On the use of affected controls to address recall bias in case-control studies of birth defects.
Inferences regarding causes of birth defects in humans are often based on results of case-control studies conducted after birth.
To address bias in these studies caused by potential differential recall of past exposures between case and control mothers, many investigatars have advocated the use of affected controls (babies with birth defects other than the one of interest).
To evaluate whether the use of affected controls is warranted for a wide range of scenarios, we analyzed data from a population-based case-control study of birth defects in Atlanta, in which there were 4,918 babies with serious defects ascertained in the first year of life and 3,029 babies without defects.
Mots-clés Pascal : Malformation, Foetus, Foetus pathologie, Nouveau né pathologie, Nouveau né, Homme, Epidémiologie, Méthodologie, Biais
Mots-clés Pascal anglais : Malformation, Fetus, Fetal diseases, Newborn diseases, Newborn, Human, Epidemiology, Methodology, Bias
Notice produite par :
Inist-CNRS - Institut de l'Information Scientifique et Technique
Cote : 94-0445281
Code Inist : 002B20F02. Création : 199406.