Disability and Health Outcomes in Geospatial Analyses of National County Health Data

Background:Persons with disabilities tend to be at risk for secondary conditions, including significantly higher morbidity and mortality. There is a need for comprehensive disability and health databases, including geographic information systems to evaluate trends in health, functioning, and employment.

Purpose: We evaluated county levels in morbidity and mortality across the United States using correlations/spatial correlations between specific health measures and disability and demographic/socioeconomic variables.

Significance: The research augments CDC Disability and Health GIS systems to measure Healthy People 2020 outcomes for persons with disabilities nationwide. Beyond OLS regression analysis, the study of spatially related county-level data tends to inflate R². Spatial regression represents a robust approach for improved analysis of geographic data for population health measures.

Methodology: We merged 2013 National County Health Rankings, U.S. Census, Veterans Administration, and the 2012 Social Security Administration’s Report on SSDI Beneficiaries, all for n = 3,221 U.S. county units. We used GeoDa to regress county-level mortality rates and other health outcomes on the independent variables percent disability, veterans, rural, uninsured, healthcare access, mental health services, minorities, and gender. Statistics included spatial R², Moran’s I, and measures of multicollinearity.

Conclusions: The principal model of factors impacting mortality yielded an adjusted R² = 0.55 (F = 481.4, p < .001) with low multicollinearity (11.5). Variables with significant beta coefficients included counties with low literacy, high minority, rural, and high diabetes rates. The results demonstrate correlations between county-level conditions and health outcomes, supporting previous research. Geospatial information can assist policymakers to apply health education interventions.