Buprenorphine-naloxone vs. extended-release naltrexone for opioid use disorder in individuals with and without criminal legal involvement: A secondary analysis of the X:BOT randomized controlled trial.

ABSTRACT

INTRODUCTION: There is uncertainty about whether criminal legal involvement (CLI) impacts the effectiveness of medications for opioid use disorder (MOUD). We aimed to determine whether CLI modifies the association between buprenorphine-naloxone (BUP-NX) vs. extended-release naltrexone (XR-NTX) and MOUD treatment outcomes. METHODS: We conducted a secondary analysis of XBOT, a 24-week multi-center randomized controlled trial comparing treatment outcomes between BUP-NX (n = 287) and XR-NTX (n = 283) in the general population. We used baseline Additional Severity-Index Lite responses to identify patients with recent CLI (n = 342), defined as active CLI and/or CLI in the past 30 days, and lifetime incarceration (n = 328). We explored recent CLI and lifetime incarceration as potential effect modifiers of BUP-NX vs. XR-NTX effectiveness on relapse, induction, and overdose. We conducted both intention-to-treat and per-protocol analyses for each outcome. RESULTS: In intention-to-treat analyses, recent CLI modified the effect of BUP-NX vs. XR-NTX on odds of successful induction (p = 0.03) and hazard of overdose (p = 0.04), but it did not modify the effect on hazard of relapse (p = 0.23). All participants experienced lower odds of successful induction with XR-NTX compared to BUP-NX, but the relative likelihood of successful induction with BUP-NX was lower than XR-NTX among individuals with recent CLI (OR 0.25, 95 % CI 0.13-0.47, p < 0.001) compared to those without recent CLI (OR 0.04, 95 % CI 0.01-0.19, p < 0.001). Participants with recent CLI experienced similar hazard of overdose with XR-NTX and BUP-NX (HR 1.12, 95 % CI 0.42-3.01, p = 0.82), whereas those without recent CLI experienced greater hazard of overdose with XR-NTX compared to BUP-NX (HR 12.60, 95 % CI 1.62-98.03, p = 0.02). In per-protocol analyses, recent CLI did not modify the effect of MOUD on hazard of overdose (p = 0.10) or relapse (p = 0.41). Lifetime incarceration did not modify any outcome. CONCLUSIONS: Compared to individuals without recent CLI, individuals with recent CLI experienced decreased relative effectiveness of BUP-NX compared to XR-NTX for induction and overdose outcomes. This highlights the importance of considering the impact of recent CLI on opioid use disorder treatment outcomes. Future research should explore the mechanisms through which recent CLI modifies MOUD effectiveness and aim to improve MOUD effectiveness for individuals with recent CLI.