A Novel prognostic scoring system to predict 3-month mortality risk in patients with acute-on-chronic liver failure in hepatitis B: a retrospective cohort study.

PURPOSE: The present study was done to establish an objective, sensitive prognostic scoring system and to determine the applicability of this model in predicting the 3-month mortality of patients with acute-on-chronic liver failure in hepatitis B (ACLFB). METHODS: We developed a novel prognostic scoring system, calculated from six clinical indices including serum total bilirubin, prothrombin activity, serum creatinine, hepatic encephalopathy, infections, and the depth of ascites from 499 patients with ACLFB. Differences in the sensitivity, specificity, and practicality of a Novel prognostic scoring system and the model of end-stage liver disease (MELD) were analyzed. RESULTS: The areas under the receiver operating characteristic curve (ROC) for the Novel scoring systems and MELD scoring systems were 0.967 (95% CI, 0.956-0.977) and 0.900 (95% CI, 0.878-0.922), respectively. The analysis of the ROC curve indicated that the Novel scoring systems were an exact, pertinent, and objective prognostic model with greater accuracy than the MELD. In the Novel scoring systems, the survival rate of these patients whose scores ranged from 2 to 6 was 98.80%, while for those whose scores point at 7 and 15, the mortality rates were 8.70% (2/23) and 95.45% (21/22), respectively, and the mortality rate of these patients whose scores were 16 and above was 100.00%. However, in the MELD prognostic scoring systems, there were no score ranges with 100.00% survival rate. CONCLUSIONS: We developed an objective, pertinent, and sensitive prognostic scoring system that predicted the 3-month mortality of patients with ACLFB with greater accuracy than the MELD.