Abstract: Objective To investigate the prognostic factors and prognosis mode in acute-on-chronic liver failure (ACLF). Methods A retrospective study was conducted and included clinical data from 190 patients diagnosed with ACLF between June 2016 and December 2019. Patients were followed up for 90 days and were stratified by prognosis into 89 cases (44.5%) in the death group and 101 cases (50.5%) in the survival group. Logistic regression was used to investigate the prognostic factors influencing the 90-day prognosis of ACLF patients and to construct a new prognostic model. The predictive performance of the model was assessed by plotting the ROC curve. Results The results of multivariate logistic regression analysis suggested that the independent predictors affecting the 90-day prognosis of ACLF patients were age(OR=1.045,P=0.007), HE(OR=4. 341,P＜0.001), PT( OR=1.241,P=0.001), PTA( OR=1.097,P=0.009) and INR( OR=0.532,P=0.027) ( P<0.05). The equation of the new model was : NEWM=-72.662 + 0.044 × age + 0.216 × PT + 0.093 × PTA-0.631 × INR + 1.468 × HE, and the ROC curve analysis showed that the AUC value of this model was 0.81( 95% CI: 0. 74 ~ 0.87) significantly higher than the MELD score of 0.67( 95% CI: 0.59 ~ 0.74), MELD-Na score 0.65( 95% CI:0. 57 ~ 0.73), iMELD score 0.70(95% CI:0. 62 ~ 0.77) and MESO score 0.66( 95% CI:0. 58 ~ 0.74), and the NEWM score had a significantly better predictive ability of 90-day survival in patients with ACLF than other scores. Conclusion Compared to the previous MELD series of scores, the NEWM is a simple and effective tool for predicting the prognosis of patients with ACLF.

Key words: Acute-on-chronic liver failure, Prognostic model, Risk factors