Abstracts

Rationale: Continuous electroencephalography (cEEG) is important for seizure detection in the neonatal intensive care unit (NICU) as the majority of electrographic seizures (ES) are subclinical. While the impact of seizures is controversial, studies suggest that higher degrees of seizure burden and tentatively earlier treatment may have an impact on developmental outcome. The aim of this study is to characterize our cohort of neonates and to assess the impact of early treatment on seizures. Methods: Retrospective study of patients less than 1 month of corrected gestational age who underwent cEEG monitoring in the NICU at Boston Children's Hospital during a three year period. Only the first monitoring session was included. Patients were excluded if they received an anti-epileptic medication prior to monitoring. Subpopulations included patients that were term and preterm, patients with hypoxic ischemic injury and hypoxic ischemic encephalopathy, stroke, sepsis, metabolic etiology, congenital heart disease, and patients requiring extracorporeal membrane oxygenation. Electrographic status epilepticus (ESE) was defined as either one seizure lasting greater than 30 minutes or seizure burden totaling 30 minutes of a one hour epoch. The degree to which patients were refractory to medication was estimated by the number of anti-seizure medications at the end of the EEG monitoring session. A hazards ratio was used to compare the time of treatment (in relation to the first identified ES) to the number of anti-epileptics at the end of monitoring. Results: Two-hundred-and-ten patients (55% male) with a mean age of 11.6 (18.45) days were monitored. The clinical indications for EEG were characterization of an event (163/210), followed by concern for non-convulsive or subclinical seizures (70/210), with overlap between the two. The leading diagnosis was stroke in 29%, followed by hypoxic injury in 20% of the cohort. 34% of the patients (72/210) had ES and 11% (8/72) had ESE. Of the seizures 38 % (27/72) were subclinical or electrographic only, 51 % (37/72) had a mix of clinical and subclinical seizures and 11 % (8/72) were electro-clinical only. The mean time from cEEG initiation to first recorded seizure was 3.5 hours (SD 5.4). Of patients with seizures 36 met inclusion/exclusion criteria. The median time to treatment was 3.2 hours (Figure 1). Number of anti-epileptic medications at the end of cEEG ranged from 0 to 4. The hazards ratio for time to treatment and outcome defined as number of anti-seizure medications was 1.897 [95% CI (1.208-2.978), P value=0.005], suggesting that a delay in seizure treatment may affect efficacy of treatment. Conclusions: Neonates in the intensive care unit are at high risk of electrographic seizures. Early detection and early treatment of ES may affect efficacy of anti-epileptic medications. Further investigations to control for confounders, such as etiology, may be needed to corroborate results. Funding: No funding.