Abstracts

Rationale: Electrographic seizures (ES) are common in the pediatric intensive care unit with electrographic status epilepticus (ESE) and high degrees of seizure burden having an association with worse neurodevelopmental outcomes. Despite, ES treatment considered standard of care at most centers, the true impact of treatment on seizure burden and outcome remains unknown. Additionally, the approach to treatment of subclinical seizures including medication choice and urgency has not been well defined. We have looked at clinical and EEG features and relationship to seizure burden. We have focused on modifiable risk factors, specifically early treatment, while attempting to understand treatment practices at our center.  Methods: Prospective study of pediatric patients (44 weeks gestational age to 21 years) with ES on cEEG in the Pediatric Intensive Care Unit (PICU) from May 2016 to October 2018. We excluded patients with a prior diagnosis of epilepsy, and those treated with an anti-seizure medication (ASM) prior to arrival. Mild background abnormality was classified as slow/disorganized, moderate as discontinuous, and severe as burst suppression/attenuated. The frequency of discharges were defined as low (rare/occasional) and high frequency (frequent abundant) as defined by ACNS critical care nomenclature.  Results:  38 patients [median 0.7 years (IQR 0.35 – 2.9 years)] were studied of which 67 %( 25/38) were male.  The most common admitting diagnosis was seizure in 18% (7/38). Of the high-risk subpopulations, 29% (11/38) had congenital heart disease (CHD), and 29 %( 11/38) required extracorporeal membrane oxygenation (ECMO).  The EEG background was classified as slow/disorganized in 71 %( 27/38) with interictal epileptiform discharges (IEDs) identified in 84 %( 32/38). The median seizure burden was 32 minutes (IQR 7 – 102) with maximum seizure burden in a 1-hour epoch in the first 24 hours being 18 minutes (IQR 3.8 – 43). Seizure burden was highest in with congenital heart disease surgery in the postoperative period with cardiac arrest. For patients who had a clinical seizure (N=12), 50 %( 6/12) received a benzodiazepines first line. The remaining 26 patients with subclinical seizures on cEEG, 54 %( 14/26) received levetiracetam first line. The time from the first identified ES to first medication was 0.63 hours (IQR 0 – 4.4) with the interval between the first and second medication being 0.11 hours (IQR 0.03 – 0.5). ES burden was lowest in patients with a severe background (p=0.19). Frequency of IEDs were not associated with higher seizure burden (p=0.5).  Patients that died had a higher seizure burden in 24 hours than those that lived (p=0.04). Maximum ES burden in 1 hour was not associated with death (p=0.46). Patients treated sooner after the first identified seizure had a lower seizure burden (0.31 hours vs 1.68 hours). Patients receiving EEG earlier in the admission to the PICU also had a lower seizure burden (0.31 hours vs 1.85 hours).   Conclusions: Electrographic seizure burden is high, above the threshold shown to impact neurodevelopmental outcome [Payne 2014] in the pediatric intensive care unit. Second, the approach to treating subclinical seizures differs in comparison to treatment of clinical seizures. Further study is needed to assess the impact of early aggressive treatment, seizure burden, and developmental outcome.  Funding: No funding