Prediction of poor outcome using detector of epileptiform EEG in ICU patients resuscitated after cardiac arrest.

Assessing the brain status of patients admitted to intensive care unit (ICU) after out-of-hospital cardiac arrest is challenging. We had earlier found wavelet subband entropy (WSE) to be a useful tool for quantifying the epileptiform content of EEG during anesthesia. In this paper, WSE was applied for EEG of ICU patients to study its prognostic value. During their stay in ICU, EEG was recorded from 20 patients resuscitated after out-of-hospital cardiac arrest. For the analysis, the patients were divided into subgroups of poor outcome (persistent vegetative state, N=4) and good outcome (regain of consciousness, N=16). WSE for each 5-sec segment of EEG was calculated and also the average of WSE for each hour. Also, similar results were calculated for EEG powers in the bands 16-32 Hz and 1-60 Hz to be used as references. The statistical analysis was made by comparing the medians of the distributions of average WSE of each hour between poor and good outcome groups. The median of WSE of poor outcome group was significantly lower than that of good outcome group. The reference indicators did not show significant differences between the groups. The results suggest that WSE can be a valuable prognostic indicator for detecting the patients with poor outcome.