Outcomes in Bloodless Medicine Patients With Acute Gastrointestinal Bleeding

Abstract

Introduction: Red blood cell transfusion is often indicated in acute gastrointestinal bleeding (GIB), but some patients refuse blood transfusion for personal or religious reasons. Limited data exists regarding the clinical management and outcomes in this population. We describe the in-hospital mortality and endoscopic outcomes of patients with overt GIB. Methods: A single center retrospective cohort study was performed using the clinical database of the Bloodless Medicine Referral Center at a university-affiliated hospital. Inclusion criteria were age 18 years or older, inpatient admission between January 2007 and January 2014, an ICD-9 diagnosis code for GIB, and documented refusal of blood transfusion. Results: Ninety-two patients (63% female, 37% male) met inclusion criteria. Mean age was 68 years (range 20-97) with a median length of stay of 5 days (range 0-48 days). A source of bleeding was documented in 86% of patients (35% upper, 51% lower, and 14% undetermined). For medical therapy, 76% received intravenous iron, 75% received erythropoietin, and 30% received aminocaproic acid. In-hospital mortality was 8.6%. Fifty-two patients (56%) underwent endoscopy (45% EGD, 30% colonoscopy, and 25% both). In the endoscopy group, mean admission hemoglobin was 9.2 g/dL (range 4.5-14.7), mean nadir hemoglobin was 7.8 g/dL (range 3.5-14.3) and in-hospital mortality was 1.9%. A source of bleeding was identified on endoscopy in 61% of patients. Of the 40 patients who did not undergo endoscopy, mean admission hemoglobin was 8.9 g/dL (range 3.5-14.9), mean nadir hemoglobin was 7.3 g/dL (range 3-13.2), and in-hospital mortality was 17.5%. Conclusion: In carefully selected Bloodless Medicine patients, endoscopy is safe and effective in the evaluation and treatment of acute GIB.