Facilitators and Challenges to Adoption of a Digital Health Tool for Opioid Use Disorder Treatment in Primary Care: Mixed Methods Study.

Background: The United States is facing an opioid overdose epidemic resulting in an unprecedented number of preventable deaths. The use of medications including buprenorphine and methadone has proven effective for opioid use disorder (OUD), but many patients struggle to stay in treatment. Novel solutions, such as digital health tools, offer one option to help improve clinic management and improve treatment engagement. Objective: Using a mixed methods approach, we investigated facilitators and barriers to the use of a third-party digital health platform called Opioid Addiction Recovery Support (OARS) to aid OUD treatment engagement and adherence in a primary care setting. Methods: Patient and provider use of OARS was observed for 10 months and summarized using descriptive statistics. Differences in use were assessed using Wilcoxon signed rank tests. Additionally, key informant interviews were conducted with providers who prescribe medication for opioid use disorder (MOUD) and their case managers to understand the facilitators and barriers to implementation. Qualitative data were analyzed using a coding reliability thematic analysis approach. Results: Among 205 patients invited to use OARS, the median age was 37 (IQR 31-44) years, 130 (63.4%) identified as men, and 193 (94.1%) identified as non-Hispanic White. Of these 205 patients, 158 (77.1%) used the app at least 1 time. The median number of days the 158 patients viewed test results was 1 (IQR 1-3), progress was 1 (IQR 0-2), and educational content was 0 (IQR 0-1). The 55 patients whose providers had manually entered their results into OARS when the electronic health record (EHR) integration failed viewed test results (P=.002), progress (P<.001), and educational content (P<.001) more days than the 103 patients who could not view their results in OARS. Providers and the lead case manager reported that OARS increased patient-provider communication, allowed patients to better track their overall MOUD treatment, and enhanced providers' ability to identify patients at risk for relapse. They also acknowledged that the lack of integration between OARS with the EHR resulted in administrative burdens, which impacted provider use of the system. Conclusions: Findings from this study highlight the challenges of successfully implementing OARS with patients who receive MOUD in primary care settings. Our results show a lack of OARS uptake among providers, case managers, and patients, despite positive assessments made by participants. We also show several barriers that impacted provider use, including the lack of integration between OARS and EHR. Future research is needed (1) to determine whether digital health tools like OARS are efficacious in improving OUD outcomes and, if proved efficacious, (2) to identify ways to routinize the use of digital health tools in MOUD treatment, primarily by solving technical and organizational challenges associated with EHR integration and patient engagement.