The clinical value of amplitude-integrated EEG for neonatal seizures: comparison of short-term (6-12 hours) and long-term (24-48 hours) monitoring

Objective. To investigate the predictive value of amplitude-integrated electroencephalography (aEEG) monitoring time for seizure detection in neonates with encephalopathy. Method. The study included 105 newborns with neonatal encephalopathy in the neonatal intensive care unit at Ege University Hospital between 2010 and 2015 (term:54, preterm:51). As a departmental protocol, aEEG monitoring was performed within the first 24 hours; aEEG and a one-hour routine conventional EEG (RC-EEG) recording were performed simultaneously. Based on the monitoring duration, two study groups consisted of group I (n=36): short-term recording (6-12 hours), and group 2 (n=52): long-term monitoring (24-48 hours). The aEEG and RC-EEG recordings were evaluated by two researchers as double-blind. Results: Long-term aEEG monitoring recorded seizures in 15 of 36 (41.6%) neonates. However, short-term aEEG monitoring recorded seizures in 17 (32%) of 53 neonates. The sensitivity and PPV of long-term aEEG and short-term aEEG monitoring for seizure detection were 73.3%, 35.4%, 64.7%, and 42.3%, respectively. During a prolonged period of EEG monitoring, neonates with seizures had a lower aEEG score than neonates without seizures; 4.263.17 versus 6.292.6, respectively (p=0.015). There was consistency between the aEEG and the RC-EEG with respect to background characteristics in both cohorts (r=-0.452). Abnormal background patterns were identified in 27 of 36 (75%) with short-term aEEG and 36 of 52 (69%) with long-term aEEG monitoring. Conclusion. This study provides that long-term aEEG monitoring up to 48 hours in preterm and term neonates improves the diagnostic yield of aEEG for seizure detection.
Keywords: neonatal seizure, aEEG, EEG monitoring