Abstracts

Rationale: EEG characteristics and the incidence of acute seizures in comatose cardiac arrest (CA) patients treated with therapeutic hypothermia (TH) are unknown. EEG features associated with poor outcome in prior studies have not been characterized in CA patients undergoing TH. Continuous EEG (cEEG) monitoring is increasingly being performed to monitor for electrographic seizures. We sought to determine the incidence of electrographic seizures in CA patients undergoing TH and monitored with cEEG. We hypothesized that features of cEEG correlate with short-term outcome. Methods: We performed a retrospective review of 49 consecutive comatose patients undergoing TH after CA at a single institution from January 2005 to January 2009. cEEG was reviewed daily by 2 electroencephalographers blinded to discharge outcome, and each 2-hour segment was reinterpreted indepdent of the daily reports. The best (“most normal”) background feature (i.e. alpha, theta-alpha, theta-delta, delta, alpha coma, burst suppression and diffuse voltage suppression) present in the cEEG, and the appearance of other features (i.e. electrographic seizure, generalized periodic epileptiform discharge [GPED], periodic lateralizing epileptiform discharge [PLED], other interictal epileptiform discharge, and triphasic waves) were recorded. Outcome was assessed by discharge status from the hospital: death; discharge to hospice, skilled nursing, acute care hospital, and rehabilitation facility; or discharge to home. Results: 38 patients had cEEG monitoring (mean age 55 ± 16 yr, range 20-82 yr). 22 (58%) died in-hospital, 8 (21%) were discharged to acute care hospitals or chronic care facilities, 1 (3%) was discharged to an acute-care rehabilitation facility, and 6 (16%) were discharged to home. 8 patients had electrographic seizures (mean onset from EEG recording to seizure appearance 22 ± 33 hrs; range 0 - 96 hrs). 6/8 patients with electrographic seizures had GPEDs, 2/8 patients had PLEDs, and 7/8 patients had status epilepticus. All patients with electrographic seizures died in-hospital. Features of the cEEG associated with acute mortality (i.e. >75% in-hospital deaths) were: electrographic seizures, burst suppression, GPEDs, diffuse attenuation, theta-delta background, and triphasic waves. Conclusions: Acute seizures were detected by cEEG in 21% of patients undergoing TH; they occured with onset hours to days after CA, most as status epilepticus, and were associated with acute mortality. Classic EEG features of comatose CA patients associated with poor outcome agreed with those in CA patients treated with TH. Of note, triphasic waves were associated with a high rate of mortality and alpha coma was not. Further studies examining the critical period to accurately observe these prognostic features and associations with other covariates are warranted. Therapies targeted towards cEEG in these patients may improve outcome.