Factors for unsuccessful endoscopic hemostasis in patients with severe peptic ulcer bleeding

Objectives Although the first approach for peptic ulcer bleeding is endoscopic hemostasis, quick determination of a hemostatic strategy is important in patients with vitals indicating shock. However, the unsuccessful factors for endoscopic treatment have yet to be sufficiently examined. We aimed to investigate the factors for unsuccessful endoscopic hemostasis in severe peptic ulcer bleeding.Materials and methods Unsuccessful factors were retrospectively investigated in 150 eligible patients who underwent endoscopic hemostasis for shock-presenting peptic ulcer bleeding at our critical care center between April 2007 and March 2021.Results There were 123 and 27 cases of successful and unsuccessful endoscopic hemostasis, respectively. Causative diseases included gastric ulcer bleeding in 124 patients (82.7%) and duodenal ulcer bleeding in 26 patients (17.3%). Shock index (SI) (1.46 vs. 1.60) (p = .013), exposed blood vessel diameter (1.4 mm vs. 3.1 mm) (p < .001) identified on contrast-enhanced computed tomography (CE-CT), duodenal ulcer bleeding (p = .012), and Forrest classification Ia (p = .004) were extracted as independent factors for unsuccessful endoscopic hemostasis. In receiving operating curve analysis, when the cut-off value for the SI was set at 1.53, the sensitivity and specificity were 70.4% and 63.4%, respectively. When the cut-off value for the exposed blood vessel diameter was set at 1.9 mm, these were 88.9% and 83.7%, respectively.Conclusions When these factors (SI ≥ 1.53, exposed blood vessel diameter ≥1.9 mm identified on CE-CT, duodenal ulcer bleeding, and Forrest Ia) are present in patients with severe peptic ulcer bleeding, non-endoscopic hemostasis, such as interventional radiology (IVR) and surgery, should be considered.