MELD score is better than Child-Pugh score in predicting 3-month survival of patients undergoing transjugular intrahepatic portosystemic shunt

Background/Aims: Patients undergoing transjugular intrahepatic portosystemic shunt (TIPS) are at risk of early death due to end-stage liver failure. The aim of this study was to compare model of end-stage liver disease (MELD) and Child-Pugh scores as predictors of survival after TIPS. Methods: We studied 140 cirrhotic patients treated with elective TIPS. Concordance (c)-statistic was used to assess the ability of MELD or Child-Pugh scores to predict 3-month survival. The prediction of overall survivals was estimated by comparing actuarial curves of subgroups of patients stratified according to either Child-Pugh scores or MELD risk scores. Results: During a median follow-up of 23.7 months, 55 patients died, 14 underwent liver transplantation and seven were lost to follow-up. For 3-month survival, the discrimination power of MELD score was superior to Child-Pugh score (0.84 vs. 0.70, z = 2.07; P = 0.038). Unlike Pugh score, MELD score identified two subgroups of Child C patients with different overall survivals (P = 0.027). The comparison between observed and predicted survivals showed that MELD score overrates death risk. Conclusions: MELD score is superior to Child-Pugh score as predictor of short-term outcome after TIPS. Its accuracy, however, decreases for long-term predictions.