Use of Neural Network Models to Predict Liver Transplant Waitlist Mortality

Document Type


Publication Date


Publication Title

Liver transplantation


Current liver transplant (LT) organ allocation relies on MELD-Na scores to predict mortality in patients awaiting LT. This study aims to develop Neural Network (NN) models that more accurately predict LT waitlist mortality. The study evaluates patients listed for LT between February 27(th), 2002 and June 30(th), 2021 using the OPTN/UNOS registry. We excluded patients listed with MELD exception scores and those listed for multi-organ transplant, except for liver-kidney transplant. Subset of data from the waitlist was used to create a mortality prediction model at 90 days after listing with 105,140 patients. A total of 28 variables were selected for model creation. The data was split using random sampling into training, validation, and test datasets in a 60:20:20 ratio. The performance of the model was assessed using Area Under Receiver Operating Curve (AUC-ROC) and Area Under Precision-Recall curve (PR-AUC). AUC-ROC for 90-Day mortality was 0.936 (0.934-0.937, 95% CI), and PR-AUC was 0.758 (0.754-0.762, 95% CI). The NN 90-Day Mortality model outperformed MELD-based models for both AUC-ROC and PR-AUC. The 90-Day Mortality model specifically identified more waitlist deaths with a higher Recall (Sensitivity) of 0.807 (0.803-0.811, 95% CI) vs 0.413 (0.409 - 0.418, 95% CI) (P<0.001). The performance metrics were compared by breaking the test dataset into multiple patient subsets by Ethnicity, Gender, Region, Age, Diagnosis Group, and Year of listing. The NN 90-Day Mortality model outperformed MELD-based models across all subsets in predicting mortality. In conclusion, organ allocation based on NN modeling has the potential to decrease waitlist mortality and lead to more equitable allocation systems in LT.

Medical Subject Headings

Transplant and Abdominal Surgery

PubMed ID



ePub ahead of print