Objective — To study the validity and accuracy of an adjusted questionnaire on medical drug use during pregnancy eight years after the pregnancy.
Methods — The ability of a questionnaire on medication during pregnancy to detect actual use (= sensitivity) was tested against information collected 8 years previously (in 1983–1984) from 473 women with high-risk pregnancies who delivered at the University Hospital Nijmegen, the Netherlands.
Results — For separate drug groups, the sensitivity varied between 5% and 91%. The timing of use was recalled moderately well. Although specific questions on drug groups did improve the sensitivity as compared to an earlier questionnaire, the improvement was not enough to make the questionnaire valid. High maternal education, low birth weight, low gestational age and a low 5-min Apgar score were related to better recall. The sensitivity of the questionnaire depended on the behavioural score of the child, implying recall bias.
Conclusion — Questionnaire data on drug use during pregnancy obtained eight years after delivery are not a valid source of information.