To compare interview data on drug use during pregnancy with data identified from a register of prescriptions.
We compared information from the Swedish Prescribed Drug Register with the Swedish Medical Birth Register on antidepressant use. In order to evaluate the clinical significance of the difference in ascertainment with the two methods, the rate of preterm births among singletons and of neonatal symptoms were studied.
During the year before the last menstrual period, 1.5% of the women filled prescriptions for antidepressants each month. Already before the pregnancy was known, the rate of filled prescriptions decreased and reached 0.5% towards the end of the pregnancy. Twenty-two percent of first-trimester use of antidepressants was unidentified using interview data and prescriptions during the 2nd and 3rd pregnancy months covered only 55% of actual use. When women who filled prescriptions 1 or 3 months before the last menstrual period were included, 17 and 43% respectively of women were included who probably did not use the drugs in the first trimester. Prescriptions gave a more complete ascertainment of drug use after the first trimester than data from antenatal care, which seemed to overestimate drug use.
Interview data seem to give the most valid results on early use. When interview data are not available, prescription data could be used, but should not include prescriptions given earlier than 1 month before the last menstrual period. Studies of drug use later in pregnancy are best based on prescription data in the absence of interview data.