LAUSR

Introduction Prognostication after cardiac arrest (CA) poses a challenge to physicians, as it may lead to futile treatment in some cases, and withdrawal of life support therapy (WLST) in others. Clinical and neurophysiological data – electroencephalogram (EEG) and somatosensory evoked potential (SSEP) – are used as prognostic features. In this study, we aimed at analyzing data of patients who survived CA, in a hospital setting in Brazil, where care is not withdrawn even in patients with expected poor neurological outcome. Methods We reviewed clinical and neurophysiological data of patients who had EEG after CA from 2006 to 2016. We considered background suppression, burst-suppression, periodic discharges, and status epilepticus (SE) as ‘malignant’ EEG patterns. Neurologic disability was measured according to the modified Rankin scale (mRS). Scores from 4 to 6 were considered poor outcome. Results 32 patients had EEG after CA. Age ranged from 47 to 98 years (mean 71.6). The majority of patients had in-hospital CA (75%) and the most common cause was hypoxia. 37% received therapeutic hypothermia (TH) as standard care. 73% of patients had poor outcome; 63% of these had a ‘malignant’ EEG pattern after CA. However, 45% of patients who had favorable outcome also had a ‘malignant’ EEG pattern after CA. Patients who received TH presented more ‘malignant’ EEG patterns, compared with those who did not (66% vs 50%). ‘Malignant’ EEG patterns were seen in most of the patients, regardless of CA local (in-hospital 54% vs out-of-hospital 62%). Conclusion Our results emphasize that good outcome may be achieved even if face of ‘malignant’ EEG patterns, and this should be considered when WLST is evaluated.