, and functional balance impair- patient. Thorax. 2009;64(Suppl 1):i1\u2013i51. ment in obstructive lung disease compared to healthy subjects. Heart Lung. 2019;48:351\u2013355. 65. Martin-Valero R, Jimenez-Cebrian AM, Moral-Munoz JA, de-la-Casa-Almeida M, Rodriguez-Huguet M, Casuso-Holgado MJ. The ef\ufb01cacy of therapeutic respiratory 76. Handley E, Nicolson CHH, Hew M, Lee AL. Prevalence and clinical implications of muscle training interventions in people with bronchiectasis: a systematic review chronic rhinosinusitis in people with bronchiectasis: a systematic review. J Allergy and meta-analysis. J Clin Med. 2020;9:1017. Clin Immunol Pract. 2019;7:2004\u20132012. 66. Ozalp O, Inal-Ince D, Cakmak A, Calik-Kutukcu E, Saglam M, Savci S, et al. High- 77. Brunner JP, Riley CA, McCoul ED. A systematic review of the treatment of chronic intensity inspiratory muscle training in bronchiectasis: a randomised controlled rhinosinusitis in adults with primary ciliary dyskinesia. Sinusitis. 2017;2:1\u20137. trial. Respirology. 2019;24:246\u2013253. 78. Vallilo CC, Terra RM, De Albuquerque ALP, Suesada MM, Mariani AW, Salge JM, et al. 67. McCreery JL, Mackintosh KA, Mills-Bennett R, McNarry MA. The effect of a high- Lung resection improves the quality of life of patients with symptomatic bron- intensity PrO2Fit inspiratory muscle training intervention on physciological and chiectasis. Ann Thorac Surg. 2014;98:1034\u20131041. psychological health in adults with bronchiectasis: a mixed-methods study. Int J Environ Res Public Health. 2021;18:3051. 79. Baddeley RA. Physiotherapy for enhanced recovery in thoracic surgery. J Thorac Dis. 2016;8(Suppl 1):S107\u2013S110. 68. Newall C, Stockley RA, Hill SL. Exercise training and inspiratory muscle training in patients with bronchiectasis. Thorax. 2005;60:943\u2013948. 80. Hume E, Ward L, Wilkinson M, Mani\ufb01eld J, Clark S, Vogiatzis I. Exercise training for lung transplant candidates and recipients: a systematic review. Eur Respir Rev. 69. Hester KLM, Newton J, Rapley T, Ryan V, De Soyza A. Information and education 2020;29:200053. provision in bronchiectasis: co-development and evaluation of a novel patient- driven resource in a digital era. Eur Respir J. 2018;51:1702402. 81. Wickerson L, Rozenberg D, Janaudis-Ferreira T, Deliva R, Lo V, Beauchamp G, et al. Physical rehabilitation for lung transplant candidates and recipients: an evidence- 70. Tsang A, Lynes D, McKenzie H, Spencer S, Kelly C. Self-management programmes informed clinical approach. World J Transplant. 2016;6:517\u2013531. for adult patients with bronchiectasis: a systematic review and realist synthesis. Disabil Rehabil. 2021;44(23):6939\u20136948. 82. Cooper L, Johnston K, Williams M. Airway clearance services (ACSs) in Australia for adults with chronic lung conditions: scoping review of publicly available web- 71. Lee AL, Hill CJ, Cecins N, Jenkins S, McDonald CF, Burge AT, et al. The short and long based information. BMC Health Ser Res. 2019;19:808. term effects of exercise training in non-cystic \ufb01brosis bronchiectasis-a randomised controlled trial. Respir Res. 2014;15:44. 83. Hamzeh H, Spencer S, Kelly C. Development of a core outcome set and outcome measurement set for physiotherapy trials in adults with bronchiectasis (COS- 72. Wynne SC, Patel S, Barker RE, Jones SE, Walsh JA, Kon S, et al. Anxiety and PHyBE study): a protocol. PLoS One. 2022;17:e0263695. depression in bronchiectasis: Response to pulmonary rehabilitation and minimal clinically important difference of the Hospital Anxiety and Depression Scale. Chron 84. McLeese RH, Spinou A, Alfahl Z, Tsagris M, Elborn JS, Chalmers JD, et al. Psycho- Respir Dis. 2020;17:1\u20139. metrics of health-related quality of life questionnaires in bronchiectasis: a sys- tematic review and meta-analysis. Eur Respir J. 2021;2021(58):2100025.","Journal of Physiotherapy 69 (2023) 61\u201364 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: Deriving latent trajectories in health research Longitudinal data from repeated measures obtained from an in- understanding latent class approaches, as LCGA and LGMM can be dividual at successive time points is essential for understanding seen as extensions of this approach. Box 1 outlines the key elements changes in health phenomena. Longitudinal data can re\ufb02ect \ufb02uctua- of each approach and Figure 1 provides a graphical description of tions over time (such as variations in symptoms of a long-term health each. The critical assumption pertaining to the choice of a mixed condition) or more systematic changes (such as response to a treat- model approach is that individual variations (random effects) arise ment). The time frame for collection of longitudinal data can range from a multivariate, normal distribution. Therefore, if it is reasonable from short-term (such as heart rate change during exercise testing) to to assume that all individuals within a population follow a similar long-term (such as age-related growth or development). A pattern of underlying trajectory (eg, increase in height over childhood), with change over time is often termed a trajectory. variation between individuals suf\ufb01ciently captured by normally distributed random effects, then mixed models are an appropriate Individual level longitudinal data are increasingly becoming choice. However, if there is potential for very different trajectory available across biological, psychosocial and behavioural domains. shapes to exist within the population (eg, clinical course of muscu- Many techniques are available for longitudinal data analysis; loskeletal pain), then latent trajectory approaches may be more choosing between them depends on the research questions alongside appropriate, as these allow estimation of trajectories unconstrained underlying theories about patterns of change in the reference pop- by assumptions of normally distributed variation in change over time. ulation. An appreciation of the distinction between \u2018within-person\u2019 and \u2018between-person\u2019 variation is essential for formulating research Both LCGA and LGMM are approaches whereby subgroups are questions and analysing longitudinal data. Fundamentally, most identi\ufb01ed from the data by performing a latent class classi\ufb01cation on research questions pertain to the estimation of within-person change individuals\u2019 \u2018growth\u2019 parameters (ie, the coef\ufb01cients for time, such as (\u2018How do people change over time?\u2019) or between-person variation in initial value, slope and curvature). While both approaches subgroup within-person change (\u2018Are there differences between people in how individuals according to similar patterns of change over time, LCGA they change over time?\u2019). does not model further variability within each subgroup. This is somewhat analogous to collapsing a continuous variable, such as age, Latent trajectory analysis has increasingly become popular in whereby all people aged 18 to 25 years might be categorised by one health research, including physiotherapy. A previous Research Note value representing \u2018young adults\u2019 for analysis. In contrast, LGMMs reviewed cross-sectional latent class analysis and is a helpful estimate further individual variance around the mean growth pa- accompaniment to this one.1 Latent class analysis estimates distinct rameters within each subgroup. The relative appropriateness of the patterns of within-person variability in a population. In latent class two methods is widely debated.9 The term \u2018mixture\u2019 in LGMM arises analysis, people are classi\ufb01ed into distinct subgroups (classes) ac- from the assumption that the observed growth parameters arise cording to their pro\ufb01les on multiple variables. In the context of lon- from a mixture of two or more truly discrete subgroups.10 On the gitudinal data, people are classi\ufb01ed into subgroups based on repeated other hand, LCGA is described as summarising a non-normal measures over time, and it is the patterns of change over time that are continuous population distribution of trajectories, such that a sub- the \u2018latent\u2019 construct estimated. In health research, latent trajectory group is a collection of individuals following approximately the same analysis has helped to identify distinct patterns of clinical course of trajectory, like contour lines on a topographic map, rather than various health conditions,2,3 described the development or evolution assuming that the subgroups represent distinct populations.11 LCGA of health conditions over the life-course4 and described patterns of is often preferred purely due to the extra computational dif\ufb01culties of change in health-related behaviours.5 Furthermore, latent trajectory LGMM.12 Because LGMM allows variation within trajectory sub- subgroups that are identi\ufb01ed can be: pro\ufb01led on various factors of groups and LCGA does not, the optimal number of identi\ufb01ed sub- interest,5 considered as risk factors for future outcomes,6 outcomes groups is usually fewer when using LGMM compared with LCGA. themselves7 or related to trajectories of other health outcomes.8 Ideally, both LCGA and LGMM models should be \ufb01tted to the data and Latent trajectory analysis encompasses a range of techniques and compared using the methods described in the next section.9 Lastly, can be implemented in several software packages. This Research Note the third type of latent trajectory analysis \u2013 L-LCA (also called gives a non-technical review of latent trajectory analysis methods, repeated measures latent class analysis) \u2013 performs subgrouping provides guidance on key considerations for their use and discusses using the repeated measures of the health phenomenon themselves, clinical implications of latent class trajectory models. treating the data as if it were collected cross-sectionally rather than analysing it as longitudinal data accounting for time ordering. Approaches to estimating latent trajectories Although the time ordering of data is ignored, L-LCA has the advantage of not forcing any parametric relationship between the Three common approaches to estimating latent trajectories within repeated measures and time.13 This approach is useful when there a population are latent class growth analysis (LCGA), latent growth may be complex patterns of change, such as the clinical course of mixture modelling (LGMM) and longitudinal latent class analysis (L- long-term conditions,2 or there are repeated measures over time LCA). However, another widely used approach to modelling changes pertaining to a number of different facets of a construct, such as pain in a health factor over time is the mixed model approach. An impacts.14 understanding of the mixed model approach is helpful for https:\/\/doi.org\/10.1016\/j.jphys.2022.11.001 1836-9553\/\u00a9 2022 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 Box 1. Common approaches to trajectory modelling in health research. Mixed Models \u000f also called multilevel, hierarchical, random-effects or random coefficient models \u000f the repeated measures of the health outcome of interest is the dependent variable \u000f one or more time parameters are included as independent variables, to capture average linear or nonlinear change as \u2018fixed effects\u2019 of time \u000f \u2018random effects\u2019 for time parametersa can be added to allow individual variation around the average (fixed) effects for time \u000f further variation between people can be captured by adding between-person factors and including interaction terms with time (eg, different trajectories for males versus females). For this reason, mixed models are considered a variable-centred approach, as opposed to latent class techniques below that do not rely on other measured variables, and are hence considered \u2018unsupervised\u2019, person-centred approaches (Figure 1A). Latent Growth Mixture Modelling (LGMM) \u000f extension to mixed models, where the assumption is that there is a \u2018mixture\u2019 of \u2018latent\u2019 populations \u000f two or more latent subgroups are estimated with each latent subgroup having unique time parameters, representing different patterns of initial status and change over time \u000f Within each latent subgroup, further variation in time parameters between individuals is estimated (ie, individuals in the same subgroup share similar, but not exactly the same, trajectory shapes) (Figure 1B). Latent Class Growth Analysis (LCGA) \u000f also called group-based trajectory modelling \u000f like LGMM, two or more latent subgroups are estimated, each having unique time parameters representing different patterns of initial status and change over time \u000f No variation in time parameters between individuals within each latent subgroup is assumed (ie, individuals in the same subgroup follow exactly the same trajectory) (Figure 1C). Longitudinal Latent Class Analysis (L-LCA) \u000f also called repeated-measures latent class analysis \u000f The repeated measures of the health outcome of interest are the input variables that are subgrouped, rather than parameters for time (Figure 1D). a Footnote: One important distinction is that mixed models can be similarly estimated in a structural equation framework (then termed \u2018latent growth curve models\u2019 or \u2018latent curve analysis\u2019). In this context, the random effects for the intercept, slope etc. are termed \u2018latent\u2019 (unobserved) variables, but these are not the same thing as the latent subgroups estimated in repeated-measures latent class analysis, latent class growth analysis, or latent growth mixture modelling. For a good review of how mixed (multilevel) models and structural equation latent growth curve models are equivalent, see Curran.20 Determining the latent trajectory shape and number subgroup is helpful to assess how well the trajectory groups capture individual variation. Meaningful differences between trajectory In addition to the choice of method, another challenge in estimating groups in terms of pre-existing characteristics, clinical examination latent trajectories is deciding upon the number of trajectory subgroups \ufb01ndings, subsequent outcomes, response to treatment, or relation- and their shapes. Useful guides to the modelling strategies for LGMM, ship between the trajectories and other outcomes provide support for LCGA and L-LCA, along with code examples for software packages, are the usefulness of identi\ufb01ed latent trajectory groups.11 provided by Herle et al13 and Lennon et al.15 Typically, a series of models are \ufb01tted, starting with a single group model and increasing in number Other considerations for latent class trajectory methods until certain thresholds are reached, in relation to measures of model \ufb01t and the size of the smallest subgroup. For LCGA and LGMMs, the Trajectory subgroups identi\ufb01ed by latent class trajectory analysis optimal shape of the trajectories is usually evaluated by allowing for the are frequently related to individual or social factors, or outcomes most complex shape possible for each trajectory, given the number of that occur at a later point in time; such relationships can be time points. To avoid over\ufb01tting, at least three time points are needed quanti\ufb01ed using one-step or three-step approaches. In the one-step for linear slopes and the minimum is four for quadratic terms (note: approach, the latent class model used to estimate the trajectory alternative shapes can be modelled by other functions of time such as groups is extended to incorporate covariates of interest. In contrast, piecewise linear functions). After the optimal number of trajectory the three-step approach involves estimating the trajectory groups, subgroups has been tentatively decided, attention can be given to allocating individuals to the trajectory groups for which they have simplifying the time functions for each trajectory where possible (eg, the highest probability of membership based on their observed \ufb01rst allowing for quadratic or cubic terms, then testing whether model data, then using the trajectory groups in further analyses either as a \ufb01t substantially worsens by only including simple linear terms). The predictor or an outcome. Ideally, further analyses account for un- selection of the optimal latent class trajectory model is supported by certainty of subgroup membership when the three-step method is statistical measures of model \ufb01t, including Akaike\/Bayesian informa- used (as it is highly unlikely that a person will have a probability of tion criterion and various likelihood ratio tests, evaluated in combi- 1 for a particular trajectory subgroup and 0 for others). However, if nation with metrics based on each person\u2019s probability of membership class membership is well predicted, the in\ufb02uence of this uncer- for each trajectory. Both types of metrics are used in latent class tech- tainty on results will be minor.9 van der Schoot discusses advan- niques more generally and are well described in the previous Research tages and disadvantages of one-step and three-step approaches.9 Note on latent class analysis.1 Trajectory groups can also be related to other time-varying cova- riates, for example: latent trajectory groups for two different out- Statistical indices do not always clearly indicate one optimal comes can be estimated separately, then resultant subgroups model and researchers often need to choose between models with associated via linking membership probabilities (dual trajectory different numbers of trajectory subgroups but comparable \ufb01t statis- analysis) or trajectory subgroups can be estimated based on tics. Therefore, evaluating the validity and utility of trajectories repeated measures of multiple outcomes rather than a single considering clinical and research knowledge is equally important.11,15 outcome (multi-trajectory analysis).11 This may involve evaluating potential models using \ufb01t statistics in conjunction with prior \ufb01ndings, pre-existing theory, patient experi- Sample size is an important, but complex, consideration for latent ence or clinical observation. Plotting mean trajectory patterns in class trajectory studies. Simulation studies suggest that samples of at combination with observed individual trajectories within each least 200 and preferably . 500 individuals are warranted, with model","Appraisal Research Note 63 Figure 1. Illustration of four approaches to trajectory modelling in health research, using hypothetical simulated data. A. Illustrates how a mixed model approach can use other measured variables (in this case sex) to model between-person variation in within-person change over time. B. The latent growth mixture model estimates \u2018latent\u2019 subgroups based on parameters for time. Within each latent subgroup, further variation in time parameters between individuals is estimated. Everyone in the same subgroup shares similar, but not exactly the same, trajectory shapes. To demonstrate this, added dots represent simulated observed values for one individual within each subgroup, and the added dashed lines represent that person\u2019s corresponding estimated trajectory. C. Latent class growth analysis also estimates \u2018latent\u2019 subgroups based on parameters for time, but every person assigned to a particular trajectory subgroup is assumed to have exactly the same trajectory shape, and deviations of their observations from the overall subgroup trajectory are considered random error. To demonstrate this, added dots represent simulated observed values for one individual within each subgroup; that person\u2019s corresponding estimated trajectory is the subgroup mean. D. Longitudinal latent class analysis estimates subgroups of people by using the repeated measures of the health outcome of interest directly (in this case severity score), rather than \u2018growth\u2019 parameters for change over time (ie, the coef\ufb01cients for time, such as initial value, slope and curvature). Subgrouping uses the repeated measures of the health phenomenon themselves, treating the data as if it were collected cross-sectionally, rather than analysing it as longitudinal data accounting for time ordering. Dots represent the mean value at each time point for each subgroup. performance also being a function of number of time points, and the 16-item reporting tool that can facilitate the transparency and number, relative size and distinctness of underlying population sub- robustness of latent trajectory \ufb01ndings.9 groups.16 The main problems of insuf\ufb01cient sample size are the risk of over-\ufb01tting the data, which means that spurious subgroups might be Discussion and Summary identi\ufb01ed as a result of sample variation or alternatively, small but important population subgroups fail to be identi\ufb01ed. In addition, One of the reasons why latent class trajectory methods have there are often problems with models failing to converge on a solu- become so popular is because the resultant trajectories are intuitively tion when there are too many parameters to estimate compared with easy to understand, and once estimated provide a helpful way to the number of data points available. Latent class trajectory analysis relate complex longitudinal patterns of change in a health outcome to uses maximum likelihood estimation, which accommodates missing other health variables. In addition, they are relatively easy to imple- time points when they are missing at random. This means that people ment in various popular software packages. With the increasing missing some time points can still be included in, and be informative availability and ease of collection of longitudinal data from mobile for, the analysis. Furthermore, as those with less data are allocated to devices and sensors, it is likely that latent class trajectory methods a trajectory subgroup with less certainty, uncertainty due to missing will be even more commonly utilised. However, it is important that data can be incorporated in subsequent analyses using trajectory latent class trajectory methods are used judiciously, pro\ufb01ciently and groups by accounting for individuals\u2019 probability of membership. in conjunction with substantive knowledge of the particular \ufb01eld of However, large amounts of missing data, insuf\ufb01cient sample size or enquiry.15 As inferences about the number and shape of latent tra- limited number of time points can compromise validity and subgroup jectories can be heavily in\ufb02uenced by sample size, nature of the data detection.17 and methodological choice, there have been strong calls for caution in the interpretation of \ufb01ndings, with researchers urged to consider Because there are so many choices involved in latent trajectory latent trajectories as abstract rather than concrete, and an individual\u2019s modelling, and the software used to implement them have different assignment to a particular trajectory group as an approximation capabilities, the \ufb01ndings from latent class trajectory analyses can vary, only.17,18 An individual will be allocated to the subgroup for which sometimes greatly;18 therefore, quality reporting of the methodo- they have the highest probability of membership, but they will have a logical process is essential. The Guidelines for Reporting on Latent probability of membership, albeit lower, for all trajectory subgroups Trajectory Studies (GRoLTS) checklist is a useful and recommended","64 Appraisal Research Note and their particular trajectory pattern will not exactly match that of Provenance: Invited. Peer reviewed. the mean trajectory for their allocated subgroup. Anne Smitha and Kate M Dunnb Replication of latent class trajectory studies is important to un- aCurtin School of Allied Health & Curtin Enable Institute, Australia derstand commonalities and differences across populations, for example: in the \ufb01eld of low back pain, a large body of trajectory bSchool of Medicine, Keele University, UK research across different settings and countries has shifted under- standing of the condition away from a recovery\/nonrecovery para- References digm to a more subtle classi\ufb01cation that includes patterns of long-term \ufb02uctuating or episodic pain, such that contemporary un- 1. Kongsted A, Nielsen AM. J Physiother. 2017;63:55\u201358. derstanding is of a long-lasting condition with varying trajectories.1,19 2. Dunn KM, et al. BMJ Open. 2013;3. As further studies are unlikely to identify different trajectory patterns, 3. Walton DMP, et al. Arch Phys Med Rehabil. 2014;95:303\u2013308. focus could now shift to investigation of trajectory patterns as 4. Aili K, et al. Pain. 2021;162:1511\u20131520. prognostic markers, patient communication aids, outcome measures 5. Mose S, et al. Clin Epidemiol. 2021;13:825\u2013843. or potential treatment effect modi\ufb01ers (ie, particular trajectory sub- 6. Radoj\u0014ci\u0013c MR, et al. J Clin Epidemiol. 2022;141:54\u201363. groups as targets for speci\ufb01c treatments).1 7. Dowsey MM, et al. Osteoarthr Cartil. 2015;23:2141\u20132149. 8. Beales D, et al. In\ufb02amm Res. 2021;70:799\u2013809. In summary, latent class trajectory analysis is a valuable person- 9. van de Schoot R, et al. Struct Equ Modeling. 2017;24:451\u2013467. centred approach for longitudinal health data. Findings can poten- 10. Muth\u00e9n BO, Muth\u00e9n LK. Alcohol: Clin Exp Res. 2000;24:882\u2013891. tially facilitate personalised approaches to health care, if re- 11. Nagin DS, Odgers CL. Annu Rev Clin Psychol. 2010:109\u2013138. searchers can identify and test ways to incorporate them into 12. Twisk J, Hoekstra T. J Clin Epidemiol. 2012;65:1078\u20131087. clinical practice. With the rapidly growing application of these 13. Herle M, et al. Eur J Epidemiol. 2020;35:205\u2013222. techniques, there is a need to balance enthusiasm for use with 14. Coenen P, et al. Arthritis Care Res. 2017;69:403\u2013412. caution. This involves both recognition that inferences regarding 15. Lennon H, et al. BMJ Open. 2018;8:e020683. trajectory subgroups are affected by choice of methods, and the 16. Sijbrandij JJ, et al. Adv Life Course Res. 2019;42:100288. avoidance of overinterpretation of identi\ufb01ed subgroups from a 17. Bauer DJ. Multivar Behav Res. 2007;42:757\u2013786. sample as real subpopulations versus useful summaries of complex 18. Warren JR, et al. Am J Sociol. 2015;120:1809\u20131856. longitudinal data. 19. Hartvigsen J, et al. Lancet. 2018;391:2356\u20132367. 20. Curran PJ. Multivar Behav Res. 2003;38:529\u2013569."]
Search
