Abstracts

Rationale: Adult intensive care units (ICUs) report 19% of patients in coma have subclinical seizures(1) and 48% of patients with treated convulsive status epilepticus have nonconvulsive seizures(2). Neonatal ICUs report 42% of babies with HIE have clinical and subclinical seizures(3). Studies from pediatric ICUs (PICU) are few, retrospective, biased to children with suspected seizures, and report prevalence from 0.5% to 44%. Most ICU studies employ short 30 minute EEG recordings without video. We aimed to determine the prevalence and characteristics of seizures in comatose children in PICU, using continuous video- EEG monitoring (vEEG). Methods: We performed continuous vEEG in consecutive comatose children aged 2 months to 17 years, admitted to our PICU between February 2007 and May 2008. Two compact, wireless video and wireless EEG units were used to record and transmit vEEG data to a remote computer. Whole recordings were reviewed offline by two neurologists for seizures. Seizures were defined as abnormal rhythmic activity lasting ≥10 seconds with or without spikes/sharp waves with an incremental onset and decremental offset. Results: 102 vEEG recordings in 100 patients were obtained and reviewed. Median and mean duration of recordings was 20 and 19 hours respectively with 76% of studies more than 12 hours. Diagnoses included: respiratory failure (17), cardiac disorders (14), cardiorespiratory arrest (13), multisystem involvement (12), traumatic brain injury (12), status epilepticus (8), sepsis (7), intracranial hemorrhage (6), meningitis (5), encephalitis (2), encephalopathy (2) and burns (2).Epileptic seizures were identified in only 7 patients and non-epileptic events thought to be seizures in 12. All but one had epilepsy and seizures prior to PICU admission. Seizures were focal in all, subclinical in 4, clinical in 2 and both in 1. Seizures were identified within 3 hours of vEEG. Conclusions: Our prospective study of continuous vEEG in comatose children in PICU revealed a low prevalence of seizures (7%), occurring almost exclusively in children with epilepsy and being detected early during monitoring. Routine use of continuous vEEG in comatose children for seizure detection seems unwarranted. Short duration vEEG in patients with prior seizures/epilepsy or suspected clinical seizures seems appropriate. The low prevalence of seizures in children in PICU, compared to adults and neonates, likely reflects the unbiased ascertainment in this study, and a different spectrum of disease with tendency to rapid recovery. References: (1) Claassen et al (2004) Detection of electrographic seizures with continuous EEG monitoring in critically ill patients. Neurology 62:1743-1748. (2) DeLorenzo et al (1998) Persistent nonconvulsive status epilepticus after the control of convulsive status epilepticus. Epilepsia 39:833-840. (3) Murray et al (2006) Prediction of seizures in asphyxiated neonates: correlation with continuous video-electroencephalographic monitoring. Pediatrics 118:41-46.