Epilepsy in children is often evaluated and diagnosed with the assistance of neuroimaging. However, the yield of neuroimaging studies in children with newly diagnosed epilepsy is unknown. Berg and associates evaluated the yield of such tests in an observational, community-based study.

Children were recruited when they were first diagnosed with epilepsy. The type of seizure was determined after review of medical records and interviews with parents. Computed tomographic (CT) scans and magnetic resonance images (MRI) were reviewed if they had been obtained as part of the initial evaluation for epilepsy.

Of the 613 children included in the study, 488 had neuroimaging studies. About 32 percent of the children had CT scans and 63 percent had MRIs. Whether or not imaging was performed was strongly correlated with the type of epilepsy syndrome diagnosed. Only 49.2 percent of children with idiopathic generalized epilepsy had neuroimaging, but an average of 87.0 percent of children with other types of epilepsy had neuroimaging. In 12.7 percent of the images, an abnormality related to the cause of the epilepsy was found. The classification of the epilepsy syndrome changed in five children because of the neuroimaging study. Children with abnormal motor examinations were more likely to have imaging abnormalities, while those with soft neurologic signs were not likely to have imaging abnormalities. Status epilepticus, partial seizures, focal spikes and focal slowing on the electroencephalogram (EEG) were also more likely to be associated with imaging abnormalities. Of the 125 children who did not have imaging studies initially, 11 have since had studies. Eight of these were normal, one showed an old infarct, one showed a Chiari I malformation (not related to the seizure) and one showed mild hippocampal asymmetry.

The authors conclude that children with epilepsy and an abnormal neurologic examination are the most likely to benefit from neuroimaging. Patients whose studies are likely to be abnormal are those with partial seizures and focal findings on EEG. Some children with idiopathic epilepsy may not need imaging studies if a thorough history, physical examination and EEG were obtained initially. In children whose epilepsy is not clearly idiopathic, neuroimaging should be strongly considered, especially if there are focal findings.