Survival Before and After Model for End-stage Liver Disease Score Introduction on the Brazilian Liver Transplant Waiting List

Abstract Introduction To examine whether the official adoption of Model for End-Stage Liver Disease (MELD) as a criterion for organ allocation was effective, we studied risk factors for patient deaths and the accuracy of the MELD score to predict mortality. Methods Patients on the waiting list for liver transplantation were divided into two periods depending on whether they were on the waiting list before (period 1) or after (period 2) the MELD introduction in Brazil. The Kaplan-Meier method with log-rank tests were used to study patient survivals. Predictive factors were identified using the Cox regression method. A receiver operating characteristic (ROC) curve was used to analyze Child-Turcotte-Pugh (CTP) and MELD accuracy. Results We analyzed 295 patients in period 1 and 240 in period 2. The survivals after 3, 6, 9, and 12 months in periods 1 and 2, were 95.6%, 90.5%, 84.9%, and 69.6% vs 95.7%, 92.1%, 85.3%, and 83.3%, respectively ( P = NS). Multivariate analysis showed CTP, MELD-Na, and albumin levels, besides spontaneous bacterial peritonitis (SBP), to be independent factors related to survival in period 1. In period 2, CTP, creatinine levels, international normalized ratio, besides spontaneous bacterial peritonitis, were the independent factors. The ROC curve for CTP was 0.676 and for MELD, 0.644 ( P = .4) in period 1. In period 2, the ROC curve for CTP was 0.680 and for MELD, 0.718 ( P = .4). Conclusion Patient survival on the waiting list for liver transplantation did not change at 1 year after the introduction of the MELD.