and repeated, persistently unifocal clonic seizures. None had generalized tonic or myoclonic seizures. In neonates recorded in our nursery using bedside EEGlpolygraphidvideo monitoring techniques, stroke has been found to be associated primarily with focal clonic seizures 131. Studies of the correlation of neonatal seizures with CT scan and other laboratory and clinical data can have clinical utility only when the seizures themselves are precisely characterized. Thus, the recommendation made by Levy and associates that repeated CT scans be performed in neonates with either generalized or focal seizures to make a definitive diagnosis of stroke might be considered premature. This strategy may best be reserved for those infants with unifocal clonic seizures who may have an increased probability of cerebral infarction.
The precise characterization and classification of seizures of older children and adults provides a sound basis for accurate diagnosis and therapy and the determination of cause and prognosis. Similar care in characterizing and classifying neonatal seizures is equally important and should be more widely applied to the investigation and clinical care of newborns.