Customize your JAMA Network experience by selecting one or more topics from the list below.
Blood for transfusion is a drug or, more specifically, a biological medicine.1 The US Food and Drug Administration regulates blood for purity, potency, and safety under the Food, Drug, and Cosmetic Act of 1938. Regulatory authorities in most developed countries prescribe blood collection, storage, testing, labeling, and intended use. Facilities for blood component “manufacture” are licensed and inspected for compliance with current good manufacturing practices. Where blood differs from small-molecule pharmaceuticals and recombinant biologics is in its unit-to-unit heterogeneity (batch variability). Although blood donors are qualified as suppliers of “raw material” by rigorous selection and testing standards, each donor is biologically distinct. Therefore, every unit of blood is considered its own batch. Despite the acknowledged batch variability, unit-to-unit differences are not expected to be so significant as to affect transfusion outcomes. However, the association of donor characteristics with transfusion outcomes has not been carefully investigated.
In this issue of JAMA Internal Medicine, Chassé and colleagues2 report the association of blood donor age and sex with recipient survival after red blood cell (RBC) transfusion. Their longitudinal cohort study links blood donor data routinely collected by the Canadian Blood Services to RBC recipient clinical information gathered from databases containing the short-term and long-term patient outcomes from 4 academic hospitals. The investigators predetermined primary and secondary exposures (donor age and donor sex) and the primary outcome (recipient survival). The study period spanned 7 years and included 80 755 blood donors, 30 503 patients, and 187 960 RBC units. The investigators constructed a robust model supported by generally accepted statistical principles. Yet, even with rigorous analyses and sensitivity testing, their study remains observational and subject to the risks of bias and confounding. The findings should be considered hypotheses, not evidence of causality.
The observation that donor demographic factors might be associated with transfusion outcome is not new. More than 50 years ago, the risk of transfusion-transmitted hepatitis was correlated with donor sex, age, socioeconomic status, and ethnicity.3 All of these associations were eventually found to be related to the donor’s risk of infection with hepatitis B or C viruses. However, the findings by Chassé and colleagues,2 particularly the primary exposure that transfused RBCs from younger donors are associated with an increased mortality risk of 6% to 8%, with the greatest risk in male recipients, are hardly expected or intuitive. Why should old men be safer donors but more vulnerable recipients? The only other published study4 of RBC donor age and recipient mortality, a large, matched cohort investigation from the Danish and Swedish Scandinavian Donations and Transfusions (SCANDAT) 2 database, found no association between donor age and patient 30-day or 1-year mortality after RBC or plasma transfusion. The studies are not precisely comparable. Patients in the SCANDAT2 analysis had a lower RBC exposure. Similarly, in a single-center US study,5 no association was found between donor age and recipient outcome after plasma transfusion in a large database-linked study of coronary artery bypass surgery. However, the absence of an obvious mechanism or of confirmatory studies does not refute the findings by Chassé and colleagues2 but indicates that verification is necessary before such findings can be acted on.
With the secondary exposure, Chassé and colleagues2 found a strong association between female RBC donors and recipient mortality. The transfusion of each additional RBC unit from a female donor compared with a male donor was associated with an increased risk of death of 8%. Numerous biological differences distinguish blood from female and male donors, from lower hemoglobin levels and different protein concentrations to the frequency of HLA, platelet, and granulocyte alloantibodies. None of these factors are used in routine blood donor screening to determine the safety of RBC transfusion. However, plasma from female donors is routinely avoided, as are single-donor platelet concentrates from multiparous women who have not been screened for HLA alloantibodies, because of their association with transfusion-related acute lung injury (TRALI). Still, the risk of TRALI, especially from blood components that contain minimal amounts of plasma, cannot explain an 8% mortality difference. Other genetic factors appear to be involved in the documented differences in donor RBC in vivo survival and ex vivo storage, although neither of these examples of batch variability has been associated with donor sex,6,7 nor have such differences been shown to affect survival of blood component recipients.
Several features of the study by Chassé and colleagues2 deserve emphasis. Of course, the specific conclusions are tantalizing and clearly merit further exploration. More generally, this study highlights the need to further investigate which demographic and genetic donor factors aside from blood groups might influence the safety and efficacy of the different blood components. Such research has been overshadowed by the understandable focus on pathogen transmission. Perhaps the most significant message may be recognition, as the authors point out, that to perform such studies one has to be able to track blood from “vein to vein.” The most efficient way to do this on a large scale is by exploiting existing databases, the best example of which is the previously cited SCANDAT database, a collaboration between Sweden and Denmark.4
In the United States, a variation of this approach has been undertaken. The third phase of the Recipient Epidemiology and Donor Evaluation Study (REDS-III) is a 7-year multicenter transfusion safety research initiative launched in 2011 by the National Heart, Lung, and Blood Institute.8 REDS-III brings together 4 domestic hubs that consist of geographically disperse blood center–transfusion service consortia that service mixed hospital bases, including 12 tertiary care and community hospitals. REDS-III has created a detailed multisite research database that links data from blood donors and their donations, the components made from these donations, and data extracts from the electronic medical records of the recipients of these components in the participating hospitals. REDS-III contains a wealth of data, much of it early in analysis. Three years (2013-2015) of recipient data with an estimated 8 billion data points will be available once the database is finalized. Clinical and laboratory data on 340 000 patient encounters are now accessible. Examples of planned REDS-III analyses include evaluating the effect of donor demographics on adult recipients’ clinical outcomes to further address the questions raised by Chassé and colleagues,2 as well as the effect of donor and component characteristics on the likelihood of recipient alloantibody formation, in addition to the donor and recipient risk factors associated with transfusion reactions. REDS-III should serve as a model for more extensive linked donor-recipient research in the United States and internationally. Such vein-to-vein databases could be expanded in future years to accommodate temporal analyses, as well as evaluations of donor and blood component factors in neglected vulnerable populations, including neonates and children. The expanded databases would serve as invaluable resources to unaffiliated clinical trialists, as well as public health officials interested in evaluating and improving the safety and effectiveness of transfusion therapies. Combining clinical trials research with database analysis should provide a powerful tool for tackling fundamental questions regarding blood transfusion, providing rapid responses to emerging transfusion issues and informing blood policy decisions that affect patient care.
Corresponding Author: Harvey G. Klein, MD, Department of Transfusion Medicine, Clinical Center, National Institutes of Health, Bldg 10, Room 1C-711, Bethesda, MD 20892 (email@example.com).
Published Online: July 11, 2016. doi:10.1001/jamainternmed.2016.3355.
Conflict of Interest Disclosures: None reported.
Disclaimer: The opinions expressed are those of the author and do not represent the positions of the National Institutes of Health, the Department of Health and Human Services, or the US Government.
Klein HG. Blood Donor Demographics and Transfusion Recipient Survival—No Country for Old Men? JAMA Intern Med. 2016;176(9):1315–1316. doi:10.1001/jamainternmed.2016.3355
Create a personal account or sign in to: