Abstracts

Rationale: Therapeutic hypothermia (TH) and medically induced coma are used frequently to increase survival rate in patients who suffered from cardiac arrest (CA). Previous studies have shown the prognosis of most of these patients to be poor. Continuous EEG (cEEG) is used frequently to provide real-time neurological information in otherwise comatose patients and prognostic information. We decided to investigate the correlation between specific EEG patterns and outcomes in post CA patients who have undergone TH and medically induced coma Methods: We retrospectively identified 83 consecutive post CA patients who have undergone TH and cEEG from January 2010 through December 2012 at our institution. cEEG was started within six hours of initiating TH and was continued past 24 hours after the patient had achieved normal body temperature. cEEG data was reviewed independently by two neurophysiologists who were blinded to patient outcomes, and the findings were categorized into the following groups: diffuse slowing, epileptiform abnormality/seizure, burst suppression, background attenuation, and isoelectric background. Neurologic outcomes were measured at the time of hospital discharge according to the following Glasgow-Pittsburg Cerebral Performance Category (CPC) 5-point scale. Results: Of the 83 post CA patients included in this study, 29 had initial cEEG pattern of diffuse slowing, 24 burst suppression, 23 background attenuation, 3 epileptiform abnormality/seizures, and 4 isoelectric background. Seven of 29 (24.1%) patients with initial cEEG showing diffuse slowing did not survive. The mortality rate of patients with initial cEEG patterns of isoelectric background, epileptiform abnormality/seizure, burst suppression, and background attenuation were 100% (4/4), 66.7% (2/3), 66.7% (16/24), and 87% (20/23), respectively. Among the 34 patients who survived (41%), 12 of 22 (54.5%) patients with initial cEEG pattern of diffuse slowing had good neurological outcome (CPC scores 1-2) whereas the ratios for those with isoelectric background, burst suppression, and background attenuation were 0%, 25%, 33.3%, respectively. One of three patients with epileptiform abnormality/seizure on initial cEEG survived and had good neurological outcome. Conclusions: Similar to previous studies of post CA patients who have undergone TH, the survival in our patients was poor (41%). Of them, 64.7% had initial cEEG patterns of diffuse slowing and 54.5% among these patients had good neurological outcome. Thus, the findings of diffuse slowing as an initial cEEG pattern may be an adjunctive indicator for prognosis at the onset of TH for post CA patients in whom the neurologic exam findings are of limited reliability. However, the number of survivors with initial cEEG pattern of epileptiform abnormality/seizure (1) and background attenuation (3) is too small for adequate analysis and as a prognostic indicator. Larger population study is underway.