Prescription opioid spillovers from retail pharmacies

Publication id: hodges-2025-prescription-opioid-spillovers
Status: verified

Citation: Hodges, C. & Shakya, S. (2025). Prescription opioid spillovers: Retail pharmacy-level analysis. Journal of Substance Use and Addiction Treatment, 175, 209725. DOI. Download PDF

Facts¶

Policy hook¶

Prescription drug monitoring programs (PDMPs) are state-level initiatives designed to curb opioid misuse. States vary in whether they mandate providers to consult the database before prescribing (“must-access”) or allow voluntary consultation. The question is whether must-access mandates reduce retail opioid sales and whether state-level policy variation generates unintended spillover effects in neighboring states with different PDMP structures.

Main finding¶

Must-access PDMP implementation is associated with a decrease in retail opioid pills dispensed within enacting states. Pharmacies in must-access states show an average decline of 500-1,300 pills per month relative to voluntary PDMP states (depending on panel specification). Geographic spillover effects are present: must-access pharmacies close to voluntary PDMP state pharmacies show a larger decline (approximately 900 additional pills per month) than must-access pharmacies in interior locations. Effects persist across multiple distance specifications (10-35 miles).

Data and setting¶

Monthly pharmacy-level opioid sales from 2006-2014 covering 41,467 unique retail pharmacies in the contiguous United States and DC. Data encompasses all oxycodone and hydrocodone pills, patches, and lozenges shipped as part of each pharmacy transaction, sourced from DEA ARCOS database via Washington Post publication. Treatment states include West Virginia, Kentucky, New Mexico, Tennessee, New York, Indiana, Massachusetts, and Louisiana, which transitioned from voluntary to must-access PDMP structures at different times between 2012-2015.

Research design (plain language)¶

Generalized difference-in-differences design with pharmacy and time fixed effects comparing opioid sales in must-access versus voluntary PDMP states before and after policy implementation. Geographic spillover analysis examines whether must-access pharmacies within various distances (10-35 miles) of voluntary PDMP state pharmacies show different effects than interior must-access pharmacies. Event study framework examines dynamics before and after implementation. Analysis uses both unbalanced and balanced panel specifications, with conservative specification assigning zero sales to pharmacies with no recorded transactions.

One caveat¶

No obvious conclusion regarding the one-year delay in observing effects; could reflect provider and patient adjustment time to new regulatory requirements. Cannot disentangle price versus quantity responses without individual transaction-level data. Relies on pharmacy designation as treatment unit, which may not capture all relevant geographic variation. Small sample sizes for some time periods (Indiana, Massachusetts, Louisiana had limited post-implementation data).

PDF or DOI¶

DOI: 10.1016/j.josat.2025.209725. Download PDF

Why it matters¶

State-level policy variation in PDMP structures generates unintended cross-border consequences. Individuals in must-access states may travel to neighboring voluntary PDMP states to obtain opioid prescriptions, partly undoing the benefits of the state-level mandate. The larger decline in opioid sales at must-access pharmacies close to voluntary PDMP pharmacies is consistent with this cross-border substitution pattern. These findings highlight the need for interstate coordination in opioid policy. Policymakers should recognize that state-level mandates alone cannot address the opioid epidemic without complementary interstate data sharing and coordination, particularly in border regions.