Investigators from Aristotle University of Thessaloniki, Greece, identified prognostic factors for subsequent epilepsy in a prospective study of 501 children with a first febrile seizure (FS). Patients were followed for a median of 5.3 years. Age at onset of FS ranged between 3 months and 5 years old (median age at onset 23 months +/- 12 months). Viral infection was the cause of the fever in 88.6% cases; bacterial infections were associated in 10.4%, and immunizations in 1%. EEG at the first episode of FS was pathologic in 105 (21%) of 501 children; 58 had generalized S/W complexes, 32 had focal epileptiform discharges (mostly centrotemporal and occipital); and 15 showed nonepileptiform abnormalities; epilepsy occurred in 4 patients with abnormal EEGs.
Epilepsy occurred in 27 (5.4%) of 501 FS patients. A median of 30 months (range 6 – 42 months) elapsed between the time of first FS and occurrence of epilepsy. Of 221 children with a first FS recurrence, 23 (10%) developed epilepsy; of 106 with a second FS recurrence, 19 (18%) had later epilepsy. Significant prognostic markers for subsequent epilepsy were as follows: 1) positive family history of epilepsy, especially maternal, 2) complex FS, 3) focal FS, 4) Todd paresis, 5) short duration of fever before FS, 6) late age at onset of FS (>3 years), and 7) multiple FSs (4 or more). Multiple FS increased the risk of epilepsy 10 times, positive family history of epilepsy 7.3 times, age at onset of FS after third year of life 3.8 times, and complex FS 3.6 times. Focality at the first and second FS recurrence increased the risk of epilepsy about 9.7 and 11.7 times, respectively. From the third FS recurrence and beyond, only focality of FS continued to have prognostic value. 
COMMENTARY. A review of the World literature concerning febrile convulsions found 37 reports dating from 1929 to 1964, totalling 5,576 patients with febrile convulsions; a mean of 20% had frequent recurrence of afebrile seizures or epilepsy (range, 2.6% - 100%), and a mean of 27% had electrographic seizure discharges (range, 3% - 86%) . The wide variations in incidence of epilepsy are explained by differences in diagnostic criteria and selection of patients. Those with short febrile seizures were mainly in prospective studies and rarely developed spontaneous seizures, whereas patients with complicated febrile seizures had often presented at a later age with epilepsy, and 17% had records of brain injury at birth . In a prospective study of 110 unselected FS patients followed for ∼2 years by the author, spontaneous nonfebrile seizures occurred in 17%, they were recurrent in 12%, and were frequent in 4% . The duration of the FS, a factor not analyzed specifically in the Pavlidou study, and abnormal EEG were most predictive of subsequent epilepsy.
In 1706 children in the US who had experienced at least one FS and were followed to the age of 7 years, epilepsy developed in 2%. In children whose first seizure was complex febrile, epilepsy developed at a rate 18 times higher than in children with no febrile seizures. In patients with simple FS, epilepsy developed in 11 per 1000 (1.1%) .
In a UK national population based study of 14,676 children, 398 (2.7%) had at least one FS. The first FS was simple in 80% and complex in 20%. Epilepsy developed in 9 (2.3%), at a rate similar to that in the US . The early recognition of a heightened susceptibility to epilepsy in a child with FS should lead to the introduction of seizure precautions and EEG surveillance at intervals.