Abstract 111: Early Processed EEG Data Predicts Outcome Better Than Age, Rhythm, or Time to Recovery of Spontaneous Circulation After Cardiac Arrest
Background: Prediction of outcome in the early hours after cardiac arrest remains challenging. We compared processed EEG data from admission and 6 hours post-ROSC for the bispectral index (BIS) and suppression ratio (SR) with other early data from two regional Therapeutic Hypothermia (TH) programs.
Objectives: Evaluate the accuracy of early EEG data to predict outcome after cardiac arrest (CA) treated with TH.
Methods: A multi-center study of 187 patients prospectively monitored with BIS during TH with continued support for ≥72 hours. Good outcome (GO) was defined as a discharge cerebral performance category (CPC) score of 1 or 2, poor (PO) as CPC 3-5. BIS and SR data were assessed blinded to outcome for value after first neuromuscular blockade (BISi, SRi) and 6 hours post-ROSC (BIS6, SR6) and compared to other parameters for ability to predict GO using Receiver Operator Characteristic (ROC) curves.
Results: 187 patients had a mean age of 60 years, 75% were male, 83% witnessed CA, initial rhythm was VT/VF in 52%, and median time to ROSC was 17 (11-26) mins. 88 (47%) had GO (80% for HACA criteria patients). GO patients more often had VT/VF initial rhythm (p<0.001) and shorter time to ROSC (18.7 v 22.9 mins, p=0.05). To predict GO, initial EEG data (BISi AUC=0.87, SRi=0.88) performed better than age (0.54) or time to ROSC (0.62), but EEG data from 6 hours post-ROSC performed best (BIS6 AUC=0.91 and SR6 AUC=0.95).
Conclusions: Among TH treated CA patients fully supported for at least 72 hours, 6 hour processed EEG data, especially the suppression ratio (SR), predict outcome better than admission processed EEG data and much better than time to ROSC, initial rhythm, or age. Processed EEG appears promising as a tool to predict outcome very early after CA.
- © 2013 by American Heart Association, Inc.