Abstracts

When patients receive paralytic medications, seizure activity cannot be readily detected clinically. Intermittent bedside EEG monitoring can detect clinically occult seizure activity in these patients. We sought to determine the occurrence of recurrent seizures or status epilepticus in a series of monitored pediatric patients receiving paralytics during mechanical ventilation, and to look for patient and/or EEG characteristics that might identify which patients are at highest risk for developing these complications. The Baylor IRB approved our protocol for a retrospective review of consecutive bedside monitoring studies performed between January 2001 and December 2003 on pediatric ICU patients receiving paralytic agents. We divided patients into four groups based on indications for monitoring. Within each group we recorded the occurrence of electrical seizures, the presence of interictal epileptiform activity and the timing of the first recorded electrical seizure. We reviewed ninety-eight monitoring sessions recorded from 82 patients aged 6 weeks to 24 years. Five patients had two sessions and two patients had three sessions. Monitoring duration ranged from 6 hours to 48 days, mean 150 hours. Twenty nine sessions were performed for a history of epilepsy, 24 for occurrence of seizure within 24 hours prior to monitoring, 26 for status epilepticus, and 19 for other conditions thought to increase risk for seizures. Thirty of the 98 monitoring sessions captured seizure activity. Six of these 30 positive studies captured only one seizure, one captured 2 seizures and one captured 6 seizures; twenty-two captured status epilepticus. Fourteen of 26 patients (54%) monitored because of status epilepticus manifested status during monitoring, compared to 5 monitored for seizure within 24 hours (21%), 1 monitored for other conditions (5%), and 2 monitored for history of epilepsy (7%). Eleven patients had ongoing status epilepticus at the onset of monitoring. Latency to development of status in the remaining 11 cases ranged from 3 hours to 20 days, with 5 occurring in less than 72 hours. The absence of interictal epileptiform activity predicted that no seizures would be captured with a negative predictive value of 84%. The yield of intermittent bedside EEG monitoring in paralyzed pediatric patients is highest in patients who are monitored because of recent seizure activity or status epilepticus. Patients without interictal epileptiform activity were unlikely to develop status epilepticus, and patients monitored due pre-existing epilepsy or due to an acute illness or insult rarely did. We plan to investigate the patients in these low-risk groups to identify factors that might guide the utilization of this costly and labor-intensive procedure.