A hierarchy of state-level medical desert factors predicting mortality in the United States, 2023

Background

Medical deserts in the United States represent regions where populations face significant barriers to accessing adequate healthcare services, leading to disparities in health outcomes. Whereas previous medical desert models focused on isolated factors such as provider shortages, this study validates a hierarchical model that integrates several domains to better predict mortality outcomes.

Methods

This ecological study utilized state-level, cross-sectional data from calendar year 2023 to examine the relationship between factors associated with medical deserts and mortality outcomes. Standardized predictor variables were grouped into four domains: lack of facilities, insufficient staffing, lack of system access, and disproportionate costs. Hierarchical regression models predicted preventable death, all-cause mortality, and infant mortality, with each domain entered sequentially.

Results

Hierarchical models demonstrated significant predictive utility, accounting for 83% of the variance in preventable death, 71% in all-cause mortality, and 51% in infant mortality. Variables related to system access and disproportionate costs significantly improved predictive power beyond facility and staffing metrics. Key predictors across models included lack of broadband access and medical debt, both of which were associated with higher mortality. The percentage of health provider shortage areas in a state was a significant predictor of preventable death and infant mortality.

Discussion

Findings from the current study provide initial validation of the predictive utility of a hierarchical model in providing actionable steps for addressing medical deserts. Results further challenge previous models that focused primarily on facility- and staffing-based metrics, emphasizing the importance of systemic factors such as affordability and accessibility. Prioritizing interventions that address access and cost barriers could more effectively reduce disparities in mortality.

Key messages

• Variables assessing lack of access and disproportionate cost significantly improved the variance accounted for in mortality above models solely based on a lack of facilities or lack of providers.

• Whereas availability of facilities and providers are indicators of medical deserts, these factors may be insufficient to account for disparities in outcomes when compared to access and cost barriers.