Abstracts

Rationale: EEG monitoring in patients treated with therapeutic hypothermia after cardiac arrest (CA) may assist in early outcome prediction. Recently, we found with high predictive values in a prospective cohort study of 56 patients that low voltage or iso-electric EEG patterns at 24 hours after CA are associated with poor outcome, and normal or diffusely slowed EEG patterns at 12 hours after CA with good outcome (Cloostermans, Crit Care Med 2012). We now present data from 141 patients and improved EEG criteria.Methods: All adult patients (aged > 18 years), who were resuscitated after CA, remained comatose, and were admitted to the intensive care unit (ICU) to receive therapeutic hypothermia were included. Patients with additional neurological injuries were excluded. In all patients, continuous EEG data (21-channels) were recorded. In the first 56 patients daily somatosensory evoked potentials (SSEP) were measured as well. We evaluated EEG recordings by visual analysis at 12 and 24 hours after CA, where the EEG was classified as isoelectric, low-voltage (<20 V), epileptiform (including generalized periodic discharges), burst suppression or diffusely slowed. Burst suppression patterns were further classified into patterns with and without identical bursts (Figure 1) (Van Putten, Clin Neurophysiol 2010). All EEG epochs were scored by two independent reviewers blinded for patient outcome. The primary outcome measure was the best score on the cerebral performance category (CPC) within 6 months dichotomized as good (CPC=1-2) or poor (CPC=3-5).Results: In our prospective observational study cohort, 141 patients are now included. Sixty-four patients (45%) achieved good outcome. Normal or diffusely slowed EEG patterns at 12 hours after CA, are associated with a good outcome (Table 1). Iso-electric, low voltage, and burst-suppression patterns with identical bursts at 24 hours after CA is associated with a poor outcome (Table 1). Conclusions: In comatose patients after CA treated with hypothermia, EEG monitoring during the first 24 hours after CA can contribute to reliable prediction of both good and poor neurological outcome.