We thank Einarson 1 for her commentary on our study on the risk of spontaneous abortion associated with antidepressant use. 2 Although it is true that we cannot assess causality based on one observational study owing to the inherent potential biases, repetition of findings, which is a strong tenet of epidemiology, leads to cause. At least one other study, by Einarson and colleagues, 3 has shown an almost identical increase in risk. We disagree that case–control studies, as opposed to cohort studies, have a confusing array of statistics. Both designs, in population-based cohorts, use similar statistics. The cohort design suggested by Einarson is not efficient, and perinatal pharmacoepidemiologists would agree that the case–control approach is the best design for studying adverse perinatal events. We also disagree with Einarson’s comment that our adjustment for underlying indication was impossible. We included a group of women with (a) depression who were not using antidepressants, (b) depression who were using antidepressants and (c) no depression. With rates of depression during gestation increasing, it is imperative that we fully understand the full spectrum of risks and benefits of antidepressant use for mothers and fetuses.
Footnotes
-
For the full letter, go to: www.cmaj.ca/cgi/eletters/cmaj.100507v1