Accumulating evidence from clinical trials suggests that administering blood transfusions once the hemoglobin falls below either 7 or 8 g/dL—a restrictive transfusion strategy—is safe in most clinical settings.(1, 2, 3, 4) However, is it safe in every one? Specifically, is a low hemoglobin concentration safe in patients with severe vascular disease and acute coronary syndrome, because of the heart’s dependence on oxygen delivery?
Safely limiting exposure to blood is important because transfusions are common, with more than 16 million red blood cell units transfused annually to 3.4 million people in the US5 and 85 million units worldwide.(6) Transfusions also are expensive, and may be associated with both infectious and noninfectious complications. It is therefore important to administer blood to patients where benefits outweigh risks.
As documented in a systematic review in this issue of the Journal, we now have a number of clinical trials to help address at what point to administer a blood transfusion. Indeed, most of the modern trials have used either a 7-g/dL or 8-g/dL threshold in the restrictive transfusion strategy. Clinicians might ask if the lower 7-g/dL threshold is superior to the higher threshold of 8 g/dL.
Salpeter et al(7) combined the results of the 3 published trials (in 2364 patients) that used a 7-g/dL hemoglobin threshold as a restrictive transfusion strategy. Two trials were in adults(1) and pediatric intensive care patients,8 and one trial involved patients with acute gastrointestinal bleeding.4 When the 3 trials were combined, the mortality was 20% lower in the 7-g/dL restrictive transfusion group compared with the 9–10-g/dL liberal transfusion group (risk ratio 0.80; 95% confidence interval, 0.65-0.98). Hospital mortality, pulmonary edema, and acute coronary syndrome also were significantly less frequent in the 7-g/dL transfusion group. In contrast, the outcomes were not improved in trials using a less restrictive transfusion threshold.
There are at least 2 possible explanations for the finding that a 7-g/dL threshold is superior to an 8-g/dL threshold. If we believe that red blood cell transfusion is harmful, administration of less of a harmful treatment should result in better outcomes. Alternatively, it is possible that the different populations of patients enrolled in these trials explain the observed differences in outcomes in 7-g/dL trials and 8-g/dL trials.
To read this article in its entirety, please visit our website.
– Jeffrey L. Carson, MD, Paul C. Hebert, MD
This article originally appeared in the February 2014 issue of The American Journal of Medicine.
