ABSTRACT
PURPOSE: Upper gastrointestinal bleeding (UGIB) is one of the most common causes of emergency department (ED) presentation which can lead to a fatal condition. Many clinical scoring systems intended to predict the prognosis of UGIB patients were developed and validated, including Glasgow-Blatchford score (GBS) and Rockall score (RS). In particular, GBS has shown its superiority in prediction of mortality, the necessity of endoscopic intervention and admission, compared with other scoring systems, in recent studies. However, GBS does not include the age of the patient as its component and has clearly shown its efficacy only in subjects under age 70. Hence, we aimed to assess whether GBS could also be used in old age UGIB patients as a useful risk stratifying method as in younger age. METHODS: UGIB patients who visited our ED for one year were retrospectively enrolled in the analysis. Medical records of the subjects were reviewed, and their GBS and clinical RS were calculated. Receiver-operating characteristics (ROC) curve of each score in prediction of high risk UGIB was drawn and area under curve (AUC) was calculated. Correlation analysis of each score and hospital length of stay was also performed. To assess the validity of each score for use in old age patients, all analyses were also performed in subgroups of age over 60 years, and under that. RESULTS: ROC curves suggest that GBS has significant detecting power for high risk UGIB in overall subjects, subgroups of age over 60 and under (p=<0.001 for all, AUC=0.919, 0.935, 0.901, respectively). Otherwise, clinical RS only showed significant results in overall group and subgroup of age over 60 with lower AUC. CONCLUSION: GBS may also be used safely as an initial risk stratifying method in old age UGIB patients visiting the ED, as in other age groups.