A novel measure of spatial accessibility to transplant services predicts population-based access to liver transplant

Geographic variation in liver transplant access in the United States has spurred interest in spatial accessibility to care. There is currently no consensus about which measure should be used for spatial accessibility. We used 2015-2022 data from the Scientific Registry of Transplant Recipients and the National Center for Health Statistics to calculate county listing-to-death ratios (LDRs) for liver transplants. We used a 2-step floating catchment area approach to define a novel measure of spatial accessibility (Spatial Accessibility Ratio, SPAR). We compared this measure to other accessibility measures using generalized linear models and Vuong's non-nested hypothesis test. Across 3108 included counties, SPAR ranged from 0.56 to 9.98; 29% of counties and 65% of the population had a SPAR ≥1 (mean or better accessibility to liver transplant). SPAR outperformed distance ( p <0.001), rurality ( p <0.001), and healthcare resource-based measures ( p <0.001) in predicting population-based transplant access; SPAR remained significantly associated with LDR after adjustment for other county-level factors. Sensitivity analyses revealed that the association between SPAR and LDR was modified by socioeconomic characteristics and geographic region. This measure may be used in future research on spatial accessibility, including developing interventions to improve access to liver transplant for patients in low-accessibility areas.