Is MELD really the definitive score for liver allocation?



The best system for organ allocation is still a controversial issue. The aim of this study was to study the accuracy of four different scores to predict mortality on the waiting list and, thus, their usefulness to determine organ allocation. We retrospectively compared two groups of patients, those who died on waiting list (group D) and those who successfully underwent transplantation (group T) during the same time period. Four scores, at the time of entering the waiting list and just before liver transplantation or death, were evaluated. The evaluated scores were as follows: (1) the Child-Pugh classification; (2) the Model for End-Stage Liver Disease (MELD) score; (3) the Freeman scale; and (4) the Guardiola et al index. The mortality rate on waiting list was 15.9%. All studied scores, except Freeman scale, were higher in group D at the time of entrance on waiting list (MELD, 17.4 ± 8 v 12.3 ± 6, P = .02; Child, 9.9 ± 2 v 7.7 ± 2, P = .002; Freeman, 9.7 ± 4 v 7.3 ± 3.9, P = .09; Guardiola, 2.6 ± 0.9 v 1.7 ± 0.7, P = .001). C-statistics of all scores were similar and in all cases lower than 0.8 (MELD, 0.75; Child, 0.78; Freeman, 0.65; Guardiola, 0.79). None of the studied scores have an excellent accuracy to predict prognosis of patients on waiting list, mainly in case of populations with high proportion of hepatocellular carcinoma. Although the MELD score is rapidly available, standardized, and objective, it does not reflect the severity of patients with cancer or metabolic disorders.