RESUMO
OBJECTIVE: The goals of this study were to explore the diverse criteria surrounding indications for antiepileptic therapy and to establish a consensus on drug selection for initial monotherapy in adult patients with epilepsy. METHODS: The study was performed using the modified Delphi method, which aims to achieve professional consensus by means of a series of questionnaires. Three different groups of items were evaluated: the beginning of antiepileptic treatment, the drug selected for initial monotherapy with respect to the type of epilepsy, and the drug selected for initial monotherapy with respect to comorbidity. RESULTS: Sixty experts completed two rounds of a questionnaire. In the first round, consensus was reached on 135 of the 194 questions analyzed. After the second round, consensus was reached on 148 items. The main findings of the survey revealed a consensus on beginning treatment after the first seizure when the EEG showed abnormalities such as generalized spike-wave discharges, when MRI demonstrated an epileptogenic brain lesion, and in elderly patients. Regarding to the antiepileptic drug selected for initial monotherapy with respect to type of epilepsy, levetiracetam and lamotrigine were recommended for generalized tonic-clonic seizures regardless of sex or age; levetiracetam was recommended for myoclonic epilepsy regardless of sex; valproic acid, ethosuximide, levetiracetam, and lamotrigine were chosen for absence epilepsy; and carbamazepine, levetiracetam, lamotrigine, and oxcarbazepine were recommended for partial epilepsy regardless of age or sex. Finally, in the evaluation of drug selection with respect to comorbidity, first-generation drugs were less recommended than second-generation drugs, which were clearly preferable. The drugs on which there was a greater consensus were levetiracetam, lamotrigine, valproic acid, and topiramate. CONCLUSIONS: There is a tendency to begin treatment after the first seizure, depending on the results of additional testing. In general, first-generation drugs are less recommended for different types of epilepsy, especially in the presence of a comorbid condition. However, the authors are conveying perceptions and opinions, the effect of which on treatment outcomes has not been evaluated.