Abstracts

Rationale: We aimed to determine the predictors of postneonatal epilepsy at 1 year of age and the utility of the first hour of EEG monitoring in predicting risk of seizures in the neonatal period. Methods: We identified all neonates admitted to our neonatal intensive care unit between January 2000 and March 2008 who underwent continuous EEG (cEEG) monitoring. Multiple birth-related variables were collected with determination of whether subjects had postneonatal epilepsy at 1 year of age. A Fischer's Exact Test was used to compare clinical variables between infants with and without postneonatal epilepsy. Baseline (first hour) EEG background graded based on Holmes and Lombroso classification (Table 1) was evaluated as a predictive test of the risk of subclinical seizures and need for further EEG monitoring. Grades 1 and 2 of EEG background were grouped together under category of mild baseline EEG background abnormality while grades 3, 4 and 5 were grouped together as significant baseline EEG background abnormality. Results: We identified 58 neonates with a mean gestational age of 37 weeks (24-42). There were 52% male neonates, 21% had a birth weight less than 2.5 kg and 2% had a birth weight less than 1.5 kg. The most common admission diagnoses were hypoxic ischemic injury (45%) and neonatal seizures (16%). There was evidence of clinical seizures in 77% of subjects. The most common MRI findings in the neonatal period consisted of hypoxic ischemic changes (42%). Baseline EEG background distribution is presented in Table 1. About 30% of patients had subclinical seizures exclusively detected using cEEG (Figure 1). About 21% of subjects were lost to follow-up, 21% died by 1 year of age, and 32% had postneonatal epilepsy at 1 year of age. There was no significant association between presence of subclinical seizures (p=0.27), baseline EEG background (p =0.23), presence of MRI abnormality (p=0.51) and admission diagnosis (p=0.06) and risk of postneonatal epilepsy at 1 year of age. The baseline EEG background was predictive of risk of subclinical seizures in the neonatal period (p=0.005). Sensitivity analysis of baseline EEG background abnormality (mild vs. significant) in predicting risk of subclinical seizures showed a sensitivity of 100%, specificity of 27%, positive predictive value of 55% and negative predictive value of 100 %. Conclusions: The risk of postneonatal epilepsy in our group is similar to the reported risk of 36% in previous studies. Our study showed a strong correlation between significant abnormality in the baseline EEG background and the increased risk of subclinical seizures in the neonatal period. This may have profound implications especially when limited resources or contraindications to cEEG monitoring exist. Our study was not able to identify any significant predictors of postneonatal epilepsy at 1 year of age. We propose that prospective multicenter studies using databases to collect data about the various variables that influence risk of postneonatal epilepsy need to be undertaken.