Leveraging Aggregate Datasets from Electronic Health Records (EHR) to Assess Burden of Illness in Chicago

This paper used Electronic Health Record (EHR) data in the Chicago Health Atlas database, coupled with the Department of Health and Human Services Inpatient Prospective Payment System (IPPS) Provider Summary for the top 100 Diagnosis –Related Groups (DRG) for 2011 and visualization software to estimate the economic burden, and visualization of diabetes in Chicago. Data for this study was obtained from the Chicago Health Atlas (CHA) database. CHA is a shared resource with IRB approval to extract data such as diagnoses, medications, and laboratory tests for patients seen at six healthcare institutions throughout Chicago. This data is extracted from the Electronic Health Record, and is de-identified prior to entry into CHA. Patients are assigned a unique cluster ID and were identified as diabetic or pre-diabetic based on the American Diabetes Federation standards of medical care in diabetes. EHR data was accessed through Structured Query Language queries, using local data extraction methods. Total covered charges were calculated, resulting in projected total covered charges, and total payments. Next, diabetic/prediabetic patients along with elements of the built environment were visualized using ESRI ArcGIS to locate diabetes hotspots along with resource density (resources are defined as grocery stores with produce sections, farmer’s markets, and parks) within Chicago. A total of 16,216 diabetic and pre-diabetic patients were identified through CHA. The volume of EHR data provided by the participating institutions offers a representative sample of the entire city of Chicago. Geographic locations with fewer built environmental resources to support a healthy lifestyle were associated with higher prevalence of diabetes/prediabetes.