Objectives & Background Transfusion thresholds for upper gastrointestinal bleeding (UGIB) are controversial. Observational studies suggest associations between liberal red blood cell (RBC) transfusion and adverse outcome. A recent trial reported increased mortality following liberal transfusion. We delivered a cluster randomised trial to evaluate the feasibility and safety of implementing a restrictive (transfusion when haemoglobin (Hb) <8 g dL) vs liberal (transfusion when Hb <10 g/dL) RBC transfusion policy for UGIB.

Methods Hospitals were randomised to a policy which was implemented through a multi-faceted educational intervention. Patients were recruited in Emergency Departments and acute admission units. All adult patients were eligible to participate, regardless of co-morbidity; the only exclusion criterion was exsanguinating haemorrhage. Feasibility and exploratory clinical outcomes were recorded up to day 28.

Results 936 patients were enrolled in six hospitals. Rockall and Blatchford risk scores were identical between policies. Protocol adherence was 96% in the restrictive policy vs 83% in the liberal policy (difference 14%, 95% CI 7 to 21%). In patients with Hb <120 g/L, Hb at discharge was lower for the restrictive policy (difference −0.7; 95% CI −1.4 to 0.0; p=0.05). For the restrictive policy fewer patients received RBCs (difference −13%, 95% CI −35 to 11%) with on average 0.8 (−1.9 to 0.3) fewer RBC units transfused. Clinical outcomes were better in the restrictive policy: 28-day further bleeding, 5% restrictive vs 9% liberal (difference −3.7%, 95% CI -12.2 to 4.8%); 28-day mortality, 5% restrictive vs 7% liberal (difference −1.3%, 95% CI −8.0 to 5.5%).; serious adverse events, 18% restrictive vs 22% liberal (difference −4.9%, 95% CI −22.6 to 12.8%). In the subgroup with IHD, there was a large observed difference for mortality (12% restrictive arm (n=6) vs. 3% liberal arm (n=2); interaction P=0.11).

Conclusion Adherence to both policies was high, resulting in a reduction in RBC transfusion and separation in the degree of anaemia and RBC exposure. There was a trend towards improved safety in the restrictive policy. We have demonstrated that a large-scale cluster randomised trial is feasible and is now warranted to determine the effectiveness of implementing restrictive RBC transfusion for all patients with AUGIB.