Policy design and implementation of jail-based medication for opioid use disorder: A five-state qualitative...

Justice-involved individuals face dramatically elevated post-release overdose risk, yet correctional facility-based medications for opioid use disorder (MOUD) implementation remains uneven despite strong effectiveness evidence. This study examined policy configurations associated with high implementation in rural correctional systems.

We used fuzzy-set Qualitative Comparative Analysis to examine five rural states: Rhode Island, Vermont, Kentucky, Alabama, and Tennessee. Seven policy domains were calibrated: statutory authority, medication coverage, funding mechanisms, workforce requirements, Medicaid integration, continuity of care, and implementation support. Outcomes reflected weighted state prison and county jail MOUD coverage.

Medicaid integration emerged as a necessary condition for high implementation within this sample (consistency = 0.94). Rhode Island and Vermont achieved universal coverage through distinct configurations: Vermont used statutory mandate with comprehensive infrastructure while Rhode Island relied on sustained appropriations without statutory requirements. Kentucky demonstrated intermediate implementation (63%) despite possessing funding and Medicaid integration, suggesting organizational fragmentation may limit policy translation in dual correctional systems. Alabama and Tennessee showed low implementation (31% and 20%) with multiple policy absences.

This exploratory analysis generates testable hypotheses about policy configurations supporting correctional facility-based MOUD. Medicaid integration may function as critical infrastructure in states with limited healthcare safety nets, though validation in larger samples is needed. Results suggest possible flexibility in legal mechanisms, with both statutory and executive pathways associated with high implementation. Substantial limited diversity (123 of 128 configurations unobserved) and cross-sectional design limit causal inference. Future research should test these patterns in expanded samples and examine mechanisms linking policy features to adoption.