Abstract
OBJECTIVE: Endoscopic therapy can reduce the risks of rebleeding, continued bleeding, need for surgery, and mortality. The objective of this systematic review was to compare the different modalities of endoscopic therapy for GI bleeding. METHODS: Studies were identified by searching electronic databases MEDLINE. We selected all available clinical studies published after 2000 that assessed efficacy and/or safety of different endoscopic hemostatic techniques in treating GI bleeding. The outcomes evaluated included initial hemostasis, rebleeding rate, and 30-day all-cause mortality. Network meta-analyses were performed to summarize the treatment effects. RESULTS: Total 20 studies involving 1845 patients were evaluated. Ten different treatment categories including mechanic, ablative, injection, and combined therapy were compared in our analysis in terms of their efficacy in stopping bleeding and complications. Band ligation [rate: 0.757; 95% Credible Interval (0.565, 0.887)] and injection therapy [rate: 0.891; 95% CI (0.791, 0.944)] had inferior efficacy in attaining initial hemostasis compared to others. Combined therapy of band ligation and HPC and hemoclip may represent the best options for preventing rebleeding and mortality respectively. No significant difference was found among other treatments in terms of complications. CONCLUSIONS: We recommend the application of hemoclips in treating GI bleeding due to its high hemostasis efficacy and low risk of 30-day mortality.