Results from a trial by Hébert et al17 changed the practice of transfusion. This study was a randomized controlled trial that compared a “liberal” transfusion strategy (defined as a post-transfusion goal of a hemoglobin concentration between 10 and 12 g/dL with a transfusion indication of a hemoglobin concentration of 10 g/dL) with a “restrictive” transfusion strategy (defined as a post-transfusion hemoglobin concentration between 7 and 9 g/dL with a transfusion indication of a hemoglobin concentration of 7 g/dL). The trial included 838 volunteers and demonstrated an overall in-hospital mortality rate significantly lower in the participants who received the restrictive transfusion. The 30-day mortality rates were not significantly different between 2 groups.17 However, clinically “less ill” volunteers (Acute Physiology and Chronic Health Evaluation score < 20) or those who were younger (<55 years of age) had significantly lower 30-day mortality rates than those who were “more severely ill” or older following the restrictive transfusion strategy.17 Thus, the study results demonstrated that a restrictive transfusion strategy is at least equivalent to a liberal transfusion strategy in all groups, except among those with severe ischemic heart disease, and the restrictive transfusion strategy was potentially better in “less ill” and younger people.17 This was the first well-structured clinical trial of RBC transfusion suggesting that maintaining patients at lower hemoglobin levels might be beneficial.17
Vincent et al18 evaluated the 28-day mortality rate of 3,534 patients from 146 western European intensive care units (ICUs). The mortality rates among study volunteers were 22.7% and 17.1% among those receiving transfusions and those not receiving transfusions, respectively.18 The study controlled for patients with a similar degree of organ dysfunction. The receipt of an RBC transfusion in the ICU increased a patient’s odds of dying by a factor of 1.37.18
Corwin et al19 analyzed anemia and blood transfusions among 4,892 study volunteers who were critically ill in US ICUs. The study results showed that the number of RBC transfusions was an independent predictor of longer ICU stay, longer length of hospital stay, and increased mortality rates.
In another trial, Carson et al20 compared the effect of a transfusion threshold of 10 g/dL with 8 g/dL in cardiovascular patients undergoing surgical hip fracture repair. The trial involved 2,016 patients older than 50 years of age, and the primary outcome was death or the inability to walk across a room without human assistance on 60-day follow-up. Rates of death or an inability to independently walk after 60 days, in-hospital morbidity rates, and in-hospital complications were similar in the 2 groups. Thus, the liberal transfusion strategy did provide clinical benefit over the restrictive (9 g/dL) strategy.
Evidence does not support a benefit to a post-transfusion hemoglobin concentration above 10 g/dL.17,21 However, this level may be helpful in pediatric patients with cancer who have acute blood loss or cyanotic heart disease due to the additional challenges of this patient population.22
Despite well-performed clinical trials, no universal RBC transfusion criterion exists. A restrictive transfusion strategy is at least equivalent to a liberal transfusion strategy in the majority of clinical scenarios. However, in clinical practice, the underlying condition of the patient and his or her transfusion goals and desired outcomes should be considered. RBC transfusion may be indicated in a patient with symptoms of anemia and a hemoglobin level below 7 g/dL. Transfusion with hemoglobin concentrations between 7 and 10 g/dL may be indicated when significant underlying comorbidities exist, such as cardiac disease, respiratory disease, bone marrow failure, or other hematological diseases; this is because anemia may not be well tolerated in these patients.23,24 Traditionally, single unit transfusions were not recommended; however, as the hemoglobin indication has decreased and transfusion has become more conservative, it has become clear that the transfusion of 1 unit can be effective and sufficient. Single unit vs 2-unit transfusions can reduce blood use as much as 25% with no adverse clinical consequences.25 Specifically, the AABB (formerly American Association of Blood Banks) recommends transfusion at a hemoglobin concentration of 7 to 8 g/dL for hospitalized patients who are stable, 8 g/dL for those with cardiovascular disease, and higher hemoglobin (unspecified) concentrations for patients with acute coronary syndromes.26
Reactions or adverse events due to RBC transfusion are uncommon and may occur in 1% to 3% of transfusions.27 The most common adverse event is febrile nonhemolytic transfusion reaction, which typically is due to human leukocyte antigen (HLA) antibodies in the recipient or an allergic reaction to plasma proteins. The most severe yet rare reaction is acute hemolysis, usually due to ABO incompatibility due to administration error. Because the changes that occur during RBC storage have become better understood, concern has developed as to whether RBCs nearing the end of the routine 42-day storage might have undergone changes, thus making them risky for certain patients. The focus of research has been on patients with cardiovascular disease or those undergoing surgery; presently, however, no data suggest this is a concern for patients with malignancy.28 For more information, please refer to the article by Drs. Qu and Triulzi in this issue.