Gastroenterology in critical care · Transfusion in gastrointestinal bleeding
Transfusion strategies for acute upper gastrointestinal bleeding
| Design | Randomised controlled trial |
|---|---|
| Treatment | Restrictive strategy, transfusion at haemoglobin below 7 g/dL |
| Control | Liberal strategy, transfusion at haemoglobin below 9 g/dL |
| Population | 921 patients with acute upper gastrointestinal bleeding |
| Follow-up | 45 days |
| Primary endpoint | Death from any cause within 45 days |
| Result | 225 patients assigned to the restrictive strategy (51%) received no transfusion versus 61 (14%) in the liberal group (p<0.001). Survival at 6 weeks was 95% versus 91% (hazard ratio for death 0.55, 95% CI 0.33 to 0.92, p=0.02); further bleeding 10% versus 16% (p=0.01); adverse events 40% versus 48% (p=0.02). The survival benefit was significant in Child-Pugh A or B cirrhosis (hazard ratio 0.30, 95% CI 0.11 to 0.85) but not in Child-Pugh C (1.04, 0.45 to 2.37) or peptic ulcer bleeding (0.70, 0.26 to 1.25). |
| Secondary endpoints | Rebleeding, adverse events, complications of transfusion |
As compared with a liberal transfusion strategy, a restrictive strategy significantly improved outcomes in patients with acute upper gastrointestinal bleeding.
The trial authors, in the published abstract
One of the few transfusion trials where restriction actively helps rather than merely proving non-inferior — plausibly because transfusion raises portal pressure and provokes rebleeding in variceal disease. The benefit did not extend to Child-Pugh C cirrhosis, so the threshold should not be applied mechanically to the most decompensated patients, nor to exsanguinating haemorrhage, which the trial excluded. Contrast the brain injury and myocardial infarction trials, where restriction is at best neutral.
Villanueva C, Colomo A, Bosch A, et al. Transfusion strategies for acute upper gastrointestinal bleeding. N Engl J Med 2013 Jan 3;368(1):11-21.