Cost Effectiveness of Sertraline for Preventing Domestic Violence Reoffending Among Highly Impulsive Violent Offenders in Australia: A Trial-Based Economic Evaluation.

Abstract

Background: Domestic violence generates significant health and justice costs, yet there is little trial-based clinical or economic evidence on pharmacological interventions for high-risk offenders. We evaluated the cost effectiveness of sertraline for reducing domestic violence reoffending among men with serious violent offending histories and high impulsivity.

Objective: The primary objective was to estimate the incremental cost per domestic violence reoffence prevented for sertraline compared with placebo from a government payer perspective over 24 months.

Methods: We conducted a cost-effectiveness analysis alongside a 24-month, double-blind, randomised controlled trial in New South Wales, Australia, comparing sertraline (n = 319) with placebo (n = 311). Both randomised groups received identical clinical support. A prespecified non-randomised group (N-Rand; n = 166) included eligible men who did not proceed to randomisation. The primary economic outcome was incremental cost per domestic violence reoffence prevented from a government perspective. Uncertainty was characterised using a probabilistic sensitivity analysis.

Results: Domestic violence reoffending was lower with sertraline than placebo (28.2% vs 35.9%; risk ratio 0.79, p = 0.039) and N-Rand (42.2%; risk ratio 0.67, p = 0.002). Relative to placebo, sertraline generated net savings of Australian dollars (A$) 4443 per participant (A$57,636 per domestic violence reoffence prevented), with criminal justice savings (A$4839) exceeding medication costs (A$396). Relative to N-Rand, net savings were A$3327 per participant (A$23,841 per domestic violence reoffence prevented). In priobabilistic sensitivity analysis, sertraline (compared to placebo) had a mean incremental cost of -A$4507 (95% uncertainty interval -A$9,433, A$249) and 0.077 domestic violence reoffences prevented (95% confidence interval 0.004, 0.149). Sertraline was dominant in 95.1% of simulations, and in 84.9% versus N-Rand.

Conclusions: Allocation to sertraline was associated with lower government costs and fewer domestic violence reoffences than placebo among men with serious violent offending histories and high impulsivity. Each reoffence prevented represents a reduction in violence experienced by victims, partners and children. These findings support consideration of sertraline as an adjunct within intervention and diversion programmes for high-risk offenders, where clinically indicated.

Clinical trial registration: Australian New Zealand Clinical Trials Registry Identifier: ACTRN12613000442707.