Background: A substantial number of hospital admissions end in patient-initiated departure before medical treatment is complete. Whether “before medically advised” (BMA) discharge increases the risk of subsequent drug overdose remains uncertain.

Methods: We performed a retrospective cohort study using administrative health data from a 20% random sample of residents of British Columbia, Canada. We focused on nonelective, nonobstetric hospital stays occurring between 2015 and 2019. We used survival analysis to compare the rate of fatal or nonfatal illicit drug overdose in the first 30 days after BMA discharge versus the rate after physician-advised discharge.

Results: Overall, 6440 of 189 808 (3.4%) hospital stays ended in BMA discharge. Among 820 overdoses occurring in the first 30 days after any hospital discharge, 755 (92%) involved patients with a history of substance use disorder. Unadjusted overdose rates were 10-fold higher after BMA discharge than after physician-advised discharge, and BMA discharge was associated with subsequent overdose even after adjustment for potential confounders (crude incidence, 2.8% v. 0.3%; adjusted hazard ratio [HR] 1.58; 95% confidence interval [CI] 1.31–1.89). Before medically advised discharge was associated with increases in subsequent emergency department visits (adjusted HR 1.92; 95% CI 1.83–2.02) and unplanned hospital readmissions (adjusted HR 2.07; 95% CI 1.96–2.19), but there was no significant association with the uncommon outcomes of fatal overdose and all-cause mortality.