Mapping Disparities in Medication Access, Utilization, and Spending across the United States

Abstract

Achieving equitable access to medicines requires measuring disparities in pharmaceutical utilization and spending and understanding the mechanisms that drive them. Yet in the United States, disparities in medication access, use, and spending—by race, geography, and disease—remain inadequately measured and insufficiently addressed. This dissertation develops new tools and applies them at scale to quantify these inequities, leveraging large-scale spending, prevalence, and geospatial data to examine underlying mechanisms and support targeted health policy action. Chapter 2 introduces a novel probabilistic framework to map prescription drugs to likely health conditions in retail pharmaceutical claims data, addressing a major limitation of existing administrative datasets. Unlike approaches based on US Food and Drug Administration–approved indications or clinical guidelines, it captures real-world prescribing practices, enabling measurement of both on- and off-label use. The resulting drug-condition map supports national, state, and county-level estimates of condition-specific utilization and spending across the US. Chapter 3 applies this drug-condition map to generate state-level estimates of pharmaceutical utilization and spending by race and ethnicity, producing the most comprehensive geographic mapping of medication inequities to date, covering 143 conditions, four payer types, and all 50 states and DC. Findings reveal persistent underutilization among Black populations relative to need, and substantial variation across states and conditions. Decomposition analyses highlight the relative contributions of disease prevalence, prescribing intensity, spending per prescription. Chapter 4 evaluates the role of physical access to community pharmacies in shaping these disparities over a ten-year period across the US. Using over a decade of geolocated pharmacy data, I develop a metric incorporating both driving and walking proximity to assess access across urban and rural census block groups nationwide. I then examine how access relates to medication use, using a disparity index building on findings from Chapter 3. I find that inadequate pharmacy access is associated with wider racial disparities in medication use across health conditions. Together, these chapters offer a data-rich framework for understanding and addressing inequities in access, use, and spending on prescribed medicines across the United States.