Acceptability and Feasibility of Implementing Accelorometry-Based Activity Monitors and a Linked Web Portal in an Exercise Referral Scheme: Feasibility Randomized Controlled Trial

Abstract

Background: Exercise referral schemes (ERSs) are recommended for patients with health conditions or risk factors. Evidence points to the initial effectiveness and cost-effectiveness of such schemes for increasing physical activity, but effects often diminish over time. Techniques such as goal setting, self-monitoring, and personalized feedback may support motivation for physical activity and maintenance of effects. Wearable technologies could provide an opportunity to integrate motivational techniques into exercise schemes. However, little is known about acceptability to exercise referral populations or implementation feasibility within exercise referral services. Objective: To determine the feasibility and acceptability of implementing an activity-monitoring device within the Welsh National ERS to inform a decision on whether and how to proceed to an effectiveness trial. Methods: We conducted a pilot randomized controlled trial with embedded mixed-methods process evaluation and an exploratory economic analysis. Adults (N=156) were randomized to intervention (plus usual practice; n=88) or usual practice only (n=68). Usual practice was a 16-week structured exercise program. The intervention group additionally received an accelerometry-based activity monitor (MyWellnessKey) and associated Web platform (MyWellnessCloud). The primary outcomes were predefined progression criteria assessing acceptability and feasibility of the intervention and proposed evaluation. Postal questionnaires were completed at baseline (time 0:T0), 16 weeks (T1), and 12 months after T0 (T2). Routine data were accessed at the same time-points. A subsample of intervention participants and scheme staff were interviewed following the initiation of intervention delivery and at T2. Results: Participants were on average aged 56.6 (SD 16.3) years and mostly female (101/156, 64.7%) and white (150/156, 96.2%). Only 2 of 5 progression criteria were met; recruitment and randomization methods were acceptable to participants, and contamination was low. However, recruitment and retention rates (11.3% and 67.3%, respectively) fell substantially short of target criteria (20% and 80%, respectively), and disproportionally recruited from the least deprived quintile. Only 57.4% of intervention participants reported receipt of the intervention (below the 80% progression threshold). Less than half reported the intervention to be acceptable at T2. Participant and staff interviews revealed barriers to intervention delivery and engagement related to the device design as well as context-specific technological challenges, all of which made it difficult to integrate the technology into the service. Routinely collected health economic measures had substantial missing data, suggesting that other methods for collecting these should be used in future. Conclusions: To our knowledge, this is the first study to evaluate short- and long-term feasibility and acceptability of integrating wearable technologies into community-based ERSs. The findings highlight device- and context-specific barriers to doing this in routine practice, with typical exercise referral populations. Key criteria for progression to a full-scale evaluation were not met.