Abstract

During a two year period prospective continuous electroencephalographic (EEG) monitoring of 275 infants identified seizure activity in 55 cases, 31 of whom were treated with anticonvulsant drugs on clinical grounds. EEG and clinical response was complete in only two and equivocal in another six. Clinical response with persistent EEG seizures occurred in 13 and neither clinical nor EEG response in 10. There was no significant improvement in the generally poor neurological outcome compared with that in 24 infants whose seizures were not treated because of limited or absent clinical manifestations. Background EEG abnormality (as an index of associated cerebral dysfunction) was a guide to potential lack of response to anticonvulsant drugs; it was also predictive of subsequent clinical outcome irrespective of treatment. This study shows that commonly used anticonvulsant drugs (phenobarbitone, paraldehyde, phenytoin, and diazepam) have little effect on seizure control or neurological outcome in neonatal seizures associated with haemorrhagic, hypoxic, or ischaemic cerebral lesions. In view of the variable clinical appearance of EEG seizure activity, continuous EEG monitoring should be an essential feature of further study of neonatal anticonvulsant treatment.