Objective:

To determine the utility of pediatric routine EEGs (rEEGs) in a single-reader high-volume tertiary children’s hospital with long-term follow-up

Background:

The value of rEEGs, especially in children, varies widely, in part due to different cohort size, short rEEG duration, years of follow-up, and inter-reader/inter-institutional variability.   

Design/Methods:

In a retrospective cohort study of children that completed a rEEG during the calendar year 2014 at our institution, all interpreted by the same experienced board-certified pediatric clinical neurophysiologist (CS), key characteristics and outcomes from electronic medical records in the intervening years up to 2023 were extracted, to include pre- and post-Covid-19 pandemic years. Seizure outcomes were categorized as 1) never developed epilepsy with follow-up, 2) no follow-up, 3) developed epilepsy on anti-seizure medication (ASM), 4) developed epilepsy off ASM, and 5) 1 event with no ASM. 

Results:

In the total cohort (977 children with 1042 rEEGs), age at the time of rEEG ranged from 1 day to 26 years (median age 6 years), with a rEEG duration standardly prolonged at 61 minutes. About 47% of rEEGs were abnormal with one or more abnormalities (interictal spikes 91%, slowing 93%, seizures during EEG 14%, hypsarrhythmia 1.2%), and 53% were normal. Approximately 39.5% of patients were lost to follow-up. In the subset with follow-up (n=630), an abnormal rEEG had a positive predictive value of 84% of developing epilepsy, and a normal rEEG had a negative predictive value of 59%, along with 64% sensitivity and 80% specificity. Additionally, a mortality rate from all causes of 2.60% was noted in the overall cohort. 

Conclusions:

Capitalizing on the unique single-center single-reader at a high-volume tertiary children’s hospital with long-term outcome, this study may elucidate more precise rEEG predictive values and patient outcomes than prior studies. More work is needed to improve pediatric rEEG diagnostic and outcome prediction.