Abstract
Drug overdose death rates are the highest recorded in New York City (NYC). Substance use disorder (SUD) treatment termination can confer increased risk of drug overdose death. Our objective was to determine the probability of, and factors associated with, drug overdose death following SUD treatment termination. Using a retrospective longitudinal cohort design, we identified those who had NYC-based SUD treatment terminated (01/2016-06/2019) using Chief Medical Examiner and SUD treatment data. Using survival analyses, we examined drug overdose deaths ≤ 14 and ≤ 90 days following SUD treatment termination, respectively. Of 51,171 patients with SUD treatment termination, 140 and 342 had a drug overdose death < 14 and ≤ 90 days, respectively. The crude drug overdose death rate was 26.7 per 1000 person-years at-risk in the ≤ 90-day period and was 71.6 per 1000 person-years at-risk in the ≤ 14-day period. In adjusted Cox proportional hazard model examining death ≤ 14 days, those unemployed (compared to employed) and those terminated from residential treatment (compared to medically supervised withdrawal, opioid treatment programs, and outpatient treatment) were more likely to have had a drug overdose death (all p-values < 0.01). In adjusted Cox proportional hazard model examining death ≤ 90 days, non-Hispanic White people (compared to non-Hispanic Black people), those not stably housed (compared to stably housed), those unemployed and those terminated from residential treatment were more likely to have had a drug overdose death (all p-values < 0.01). Strategies to improve retention including the reassessment of program treatment termination criteria along with strategies to promote ongoing OUD treatment, engagement in harm reduction, and distribution of naloxone are needed.