Baseline pain medication is associated with longer duration of high adherence in a three-month digital treatment program for hip and knee osteoarthritis.

Objective: To investigate whether baseline use of pain medication is associated with program adherence during a three-month digital treatment program for individuals with hip or knee osteoarthritis (OA). Design: An observational cohort study using registry data on weekly participant adherence from 33078 participants enrolled in a digital education and exercise therapy program. Poor adherence was defined as completing less than 80 % of the approximately 20 prescribed weekly activities for two consecutive weeks during the 13-week treatment period. Baseline analgesic use was categorized into six groups: no medication, paracetamol (with/without dietary supplements), NSAIDs (with/without supplements), paracetamol combined with NSAIDs (with/without dietary supplements), dietary supplements only, and opioids (with/without other medications). Interval-censored parametric survival models adjusted for baseline characteristics were used for statistical analysis. In sensitivity analyses, alternative definitions of poor adherence were used. A complete case analysis was conducted as a subgroup analysis. Results: Compared to the no-medication group, individuals taking paracetamol-either alone (hazard ratio 0.94, 95 % CI 0.91, 0.98) or in combination with NSAIDs (0.91, 0.87, 0.94) reached poor adherence later. In contrast, opioid users (hazard ratio 1.12, 95 % CI 1.06, 1.19) reached poor adherence earlier. Adjusted median days to reach poor adherence ranged from 39.2 (95 % CI 36.9, 41.4) for opioids to 49.2 (47.7, 50.7) for paracetamol + NSAIDs users. Alternative definitions of poor adherence and a complete-case analysis generally yielded similar findings. Conclusion: Baseline use of paracetamol, alone or with NSAIDs, was associated with longer time to reach poor adherence, whereas opioid use predicted poor adherence earlier.