Urgent endoscopy is associated with lower mortality in high-risk but not low-risk nonvariceal upper gastrointestinal bleeding

Background and study aims: The role of urgent endoscopy in high-risk nonvariceal upper gastrointestinal bleeding (NVUGIB) is unclear. The aim of this study was to determine whether esophagogastroduodenoscopy (EGD) performed sooner than the currently recommended 24 h in high-risk patients presenting with NVUGIB is associated with lower all-cause in-hospital mortality. Methods: All adult patients undergoing EGD for the indications of coffee-grounds vomitus, hematemesis or melena at a university hospital over an 18-month period were enrolled. Patients with variceal and lower gastrointestinal bleeding were excluded. Data were prospectively collected. Results: A total of 934 patients were included. The area under the receiver operating characteristic curve (AUROC) for the Glasgow-Blatchford score (GBS) was 0.813 for predicting all-cause in-hospital mortality, with a cut-off score of >= 12 resulting in 90% specificity. In low-risk patients with GBS < 12, presentation-to-endoscopy time in those who died and in those who survived was similar. In high-risk patients with GBS of >= 12, presentation-to-endoscopy time was significantly longer in those who died than in those who survived. Multivariate analysis of the high-risk cohort showed presentation-to-endoscopy time to be the only factor associated with all-cause in-hospital mortality. For high-risk patients, the AUROC for presentation-to-endoscopy time in predicting all-cause in-hospital mortality was 0.803, with a sensitivity of 100% at the cut-off time of 13 h. All-cause in-hospital mortality in high-risk patients was significantly higher in those with presentation-to-endoscopy time of > 13 h compared with those undergoing endoscopy in < 13 h from presentation (44% vs. 0%; P < 0.001). Conclusions: Endoscopy within 13 h of presentation was associated with lower mortality in high-risk but not low-risk NVUGIB.