内镜套扎与普萘洛尔预防肝硬化食管静脉曲张首次出血效能的Meta分析

目的比较内镜下套扎术（EVL）和普萘洛尔预防肝硬化食管静脉曲张首次出血的疗效和安全性。方法计算机检索1999至2012年12月31日PubMed、EBMBASE、万方数据库中关于EVL和普萘洛尔预防肝硬化食管静脉曲张首次出血的随机对照试验，同时追索纳入文献的参考文献。使用RevMan 5.1专业Meta分析软件对治疗后首次出血率，出血病死率，总病死率和不良反应发生率进行Meta分析。结果纳入6个随机对照试验，包括EVL组280例和普萘洛尔组301例患者。文献的Cochrane质量评价均为B级，属于较高质量文献。Meta分析结果显示，EVL在预防肝硬化食管静脉曲张患者的首次出血率、出血病死率、总病死率和不良反应发生率分别为16.4%、4.3%、22.5%和25.0%，与普萘洛尔组的21.6%（P=0.07）、6.3%（P=0.27）、20.9%（P=0.56）和31.9%（P=0.33）相比均无显著性差异；漏斗图存在不对称的现象，提示可能存在发表性偏倚；敏感性分析表明纳入研究具有稳定性。结论EVL与普萘洛尔在预防肝硬化食管静脉曲张破裂的效能无显著性差异。

Endoscopic variceal ligation versus propranolol for prevention of primary variceal bleeding in cirrhotics:a Meta analysis

Objective To compare the efficacy and safety of endoscopic variceal ligation and propranolol administration for primary prophylaxis of variceal bleeding in cirrhotics. Methods We searched randomized con-trolled trials （RCT）data from PubMed,EBMBASE,Wanfang Database and CNKI database （1999 to 31 October, 2012）to compare endoscopic variceal ligation with propranolol administration for primary prophylaxis of variceal bleeding in cirrhotics,and performed a Meta-analysis using the Cochrane Collaboration＇s RevMan 5.1 software. Results Six RCT involving 581 patients（variceal banding in 280 and propranolol in 301）with cirrhotics were in-cluded in this study. All included trials had high quality for their grade B Cochrane quality evaluation. No dif-ferences were seen on first variceal bleeding rate （EVL/PPL=16.4%/21.6%,P=0.07）,bleeding-related death rate （EVL/PPL=4.3%/6.3%,P=0.27）,overall mortality （EVL/PPL=22.5%/20.9%,P=0.56）and adverse events rate（EVL/PPL=25.0%/31.9%,P=0.33）between the two groups;Publication bias was present with asymmetric funnel plot,but included studies remained stable after sensitivity analyses. Conclusion There are no significant differences in first esophageal variceal bleeding,mortality of bleeding,total mortality and adverse events between endoscopic variceal ligation and propranolol for prevention of primary variceal bleeding in cirrhotics.