Comparison of Four Scoring Systems for Patients With Nonvariceal Upper Gastrointestinal Bleeding.

Introduction Upper gastrointestinal bleeding (UGIB) is a common medical emergency that causes significant deaths and morbidity. Effective risk classification is crucial for clinical decision-making and resource allocation. Several risk assessments, including the Glasgow-Blatchford score (GBS), AIMS65, National Early Warning Score (NEWS), and National Early Warning Score + Lactate (NEWS+L), are widely used, but each has unique strengths and disadvantages. The purpose of this study is to examine the predictive performance of different scoring systems for critical outcomes, including blood transfusion requirements, inpatient admission, and 90-day mortality, in patients with nonvariceal upper GI bleeding (NVUGIB). Method We performed a retrospective review of 229 individuals who presented with nonvariceal upper GI hemorrhage. Baseline demographics, clinical presentations, laboratory values, and vital signs were gathered. For each patient, GBS, AIMS65, NEWS, and NEWS+L scores were calculated. The predictive accuracy of these scores for blood transfusion, inpatient admission, and 90-day mortality was evaluated using the area under the receiver operating characteristic curves (AUCs). Results The results show that the GBS had the highest predictive accuracy for blood transfusion (AUC: 75.7%), while NEWS was the best predictor for inpatient admission (AUC: 84.04%). For 90-day mortality, NEWS and NEWS+L performed similarly, with AUCs of 77.25% and 77.52%, respectively. AIMS65 demonstrated low predictive capacity across outcomes, although it was less successful than other ratings for specific outcomes. Conclusion Our results show that each risk score has distinct predictive strengths: GBS for transfusion, NEWS for admission, and NEWS/NEWS+L for mortality. Combining these scores may improve risk classification and direct-focused therapies, hence improving patient outcomes in UGIB.