Introduction: This study aims to assess the effect of individual and geographic-level social determinants of health on risk of hospitalization in the Veterans Health Administration primary care clinics known as the Patient Aligned Care Team. Methods: For a population of Veterans enrolled in the primary care clinics, the study team extracted patient-level characteristics and healthcare utilization records from 2015 Veterans Health Administration electronic health record data. They also collected census data on social determinants of health factors for all U.S. census tracts. They used generalized estimating equation modeling and a spatial-based GIS analysis to assess the role of key social determinants of health on hospitalization. Data analysis was completed in 2018. Results: A total of 6.63% of the Veterans Health Administration population was hospitalized during 2015. Most of the hospitalized patients were male (93.40%)and white (68.80%); the mean age was 64.5 years. In the generalized estimating equation model, white Veterans had a 15% decreased odds of hospitalization compared with non-white Veterans. After controlling for patient-level characteristics, Veterans residing in census tracts with the higher neighborhood SES index experienced decreased odds of hospitalization. A spatial-based analysis presented variations in the hospitalization rate across the Veterans Health Administration primary care clinics and identified the clinic sites with an elevated risk of hospitalization (hotspots)compared with other clinics across the country. Conclusions: By linking patient and population-level data at a geographic level, social determinants of health assessments can help with designing population health interventions and identifying features leading to potentially unnecessary hospitalization in selected geographic areas that appear to be outliers.
|Original language||English (US)|
|Number of pages||8|
|Journal||American journal of preventive medicine|
|State||Published - Jun 2019|
Bibliographical noteFunding Information:
We would like to thank our colleagues at the Department of Veterans Affairs Clinical Systems Development and Evaluation and Veterans Affairs Puget Sound Health Care System in Seattle, Washington for their support during this project. The Johns Hopkins University School of Public Health, Center for Population Health IT performed this research under contract to the U.S. Department of Veterans Affairs as a case study that included data assessments and analysis, as well as stakeholder input as necessary. Any recommendations put forth in this paper are the views of the authors and do not necessarily represent the views of the U.S. Department of Veterans Affairs. Veterans Affairs had no disclosure of potential conflicts of interest. No financial disclosures were reported by the authors of this paper.
© 2019 American Journal of Preventive Medicine