Implementation of a Mobile Digital Tool Supporting Medication for Opioid Use Disorder Treatment Improves Retention: Stepped-Wedge Cluster Randomized Controlled Trial

Background: Despite its proven efficacy, retention in medication for opioid use disorder (MOUD) remains low, with structural and systemic barriers—such as access to care and treatment setting—alongside individual factors, including personalization and motivation, contributing to high rates of discontinuation. Digital interventions offer a promising approach to address many of these barriers; however, robust evidence for their effectiveness in improving retention and engagement with treatment remains scarce. Objective: This study aims to evaluate the impact of Recovery Connect—a white-labeled version of Recovery Path and a digital remote patient monitoring app used as part of a blended treatment model for opioid use disorder—on patient retention, treatment continuance, and medication adherence. Methods: A stepped-wedge cluster randomized trial was conducted across 9 outpatient MOUD clinics, organized into 8 clusters. Clusters were sequentially transitioned from usual care to a digitally enhanced model incorporating Recovery Connect, which provided real-time monitoring, psychoeducational and skill-based content, and messaging between patients and clinicians. The primary outcome was 30-day retention in treatment following exposure (implementation of the app in the clinic), linkage (downloading and connecting to the app), or engagement (levels of app usage). Secondary outcomes included treatment continuance—defined as receiving at least 75% of expected doses—and the number of daily doses taken within the first 3, 7, and 30 days after admission. Cluster-controlled discrete-time survival analyses were conducted, adjusting for patient- and clinic-level covariates. Results: Patients admitted to clinics that had implemented the app (n=1205) showed increased retention (922/1205, 75.5%) compared with those in clinics that had not (203/319, 63.6%, P<.001). Patients who downloaded and linked with a mental health professional on Recovery Connect had an 81.3% likelihood of retention, compared with 72.0% (P<.001) among those not linked. Linkage also significantly predicted higher treatment continuance and a greater number of daily doses taken during the first 7 and 30 days (P<.001). Low, moderate, and high engagement levels were associated with progressively higher 30-day retention compared with no engagement (P<.001). Conclusions: This study provides evidence that implementing Recovery Connect (Recovery Path) significantly enhances patient retention and treatment continuity in outpatient opioid use disorder care. Early linkage and engagement during the first week were strong predictors of positive outcomes, underscoring the value of early, proactive digital support. These findings reinforce the effectiveness of blended digital-clinical models, aligning with broader evidence that integrating remote monitoring enhances continuity of care and supports recovery. Policy implications include the need for reimbursement mechanisms, workflow integration, and ethical, privacy-preserving implementation to enable scalable and equitable adoption of digital tools in substance use treatment. Trial Registration: ClinicalTrials.gov NCT07140926; https://clinicaltrials.gov/ct2/show/NCT07140926