Purpose: To explore the feasibility of EMR data in epidemiological surveillance of population obesity.

Background: Understanding the distribution and risk factors for obesity in small populations (i.e. cities and counties) has been limited due to small sample sizes in national data that make any type of local geographic inferences impossible. Electronic medical record data provides the opportunity to assess the distribution of obesity and other chronic conditions that may differentially impact certain groups within a city or county geography.

Methods: EMR data was extracted from the Texas Tech University Health Sciences Center El Paso (TTUHSC-EP) and University Medical Center El Paso (UMC) outpatient clinics. A total of 143,524 patient records were geocoded using EMR addresses. Census tract average Body Mass Index (BMI) and obesity prevalence was mapped using ArcGIS. Census tract prevalence was then compared to Behavioral Risk Factor Surveillance Survey (BRFSS) estimates for El Paso MSA.

Results: Tract population EMR representation varied from less than 1% to 34%. Areas with the lowest representation were Ft. Bliss (military installation) and the far East area of El Paso County where there is limited representation of TTUHSC-EP or UMC. Census-tract level obesity prevalence varied from 26.6% to 57.6%.The EMR and BRFSS difference varied from -4.0% to 26.4%. Further analysis will examine significant obesity clusters in El Paso County.

Discussion: Using EMR data for obesity and other chronic disease conditions may effectively pinpoint high risk neighborhoods/areas within small geographies and help inform targeted efforts in prevention. Strengths and limitations will be discussed.