Center geography or center practice? Decomposing geographic variation in access to deceased donor liver...

Disparities in deceased donor liver transplant (DDLT) may be due to geographic imbalances of organ supply-to-demand ratios and/or center organ acceptance practices. We evaluated the relative contribution of centers versus broader geographic units (donor service areas [DSAs]/states/organ procurement and transplantation network regions/census divisions) to variability in DDLT rate. Using Scientific Registry of Transplant Recipients data on adult, first-time, non-Status1A, period-prevalent waitlist candidates, we conducted multilevel Poisson regression with center-level and DSA-level intercepts to model DDLT rates across 3 eras: preacuity circles (pre-AC) (May 1, 2017 to April 30, 2019), initial post-AC (February 4, 2020 to February 3, 2022), and recent post-AC (February 4, 2022 to December 31, 2024), adjusting for candidate factors. We calculated center-level and DSA-level median incidence rate ratios (MIRRs) and their relative contributions to total geographic variation. DSA-level MIRR declined from _1.471.52_1.58 to _1.321.37_1.42 to _1.171.23_1.28 from pre-AC to initial post-AC to recent post-AC. The center-level, within-DSA MIRR increased from _1.561.61_1.64 to _1.761.82_1.84 to _1.921.96_2.05 across eras. The DSA-level contribution to total center-level variation declined from 43.7% to 21.5% to 8.4% across eras. Similarly, the contribution of other broad, geographic units (states/organ procurement and transplantation network regions/census divisions) to total geographic variation in DDLT rates decreased after AC. Further allocation changes are unlikely to substantially reduce geographic disparities because DDLT rate variation is predominantly driven by center-level differences, which were associated with highly variable advanced preservation practices.