Reducing Substance Use-Related Harms: A Simulation-Optimization Framework for the Design and Evaluation of Harm Reduction Vending Machines.

This article presents a novel simulation-optimization framework developed to optimize the allocation of harm reduction supplies in opioid-focused harm reduction vending machines (HRVMs) aimed at persons who inject drugs (PWID). Using agent-based modeling combined with mathematical optimization, the study evaluates the impact of distributing naloxone kits, fentanyl test strips (FTS), and safer injection kits on reducing HIV, hepatitis C virus (HCV), and overdose-related health burdens measured in disability-adjusted life-years (DALYs). Results indicate that optimal HRVM inventory allocation varies with local fentanyl prevalence and user engagement with FTS, with approximately half the capacity allocated to FTS, about a quarter to naloxone, and a quarter to safer injection kits under baseline conditions. Scenarios including safer injection kits and referrals to addiction treatment services showed the greatest reductions in disease burden and societal costs. The framework supports adaptable, data-driven HRVM deployment strategies to improve health outcomes among PWID while addressing logistical and policy challenges.