Establishment and validation of a 90 d mortality prediction model for HBV-associated acute-on-chronic liver failure
Objective To establish a predictive model for 90 d mortality in patients with HBV associated acute on-chronic liver failure(HBV-ACLF)and to evaluate its performance.Methods A total of 276 patients with HBV-ACLF treated in our hospital from Jan.2019 to Jun.2022 were analysis retrospectively.Based on the prognosis at 90 d of ad-mission,276 patients with HBV-ACLF were divided into death group(n=126,45.65％)and survival group(n=150,54.35％).Clinical data at admission were collected from both groups,and a multi-factor Logistic regression model was used to screen the risk factors for death in HBV-ACLF patients,and a prediction model was constructed accordingly.The ROC curve method was used to evaluate the efficacy of the prediction model.Results Univariate analysis showed that compared with the survival group,patients in the death group had higher age(＞75 years old),proportion of hepatic encephalopathy,D-D,prothrombin time,lactate,NLR,RDW,CTP score and MELD score,and they also had lower albumin.The difference was statistically significant(P＜0.05).Multi-factorial Logistic regression analysis showed that age＞75 years old,D-D＞2.10 mg/L,lactate＞5.25 mmol/L,MELD score＞25,hepatic encephalopathy and NLR＞5 were independent risk factors for death at 90 d in patients with HBV-ACLF.ROC curve analysis showed that the AUC of the model predicting 90 d death in HBV-ACLF patients was 0.880(95％CI:0.837-0.922)with 81.52％accuracy at internal validation and 0.828(95％CI:0.773-0.884)with 76.74％accuracy at external validation.Conclusion The model constructed based on age,D-D,lactate,MELD score,hepatic encephalopathy and NLR can predict the risk of death at 90 d in patients with HBV-ACLF,which can help clinical recognition of critical illness and assist in the refined management of patients.
NETL
NSTL
万方数据
