Predictive Value of Sequential Electroencephalogram (EEG) in Neonates With Seizures and Its Relation to Neurological Outcome

Abstract

The aim of this study was to evaluate the relationship of sequential neonatal electroencephalography (EEG) and neurological outcome in neonates with seizures to identify polysomnographic features predictive of outcome. Sequential EEGs recordings of 58 neonates that belonged to 2 historical cohorts of newborns with seizures from the same neonatal intensive care unit and who had follow-up at the Neurodevelopment Clinic of the Hospital São Lucas, Pontifícia Universidade Católica do Rio Grande do Sul (PUCRS) in Porto Alegre, Brazil, were analyzed and classified into 4 groups: normal-normal, abnormal-normal, abnormal-abnormal, normal-abnormal. In patients with more than 2 recordings, during the neonatal period, the first EEG was compared with the following more abnormal. A total of 58 pairs of 2 sequential EEGs were analyzed. Considering the first EEG, a statistically significant difference was observed between the relationship of the result of this exam, if it was abnormal, with developmental delay ( P = .030) and postnatal death ( P = .030). Abnormal background activity was also related to neurodevelopment delay ( P = .041). EEG sequences abnormal-abnormal and normal-abnormal significantly correlated to the outcome epilepsy ( P = .015). Abnormal sequential background activity was associated with neurodevelopment delay ( P = .006) and epilepsy ( P = .041). The burst suppression pattern when present in any EEG correlated with epilepsy ( P = .013) and postnatal death ( P = .034). Sequential abnormal background patterns in the first and second EEG increased the risk for epilepsy (relative risk [RR] = 1.8; 95% confidence interval [CI] = 1.03-3.0) and neurodevelopment delay (RR = 2.20; 95% CI = 1.3-3.0). Abnormal background activity only in the second electroencephalogram increased the risk for neurodevelopment delay (RR = 2.20; 95% CI = 1.3-3.0). All the neonates (n = 33) with seizures related to probable hypoxic ischemic encephalopathy had abnormalities in the first EEG ( P = .030). Postnatal epilepsy was diagnosed in 24 infants (41.4%). Five (20.8%) presented West syndrome, 7 (29.2%) focal symptomatic epilepsy, 6 (25%) generalized symptomatic epilepsy, 2 (8.3%) early myoclonic encephalopathy, 1 (4.2%) early infantile epileptic encephalopathy, and in 3, the epileptic syndrome was undefined (12.5%). All infants (n = 5) with West syndrome had some degree of neurodevelopment delay. In conclusion, our findings suggest that sequential EEG in neonates with seizures has more predictive value to estimate the outcomes of neurodevelopment delay, epilepsy, and postnatal death than a single EEG recording. The abnormal background activity in even 1 EEG of the sequential recordings was more significant to determine neurological outcome than abnormal ictal activity or abnormalities in the organization of sleep state.