Premium
The healthy donor effect: a matter of selection bias and confounding
Author(s) -
Atsma Femke,
de Vegt Femmie
Publication year - 2011
Publication title -
transfusion
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.045
H-Index - 132
eISSN - 1537-2995
pISSN - 0041-1132
DOI - 10.1111/j.1537-2995.2011.03270.x
Subject(s) - biostatistics , selection bias , confounding , library science , family medicine , citation , medicine , selection (genetic algorithm) , epidemiology , computer science , artificial intelligence , pathology
T he healthy donor effect (HDE) is an important methodologic pitfall in health research among blood donors. The HDE is a term applied to the phenomenon of decreased morbidity rates and a healthier lifestyle in donors when compared to the general population, but also in the comparison between donors and nondonors and between active donors and lapsed donors. As a result, effect estimations obtained from studies using such comparisons may be biased. A major topic in donor health research is the protective effect of blood donation against cardiovascular disease. This presumed beneficial effect of blood donation has often been investigated by comparing donors with nondonors. Contrasting results were found; some studies did not find a significant effect, while others found a decreased cardiovascular risk in blood donors and suggested that blood donation may yield a protective cardiovascular effect. However, it is very likely that this effect was to a great deal overestimated. The beneficial cardiovascular profile was probably mainly caused by the inclusion of healthy and health conscious blood donors during the blood donation process, i.e., the HDE. Donor health studies are very important in learning more about positive and negative effects of blood donation. To design and perform this type of research in a valid way, a thorough understanding of methodologic features of HDE is imperative. To what type of bias does the HDE belong? And how does it influence study results? A proper knowledge of the methodology of the HDE is of vital importance for a correct treatment of the bias. In this report, we present a brief overview of different biases in epidemiologic research in relation to the HDE.