Predictive Value of Amplitude-Integrated Electroencephalography in Clinically Identified Neonatal Seizures in Term Neonates

Amarpal Singh, Gopi Kishan Sharma, Prabhudev Basavaraj Hasbi, Rakesh Kumar Gulati

Abstract

Aim: To assess the predictive value of amplitude-integrated electroencephalography (aEEG) in clinically identified neonatal seizures in term neonates admitted in the NICU

Materials and Methods: All term neonates admitted to the NICU with clinically identified seizures were enrolled into the study. The accurate description of the clinical seizure was noted. The aEEG was done once the seizures subsided, the neonate became hemodynamically stable, and within 24 hours of admission. The aEEG was assessed by a single competent examiner. The primary outcome was to identify the predictive value of aEEG in clinically identified seizures.

Results: A total of 100 neonates were included in the study, of which, 50 neonates had seizures. Of these 50 neonates, 35 neonates had an abnormal aEEG, while the remaining 15 neonates had a normal aEEG. Among the 50 neonates without seizures, only 3 had an abnormal aEEG, while the remaining 47 neonates had a normal aEEG. The sensitivity of aEEG in predicting clinically detected neonatal seizures was 70%, specificity was 94%, positive predictive value (abnormal aEEG) was 92.1%, and negative predictive value (normal aEEG) was 75.8%.

Conclusion: aEEG is a good predictor of clinically detected neonatal seizures. It is useful in confirming the clinical diagnosis of suspicious neonatal seizures and acts as a bedside tool for the prediction of neonatal seizures in the NICU.