In a paper of the North American AED Pregnancy Registry 10.7% (16 / 149) of the pure prospective pregnancies exposed to valproic acid monotherapy ended in major malformations, in Sweden 13.1% (35 / 268) exposed were malformed, The UK Epilepsy and Pregnancy Registry found 6.2% (44 / 715) children with major malformations and in 2006 the Neurodevelopmental effects of AED (NEAD) study found 17.4% with a malformation if exposed to valproic acid monotherapy. Not all studies use the same definition for malformations, but in a very crude meta-analysis including these four studies the risk for (major) malformations due to valproic acid monotherapy is 8.9% [95% CI 7.43-10.65%].
Although there is evidence about the teratogenicity of valproic acid, still very few is known about the risks of specific malformations. In all studies found in literature, the authors made some critical remarks about the size of the study population and the scarcity of datasets that can be (combined and) used for more specific studies. Clearly there is a need for additional datasets capable for a case control design, as EUROCAT is, to estimate and compare the risks of all available AEDs individually that are used in pregnancy, and not only for major congenital malformations combined but especially for specific malformations.