Abstract
INTRODUCTION: The Donor Risk Index (DRI) is a widely used liver transplant allograft risk model but does not account for the increasing adoption of machine perfusion (MP). METHODS: Using Bayesian updating, we incorporated MP into the DRI framework (DRI-MP). A Bayesian proportional hazards model with informative priors derived from the original DRI was applied to Organ Procurement and Transplantation Network data from January 2022 to June 2024. Model performance was assessed using Harrell Concordance-statistic, calibration plots, and Brier scores. RESULTS: DRI-MP, defined as DRI × 0.7 for MP cases, improved 90-day graft survival discrimination (Harrell Concordance-statistic: = 0.546 vs 0.535, P = 0.040), while maintaining robust calibration. DISCUSSION: The Bayesian-updated DRI-MP modestly improves donor risk discrimination, reflecting contemporary transplant practice and providing an implementable tool with continuity from the original DRI.