Endoscopic clipping for acute nonvariceal upper-GI bleeding: a meta-analysis and critical appraisal of randomized controlled trials
Journal Articles
Overview
Research
Identity
Additional Document Info
View All
Overview
abstract
BACKGROUND: Acute nonvariceal upper-GI bleeding (NVUGIB) is common, with a high rate of recurrent bleeding and substantial mortality rate. Endoscopic clipping has the theoretical advantage of minimizing tissue injury and is increasingly used. OBJECTIVE: We conducted a systematic review and meta-analysis to investigate any potential benefits of clipping over other endoscopic techniques for NVUGIB. DESIGN: Randomized controlled trials (RCT) that compared clipping with other endoscopic hemostatic methods to treat NVUGIB were included. Summary effect size was estimated by odds ratio (OR) with a random-effects model. RESULTS: Twelve RCTs met inclusion criteria. For peptic ulcer bleeding (PUB), the hemoclip (n = 351 patients) was compared with the heat probe alone, thermal therapy plus injection, and injection alone in 2, 2, and 5 studies, respectively (n = 348 patients). The rate of the initial hemostasis was nonsignificantly increased in the control group compared with the hemoclip group (92% vs 96%, OR 0.58 [95% CI, 0.19-1.75]). The rebleeding rate was nonsignificantly decreased with hemoclips compared with controls (8.5% vs 15.5%, OR 0.56 [95% CI, 0.30-1.05]). Emergency surgery and the mortality rate were not significantly different between the hemoclip and controls. Subgroup analysis conducted in studies that compared hemoclips with injection alone show similar results. Two studies and one study reported outcomes of interest for Dieulafoy's lesions and Mallory-Weiss syndrome, respectively. CONCLUSIONS: RCTs that compared clipping alone with other endoscopic hemostatic techniques for NVUGIB were limited. Current evidence suggests that the hemoclip is not superior to other endoscopic modalities in terms of initial hemostasis, rebleeding rate, emergency surgery, and the mortality rate for treatment of PUB.
