A columnar graphical prediction model for hepatic encephalopathy secondary to decompensated cirrhosis in hepatitis B cirrhosis.

BACKGROUND AND AIM: To explore the risk factors for secondary hepatic encephalopathy in the decompensated phase of hepatitis B cirrhosis, and to apply column-line diagrams to construct and validate the clinical prediction. METHODS: A retrospective design was conducted on patients with hepatitis B cirrhosis in the decompensated stage who were hospitalized in the Second Hospital of Hebei Medical University between June 2018 and June 2023. Independent risk factors, identified through Lasso regression and multivariable logistic analysis, were utilized to construct a nomogram. The model’s performance was evaluated by plotting ROC, calibration, and decision curve analysis (DCA) curves, and by calculating metrics including the area under the curve (AUC), sensitivity, specificity, accuracy, positive predictive value (PPV), and negative predictive value (NPV). RESULTS: History of diabetes, upper gastrointestinal bleeding, lung infection, renal insufficiency, blood ammonia (AMMO), and Child-Turcotte-Pugh (CTP) were identified as independent risk factors for secondary hepatic encephalopathy in the decompensated phase of hepatitis B cirrhosis (P < 0.05). The AUC values of the constructed prediction model in the training set, the internal validation set, and the external validation set were 0.886 [95% CI: 0.8488–0.9232], 0.856 [95% CI: 0.7862–0.9258], and 0.844 [95% CI: 0.7801–0.9078], respectively, which showed good prediction performance. The model fit the calibration curve well, and the DCA threshold was good, indicating that the model has high clinical application value. CONCLUSION: The LASSO-logistic regression prediction model can better individualize the assessment of patients with hepatitis B cirrhosis in the decompensated stage, which has practical application value and generalizability. CLINICAL TRIAL: Not applicable.