Background and aims: MELD allocation system has changed the clinical consequences on waiting list (WL) for LT, but its impact on mortality has been seldom studied. We aimed to assess the ability of MELD and other prognostic scores to predict mortality after LT.
Methods: 301 consecutive patients enlisted for LT were included, and prioritized within WL by using the MELD-score according to: hepatic insufficiency (HI), refractory ascites (RA) and hepatocellular carcinoma (HCC). The analysis was performed to predict early mortality after LT (8 weeks).
Results: Patients were enlisted as HI (44.9%), RA (19.3%) and HCC (35.9%). The major aetiologies of liver disease were HCV (45.5%). Ninety-four patients (31.3%) were excluded from WL, with no differences among the three groups (p = 0.23). The remaining 207 patients (68.7%) underwent LT, being HI the most frequent indication (42.5%). HI patients had the shortest length within WL (113.6 days vs 215.8 and 308.9 respectively; p<0.001), but the highest early post-LT mortality rates (18.2% vs 6.8% and 6.7% respectively; p<0.001). The independent predictors of early post-LT mortality in the HI group were higher bilirubin (OR = 1.08; p = 0.038), increased iMELD (OR = 1.06; p = 0.046) and non-alcoholic cirrhosis (OR = 4.13; p = 0.017). Among the prognostic scores the iMELD had the best predictive accuracy (AUC = 0.66), which was strengthened in non-alcoholic cirrhosis (AUC = 0.77).
Conclusion: Patients enlisted due to HI had the highest early post-LT mortality rates despite of the shortest length within WL. The iMELD had the best accuracy to predict early post-LT mortality in patients with HI, and thus it may benefit the WL management.