Determinants of empiric transfusion in gastrointestinal bleeding in the emergency department.

Introduction: Current guidelines for the management of GI bleeding (GIB) recommend restrictive transfusion triggers unless patients have shock or specific comorbidities. However, these studies may not be applicable to Emergency Department (ED) patients. Factors determining transfusion decisions in the ED are poorly understood. We compared baseline characteristics and outcomes between ED patients with GI bleeding transfused at lower or higher empiric hemoglobin levels.

Methods: Single center, retrospective analysis of hospital records from a large tertiary care center of ED patients diagnosed with GIB who underwent red blood cell transfusion in the ED. A pre-transfusion hemoglobin cutoff of 7 g/dl was used to divide patients into restrictive and empirically transfused groups. Demographics, mortality, hospital length-of-stay, and mortality risk estimates were compared between groups.

Results: 175 patients met inclusion criteria, with 120 restrictive patients (68.5%) and 55 liberal patients (31.4%). The sample was 49.7% male, with mean age 67.2 years, similar between groups. Patients in the empiric transfusion group had more acute emergency severity index scores (2.09 vs. 2.3). No difference was found between groups in triage vital signs, pre-endoscopy Rockall scores or mortality estimates, or length of stay. Most common reasons for empiric transfusion from chart review were hypotension and witnessed large hemorrhage.

Conclusions: Patients that were empirically transfused had similar presentations to patients meeting restrictive guidelines, based on review of triage data. Transfusions above restrictive thresholds occurred frequently in our population. Additional studies are required to clarify appropriate criteria to guide transfusions for GIB in the ED.