Researchers in pediatric neurology and electroencephalography at the University of Yamanashi, Japan, studied the EEG in children referred within 7-20 days after a febrile seizure (FS), and determined the utility of localization of paroxysmal discharges as a predictor for subsequent epilepsy. Of 119 patients with FS, 26 (21.8%) had paroxysmal abnormalities in the EEG and 9 (7.6%) patients developed epilepsy. Patients with complex FS (n=20, 17%) had a significantly higher risk of development of epilepsy than those with simple FS (n=99, 83%); p<0.05; they also had a higher incidence of abnormal EEGs (9/20, [45%], vs 17/99, [17%]; p<0.05). Patients with EEG abnormality had a significantly higher risk of development of epilepsy than those without EEG abnormality (p<0.01). Risk of epilepsy varied with the localization of paroxysmal discharge: 10% in children with generalized paroxysmal spike and wave activity; 28.5% with rolandic discharges; 75% with frontal paroxysms; and none with occipital paroxysmal discharges. Compared with generalized EEG foci, the relative risk for development of epilepsy in children presenting with frontal foci was 27.0, and significantly higher than those with paroxysms in other regions (p<0.035). Serial EEG is recommended in FS patients showing frontal paroxysmal EEG abnormalities. 
COMMENT. The electroencephalogram is not included in the American Academy of Pediatrics guidelines for the neurodiagnostic evaluation of the child with a first simple FS . In contrast, several reports have documented the value of the EEG in prediction of the development of recurrent afebrile seizures and epilepsy following a complex FS (see Pediatr Neurol Briefs August 2011;25(8):59). The risk of epilepsy following a simple FS is about 2%  whereas the risk following a complex FS in a child with developmental delay is 9.2% . In children with normal development and simple FS, the risk is only 1.1% but significantly greater than for healthy children with no FS (0.5%; p=0.027). In a prospective study of 428 children with a first FS followed for 2 years or more, unprovoked seizures occurred in 26 (6%). Risk factors for unprovoked seizures included neurodevelopmental abnormalities, complex FS, a family history of epilepsy, recurrent FS, and brief duration of fever before initial FS .
A review of the literature from 1947 to 1964 uncovered 36 publications in which the EEG findings were reported in relation to FS and 23 included number and percent with paroxysmal discharges (mean 25%) . In our prospective 2-year follow-up study of 76 FS patients with EEGs, paroxysmal discharges were recorded in 18 (24%); they were more frequent in patients who developed non-FS (61%) than in those with FS alone (12%). The discharges were generalized in 11 patients and focal in 7 . Age was a significant factor in relation to the incidence of abnormal EEGs; patients with abnormal records were 3-10 years (mean 7 years) old and those with normal records were 1-7 years (mean 3 years) old. In her extensive investigations of the EEG and FS, Lennox MA also reported that paroxysmal records occurred mainly in children 5 years or older . Repeated EEGs at follow-up subsequent to a FS were recommended to determine prognosis and risk of nonfebrile seizures. The present study concludes with a similar recommendation, but more specifically in patients with EEG showing frontal paroxysmal foci.