ABSTRACT
PURPOSE: This study evaluated the utility of 20-minute spot electroencephalography (sEEG) for prognostication in patients treated with therapeutic hypothermia after cardiac arrest. METHODS: This was a retrospective, observational study using a prospectively collected hypothermia registry conducted at a single tertiary hospital over three years. We included patients who were monitored by sEEG within three days after resuscitation. We evaluated the prognostic value of continuous activity pattern (diffuse slowing or normal wave) for good outcome predictor and malignant pattern (epileptiform, burst suppression, and isoelectric pattern) for poor outcome predictors during hypothermia and normothermia period. RESULTS: Of 98 patients, continuous activity was displayed in 45 patients. Of these, 27 were included in the good outcome group, resulting in a positive predictive value (PPV) of 60.0% (sensitivity and specificity of 100.0 and 74.6%, respectively). Malignant patterns resulted in a PPV of 100.0% for poor outcomes (sensitivity and specificity of 35.2% and 100.0%, respectively). Continuous activity during hypothermia showed better prognostic performance for good outcome than those during normothermia (area under the curve (AUC)=0.900 (95% confidence interval (CI) 0.825-0.975) and AUC=0.827 (95% CI 0.692-0.962), respectively). For the poor outcome prediction, the AUC was 0.712 (95% CI 0.536-0.887) for malignant patterns during normothermia, and 0.656 (95% CI 0.522-0.789) for those during hypothermia. CONCLUSION: For patients treated with TH, continuous activity on sEEG, particularly during hypothermia, could contribute to the prediction of good outcomes with high specificity. Epileptiform discharge, burst suppression, and isoelectric patterns accurately predicted poor outcomes in this cohort.