The prevalence of epileptic seizures (clinical and subclinical) in 100 consecutive comatose children aged 2 months to 17 years in the pediatric ICU was determined using continuous video-EEG (v-EEE) monitoring, in a prospective study at The Royal Children’s Hospital, Melbourne, Australia. Half of the monitored patients had primary brain disorders, including traumatic brain injury, hypoxic ischemic encephalopathy, and status epilepticus; half had brain dysfunction secondary to systemic sepsis or organ failure. Epileptic seizures were identified in 7 patients, 6 having a history of epilepsy and a seizure witnessed immediately prior to the v-EEG. Median length of ICU stay was 5 days, and median duration of v-EEG was 20 hours. All seizures were recorded in the first 3 hours of v-EEG. Of 18 monitored patients suspected of having epileptic seizures by ICU staff, only 4 (22%) had confirmed seizures. Short-duration v-EEG is more appropriate than continuous v-EEG in comatose PICU patients with a history of seizures, epilepsy, or clinical events suspected to be seizures. 
COMMENT. The authors conclude that epileptic seizures are relatively uncommon (7%) in comatose PICU patients, and v-EEG should be short in duration and limited to those with clinical seizures prior to admission, or in patients suspected of having clinical seizures by medical or nursing staff. Further studies are recommended. These findings and recommendations for PICU patients are in contrast to neonatal ICU comatose patients with a higher prevalence of epileptic seizures and a longer NICU stay. Continuous EEG with simultaneous video recording is recommended for detection of seizures in comatose NICU patients. In studies of neonates involving HIE, seizures occur in 22-59% ; 70-88.5% of seizures are NCS.