164 Comparison of aEEG Classification with the Underlying Single Channel Raw EEG Background Activity in Neontal Encephalopathy

Abstract

Background: The amplitude-integrated electroencephalogram (aEEG) is increasingly used in neonatal units. A simple aEEG classification has been shown to be predictive for neurodevelopmental outcome after neonatal encephalopathy (NE) and has been used to select infants for studies of neuroprotective therapy. New equipment has been developed with the possibility of combining the aEEG with the continuous EEG.Aim: To compare an aEEG classification with the underlying raw single channel EEG activity in NE.Subjects and Methods: The aEEG was recorded for median 61 (IQR 36–95) hours in 8 term infants with NE using a digital aEEG monitor (CFM 6000 Olympic Medical, USA). The aEEG was classified as normal, moderately abnormal or severely suppressed according to the voltage of the lower and upper margin of the trace (1). The EEG was classified as continuous, moderately discontinuous (interburst interval <20 seconds), severely discontinuous (interburst interval 20–40 seconds) and extremely discontinuous (interburst interval >40 seconds) (2). 10 hours of each aEEG classification were assessed (total 30 hours). Traces with high impedance were excluded.Results: Normal aEEG records showed continuous EEG activity throughout the recordings. In the moderately abnormal aEEG group, the underlying EEG was moderately discontinuous for 97% and severely discontinuous for 3% of the recordings, and the median interburst interval was 4 (IQR 3–6) seconds. In the severely suppressed aEEG records, the underlying EEG was extremely discontinuous for 83%, severely discontinuous for 9% and moderately discontinuous for 8% of the recordings, and the median interburst interval was 20 (IQR 10–49) seconds, significantly longer than in the moderate aEEG group (p=0.008).Conclusion: A good correlation of the aEEG classification with the underlying EEG was shown. Combining assessment of aEEG and the degree of discontinuity of the underlying EEG may further refine the prognostic accuracy of aEEG.