From the Journals

Blood donor age, sex do not affect recipient survival

View on the News

Current transfusion practice appears safe

The findings of Edgren et al. provide reassurance regarding the safety of current transfusion practice.

They present a convincing argument that differences in the statistical approach for controlling confounding likely explained the discrepant results of the Canadian study and their study.

This subtle confounding stems from the fact that increased transfusions expose the recipient to a greater total number of blood products, which in turn is associated with higher comorbidity, greater severity of illness, and higher mortality.

Nareg Roubinian, MD, is at the Blood Systems Research Institute, San Francisco, and in the division of research at Kaiser Permanente Northern California, Oakland. He and his associates reported having no relevant financial disclosures. They made these remarks in an invited commentary accompanying Dr. Edgren’s report (JAMA Intern. Med. 2017 April 24. doi: 10.1001/jamainternmed.2017.0914).


The age and sex of blood donors do not affect the recipient’s survival and do not need to be considered in blood allocation, according to a report published online April 24 in JAMA Internal Medicine.

A recent observational Canadian study suggested that blood from young donors and female donors increased the recipients’ risk of death – a finding which, if confirmed, would have immediate implications for medical practice.

A separate group of Scandinavian researchers attempted to replicate these findings by performing a retrospective cohort study using similar but more nuanced statistical methods. Gustaf Edgren, MD, PhD, of the department of medical epidemiology and biostatistics, Karolinska Institutet, Stockholm, and his associates analyzed information collected on 968,264 patients over a 10-year period from a Swedish and Danish transfusion database.

In initial, unadjusted analyses, both extremes of age (young and old) and female sex in the donor were associated with reduced survival in the recipient. However, that association disappeared when the data were adjusted to account for the total number of transfusions a patient received, a marker of their severity of illness. The hazard ratio per transfusion from a donor younger than age 20 was 0.98, and the hazard ratio per transfusion from a female donor was 0.99. This pattern also occurred in sensitivity analyses, the investigators noted (JAMA Intern. Med. 2017 April 24. doi: 10.1001/jamainternmed.2017.0890).

“When studying associations between ... transfusions with a particular characteristic and the risk of death in the recipient, [the] underlying disease severity ... may still confound the association. However, with meticulous adjustment for total number of transfusions, it should be possible to block the confounding effect of patient disease severity entirely,” they noted.

“We believe that, rather than reflecting true biologic effects, the Canadian results can be explained by residual confounding (i.e., that the observations resulted from incomplete adjustment for the number of transfusions),” Dr. Edgren and his associates said.

“In addition, we believe these data reinforce the importance of extreme caution in assessing epidemiologic analyses in this field, given the tremendous clinical and logistical implications of false-positive findings,” they added.

Next Article: