Introduction. Liver cirrhosis is a chronic condition associated with severe complications, including esophageal variceal bleeding, which carries high morbidity and mortality rates. Approximately 60% of patients with cirrhosis experience recurrent bleeding within one year, with a mortality rate of 20–60% within six weeks after the initial bleeding episode. Therefore, identifying predictors of recurrent bleeding is essential for improving preventive strategies. This study aimed to identify predictors of recurrent esophageal variceal bleeding in patients with liver cirrhosis and to develop a predictive model. Methods. This retrospective study included 220 patients with liver cirrhosis and a history of esophageal variceal bleeding based on medical record data. The variables analyzed included age, platelet count, red color sign, spontaneous bacterial peritonitis, splenomegaly, number of endoscopic variceal ligation (EVL) bands, portal hypertensive gastropathy, and Child–Pugh score. Statistical analyses were performed to identify independent predictors and to develop a prediction model. Results. Recurrent esophageal variceal bleeding occurred in 33.6% of patients. Patients with recurrent bleeding were more likely to be male, younger, have hepatitis B as the underlying etiology, a Child–Pugh score ≥7, a platelet count <150,000/µL, and ≥3 EVL bands placed during treatment. Bivariate analysis demonstrated that a Child–Pugh score ≥7, platelet count <150,000/µL, ≥3 EVL bands, the presence of a red color sign, and splenomegaly were significantly associated with recurrent esophageal variceal bleeding in patients with liver cirrhosis. The final prediction model consisted of three variables: Child–Pugh score, platelet count, and number of EVL bands, with a maximum total score of 4 points. The model demonstrated good predictive performance, with an area under the receiver operating characteristic curve (AUC) of 0.81. Conclusions. The predictors of recurrent esophageal variceal bleeding in patients with liver cirrhosis were Child–Pugh score, platelet count, and the number of EVL bands. The resulting prediction model demonstrated good statistical performance and may be useful for risk stratification in clinical practice.
Copyrights © 2026