The Efficiency and Cost-Effectiveness of Wearable Sensors in a Digital Physiotherapeutic Total Hip Arthroplasty-Specific Training System for Patients After Total Hip Arthroplasty: Randomized Controlled Trial.

Background: With total hip arthroplasty (THA) volumes rising worldwide, scalable home rehabilitation strategies are needed. Home-based digital physiotherapy can improve recovery after THA, and adding wearable motion sensor feedback may further enhance exercise performance and adherence. However, the impact of sensor-augmented digital rehabilitation on patient outcomes and cost-effectiveness remains unclear. Objective: This study aimed to evaluate the clinical effectiveness (including adherence) and cost-effectiveness of adding wearable sensor feedback to a home-based digital THA rehabilitation program compared with the same digital program without sensors. Methods: We conducted a single-center randomized controlled trial in Shanghai from June 2023 to June 2024. A total of 240 patients who had undergone primary THA were randomized (1:1) to a 12-week home exercise program delivered via a mobile app either with wearable motion sensors for real-time feedback (intervention) or without sensors (control). Both groups received identical exercise content and weekly teleconsultations. Outcomes were assessed at 6, 12, and 24 weeks by blinded evaluators. The primary outcome was the Hip Disability and Osteoarthritis Outcome Score (HOOS) at 24 weeks. Secondary outcomes included HOOS subscales, timed up-and-go (TUG), Berg Balance Scale, 36-item Short Form Health Survey (SF-36) physical and mental scores, Hospital Anxiety and Depression Scale (HADS) anxiety/depression, patient satisfaction, adherence metrics, and total 24-week costs. Intention-to-treat analyses were used. Mixed effects models and χ² tests were used for group comparisons. Cost-effectiveness was evaluated from a societal perspective. Results: The sensor-based rehabilitation group showed significantly greater improvement in the primary outcome (HOOS overall) than the control group at 24 weeks (P=.01). Early postoperative gains were larger: At 6 weeks, 4 of the 5 HOOS subscales, TUG time, and all patient-reported outcomes remained significant after Bonferroni correction for 47 secondary comparisons (adjusted P<.001). By 24 weeks, hip-specific functional differences had narrowed, but SF-36 physical and mental scores (both P<.001) and HADS anxiety and depression scores (both P<.001) remained significant after correction. Total 24-week costs were similar between the groups (¥87,967 vs ¥94,396 [US $12,879.80 vs $13,821.10] per patient; P=.32). The sensor intervention was associated with numerically lower costs on average (¥6428.98 [US $941.31] less per patient), yielding negative incremental cost-effectiveness ratios, though the cost difference was not statistically significant. Adherence was high in both groups but higher with sensors: The intervention group completed more exercise sessions (P=.002) and showed greater participation in weekly assessments and follow-up calls (both P<.001). Adverse events were uncommon in both groups (5.8% vs 7.5%) and mostly minor, with no serious events reported. Conclusions: Augmenting home-based digital rehabilitation with wearable sensor feedback led to statistically significant improvements in the primary hip function outcome and in patient-reported quality of life and psychological outcomes that persisted after Bonferroni correction through 24 weeks. Several early hip-specific secondary outcome differences did not survive correction for multiple comparisons. The between-group HOOS differences were below the established minimal clinically important difference. Exercise adherence was significantly higher in the sensor group. These benefits were achieved with no increase in cost, suggesting the sensor-enhanced program is a safe and feasible approach to THA rehabilitation that enhances adherence and accelerates early recovery.