[Prognosis of patients with bleeding caused by rupture of esophageal and gastric varices: a cohort study]

Abstract

Objectives: To determine the factors of long-term survival in patients with bleeding caused by rupture of esophageal and gastric varices and evaluate the effects of surgical treatment, endoscopic treatment, TIPS, and combined therapy on the long-term survival.

Methods: Retrospective cohort study was used to analyze 202 patients with bleeding caused by rupture of esophageal and gastric varices who were admitted to Zhongshan Hospital from 1992 to 1999. All of the patients were divided into several groups according to the intervention they received: no-intervention group, surgery group, endoscopy group, TIPS group, and combined treatment group. The survival curves of different groups were compared. Multivariate analysis was conducted.

Results: Kaplan-Meier curves demonstrated that the combined treatment group and surgery group had better prognosis than the endoscopy group and control group (P < 0.01). There was no statistically significant difference in survival rate between the combined treatment group and surgery group and between the endoscopy group and control group. Child-Pugh stratification showed that the results were the same for patients in Child-Pugh A class and B class. Multivariate analysis based on Cox proportional hazard model showed that there were four independent variables associated with prognosis: prothrombin time, HBV-DNA, ascites, and operation.

Conclusions: Surgery, sometimes combined with other treatment, may improve the prognosis of patients with bleeding caused by rupture of esophageal and gastric varices. The four factors associated with the prognosis of patients with bleeding caused by rupture of esophageal and gastric varices are: prothrombin time, HBV-DNA, ascites, and operation.