Opioid agonist therapy switching among individuals with prescription-type opioid use disorder: Secondary analysis of a pragmatic randomized trial

Engagement and retention in opioid agonist therapy (OAT) remains a challenge. This study evaluated the impact of initial randomized OAT allocation on subsequent switching among people with prescription-type opioid use disorder (POUD). Secondary analysis of a 24-week Canadian multicenter, pragmatic, randomized trial conducted between 2017 and 2020 comparing flexible take-home buprenorphine/naloxone versus supervised methadone models of care for POUD. We used Cox Proportional Hazards modeling to assess for impact of treatment assignment on time to OAT switching, adjusting for important confounders. For clinical correlates, we analyzed data from baseline questionnaires on demographic, substance use, and health factors as well as urine drug screen. Of 272 randomized participants, 210 initiated OAT within 14 days per trial protocol, of whom 103 participants were randomized to buprenorphine/naloxone and 107 to methadone. Within 24-week follow-up, 41 (20.5%) of all participants switched OAT with 25 (24.3%, median 27 days, 88.4 per 100 person-years) and 16 participants (15.0%, median 53.5 days, 46.1 per 100 person-years) switching from buprenorphine/naloxone and methadone arms, respectively. In adjusted analysis, allocation to buprenorphine/naloxone was associated with significantly higher risk of switching (aHR = 2.31, 95% CI 1.22 – 4.38). OAT switching was common in this sample of individuals with POUD, with individuals randomly allocated to buprenorphine/naloxone being more than twice as likely to switch versus methadone. This may reflect a stepped care approach in OUD management. More research is needed to evaluate overall retention and outcomes with the different observed risks of switching between methadone and buprenorphine/naloxone. •Opioid agonist therapy rotation within 6 months was common in the OPTIMA trial.•Participants rotated away from buprenorphine/naloxone twice as often as methadone.•Substance use correlations suggest the pattern is due to medication intensification.•The pattern may additionally reflect the relative ease of switching to methadone.•Regional variability of rotation may indicate clinician and systems-level factors.