Assessment of risk stratification scoring systems in upper gastrointestinal bleeding patients in the emergency department

Background Acute upper gastrointestinal bleeding (UGIB) is a common medical emergency presented to the emergency department that requires early assessment and management. Many risk stratification scores have been developed to predict the clinical outcomes in patients with UGIB. The commonly used risk scores the Rockall scoring systems (PRS and FRS), Glasgow–Blatchford score (GBS) and AIMS65 score. Aim The aim of the present study was to assess and compare the ability of the wildly used risk scores the RS, GBS, and AIMS65 to predict the clinical outcomes in UGIB patients Patients and methods One hundred patients (age >18 years) with acute UGIB in the emergency department of Alexandria Main University Hospital were prospectively studied. All the study scores were calculated and compared using the area under the receiver operating characteristic curve (AUC) method to evaluate the performance of each score to predict the mortality, blood transfusion, endoscopic intervention, ICU admission, rebleeding, and length of hospital stay. Results Among the one hundred patients included in the study, 65% were males with a median of age 58 years. 56% had esophageal varices and 63% with liver disease. All the used scores were statistically significant in predicting all clinical outcomes. GBS had the best AUC among the AIMS65, PRS, and FRS scores in predicting mortality with (AUC= 0.80 vs. 0.76, 0.69), blood transfusion need with (AUC= 0.92 vs. 0.88, 0.87), ICU admission with (AUC= 0.86 vs. 0.83, 0.81), rebleeding with (AUC= 0.81 vs. 0.77, 0.69), and length of hospital stay with (AUC= 0.81 vs. 0.75. 0.79). Conclusion All the study scores (GBS, AIMS65, PRS, and FRS) were able to predict the clinical outcomes in the UGIB patients. GBS was the best performing risk score among the four scores for predicting all the clinical outcomes (mortality, blood transfusion, rebleeding, ICU admission, and length of hospital stay) except the prediction of endoscopic intervention in our study population.