WIP 71 - Evaluating Catchment Area Methodologies to Assess Inclusivity of a Large Academic Pediatric Emergency Room

Poster Number: WIP 71
Publication Number: WIP 71.979

Background: There has been increasing focus on health system accountability to healthcare equity. Hospital “inclusivity,” is a concept proposed by the Lown Institute’s “Hospital Ranking for Social Responsibility” that captures the degree to which the demographics of a hospital’s patient population matches the demographics of its community/catchment area, but how best to measure a hospital’s catchment area for inclusivity is unclear.

Objective: We aim to (1) describe twelve definitions of catchment area for our children’s emergency department (ED) and (2) examine inclusivity by race, ethnicity, language, and insurance across each catchment area.

Design/Methods: We conducted a cross-sectional study of all pediatric ED visits (2021-2022) in our large academic children’s hospital. We geocoded visits to census tracts by patient address. We extracted patient-level demographic data (race, ethnicity, language, and insurance) from the EMR.

Our definitions of catchment area were adapted from Lown Institute methods. Census tracts were ordered by the number of unique patients from each tract to determine the last census tract that added significant patients to the cumulative count. The distance from the hospital to the geographic centroid of the last census tract was measured as either the (a) straight-line distance or (b) shortest drive time. We varied this distance by using (i) 100%, (ii) 50%, or (iii) 25% of that distance. The area was defined by (a) all tracts within the straight-line distance or drive time or (b) the smallest convex polygon that enclosed all selected tracts with patients (convex hull).

Prior to April 2024, we will map these twelve catchment areas to depict the distribution of our patient population. We will describe the demographic characteristics of each catchment area using census-tract estimates of neighborhood data from the American Community Survey. Inclusivity indices for race, ethnicity, language and insurance will be calculated as the ratio between the patient-level and census tract-level proportion, weighting census tracts by distance to the hospital.