Long-term buprenorphine treatment for OUD and adverse events following discontinuation among Medicaid beneficiaries

Arthur Robin Williams, M.D.
Department of Psychiatry, Columbia University

ABSTRACT:
Background: Buprenorphine reduces the risk of overdose and death among patients with opioid use disorder, yet to date, an empiric basis for the optimal length of treatment is lacking. Adverse health outcomes following buprenorphine discontinuation were compared among patients who were successfully retained beyond six months of continuous treatment akin to the "Retention" stage of the OUD Cascade of Care.

Methods: Retrospective longitudinal cohort within MarketScan multi-state US Medicaid claims (2013-2017) covering 12 million beneficiaries annually. The sample included adults 18-64 years old who received buprenorphine continuously for ≥180 days by cohorts retained for 6-9 months (n=4,126), 9-12 months (n=2,440), 12-15 months (n=1,499), and 15-18 months (n=931) with claims extending 6 months following discontinuation to assess outcome events. Primary adverse outcomes included all-cause emergency department visits, inpatient hospital admissions, filled opioid prescriptions, and overdose, controlling for demographic characteristics and comorbid mental health and substance use diagnoses.

Results: Rates of adverse events were high across all cohorts following buprenorphine discontinuation with approximately half of patients (42-50%) seen in the emergency department at least once during the 6 months following discontinuation. Compared to those retained for 6-9 months, patients retained for 15-18 months had lower odds of all-cause emergency department visits (OR 0.70, 95%CI 0.60-0.90, p<0.001), all-cause inpatient hospitalizations (OR 0.80, 95% CI 0.6-1.0, p<0.05), and filling opioid prescriptions (OR 0.70, 95%CI 0.60-0.80, p<0.001) in the six months following discontinuation. Approximately 5% of persons across all cohorts experienced one or more medically treated overdoses in the 6 months following buprenorphine discontinuation. Analysis was limited to beneficiaries in multiple unidentified states and could not account for mortality outcomes.

Conclusion: Risk of acute care service use and overdose were high following buprenorphine discontinuation irrespective of treatment duration. Superior outcomes became significant with treatment duration beyond 15 months but remained concerning suggesting a minimum of 6 months is insufficient. Improvements at all levels of care are needed to increase retention of patients on buprenorphine.