Objective To analyze the influencing factors of hepatic encephalopathy (HE) after transjugular intrahepatic portosystemic shunt (TIPS) and construct a nomogram prediction model based on these factors.

Methods A total of 290 patients with cirrhotic portal hypertensive variceal gastrointestinal bleeding in the Yuncheng Central Hospital Affiliated to Shanxi Medical University from January 2019 to December 2023 were selected and randomly divided into training set of 145 cases and validation set of 145 cases. All patients underwent TIPS treatment, and the incidence of HE within 3 months after TIPS was recorded. In the training set, patients were divided into HE group (n=42) and non-HE group (n=103) based on the occurrence of HE. Clinical materials were compared between the two groups, and Lasso-Logistic regression analysis was applied to explore the influencing factors of HE after TIPS. A nomogram prediction model was constructed based on the influencing factors and validated in both the training set and the validation set for its clinical value in predicting HE after TIPS.

Results The overall incidence of HE was 29.31%, with incidence rates of 28.97% and 29.66% respectively in the training set and the validation set. In the training set, the HE group had significantly higher age, C grading of preoperative Child-Pugh ratio, diabetes mellitus ratio, total bilirubin (TBIL), prothrombin time (PT), serum sodium, serum creatinine, interleukin-6 (IL-6), interleukin-18 (IL-18), blood ammonia, monocyte chemotactic protein-1 (MCP-1), postoperative portal venous pressure, and intestinal flora disturbance ratio when compared to the non-HE group, while the preoperative glial fibrillary acidic protein (GFAP) level was significantly lower in the HE group (P < 0.05). Lasso-Logistic regression analysis showed that preoperative C grading of Child-Pugh grading, diabetes mellitus, TBIL, PT, IL-6, IL-18, blood ammonia, GFAP, MCP-1 level, and postoperative intestinal flora disturbance were influencing factors for HE after TIPS (P < 0.05). A nomogram prediction model was constructed based on ten influencing factors selected by Lasso-Logistic regression analysis. The area under the curve (AUC) of this model for predicting HE after TIPS was 0.933 (95%CI, 0.889 to 0.976) in the training set and 0.944 (95%CI, 0.893 to 0.995) in the validation set, with good consistency between the model's prediction and actual observation.

Conclusion The nomogram prediction model for HE after TIPS, constructed based on the influencing factors selected by Lasso-Logistic regression analysis, has high predictive efficacy and accuracy.