88EMF Freestanding Versus Hospital-Based Emergency Departments: Comparing Geographic Catchment Areas and Visit Characteristics

Abstract

We compared the geographic catchment areas and patient visit characteristics at hospital-affiliated freestanding emergency departments (FREDs) versus hospital-based emergency departments (HEDs). We performed a retrospective observational analysis of adult patient visits to 19 hospital-affiliated FREDs and 5 HEDs in Colorado over a 1-year period. Catchment areas were defined using a straight marginal approach as Colorado zip codes that served as the zip code of residence for at least 0.5% of an emergency department’s (ED's) annual census. Catchment area characteristics were obtained from the 2016 American Community Survey, and median demographics were compared using the 2-sample Wilcoxon rank-sum test. Patient visit characteristics were abstracted from 2016-17 electronic medical record data, and Chi-squared test was used to compare visit characteristics between FREDs and HEDs. The FRED catchment area included 150 zip codes and the HED catchment area included 82 zip codes, with 70 zip codes overlapping. The catchment areas were similar in median zip code population, percent Hispanic residents, percent of residents with health insurance, educational attainment, and percent non-English speakers. Compared to HED catchment areas, FRED catchment areas had a lower proportion of non-White residents (12.4% versus 17.3%, p=0.005), higher median household incomes ($70,356 versus 58,762, p=0.008), a lower proportion of residents with public health insurance (26.1% versus 30.6%, p=0.033), and a lower proportion of residents below the federal poverty level (8.1% versus 10.8%, p=0.013). On analysis of patient visit characteristics, FREDs had a larger proportion of visits from patients between the ages of 18 and 44 (63% versus 54%, p< 0.001) and fewer visits from patients 80 years and older (3% versus 6%, p<0.001). Similar to catchment area patterns, FREDs had significantly lower proportions of non-White patient visits (24% versus 34%, p<0.001) and patient visits with public insurance (29% versus 39% Medicaid and 14% versus 23% Medicare, p<0.001). FREDs did have a larger proportion of uninsured patient visits than HEDs (12% versus 10%, p<0.001). FREDs had more visits from infrequent ED utilizers, with fewer than 4 visits to any study ED over the year (84% versus 72%, p<0.001), and they had fewer visits from super-utilizers, with 10 or more ED visits (3% versus 7%). Hospital-affiliated FREDs tend to serve communities with higher incomes, fewer minorities, and higher rates of private insurance than do HEDs. These patterns are largely reflected in patient visit characteristics as well.