乙型肝炎性肝硬化患者轻微型肝性脑病预测模型的构建

Abstract: Objective To improve the early screening of mild hepatic encephalopathy (MHE) in patients with hepatitis B cirrhosis by constructing a predictive model for MHE. Methods A retrospective analysis was conducted on 380 patients with hepatitis B cirrhosis admitted to our hospital from January 2020 to June 2023. All patients were followed up for one year and divided into an MHE group and a control group based on whether MHE occurred. The clinical characteristics and cognitive function status of the two groups were compared. Multivariate logistic regression analysis was used to study the risk factors of MHE, and a predictive model for MHE was constructed and validated based on the Nomogram framework. Results Child Pugh C liver function, blood ammonia＞72 μmol/L, liver fibrosis index 4 (FIB-4 index) ＞3.25, upper gastrointestinal bleeding, Stroop Color Word Test (CWT) - total time＞130.5 s, and interference effect (SIE) accuracy＞5 were independent risk factors for MHE (P＜0.05). The area under the curve (AUC) of the training set was 0.891 (95% confidence interval: 0.847～0.935), and the AUC of the validation set was 0.893 (95% confidence interval: 0.827～0.960). Conclusion The Nomogram prediction model for MHE in patients with hepatitis B cirrhosis based on relevant risk factors has good predictive performance and credibility.