Endoscopic clipping versus injection and thermo-coagulation in the treatment of non-variceal upper gastrointestinal bleeding: a meta-analysis
- Correspondence to:
Professor Joseph J Sung
Department of Medicine and Therapeutics, Prince of Wales Hospital, Shatin, NT, Hong Kong;
- Accepted 8 May 2007
- Revised 30 April 2007
- Published Online First 12 June 2007
Background: Hemoclips, injection therapy and thermocoagulation (heater probe or electrocoagulation) are the most commonly used types of endoscopic hemostasis for the control of non-variceal gastrointestinal bleeding.
Aim: To compare the efficacy of hemoclips versus injection or thermocoagulation in endoscopic hemostasis by pooling data from the literature.
Method: Publications in the English literature (MEDLINE, EMBASE and Cochrane Library) as well as abstracts in major international conferences were searched using the keywords “hemoclips” and “bleeding”, and 15 trials fulfilling the search criteria were found. Outcome measures included: initial hemostasis (after endoscopic intervention); recurrent bleeding; definitive hemostasis (no recurrent bleeding until the end of follow-up); the requirement for surgical intervention; and all-cause mortality. The heterogeneity of trials was examined and the effects were pooled by meta-analysis.
Results: Of 1156 patients recruited in the 15 studies, 390 were randomly assigned to receive clips alone, 242 received clips combined with injection, 359 received injection alone, and 165 received thermocoagulation with or without injection. Definitive hemostasis was higher with hemoclips (86.5%) than injection (75.4%; RR 1.14, 95% CI 1.00–1.30), or endoscopic clips with injection (88.5%) compared with injections alone (78.1%; RR 1.13, 95% CI 1.03–1.23), leading to a reduced requirement for surgery but no difference in mortality. Compared with thermocoagulation, there was no improvement in definitive hemostasis with clips (81.5% versus 81.2%; RR 1.00, 95% CI 0.77–1.31). These estimates were robust in sensitivity analyses. There was also no difference between clips and thermocoagulation in rebleeding, the need for surgery and mortality. The reported locations of failed hemoclip applications included posterior wall of duodenal bulb, posterior wall of gastric body and lesser curve of the stomach.
Conclusion: Successful application of hemoclips is superior to injection alone but comparable to thermocoagulation in producing definitive hemostasis. There was no difference in all-cause mortality irrespective of the modalities of endoscopic treatment.
Published Online First 12 June 2007
Conflict of interest: none declared.