With small diameter endoscopes, transnasal esophagogastroduodenoscopy (t-EGD) is routinely performed. The aim of this prospective observational study was to evaluate the role of t-EGD for upper gastrointestinal bleeding (UGIB).

One hundred and forty-five consecutive patients (mean age, 66±18.4 years) with suspicion of UGIB were classified a priori into 3 groups according to initial clinical presentation: (1) intensive care unit with EGD under sedation, (2) endoscopy unit with EGD under transient sedation and (3) unsedated t-EGD as “first look”. Demographic, clinical and biological parameters, Rockall and Blatchford scores, endoscopic diagnosis and treatment, and outcome were analysed.

Unsedated t-EGD was attempted in 89 patients, performed in 52 (5 failures, 28 contraindications) and the procedure was converted under sedation for 2 patients. Based on ASA classification, clinical (blood pressure, hemodynamical failure) and biological variables (hemoglobin, platelets, creatinine), these patients were less severe than in the other groups. Pre-endoscopic Rockall and Blatchford scores were significantly lower in this group. More patients in this group presented significant cardiovascular co-morbidity (47.2%), taking aspirin, clopidogrel and/or anticoagulant.

Our results strongly support that “first look” unsedated t-EGD can avoid unnecessary sedation in selected patients with UGIB, presenting a low probability for endoscopic haemostatic treatment and high sedation risks.