Studies in patients whose medication is discontinued Assessing cognitive and behavioral effects of AEDs by testing seizure-free patients before and after discontinuation of a monotherapy eliminates many confounding factors, in particular the possible contribution of on going seizure activity. A possible shortcoming of such a design is that it is not entirely prospective and may not unmask potential chronic cumulative cognitive or behavioral effects that are not or only slowly reversible. A group of 25 children with a history of febrile seizures who had received PB for an average of 35 months were tested before PB discontinuation over 1-2 months and again after a mean interval of 13 weeks (49). The control group consisted of 25 untreated matched children with a history of febrile seizures. The daily PB dose was 3-4 mg/kg, and 45% of the children had levels >15 mgL There were no differences between treated patients and controls either before of after discontinuation of PB. The same group of authors had previously described behavioral disturbances in children treated with PB for febrile seizures (50). Gallassi et al. used a discontinuation de- sign to study the effects of PB, CBZ, PHT, and VPA (51-53). A group of 16 young adults seizure free on either PB or CBZ monotherapy were tested before, during, and up to 1 year after discontinuation of the medi- cation (5 1). Before discontinuation, patients on CBZ per- formed better than those on PB on one measure only (spatial span). One year after discontinuation, patients discontinued from PB had improved on this same measure, patients discontinued from CBZ had a shorter “simple reaction time,” and there were no differences between the two groups. In a similar study of 10 adults discontinued from PHT monotherapy, patients performed worse than controls on three measures before discontinuation (52). No differences between patients