Effectiveness of Internet-Based Exercises Aimed at Treating Knee Osteoarthritis: The iBEAT-OA Randomized Clinical Trial | Geriatrics | JAMA Network Open | JAMA Network
[Skip to Navigation]
Sign In
Visual Abstract. Effectiveness of Internet-based Exercises Aimed at Treating Knee Osteoarthritis
Effectiveness of Internet-based Exercises Aimed at Treating Knee Osteoarthritis
Figure 1.  Flowchart of Participants
Flowchart of Participants

BMI indicates body mass index (calculated as weight in kilograms divided by height in meters squared); COVID-19, coronavirus disease 2019.

Figure 2.  Change in Numerical Rating Scale (NRS) Pain Scores From Baseline to 6 Weeks
Change in Numerical Rating Scale (NRS) Pain Scores From Baseline to 6 Weeks
Figure 3.  Change in 30-Second Chair Sit-to-Stand and Timed Up-and-Go Scores From Baseline to 6 Weeks
Change in 30-Second Chair Sit-to-Stand and Timed Up-and-Go Scores From Baseline to 6 Weeks
Table 1.  Participant Baseline Characteristics
Participant Baseline Characteristics
Table 2.  Change in Primary and Secondary Outcomes
Change in Primary and Secondary Outcomes
1.
Hunter  DJ, Bierma-Zeinstra  S. Osteoarthritis. Lancet. 2019;393(10182):1745-1759. doi:10.1016/S0140-6736(19)30417-9PubMed
2.
McCormick  A, Fleming  D, Charlton  J.  Morbidity Statistics From General Practice: Fourth National Study 1991-1992. HMSO; 1995.
3.
Morgan  OJ, Hillstrom  HJ, Ellis  SJ,  et al.  Osteoarthritis in England: incidence trends from National Health Service hospital episode statistics.   ACR Open Rheumatol. 2019;1(8):493-498. doi:10.1002/acr2.11071PubMedGoogle ScholarCrossref
4.
Turkiewicz  A, Gerhardsson de Verdier  M, Engström  G,  et al.  Prevalence of knee pain and knee OA in southern Sweden and the proportion that seeks medical care.   Rheumatology (Oxford). 2015;54(5):827-835. doi:10.1093/rheumatology/keu409 PubMedGoogle ScholarCrossref
5.
Conaghan  PG, Dickson  J, Grant  RL; Guideline Development Group.  Care and management of osteoarthritis in adults: summary of NICE guidance.   BMJ. 2008;336(7642):502-503. doi:10.1136/bmj.39490.608009.AD PubMedGoogle ScholarCrossref
6.
Ekman  B, Nero  H, Lohmander  LS, Dahlberg  LE.  Costing analysis of a digital first-line treatment platform for patients with knee and hip osteoarthritis in Sweden.   PLoS One. 2020;15(8):e0236342. doi:10.1371/journal.pone.0236342 PubMedGoogle Scholar
7.
Nero  H, Dahlberg  J, Dahlberg  LEA.  6-Week web-based osteoarthritis treatment program: observational quasi-experimental study.   J Med Internet Res. 2017;19(12):e422. doi:10.2196/jmir.9255 PubMedGoogle Scholar
8.
Dahlberg  LE, Dell’Isola  A, Lohmander  LS, Nero  H.  Improving osteoarthritis care by digital means—effects of a digital self-management program after 24- or 48-weeks of treatment.   PLoS One. 2020;15(3):e0229783. doi:10.1371/journal.pone.0229783 PubMedGoogle Scholar
9.
Cronström  A, Nero  H, Dahlberg  LE.  Factors associated with patients’ willingness to consider joint surgery after completion of a digital osteoarthritis treatment program: a prospective cohort study.   Arthritis Care Res (Hoboken). 2019;71(9):1194-1201. doi:10.1002/acr.23772 PubMedGoogle ScholarCrossref
10.
Cronström  A, Dahlberg  LE, Nero  H, Ericson  J, Hammarlund  CS.  ‘I would never have done it if it hadn’t been digital’: a qualitative study on patients’ experiences of a digital management programme for hip and knee osteoarthritis in Sweden.   BMJ Open. 2019;9(5):e028388. doi:10.1136/bmjopen-2018-028388 PubMedGoogle Scholar
11.
Cronström  A, Dahlberg  LE, Nero  H, Hammarlund  CS.  “I was considering surgery because I believed that was how it was treated”: a qualitative study on willingness for joint surgery after completion of a digital management program for osteoarthritis.   Osteoarthritis Cartilage. 2019;27(7):1026-1032. doi:10.1016/j.joca.2019.04.004PubMedGoogle ScholarCrossref
12.
World Medical Association.  World Medical Association Declaration of Helsinki: ethical principles for medical research involving human subjects.   JAMA. 2013;310(20):2191-2194. doi:10.1001/jama.2013.281053PubMedGoogle ScholarCrossref
13.
Gohir  SA, Greenhaff  P, Abhishek  A, Valdes  AM.  Evaluating the efficacy of Internet-Based Exercise Programme Aimed At Treating Knee Osteoarthritis (iBEAT-OA) in the community: a study protocol for a randomised controlled trial.   BMJ Open. 2019;9(10):e030564. doi:10.1136/bmjopen-2019-030564 PubMedGoogle Scholar
14.
Altman  RD.  Classification of disease: osteoarthritis.   Semin Arthritis Rheum. 1991;20(6)(suppl 2):40-47. doi:10.1016/0049-0172(91)90026-VPubMedGoogle ScholarCrossref
15.
Kellgren  JH, Lawrence  JS.  Radiological assessment of osteo-arthrosis.   Ann Rheum Dis. 1957;16(4):494-502. doi:10.1136/ard.16.4.494PubMedGoogle ScholarCrossref
16.
Schiphof  D, Boers  M, Bierma-Zeinstra  SMA.  Differences in descriptions of Kellgren and Lawrence grades of knee osteoarthritis.   Ann Rheum Dis. 2008;67(7):1034-1036. doi:10.1136/ard.2007.079020 PubMedGoogle ScholarCrossref
17.
Versus Arthritis. Osteoarthritis (OA) of the knee. Accessed January 13, 2021. https://www.versusarthritis.org/about-arthritis/conditions/osteoarthritis-of-the-knee/
18.
Hawker  GA, Mian  S, Kendzerska  T, French  M.  Measures of adult pain: Visual Analog Scale for Pain (VAS Pain), Numeric Rating Scale for Pain (NRS Pain), McGill Pain Questionnaire (MPQ), Short-Form McGill Pain Questionnaire (SF-MPQ), Chronic Pain Grade Scale (CPGS), Short Form-36 Bodily Pain Scale (SF-36 BPS), and Measure of Intermittent and Constant Osteoarthritis Pain (ICOAP).   Arthritis Care Res (Hoboken). 2011;63(suppl 11):S240-S252. doi:10.1002/acr.20543 PubMedGoogle ScholarCrossref
19.
Alghadir  AH, Anwer  S, Iqbal  A, Iqbal  ZA.  Test-retest reliability, validity, and minimum detectable change of visual analog, numerical rating, and verbal rating scales for measurement of osteoarthritic knee pain.   J Pain Res. 2018;11:851-856. doi:10.2147/JPR.S158847 PubMedGoogle ScholarCrossref
20.
Bellamy  N, Buchanan  WW, Goldsmith  CH, Campbell  J, Stitt  LW.  Validation study of WOMAC: a health status instrument for measuring clinically important patient relevant outcomes to antirheumatic drug therapy in patients with osteoarthritis of the hip or knee.   J Rheumatol. 1988;15(12):1833-1840.PubMedGoogle Scholar
21.
McConnell  S, Kolopack  P, Davis  AM.  The Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC): a review of its utility and measurement properties.   Arthritis Rheum. 2001;45(5):453-461. doi:10.1002/1529-0131(200110)45:5<453::AID-ART365>3.0.CO;2-W PubMedGoogle ScholarCrossref
22.
Collins  NJ, Misra  D, Felson  DT, Crossley  KM, Roos  EM.  Measures of knee function: International Knee Documentation Committee (IKDC) Subjective Knee Evaluation Form, Knee Injury and Osteoarthritis Outcome Score (KOOS), Knee Injury and Osteoarthritis Outcome Score Physical Function Short Form (KOOS-PS), Knee Outcome Survey Activities of Daily Living Scale (KOS-ADL), Lysholm Knee Scoring Scale, Oxford Knee Score (OKS), Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC), Activity Rating Scale (ARS), and Tegner Activity Score (TAS).   Arthritis Care Res (Hoboken). 2011;63(0-11)(suppl 11):S208-S228. doi:10.1002/acr.20632PubMedGoogle ScholarCrossref
23.
Gill  SD, de Morton  NA, Mc Burney  H.  An investigation of the validity of six measures of physical function in people awaiting joint replacement surgery of the hip or knee.   Clin Rehabil. 2012;26(10):945-951. doi:10.1177/0269215511434993 PubMedGoogle ScholarCrossref
24.
Podsiadlo  D, Richardson  S.  The timed “Up & Go”: a test of basic functional mobility for frail elderly persons.   J Am Geriatr Soc. 1991;39(2):142-148. doi:10.1111/j.1532-5415.1991.tb01616.xPubMedGoogle ScholarCrossref
25.
Jones  CJ, Rikli  RE, Beam  WCA.  A 30-s chair-stand test as a measure of lower body strength in community-residing older adults.   Res Q Exerc Sport. 1999;70(2):113-119. doi:10.1080/02701367.1999.10608028 PubMedGoogle ScholarCrossref
26.
Kennedy  DM, Hanna  SE, Stratford  PW, Wessel  J, Gollish  JD.  Preoperative function and gender predict pattern of functional recovery after hip and knee arthroplasty.   J Arthroplasty. 2006;21(4):559-566. doi:10.1016/j.arth.2005.07.010 PubMedGoogle ScholarCrossref
27.
Gill  S, McBurney  H.  Reliability of performance-based measures in people awaiting joint replacement surgery of the hip or knee.   Physiother Res Int. 2008;13(3):141-152. doi:10.1002/pri.411 PubMedGoogle ScholarCrossref
28.
Zeni  JA  Jr, Axe  MJ, Snyder-Mackler  L.  Clinical predictors of elective total joint replacement in persons with end-stage knee osteoarthritis.   BMC Musculoskelet Disord. 2010;11:86. doi:10.1186/1471-2474-11-86 PubMedGoogle ScholarCrossref
29.
Halket  A, Stratford  PW, Kennedy  DM, Woodhouse  LJ.  Using hierarchical linear modeling to explore predictors of pain after total hip and knee arthroplasty as a consequence of osteoarthritis.   J Arthroplasty. 2010;25(2):254-262. doi:10.1016/j.arth.2009.01.007 PubMedGoogle ScholarCrossref
30.
French  HP, Fitzpatrick  M, FitzGerald  O.  Responsiveness of physical function outcomes following physiotherapy intervention for osteoarthritis of the knee: an outcome comparison study.   Physiotherapy. 2011;97(4):302-308. doi:10.1016/j.physio.2010.03.002 PubMedGoogle ScholarCrossref
31.
Mizner  RL, Petterson  SC, Clements  KE, Zeni  JA  Jr, Irrgang  JJ, Snyder-Mackler  L.  Measuring functional improvement after total knee arthroplasty requires both performance-based and patient-report assessments: a longitudinal analysis of outcomes.   J Arthroplasty. 2011;26(5):728-737. doi:10.1016/j.arth.2010.06.004 PubMedGoogle ScholarCrossref
32.
Bennell  K, Dobson  F, Hinman  R.  Measures of physical performance assessments: Self-Paced Walk Test (SPWT), Stair Climb Test (SCT), Six-Minute Walk Test (6MWT), Chair Stand Test (CST), Timed Up & Go (TUG), Sock Test, Lift and Carry Test (LCT), and Car Task.   Arthritis Care Res (Hoboken). 2011;63(suppl 11):S350-S370. doi:10.1002/acr.20538 PubMedGoogle ScholarCrossref
33.
Hill  JC, Kang  S, Benedetto  E,  et al.  Development and initial cohort validation of the Arthritis Research UK Musculoskeletal Health Questionnaire (MSK-HQ) for use across musculoskeletal care pathways.   BMJ Open. 2016;6(8):e012331. doi:10.1136/bmjopen-2016-012331 PubMedGoogle Scholar
34.
Tan  J, Balci  N, Sepici  V, Gener  FA.  Isokinetic and isometric strength in osteoarthrosis of the knee: a comparative study with healthy women.   Am J Phys Med Rehabil. 1995;74(5):364-369. doi:10.1097/00002060-199509000-00008 PubMedGoogle ScholarCrossref
35.
Lienhard  K, Lauermann  SP, Schneider  D, Item-Glatthorn  JF, Casartelli  NC, Maffiuletti  NA.  Validity and reliability of isometric, isokinetic and isoinertial modalities for the assessment of quadriceps muscle strength in patients with total knee arthroplasty.   J Electromyogr Kinesiol. 2013;23(6):1283-1288. doi:10.1016/j.jelekin.2013.09.004 PubMedGoogle ScholarCrossref
36.
Wessel  J.  The reliability and validity of pain threshold measurements in osteoarthritis of the knee.   Scand J Rheumatol. 1995;24(4):238-242. doi:10.3109/03009749509100881 PubMedGoogle ScholarCrossref
37.
Gagliese  L, Weizblit  N, Ellis  W, Chan  VWS.  The measurement of postoperative pain: a comparison of intensity scales in younger and older surgical patients.   Pain. 2005;117(3):412-420. doi:10.1016/j.pain.2005.07.004 PubMedGoogle ScholarCrossref
38.
Mutlu  EK, Ozdincler  AR.  Reliability and responsiveness of algometry for measuring pressure pain threshold in patients with knee osteoarthritis.   J Phys Ther Sci. 2015;27(6):1961-1965. doi:10.1589/jpts.27.1961 PubMedGoogle ScholarCrossref
39.
Middlebrook  N, Heneghan  NR, Evans  DW, Rushton  A, Falla  D.  Reliability of temporal summation, thermal and pressure pain thresholds in a healthy cohort and musculoskeletal trauma population.   PLoS One. 2020;15(5):e0233521. doi:10.1371/journal.pone.0233521 PubMedGoogle Scholar
40.
Pud  D, Granovsky  Y, Yarnitsky  D.  The methodology of experimentally induced diffuse noxious inhibitory control (DNIC)-like effect in humans.   Pain. 2009;144(1-2):16-19. doi:10.1016/j.pain.2009.02.015 PubMedGoogle ScholarCrossref
41.
Yarnitsky  D, Arendt-Nielsen  L, Bouhassira  D,  et al.  Recommendations on terminology and practice of psychophysical DNIC testing.   Eur J Pain. 2010;14(4):339. doi:10.1016/j.ejpain.2010.02.004 PubMedGoogle ScholarCrossref
42.
Wylde  V, Palmer  S, Learmonth  ID, Dieppe  P.  Test-retest reliability of Quantitative Sensory Testing in knee osteoarthritis and healthy participants.   Osteoarthritis Cartilage. 2011;19(6):655-658. doi:10.1016/j.joca.2011.02.009PubMedGoogle ScholarCrossref
43.
Lewis  GN, Rice  DA, McNair  PJ.  Conditioned pain modulation in populations with chronic pain: a systematic review and meta-analysis.   J Pain. 2012;13(10):936-944. doi:10.1016/j.jpain.2012.07.005 PubMedGoogle ScholarCrossref
44.
Valencia  C, Kindler  LL, Fillingim  RB, George  SZ.  Stability of conditioned pain modulation in two musculoskeletal pain models: investigating the influence of shoulder pain intensity and gender.   BMC Musculoskelet Disord. 2013;14:182. doi:10.1186/1471-2474-14-182 PubMedGoogle ScholarCrossref
45.
Biurrun Manresa  JA, Fritsche  R, Vuilleumier  PH,  et al.  Is the conditioned pain modulation paradigm reliable: a test-retest assessment using the nociceptive withdrawal reflex.   PLoS One. 2014;9(6):e100241. doi:10.1371/journal.pone.0100241 PubMedGoogle Scholar
46.
Kennedy  DL, Kemp  HI, Ridout  D, Yarnitsky  D, Rice  ASC.  Reliability of conditioned pain modulation: a systematic review.   Pain. 2016;157(11):2410-2419. doi:10.1097/j.pain.0000000000000689 PubMedGoogle ScholarCrossref
47.
Fernandes  C, Pidal-Miranda  M, Samartin-Veiga  N, Carrillo-de-la-Peña  MT.  Conditioned pain modulation as a biomarker of chronic pain: a systematic review of its concurrent validity.   Pain. 2019;160(12):2679-2690. doi:10.1097/j.pain.0000000000001664 PubMedGoogle ScholarCrossref
48.
Health Research Authority. Informing participants and seeking consent. Updated September 4, 2019. Accessed January 13, 2021. https://www.hra.nhs.uk/planning-and-improving-research/best-practice/informing-participants-and-seeking-consent/
49.
Fransen  M, McConnell  S, Harmer  AR, Van der Esch  M, Simic  M, Bennell  KL.  Exercise for osteoarthritis of the knee: a Cochrane systematic review.   Br J Sports Med. 2015;49(24):1554-1557. doi:10.1136/bjsports-2015-095424 PubMedGoogle ScholarCrossref
50.
National Institute for Health and Care Excellence. Osteoarthritis: care and management. Updated December 2020. Accessed January 13, 2021. https://www.nice.org.uk/guidance/cg177
51.
Clifton  L, Clifton  DA.  The correlation between baseline score and post-intervention score, and its implications for statistical analysis.   Trials. 2019;20(1):43. doi:10.1186/s13063-018-3108-3 PubMedGoogle ScholarCrossref
52.
Mickey  RM, Greenland  S.  The impact of confounder selection criteria on effect estimation.   Am J Epidemiol. 1989;129(1):125-137. doi:10.1093/oxfordjournals.aje.a115101 PubMedGoogle ScholarCrossref
53.
Bannuru  RR, Osani  MC, Vaysbrot  EE,  et al.  OARSI guidelines for the non-surgical management of knee, hip, and polyarticular osteoarthritis.   Osteoarthritis Cartilage. 2019;27(11):1578-1589. doi:10.1016/j.joca.2019.06.011PubMedGoogle ScholarCrossref
54.
Pham  T, van der Heijde  D, Altman  RD,  et al.  OMERACT-OARSI initiative: Osteoarthritis Research Society International set of responder criteria for osteoarthritis clinical trials revisited.   Osteoarthritis Cartilage. 2004;12(5):389-399. doi:10.1016/j.joca.2004.02.001PubMedGoogle ScholarCrossref
55.
Devji  T, Guyatt  GH, Lytvyn  L,  et al.  Application of minimal important differences in degenerative knee disease outcomes: a systematic review and case study to inform BMJ Rapid Recommendations.   BMJ Open. 2017;7(5):e015587. doi:10.1136/bmjopen-2016-015587 PubMedGoogle Scholar
56.
MacKay  C, Clements  N, Wong  R, Davis  AM.  A systematic review of estimates of the minimal clinically important difference and patient acceptable symptom state of the Western Ontario and McMaster Universities Osteoarthritis Index in patients who underwent total hip and total knee replacement.   Osteoarthritis Cartilage. 2019;27(10):1408-1419. doi:10.1016/j.joca.2019.05.002PubMedGoogle ScholarCrossref
57.
Juhl  C, Christensen  R, Roos  EM, Zhang  W, Lund  H.  Impact of exercise type and dose on pain and disability in knee osteoarthritis: a systematic review and meta-regression analysis of randomized controlled trials.   Arthritis Rheumatol. 2014;66(3):622-636. doi:10.1002/art.38290 PubMedGoogle ScholarCrossref
58.
Thorstensson  CA, Garellick  G, Rystedt  H, Dahlberg  LE.  Better management of patients with osteoarthritis: development and nationwide implementation of an evidence-based supported osteoarthritis self-management programme.   Musculoskeletal Care. 2015;13(2):67-75. doi:10.1002/msc.1085 PubMedGoogle ScholarCrossref
59.
Jönsson  T, Eek  F, Dell’Isola  A, Dahlberg  LE, Ekvall Hansson  E.  The better management of patients with osteoarthritis program: outcomes after evidence-based education and exercise delivered nationwide in Sweden.   PLoS One. 2019;14(9):e0222657. doi:10.1371/journal.pone.0222657 PubMedGoogle Scholar
60.
Thorlund  JB, Roos  EM, Goro  P, Ljungcrantz  EG, Grønne  DT, Skou  ST.  Patients use fewer analgesics following supervised exercise therapy and patient education: an observational study of 16 499 patients with knee or hip osteoarthritis.   Br J Sports Med. 2020;bjsports-2019-101265. doi:10.1136/bjsports-2019-101265 PubMedGoogle Scholar
61.
Polaski  AM, Phelps  AL, Kostek  MC, Szucs  KA, Kolber  BJ.  Exercise-induced hypoalgesia: a meta-analysis of exercise dosing for the treatment of chronic pain.   PLoS One. 2019;14(1):e0210418. doi:10.1371/journal.pone.0210418 PubMedGoogle Scholar
62.
Castro Sweet  CM, Chiguluri  V, Gumpina  R,  et al.  Outcomes of a digital health program with human coaching for diabetes risk reduction in a Medicare population.   J Aging Health. 2018;30(5):692-710. doi:10.1177/0898264316688791 PubMedGoogle ScholarCrossref
63.
Toelle  TR, Utpadel-Fischler  DA, Haas  K-K, Priebe  JA.  App-based multidisciplinary back pain treatment versus combined physiotherapy plus online education: a randomized controlled trial.   NPJ Digit Med. 2019;2:34. doi:10.1038/s41746-019-0109-x PubMedGoogle ScholarCrossref
64.
Shebib  R, Bailey  JF, Smittenaar  P, Perez  DA, Mecklenburg  G, Hunter  S.  Randomized controlled trial of a 12-week digital care program in improving low back pain.   NPJ Digit Med. 2019;2:1. doi:10.1038/s41746-018-0076-7 PubMedGoogle ScholarCrossref
65.
Fowkes  J, Fross  C, Gilbert  G, Harris  A. Virtual health: a look at the next frontier of care delivery. McKinsey & Company. June 11, 2020. Accessed January 13, 2021. https://www.mckinsey.com/industries/healthcare-systems-and-services/our-insights/virtual-health-a-look-at-the-next-frontier-of-care-delivery
66.
Skou  ST, Roos  EM, Laursen  MB,  et al.  A randomized, controlled trial of total knee replacement.   N Engl J Med. 2015;373(17):1597-1606. doi:10.1056/NEJMoa1505467PubMedGoogle ScholarCrossref
Limit 200 characters
Limit 25 characters
Conflicts of Interest Disclosure

Identify all potential conflicts of interest that might be relevant to your comment.

Conflicts of interest comprise financial interests, activities, and relationships within the past 3 years including but not limited to employment, affiliation, grants or funding, consultancies, honoraria or payment, speaker's bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued.

Err on the side of full disclosure.

If you have no conflicts of interest, check "No potential conflicts of interest" in the box below. The information will be posted with your response.

Not all submitted comments are published. Please see our commenting policy for details.

Limit 140 characters
Limit 3600 characters or approximately 600 words
    Original Investigation
    Orthopedics
    February 23, 2021

    Effectiveness of Internet-Based Exercises Aimed at Treating Knee Osteoarthritis: The iBEAT-OA Randomized Clinical Trial

    Author Affiliations
    • 1National Institute for Health Research, Nottingham Biomedical Research Centre, School of Medicine, University of Nottingham, Nottingham, United Kingdom
    • 2Department of Health Sciences, Lund University, Lund University, Sweden
    JAMA Netw Open. 2021;4(2):e210012. doi:10.1001/jamanetworkopen.2021.0012
    Key Points

    Question  What is the effectiveness of an internet-based exercise program vs routine self-management on pain outcomes among patients with knee osteoarthritis?

    Findings  This randomized clinical trial including 105 patients compared an internet-based program, including recommended information and exercises, with usual care for patients with knee osteoarthritis. Patients receiving the internet-based program experienced decreased pain and improved function at 6 weeks vs the usual care group.

    Meaning  These finding suggest that digitally delivered treatment information provided an important patient benefit and may decrease the burden of treatment for knee osteoarthritis on both patients and health care systems.

    Abstract

    Importance  Osteoarthritis is a prevalent, debilitating, and costly chronic disease for which recommended first-line treatment is underused.

    Objective  To compare the effect of an internet-based treatment for knee osteoarthritis vs routine self-management (ie, usual care).

    Design, Setting, and Participants  This randomized clinical trial was conducted from October 2018 to March 2020. Participants included individuals aged 45 years or older with a diagnosis of knee osteoarthritis recruited from an existing primary care database or from social media advertisements were invited. Data were analyzed April to July 2020.

    Interventions  The intervention and control group conformed to first-line knee osteoarthritis treatment. For the intervention group, treatment was delivered via a smartphone application. The control group received routine self-management care.

    Main Outcomes and Measures  The primary outcome was change from baseline to 6 weeks in self-reported pain during the last 7 days, reported on a numerical rating scale (NRS; range, 0-10, with 0 indicating no pain and 10, worst pain imaginable), compared between groups. Secondary outcomes included 2 physical functioning scores, hamstring and quadriceps muscle strength, the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC), and quantitative sensory testing.

    Results  Among a total of 551 participants screened for eligibility, 146 were randomized and 105 were analyzed (mean [SD] age, 66.7 [9.2] years, 71 [67.1%] women), including 48 participants in the intervention group and 57 participants in the control group. There were no significant differences in baseline characteristics between the groups. At the 6-week follow-up, the intervention group showed a greater NRS pain score reduction than the control group (between-group difference, −1.5 [95% CI, −2.2 to −0.8]; P < .001). Similarly, the intervention group had better improvements in the 30-second sit-to-stand test (between-group difference, 3.4 [95% CI, 2.2 to 4.5]; P < .001) and Timed Up-and-Go test (between-group difference, −1.8 [95% CI, −3.0 to −0.5] seconds; P = .007), as well as the WOMAC subscales for pain (between-group difference, −1.1 [95% CI, −2.0 to −0.2]; P = .02), stiffness (between-group difference, −1.0 [95% CI, −1.5 to −0.5]; P < .001), and physical function (between-group difference, −3.4 [95% CI, −6.2 to −0.7]; P = .02). The magnitude of within-group changes in pain (d = 0.83) and function outcomes (30 second sit-to-stand test d = 1.24; Timed Up-and-Go test d = 0.76) in the intervention group corresponded to medium to very strong effects. No adverse events were reported.

    Conclusions and Relevance  These findings suggest that this internet-delivered, evidence-based, first-line osteoarthritis treatment was superior to routine self-managed usual care and could be provided without harm to people with osteoarthritis. Effect sizes observed in the intervention group corresponded to clinically important improvements.

    Trial Registration  ClinicalTrials.gov Identifier: NCT03545048

    Introduction

    Osteoarthritis is the most common joint disease and among the most prevalent chronic conditions.1 It causes pain and disability in adult and elderly populations, representing a heavy burden on health care systems and society. In the United Kingdom, as in rest of the Western world, 10% to 15% of adults consult their general practitioners about osteoarthritis every year.1-4 The UK National Institute for Health and Care Excellence (NICE) recommends that the first-line treatment for knee osteoarthritis should include disease information and a long-term exercise program.5

    Long-term treatment in chronic diseases is not compatible with the need for cutting health care spending or for reducing face-to-face consultations during the ongoing coronavirus disease 2019 (COVID-19) pandemic. Therefore, internet-based remote treatment may be an efficacious and cost-effective alternative compared with routine in-person treatment.6 The advantages of internet-delivered health care in user flexibility and the ability to receive care at home (thus avoiding travel) have prompted the development of an internet-based first-line osteoarthritis management program consisting of exercises, informational lessons, an asynchronous dialogue with a physiotherapist, and outcome monitoring. A 2017 observational study7 reported improved physical function and decreased pain levels at 6 weeks among patients receiving an internet-based exercise program, and these improvements were confirmed at the 48-week follow-up.8 Additionally, some studies have shown that participants who took part in the 6-week digital program changed their mind regarding undergoing surgical treatment and expressed support for digital osteoarthritis treatment.9-11 However, no randomized clinical trials have compared this digital program with usual care. Previous studies using online interventions suggest that there is a lack of high-quality studies without methodological flaws (eTable 1 in Supplement 1); hence, there is a need for a randomized clinical trial assessing a digital intervention for knee osteoarthritis vs usual care. The aim of this study was to determine the efficacy of a 6-week internet-based exercise intervention7 to modulate pain, muscle strength, and function in individuals with knee osteoarthritis compared with self-managed usual care.

    Methods

    This randomized clinical trial was approved by the sponsors and performed according to the Declaration of Helsinki.12 The study received approval from the Research Ethics Committee of Nottingham University, Health Research Authority, and the Nottingham University Hospitals National Health Service Trust Research and Innovation department. All participants provided written informed consent. This study is reported following the Consolidated Standards of Reporting Trials (CONSORT) reporting guideline.

    Trial Design

    The Internet-Based Exercise Programme Aimed at Treating Knee Osteoarthritis (iBEAT-OA) is a parallel-group randomized clinical trial, performed in the primary care setting and managed from Nottingham City Hospital, Nottingham, United Kingdom, in participants with knee osteoarthritis, comparing a digitally delivered intervention with usual care self-management randomized in a 1-to-1 ratio. The trial protocol has been published elsewhere13 and can be found in Supplement 2.

    Participants

    Participants were recruited from an existing database of community-dwelling adults with knee pain who had previously agreed to be contacted for future studies on osteoarthritis or from answering advertisements posted on social media.

    Inclusion criteria were age 45 years or older, a clinical diagnosis of knee osteoarthritis (defined as knee pain for ≥3 months, early morning stiffness <30 minutes, crepitus, bony tenderness, and no palpable warmth),14 as well as radiographically established knee osteoarthritis (determined by Kellgren and Lawrence grade ≥1 on the 1-4 scale).15,16 Further inclusion criteria were the ability to read and write English and having access to and ability to use a smartphone or tablet.

    Exclusion criteria were inability to give informed consent, terminal or mental illness, neurological conditions (ie, stroke, multiple sclerosis, Parkinson disease, motor neuron disease, muscular dystrophy, or Huntington disease), inflammatory joint diseases (eg, rheumatoid arthritis, gout, or calcium pyrophosphate deposition disease), or dementia. Further exclusion criteria were having been diagnosed with sleep apnea by physician, acute soft tissue injury to the knee within last 3 months before inclusion, and unstable heart condition or rapid fluctuations in hypertension, or a body mass index (BMI; calculated as weight in kilograms divided by height in meters squared) greater than 50.

    Study Groups

    Both study arms relied on osteoarthritis self-management programs. The patients in the control arm received usual care, including exercise and information in accordance to the NICE guidelines5 and Versus Arthritis,17 delivered by their general practitioner and physiotherapists. The intervention group had a structured exercise and osteoarthritis disease information program delivered digitally via a smartphone application (hereafter, app) provided by Joint Academy.

    Intervention Group

    The treatment to the intervention group consisted of a 6-week digitally delivered program accessed via an iOS (Apple) or Google Play (Alphabet) app. It provided the intervention group with daily exercises and informative texts. The open- and closed-chain exercise instructions focused on neuromuscular leg strengthening and core stability and performance, as well as balance enhancement, as exemplified by doing sit-to-stand and stair-climbing exercises. These exercises were adjusted by the program in regard to degrees of complexity, load, and difficulty in relation to each participant’s response after doing the exercise, classified as too easy, good, or too difficult.10 The educational sessions covered the basics of osteoarthritis, its treatment, self-managing symptoms, the benefits of behavioral change and maintaining a healthy lifestyle. Each educational session was followed by a quiz to ensure that participants had understood the key messages. Adherence was encouraged by daily emails or smartphone notifications, or by the physiotherapist via asynchronous chat or telephone during the study period.

    Usual Care Group

    The usual care group was advised to continue with management of knee osteoarthritis as recommended by their general practitioner prior to trial recruitment. This involves use of core and adjunctive treatments, per NICE guidelines,5 and a self-management plan was developed, with patients able to initiate further consultations with general practitioners and therapists and referred to hospital specialists as required. Participants in the usual care group could continue to seek health care input for their knee pain as required during the duration of their study participation. Additionally, some participants in the usual care group were given a patient information leaflet on knee osteoarthritis developed by Versus Arthritis17 by their general practitioner or therapist.

    Outcomes

    All participants were monitored similarly and had a face-to-face meeting at enrollment (ie, baseline) and after 6 weeks (ie, end of study) to assess function, muscle strength, and pain sensitization and to complete questionnaires. At the first session, participants who had not had a radiographic examination in the previous 12 months underwent standing radiographic imaging (posteroanterior view) for Kellgren and Lawrence grading.15,16 Participants were asked to provide information on analgesic medication use (over the counter or prescribed) at both sessions.

    Monitoring

    The intervention group was actively monitored via the app for their exercise adherence, weight monitoring, and progress to next set of exercises. Additionally, support by a physiotherapist was available as needed (online or via telephone). In comparison, the usual care group was not monitored, motivated, or supported by the research physiotherapist.

    Following randomization and allocation, participants in the intervention group were asked to answer the standard online questionnaire in the digital program covering baseline information, such as joint pain intensity, health-related quality of life, and physical function, as well as performing physical tests. The answer to this questionnaire established a baseline fitness level of participants to tailor the exercises (number of repetitions, specific performance). The results of online questionnaire to establish their baseline fitness level are not a part of assessment for this study.

    Primary End Point

    The primary outcome was the between-group difference in change of participant-reported pain assessed by the question “How much pain have you had in the past week in your most painful knee? Can you kindly mark it on this scale?” reported on a numerical rating scale (NRS; range, 0-10, with 0 indicating no pain and 10, the worst pain imaginable)18,19 between baseline and 6 weeks.

    Secondary Outcomes

    Secondary outcomes included the between-group differences in change of the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC), 2 physical functioning tests (ie, the 30-second sit-to-stand test and the Timed Up-and-Go [TUG] test), the Arthritis Research UK Musculoskeletal Health Questionnaire (MSK-HQ), maximum voluntary contraction of quadriceps and hamstring muscles, as well as quantitative sensory testing between baseline and 6 weeks.

    The WOMAC self-administered questionnaire is a widely used, validated participant-reported outcome in hip and knee osteoarthritis.20-23 It contains 24 items divided into 3 subscales and scored on a scale of 0 to 4, with lower scores indicating lower levels of symptoms or physical disability, including pain (5 items; range, 0-20), stiffness (2 items; range, 0-8), and physical function (17 items; range, 0-68).

    The 30-second sit-to-stand test was assessed as the number of times the participant could rise from a sitting position on a chair to a full standing position in 30 seconds, and the TUG test was assessed as the time (measured in seconds) required for the participant to stand up on therapist’s command, walk 3 m, turn around, walk back to the chair, and sit down again. Both are validated methods used to assess participants’ physical function.24-32 At enrollment, participants had a demonstration of the tests and practiced once before doing each test. The TUG was repeated 3 times and the mean time was used. The 30-second sit-to-stand test was conducted once to avoid fatigue. A break was allowed between both tests.

    The MSK-HQ is a self-administered validated questionnaire allowing people with musculoskeletal conditions to report symptoms and quality of life in a standardized way, including stiffness, generalized well-being, difficulty with sleeping, and understanding of the diagnosis and treatment.33 The 14 questions are scored on a scale of 0 to 4 (range, 0-54), with lower scores indicating lower levels of symptoms.

    Isokinetic peak torque of quadriceps and hamstring muscles was measured as newton meters (Nm) at 60°/s and 180°/s using a HUMAC / NORM Testing and Rehabilitation System model 7709 (Computer Sports Medicine ). A standardized protocol34 was used and is a valid and reliable assessment method of muscle strength in quadriceps and hamstrings.35

    The quantitative sensory testing included pressure pain threshold (PPT) to measure tenderness around the knee joint (with 0 indicating maximum tenderness),36-39 temporal summation (TS) to measure element of central sensitization (range, 0-10, with 10 indicating maximum sensitization),40-43 and conditional pain modulation (CPM) to assess the function of endogenous pain inhibitory pathways in humans (with 0 indicating maximum inhibitory pathways).44-47 In addition, sleeping assessment with actigraphy, Pittsburgh Sleep Quality Index, musculoskeletal ultrasonographic assessment, and serum biomarkers of inflammation were measured and will be reported separately in future studies.

    Problems and Harms

    Participants were informed via the participant information sheet required by Health Research Authority,48 that if they have any concerns, they should speak to one of us (S.G.). Any adverse event or harm would be escalated to principal investigator (A.V.) and ultimately reported to the sponsor of this study.

    Sample Size

    A systematic review of 44 high-quality exercise trials for knee osteoarthritis pain (3537 participants) found a mean effect of 12 of 100 points on a visual analogue scale corresponding to a standardized effect size (Cohen d) of 0.49.49 Based on this estimated effect size (corresponding to 1.2 points on the NRS scale), a sample size of 60 participants per group was needed to achieve 75% power, with α level set to .05. The estimated dropout rate for exercise interventions was 12%; therefore, a sample of 67 participants per group was planned.

    Randomization

    Participants were randomized by the research team using Sealed Envelope randomization software version 1.19.1 at a 1-to-1 ration for the intervention or usual care. The code break was kept secure by the research manager.

    Blinding and Similarities of Interventions

    Owing to the nature of the study, no blinding of participants or observers occurred. The intervention and usual care protocols followed the NICE management recommendations for knee osteoarthritis.50

    Statistical Analysis

    Analyses were performed as intention-to-treat, defined as all participants randomly assigned to a treatment group and being assessed after randomization. Any missing values of questionnaire items for participants who completed both baseline and 6 week follow up sessions were imputed using mean imputation method for both groups (eTable 2 in Supplement 1).

    Baseline and follow-up scores are presented for each group as means and SDs, since there were no extreme outliers affecting those measures. For each outcome variables, a change score between baseline and the 6-week follow-up was calculated and used as the outcome variables in the comparison between groups. Since the change scores followed an approximately normal distribution, comparisons between the intervention and control group were performed by analysis of variance, adjusted for baseline scores (analysis of covariance).51 Sex, age, BMI, Kellgren and Lawrence grade, and use of medication were included in the models to evaluate potential confounding that could have occurred from unbalanced distribution between groups, despite randomization. The evaluation was based on a change in estimate approach.52 Since the inclusion did not result in any change in estimate more than 15%, the variables were excluded from the final analyses. Estimated mean differences in change scores between the groups, adjusted for baseline scores, are presented with accompanying 95% CIs. Within-group changes were evaluated by estimated mean change with accompanying 95% CIs. For within-group comparison, statistical significance was determined by 95% CIs not crossing 0. Standardized effect sizes (Cohen d) were computed for within-group changes as observed mean change baseline-follow up divided by pooled SD ([baseline SD + follow-up SD] / 2). The data were analyzed using SPSS statistical software version 24 (IBM) by an author who was unaware of group assignment (F.E.), per the statistical analysis plan in Supplement 2.

    Results

    The trial started in late 2018. The control and intervention participant flowchart, including exclusions and losses to follow-up, are shown in Figure 1. Among a total of 551 participants screened for eligibility, 146 were randomized, including 79 participants randomized to the intervention group and 67 participants randomized to the usual care group. Nine participants in each group did not attend the baseline appointment, and no baseline data and demographic characteristics were obtained for these participants. Owing to the COVID-19 lockdown in the United Kingdom in March 2020, the study ended before 27 participants had their face-to-face follow-up visit at 6 weeks (Figure 1). In addition, 3 participants in the usual care group and 2 participants in the intervention group were lost to follow-up (Figure 1) and were therefore excluded from the analyses. A total of 105 participants were analyzed (mean [SD] age, 66.7 [9.2] years, 71 [67.1%] women), including 48 participants in the intervention group (mean [SD] age, 65.2 [9.7] years; 34 [70.8%] women; mean [SD] BMI, 30.4 [5.5]) and 57 participants in the usual care group (mean [SD] age, 68.0 [8.6] years; 37 [64.9%] women; mean [SD] BMI, 31.9 [5.9]) completed this study. No significant difference was observed between participants groups in terms of use of analgesic medications before or after the intervention. In the intervention group, the mean (SD) adherence with the internet-based exercise program was 87.9% (14.3%) of sessions completed.

    Baseline Data

    The mean baseline demographic and clinical variables were similar for participants in both groups (Table 1). The mean baseline demographic and clinical variables for 17 participants in the control group and 15 participants in the intervention group who could not finish follow-up session were similar to the participants who completed both sessions (Table 1). Owing to COVID-19 lockdown restrictions, baseline data for WOMAC and MSK-HQ for these 32 participants were not retrieved for analysis.

    Primary Outcome

    The intervention group showed a greater decrease in NRS pain score from baseline to 6 weeks compared with the control group (between-group difference, −1.5 [95% CI, −2.2 to −0.8]; P < .001) (Table 2; Figure 2). Between baseline and the 6-week follow-up, there was a statistically significant improvement in NRS pain scores in the intervention group (mean change, −1.8 [95% CI, −2.4 to −1.3]; d = −0.83) but not in the usual care group (mean change, −0.3 [95% CI, −0.8 to 0.2]; d = −0.2) (Figure 2).

    Secondary Outcomes

    The between-group analysis of mean change from baseline to 6 weeks showed that the intervention group improved statistically significantly more than the control group in the WOMAC subscales for pain (between-group difference, −1.1 [95% CI, −2.0 to −0.2]; P = .02), stiffness (between-group difference, −1.0 [95% CI, −1.5 to −0.5]; P < .001) and physical function (between-group difference, −3.4 [95% CI, −6.2 to −0.7]; P = .02), the 30-second sit-to-stand test (between-group difference, 3.4 [95% CI, 2.2 to 4.5]; P < .001), the TUG test (between-group difference, −1.8 [95% CI, −3.0 to −0.5] seconds; P = .007), and hamstring isokinetic strength at 60°/s (between-group difference, 6.9 [95% CI, 1.6 to 12.3] Nm; P = .01). There were no statistically significant between-group differences regarding change in remaining strength measures, PPT, TS, CPM, or MSK-HQ.

    The within-group mean changes showed that the intervention group improved significantly in the 3 WOMAC subscales (pain change, −2.2 [95% CI, −2.9 to −1.6]; d = −0.60; stiffness change, −0.8 [95% CI, −1.2 to −0.4]; d = −.51; and function change, −7.8 [95% CI, −9.8 to −5.7]; d = −0.60), 30-second sit-to-stand test (change, 4.5 [95% CI, 3.7 to 5.4]; d = 1.24), TUG test (change, −1.4 [95% CI, −2.3 to −0.5] seconds; d = −0.76), quadriceps isokinetic strength at 60°/s (change, 9.6 [95% CI, 4.4 to 14.7] Nm; d = 0.20), and hamstring isokinetic strength at 60°/s (change, 10.5 [95% CI, 6.6 to 14.6] Nm; d = 0.35) and at 180°/s (change, 5.8 [95% CI, 2.5 to 9.1] Nm; d = 0.26) (Table 2, Figure 3). For the usual care group between baseline and follow up, there was a statistically significant change in WOMAC subscale scores for pain (change, −1.2 [95% CI, −1.8 to −0.5]; d = −0.33) and function (change, −4.3 [95% CI, −6.2 to −2.4]; d = −0.37), PPT (superolateral patella change, −55.6 [95% CI, −86.1 to −25.2]; d = −0.36; superomedial patella change, −44.0 [95% CI, −72.4 to −15.7]; d = −0.32; medial joint line change, −49.1 [95% CI, −81.4 to −16.7]; d = −0.31]; tibialis anterior muscle, −54.6 [95% CI, −86.1 to −23.0]; d = −0.30), and the 30-second sit-to-stand test (change, 1.2 [95% CI, 0.4 to 2.0]; d = 0.26) (Table 2). There were no within-group changes in TS, CPM, or MSK-HQ questionnaire scores in any of the groups. No serious adverse events were reported in any of the study groups.

    Discussion

    In this randomized clinical trial, we found that an internet-based first-line knee osteoarthritis management program was superior to routine self-managed care in the primary outcome, NRS pain score, as well as secondary outcomes of function performance. Participants in both groups were treated in accordance with guidelines.5,53

    Between-group differences in change from baseline to 6 weeks suggest a clinically relevant benefit resulting from participation in the internet-based program compared with the self-management program. The minimally important change between treatments and responder criteria are debated concepts challenging their application.54-56 However, the improvements seen in the internet-based intervention group, with standardized effect sizes (ie, Cohen d) corresponding to medium to very strong effects, strongly suggest a clinically important improvement. In comparison, the improvements among participants in the self-managed usual care group were small or absent, with standardized effect sizes corresponding to minor or weak effects. The effect sizes attained with the internet-based program used in this study are comparable with or greater than those presented in systematic reviews of face-to-face exercise programs.49,57

    The internet-based program used in this study was derived from the better management of patients with osteoarthritis (BOA) program, initiated in Sweden in 2008 to implement existing osteoarthritis treatment guidelines for education and exercise face-to-face in primary care facilities.58,59 Observational studies from the BOA and Good Life with Osteoarthritis in Denmark (GLA:D) programs subsequently launched in Denmark and other countries reported that among patients with knee and hip osteoarthritis, participating in these programs was associated with reductions in symptoms, improved function, decrease in willingness to undergo surgical treatment, decrease in use of osteoarthritis medication, and reduced use of sick leave.58-60

    The factors associated with the superiority of the internet-based program vs self-management in this study are not clear. Possibly, daily delivery of individualized treatment, together with support, engagement, and nudging from health care professionals, may have played a role.7,8,10 Other studies have suggested that increasing exercise frequency, but not necessarily length of sessions, can improve pain relief and that exercise should be performed at least 3 times a week.57,61 The superior outcome in the intervention group may depend on the content and context in the app, including a combination of standardized exercises and information, as well as using a digital delivery system.

    Our results are consistent with studies showing efficacy and effectiveness of digital management for other chronic conditions.62-64 Advantages of a digitally delivered treatment may include lower costs and easier access for patients living in remote areas where transport may be an added obstacle. In a pandemic, such as the ongoing COVID-19 pandemic, digitally delivered care can continue without interruption.

    Generalizability

    The demographic and basic osteoarthritis characteristics of participants in our randomized clinical trial (ie, age, sex, pain, function) were similar to participants in the population-based osteoarthritis registers of BOA, GLA:D, and Joint Academy.7,8,59,60 Likewise, patients eligible for total knee joint replacement randomized to operation or first-line treatment, had similar average pain level as in the present study.66

    Limitations

    This study has some limitations. Owing to the COVID-19 lockdown, 27 participants could not attend the follow-up visit, preventing the study from reaching the planned statistical power. A further 5 participants discontinued treatment, but this loss to follow-up was less than the expected loss of 12%. Numbers of participants not attending the 6-week follow-up were similar in both groups, and they had similar demographic and baseline characteristics compared with the analyzed set of participants. Additionally, blinding was not possible for this study, and potential placebo effects associated with internet-based treatment compared with usual care are not accounted for.

    Conclusions

    This parallel-group randomized clinical trial compared knee osteoarthritis management via a digital platform with a self-management usual care program. Both forms of management conformed to clinical guidelines. Digital delivery was superior to routine self-management. No serious harms were reported. Our findings suggest that digital treatment has the potential to decrease the osteoarthritis burden on both the health care systems and patients.65

    Back to top
    Article Information

    Accepted for Publication: January 3, 2021.

    Published: February 23, 2021. doi:10.1001/jamanetworkopen.2021.0012

    Correction: This article was corrected on March 25, 2021, to fix transposed data for the intervention and control groups in the visual abstract.

    Open Access: This is an open access article distributed under the terms of the CC-BY License. © 2021 Gohir SA et al. JAMA Network Open.

    Corresponding Author: Ana M. Valdes, PhD, National Institute for Health Research, Nottingham Biomedical Research Centre, University of Nottingham, Academic Rheumatology, Clinical Sciences Building, City Hospital Nottingham, Nottingham NG5 1PB, United Kingdom (ana.valdes@nottingham.ac.uk).

    Author Contributions: Drs Valdes and Gohir had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

    Concept and design: Gohir, Abhishek, Valdes.

    Acquisition, analysis, or interpretation of data: Gohir, Eek, Kelly, Valdes.

    Drafting of the manuscript: Gohir, Abhishek, Valdes.

    Critical revision of the manuscript for important intellectual content: Gohir, Eek, Kelly, Valdes.

    Statistical analysis: Gohir, Eek, Valdes.

    Obtained funding: Valdes.

    Administrative, technical, or material support: Gohir, Kelly.

    Supervision: Gohir, Abhishek, Valdes.

    Conflict of Interest Disclosures: Dr Abhishek reported receiving personal fees from Up-to-Date, Springer, NGM Biopharmaceuticals, and Inflazome outside the submitted work. Dr Valdes reported receiving grants from Pfizer during the conduct of the study. No other disclosures were reported.

    Funding/Support: This study was funded in part by grants 21960 and 18769 from the Versus Arthritis UK Pain Centre (paid to the University of Nottingham) and by the National Institute for Health Research Nottingham Biomedical Research Centre. The Joint Academy provided their software for free for this study. This work was also supported by Pfizer Global medical grant No. WI243608, Global Awards for Advancing Chronic Pain Research (ADVANCE) 2018.

    Role of the Funder/Sponsor: The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

    Data Sharing Statement: See Supplement 3.

    Additional Contributions: Bonnie Millar, PhD (National Institute for Health Research Nottingham Biomedical Research Centre), provided support in data collection and storage. Leif Dahlberg, MD, PhD, read and gave advice for the manuscript. They were not compensated for their contributions.

    References
    1.
    Hunter  DJ, Bierma-Zeinstra  S. Osteoarthritis. Lancet. 2019;393(10182):1745-1759. doi:10.1016/S0140-6736(19)30417-9PubMed
    2.
    McCormick  A, Fleming  D, Charlton  J.  Morbidity Statistics From General Practice: Fourth National Study 1991-1992. HMSO; 1995.
    3.
    Morgan  OJ, Hillstrom  HJ, Ellis  SJ,  et al.  Osteoarthritis in England: incidence trends from National Health Service hospital episode statistics.   ACR Open Rheumatol. 2019;1(8):493-498. doi:10.1002/acr2.11071PubMedGoogle ScholarCrossref
    4.
    Turkiewicz  A, Gerhardsson de Verdier  M, Engström  G,  et al.  Prevalence of knee pain and knee OA in southern Sweden and the proportion that seeks medical care.   Rheumatology (Oxford). 2015;54(5):827-835. doi:10.1093/rheumatology/keu409 PubMedGoogle ScholarCrossref
    5.
    Conaghan  PG, Dickson  J, Grant  RL; Guideline Development Group.  Care and management of osteoarthritis in adults: summary of NICE guidance.   BMJ. 2008;336(7642):502-503. doi:10.1136/bmj.39490.608009.AD PubMedGoogle ScholarCrossref
    6.
    Ekman  B, Nero  H, Lohmander  LS, Dahlberg  LE.  Costing analysis of a digital first-line treatment platform for patients with knee and hip osteoarthritis in Sweden.   PLoS One. 2020;15(8):e0236342. doi:10.1371/journal.pone.0236342 PubMedGoogle Scholar
    7.
    Nero  H, Dahlberg  J, Dahlberg  LEA.  6-Week web-based osteoarthritis treatment program: observational quasi-experimental study.   J Med Internet Res. 2017;19(12):e422. doi:10.2196/jmir.9255 PubMedGoogle Scholar
    8.
    Dahlberg  LE, Dell’Isola  A, Lohmander  LS, Nero  H.  Improving osteoarthritis care by digital means—effects of a digital self-management program after 24- or 48-weeks of treatment.   PLoS One. 2020;15(3):e0229783. doi:10.1371/journal.pone.0229783 PubMedGoogle Scholar
    9.
    Cronström  A, Nero  H, Dahlberg  LE.  Factors associated with patients’ willingness to consider joint surgery after completion of a digital osteoarthritis treatment program: a prospective cohort study.   Arthritis Care Res (Hoboken). 2019;71(9):1194-1201. doi:10.1002/acr.23772 PubMedGoogle ScholarCrossref
    10.
    Cronström  A, Dahlberg  LE, Nero  H, Ericson  J, Hammarlund  CS.  ‘I would never have done it if it hadn’t been digital’: a qualitative study on patients’ experiences of a digital management programme for hip and knee osteoarthritis in Sweden.   BMJ Open. 2019;9(5):e028388. doi:10.1136/bmjopen-2018-028388 PubMedGoogle Scholar
    11.
    Cronström  A, Dahlberg  LE, Nero  H, Hammarlund  CS.  “I was considering surgery because I believed that was how it was treated”: a qualitative study on willingness for joint surgery after completion of a digital management program for osteoarthritis.   Osteoarthritis Cartilage. 2019;27(7):1026-1032. doi:10.1016/j.joca.2019.04.004PubMedGoogle ScholarCrossref
    12.
    World Medical Association.  World Medical Association Declaration of Helsinki: ethical principles for medical research involving human subjects.   JAMA. 2013;310(20):2191-2194. doi:10.1001/jama.2013.281053PubMedGoogle ScholarCrossref
    13.
    Gohir  SA, Greenhaff  P, Abhishek  A, Valdes  AM.  Evaluating the efficacy of Internet-Based Exercise Programme Aimed At Treating Knee Osteoarthritis (iBEAT-OA) in the community: a study protocol for a randomised controlled trial.   BMJ Open. 2019;9(10):e030564. doi:10.1136/bmjopen-2019-030564 PubMedGoogle Scholar
    14.
    Altman  RD.  Classification of disease: osteoarthritis.   Semin Arthritis Rheum. 1991;20(6)(suppl 2):40-47. doi:10.1016/0049-0172(91)90026-VPubMedGoogle ScholarCrossref
    15.
    Kellgren  JH, Lawrence  JS.  Radiological assessment of osteo-arthrosis.   Ann Rheum Dis. 1957;16(4):494-502. doi:10.1136/ard.16.4.494PubMedGoogle ScholarCrossref
    16.
    Schiphof  D, Boers  M, Bierma-Zeinstra  SMA.  Differences in descriptions of Kellgren and Lawrence grades of knee osteoarthritis.   Ann Rheum Dis. 2008;67(7):1034-1036. doi:10.1136/ard.2007.079020 PubMedGoogle ScholarCrossref
    17.
    Versus Arthritis. Osteoarthritis (OA) of the knee. Accessed January 13, 2021. https://www.versusarthritis.org/about-arthritis/conditions/osteoarthritis-of-the-knee/
    18.
    Hawker  GA, Mian  S, Kendzerska  T, French  M.  Measures of adult pain: Visual Analog Scale for Pain (VAS Pain), Numeric Rating Scale for Pain (NRS Pain), McGill Pain Questionnaire (MPQ), Short-Form McGill Pain Questionnaire (SF-MPQ), Chronic Pain Grade Scale (CPGS), Short Form-36 Bodily Pain Scale (SF-36 BPS), and Measure of Intermittent and Constant Osteoarthritis Pain (ICOAP).   Arthritis Care Res (Hoboken). 2011;63(suppl 11):S240-S252. doi:10.1002/acr.20543 PubMedGoogle ScholarCrossref
    19.
    Alghadir  AH, Anwer  S, Iqbal  A, Iqbal  ZA.  Test-retest reliability, validity, and minimum detectable change of visual analog, numerical rating, and verbal rating scales for measurement of osteoarthritic knee pain.   J Pain Res. 2018;11:851-856. doi:10.2147/JPR.S158847 PubMedGoogle ScholarCrossref
    20.
    Bellamy  N, Buchanan  WW, Goldsmith  CH, Campbell  J, Stitt  LW.  Validation study of WOMAC: a health status instrument for measuring clinically important patient relevant outcomes to antirheumatic drug therapy in patients with osteoarthritis of the hip or knee.   J Rheumatol. 1988;15(12):1833-1840.PubMedGoogle Scholar
    21.
    McConnell  S, Kolopack  P, Davis  AM.  The Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC): a review of its utility and measurement properties.   Arthritis Rheum. 2001;45(5):453-461. doi:10.1002/1529-0131(200110)45:5<453::AID-ART365>3.0.CO;2-W PubMedGoogle ScholarCrossref
    22.
    Collins  NJ, Misra  D, Felson  DT, Crossley  KM, Roos  EM.  Measures of knee function: International Knee Documentation Committee (IKDC) Subjective Knee Evaluation Form, Knee Injury and Osteoarthritis Outcome Score (KOOS), Knee Injury and Osteoarthritis Outcome Score Physical Function Short Form (KOOS-PS), Knee Outcome Survey Activities of Daily Living Scale (KOS-ADL), Lysholm Knee Scoring Scale, Oxford Knee Score (OKS), Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC), Activity Rating Scale (ARS), and Tegner Activity Score (TAS).   Arthritis Care Res (Hoboken). 2011;63(0-11)(suppl 11):S208-S228. doi:10.1002/acr.20632PubMedGoogle ScholarCrossref
    23.
    Gill  SD, de Morton  NA, Mc Burney  H.  An investigation of the validity of six measures of physical function in people awaiting joint replacement surgery of the hip or knee.   Clin Rehabil. 2012;26(10):945-951. doi:10.1177/0269215511434993 PubMedGoogle ScholarCrossref
    24.
    Podsiadlo  D, Richardson  S.  The timed “Up & Go”: a test of basic functional mobility for frail elderly persons.   J Am Geriatr Soc. 1991;39(2):142-148. doi:10.1111/j.1532-5415.1991.tb01616.xPubMedGoogle ScholarCrossref
    25.
    Jones  CJ, Rikli  RE, Beam  WCA.  A 30-s chair-stand test as a measure of lower body strength in community-residing older adults.   Res Q Exerc Sport. 1999;70(2):113-119. doi:10.1080/02701367.1999.10608028 PubMedGoogle ScholarCrossref
    26.
    Kennedy  DM, Hanna  SE, Stratford  PW, Wessel  J, Gollish  JD.  Preoperative function and gender predict pattern of functional recovery after hip and knee arthroplasty.   J Arthroplasty. 2006;21(4):559-566. doi:10.1016/j.arth.2005.07.010 PubMedGoogle ScholarCrossref
    27.
    Gill  S, McBurney  H.  Reliability of performance-based measures in people awaiting joint replacement surgery of the hip or knee.   Physiother Res Int. 2008;13(3):141-152. doi:10.1002/pri.411 PubMedGoogle ScholarCrossref
    28.
    Zeni  JA  Jr, Axe  MJ, Snyder-Mackler  L.  Clinical predictors of elective total joint replacement in persons with end-stage knee osteoarthritis.   BMC Musculoskelet Disord. 2010;11:86. doi:10.1186/1471-2474-11-86 PubMedGoogle ScholarCrossref
    29.
    Halket  A, Stratford  PW, Kennedy  DM, Woodhouse  LJ.  Using hierarchical linear modeling to explore predictors of pain after total hip and knee arthroplasty as a consequence of osteoarthritis.   J Arthroplasty. 2010;25(2):254-262. doi:10.1016/j.arth.2009.01.007 PubMedGoogle ScholarCrossref
    30.
    French  HP, Fitzpatrick  M, FitzGerald  O.  Responsiveness of physical function outcomes following physiotherapy intervention for osteoarthritis of the knee: an outcome comparison study.   Physiotherapy. 2011;97(4):302-308. doi:10.1016/j.physio.2010.03.002 PubMedGoogle ScholarCrossref
    31.
    Mizner  RL, Petterson  SC, Clements  KE, Zeni  JA  Jr, Irrgang  JJ, Snyder-Mackler  L.  Measuring functional improvement after total knee arthroplasty requires both performance-based and patient-report assessments: a longitudinal analysis of outcomes.   J Arthroplasty. 2011;26(5):728-737. doi:10.1016/j.arth.2010.06.004 PubMedGoogle ScholarCrossref
    32.
    Bennell  K, Dobson  F, Hinman  R.  Measures of physical performance assessments: Self-Paced Walk Test (SPWT), Stair Climb Test (SCT), Six-Minute Walk Test (6MWT), Chair Stand Test (CST), Timed Up & Go (TUG), Sock Test, Lift and Carry Test (LCT), and Car Task.   Arthritis Care Res (Hoboken). 2011;63(suppl 11):S350-S370. doi:10.1002/acr.20538 PubMedGoogle ScholarCrossref
    33.
    Hill  JC, Kang  S, Benedetto  E,  et al.  Development and initial cohort validation of the Arthritis Research UK Musculoskeletal Health Questionnaire (MSK-HQ) for use across musculoskeletal care pathways.   BMJ Open. 2016;6(8):e012331. doi:10.1136/bmjopen-2016-012331 PubMedGoogle Scholar
    34.
    Tan  J, Balci  N, Sepici  V, Gener  FA.  Isokinetic and isometric strength in osteoarthrosis of the knee: a comparative study with healthy women.   Am J Phys Med Rehabil. 1995;74(5):364-369. doi:10.1097/00002060-199509000-00008 PubMedGoogle ScholarCrossref
    35.
    Lienhard  K, Lauermann  SP, Schneider  D, Item-Glatthorn  JF, Casartelli  NC, Maffiuletti  NA.  Validity and reliability of isometric, isokinetic and isoinertial modalities for the assessment of quadriceps muscle strength in patients with total knee arthroplasty.   J Electromyogr Kinesiol. 2013;23(6):1283-1288. doi:10.1016/j.jelekin.2013.09.004 PubMedGoogle ScholarCrossref
    36.
    Wessel  J.  The reliability and validity of pain threshold measurements in osteoarthritis of the knee.   Scand J Rheumatol. 1995;24(4):238-242. doi:10.3109/03009749509100881 PubMedGoogle ScholarCrossref
    37.
    Gagliese  L, Weizblit  N, Ellis  W, Chan  VWS.  The measurement of postoperative pain: a comparison of intensity scales in younger and older surgical patients.   Pain. 2005;117(3):412-420. doi:10.1016/j.pain.2005.07.004 PubMedGoogle ScholarCrossref
    38.
    Mutlu  EK, Ozdincler  AR.  Reliability and responsiveness of algometry for measuring pressure pain threshold in patients with knee osteoarthritis.   J Phys Ther Sci. 2015;27(6):1961-1965. doi:10.1589/jpts.27.1961 PubMedGoogle ScholarCrossref
    39.
    Middlebrook  N, Heneghan  NR, Evans  DW, Rushton  A, Falla  D.  Reliability of temporal summation, thermal and pressure pain thresholds in a healthy cohort and musculoskeletal trauma population.   PLoS One. 2020;15(5):e0233521. doi:10.1371/journal.pone.0233521 PubMedGoogle Scholar
    40.
    Pud  D, Granovsky  Y, Yarnitsky  D.  The methodology of experimentally induced diffuse noxious inhibitory control (DNIC)-like effect in humans.   Pain. 2009;144(1-2):16-19. doi:10.1016/j.pain.2009.02.015 PubMedGoogle ScholarCrossref
    41.
    Yarnitsky  D, Arendt-Nielsen  L, Bouhassira  D,  et al.  Recommendations on terminology and practice of psychophysical DNIC testing.   Eur J Pain. 2010;14(4):339. doi:10.1016/j.ejpain.2010.02.004 PubMedGoogle ScholarCrossref
    42.
    Wylde  V, Palmer  S, Learmonth  ID, Dieppe  P.  Test-retest reliability of Quantitative Sensory Testing in knee osteoarthritis and healthy participants.   Osteoarthritis Cartilage. 2011;19(6):655-658. doi:10.1016/j.joca.2011.02.009PubMedGoogle ScholarCrossref
    43.
    Lewis  GN, Rice  DA, McNair  PJ.  Conditioned pain modulation in populations with chronic pain: a systematic review and meta-analysis.   J Pain. 2012;13(10):936-944. doi:10.1016/j.jpain.2012.07.005 PubMedGoogle ScholarCrossref
    44.
    Valencia  C, Kindler  LL, Fillingim  RB, George  SZ.  Stability of conditioned pain modulation in two musculoskeletal pain models: investigating the influence of shoulder pain intensity and gender.   BMC Musculoskelet Disord. 2013;14:182. doi:10.1186/1471-2474-14-182 PubMedGoogle ScholarCrossref
    45.
    Biurrun Manresa  JA, Fritsche  R, Vuilleumier  PH,  et al.  Is the conditioned pain modulation paradigm reliable: a test-retest assessment using the nociceptive withdrawal reflex.   PLoS One. 2014;9(6):e100241. doi:10.1371/journal.pone.0100241 PubMedGoogle Scholar
    46.
    Kennedy  DL, Kemp  HI, Ridout  D, Yarnitsky  D, Rice  ASC.  Reliability of conditioned pain modulation: a systematic review.   Pain. 2016;157(11):2410-2419. doi:10.1097/j.pain.0000000000000689 PubMedGoogle ScholarCrossref
    47.
    Fernandes  C, Pidal-Miranda  M, Samartin-Veiga  N, Carrillo-de-la-Peña  MT.  Conditioned pain modulation as a biomarker of chronic pain: a systematic review of its concurrent validity.   Pain. 2019;160(12):2679-2690. doi:10.1097/j.pain.0000000000001664 PubMedGoogle ScholarCrossref
    48.
    Health Research Authority. Informing participants and seeking consent. Updated September 4, 2019. Accessed January 13, 2021. https://www.hra.nhs.uk/planning-and-improving-research/best-practice/informing-participants-and-seeking-consent/
    49.
    Fransen  M, McConnell  S, Harmer  AR, Van der Esch  M, Simic  M, Bennell  KL.  Exercise for osteoarthritis of the knee: a Cochrane systematic review.   Br J Sports Med. 2015;49(24):1554-1557. doi:10.1136/bjsports-2015-095424 PubMedGoogle ScholarCrossref
    50.
    National Institute for Health and Care Excellence. Osteoarthritis: care and management. Updated December 2020. Accessed January 13, 2021. https://www.nice.org.uk/guidance/cg177
    51.
    Clifton  L, Clifton  DA.  The correlation between baseline score and post-intervention score, and its implications for statistical analysis.   Trials. 2019;20(1):43. doi:10.1186/s13063-018-3108-3 PubMedGoogle ScholarCrossref
    52.
    Mickey  RM, Greenland  S.  The impact of confounder selection criteria on effect estimation.   Am J Epidemiol. 1989;129(1):125-137. doi:10.1093/oxfordjournals.aje.a115101 PubMedGoogle ScholarCrossref
    53.
    Bannuru  RR, Osani  MC, Vaysbrot  EE,  et al.  OARSI guidelines for the non-surgical management of knee, hip, and polyarticular osteoarthritis.   Osteoarthritis Cartilage. 2019;27(11):1578-1589. doi:10.1016/j.joca.2019.06.011PubMedGoogle ScholarCrossref
    54.
    Pham  T, van der Heijde  D, Altman  RD,  et al.  OMERACT-OARSI initiative: Osteoarthritis Research Society International set of responder criteria for osteoarthritis clinical trials revisited.   Osteoarthritis Cartilage. 2004;12(5):389-399. doi:10.1016/j.joca.2004.02.001PubMedGoogle ScholarCrossref
    55.
    Devji  T, Guyatt  GH, Lytvyn  L,  et al.  Application of minimal important differences in degenerative knee disease outcomes: a systematic review and case study to inform BMJ Rapid Recommendations.   BMJ Open. 2017;7(5):e015587. doi:10.1136/bmjopen-2016-015587 PubMedGoogle Scholar
    56.
    MacKay  C, Clements  N, Wong  R, Davis  AM.  A systematic review of estimates of the minimal clinically important difference and patient acceptable symptom state of the Western Ontario and McMaster Universities Osteoarthritis Index in patients who underwent total hip and total knee replacement.   Osteoarthritis Cartilage. 2019;27(10):1408-1419. doi:10.1016/j.joca.2019.05.002PubMedGoogle ScholarCrossref
    57.
    Juhl  C, Christensen  R, Roos  EM, Zhang  W, Lund  H.  Impact of exercise type and dose on pain and disability in knee osteoarthritis: a systematic review and meta-regression analysis of randomized controlled trials.   Arthritis Rheumatol. 2014;66(3):622-636. doi:10.1002/art.38290 PubMedGoogle ScholarCrossref
    58.
    Thorstensson  CA, Garellick  G, Rystedt  H, Dahlberg  LE.  Better management of patients with osteoarthritis: development and nationwide implementation of an evidence-based supported osteoarthritis self-management programme.   Musculoskeletal Care. 2015;13(2):67-75. doi:10.1002/msc.1085 PubMedGoogle ScholarCrossref
    59.
    Jönsson  T, Eek  F, Dell’Isola  A, Dahlberg  LE, Ekvall Hansson  E.  The better management of patients with osteoarthritis program: outcomes after evidence-based education and exercise delivered nationwide in Sweden.   PLoS One. 2019;14(9):e0222657. doi:10.1371/journal.pone.0222657 PubMedGoogle Scholar
    60.
    Thorlund  JB, Roos  EM, Goro  P, Ljungcrantz  EG, Grønne  DT, Skou  ST.  Patients use fewer analgesics following supervised exercise therapy and patient education: an observational study of 16 499 patients with knee or hip osteoarthritis.   Br J Sports Med. 2020;bjsports-2019-101265. doi:10.1136/bjsports-2019-101265 PubMedGoogle Scholar
    61.
    Polaski  AM, Phelps  AL, Kostek  MC, Szucs  KA, Kolber  BJ.  Exercise-induced hypoalgesia: a meta-analysis of exercise dosing for the treatment of chronic pain.   PLoS One. 2019;14(1):e0210418. doi:10.1371/journal.pone.0210418 PubMedGoogle Scholar
    62.
    Castro Sweet  CM, Chiguluri  V, Gumpina  R,  et al.  Outcomes of a digital health program with human coaching for diabetes risk reduction in a Medicare population.   J Aging Health. 2018;30(5):692-710. doi:10.1177/0898264316688791 PubMedGoogle ScholarCrossref
    63.
    Toelle  TR, Utpadel-Fischler  DA, Haas  K-K, Priebe  JA.  App-based multidisciplinary back pain treatment versus combined physiotherapy plus online education: a randomized controlled trial.   NPJ Digit Med. 2019;2:34. doi:10.1038/s41746-019-0109-x PubMedGoogle ScholarCrossref
    64.
    Shebib  R, Bailey  JF, Smittenaar  P, Perez  DA, Mecklenburg  G, Hunter  S.  Randomized controlled trial of a 12-week digital care program in improving low back pain.   NPJ Digit Med. 2019;2:1. doi:10.1038/s41746-018-0076-7 PubMedGoogle ScholarCrossref
    65.
    Fowkes  J, Fross  C, Gilbert  G, Harris  A. Virtual health: a look at the next frontier of care delivery. McKinsey & Company. June 11, 2020. Accessed January 13, 2021. https://www.mckinsey.com/industries/healthcare-systems-and-services/our-insights/virtual-health-a-look-at-the-next-frontier-of-care-delivery
    66.
    Skou  ST, Roos  EM, Laursen  MB,  et al.  A randomized, controlled trial of total knee replacement.   N Engl J Med. 2015;373(17):1597-1606. doi:10.1056/NEJMoa1505467PubMedGoogle ScholarCrossref
    ×