TY - JOUR
T1 - SOCIAL DETERMINANTS OF HEALTH AND CLINICAL HYPERTENSION AND DIABETES PREVALENCE ON CHICAGO'S SOUTH SIDE
AU - Kolak, Marynia
AU - Chua, Rhys
AU - Kamber, Isaac
AU - Tung, Elizabeth L.
AU - Besser, Stephanie
AU - Anyanwu, Emeka
AU - Liao, James
AU - Anselin, Luc
AU - Tabit, Corey
N1 - Publisher Copyright:
© 2019 American College of Cardiology Foundation
PY - 2019/3/12
Y1 - 2019/3/12
N2 - Background: Social determinants of health (SDOH) critically impact healthcare access and outcomes. However, few studies have examined spatial relationships between SDOH and disease prevalence at neighborhood scales. In this study, we merged electronic health record (EHR) data from a major Chicago hospital system with social data at the census tract level. We hypothesized that census tracts with disadvantageous SDOH have greater prevalence of hypertension and diabetes than other nearby tracts. Methods: We calculated a clinical population prevalence for patients diagnosed with hypertension, diabetes, and heart failure for each census tract. We implemented a linear and spatial regression to model prevalence rates for each disease as a function of SDOH, including poverty, minority status, food access, walkability, foreclosure risk, crime, and others. Results: The sample consisted of 279,500 true outpatient visits/encounters and 52,486 unique patients. Clinical population prevalence rates for patients with hypertension had a strong positive association with areas with high rates of poverty, minority, and disability status (Spatial Pseudo R2=0.72). Neighboring tracts with high disease rates are the strongest predictor of cardiovascular-related chronic disease by several orders of magnitude, more than the effects of age and race combined. Diabetes had similar results, though heart failure was not as well predicted using SDOH alone. Conclusion On Chicago's South Side, segregated areas hit hardest by foreclosures, largely comprised of minority homeowners, had high rates of hypertension and diabetes. Neighboring tracts with high rates of hypertension were the strongest predictor of hypertension prevalence. Advanced spatial analysis can identify especially vulnerable areas within a broader population and could be used to target interventions to the patients most in need.
AB - Background: Social determinants of health (SDOH) critically impact healthcare access and outcomes. However, few studies have examined spatial relationships between SDOH and disease prevalence at neighborhood scales. In this study, we merged electronic health record (EHR) data from a major Chicago hospital system with social data at the census tract level. We hypothesized that census tracts with disadvantageous SDOH have greater prevalence of hypertension and diabetes than other nearby tracts. Methods: We calculated a clinical population prevalence for patients diagnosed with hypertension, diabetes, and heart failure for each census tract. We implemented a linear and spatial regression to model prevalence rates for each disease as a function of SDOH, including poverty, minority status, food access, walkability, foreclosure risk, crime, and others. Results: The sample consisted of 279,500 true outpatient visits/encounters and 52,486 unique patients. Clinical population prevalence rates for patients with hypertension had a strong positive association with areas with high rates of poverty, minority, and disability status (Spatial Pseudo R2=0.72). Neighboring tracts with high disease rates are the strongest predictor of cardiovascular-related chronic disease by several orders of magnitude, more than the effects of age and race combined. Diabetes had similar results, though heart failure was not as well predicted using SDOH alone. Conclusion On Chicago's South Side, segregated areas hit hardest by foreclosures, largely comprised of minority homeowners, had high rates of hypertension and diabetes. Neighboring tracts with high rates of hypertension were the strongest predictor of hypertension prevalence. Advanced spatial analysis can identify especially vulnerable areas within a broader population and could be used to target interventions to the patients most in need.
UR - http://www.scopus.com/inward/record.url?scp=85144728520&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85144728520&partnerID=8YFLogxK
U2 - 10.1016/S0735-1097(19)32430-1
DO - 10.1016/S0735-1097(19)32430-1
M3 - Conference article
AN - SCOPUS:85144728520
SN - 0735-1097
VL - 73
SP - 1824
JO - Journal of the American College of Cardiology
JF - Journal of the American College of Cardiology
IS - 9
T2 - The American College of Cardiology 68th Annual Scientific Sessions
Y2 - 16 March 2019 through 18 March 2019
ER -