JMIR Publications

ABSTRACT

Background:

Only about one in two patients with knee osteoarthritis (OA) receives a referral or recommendation for exercise. Digital health technologies could counteract this undersupply.

Objective:

We aimed to investigate the effectiveness of a self-directed m-Health exercise intervention (re.flex) compared to usual care in patients with knee OA.

Methods:

This monocentric, two-arm, randomized controlled parallel-group trial included patients with moderate to severe knee OA. Participants were mainly recruited via newspapers. Randomization was 1:1 into an intervention group (re.flex) and a control group (usual care) using computer-generated blocks (stratified by etiology, medication, laterality). Participants were unmasked to group assignment. The re.flex group conducted a 12-week self-directed app-based and sensor-assisted exercise program with three sessions per week in addition to usual care. Primary outcomes were OA-specific knee pain and physical function (using the Knee Osteoarthritis Outcome Score 0-100) at three months. Secondary outcomes included adherence and safety. The “jump to reference” method was used for the primary analysis using multiple imputation to account for missing data. Intervention effects were calculated using a baseline-adjusted ANCOVA. Bonferroni correction with an alpha level of 0.025 was applied.

Results:

Between Jan 25, 2023 and Aug 11, 2023, 195 participants were enrolled. 98 participants were allocated to re.flex, 97 participants to usual care. The primary analysis included 194 participants. The mean age was 61.9 years (SD 7.7) and the majority were female (68 %, 132/194). Pain reduction was significantly larger in re.flex than in usual care, with an adjusted mean difference between study groups of 4.8 points (95% CI 0.7 to 8.9, P=.02, Cohen’s d 0.35). Improvement in physical function was not statistically significant (β-Coefficient 3.9 points, 95% CI 0.0 to 7.9, P=.049). 12 adverse events were linked to re.flex, none were serious. Participants adhered to 77% (2705/3528) of all scheduled exercise sessions.

Conclusions:

Between Jan 25, 2023 and Aug 11, 2023, 195 participants were enrolled. 98 participants were allocated to re.flex, 97 participants to usual care. The primary analysis included 194 participants. The mean age was 61.9 years (SD 7.7) and the majority were female (68 %, 132/194). Pain reduction was significantly larger in re.flex than in usual care, with an adjusted mean difference between study groups of 4.8 points (95% CI 0.7 to 8.9, P=.02, Cohen’s d 0.35). Improvement in physical function was not statistically significant (β-Coefficient 3.9 points, 95% CI 0.0 to 7.9, P=.049). 12 adverse events were linked to re.flex, none were serious. Participants adhered to 77% (2705/3528) of all scheduled exercise sessions. Clinical Trial: German Clinical Trial Register (DRKS) DRKS00030932; https://drks.de/search/de/trial/DRKS00030932.