[Sclerotherapy versus ligation in hemorrhage caused by rupture of esophageal varices. Direct meta-analysis of randomized trials].

OBJECTIVES To compare the advantages of endoscopic ligation and endoscopic sclerotherapy for bleeding esophageal varices, published randomized clinical trials were critically reviewed by meta-analysis. Only ten clinical trials concerning a history of recent or active bleeding esophageal varices were included. METHODS The methodology, population, treatment and outcomes of each relevant trial were evaluated by duplicate independent review. RESULTS Endoscopic sclerotherapy compared to banding ligation significantly increased the rate of rebleeding (OR: 1.6; 95% IC: 1.1-2.3) without increasing early mortality compared to endoscopic banding ligation (OR: 1.3; 95% IC: 0.8-1.9). The rate of varice eradication associated with these two types of treatment was not different (OR: 0.9; 95% IC: 0.6-1.3) but was obtained more quickly with endoscopic banding ligation (3.8 +/- 1.6 versus 5.8 +/- 2.2; P < 0.05). The rate of complications was higher after sclerotherapy (OR: 2.5; 95% IC: 1.7-3.7), in those cases with a positive heterogeneity test. CONCLUSIONS This meta-analysis shows a lower morbidity with endoscopic banding ligation in patients with variceal hemorrhage. The most important advantage of endoscopic banding ligation was the reduction of the rate of rebleeding.