Abstracts

Rationale: Continuous electroencephalographic monitoring (cEEG) is frequently used in critically ill children admitted to the pediatric intensive care unit (PICU). Approximately 10-50% of children that undergo a clinically indicated cEEG have electrographic seizures, the majority of which are subclinical. cEEG is resource intense and limited in many hospitals. Early EEG biomarkers for seizures are needed to allocate this resource towards patents most likely to have electrographic seizures. The aim of this study is to assess the utility of early epileptiform discharges as predictors of electrographic seizure detection in the pediatric intensive care unit. Methods: Prospective study of pediatric patients from 44 weeks gestational age to 21 years who underwent a clinically indicated cEEG in the PICU. cEEG was defined as greater than three hours of uninterrupted EEG. In patients with multiple cEEG procedures, only the first recording was considered. Patients were excluded if they were admitted in the setting of epilepsy surgery. Electrographic seizures were defined as any seizure detected on cEEG, whether electro-clinical or electrographic-only. Electrographic status epilepticus (ESE) was defined as a continuous seizure lasting greater than 30 minutes or seizures totaling at least 50% of a 1 hour epoch. EEG background was categorized into normal, slow disorganized, attenuated and featureless, discontinuous, and burst suppression. The presence of epileptiform discharges was also documented as being either present or absent. The time of the first epileptiform discharge was identified in relation to the start of cEEG monitoring. Results: Thirty two patients satisfied the inclusion-exclusion criteria. Patient ages ranged from 44 weeks gestation to 19 years. Sixteen percent (5/32) of patients had electrographic seizures (ES). Epileptiform discharges (ED) were documented in 34% (11/32). Of the patients with ES, only 1 patient did not have ED. ED or ES were detected within the first hour of recording in all patients that developed ES. Thirteen percent (4/32) of patients died. Three of those patients had an attenuated/featureless background at the start of recording. Ten patients had a normal background of which there were no detected seizures. Conclusions: This study suggests that the decision to transition to cEEG for seizure detection can be made by identification of epileptiform discharges on a routine or prolonged study. This will ultimately help allocate resources towards patients that are at greatest risk of developing electrographic seizures. An initial normal EEG is also predictive of not identifying seizures. Funding: None