A prospective comparison of 3 scoring systems in emergency department patients with UPPER gastrointestinal bleeding

Background / Aims: Acute gastrointestinal bleeding is a potentially life-threatening abdominal emergency that remains a common cause of hospitalization. Upper gastrointestinal bleeding (UGIB) is highly associated with mortality. Several scoring system have been produced for predicted mortality in patients with UIGB. The aim of our study was to compare Glasgow-Blatchford Score (GBS), pre-endoscopic Rockall Score (PRS), and Rockall Score (RS) in predicting 28-day mortality caused by UIGB in emergency department (ED). Methods: We designed a prospective study to compare the performance of the GBS, PRS, and RS in predicting 28-day mortality. Study was conducted at the ED of Celal Bayar University Hafsa Sultan Hospital. A total of 210 consecutive UGIB patients were applied to our hospital in period of 1-year (May 2014-May 2015). The area under the receiver operating characteristic curve (AUROC) and sensitivity and specificity for each system were analyzed. Results: The study included a total of 210 patients, with 65 (31%) women and 145 (69%) men. All-cause mortality at the 28-day follow-up evaluation was 17.1%. A significant assosiation between this 3 scores (GBS, PRS, and RS) and mortality was found. For prediction of mortality, AUROC was obtained for GBS (0.705), PRS (0.768), and RS (0.814) (p <0.001, p < 0.001, p <0.001 respectively). A sensitivity of a GBS score of ≥10.5 for mortality was 78%, with a specificity of 53% (p <0.001). A sensitivity of a PRS score of ≥3 for mortality was 85%, with a specificity of 59% (p <0.001). A sensitivity of a RS score of ≥6 for mortality was 76%, with a specificity of 77% (p ≤0.001). Conclusions: GBS, PRS, and RS are all valuable for predicting 28-day mortality caused by UGIB at ED presentation. The 3 scores were found similar in predicting mortality yet RS higher specificity and higher AUROC values among others.