Estimating a treatment effect on reoffending by people in prison with an alcohol use disorder: a national matched propensity score analysis

Abstract Little is known as to the efficacy of treatment interventions for an alcohol use disorder (AUD) to reduce reoffending. A quasi-experimental propensity score matched observational study was deployed to calculate the average treatment effect on reoffending for people in prison with an AUD exposed to prison-based treatment compared to a matched control group. A one-to-one match without replacement followed by a marginal Cox proportional hazards time-to-event model with the treatment group only were also run. No statistically significant difference in reoffending rates and risk of reoffending on release were noted. Analyses suggest that the treatment group were more likely to be treated for binge drinking than individuals assessed who commit crime due to their consumption of alcohol. The English treatment model is an opportunity to create an enhanced system that integrates public health whilst addressing recidivism. Further work is required assessing the effect of multiple treatments on reoffending.