Declining Prescriptions, Rising Deaths: Community Predictors of Opioid Mortality Hotspots In Louisiana (2016-2023)

Background and Purpose: Despite declining opioid prescription rates, Louisiana's opioid mortality rate tripled from 2016 to 2023, suggesting that prescribing policies have not translated into reduced mortality. During this period, accidental illicit opioid deaths rose from 36.5% to 56.5% of all accidental opioid deaths, while prescription opioid deaths fell from 63.1% to 43.4%. In this study, I examined the geographic distribution of opioid mortality and community characteristics associated with spatial clustering across periods of policy change and the COVID-19 pandemic. The goals were to 1) identify how opioid mortality is spatially distributed across Louisiana parishes and ZCTAs and 2) examine community characteristics associated with mortality clustering across four study periods: 2016–17, 2018–19, 2020–21, and 2022–23.

Methods: Using the American Community Survey 5-Year Estimates and the Louisiana Opioid Data Surveillance System, I conducted a repeated cross-sectional analysis. Local indicators of spatial association (LISA) identified spatial clusters of hotspots, outliers, and cold spots in opioid mortality rates. Multinomial logistic regression then examined how community characteristics (rurality, poverty, unemployment, uninsurance, educational attainment, racial composition, and opioid prescription rate) predicted cluster classification from 2016 to 2023. Analyses were conducted separately for each period at both geographic levels; regression was conducted only at the ZCTA level.

Results: Of 6,754 opioid-involved deaths, 6,408 with valid Louisiana ZIP codes were included in spatial analyses. The statewide mortality rate increased from 15.78 to 52.35 per 100,000 residents while prescription rates declined from 115.8 to 79.0 per 100 residents, an inverse relationship that persisted across all four periods. LISA identified significant spatial clustering at both levels, with greater variation at the ZCTA level. Persistent high-high hotspots included ZCTA 70427 (Bogalusa, Washington Parish), where mortality rates rose from 129.8 to 239.1 per 100,000. Community predictors shifted across periods. Unemployment was a significant positive predictor of high-high classification in 2018–19 (OR=1.11, p=0.034) and 2020–21 (OR=1.07, p=0.027). Greater rurality was negatively associated with high-high classification in 2020–21 (OR=0.65, p=0.016), indicating disproportionate urban clustering risk during the COVID-19 transition. Opioid prescription rate did not consistently predict hotspot classification, reaching significance only in 2022–23 (OR=1.05, p=0.002).

Conclusions and Implications: Policies intended to reduce access to prescription opioids have been only partially successful. Despite declining prescriptions, opioid-related deaths continued to rise, and the increased proportion of illicit opioid deaths directly offset reductions in prescription deaths. Unemployment and urban geography were more consistent predictors of hotspot classification than prescription availability, suggesting structural economic conditions shape community vulnerability independent of prescribing patterns. Effective responses require investment in structural neighborhood conditions rather than relying primarily on prescription control. These findings provide direction for social work practitioners and policymakers to prioritize resource allocation in persistently high-mortality communities.