US Population Size and Outcomes of Adults on Liver Transplant Waiting Lists

Abstract

Importance: Disparities in organ supply and demand led to geographic inequities in the score-based liver transplant (LT) allocation system, prompting a change to allocation based on acuity circles (AC) defined by fixed distances. However, fixed distances do not ensure equivalent population size, potentially creating new sources of disparity.

Objective: To estimate the association between population size around LT centers and waiting list outcomes for critically ill patients with chronic end-stage liver disease and high Model for End-stage Liver Disease (MELD) scores or acute liver failure (ALF).

Design, setting, and participants: This US nationwide retrospective cohort study included adult (aged ≥18 years) candidates for deceased donor LT wait-listed between June 18, 2013, and May 31, 2023. Follow-up was completed June 30, 2023. Participants were divided into pre-AC and post-AC groups.

Exposure: Population size within defined radii around each LT center (150 nautical miles [nm] for participants with high MELD scores and 500 nm for those with ALF) based on AC allocation policy.

Main outcomes and measures: LT candidate waiting list mortality and dropout rate were analyzed using generalized linear mixed-effect models with random intercepts for center and listing date before and after AC implementation. Fine-Gray competing risk regression, accounting for clustering, was used as a secondary model.

Results: The study analyzed 6142 LT candidates (1581 with ALF and 4561 with high MELD scores) during the pre-AC era and 4344 candidates (749 with ALF and 3595 with high- MELD scores) in the post-AC era, for a total of 10 486 participants (6331 male [60.5%]; mean [SD] age, 48.5 [7.1] years). In the high-MELD cohort, being listed at a center in the lowest tertile of population size was associated with increased waiting list mortality in the AC era (adjusted odds ratio [AOR], 1.68; 95% CI, 1.14-2.46). Doubling of the population size was associated with a 34% reduction in the odds of mortality or dropout (AOR, 0.66; 95% CI, 0.49-0.90). These results were consistent with those of the extended Fine-Gray models and were also corroborated by multiple sensitivity analyses. However, there were no significant population density-associated disparities in the ALF cohort.

Conclusions and relevance: In this retrospective nationwide cohort study, being wait-listed in less populated regions was associated with greater mortality among critically ill LT candidates with high MELD scores, underscoring the limitations of allocation systems based purely on fixed distances.

Figure.. Directed Acyclic Graph (DAG) to Inform Theory-Based Variable Selection for Study Models

The exposure (population size within the respective acuity circle), outcomes (death or dropout on the waiting list [WL]), and other covariates in this study are illustrated. An arrow (edge) represents a possible association. For DAGs, the selection of variables (nodes) and the direction of edges are guided by a theory-based approach grounded in empirical literature. The arrows represent adjusted paths between variables. This DAG was drawn receiwith DAGitty, which provides a graphic user interface tailored to draw and analyze causal diagrams. OPO indicates organ procurement organization; SES, socioeconomic status. The directional symbol in the population size circle indicates the exposure of interest, and the line symbol in the WL outcomes circle indicates the outcomes of interest.