Recalibrated MELD and hepatic encephalopathy are prognostic factors in cirrhotic patients with acute variceal bleeding

Abstract Background & Aims Early TIPS placement must be considered in patients with Child‐Pugh B and active bleeding at endoscopy or in patients with Child‐Pugh C 10‐13 and variceal bleeding. However, active bleeding at endoscopy is a subjective criterion. Moreover, a previous study has shown that a MELD‐based score accurately predicted 6‐week mortality and helped to stratify patients. Using a prospective series of patients included in a multicentre study before the era of early TIPS, we aimed (i) to identify factors associated with 6‐week mortality, focusing on the prognostic value of active bleeding; and (ii) to assess whether a recalibrated MELD‐based score accurately predicted 6‐week mortality. Methods Ancillary study of the prospective multicentre Baveno IV study, including patients with acute variceal bleeding. Results Two hundred and nineteen patients were analysed (Child‐Pugh A/B/C = 18/45/37%). The overall actuarial likelihood of survival on day 42 was 84%. The variability for the diagnosis of active bleeding at endoscopy was high (range, 41.4% to 84.6% among the centres). Active bleeding at endoscopy was not associated with 6‐week mortality in the entire population or in Child‐Pugh B patients. In a multivariate analysis, independent factors associated with mortality were liver function, infection, HE and HCC. The recalibrated MELD‐based score was accurate in predicting 6‐week mortality (AUROC = 0.787). The recalibrated MELD‐based score demonstrated better performance compared to the MELD score. Conclusion The recalibrated MELD‐based score accurately predicted mortality in our prospective cohort. Active bleeding at endoscopy had no prognostic value in cirrhotic patients presenting with acute variceal bleeding. Standardizing active bleeding assessment at endoscopy is warranted.