Validation of MELD 3.0 and ReMELD-Na scoring systems: a German clinical cohort study
Discuss this preprint
Start a discussion What are Sciety discussions?Listed in
This article is not in any list yet, why not save it to one of your lists.Abstract
Background
The Model for End-Stage Liver Disease (MELD) score, introduced in 2002, has since been refined. MELD 3.0, launched in 2023 in the United States, improves mortality prediction by incorporating sex and albumin. Since March 2025, Germany started to use reMELD-Na to improve prioritization of patients on the waiting list for liver transplantation (LT). This study compares the performance of original MELD, MELD-Na, MELD 3.0 with and without albumin, and reMELD-Na for patients waitlisted for LT in a large German transplant center.
Methods
This retrospective single-center study included 206 listed patients from 2017 to 2021 for LT. Reclassification patterns along with predictive accuracy for three-month survival and overall survival (OS) of five different MELD scores were assessed using Harrell’s c-index and integrated area under the curve (iAUC).
Results
Over a median follow-up of 33.9 months, 100 patients (51.5%) underwent LT, with a post-transplant survival rate of 70%. Thirty-eight patients (18.4%) received LT and sixteen patients died within the first three months after listing. ReMELD-Na and MELD 3.0 without albumin demonstrated the highest discrimination for three-month survival (c-index 0.848 and 0.827, respectively). Original MELD showed the poorest discrimination. MELD 3.0 without albumin showed the best overall performance in predicting OS (c-index 0.827) in males, while reMELD-Na performed best in females (c-index 0.705). Females with higher original MELD scores tended to receive even higher scores with MELD 3.0.
Conclusion
This is the first German study to validate reMELD-Na and MELD 3.0, showing superior predictive performance over original MELD. MELD 3.0 may better reflect disease severity in women at advanced stages due to more upward reclassification.
