A Geographic Weighted Regression Analysis of the Health Opportunity Index and Stroke Prevalence in Health and Human Services Region 3.

Although stroke prevalence remains one of the leading causes of death and morbidity in the United States, there is paucity of ecological studies at the census tract level that elucidate geospatial associations between predictors of stroke prevalence in states across U.S. Health and Human Services Region 3 (HHS Region 3: Delaware, Maryland, Pennsylvania, Virginia, West Virginia, and the District of Columbia). This study operationalized the Health Opportunity Index (HOI) by exploring the geospatial relationship between the 13 indicators of the HOI and stroke prevalence at the census tract level in HHS Region 3 using four HOI indicator profiles: (a) neighborhood and built environment profile, (b) social and community context profile, (c) resource profile, and (d) economic profile. The methodological approach was quantitative using secondary data. The sample size was 8021 census tracts. The HOI was estimated for each census tract in the study area. Geographic weighted regression model was run to examine the varying strengths and direction of geospatial relationship of 13 HOI indicators and stroke prevalence across census tracts in HHS Region 3. The results showed variation in the geographic weighted regression (GWR) local estimated coefficients for each indicator across the study area, reflecting variation in the strength and direction of the associations. The findings of our study can guide the identification of geographic priorities for resource allocation, design of quality improvement interventions, inform policy creation and targeted local strategies for stroke prevention services across neighborhoods, support grant applications, and inform future research on stroke prevalence in HHS Region 3.