Pre-endoscopic assessment to predict outcome of acute upper gastrointestinal bleeding--a retrospective study in patients subject to emergency endoscopy outside regular working hours

Objective: To evaluate the predictive value of pre-endoscopic risk scores in patients with upper gastrointestinal bleeding (UGI-B).

Methods: The medical records of patients evaluated by emergency esophago-gastro-duodenoscopy (E-EGD) for suspected UGI-B outside regular working hours were retrospectively analysed.

Results: During the 75 months of the study period 112 E-EGDs met the inclusion criteria. Mean age was 65.5 ± 14.7 years. 38.4 % of patients were female. Endoscopy revealed 41 gastro-duodenal ulcers, 16 Mallory-Weiss, 13 varices, 4 neoplasia. 72 patients received transfusions, 39 had endoscopic interventions. 2 patients were surgically treated, 16 had recurrent bleeding. 16 of the 110 patients died during hospitalisation. The following sensitivities were found for the Blatchford score (cut-off > 1), the clinical Rockall score (cut-off > 0) and the Adamopoulos score (cut-off > 2) in predicting need for clinical intervention (endoscopic or surgical intervention or transfusion): 100/97.7/93%, recurrent bleeding: 100/100/93.8%, in-hospital mortality: 100/93.8/93.8%, respectively.

Conclusions: The Blatchford score is a suitable tool in determining the need for clinical intervention and the risk of recurrent bleeding and death in patients with UGI-B. The clinical Rockall score and the Adamopoulos score (the latter had originally been developed to predict active UGI-B at endoscopy and was used with a lower cut-off in our study) are inferior alternatives.