Several publications have reported an increased risk of fetal malformation in the offspring of mothers who during pregnancy took more than one antiepileptic drug (AED) simultaneously as compared to those taking AED monotherapy (Tanganelli & Regesta, 1992;Olafsson et al., 1998;Arpino et al., 2000;Lowe, 2001;Hart, 2003;Morrow & Craig, 2003;Pack, 2006). A similar increased risk in patients taking polytherapy was not found in an earlier analysis of data from the Australian Register of AEDs in Pregnancy (Vajda et al., 2003). Here we have specifically reexamined the issue of the relative teratogenic risk of AED polytherapy in the larger cohort of women now enrolled on the register, with particular attention to the nature and dose of the drugs taken.
Materials and Methods
Analysis of data from the Australian Pregnancy RegisterThe Australian Register of Antiepileptic Drugs in Pregnancy has been collecting data concerning the use of AEDs in pregnant women since 1999. Recruitment is nationwide and voluntary, and contact with the register is by telephone. Potentially eligible women are informed about the register by a variety of methods, but most commonly by their treating medical practitioners, nurses, or allied health professionals, or by other pregnant women. Relevant details are obtained from pregnant women on recruitment (usually in the first or second trimester), at 7 months of pregnancy, in the first postnatal month, and at the end of the first postnatal year. Treating doctors are contacted to confirm medical details. Fetal malformations are classified according to the Birth Defects Registry of Victoria (Riley & Haliday, 1988). The register database was housed initially at St Vincent's Hospital, Melbourne, and subsequently at Monash University, Melbourne, and has been under the consecutive ethical oversight of the ethics committees of these institutions.For the purposes of this analysis the incidence of the malformations was determined on the basis of all information available at the end of the first postnatal month and the first postnatal year. The 1-year data were used for all analyses except for comparisons with the international literature, for which the 1-month data were also used.The data relevant to the present article were extracted from the register's Microsoft Access database into an Excel spreadsheet and then analyzed by confidence interval (CI)