Abstract P343: E-Epidemiology to study the Associations of Socioeconomic Status with Risk of Cardiovascular Disease Hospitalization Among Patients with Chronic Kidney Disease
Introduction: There has been a paradigm shift in epidemiology to integrate large, disparate sources of information, including electronic health record (EHR) data, to identify, prevent and treat individuals at risk for cardiovascular disease (CVD). This developing field, termed “e-epidemiology,” has the potential to enhance our ability to utilize these resources more efficiently, however limited data exists on how results from EHR based studies compare to traditional cohort studies.
Objective: Quantify the association between socioeconomic status (SES) and CVD hospitalization among patients with chronic kidney disease (CKD) using EHR clinical and public SES data.
Methods: We quantified the odds of incident or recurrent CVD hospitalization for a fatal/nonfatal heart disease or stroke event (defined by ICD-9 codes) among patients with lower vs. higher SES. We included patients ≥40 years old with prevalent CKD receiving care in 2 Durham NC health systems from 2007-13. Annual cross sectional cohorts of patients (new and returning) were created to calculate the probability of a CVD hospitalization in the following year. We measured SES using the AHRQ SES score (based on income, education, occupation, wealth and housing data). In logistic regression models, we estimated the odds of CVD hospitalization in the next year per quartile of SES score, adjusting for demographic and clinical covariates.
Results: Of 60463 patients, 61% were female, 45% were Black, 89% had hypertension, and 41% had diabetes. Examined continuously, the probability of a CVD hospitalization was inversely associated with SES score (Figure 1). Patients in the lowest quartile of SES had a 32% greater odds of CVD hospitalization in the next year as compared to patients in the highest quartile of SES (OR = 1.32, 95% CI=1.08-1.61).
Conclusions: Our results support published data from traditional cohort studies that lower SES is associated with CVD hospitalizations. The EHR is a feasible, efficient and robust data source that may have great potential to advance CVD research.
Author Disclosures: N.A. Bhavsar: None. B.A. Goldstein: None. M. Phelan: None. L.E. Boulware: None.
- © 2017 by American Heart Association, Inc.