[A nomogram prediction model for individualized prediction of the risk of covert (minimal) hepatic encephalopathy occurrence in patients with liver cirrhosis].

Objective: To construct an individualized nomogram prediction model for predicting the risk of the occurrence of covert hepatic encephalopathy (CHE) in patients with liver cirrhosis. Methods: 325 cases of liver cirrhosis admitted from January 2020 to December 2022 were selected as the study subjects. Patients were divided into training (n=213) and validation (n=112) sets using a cluster randomization method. The risk factors for CHE occurrence in patients with cirrhosis in the training set were analyzed by univariate and multivariate logistic regression. A prediction model related to the nomogram was established. Results: Independent risk factors for the occurrence of CHE in patients with cirrhosis were a history of hepatic encephalopathy, co-infection, gastrointestinal bleeding, severe ascites, prothrombin time ≥16 seconds, high total bilirubin, and high blood ammonia levels (P<0.05). Nomogram model validation results: The model had a net benefit for the training and validation sets, with C-indices of 0.830 (95%CI: 0.802-0.858) and 0.807 (95%CI: 0.877-0.837), respectively, within the range of 0-96%. The calibration curves of both sets were evenly close to the ideal curves. The AUCs for the ROC curves in both sets were 0.827 (95%CI: 0.796-0.858) and 0.811 (95%CI: 0.787-0.836), respectively. Conclusion: Patients with cirrhosis have many risk factors for CHE occurrence. The nomogram model constructed based on these risk factors possesses a good predictive value for assessing CHE occurrence in cirrhotic patients.