Online Program

Prescription drug monitoring programs: identifying local sources of variation in policy impact (306621)

Keywords: Bayesian hierarchical models, prescription drug monitoring programs, opioid overdoses

Opioid policy evaluations often focus on the average effects that state-level changes in laws and policies have on opioid-related harm. However, the benefit of opioid policies is likely not uniform, and may depend on local sources and motives for opioid use, and access to treatment for opioid use disorder. We describe how Bayesian hierarchical models can be leveraged to characterize spatial patterns of opioid overdoses in small areas, and to estimate how the impact of state-level opioid policies may vary by small-area characteristics within states. We focus on a leading response to the opioid crisis: prescription drug monitoring programs (PDMP). We combine state-level policy information with county-level data on opioid overdoses, the opioid supply, economic characteristics, and treatment in 50 states in 2007-2017, to test whether the relationship between adoption of PDMP “best practices” and opioid overdoses differs by county-level access to opioids, socioeconomic characteristics, and treatment. This study will show how a multi-level, spatial approach can be used to identify the optimal combination of resources needed at the state and local level to address the opioid epidemic.