Validation studies of drug exposures in pregnant women

The aim of these studies was to check the quality of data collection concerning drug use during pregnancy in the Hungarian Case–Control Surveillance of Congenital Abnormalities from four different aspects. (a) The parents of non‐respondent cases were visited at home to collect data, but the Ethics Committee did not allow collection of comparable data from non‐respondent control women. Thus only 200 selected non‐respondent controls were visited at home and questioned about drug use as part of a validation study. There was no difference in the distribution and occurrence of frequently used drugs between the respondent and non‐respondent control groups. (b) Data collection was compared between the usual method of the Case–Control Surveillance and personal interview in cases with congenital limb deficiencies. In general, the two methods produced similar results with respect to drug use, but the personal interview revealed a higher use of unintended contraceptive pills while the use of antenatal care logbooks in the Case–Control Surveillance showed a higher use of drugs to prevent threatened abortion and preterm delivery. (c) A check of source and completeness of drug exposure showed that about one‐quarter of the drugs used during pregnancy were not recorded in the Case–Control Surveillance, and the proportion of lack of recall was about 23%. A small group of women (2.4%) did not use prescribed drugs due to the suspected teratogenic risk. (d) The teratogenic labelling of drugs deters pregnant patients from the use of prescribed drugs, or at least shortens the duration of drug intake. Thus the evaluation of drug use on the basis of medically recorded prescribed drugs and/or usual recommendations produces a bias thus it is necessary to complete it with information from the patients. Copyright © 2003 John Wiley & Sons, Ltd.