Treatment of infants with epilepsy: Common practices around the world

High quality data to guide recommendations for infants with epilepsy are lacking. This study aimed to develop an understanding of common practice and regional variations in the treatment interventions of infants with epilepsy, and also to identify areas for further study and to highlight where common practice occurs without sound evidence.

A survey addressed clinical treatment practice for infants with epilepsy. Alternative interventions were included.

The survey found that most regions had similar practice for first-line interventions, except for North America, where more levetiracetam was prescribed. There was a preference for valproate as first-line therapy for generalized seizures, myoclonic seizures, and Dravet syndrome; only Oceania differed for generalized and myoclonic seizures. Phenobarbital was used for generalized and focal seizures in resource-poor and resource-equipped regions. Carbamazepine and oxcarbazepine were the preferred agents for focal seizures from all regions except North America, which uses more levetiracetam. For second- and third-line interventions, the range of choices was diverse, often with little correlation across regions. The ketogenic diet, vagus nerve stimulation, and epilepsy surgery were considered viable choices in most settings, but usually only once seizures were considered medically refractory. The survey highlighted the marked discrepancy in Africa, the one region that consistently confirmed a lack of access to these alternative interventions and to the newer antiepileptic drugs.

More randomized controlled trials in infants with seizures are needed to permit useful recommendations. The survey identified widespread use of levetiracetam in North America, which may be the result of effective marketing or based on good clinical practice. The widespread use of valproate may have safety implications. The lack of access to care in the African region highlighted the need for more sustained resources. Although the survey was not evidence based, the findings could be useful to support additional well-designed studies.