Logo image
Back
Impact of Model for End-stage Liver Disease Cutoffs on Liver Transplant Access in the United States: A Regression Discontinuity Design
Journal article   Peer reviewed

Impact of Model for End-stage Liver Disease Cutoffs on Liver Transplant Access in the United States: A Regression Discontinuity Design

Tomohiro Tanaka, Michael P Jones, George Wehby and David Axelrod
Transplantation
02/02/2026
DOI: 10.1097/TP.0000000000005650
PMID: 41628055

View Online

Abstract

Since Share-35 and continuing under the Acuity Circle (AC) policy, US liver transplant (LT) allocation has relied on Model for End-Stage Liver Disease (MELD) thresholds that may distort fair and seamless access. We analyzed adult LT candidates listed from January 2016 to June 2024. Regression discontinuity analysis using Poisson regression for top-5 offer rates and discrete-time hazard regression for transplant rates and waitlist mortality estimated policy effects at a MELD score of 35 (MELD-35; Share-35) and MELD scores of 33 and 37 (MELD 33/37; AC). Secondary analyses were used to analyze the offer decline rates and donor characteristics at each MELD threshold. In total, 98 896 candidates were analyzed. During Share-35, crossing the MELD-35 threshold was associated with significant increases in offers (rate ratio [RR], 3.39; 95% confidence interval [CI], 3.22-3.57) and transplants (odds ratio [OR], 2.13; 95% CI, 1.89-2.41). Under AC, smaller but significant effects were observed at MELD-37 (RR, 1.37; 95% CI, 1.31-1.44; OR, 1.15; 95% CI, 1.01-1.30) and at MELD-33 (RR, 1.48; 95% CI, 1.21-1.38 and OR, 1.16; 95% CI, 1.03-1.30). Candidates with an MELD score <37 experienced significantly higher waitlist mortality (OR, 0.40; 95% CI, 0.24-0.67). Secondary analyses showed higher decline rates at MELD-35 under Share-35, but lower rates at MELD 33/37 under AC with reduced marginal donor use. MELD thresholds continue to influence LT access and donor quality/selection, creating artificial discontinuities and unintended disparity that misalign allocation with true disease severity. These findings support avoiding rigid score-based cutoffs not only in future US allocation models (eg, continuous distribution), but also in broader international allocation frameworks moving forward.

Details

Metrics

1 Record Views
Logo image