目的 旨在通过对比单用卡维地洛与卡维地洛联合内镜下静脉曲张套扎术（endoscopic variceal ligation，EVL）预防食管胃静脉再出血的治疗效果，为临床治疗提供策略。 方法 回顾性选取2010年10月—2023年6月于山东大学附属省立医院治疗，单用卡维地洛或卡维地洛联合EVL预防食管胃静脉曲张再出血的178例患者，分为单用卡维地洛组（n=47）和联合治疗组（n=131）；收集两组患者再出血及死亡情况。符合正态分布的计量资料两组间比较采用成组t检验，非正态分布的计量资料两组间比较采用Mann-Whitney U检验。计数资料两组间比较采用χ²检验。Cox比例风险模型用于单变量和多变量分析。采用Kaplan-Meier方法估计再出血和死亡的累积发生率。通过倾向性评分匹配（PSM）对卡维地洛组和联合治疗组进行匹配，使得两组患者基线相似，减小选择偏差并增强因果推断的可信度。 结果 联合治疗组再出血率显著低于卡维地洛组（10年累积发生率：29.8% vs 36.2%，HR=0.505，95%CI：0.292～0.847，P=0.015），但两组在肝脏相关死亡率方面差异无统计学意义（10年累积发生率：21.3% vs 21.4%，HR=0.799，95%CI：0.406～1.578，P=0.518），PSM分析结果进一步验证了上述结论的稳定性。Cox回归分析显示，肌酐是影响再出血（HR=1.004，95%CI：1.001～1.008，P=0.011）和肝脏相关死亡率（HR=1.004，95%CI：1.001～1.007，P=0.019）的独立危险因素。 结论 卡维地洛联用EVL预防食管胃静脉曲张出血效果优于卡维地洛单用。

Objective To compare the therapeutic effects of carvedilol alone and carvedilol combined with endoscopic variceal ligation （EVL） in the prevention of re-bleeding from gastroesophageal varices，and to provide strategies for clinical treatment. Methods We retrospectively included 178 patients who had received carvedilol alone or carvedilol plus EVL to prevent gastroesophageal variceal re-hemorrhage from October 2010 to June 2023. They were divided into carvedilol alone group （47 cases） and carvedilol+EVL group （131 cases）. Between-group comparisons were conducted using the paired t test for normally distributed continuous data，the Mann-Whitney U test for non-normally distributed continuous data，and the chi-square test for categorical data. A Cox proportional hazards model was employed for univariable and multi-variable analyses. The cumulative incidence rates of re-bleeding and mortality were estimated using the Kaplan-Meier method. The baseline characteristics of the two groups were matched through propensity score matching （PSM） to reduce selection bias and enhance the credibility of causal inference. Results The re-bleeding rate of the carvedilol+EVL group was significantly lower than that of the carvedilol alone group （10-year cumulative incidence：29.8% vs 36.2%，hazard ratio ［HR］=0.505，95% confidence interval ［CI］：0.292‍ ‍—‍ ‍0.847，P=0.015）. There was no significant difference in liver-related mortality （10-year cumulative incidence：21.3% vs 21.4%，HR=0.799，95%CI：0.406‍ ‍—‍ ‍1.578，P=0.518）. The results were stable with PSM analysis. The Cox regression analysis revealed that creatinine was an independent risk factor affecting re-bleeding （HR=1.004，95%CI：1.001‍ ‍—‍ ‍1.008，P=0.011） and liver-related mortality （HR=1.004，95%CI：1.001‍ ‍—‍ ‍1.007，P=0.019）. Conclusion Carvedilol combined with EVL is better than carvedilol alone in the prevention of gastroesophageal variceal re-bleeding.