Andrew Dunn, MD
Is a restrictive transfusion strategy more effective and safer than a liberal transfusion strategy in patients with acute upper gastrointestinal (GI) bleeding?
Randomized controlled trial. ClinicalTrials.gov NCT00414713.
Hospital de la Santa Creu i Sant Pau, Barcelona, Spain.
921 patients > 18 years of age (mean age 65 y, 68% men) who were hospitalized with severe, acute upper GI bleeding (hematemesis, bloody nasogastric aspirate, or melena). Exclusion criteria included massive exsanguinating or lower GI bleeding; recent trauma or surgery; Rockall score = 0 (low rebleeding risk) with hemoglobin (Hgb) level > 12 g/dL; or an acute coronary syndrome, symptomatic peripheral vasculopathy, stroke, transient ischemic attack, or transfusion in the past 90 days.
Restrictive transfusion strategy (red blood cell transfusion given only at Hgb level < 7 g/dL, with a posttransfusion target of 7 to 9 g/dL) (n = 461); or liberal transfusion strategy (transfusion given at Hgb level < 9 g/dL, with a posttransfusion target of 9 to 11 g/dL) (n = 460). Transfusion strategies were used until hospital discharge or death.
Primary outcome was all-cause mortality at 45 days. Secondary outcomes were further bleeding (hematemesis or fresh melena, with either hemodynamic instability or ≥ 2 g/dL reduction in Hgb level over 6 h) and in-hospital complications.
97% (intention-to-treat analysis).
49% of patients in the restrictive-strategy group and 86% in the liberal-strategy group had transfusions. At 45 days, the restrictive strategy reduced all-cause mortality, further bleeding, and in-hospital complications compared with the liberal strategy (Table).
A restrictive transfusion strategy was better than a liberal strategy for reducing 45-day mortality and further bleeding in patients with acute upper gastrointestinal bleeding.
Restrictive vs liberal transfusion strategy in patients with acute upper gastrointestinal bleeding†
†Abbreviations defined in Glossary. RRR, NNT, and CI calculated from hazard ratios and control event rates in article.
‡Adjusted for age, in-hospital bleeding, cirrhosis status, and Rockall score.
§Hematemesis or fresh melena, with either hemodynamic instability or ≥ 2 g/dL reduction in hemoglobin level over 6 h.
||Adjusted for age, in-hospital bleeding, cirrhosis status, Rockall score, baseline hemoglobin level, and shock at admission.
¶Unadjusted analysis confirmed by author.
Andrew Dunn. A restrictive transfusion strategy reduced 45-day mortality in patients with acute upper GI bleeding. Ann Intern Med. 2013;158:JC6. doi: 10.7326/0003-4819-158-6-201303190-02006
Download citation file:
Published: Ann Intern Med. 2013;158(6):JC6.
Gastroenterology/Hepatology, Hospital Medicine.
Results provided by:
Copyright © 2017 American College of Physicians. All Rights Reserved.
Print ISSN: 0003-4819 | Online ISSN: 1539-3704
Conditions of Use