Electroencephalograms, clinical observations and the monitoring of neonatal seizures

To identify neonatal seizures and evaluate the efficiency of clinical observations and short duration electroencephalograms (EEG).

Sixty-three neonates were investigated using prolonged video/EEG monitoring. Patients with confirmed seizures were treated with sequential doses of phenobarbitone and, if seizures persisted, phenytoin. The likelihood of correct management if short duration EEG and clinical observations had been employed was determined.

Thirty-two patients had confirmed seizures. After administration of anticonvulsants, clinical observations identified seizures in a mean of 66% (s.d. 7.3%) of the cohort. A 60 min EEG after each stage of phenobarbitone therapy would guarantee electrographic seizure capture in a mean of 76% (s.d. 10%) of the cohort. A 60 min EEG after addition of phenytoin would guarantee capture in 50%.

An EEG would avoid misdiagnoses in most patients with ambiguous clinical signs. After anticonvulsant infusions, EEG add substantial information to that gained by clinical observations.