Abstract 54: Seizure Detection During Therapeutic Hypothermia After Resuscitation From Cardiac Arrest
Background: Seizures occur in up to 25% of patients after initial resuscitation from cardiac arrest, with most occurring within 24 hours of arrest. International guidelines have suggested the post-arrest use of electroencephalographic (EEG) monitoring, yet continuous EEG is not readily available in many hospitals. Several case reports have suggested that monitoring via Bispectral Index (BIS), which represents a calculated summary of raw EEG data including frequency and amplitude, may have utility in detecting seizures. The utility of BIS to detect seizures, especially in the setting of therapeutic hypothermia (TH), is unknown.
Objectives: We hypothesized that seizures would be accompanied by a noticeable increase in numeric BIS values compared to before electrical convulsive activity, in post-arrest patients receiving TH.
Methods: We performed a single center retrospective analysis of consecutive post-arrest patients treated between 5/2005–5/2009. Inclusion required the utilization of TH post-arrest as well as concurrent EEG and BIS monitoring throughout the TH period. Per hospital protocol, BIS values were documented hourly during TH. Continuous EEG was interpreted by two neurologists for seizure activity. The three documented BIS values prior to seizure activity were averaged and compared to BIS values after seizure onset.
Results: Post-arrest TH with both EEG and BIS monitoring was performed in 54 patients. The mean age was 54±16, 24/54 (44%) were female, and 22/54 (41%) survived to discharge. Nine patients were identified that had EEG evidence of seizures and had BIS documented before and after onset of electrical convulsive activity. Mean seizure duration was 1044±805 min. When BIS values were compared before and after seizure onset in each patient, no statistically significant changes were found. Averaged across the nine patient subset, mean BIS prior to seizure was 34±17, and after seizure onset mean BIS was 34±12. (p = 0.74).
Conclusions: There was not a significant difference in hourly BIS values before and after seizure onset. Whether other aspects of BIS recording could indicate seizure activity will require further study. During TH, following the numeric BIS values may not be an adequate alternative for EEG monitoring to detect seizures.
- © 2010 by American Heart Association, Inc.