Optimizing Electroencephalographic Studies for Epilepsy Diagnosis in Children With New-Onset Seizures

Objectives: To establish whether early electroencephalography (EEG) or later sleep-deprived EEG (SD-EEG) has a higher yield of epileptiform and background abnormalities in children with new-onset seizures, and to use EEG results to assist in diagnosis of electroclinical epilepsy syndromes at presentation. Design: Prospective analysis blinded to EEG protocol and epilepsy diagnosis. Setting: Regional service capturing a pediatric population of 121 000. Patients: Consecutive untreated children aged 2 to 16 years presenting to emergency departments with new-onset seizures (excluding myoclonic and absence seizures). Intervention: Each child had 2 EEG protocols: an early EEG study (within 24 hours following a seizure) and an SD-EEG study (48 hours to 4 weeks following a seizure). Epilepsy diagnosis was made independently by 2 pediatric epileptologists. Main Outcome Measures: Rate of epileptiform abnormalities and slowing in the 2 EEG studies. The secondary outcome measure was diagnosis of epilepsy syndrome where possible. Results: Of 92 children studied, 50 (54%) had a single seizure; 42 (46%) had 2 or more seizures at presentation. Seizures were focal in 61 children (66%) and generalized in 19 (21%). Epileptiform discharges occurred in 56 SD-EEGs (61%) and 52 early EEGs (57%) (P=.27). Background slowing occurred in 26 SD-EEGs (28%) and 42 early EEGs (46%) (P<.001). Parents preferred early EEG (65 parents [71%]) to later SD-EEG (14 parents [15%]) because of availability of earlier results and epilepsy diagnosis. Forty-two of 92 children (46%) were diagnosed with a specific electroclinical syndrome. Conclusions: Early EEG and SD-EEG studies have a similar yield of epileptiform abnormalities. Background abnormalities are more frequent in early EEGs. The EEG results at presentation in new-onset seizures support epilepsy diagnosis, with electroclinical syndromes diagnosed in almost 50% of children.