Electroencephalographic patterns and prediction of outcome in comatose survivors after cardiopulmonary resuscitation / 中国脑血管病杂志

ABSTRACT

Objective: To explore the opportunity of electroencephalographic (EEG) evaluation and the prognostic value of the EEG patterns in comatose survivors after cardiopulmonary resuscitation (CPR). Methods: Sixty-four comatose survivors were selected for EEG evaluation after CPR. Acccording to the time intervals of the first EEG monitoring, the patients were divided into 1-3 day, 4-7 day, and >7 day groups. The EEG monitoring included generalized suppression, burst-suppression, alpha or theta coma, and slow wave increase patterns. The endpoint of clinical outcome (Glasgow Outcome Scale score) was evaluated at 6 months. The 4 types of EEG patterns were calculated respectively, and the prognostic sensitivity, specificity, positive predictive value, false positive rates, and total consistent rates during the 3 time intervals were predicted. Results: The generalized suppression pattern had a higher sensitivity (67%-80%) and specificity (100%) in predicting unfavorable prognosis within 7 days after CPR; the burst suppression pattern had a higher specificity (100%) in predicting unfavorable prognosis within 7 days, however, its sensitivity was low (6%-8%); The slow wave increase pattern had high sensitivity (100%) and specificity (91%-94%) in predicting favorable prognosis within 7 days; and the sensitivity was 3%-40%, and specificity was 50%-67% in predicting unfavorable prognosis in the alpha or theta coma pattern. Conclusion: The early (within 7 days) EEG after CPR may accurately and reliably show that both the generalized suppression and burst-suppression patterns predict the unfavorable prognosis; the slow wave increase pattern can accurately and reliably predict favorable prognosis; and the alpha or theta coma pattern cannot favorably predict prognosis.