Factors associated with failure of initial endoscopic hemoclip hemostasis for upper gastrointestinal bleeding

Background: Endoscopic hemoclip is widely used for the management of bleeding peptic ulcers. The major difficulty in clinical application of the hemoclip is deployment to the lesion during initial hemostasis. The aim of this study was to define factors associated with the failure of endoscopic hemoclip for initial hemostasis of upper GI bleeding. Patients and Methods: From January to December 2003, we prospectively studied 77 randomized patients with clinical evidence of upper GI bleeding due to either active bleeding or a visible vessel identified by upper GI endoscopy in our emergency department. Results: Among the 77 patients, 13 (16.9%) failed treatment (Group 1) and 64 (83.1%) were successfully (Group 2) treated by endoscopic hemoclip for lesions related to upper GI bleeding. There were no differences due to gender, blood pressure, initial heart rate, and hemo-globulin before or after endoscopic treatment, platelet count, serum creatinine, and albumin between groups. The mean age of Group I was higher than that of Group 2 (73.31 +/- 9.38 years vs. 65.41 +/- 16.45 years, respectively; P = 0.083). Most patients who did not achieve initial hemostasis by endoscopic hemoclip had upper GI lesions over the gastric antrum and duodenal bulb. Among the 13 patients who failed to achieve endoscopic hemoclip initial hemostasis, four lesions were located over the posterior wall of the antrum, and four lesions over the lesser curvature side of the duodenal bulb. Conclusion: Endoscopic hemoclip is an effective hemostatic method for upper GI bleeding. Age, gastric antrum, and duodenal bulb lesions may be associated with the failure of initial hemostasis by endoscopic hemoclip.