ABCDXXX: The obscenity of postmarketing surveillance for teratogenic effects.

Our current system of postmarketing surveillance, which is based on voluntary reporting of suspected teratogenic effects, is a failure. Postmarketing surveillance should, at a minimum, provide reassurance that every approved drug treatment does not produce a teratogenic effect as great as thalidomide embryopathy or fetal alcohol syndrome. This means that postmarketing surveillance should be able to detect a twofold or greater increase in the frequency of major congenital anomalies, a fivefold or greater increase in the frequency of intellectual disability, or a characteristic pattern of minor anomalies and functional abnormalities that occurs with a frequency of at least 10% among the children of women who were treated with the drug during pregnancy. Effective surveillance for teratogenic effects could be accomplished through a complementary set of mechanisms that includes pregnancy exposure registries or cohorts as well as direct examination of a small subset of infants whose mothers received the treatment during various periods of pregnancy. If this routine surveillance reveals a "signal" (i.e., an indication suggesting a possible teratogenic effect), further study would be needed to establish whether the observed effect is real and causal. Once a signal of possible teratogenicity in humans has been recognized, validating or refuting it would become an urgent matter.