The severity and nature of side effects and doses and plasma levels of phenobarbital (PB), primidone (PRM), phenytoin (PHT), carbamazepine (CBZ), and valproate (VPA) were examined in 392 pediatric outpatients at the Division of Paediatric Neurology, National Hospital “Marques de Valdecilla”, University of Cantabria, Santander, Spain. Side effects occurred in 50%, necessitating changes in medication in 18% and withdrawal of drug in 7%. The incidence of side effects was highest with PHT (71%) and lowest with PRM (29%). Serious side effects requiring drug withdrawal occurred with PHT (10%), VPA (8%), and PRM (8%), and less frequently with PB (4%) and CBZ (3%). The best tolerated drug was CBZ, and the least tolerated was PHT. Behavioral side effects were noted most commonly with PB (60%), neurological abnormalities such as ataxia and nystagmus with PHT (22%), digestive tract disorders with VPA (28%), and gingival hyperplasia and hirsutism with PHT (58%). The side effects that most often necessitated changes in treatment were behavioral disorders, especially excitement and hyperactivity, with PB and PRM, hirsutism and gingival hyperplasia with PHT, restless sleep and vomiting with CBZ, and digestive disorders with VPA. Behavioral disorders produced by PB and PRM disappeared in half of the patients if the dosage of the drug was increased. 
COMMENT. A collaborative group for epidemiology of epilepsy reports a 42% incidence of adverse reactions to antiepileptic drugs in 355 patients followed for an average of 11 months in 15 university and hospital centers in Italy . Clinical judgment provided the most valid basis for the evaluation of drug toxicity, and “toxic” plasma drug levels were not correlated with adverse reactions. Plasma levels were within normal limits in 78% of cases with and in 81% of cases without adverse drug reactions. Of 31 patients with “abnormal” plasma levels, only 1 had an adverse drug reaction. The authors stress the importance of reporting adverse drug reactions as a means of improving the quality of care for the epileptic in routine clinical practice. Physicians are sometimes reluctant to get involved, fearing legal repercussions or lengthy and tedious questionnaires from drug companies. Perhaps a system permitting anonymity might encourage more active physician participation. Additional, more sensitive, methods of monitoring drug treatment should expand the concept of “intolerable” side effects to include subtle psychological and behavioral effects.