D’Amico G, Maruzzelli L, Airoldi A, et al. - Although discrimination of the model for end stage liver disease (MELD) is generally regarded as acceptable, its calibration remains unknown, researchers evaluated the discrimination and calibration performance of 3 versions of the model in a validation study: original MELD-TIPS, used to predict survival after transjugular intra-hepatic portosystemic shunt (TIPS); classic MELD-Mayo; MELD-UNOS, used by United Network for Organ Sharing (UNOS). In addition, recalibration and model updating were explored. Seven hundred seventy-six patients submitted to elective TIPS (TIPS cohort), and 445 unselected patients (non-TIPS cohort) were involved. The MELD performance in this validation study was largely unsatisfactory, particularly in non-viral cirrhosis. Candidate variables for MELD updating and MELD recalibration are proposed.