Physical therapy vs internet-based exercise training for patients with knee osteoarthritis: results of a randomized controlled trial

  • K.D. Allen
    Correspondence
    Address correspondence and reprint requests to: K.D. Allen, Thurston Arthritis Research Center, University of North Carolina at Chapel Hill, 3300 Thurston Bldg., CB# 7280, Chapel Hill, NC 27599, USA.
    Affiliations
    Thurston Arthritis Research Center, University of North Carolina at Chapel Hill, 3300 Thurston Bldg., CB# 7280, Chapel Hill, NC 27599, USA

    Department of Medicine, University of North Carolina at Chapel Hill, 125 MacNider Hall, CB# 7005 Chapel Hill, NC 27599, USA

    Center for Health Services Research in Primary Care, Durham VA Medical Center, Durham, NC, USA
    Search for articles by this author
  • L. Arbeeva
    Affiliations
    Thurston Arthritis Research Center, University of North Carolina at Chapel Hill, 3300 Thurston Bldg., CB# 7280, Chapel Hill, NC 27599, USA

    Department of Medicine, University of North Carolina at Chapel Hill, 125 MacNider Hall, CB# 7005 Chapel Hill, NC 27599, USA
    Search for articles by this author
  • L.F. Callahan
    Affiliations
    Thurston Arthritis Research Center, University of North Carolina at Chapel Hill, 3300 Thurston Bldg., CB# 7280, Chapel Hill, NC 27599, USA

    Department of Medicine, University of North Carolina at Chapel Hill, 125 MacNider Hall, CB# 7005 Chapel Hill, NC 27599, USA
    Search for articles by this author
  • Y.M. Golightly
    Affiliations
    Thurston Arthritis Research Center, University of North Carolina at Chapel Hill, 3300 Thurston Bldg., CB# 7280, Chapel Hill, NC 27599, USA

    Injury Prevention Research Center, University of North Carolina at Chapel Hill, Chapel Hill, NC, USA

    Department of Epidemiology, University of North Carolina at Chapel Hill, Chapel Hill, NC, USA
    Search for articles by this author
  • A.P. Goode
    Affiliations
    Duke Clinical Research Institute, Durham, NC, USA

    Department of Orthopedic Surgery, Division of Physical Therapy, Duke University Medical Center, USA
    Search for articles by this author
  • B.C. Heiderscheit
    Affiliations
    Department of Orthopedics and Rehabilitation, University of Wisconsin–Madison, Madison, WI, USA
    Search for articles by this author
  • K.M. Huffman
    Affiliations
    Department of Medicine, Division of Rheumatology, Duke University Medical Center, Durham, NC, USA

    Physical Medicine and Rehabilitation Service, Durham VA Medical Center, Durham, NC, USA
    Search for articles by this author
  • H.H. Severson
    Affiliations
    Oregon Research Institute, Eugene, OR, USA
    Search for articles by this author
  • T.A. Schwartz
    Affiliations
    Thurston Arthritis Research Center, University of North Carolina at Chapel Hill, 3300 Thurston Bldg., CB# 7280, Chapel Hill, NC 27599, USA

    Department of Biostatistics, Gillings School of Global Public Health, University of North Carolina at Chapel Hill, Chapel Hill, NC, USA

    School of Nursing, University of North Carolina at Chapel Hill, Chapel Hill, NC, USA
    Search for articles by this author
Open ArchivePublished:January 04, 2018DOI:https://doi.org/10.1016/j.joca.2017.12.008

      Summary

      Objective

      To compare the effectiveness of physical therapy (PT, evidence-based approach) and internet-based exercise training (IBET), each vs a wait list (WL) control, among individuals with knee osteoarthritis (OA).

      Design

      Randomized controlled trial of 350 participants with symptomatic knee OA, allocated to standard PT, IBET and WL control in a 2:2:1 ratio, respectively. The PT group received up to eight individual visits within 4 months. The IBET program provided tailored exercises, video demonstrations, and guidance on progression. The primary outcome was the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC, range 0 [no problems]–96 [extreme problems]), assessed at baseline, 4 months (primary time point) and 12 months. General linear mixed effects modeling compared changes in WOMAC among study groups, with superiority hypotheses testing differences between each intervention group and WL and non-inferiority hypotheses comparing IBET with PT.

      Results

      At 4-months, improvements in WOMAC score did not differ significantly for either the IBET or PT group compared with WL (IBET: −2.70, 95% Confidence Interval (CI) = −6.24, 0.85, P = 0.14; PT: −3.36, 95% (CI) = −6.84, 0.12, P = 0.06). Similarly, at 12-months mean differences compared to WL were not statistically significant for either group (IBET: −2.63, 95% CI = −6.37, 1.11, P = 0.17; PT: −1.59, 95% CI = −5.26, 2.08, P = 0.39). IBET was non-inferior to PT at both time points.

      Conclusions

      Improvements in WOMAC score following IBET and PT did not differ significantly from the WL group. Additional research is needed to examine strategies for maximizing benefits of exercise-based interventions for patients with knee OA.

      Trial registration

      Keywords

      Introduction

      Exercise is a recommended component of treatment for knee osteoarthritis (OA), based on studies showing improvements in pain, function and other outcomes
      • Meneses S.R.
      • Goode A.P.
      • Nelson A.E.
      • Lin J.
      • Jordan J.M.
      • Allen K.D.
      • et al.
      Clinical algorithms to aid osteoarthritis guideline dissemination.
      • Nelson A.E.
      • Allen K.D.
      • Golightly Y.M.
      • Goode A.P.
      • Jordan J.M.
      A systematic review of recommendations and guidelines for the management of osteoarthritis: the Chronic Osteoarthritis Management Initiative of the U.S. Bone and Joint Initiative.
      • Fransen M.
      • McConnell S.
      • Harmer A.R.
      • Van der Esch M.
      • Simic M.
      • Bennell K.L.
      Exercise for osteoarthritis of the knee: a Cochrane systematic review.
      . However, the majority of adults with OA are inactive, highlighting the continued need for increasing regular engagement in exercise
      • Fontaine K.R.
      • Heo M.
      Changes in the prevalence of U.S. adults with arthritis who meet physical activity recommendations, 2001-2003.
      • Shih V.C.
      • Song J.
      • Chang R.W.
      • Dunlop D.D.
      Racial differences in activities of daily living limitation onset in older adults with arthritis: a national cohort study.
      . Physical therapists can play a key role in instructing patients with OA in an appropriate exercise program (as well as deliver other treatments such as orthotics, braces, gait aids and manual therapy), and physical therapy (PT) care is a recommended, evidence-based component of knee OA treatment
      • Nelson A.E.
      • Allen K.D.
      • Golightly Y.M.
      • Goode A.P.
      • Jordan J.M.
      A systematic review of recommendations and guidelines for the management of osteoarthritis: the Chronic Osteoarthritis Management Initiative of the U.S. Bone and Joint Initiative.
      • Wang S.Y.
      • Olson-Kellogg B.
      • Shamliyan T.A.
      • Choi J.Y.
      • Ramakrishnan R.
      • Kane R.L.
      Physical therapy interventions for knee pain secondary to osteoarthritis: a systematic review.
      . However, PT is underutilized for knee OA
      • Dhawan A.
      • Mather 3rd, R.C.
      • Karas V.
      • Ellman M.B.
      • Young B.B.
      • Bach Jr., B.R.
      • et al.
      An epidemiologic analysis of clinical practice guidelines for non-arthroplasty treatment of osteoarthritis of the knee.
      • Cisternas M.G.
      • Yelin E.
      • Katz J.N.
      • Solomon D.H.
      • Wright E.A.
      • Losina E.
      Ambulatory visit utilization in a national, population-based sample of adults with osteoarthritis.
      • Allen K.D.
      • Bosworth H.B.
      • Chatterjee R.
      • Coffman C.J.
      • Corsino L.
      • Jeffreys A.S.
      • et al.
      Clinic variation in recruitment metrics, patient characteristics and treatment use in a randomized clinical trial of osteoarthritis management.
      , partly due to health care access-related issues, particularly for uninsured and under-insured patients and those in medically underserved areas. Individuals with lower socioeconomic status likely have the least access to a physical therapist or a supervised exercise program, yet these individuals also bear a greater burden of OA
      • Vavken P.
      • Dorotka R.
      The burden of musculoskeletal disease and its determination by urbanicity, socioeconomic status, age, and gender - results from 14507 subjects.
      • Ackerman I.N.
      • Busija L.
      Access to self-management education, conservative treatment and surgery for arthritis according to socioeconomic status.
      .
      Internet-based programs are an alternative, low-cost method for providing instruction and support in appropriate exercise. However, there has been little research on internet-based exercise programs for patients with OA
      • Li L.C.
      • Lineker S.
      • Cibere J.
      • Crooks V.A.
      • Jones C.A.
      • Kopec J.A.
      • et al.
      Capitalizing on the teachable moment: osteoarthritis physical activity and exercise net for improving physical activity in early knee osteoarthritis.
      • Bennell K.L.
      • Nelligan R.
      • Dobson F.
      • Rini C.
      • Keefe F.
      • Kasza J.
      • et al.
      Effectiveness of an internet-delivered exercise and pain-coping skills training intervention for persons with chronic knee pain: a randomized trial.
      • Brooks M.A.
      • Beaulieu J.E.
      • Severson H.H.
      • Wille C.M.
      • Cooper D.
      • Gau J.M.
      • et al.
      Web-based therapeutic exercise resource center as a treatment for knee osteoarthritis: a prospective cohort pilot study.
      • Bossen D.
      • Veenhof C.
      • Van Beek K.E.
      • Spreeuwenberg P.M.
      • Dekker J.
      • De Bakker D.H.
      Effectiveness of a web-based physical activity intervention in patients with knee and/or hip osteoarthritis: randomized controlled trial.
      or older adults
      • Irvine A.B.
      • Gelatt V.A.
      • Seeley J.R.
      • Macfarlane P.
      • Gau J.M.
      Web-based intervention to promote physical activity by sedentary older adults: randomized controlled trial.
      • Carr L.J.
      • Dunsiger S.I.
      • Lewis B.
      • Ciccolo J.T.
      • Hartman S.
      • Bock B.
      • et al.
      Randomized controlled trial testing an internet physical activity intervention for sedentary adults.
      . Further, there have been no direct comparisons of internet-based exercise programs with supervised exercise-based therapies for knee OA. This study compared the effectiveness of an internet-based exercise training (IBET) program to in-person PT among individuals with symptomatic knee OA. Specifically, this study tested whether PT or IBET were superior to a wait list (WL) control group at 4-month (primary time point) and 12-month follow-up. Additionally, analyses examined whether IBET was non-inferior to PT.

      Methods

      This study was approved by the Institutional Review Boards of the University of North Carolina at Chapel Hill (UNC) and Duke University Medical Center. Detailed methods have been published
      • Williams Q.I.
      • Gunn A.H.
      • Beaulieu J.E.
      • Benas B.C.
      • Buley B.
      • Callahan L.F.
      • et al.
      Physical therapy vs internet-based exercise training (PATH-IN) for patients with knee osteoarthritis: study protocol of a randomized controlled trial.
      . Recruitment occurred from November 2014 to February 2016, and follow-up assessments were completed in February 2017.

       Study design

      The PT vs IBET for Patients with Knee Osteoarthritis (PATH-IN) study was a pragmatic randomized controlled trial with participants assigned to standard PT, IBET and WL control, with allocation of 2:2:1, respectively. Randomization schedules were computer generated by a statistician with stratification by recruitment source (UNC Healthcare system, Johnston County Osteoarthritis Project
      • Jordan J.M.
      • Helmick C.G.
      • Renner J.B.
      • Luta G.
      • Dragomir A.D.
      • Woodard J.
      • et al.
      Prevalence of knee symptoms and radiographic and symptomatic knee osteoarthritis in African Americans and Caucasians: the Johnston County Osteoarthritis Project.
      and self-referral). Participants continued with usual medical care for OA. Participants in the WL group did not receive PT or IBET during the study but were offered two PT visits and access to IBET following 12-month assessments.

       Participants and recruitment

      Study inclusion criteria were: (1) Radiographic evidence of knee OA, physician diagnosis of knee OA in the medical record, or self-report of physician diagnosis along with items based on the American College of Rheumatology clinical criteria
      • Altman R.
      • Asch D.
      • Bloch G.
      • Bole D.
      • Borenstein K.
      • Brandt K.
      • et al.
      The American College of Rheumatology criteria for the classification and reporting of osteoarthritis of the knee.
      . (2) Self-report of pain, aching or stiffness in one or both knees on most days of the week. Exclusion criteria are shown in Box 1. Participants were recruited using two methods: (1) Active recruitment of patients with evidence of knee OA in the UNC medical record, as well as participants with knee OA in the Johnston County Osteoarthritis Project
      • Jordan J.M.
      • Helmick C.G.
      • Renner J.B.
      • Luta G.
      • Dragomir A.D.
      • Woodard J.
      • et al.
      Prevalence of knee symptoms and radiographic and symptomatic knee osteoarthritis in African Americans and Caucasians: the Johnston County Osteoarthritis Project.
      ; these individuals were mailed an introductory letter, with telephone follow-up. (2) Advertisement within UNC and the surrounding communities. All individuals who met eligibility criteria based on telephone-based screening completed consent and baseline assessments in person. Participants were then given their randomization assignment via telephone by the study coordinator.
      Exclusion Criteria.
      • No regular internet access
      • Currently meeting Department of Health and Human Services Guidelines for physical activity
      • Currently completing series of PT visits for knee OA
      • Diagnosis of gout in the knee, rheumatoid arthritis, fibromyalgia, or other systemic rheumatic disease
      • Severe dementia or other memory loss condition
      • Active diagnosis of psychosis or current uncontrolled substance abuse disorder
      • On waiting list for arthroplasty
      • Hospitalization for a stroke, heart attack, heart failure, or had surgery for blocked arteries in the past 3 months
      • Total joint replacement knee surgery, other knee surgery, meniscus tear, or ACL tear in the past 6 months
      • Severely impaired hearing or speech
      • Unable to speak English
      • Serious or terminal illness as indicated by referral to hospice or palliative care
      • Other health problem that would prohibit participation in the study
      • Nursing home residence
      • Current participation in another OA intervention
      • Fall history deemed by a study physical therapist co-investigator to impose risk for potential injury with participation in a home-based exercise program study

       IBET program

      The IBET program was developed by Visual Health Information and a multidisciplinary team, including physical therapists, physicians and patients; details have been described
      • Brooks M.A.
      • Beaulieu J.E.
      • Severson H.H.
      • Wille C.M.
      • Cooper D.
      • Gau J.M.
      • et al.
      Web-based therapeutic exercise resource center as a treatment for knee osteoarthritis: a prospective cohort pilot study.
      . Features of the IBET program include: (1) Tailored Exercises based on measures regarding pain, function and current activity, along with an algorithm that assigns participants to one of seven different exercise levels. Exercise routines include strengthening, stretching and aerobic activity recommendations. (2) Exercise Progression recommendations, based on serial measures of pain and function. (3) Video Display of Exercises (and photographs) to demonstrate proper exercise performance. (4) Automated Reminders to engage with the website and remain active if participants have not logged in for 7 days. (5) Progress Tracking, including graphs of pain, function, and exercise over time. Participants were asked to access the IBET site as soon as they were randomized and to continue through the 12-month follow-up assessment. In accordance with current Department of Health and Human Services and other guidelines for physical activity
      • U.S. Department of Health and Human Services
      Physical Activity Guidelines for Americans.
      , participants were encouraged to complete strengthening and stretching exercises at least 3 times per week and to engage in aerobic exercises daily, or as often as possible.

       PT

      Physical therapists (with experience in treating OA) at multiple clinics administered the intervention following training by PT co-investigators (YMG, APG), who also performed periodic fidelity checks. The PT intervention, described in detail elsewhere
      • Williams Q.I.
      • Gunn A.H.
      • Beaulieu J.E.
      • Benas B.C.
      • Buley B.
      • Callahan L.F.
      • et al.
      Physical therapy vs internet-based exercise training (PATH-IN) for patients with knee osteoarthritis: study protocol of a randomized controlled trial.
      , was modeled after recommended elements of care provided to patients with knee OA
      • Hertling D.
      • Kessler R.M.
      Knee.
      , including: (1) evaluation of strength, flexibility, mobility, balance, function, knee alignment, and possible limb length inequality; (2) evaluation of the need for assistive devices, knee braces, patellar taping, heel lifts, shoe wedges and other footwear modifications; (3) instruction in an appropriate home exercise program (including strengthening, stretching/range of motion, and aerobic exercises); (4) instruction in activity pacing and joint protection; (5) manual therapy, if appropriate; (6) modalities for pain management, if appropriate. Emphasis was placed on the home exercise program, which was initiated at the first visit. To mirror standard clinical practice, physical therapists were permitted to tailor visits to patients' needs and functional limitations. Based on a typical range of outpatient PT visits for knee OA, study participants could receive up to eight one-hour sessions. At the first visit, physical therapists completed a standardized evaluation form and documented treatment provided. At subsequent visits, physical therapists completed progress notes including documentation of treatment provided. The Appendix lists the guidance given to physical therapists.

       Measures

      Baseline, 4-month and 12-month assessments were conducted by trained research assistants blinded (via database restrictions) to participants' randomization assignment. Assessments were typically conducted in person, though telephone-based follow-up assessments were permitted in cases where participants are unable to return to the study site. Participants were paid $30 for completion of assessments at each time point.

       Primary outcome: Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) total score

      The WOMAC is a measure of lower extremity pain (five items), stiffness (two items), and function (17 items)
      • Bellamy N.
      WOMAC: a 20-year experiential review of a patient-centered self-reported health status questionnaire.
      • Bellamy N.
      • Buchanan W.W.
      • Goldsmith C.H.
      • Campbell J.
      • Stitt L.W.
      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.
      . All items were rated on a Likert scale of 0 (no symptoms)–4 (extreme symptoms), with a total range of 0–96 and higher scores indicating worse symptoms.

       Secondary outcomes

      We examined the WOMAC pain (range 0 [no pain]–20 [extreme pain]) and function (range 0 [no difficulty]−68 [extreme difficulty]) subscales separately. We also conducted four tests of physical function: the 30-second chair stand
      • Jones C.
      • RIkli R.
      • Beam W.A.
      30-s chair-stand test as a measure of lower body strength in community-residing older adults.
      , the Timed Up and Go Test (TUG)
      • Wall J.C.
      • Bell C.
      • Campbell S.
      • Davis J.
      The Timed Get-up-and-Go test revisited: measurement of the component tasks.
      • Podsiadlo D.
      • Richardson S.
      The Timed up & Go: a test of basic functional mobility for frail elderly persons.
      , a two-minute step test
      • RIkli R.
      • Jones J.
      Development and validation of a functional fitness test for community-residing older adults.
      , and unilateral stand time
      • Scott V.
      • Votova K.
      • Scanlan A.
      • Close J.
      Multifactorial and functional mobility assessment tools for fall risk among older adults in community, home-support, long-term and acute care settings.
      • Rossiter-Fornoff J.E.
      • Wolf S.L.
      • Wolfson L.I.
      • Buchner D.M.
      A cross-sectional validation study of the FICSIT common data base static balance measures. Frailty and injuries: Cooperative studies of intervention techniques.
      . The latter was part of the Four-Stage Balance Test
      • Rossiter-Fornoff J.E.
      • Wolf S.L.
      • Wolfson L.I.
      • Buchner D.M.
      A cross-sectional validation study of the FICSIT common data base static balance measures. Frailty and injuries: Cooperative studies of intervention techniques.
      , and participants scored a “0” if they were unable to stand for 10 s in side-by-side, semi-tandem or tandem positions. Self-reported physical activity was assessed with the Physical Activity Scale for the Elderly (PASE), which measures occupational, household, and leisure activities during a 1-week period; the typical range for the total PASE score is 0–400, with higher scores indicating greater activity
      • Washburn R.A.
      • Smith K.W.
      • Jette A.M.
      • Janney C.A.
      The physical activity scale for the elderly (PASE): development and evaluation.
      . In addition, participants self-reported their current minutes per week of stretching, strengthening and aerobic exercise. Participants' Global Assessment of Change in right and left knee (separately) pain, aching and stiffness was reported at follow-up assessments. This scale ranged from −6 (a very great deal worse) to +6 (a very great deal better); data were coded as missing if participants never had symptoms in that knee or responded “don't know.”

       Intervention delivery

      We report the number of days on which participants in the IBET group logged into the website and the number of PT visits attended. For each participant in the PT group we calculated the proportion of visits at which the therapist reported delivering specific interventions.

       Non-study PT visits

      At 4-month and 12-month follow-up, we asked participants whether they received PT care for knee OA outside the study since their last visit. This informed per-protocol analyses described below.

       Demographic and clinical characteristics

      Self-reported participant characteristics included age, race/ethnicity, gender, household financial state (with low income defined as self-report of “just meeting basic expenses” or “don't even have enough to meet basic expenses”), education level (bachelor's degree vs less education), work status (employed vs not working), marital status, joints affected by OA, duration of OA symptoms, self-rated health (excellent, very good or good vs fair or poor), and depressive symptoms (Patient Health Questionnaire-8)
      • Kroenke K.
      • Strine T.W.
      • Spitzer R.L.
      • Williams J.B.
      • Berry J.T.
      • Mokdad A.H.
      The PHQ-8 as a measure of current depression in the general population.
      . Height and weight were measured at baseline to calculate body mass index (BMI). Participants also self-reported use of other OA treatments at both follow-up time points, including pain medication use, knee injections, knee brace use, and topical creams.

       Adverse event assessment

      Adverse events were identified through regular reports of participants' visits to the UNC healthcare system, as well as through participants' reports to study team members.

       Sample size

      As detailed elsewhere
      • Meneses S.R.
      • Goode A.P.
      • Nelson A.E.
      • Lin J.
      • Jordan J.M.
      • Allen K.D.
      • et al.
      Clinical algorithms to aid osteoarthritis guideline dissemination.
      , the sample size estimate of N = 350 was based on the hypothesis of non-inferiority, which is the most conservative
      • Greene C.J.
      • Morland L.A.
      • Durkalski V.L.
      • Frueh B.C.
      Noninferiority and equivalence designs: issues and implications for mental health research.
      • Blackwelder W.C.
      “Proving the null hypothesis” in clinical trials.
      • Pocock S.J.
      The pros and cons of noninferiority trials.
      , and on the 2:2:1 randomization ratio
      • Piantadosi S.
      Clinical Trials: A Methodological Perspective.
      . A one-sided, two-sample t test sample size calculation was used at the 0.025 significance level for the difference in mean WOMAC between IBET and PT to be less than five points at 4-month follow-up, with an adjustment to the variance to account for repeated measures
      • Borm G.F.
      • Fransen J.
      • Lemmens W.
      A simple sample size formula for analysis of covariance in randomized clinical trials.
      and potential 10% attrition.

       Data analyses

      We tested four hypotheses: H1: Participants who receive either IBET or standard PT will have clinically relevant improvements in WOMAC at 4-month follow-up, compared with WL control group. H2: IBET will be non-inferior to PT (an intervention with established evidence
      • Wang S.Y.
      • Olson-Kellogg B.
      • Shamliyan T.A.
      • Choi J.Y.
      • Ramakrishnan R.
      • Kane R.L.
      Physical therapy interventions for knee pain secondary to osteoarthritis: a systematic review.
      ) at 4-month follow-up, indicated by a mean WOMAC score less than five points higher (worse) than PT. For total WOMAC scores, a five point non-inferiority margin was selected because it is on the border of what would be considered a clinically important effect in this context
      • Angst F.
      • Aeschlimann A.
      • Michel B.A.
      • Stucki G.
      Minimal clinically important rehabilitation effects in patients with osteoarthritis of the lower extremity.
      • Bellamy N.
      • Hochberg M.
      • Tubach F.
      • Martin-Mola E.
      • Awada H.
      • Bombardier C.
      • et al.
      Development of multinational definitions of minimal clinically important improvement and patient acceptable symptomatic state in osteoarthritis.
      . H3 and H4 mirrored H1 and H2 but at the 12-month follow-up time point. For the superiority hypotheses (H1, H3), primary conclusions were based on intention-to-treat (ITT) analyses, with participants assigned to the arm to which they were randomized, regardless of adherence, using all available follow-up data
      • ICH E9 Expert Working Group
      ICH harmonised tripartite guideline - statistical principles for clinical trials.
      . For non-inferiority hypotheses (H2, H4), the ITT analysis would not necessarily be the conservative approach
      • Snapinn S.M.
      Noninferiority trials.
      . We therefore performed analyses on both an ITT and per-protocol basis
      • Pocock S.J.
      The pros and cons of noninferiority trials.
      • Piaggio G.
      • Elbourne D.R.
      • Altman D.G.
      • Pocock S.J.
      • Evans S.J.
      • Group C.
      Reporting of noninferiority and equivalence randomized trials: an extension of the CONSORT statement.
      . For the latter we excluded individuals who did not adhere to their assigned study group, including those in the PT group (N = 9) who attended no visits, those in the IBET group who did not log on to the website (N = 28), and those in the IBET (N = 5) and WL (N = 4) groups who received PT outside the study.
      A general linear mixed effects model was fitted with changes from baseline in WOMAC scores as the dependent variables with an unstructured covariance matrix to account for the two follow-up repeated measures. Fixed effects included follow-up time, intervention group, their interaction, baseline WOMAC score, and enrollment source. The SAS MIXED procedure (Cary, NC) was used to fit these models and to test linear contrasts corresponding to each hypothesis. Participants missing either follow-up measurement were still included in the model under a ‘missing at random' paradigm. Sensitivity analyses were also conducted through multiple imputation of missing values (see below).
      To test the null hypothesis of non-inferiority of IBET vs standard PT at 4 months in management of OA symptoms, the 95% confidence interval (CI) of the appropriate linear contrast was constructed; non-inferiority was concluded if the upper limit of the interval was less than the non-inferiority margin of five points
      • Piaggio G.
      • Elbourne D.R.
      • Altman D.G.
      • Pocock S.J.
      • Evans S.J.
      • Group C.
      Reporting of noninferiority and equivalence randomized trials: an extension of the CONSORT statement.
      . Superiority hypotheses involved two comparisons vs WL control, so each was conducted at the two-sided 0.025 significance level. The non-inferiority hypotheses involve only one comparison and were tested at the full one-sided 0.025 significance level.
      We had several strategies for handling missing data. When individual items were missing from self-report scales, we followed guidelines regarding when to impute scores
      • Bellamy N.
      WOMAC Osteoarthritis Index User Guide.
      . When guidelines were unavailable, we treated the scale as missing if >1 item was missing; when one item was missing we substituted with the mean of available items. When participants declined or could not complete function tests they were assigned the lowest value for that test; when participants ran out of time to complete function tests or assessments were completed via telephone, data were treated as missing. In some cases (four at 4-months, five at 12-months), our data coding scheme did not allow us to differentiate between these two situations; these were treated as missing. For sensitivity analysis for the ITT approach, we performed multiple imputation to deal with missing data at follow-up assessments via the SAS MI and MIANALYZE procedures, specifying 30 imputations. First, three missing race values were imputed based on other participant characteristics. Then we identified baseline characteristics that differed between completers and non-completers at follow-up at the P ≤ 0.25 level. These characteristics were used to impute missing baseline WOMAC values. Next, missing 4-month WOMAC scores were imputed as a function of these baseline characteristics, baseline WOMAC score, and treatment group; imputation of 12-month WOMAC score also included 4-month WOMAC scores. The same imputation process was followed for secondary outcomes.
      Corresponding analytic strategies were used for secondary outcomes, though there was insufficient information in the literature to define a non-inferiority margin for these measures. Additionally, as the Global Assessment of Change variables do not have baseline values, the actual values (rather than change from baseline) were managed as the response variable, with no baseline score as a covariate. A square root transformation was used for the weekly minutes of exercise variables to improve the residuals with respect to the normality assumption. To provide comparison with prior studies, we calculated standardized mean differences (SMDs) for WOMAC total and subscale scores for both intervention groups compared to WL (ratio of model-predicted mean group differences to their pooled standard deviation (SD)).

      Results

       Participants and retention

      We identified 11,274 potential participants from all recruitment sources (Fig. 1). Of 683 who completed telephone screening, 350 (51%) were eligible, enrolled and randomized. Because randomization was stratified by enrollment source, allocation across groups was slightly different than the 2:2:1 ratio, with 142 participants assigned to the IBET group, 140 to the PT group and 68 to the WL group. At both 4-month and 12-month follow-up, 86% of participants completed primary outcomes (Fig. 1). Compared with participants who completed follow-up assessments for the primary outcome at 12-months, non-completers had a higher baseline mean WOMAC total score (31.2, SD = 17.6 vs 37.6, SD = 19.1, respectively). Participant characteristics are shown in Table I. Participants' use of other OA treatments at follow-up was similar across groups (Appendix Table I).
      Table IParticipant characteristics at baseline
      Values are Mean (SD) unless otherwise specified.
      CharacteristicAll participants (N = 350)IBET group (N = 142)PT group (N = 140)WL control (N = 68)
      Age, years65.3 (11.1)65.3 (11.5)65.7 (10.3)64.3 (12.2)
      Women N (%)251 (71.7%)98 (69%)100 (71.4%)53 (77.9%)
      Non-white race N (%)95 (27.4%)48 (33.8%)29 (21%)18 (26.9%)
      Married or living with partner N (%)215 (61.4%)93 (65.5%)80 (57.1%)42 (61.8%)
      Bachelors degree N (%)208 (59.4%)80 (56.3%)86 (61.4%)42 (61.8%)
      Employed N (%)141 (40.3%)51 (35.9%)59 (42.1%)31 (45.6%)
      Household financial status: low income N (%)62 (17.8%)29 (20.6%)20 (14.3%)13 (19.1%)
      Fair or poor health N (%)48 (13.7%)22 (15.5%)14 (10%)12 (17.6%)
      BMI, kg/m231.4 (8)31.5 (7.8)31.9 (8.6)30.1 (7.3)
      Joints with OA symptoms5.4 (3.2)5.2 (3.1)5.5 (3)5.5 (3.9)
      Duration of OA symptoms, years13.1 (11.7)11.6 (11)14.1 (11.6)14.2 (13)
      PHQ-8 score3.8 (4.1)3.7 (4.1)4 (4.5)3.6 (3.5)
      WOMAC total32.0 (17.9)31.3 (17.5)32 (17.7)33.6 (19.2)
      WOMAC pain subscale6.1 (3.8)6.0 (3.9)6.1 (3.5)6.2 (4.0)
      WOMAC function subscale22.5 (13.0)21.8 (12.7)22.6 (12.9)23.9 (13.8)
      PASE total score126.9 (72.7)132.3 (71.2)121.4 (72)126.9 (77.2)
      PASE household score75.2 (40.7)81.6 (41.3)70.4 (40.4)71.8 (38.8)
      PASE leisure score21.6 (21.9)22.4 (21.9)20.9 (23.2)21.5 (19.7)
      PASE work score30.7 (51.1)30.5 (51.5)29.1 (48.4)34.2 (55.9)
      Timed Up and Go, seconds11.9 (4.3)12 (4.6)11.9 (4.2)11.6 (3.7)
      Unilateral stand test, seconds7.2 (3.6)7.3 (3.5)7.3 (3.6)6.7 (3.7)
      30 s chair stand9.5 (3.9)9.5 (3.8)9.5 (4.2)9.6 (3.5)
      Missing Data: non-white race = 3, household financial status = 1, WOMAC = 2, Timed Up and Go = 4, PASE total = 10, PASE Household = 5, PASE work = 1, PASE Leisure = 6.
      PHQ = Patient Health Questionnaire; WOMAC = Western Ontario and McMaster Universities Osteoarthritis Index; PASE = Physical Activity Scale for the Elderly.
      Values are Mean (SD) unless otherwise specified.

       Adverse events

      There were four non-serious study-related events in the PT group (one fall, three increased knee pain) and four in the IBET group (two increased knee pain, one shoulder pain, one ankle pain).

       Intervention delivery

      Between baseline and 4-month follow-up, 114 (80%) of participants in the IBET group logged onto the website; the mean (SD) number of days logged on was 20.7 (24.6), median = 9.5. Between baseline and 12-month follow-up, 115 (81%) of participants in the IBET group logged onto the website with a mean (SD) number of days logged on of 40.5 (59.8) median = 10.5. The mean number of days logged onto the website between 4-month and 12-month follow-up was 19.8 (37.7), median = 0. Seven physical therapists contributed to intervention delivery, with numbers of participants treated by each PT ranging from 2 to 40; this wide range was primarily due to participants' geographic proximity to the different study PT clinics. Among participants in the PT group, 131 (94%) attended at least one visit; 51% attended 6–8 visits. The mean (SD) number of visits was 5.7 (2.5), with a median of 7.0 visits. The mean proportions of visits per patient at which therapists reported delivering specific intervention components were: Therapeutic Exercise – 94%; Balance/Neuromuscular Education – 38%; Manual Therapy – 43%; Gait/Strength Training – 44%; Modalities – 29%; and Shoes/Wedges – 20%.

       Primary outcome: WOMAC total score

       ITT analyses

       Superiority hypotheses

      Neither IBET nor PT were superior to WL at 4 months or 12 months at the specified P < 0.025 (Table II, Fig. 2). Multiple imputation analyses showed similar results for both interventions (Appendix Table II). At 4-months, the SMDs for PT and IBET group, respectively were −0.26 (−0.53, 0.00) and −0.20 (−0.48, 0.07), compared to the WL group (Table III). At 12-months, the SMDs for PT and IBET group, respectively were −0.12 (−0.40, 0.16) and −0.19 (−0.46, 0.09), compared to the WL group.
      Table IIWithin- and between-group mean changes in outcomes and 95% CIs: Results of ITT analyses
      OutcomeBaseline to 4-month difference (95% CI)Difference in baseline to 4-month vs WL (95% CI), P-valueBaseline to 12-month difference (95% CI)Difference in baseline to 12-month vs WL (95% CI), P-value
      WOMAC total (N = 348)
      Indicates number included in the statistical model for that outcome.
       WL−3.37 (−6.33, −0.41)−2.83 (−5.93, 0.27)
       PT−6.73 (−8.86, −4.6)−3.36 (−6.84, 0.12), 0.06−4.42 (−6.66, −2.17)−1.59 (−5.26, 2.08), 0.39
       IBET−6.06 (−8.29, −3.84)−2.70 (−6.24, 0.85), 0.14−5.46 (−7.82, −3.09)−2.63 (−6.37, 1.11), 0.17
      WOMAC function (N = 348)
       WL−2.3 (−4.46, −0.14)−1.51 (−3.76, 0.74)
       PT−4.77 (−6.32, −3.23)−2.48 (−5.02, 0.07), 0.06−3.3 (−4.91, −1.68)−1.79 (−4.45, 0.87), 0.19
       IBET−3.74 (−5.36, −2.12)−1.44 (−4.03, 1.15), 0.27−3.4 (−5.11, −1.7)−1.90 (−4.61, 0.82), 0.17
      WOMAC pain (N = 350)
       WL−0.66 (−1.41, 0.09)−0.64 (−1.38, 0.09)
       PT−1.11 (−1.65, −0.58)−0.45 (−1.33, 0.42), 0.31−0.7 (−1.23, −0.16)−0.05 (−0.92, 0.81), 0.90
       IBET−1.59 (−2.15, −1.02)−0.93 (−1.82, −0.03), 0.04−1.15 (−1.71, −0.59)−0.51 (−1.39, 0.38), 0.26
      PASE total (N = 340)
       WL−4.7 (−21.04, 11.64)1.17 (−13.11, 15.45)
       PT2.25 (−9.18, 13.68)6.95 (−12.31, 26.22) , 0.488.28 (−2.01, 18.56)7.11 (−9.69, 23.91), 0.41
       IBET−11.52 (−23.79, 0.74)−6.82 (−26.55, 12.91), 0.508.19 (−2.99, 19.37)7.02 (−10.31, 24.35), 0.43
      PASE leisure (N = 344)
       WL−2.41 (−7.49, 2.66)−0.11 (−6.25, 6.04)
       PT3.27 (−0.28, 6.82)5.68 (−0.25, 11.62), 0.068.68 (4.3, 13.05)8.78 (1.46, 16.1), 0.02
       IBET−1.34 (−5.14, 2.46)1.07 (−5.70, 0.14), 0.737.66 (2.94, 12.39)7.77 (0.25, 15.29), 0.04
      PASE household (N = 345)
       WL−5.32 (−14.49, 3.84)−3.42 (−11.59, 4.75)
       PT−1.07 (−7.50, 5.36)4.25 (−6.59, 15.09), 0.442.3 (−3.61, 8.21)5.72 (−3.94, 15.38), 0.25
       IBET−9.16 (−16.11, −2.21)−3.84 (−14.99, 7.31), 0.50−3.72 (−10.13, 2.69)−0.3 (−10.26, 9.67), 0.95
      PASE Work (N = 349)
       WL2.35 (−8.45, 13.15)5.24 (−4.49, 14.98)
       PT0.98 (−6.53, 8.48)−1.37 (−14.12, 11.37), 0.83−2.62 (−9.58, 4.34)−7.86 (−19.36, 3.63), 0.18
       IBET−2.00 (−10.01, 6)−4.35 (−17.39, 8.69), 0.515.25 (−2.2, 12.69)0.00 (−11.78, 11.79), 1.00
      Unilateral stand time (N = 350)
       WL0.04 (−0.75, 0.82)−0.09 (−0.88, 0.69)
       PT−0.59 (−1.15, −0.03)−0.63 (−1.56, 0.30), 0.19−0.05 (−0.6, 0.50)0.04 (−0.89, 0.98), 0.93
       IBET0.02 (−0.57, 0.61)−0.02 (−0.97, 0.93), 0.97−0.05 (−0.64, 0.53)0.04 (−0.91, 1.00), 0.93
      30 s chair stand (N = 350)
       WL0.18 (−0.87, 1.23)0.66 (−0.27, 1.58)
       PT−0.13 (−0.87, 0.61)−0.31 (−1.55, 0.94), 0.630.16 (−0.49, 0.82)−0.49 (−1.58, 0.60), 0.37
       IBET0.50 (−0.29, 1.28)0.32 (−0.95, 1.59), 0.620.90 (0.20, 1.60)0.24 (−0.87, 1.35), 0.67
      2 m march test (N = 350)
       WL−8.43 (−14.61, −2.24)0.00 (−6.49, 6.48)
       PT−0.68 (−5.07, 3.71)7.75 (0.43, 15.07), 0.041.11 (−3.45, 5.67)1.12 (−6.59, 8.82), 0.78
       IBET−3.54 (−8.20, 1.11)4.88 (−2.56, 12.33), 0.201.12 (−3.76, 6)1.13 (−6.74, 8.99), 0.78
      Timed Up and Go (N = 346)
       WL−0.23 (−1.24, 0.78)−0.26 (−1.4, 0.87)
       PT−0.62 (−1.34, 0.09)−0.39 (−1.58, 0.8), 0.52−0.77 (−1.57, 0.04)−0.5 (−1.86, 0.85), 0.46
       IBET−0.87 (−1.63, −0.11)−−0.64 (−1.85, 0.58), 0.30−1.49 (−2.35, −0.63)−1.22 (−2.61, 0.16), 0.08
      Weekly minutes of aerobic activity
      A square root transformation was applied due to superior diagnostics in statistical models.
       WL−0.09 (−1.53, 1.35)−1.59 (−3.21, 0.04)
       PT1 (−0.02, 2.02)1.09 (−0.61, 2.8), 0.210.48 (−0.67, 1.63)2.07 (0.13, 4), 0.04
       IBET1.79 (0.71, 2.88)1.89 (0.15, 3.62), 0.030.41 (−0.82, 1.63)1.99 (0.01, 3.97), 0.05
      Weekly minutes of stretching
      A square root transformation was applied due to superior diagnostics in statistical models.
       WL−0.4 (−1.39, 0.6)−1.34 (−2.24, −0.44)
       PT1.45 (0.76, 2.15)1.85 (0.67, 3.03), 0.000.27 (−0.37, 0.92)1.62 (0.55, 2.68), 0.00
       IBET0.97 (0.23, 1.72)1.37 (0.16, 2.57), 0.030.72 (0.04, 1.41)2.07 (0.98, 3.16), 0.00
      Weekly minutes of strengthening
      A square root transformation was applied due to superior diagnostics in statistical models.
       WL0.43 (−0.69, 1.54)−0.14 (−1.32, 1.04)
       PT1.78 (0.99, 2.57)1.36 (0.05, 2.66), 0.041.07 (0.23, 1.91)1.21 (−0.18, 2.6), 0.09
       IBET1.27 (0.44, 2.11)0.85 (−0.49, 2.19), 0.221.21 (0.32, 2.1)1.35 (−0.08, 2.78), 0.06
      Patient global assessment of change – right Knee
       WL0.15 (−0.36, 0.66)−0.18 (−0.69, 0.33)
       PT1.36 (0.99, 1.73)1.21 (0.6, 1.81), 0.000.58 (0.2, 0.96)0.76 (0.15, 1.37), 0.01
       IBET0.42 (0.03, 0.82)0.27 (−0.35, 0.89), 0.390.53 (0.12, 0.94)0.71 (0.08, 1.33), 0.03
      Patient global assessment of change – left Knee
       WL−0.09 (−0.64, 0.45)−0.38 (−0.95, 0.19)
       PT0.93 (0.56, 1.31)1.03 (0.39, 1.66), 0.000.17 (−0.24, 0.59)0.56 (−0.12, 1.23), 0.11
       IBET0.46 (0.06, 0.86)0.56 (−0.1, 1.21), 0.090.57 (0.13, 1.01)0.95 (0.26, 1.64), 0.01
      Notes: Between-group comparisons refer to changes from baseline for each intervention group relative to the WL group. Results are least-squares means and mean differences (and corresponding 95% CIs) from separate general linear mixed effects models, as described in the Methods.
      WOMAC = Western Ontario and McMaster Universities Osteoarthritis Index; PASE = Physical Activity Scale for the Elderly.
      Indicates number included in the statistical model for that outcome.
      ∗∗ A square root transformation was applied due to superior diagnostics in statistical models.
      Figure thumbnail gr2
      Fig. 2Estimated mean WOMAC Total Scores and 95% CIs by Group and Time Point.
      Table IIISMDs and 95% CIs for IBET and PT vs WL control group
      Outcome4 Months12 Months
      WOMAC total
       PT−0.26 (−0.53, 0.00)−0.12 (−0.40, 0.16)
       IBET−0.20 (−0.48, 0.07)−0.19 (−0.46, 0.09)
      WOMAC function
       PT−0.27 (−0.52, −0.01)−0.19 (−0.45, 0.08)
       IBET−0.15 (−0.43, 0.13)−0.19 (−0.45, 0.08)
      WOMAC pain
       PT−0.14 (−0.39, 0.11)−0.02 (−0.31, 0.27)
       IBET−0.28 (−0.41, −0.15)−0.15 (−0.39, 0.09)

       Non-inferiority hypotheses

      Compared to PT, IBET effects were within the pre-specified non-inferiority limit of five points on the WOMAC total score at both 4-months (estimate = 0.67, 95% CI = −2.23, 3.56; P = 0.65) and 12-months (estimate = −1.04, 95% CI = −5.26, 2.08; P = 0.39), Fig. 3.
      Figure thumbnail gr3
      Fig. 3Comparison of change in WOMAC total scores between IBET and PT group (non-inferiority hypotheses): Results of intent-to-treat analyses.

       Per-protocol analyses

      Per-protocol analyses yielded similar results (Appendix Table III): The greatest difference was between PT and WL at 4 months (−3.65, 95% CI = −7.34, −0.03, P = 0.05). Differences between IBET and PT were within the pre-specified non-inferiority limit at both time points.

       Secondary outcomes

       WOMAC subscales

      In ITT analyses, changes in WOMAC pain and function did not differ significantly between either intervention group and WL at 4 or 12 months (Table II). There were also no statistically significant differences between PT and IBET (Appendix Table IV). Results were similar in multiple imputation and per-protocol analyses for both WOMAC subscales (Appendix Tables II and III). Table III shows SMDs for both interventions compared to WL.

       PASE and weekly minutes of exercise

      At 4 months there were no significant differences in PASE subscale scores across groups (Table II). At 12 months the PT group had significantly greater improvement in PASE Leisure subscale score compared to WL (P = 0.02). There were no notable differences in multiple imputation or per-protocol analyses of PASE scores (Appendix Tables II and III). There were no significant differences in weekly minutes of strengthening or aerobic exercise across groups at either time point (Table II). The PT group reported greater weekly minutes of stretching than WL at both 4 and 12 months, and the IBET reported greater minutes than WL at 12 months. Results were similar in multiple imputation analyses. In per protocol analyses, the PT group reported greater minutes of strengthening at 4 months and aerobic activity at 12 months compared to WL; the IBET group reported greater minutes of aerobic activity at 4 and 12 months.

       Functional tests

      For both unilateral stand time and the 30 s chair stand test, there were minimal within-group changes over time and no between-group differences when using ITT (Table II), multiple imputation or per-protocol analyses (Appendix Tables II and III). For the 2-minute march test, the largest difference was between the PT and WL groups at 4 months (ITT estimate = 7.75, 95% CI = 0.43, 15.07, P = 0.04), favoring the PT group; using multiple imputation, this difference was 8.97 (95% CI = 1.68, 16.26, P = 0.02). There were no statistically significant differences in the TUG test for the PT or IBET groups compared to the WL group (Appendix Table II).

       Global assessment of knee symptom change

      In ITT analyses for the right knee, the PT group reported greater improvement than WL at 4 and 12 months, and the IBET group reported greater improvement than WL at 12 months (Table II). At 4 months, IBET reported less improvement than the PT group. For the left knee, the PT group reported more improvement than WL at 4 months, and the IBET group reported more improvement than WL at 12 months. Results were similar for multiple imputed and per protocol analyses.

      Discussion

      In this study there were no statistically significant nor clinically meaningful differences in most study outcomes, including total WOMAC score, between intervention groups and the WL group. IBET was non-inferior to PT at both 4 and 12 months for the primary outcome.
      Given prior studies on the effectiveness of exercise and PT care for knee OA
      • Fransen M.
      • McConnell S.
      • Harmer A.R.
      • Van der Esch M.
      • Simic M.
      • Bennell K.L.
      Exercise for osteoarthritis of the knee: a Cochrane systematic review.
      • Wang S.Y.
      • Olson-Kellogg B.
      • Shamliyan T.A.
      • Choi J.Y.
      • Ramakrishnan R.
      • Kane R.L.
      Physical therapy interventions for knee pain secondary to osteoarthritis: a systematic review.
      • Bartholdy C.
      • Juhl C.
      • Christensen R.
      • Lund H.
      • Zhang W.
      • Henriksen M.
      The role of muscle strengthening in exercise therapy for knee osteoarthritis: a systematic review and meta-regression analysis of randomized trials.
      , it is unclear why the PT intervention was not superior to WL for most outcomes. It is challenging to compare effects across PT-based interventions due to heterogeneity in dose (e.g., number and duration of sessions), type and duration. However, a meta-analysis of PT-related interventions for knee OA found that with respect to pain, SMDs were −0.21 (−0.35, −0.08) and −0.69 (−1.24, −0.14) for programs focusing on aerobic and strengthening exercise, respectively
      • Wang S.Y.
      • Olson-Kellogg B.
      • Shamliyan T.A.
      • Choi J.Y.
      • Ramakrishnan R.
      • Kane R.L.
      Physical therapy interventions for knee pain secondary to osteoarthritis: a systematic review.
      . The SMD for pain immediately following our PT intervention was smaller than these (−0.14) and declined at 12 months. The meta-analysis found that with respect to disability/function, SMDs were −0.21 (−0.37, −0.04) and −0.16 (−0.48, −0.16), for programs focusing on aerobic and strengthening exercise, respectively. The SMD for function immediately after our PT intervention was somewhat larger than these −(0.27), but declined to 0.19 at 12 months. Therefore, our PT intervention was comparable to pooled estimates of prior PT-related studies regarding function but less effective with respect to pain; overall these effect sizes were small. We aimed for the PT intervention to mirror standard practice, but effects may have been more robust with a greater exercise dose. Recent meta-analyses of OA studies indicate that exercise-based interventions adhering to American College of Sports Medicine dose recommendations resulted in larger improvements
      • Bartholdy C.
      • Juhl C.
      • Christensen R.
      • Lund H.
      • Zhang W.
      • Henriksen M.
      The role of muscle strengthening in exercise therapy for knee osteoarthritis: a systematic review and meta-regression analysis of randomized trials.
      • Moseng T.
      • Dagfinrud H.
      • Smedslund G.
      • Osteras N.
      The importance of dose in land-based supervised exercise for people with hip osteoarthritis. A systematic review and meta-analysis.
      . Additional work is needed to develop strategies for standardizing and implementing these recommendations within the structure and limited number of visits typically allowed for routine PT care for knee OA.
      Although IBET was non-inferior to PT for most outcomes, these results should be interpreted in light of the small, non-significant effects of the PT intervention. Effect sizes for the IBET intervention were also small. There has been little research on internet-based exercise programs for knee OA, but in two prior studies, effects were somewhat greater than in our study
      • Brooks M.A.
      • Beaulieu J.E.
      • Severson H.H.
      • Wille C.M.
      • Cooper D.
      • Gau J.M.
      • et al.
      Web-based therapeutic exercise resource center as a treatment for knee osteoarthritis: a prospective cohort pilot study.
      • Bossen D.
      • Veenhof C.
      • Van Beek K.E.
      • Spreeuwenberg P.M.
      • Dekker J.
      • De Bakker D.H.
      Effectiveness of a web-based physical activity intervention in patients with knee and/or hip osteoarthritis: randomized controlled trial.
      . Both of those prior studies recruited participants via self-referral or opt-in after clinician referral, which may have resulted in more highly motivated samples with greater “readiness to change” compared to our participants, who recruited participants proactively by the study team
      • Williams Q.I.
      • Gunn A.H.
      • Beaulieu J.E.
      • Benas B.C.
      • Buley B.
      • Callahan L.F.
      • et al.
      Physical therapy vs internet-based exercise training (PATH-IN) for patients with knee osteoarthritis: study protocol of a randomized controlled trial.
      . Engagement with the IBET program was relatively low, highlighting the need for strategies to facilitate use of these types of programs and identify patients who may be most likely to benefit.
      There are several limitations to our study. First, we did not confirm OA diagnosis with standardized de novo radiographs or independent physician assessments. However, all participants had either a prior radiographic or physician diagnosis of OA (in the medical record or self-reported), so it is very unlikely there were participants without either radiographic or symptomatic OA. Second, self-reported physical activity is often over-reported. However, it is unlikely that this differed among study groups. Third, we did not assess adherence to home exercise. Fourth, because this was a pragmatic study, physical therapists were permitted to vary the intervention in terms of specific exercises assigned and intensity, based on participant needs; this approach has advantages regarding the study of real-world PT practice but presents challenges in evaluating effects of a specific exercise dose. Fifth, this study was conducted in one geographic region and only included participants with regular internet access, which may limit generalizability of findings. Sixth, this sample was relatively well educated, and results may not generalize to patient groups with lower education levels.
      In conclusion, in this pragmatic study neither the PT nor IBET intervention resulted in statistically significant or clinically relevant improvement in most outcomes, compared to a WL control group. Effects of both interventions may have been robust if the dose had been greater
      • Bartholdy C.
      • Juhl C.
      • Christensen R.
      • Lund H.
      • Zhang W.
      • Henriksen M.
      The role of muscle strengthening in exercise therapy for knee osteoarthritis: a systematic review and meta-regression analysis of randomized trials.
      • Moseng T.
      • Dagfinrud H.
      • Smedslund G.
      • Osteras N.
      The importance of dose in land-based supervised exercise for people with hip osteoarthritis. A systematic review and meta-analysis.
      . In agreement with a recent systematic review
      • Brasure M.
      • Shamliyan T.A.
      • Olson-Kellog B.
      • Butler M.E.
      • Kane R.L.
      Physical Therapy for Knee Pain Secondary to Osteoarthritis: Future Research Needs.
      , results of this study suggest additional research is needed to develop strategies for maximizing the effectiveness of PT interventions, including understanding which PT treatments work best for which patients and optimizing intervention dose in the context of real-world clinical settings.

      Competing interests

      Visual Health Information, Inc (VHI) owns the website used in the current manuscript. Heiderscheit and Seversen have received consulting fees from VHI. A patent related to the website described in this manuscript is currently under review.

      Role of the funding

      This study was funded through a Patient-Centered Outcomes Research Institute Award ( CER-1306-02043 ). The statements, opinions presented in this manuscript are solely the responsibility of the authors and do not necessarily represent the views of the Patient-Centered Outcomes Research Institute, its Board of Governors or Methodology Committee. KDA, LA, LFC, YMG, APG, and TAS receive support from National Institute of Arthritis and Musculoskeletal and Skin Diseases Multidisciplinary Clinical Research Center P60 AR062760 . KDA receives support from the Center for Health Services Research in Primary Care, Durham VA Health Care System ( CIN 13-410 ).

      Authors contributions

      KDA, LA, LFC, YMG, APG, BCH, KMH, HS and TAS contributed to the study design and protocol and helped draft the manuscript. HS and BCH contributed to the original design and evaluation of the exercise website. TS and LA contributed to plans for and conduct of statistical analyses. All authors read and approved the final manuscript. YMG and APG oversaw design and fidelity checks for the PT intervention.

      Acknowledgements

      The study team thanks all of the study participants, without whom this work would not be possible. We also thank the following team members for their contributions to the research: Caroline Nagle, Kimberlea Grimm, Ashley Gwyn, Bernadette Benas, Alex Gunn, Leah Schrubbe, and Quinn Williams. The study team also expresses gratitude to the Stakeholder Panel for this project: Ms Sandy Walker LPN (Chapel Hill Children's Clinic), Ms Susan Pedersen RN BSN, Ms Sally Langdon Thomas, Mr Ralph B. Brown, Ms Frances Talton CDA RHS Retired, Dr Katrina Donahue, MD, MPH (Department of Family Medicine at the University of North Carolina at Chapel Hill), Dr Alison Brooks, MD, MPH (Department of Orthopedics & Rehabilitation at the University of Wisconsin-Madison), Dr Anita Bemis-Dougherty, PT, DPT, MAS (American Physical Therapy Association), Dr Teresa J. Brady, PhD (Centers for Disease Control and Prevention), Ms Laura Marrow (Arthritis Foundation National Office), Ms Megan Simmons Skidmore (American Institute of Healthcare and Fitness), and Dr Maura Daly Iversen, PT, DPT, SD, MPH, FNAP, FAPTA (Department of Physical Therapy, Movement and Rehabilitation Sciences Northeastern University). The study team thanks study physical therapists and physical therapy assistants: Jennifer Cooke, PT, DPT, Jyotsna Gupta, PT, PhD and Carla Hill, PT, DPT, OCS, Cert MDT (Division of Physical Therapy, University of North Carolina at Chapel Hill), Bruce Buley, Andrew Genova, and Ami Pathak (Comprehensive Physical Therapy, Chapel Hill, NC), Chris Gridley and Aaron Kline (Pivot Physical Therapy, Smithfield, NC).

      Appendix A. Supplementary data

      The following are the supplementary data related to this article:

      Appendix. Guidance for structure and content of PT visits

      Appendix Table 1Proportions of participants reporting use of specific OA treatments by group and follow-up time point
      % of Participants self-reporting use of treatmentIBETPTWL
      Pain medications for osteoarthritis
       4-Month follow-up65%55%60%
       12-Month follow-up59%57%52%
      Knee injection since baseline
       4-Month follow-up7%7%12%
       12-Month Follow-Up14%12%21%
      PT (non-study) for knee OA since last study visit
       4-Month follow-up5%1%7%
       12-Month follow-up12%7%11%
      Knee brace (current)
       4-Month follow-up19%19%20%
       12-Month follow-up19%20%17%
      Topical creams (current)
       4-Month follow-up24%23%22%
       12-Month follow-up29%28%24%
      Appendix Table 2Within- and between-group mean changes in outcomes and 95% CIs: Results of ITT analyses with multiple imputation
      OutcomeBaseline to 4-month difference (95% CI)Difference in baseline to 4-month vs WL (95% CI), P-valueBaseline to 12-month difference (95% CI)Difference in baseline to 12-month vs WL (95% CI), P-value
      WOMAC total
       WL−3.29 (−6.29, −0.29)−2.95 (−6.04, 0.15)
       PT−6.85 (−9.01, −4.69)−3.56 (−7.16, 0.04), 0.05−4.72 (−6.96, −2.48)−1.77 (−5.38, 1.84), 0.34
       IBET−6.00 (−8.19, −3.82)−2.71 (−6.28, 0.86), 0.14−5.68 (−8.03, −3.32)−2.73 (−6.50, 1.04), 0.16
      WOMAC function
       WL−2.31 (−4.45, −0.17)−1.63 (−3.95, 0.68)
       PT−4.97 (−6.50, −3.43)−2.66 (−5.20, −0.11), 0.04−3.39 (−5.04, −1.75)−1.76 (−4.48, 0.96), 0.21
       IBET−3.97 (−5.57, −2.37)−1.66 (−4.29, 0.97), 0.22−3.75 (−5.41, −2.10)−2.12 (−4.85, 0.62), 0.13
      WOMAC pain
       WL−0.65 (−1.4, 0.10)−0.65 (−1.39, 0.09)
       PT−1.12 (−1.66, −0.58)−0.47 (−1.36, 0.41), 0.29−0.71 (−1.26, −0.17)−0.06 (−0.94, 0.81), 0.89
       IBET−1.53 (−2.12, −0.95)−0.89 (−1.8, 0.03), 0.06−1.12 (−1.65, −0.58)−0.47 (−1.33, 0.40), 0.29
      PASE total
       WL−2.72 (−19.05, 13.61)1.96 (−12.93, 16.85)
       PT2.49 (−9.08, 14.07)5.21 (−13.91, 24.33), 0.597.91 (−2.86, 18.69)5.95 (−11.31, 23.22), 0.50
       IBET−11.25 (−24.37, 1.88)8.53 (−27.33, 10.26), 0.379.43 (−2.12, 20.99)7.47 (−10.23, 25.18), 0.41
      PASE leisure
       WL−2.73 (−8.15, 2.69)−0.23 (−6.49, 6.03)
       PT4.01 (0.44, 7.58)6.74 (0.56, 12.92), 0.038.81 (4.45, 13.16)9.04 (1.67, 16.40), 0.02
       IBET−1.00 (−4.91, 2.91)1.74 (−4.59, 8.06), 0.597.69 (3.09, 12.28)7.92 (0.55, 15.29), 0.04
      PASE household
       WL−5.65 (−14.68, 3.37)−4.05 (−12.07, 3.97)
       PT−2.05 (−8.34, 4.24)3.60 (−7.05, 14.25), 0.512.02 (−3.85, 7.88)6.07 (−3.55, 15.68), 0.22
       IBET−8.83 (−15.61, −2.05)−3.18 (−14.04, 7.69), 0.57−4.12 (−10.69, 2.44)−0.07 (−10.06, 9.92), 0.99
      PASE work
       WL4.38 (−6.81, 15.56)5.67 (−4.10, 15.44)
       PT1.45 (−6.01, 8.91)−2.93 (−15.79, 9.93), 0.66−2.76 (−9.64, 4.11)−8.43 (−19.85, 2.98), 0.15
       IBET−1.32 (−9.51, 6.87)−5.70 (−19.48, 8.08), 0.425.87 (−1.30, 13.04)0.20 (−11.46, 11.86), 0.97
      Unilateral stand time
       WL−0.12 (−0.90, 0.66)−0.14 (−0.94, 0.65)
       PT−0.53 (−1.08, 0.02)−0.41 (−1.32, 0.50), 0.38−0.02 (−0.55, 0.52)0.13 (−0.81, 1.06), 0.79
       IBET0.08 (−0.53, 0.70)0.20 (−0.77, 1.18), 0.680.02 (−0.54, 0.58)0.16 (−0.78, 1.11), 0.73
      30 s chair stand
       WL0.10 (−0.95, 1.16)0.55 (−0.38, 1.49)
       PT−0.06 (−0.80, 0.68)−0.16 (−1.41, 1.09), 0.800.13 (−0.54, 0.80)−0.43 (−1.54, 0.69), 0.45
       IBET0.67 (−0.10, 1.43)0.56 (−0.72, 1.85), 0.390.86 (0.18, 1.55)0.31 (−0.84, 1.45), 0.60
      2 m march test
       WL−8.83 (−14.99, −2.67)−0.09 (−6.67, 6.50)
       PT0.14 (−4.20, 4.48)8.97 (1.68, 16.26), 0.021.06 (−3.68, 5.79)1.14 (−6.69, 8.98), 0.77
       IBET−2.38 (−6.88, 2.11)6.45 (−0.99, 13.88), 0.091.35 (−3.51, 6.20)1.43 (−6.80, 9.67), 0.73
      Timed Up and Go
       WL−0.11 (−1.14, 0.91)−0.31 (−1.43, 0.80)
       PT−0.56 (−1.30, 0.17)−0.45 (−1.63, 0.73), 0.45−0.94 (−1.75, −0.13)−0.62 (−1.98, 0.73), 0.37
       IBET−0.90 (−1.74, −0.06)−0.79 (−2.08, 0.50), 0.23−1.47 (−2.36, −0.58)−1.16 (−2.58, 0.27), 0.11
      Weekly minutes of aerobic activity
      A square root transformation was applied due to superior diagnostics in statistical models.
       WL−0.05 (−1.48, 1.38)−1.68 (−3.32, −0.05)
       PT0.98 (−0.03, 1.99)1.03 (−0.67, 2.73), 0.230.51 (−0.65, 1.66)2.19 (0.24, 4.13), 0.03
       IBET1.88 (0.76, 3)1.93 (0.25, 3.62), 0.020.49 (−0.77, 1.74)2.17 (0.15, 4.18), 0.03
      Weekly minutes of stretching
      A square root transformation was applied due to superior diagnostics in statistical models.
       WL−0.36 (−1.38, 0.65)−1.29 (−2.19, −0.38)
       PT1.45 (0.76, 2.15)1.81 (0.62, 3), 0.000.36 (−0.28, 1)1.65 (0.58, 2.72), 0.00
       IBET1.08 (0.31, 1.85)1.44 (0.22, 2.67), 0.020.8 (0.09, 1.51)2.09 (1, 3.17), 0.00
      Weekly minutes of strengthening
      A square root transformation was applied due to superior diagnostics in statistical models.
       WL0.43 (−0.69, 1.55)−0.1 (−1.28, 1.08)
       PT1.85 (1.04, 2.65)1.42 (0.11, 2.72), 0.031.17 (0.33, 2.02)1.27 (−0.12, 2.67), 0.07
       IBET1.47 (0.63, 2.32)1.04 (−0.31, 2.39), 0.131.32 (0.38, 2.26)1.42 (−0.03, 2.87), 0.05
      Patient global assessment of change – right Knee
       WL0.14 (−0.39, 0.67)−0.17 (−0.69, 0.36)
       PT1.36 (0.97, 1.74)1.22 (0.58, 1.86), 0.000.60 (0.2, 1.01)0.77 (0.14, 1.4), 0.02
       IBET0.43 (0.04, 0.83)0.30 (−0.33, 0.92), 0.350.53 (0.13, 0.94)0.70 (0.04, 1.36), 0.04
      Patient global assessment of change – left Knee
       WL−0.1 (−0.66, 0.45)−0.39 (−0.96, 0.17)
       PT0.94 (0.56, 1.33)1.05 (0.4, 1.7), 0.000.16 (−0.27, 0.59)0.55 (−0.14, 1.24), 0.11
       IBET0.50 (0.1, 0.9)0.60 (−0.07, 1.27), 0.080.58 (0.15, 1.02)0.98 (0.28, 1.68), 0.00
      A square root transformation was applied due to superior diagnostics in statistical models.
      Appendix Table 3Within- and between-group mean changes in outcomes and 95% CIs: Results of per protocol analyses
      OutcomeBaseline to 4-month difference (95% CI)Difference in baseline to 4-month vs WL (95% CI), P-valueBaseline to 12-month difference (95% CI)Difference in baseline to 12-month vs WL (95% CI), P-value
      WOMAC total
       WL−3.64 (−6.8, −0.48)−2.74 (−6, 0.53)
       PT−7.29 (−9.56, −5.03)−3.65 (−7.34, 0.03), 0.05−4.71 (−7.07, −2.35)−1.97 (−5.81, 1.86), 0.31
       IBET−6 (−8.53, −3.46)−2.36 (−6.23, 1.51), 0.23−5.84 (−8.48, −3.19)−3.1 (−7.13, 0.93), 0.13
      WOMAC function
       WL−2.48 (−4.79, −0.18)−1.37 (−3.72, 0.97)
       PT−5.2 (−6.83, −3.56)−2.71 (−5.4, −0.02), 0.05−3.58 (−5.26, −1.9)−2.21 (−4.96, 0.55), 0.11
       IBET−3.79 (−5.63, −1.95)−1.31 (−4.13, 1.52), 0.36−3.75 (−5.65, −1.86)−2.38 (−5.27, 0.51), 0.11
      WOMAC pain
       WL−0.7 (−1.49, 0.09)−0.68 (−1.46, 0.1)
       PT−1.19 (−1.75, −0.62)−0.49 (−1.41, 0.43), 0.29−0.69 (−1.25, −0.13)−0.01 (−0.92, 0.9), 0.98
       IBET−1.59 (−2.22, −0.95)−0.89 (−1.86, 0.08), 0.07−1.16 (−1.79, −0.53)−0.49 (−1.44, 0.47), 0.32
      PASE total
       WL−2.95 (−20.41, 14.51)2.36 (−12.43, 17.15)
       PT2.39 (−9.73, 14.51)5.34 (−15.13, 25.81), 0.618.33 (−2.3, 18.96)5.97 (−11.29, 23.23), 0.50
       IBET−8.85 (−22.92, 5.22)−5.9 (−27.55, 15.75), 0.597.94 (−4.41, 20.29)5.58 (−12.77, 23.93), 0.55
      PASE leisure
       WL−2.7 (−8.17, 2.77)−0.79 (−7.26, 5.68)
       PT3.39 (−0.42, 7.19)6.08 (−0.26, 12.42), 0.067.93 (3.34, 12.52)8.72 (1.07, 16.37), 0.02
       IBET−1.19 (−5.56, 3.18)1.51 (−5.18, 8.2), 0.667.54 (2.24, 12.84)8.33 (0.24, 16.42), 0.04
      PASE household
       WL−3.48 (−13.05, 6.1)−1.79 (−10.37, 6.78)
       PT−0.57 (−7.24, 6.1)2.9 (−8.35, 14.16), 0.613.4 (−2.78, 9.58)5.19 (−4.9, 15.28), 0.31
       IBET−8.55 (−16.33, −0.76)−5.07 (−17.02, 6.88), 0.40−3.1 (−10.25, 4.05)−1.3 (−12.02, 9.41), 0.81
      PASE work
       WL2.27 (−9.4, 13.93)5.25 (−4.77, 15.26)
       PT0.42 (−7.61, 8.46)−1.84 (−15.53, 11.84), 0.79−2.93 (−10.07, 4.21)−8.18 (−19.91, 3.55), 0.17
       IBET−0.95 (−10.18, 8.28)−3.22 (−17.64, 11.2), 0.664.18 (−3.94, 12.3)−1.07 (−13.41, 11.28), 0.86
      Unilateral stand time
       WL−0.03 (−0.81, 0.76)−0.21 (−1.04, 0.62)
       PT−0.6 (−1.15, −0.04)−0.57 (−1.49, 0.35), 0.22−0.09 (−0.66, 0.48)0.12 (−0.86, 1.1), 0.80
       IBET0.19 (−0.43, 0.81)0.22 (−0.75, 1.18), 0.660.01 (−0.65, 0.66)0.21 (−0.81, 1.24), 0.68
      30 s chair stand
       WL0.06 (−1.05, 1.17)0.56 (−0.43, 1.55)
       PT−0.2 (−0.98, 0.58)−0.26 (−1.57, 1.05), 0.700.12 (−0.57, 0.81)−0.44 (−1.6, 0.71), 0.45
       IBET0.62 (−0.26, 1.5)0.57 (−0.81, 1.94), 0.420.95 (0.16, 1.73)0.38 (−0.83, 1.6), 0.53
      2 m march test
       WL−8.51 (−14.94, −2.08)−0.5 (−7.28, 6.28)
       PT−0.33 (−4.88, 4.22)8.18 (0.62, 15.74), 0.031.73 (−2.94, 6.4)2.23 (−5.72, 10.18), 0.58
       IBET−2.32 (−7.42, 2.78)6.19 (−1.73, 14.12), 0.123.01 (−2.33, 8.35)3.51 (−4.86, 11.89), 0.41
      Timed Up and Go
       WL−0.15 (−1.23, 0.93)0.04 (−1.19, 1.27)
       PT−0.56 (−1.32, 0.21)−0.41 (−1.68, 0.86), 0.53−0.68 (−1.53, 0.18)−0.71 (−2.16, 0.73), 0.33
       IBET−0.82 (−1.68, 0.03)−0.68 (−2, 0.65), 0.32−1.56 (−2.53, −0.59)−1.6 (−3.12, −0.08), 0.04
      Weekly minutes of aerobic activity
       WL−0.22 (−1.73, 1.29)−1.68 (−3.37, 0)
       PT1.11 (0.04, 2.17)1.33 (−0.45, 3.1), 0.140.76 (−0.43, 1.9)2.45 (0.45, 4.44), 0.02
       IBET2.25 (1.04, 3.46)2.47 (0.61, 4.33), 0.010.83 (−0.52, 2.18)2.52 (0.42, 4.61), 0.02
      Weekly minutes of stretching
       WL−0.47 (−1.5, 0.57)−1.37 (−2.32, −0.43)
       PT1.61 (0.89, 2.33)2.08 (0.87, 3.3), 0.000.29 (−0.38, 0.97)1.67 (0.55, 2.78), 0.00
       IBET0.84 (0.02, 1.67)1.31 (0.03, 2.59), 0.040.59 (−0.18, 1.36)1.96 (0.79, 3.13), 0.00
      Weekly minutes of strengthening
       WL0.47 (−0.67, 1.61)−0.17 (−1.4, 1.05)
       PT2.02 (1.21, 2.82)1.55 (0.22, 2.88), 0.021.12 (0.25, 2)1.3 (−0.15, 2.74), 0.08
       IBET1.08 (0.17, 1.99)0.61 (−0.79, 2.01), 0.391.19 (0.2, 2.17)1.36 (−0.16, 2.88), 0.08
      Patient global assessment of change – right knee
       WL0.15 (−0.38, 0.69)−0.2 (−0.74, 0.33)
       PT1.43 (1.04, 1.81)1.27 (0.64, 1.91), 0.000.63 (0.24, 1.03)0.83 (0.2, 1.47), 0.01
       IBET0.60 (0.17, 1.03)0.45 (−0.21, 1.11), 0.180.75 (0.29, 1.20)0.95 (0.28, 1.62), 0.00
      Patient global assessment of change – left knee
       WL0.07 (−0.52, 0.65)−0.33 (−0.92, 0.26)
       PT1.03 (0.63, 1.42)0.96 (0.28, 1.64), 0.000.34 (−0.08, 0.77)0.67 (−0.03, 1.37), 0.06
       IBET0.56 (0.11, 1.01)0.49 (−0.22, 1.20), 0.170.82 (0.35, 1.29)1.15 (0.43, 1.88), 0.00
      Appendix Table 4Differences in mean changes between IBET and PT and 95% CIs
      OutcomeDifference in baseline to 4-month vs PT (95% CI), P-valueDifference in baseline to 12-month vs PT (95% CI), P-value
      WOMAC total
       ITT0.67 (−2.23, 3.56), 0.65−1.04 (−4.13, 2.05), 0.51
       Multiple Imputation
      Multiple Imputation was performed on missing values under the ITT paradigm.
      0.85 (−2.06, 3.75), 0.57−0.96 (−4.06, 2.14), 0.54
       Per Protocol1.3 (−1.9, 4.5), 0.43−1.13 (−4.49, 2.23), 0.51
      WOMAC function
       ITT1.04 (−1.07, 3.15), 0.33−0.11 (−2.34, 2.13), 0.93
       Multiple Imputation1.00 (−1.09, 3.08), 0.35−0.36 (−2.55, 1.84), 0.75
       Per Protocol1.41 (−0.92, 3.73), 0.23−0.17 (−2.58, 2.23), 0.89
      WOMAC pain
       ITT−0.47 (−1.20, 0.26), 0.20−0.45 (−1.18, 0.27), 0.22
       Multiple Imputation−0.41 (−1.16, 0.33), 0.28−0.40 (−1.13, 0.32), 0.27
       Per Protocol−0.4 (−1.2, 0.4), 0.33−0.47 (−1.27, 0.32), 0.24
      PASE total
       ITT−13.77 (−29.73, 2.19), 0.09−0.09 (−14.41, 14.23), 0.99
       Multiple Imputation−13.74 (−29.76, 2.27), 0.091.52 (−13.12, 16.16), 0.84
       Per Protocol−11.24 (−28.97, 6.49), 0.21−0.39 (−15.74, 14.96), 0.96
      PASE leisure
       ITT−4.61 (−9.49, 0.28), 0.06−1.01 (−7.20, 5.18), 0.75
       Multiple Imputation−5.00 (−9.98, −0.03), 0.05−1.12 (−7.06, 4.82), 0.71
       Per Protocol−4.57 (−10.02, 0.87), 0.09−0.39 (−7.12, 6.34), 0.91
      PASE household
       ITT−8.09 (−17.13, 0.94), 0.08−6.02 (−14.27, 2.24), 0.15
       Multiple Imputation−6.77 (−15.76, 2.21), 0.14−6.14 (−14.69, 2.42), 0.16
       Per Protocol−7.97 (−17.77, 1.82), 0.11−6.5 (−15.46, 2.46), 0.15
      PASE work
       ITT−2.98 (−13.45, 7.49), 0.587.87 (−1.79, 17.53), 0.11
       Multiple Imputation−2.77 (−13.58, 8.04), 0.628.63 (−0.73, 17.99), 0.07
       Per Protocol−1.38 (−13.1, 10.34), 0.827.11 (−3.12, 17.34), 0.17
      Unilateral stand time
       ITT0.61 (−0.16, 1.38), 0.120.00 (−0.78, 0.77), 1.00
       Multiple Imputation0.61 (−0.20, 1.43), 0.140.04 (−0.72, 0.80), 0.92
       Per Protocol0.79 (−0.01, 1.58), 0.050.09 (−0.74, 0.92), 0.83
      30 s chair stand
       ITT0.63 (−0.40, 1.66), 0.230.74 (−0.17, 1.64), 0.11
       Multiple Imputation0.73 (−0.30, 1.75), 0.160.73 (−0.17, 1.64), 0.11
       Per Protocol0.82 (−0.3, 1.95), 0.150.83 (−0.16, 1.82), 0.10
      2 m march test
       ITT−2.86 (−8.94, 3.21), 0.350.01 (−6.40, 6.42), 1.00
       Multiple Imputation−2.52 (−8.47, 3.43), 0.410.29 (−6.34, 6.92), 0.93
       Per Protocol−1.99 (−8.5, 4.52), 0.551.28 (−5.53, 8.1), 0.71
      Timed Up and Go
       ITT−0.24 (−1.23, 0.74), 0.63−0.72 (−1.85, 0.41), 0.21
       Multiple Imputation−0.34 (−1.40, 0.73), 0.53−0.53 (−1.69, 0.62), 0.36
       Per Protocol−0.27 (−1.35, 0.82), 0.63−0.89 (−2.13, 0.36), 0.16
      Weekly minutes of aerobic activity
      A square root transformation was applied due to superior diagnostics in statistical models.
       ITT0.79 (−0.62, 2.2), 0.27−0.07 (−1.69, 1.54), 0.93
       Multiple Imputation0.9 (−0.57, 2.37), 0.23−0.02 (−1.67, 1.63), 0.98
       Per Protocol1.15 (−0.38, 2.67), 0.140.07 (−1.66, 1.8), 0.93
      Weekly minutes of stretching
      A square root transformation was applied due to superior diagnostics in statistical models.
       ITT−0.48 (−1.46, 0.5), 0.330.45 (−0.44, 1.34), 0.32
       Multiple Imputation−0.37 (−1.37, 0.63), 0.460.44 (−0.49, 1.37), 0.35
       Per Protocol−0.77 (−1.82, 0.28), 0.150.3 (−0.67, 1.27), 0.55
      Weekly minutes of strengthening
      A square root transformation was applied due to superior diagnostics in statistical models.
       ITT−0.51 (−1.6, 0.58), 0.360.14 (−1.03, 1.31), 0.81
       Multiple Imputation−0.38 (−1.47, 0.72), 0.500.15 (−1.05, 1.34), 0.81
       Per Protocol−0.94 (−2.09, 0.21), 0.110.07 (−1.19, 1.32), 0.92
      Patient global assessment of change – right knee
       ITT−0.93 (−1.44, −0.42), 0.00−0.05 (−0.58, 0.48), 0.85
       Multiple Imputation−0.92 (−1.45, −0.4), 0.00−0.07 (−0.6, 0.46), 0.80
       Per Protocol−0.83 (−1.38, −0.28), 0.000.12 (−0.46, 0.69), 0.69
      Patient global assessment of change – left knee
       ITT−0.47 (−0.99, 0.05), 0.070.39 (−0.18, 0.97), 0.17
       Multiple Imputation−0.45 (−0.99, 0.09), 0.100.43 (−0.13, 0.98), 0.13
       Per Protocol−0.47 (−1.04, 0.1), 0.100.48 (−0.13, 1.08), 0.12
      Multiple Imputation was performed on missing values under the ITT paradigm.
      ∗∗ A square root transformation was applied due to superior diagnostics in statistical models.

      References

        • Meneses S.R.
        • Goode A.P.
        • Nelson A.E.
        • Lin J.
        • Jordan J.M.
        • Allen K.D.
        • et al.
        Clinical algorithms to aid osteoarthritis guideline dissemination.
        Osteoarthr Cartil. 2016 Sep; 24: 1487-1499
        • Nelson A.E.
        • Allen K.D.
        • Golightly Y.M.
        • Goode A.P.
        • Jordan J.M.
        A systematic review of recommendations and guidelines for the management of osteoarthritis: the Chronic Osteoarthritis Management Initiative of the U.S. Bone and Joint Initiative.
        Semin Arthritis Rheum. 2014; 434: 701-712
        • Fransen M.
        • McConnell S.
        • Harmer A.R.
        • Van der Esch M.
        • Simic M.
        • Bennell K.L.
        Exercise for osteoarthritis of the knee: a Cochrane systematic review.
        Br J Sports Med. 2015; 49: 1554-1557
        • Fontaine K.R.
        • Heo M.
        Changes in the prevalence of U.S. adults with arthritis who meet physical activity recommendations, 2001-2003.
        J Clin Rheumatol. 2005; 11: 13-16
        • Shih V.C.
        • Song J.
        • Chang R.W.
        • Dunlop D.D.
        Racial differences in activities of daily living limitation onset in older adults with arthritis: a national cohort study.
        Arch Phys Med Rehabil. 2005; 86: 1521-1526
        • Wang S.Y.
        • Olson-Kellogg B.
        • Shamliyan T.A.
        • Choi J.Y.
        • Ramakrishnan R.
        • Kane R.L.
        Physical therapy interventions for knee pain secondary to osteoarthritis: a systematic review.
        Ann Intern Med. 2012; 157: 632-644
        • Dhawan A.
        • Mather 3rd, R.C.
        • Karas V.
        • Ellman M.B.
        • Young B.B.
        • Bach Jr., B.R.
        • et al.
        An epidemiologic analysis of clinical practice guidelines for non-arthroplasty treatment of osteoarthritis of the knee.
        Arthroscopy. 2014; 30: 65-71
        • Cisternas M.G.
        • Yelin E.
        • Katz J.N.
        • Solomon D.H.
        • Wright E.A.
        • Losina E.
        Ambulatory visit utilization in a national, population-based sample of adults with osteoarthritis.
        Arthritis Rheum. 2009; 61: 1694-1703
        • Allen K.D.
        • Bosworth H.B.
        • Chatterjee R.
        • Coffman C.J.
        • Corsino L.
        • Jeffreys A.S.
        • et al.
        Clinic variation in recruitment metrics, patient characteristics and treatment use in a randomized clinical trial of osteoarthritis management.
        BMC Musculoskelet Disord. 2014; 15: 413
        • Vavken P.
        • Dorotka R.
        The burden of musculoskeletal disease and its determination by urbanicity, socioeconomic status, age, and gender - results from 14507 subjects.
        Arthritis Care Res. 2011; 63: 1558-1564
        • Ackerman I.N.
        • Busija L.
        Access to self-management education, conservative treatment and surgery for arthritis according to socioeconomic status.
        Best Pract Res Clin Rheumatol. 2012; 26: 561-583
        • Li L.C.
        • Lineker S.
        • Cibere J.
        • Crooks V.A.
        • Jones C.A.
        • Kopec J.A.
        • et al.
        Capitalizing on the teachable moment: osteoarthritis physical activity and exercise net for improving physical activity in early knee osteoarthritis.
        JMIR Res Protoc. 2013; 2: e17
        • Bennell K.L.
        • Nelligan R.
        • Dobson F.
        • Rini C.
        • Keefe F.
        • Kasza J.
        • et al.
        Effectiveness of an internet-delivered exercise and pain-coping skills training intervention for persons with chronic knee pain: a randomized trial.
        Ann Intern Med. 2017; 166: 453-462
        • Brooks M.A.
        • Beaulieu J.E.
        • Severson H.H.
        • Wille C.M.
        • Cooper D.
        • Gau J.M.
        • et al.
        Web-based therapeutic exercise resource center as a treatment for knee osteoarthritis: a prospective cohort pilot study.
        BMC Musculoskelet Disord. 2014; 15: 158
        • Bossen D.
        • Veenhof C.
        • Van Beek K.E.
        • Spreeuwenberg P.M.
        • Dekker J.
        • De Bakker D.H.
        Effectiveness of a web-based physical activity intervention in patients with knee and/or hip osteoarthritis: randomized controlled trial.
        J Med Internet Res. 2013; 15: e257
        • Irvine A.B.
        • Gelatt V.A.
        • Seeley J.R.
        • Macfarlane P.
        • Gau J.M.
        Web-based intervention to promote physical activity by sedentary older adults: randomized controlled trial.
        J Med Internet Res. 2013; 15: e19
        • Carr L.J.
        • Dunsiger S.I.
        • Lewis B.
        • Ciccolo J.T.
        • Hartman S.
        • Bock B.
        • et al.
        Randomized controlled trial testing an internet physical activity intervention for sedentary adults.
        Health Psychol. 2013; 32: 328-336
        • Williams Q.I.
        • Gunn A.H.
        • Beaulieu J.E.
        • Benas B.C.
        • Buley B.
        • Callahan L.F.
        • et al.
        Physical therapy vs internet-based exercise training (PATH-IN) for patients with knee osteoarthritis: study protocol of a randomized controlled trial.
        BMC Musculoskelet Disord. 2015; 16: 264
        • Jordan J.M.
        • Helmick C.G.
        • Renner J.B.
        • Luta G.
        • Dragomir A.D.
        • Woodard J.
        • et al.
        Prevalence of knee symptoms and radiographic and symptomatic knee osteoarthritis in African Americans and Caucasians: the Johnston County Osteoarthritis Project.
        J Rheumatol. 2007; 31: 172-180
        • Altman R.
        • Asch D.
        • Bloch G.
        • Bole D.
        • Borenstein K.
        • Brandt K.
        • et al.
        The American College of Rheumatology criteria for the classification and reporting of osteoarthritis of the knee.
        Arthritis Rheum. 1986; 29: 1039-1049
        • U.S. Department of Health and Human Services
        Physical Activity Guidelines for Americans.
        2008 (Washington, DC2008)
        • Hertling D.
        • Kessler R.M.
        Knee.
        in: Hertling D. Kessler R.M. Management of Common Musculoskeletal Disorders: Physical Therapy Principles and Methods. 4th edn. Lippincott Williams & Wilkins, Philadelphia2006
        • Bellamy N.
        WOMAC: a 20-year experiential review of a patient-centered self-reported health status questionnaire.
        J Rheumatol. 2002; 29: 2473-2476
        • Bellamy N.
        • Buchanan W.W.
        • Goldsmith C.H.
        • Campbell J.
        • Stitt L.W.
        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: 1833-1840
        • Jones C.
        • RIkli R.
        • Beam W.A.
        30-s chair-stand test as a measure of lower body strength in community-residing older adults.
        Res Q Exerc Sport. 1999; 70: 113
        • Wall J.C.
        • Bell C.
        • Campbell S.
        • Davis J.
        The Timed Get-up-and-Go test revisited: measurement of the component tasks.
        J Rehabil Res Dev. 2000; 37: 109-113
        • Podsiadlo D.
        • Richardson S.
        The Timed up & Go: a test of basic functional mobility for frail elderly persons.
        J Am Geriatr Soc. 1991; 39: 142-148
        • RIkli R.
        • Jones J.
        Development and validation of a functional fitness test for community-residing older adults.
        J Aging Phys Act. 1999; 7: 129-161
        • Scott V.
        • Votova K.
        • Scanlan A.
        • Close J.
        Multifactorial and functional mobility assessment tools for fall risk among older adults in community, home-support, long-term and acute care settings.
        Age Ageing. 2007; 36: 130-139
        • Rossiter-Fornoff J.E.
        • Wolf S.L.
        • Wolfson L.I.
        • Buchner D.M.
        A cross-sectional validation study of the FICSIT common data base static balance measures. Frailty and injuries: Cooperative studies of intervention techniques.
        J Gerontol A Biol Sci Med Sci. 1995; 50: M291-M297
        • Washburn R.A.
        • Smith K.W.
        • Jette A.M.
        • Janney C.A.
        The physical activity scale for the elderly (PASE): development and evaluation.
        J Clin Epidemiol. 1993; 46: 153-162
        • Kroenke K.
        • Strine T.W.
        • Spitzer R.L.
        • Williams J.B.
        • Berry J.T.
        • Mokdad A.H.
        The PHQ-8 as a measure of current depression in the general population.
        J Affect Disord. 2009; 114: 163-173
        • Greene C.J.
        • Morland L.A.
        • Durkalski V.L.
        • Frueh B.C.
        Noninferiority and equivalence designs: issues and implications for mental health research.
        J Trauma Stress. 2008; 21: 433-439
        • Blackwelder W.C.
        “Proving the null hypothesis” in clinical trials.
        Control Clin Trials. 1982; 3: 345-353
        • Pocock S.J.
        The pros and cons of noninferiority trials.
        Fundam Clin Pharmacol. 2003; 17: 483-490
        • Piantadosi S.
        Clinical Trials: A Methodological Perspective.
        John Wiley and Sons, New Jersey2005
        • Borm G.F.
        • Fransen J.
        • Lemmens W.
        A simple sample size formula for analysis of covariance in randomized clinical trials.
        J Clin Epidemiol. 2007; 60: 1234-1238
        • Angst F.
        • Aeschlimann A.
        • Michel B.A.
        • Stucki G.
        Minimal clinically important rehabilitation effects in patients with osteoarthritis of the lower extremity.
        J Rheumatol. 2002; 29: 131-138
        • Bellamy N.
        • Hochberg M.
        • Tubach F.
        • Martin-Mola E.
        • Awada H.
        • Bombardier C.
        • et al.
        Development of multinational definitions of minimal clinically important improvement and patient acceptable symptomatic state in osteoarthritis.
        Arthritis Care Res. 2015; 67: 972-980
        • ICH E9 Expert Working Group
        ICH harmonised tripartite guideline - statistical principles for clinical trials.
        Stat Med. 1999; 18: 1905-1942
        • Snapinn S.M.
        Noninferiority trials.
        Curr Control Trials Cardiovasc Med. 2000; 1: 19-21
        • Piaggio G.
        • Elbourne D.R.
        • Altman D.G.
        • Pocock S.J.
        • Evans S.J.
        • Group C.
        Reporting of noninferiority and equivalence randomized trials: an extension of the CONSORT statement.
        JAMA. 2006; 295: 1152-1160
        • Bellamy N.
        WOMAC Osteoarthritis Index User Guide.
        2002 (Version V. Brisbane, Australia)
        • Bartholdy C.
        • Juhl C.
        • Christensen R.
        • Lund H.
        • Zhang W.
        • Henriksen M.
        The role of muscle strengthening in exercise therapy for knee osteoarthritis: a systematic review and meta-regression analysis of randomized trials.
        Semin Arthritis Rheum. 2017; 47: 9-21
        • Moseng T.
        • Dagfinrud H.
        • Smedslund G.
        • Osteras N.
        The importance of dose in land-based supervised exercise for people with hip osteoarthritis. A systematic review and meta-analysis.
        Osteoarthr Cartil. 2017; 25: 1563-1576
        • Brasure M.
        • Shamliyan T.A.
        • Olson-Kellog B.
        • Butler M.E.
        • Kane R.L.
        Physical Therapy for Knee Pain Secondary to Osteoarthritis: Future Research Needs.
        Agency for Healthcare Research and Quality, Rockville, MD2013 (AHRQ Publication No. 13-EHC048-EF)