MELD or MELD-Na as a Predictive Model for Mortality Following Transjugular Intrahepatic Portosystemic Shunt Placement.

Abstract

The model for end-stage liver disease (MELD) was originally developed to predict survival after transjugular intrahepatic portosystemic shunt (TIPS). The MELD-sodium (MELD-Na) score has replaced MELD for organ allocation for liver transplantation. However, there are limited studies to compare the MELD with MELD-Na to predict mortality after TIPS. We performed a retrospective chart review of patients who underwent TIPS placement between 2006 and 2016 at our institution. The primary outcome was mortality, and the secondary outcomes sought to assess which variables could provide prognostic information for mortality after TIPS placement. We performed receiver operating characteristic (ROC) curve analysis to assess the performance of MELD and MELD-Na. There were 186 eligible patients in the analysis. The mean pre-TIPS MELD and MELD-Na were 13 and 15, respectively. Overall, mortality after TIPS was 15% at 30 days and 16.7% at 90 days. In a comparison of the areas under the ROCs for MELD and MELD-Na, MELD was superior to MELD-Na for 30-day (0.762 vs. 0.709) and 90-day (0.780 vs. 0.730) mortality after TIPS. The optimal cutoff score for 30-day mortality was 15 (0.676-0.848) for MELD and 17 (0.610-0.808) for MELD-Na, whereas the optimal cutoff score for 90-day mortality was 16 (95% CI: 0.705-0.855) for MELD and 17 (95% CI: 0.643-0.817) for MELD-Na. There were 24 patients with high MELD-Na ≥17, but with low MELD <15, and 90-day mortality in this group was 8.3%. Although MELD-Na is a superior prognostic tool to MELD for predicting overall mortality in cirrhotic patients, MELD tended to outperform MELD-Na to predict mortality after TIPS.