EEG BACKGROUND AS PREDICTOR OF ELECTROGRAPHIC SEIZURES IN HIGH-RISK NEONATES

Purpose: Experience with continuous EEG monitoring in 29 consecutive i nfants at risk of neonatal seizures demonstrated that background abnor malities on the initial EEG were strongly associated wit electrographi c seizures in the subsequent 18-24 h. To test this association prospec tively, we examined the relationship between EEG background and the pr esence of electrographic seizures in the next 22 at-risk infants monit ored for seizures. Methods: A standard EEG was recorded for each infan t at risk for seizures. Based on specific criteria, the background was graded as normal; immature for age; or mildly, moderately, or severel y abnormal. The EEG was then monitored continuously until no electrogr aphic seizures were recorded for 18-24 h. Results: Findings in the ret rospective group of 29 infants and the prospective group of 22 infants did not differ. For the combined 51 infants, a normal or immature ini tial EEG background predicted the absence of seizures in the subsequen t 18-24 h with a sensitivity of 96% [confidence interval (CI) 0.88, 1. 0] and specificity of 81% (CI 0.67, 0.96). Administration of antiepile ptic drugs (AEDs) before the start of the EEG recording did not affect this association. Conclusions: In 51 consecutively monitored infants at risk for neonatal seizures, a normal or immature EEG background str ongly predicted the absence of electrographic seizures in the subseque nt 18-24 h; background abnormalities strongly predicted eht occurrence of electrographic seizures concomitantly or in the subsequent 18-24 h of recording. Screening infants at risk for neonatal seizures with a routine EEG allows identification of infants at highest risk for seizu res, thus conserving resources required for continuous EEG monitoring and facilitating early intervention for seizures.