Clinical profile and predictors of mortality in upper gastrointestinal bleeding presenting to the emergency department of a tertiary care center ‒ a prospective observational study

Abstract

Introduction and aim. Upper gastrointestinal bleeding (UGIB) is a common and potentially life-threatening emergency associated with significant morbidity and mortality. Prospective emergency department (ED)-based data from India on factors associated with in-hospital mortality remain limited, and comparative evidence between bleeding-specific scores such as the Glasgow-Blatchford Score (GBS) and triage systems like the Emergency Severity Index (ESI) is scarce. This study aimed to describe the clinical profile of patients presenting with UGIB and to evaluate predictors of in-hospital mortality, with a focus on comparing the prognostic performance of GBS and ESI. Material and methods. This single-center prospective observational cohort study was conducted in the ED of a tertiary care center in North India from January 2024 to January 2025. Adult patients with clinically suspected or endoscopically confirmed UGIB were enrolled. Clinical, laboratory, and management data were recorded using a standardized form. GBS and ESI were assigned at presentation by trained clinicians prior to outcome assessment. The primary outcome was in-hospital mortality. Predictive performance was assessed using receiver operating characteristic (ROC) curve analysis. Results. Eighty-three patients with UGIB were included; 55.4% had a GBS >2. Patients with higher GBS demonstrated greater physiological derangement, increased transfusion requirements, higher incidence of shock, and significantly higher mortality (39.1% vs. 0%, p<0.001). All non-survivors were triaged as high acuity by ESI and had qSOFA ≥2 at presentation. GBS showed good discriminative ability for predicting mortality (AUROC=0.785), outperforming ESI (AUROC=0.723). Conclusion. GBS showed good performance in predicting in-hospital mortality and may aid early ED risk stratification. However, findings should be interpreted cautiously given the single-center design and small sample size.