Burst suppression electroencephalogram pattern in the newborn: predicting the outcome

A neonatal burst suppression electroencephalogram (EEG) is usually associated with an ominous prognosis. It is controversial whether a reactive burst suppression pattern (ie, a burst suppression pattern that can be interrupted by stimulation) is predictive of a better outcome. We retrospectively studied 22 full-term newborns with burst suppression EEGs to examine their functional outcome. Follow-up (3 to 9 years) was by record review and telephone interview. On the basis of initial EEG pattern and prognosis, three groups were identified post hoc: group 1 (n = 16) had initially nonreactive burst suppression EEGs that remained abnormal; 11 patients died, 4 remained profoundly impaired (nonambulatory, nonverbal), and 1 was moderately impaired (unassisted ambulation, limited speech). Group 2 (n = 3) had initially nonreactive burst suppression EEGs that later improved substantially (within a mean of 7 weeks). At follow-up (3 to 8 1/2 years), each child was waLking (one with braces), talking, and enrolled in special education. Group 3 (n = 3) had reactive burst suppression EEGs initially. At follow-up (ages 3 1/2 to 9 years), each child was walking unassisted and speaking in sentences. Two children in group 3 were of preschool age, and the third was in a special needs program. Although most newborns with nonreactive burst suppression EEGs have a dire neurologic outcome, of those in whom the EEG improves early, the prognosis may be slightly more favorable. Infants with reactive burst suppression EEGs during the acute phase of illness appear to have the best prognosis.