Does exposure to opioid substitution treatment in prison reduce the risk of death after release? A national prospective observational study in England

Background and Aims: People with opioid use disorder (OUD) in prison face an acute risk of death after release. We estimated whether prison-based opioid substitution treatment (OST) reduces this risk. Design: Prospective observational cohort study using prison health care, national community drug misuse treatment and deaths registers. Setting: Recruitment at 39 adult prisons in England (32 male; seven female) accounting for 95% of OST treatment in England during study planning. Participants: Adult prisoners diagnosed with OUD (recruited: September 2010-August 2013; first release: September 2010; last release: October 2014; follow-up to February 2016; n = 15141 in the risk set). Intervention and Comparator: At release, participants were classified as OST exposed (n = 8645) or OST unexposed (n = 6496). The OST unexposed group did not receive OST, or had been withdrawn, or had a low dose. Measurements: Primary outcome: all-cause mortality (ACM) in the first 4 weeks. Secondary outcomes: drug-related poisoning (DRP) deaths in the first 4 weeks; ACM and DRP mortality after 4 weeks to 1 year; admission to community drug misuse treatment in the first 4 weeks. Unadjusted and adjusted Cox regression models (covariates: sex, age, drug injecting, problem alcohol use, use of benzodiazepines, cocaine, prison transfer and admission to community treatment), tested difference in mortality rates and community treatment uptake. Findings: During the first 4 weeks after prison release there were 24 ACM deaths: six in the OST exposed group and 18 in the OST unexposed group [mortality rate 0.93 per 100 person-years (py) versus 3.67 per 100 py; hazard ratio (HR) = 0.25; 95% confidence interval (CI) = 0.10-0.64]. There were 18 DRP deaths: OST exposed group mortality rate 0.47 per 100 py versus 3.06 per 100 py in the OST unexposed group (HR = 0.15; 95% CI = 0.04-0.53). There was no group difference in mortality risk after the first month. The OST exposed group was more likely to enter drug misuse treatment in the first month post-release (odds ratio 2.47, 95% CI = 2.31-2.65). The OST mortality protective effect on ACM and DRP mortality risk was not attenuated by demographic, overdose risk factors, prison transfer or community treatment (fully adjusted HR = 0.25; 95% CI = 0.09-0.64 and HR = 0.15; 95% CI = 0.04-0.52, respectively). Conclusions: In an English national study, prison-based opioid substitution therapy was associated with a 75% reduction in all-cause mortality and an 85% reduction in fatal drug-related poisoning in the first month after release.