Febrile seizures: is the EEG a useful predictor of recurrences?

We examined the predictive value of a paroxysmal EEG in children with febrile seizures seen at the University Pediatric Clinic, Skopje, Macedonia, between 1982 and 1984. This was the only facility providing EEG or neurologic consultation for children in Macedonia, and almost all children in the area who experienced a febrile seizure were referred to this facility. EEGs were classified as epileptiform if they contained spikes and sharp waves or spike wave complexes, which were either focal or generalized, and were considered abnormal for age and state. Nonspecifically abnormal was defined as focal or generalized slowing excessive for age and state. Follow-up visits were scheduled at 6-month intervals; mean follow-up time was approximately 23 months. In order to determine whether clearly abnormal EEG features would predict recurrences, we compared the recurrences in 170 children with initial normal-appearing EEGs with 99 children with initial paroxysmal EEGs. There was no significant difference in risk of recurrence of febrile seizures between the two groups; increase in recurrence risk was determined primarily by younger age. The EEG did not add information regarding the likelihood of recurrence of febrile seizures.

Febrile seizures: clinical characteristics and initial EEG.

We examined the relationship between clinical characteristics and EEG classification in all children with febrile seizures examined at the University Pediatric Clinic, Skopje, Yugoslavia between 1982 and 1984. This is the only facility in Macedonia providing EEG or neurologic consultation for children. EEGs were classified as paroxysmally abnormal if they contained spikes, sharp waves, or spike-wave complexes considered abnormal for age. In all, 22% of the 676 children had an abnormal initial EEG. The most common basis for classification as abnormal was spike-wave complexes greater than 3 Hz; the next most common basis was the presence of spikes. Birth weight, gender, accompanying illness, and family history of seizures, and whether the index seizure was single or multiple were not associated with differences in rate of abnormal EEG. Clinically focal index seizures and longer duration were associated with EEG abnormality. Number of previous febrile seizures was associated with an increasing rate of EEG abnormality, from 18% in children with no previous seizures to 63% in those with four or more previous seizures. Age at EEG was linearly related to likelihood of paroxysmal EEG abnormality, both for the total cohort and for the 376 children with no previous seizures. In the total cohort, logistic regression identified leading predictors of abnormal initial EEG to be older age, number of previous febrile seizures, preexisting motor abnormality, and focal seizures. For children with a first febrile seizure, leading predictors were focal seizure, older age, and preexisting motor abnormality.

Febrile convulsions and later development of epilepsy.

A group of 172 epileptic children who had had prior febrile convulsions was compared with a group of 674 who had not. Children with epilepsy and prior febrile convulsions were similar in some respects (sex ratio, positive family history for seizures) to children with pure febrile convulsions and in most respects (type of epilepsy, mental status, initial EEG, and two- and four-year remission rates in the long-term outcome) to epileptic children without prior febrile convulsions. Our data do not support the current view that febrile convulsions, per se, are the main cause of mesial temporal sclerosis, le, temporal lobe epilepsy. Thus, our clinical findings support previously expressed doubts on the role of febrile seizures in temporal lobe epilepsy that were based on pathohistologic findings.