To the Editor We read with interest the Viewpoint by Huybrechts et al1 concerning the safety of medical therapies during pregnancy, commonly prescribed for off-label use. Because of obvious ethical concerns, pregnant individuals are excluded from randomized clinical trials looking at reproductive outcomes following drug exposures; thus, pregnancy exposure data must be obtained from indirect sources, mostly represented by cohort studies and pregnancy registries. However, there is a considerable lag between the time of drug approval and obtaining sufficient numbers of first-trimester exposures in the pregnancy registries to be able to rule out an increase in overall and rare birth defects. The unfortunate reality is that we learned the most about teratogenic effects only after a drug has been marketed. The well-known historical examples are of thalidomide, diethylstilboestrol, and more recently, dolutegravir.