Background: Upper gastrointestinal bleeding (UGIB) while being hospitalized is common, resulting in increased disability and mortality. Using the clinical risk scoring system to assess these patients is, therefore, important in assessing the necessity for endoscopic intervention, predicting rebleeding rate, and 30-day mortality rate.
Objective: To assess the effectiveness of the clinical risk scoring system in terms of predicting the need of endoscopic intervention, rebleeding rate, and 30-day mortality in in-hospital UGIB patients.
Materials and Methods: The patients with in-hospital UGIB who were admitted during August 2014 to August 2018 and underwent esophagogastroduodenoscopy within 72 hours after bleeding onset were included. The data were retrospectively collected. A comparison of the effectiveness of the clinically significant bleeding criteria, Glasgow-Blatchford score, and AIMS65 score was done.
Results: One hundred and sixteen patients were included into the study. The main cause of UGIB was stress-related mucosal disease (SRMD) (88.8%). The effective system of predicting the necessity for endoscopic intervention was the clinically significant bleeding criteria (sensitivity 84%, specificity 74%). The effective scoring system for predicting rebleeding rate and 30-day mortality was the AIMS65 score (sensitivity 83%, specificity 59%).
Conclusion: In-hospital UGIB is a common condition, mostly associated with SRMD. Patients with clinically significant bleeding would benefit from endoscopic intervention.

Keywords: Upper gastrointestinal bleeding, In-hospital UGIB, Clinically significant bleeding, GBS, AIMS65