Age of Red Cells for Transfusion and Outcomes in Critically Ill Adults
Packed red cells stored over time tend to develop a series of biochemical changes known as a "storage lesion". Clinicians have long been concerned that these storage lesions have an effect on outcomes for patients receiving transfusions. Although two previous RCTs have shown that this is not true, criticisms were levelled against each. Specifically, the INFORM trial had a low overall mortality potentially reducing its power while the ABLE trial evaluated only transfusions given soon after ICU admission. This multicenter RCT (TRANSFUSE) aimed to address these issues.
Nearly 5000 adults admitted to ICUs in 5 countries and receiving red cell transfusions were randomized to the freshest blood product available or the oldest blood product available. The primary outcome was 90-day mortality with a number of other secondary outcomes. Randomization was computerized and patients, caregivers, and research personnel were blinded to treatment allocation.
There was no overall difference in mortality between the two groups (24.1% vs 24.8%). A pre-specified subgroup analysis showed a higher mortality in patients with high APACHE III scores (higher risk of ICU death) receiving fresh red cells and a lower mortality in patients with low APACHE III scores receiving fresh red cells. Other subgroup analyses (blood group, SOFA score) and secondary outcomes had no difference between treatment allocations other than a slightly higher non-hemolytic febrile transfusion reaction rate in the fresh pRBC group.
Several exclusion criteria limit generalizability. One exclusion criteria was physician preference to give a particular blood product which may have subtly biased the results. No children were included and all these patients were critically ill.