Abstract Details

Title Detailed EEG pattern and spectral analysis can prognosticate good and poor neurologic outcomes after cardiac arrest

Topic Epilepsy/Clinical Neurophysiology (EEG)

Presentation(s) N6 - Neuroscience in the Clinic: Multimodal Tools for Cardiac Arrest Neuroprognostication (6:00 PM-6:10 PM)

Poster/Presentation
Number 001

Background For patients who do not regain consciousness after cardiac arrest, guiding families and healthcare proxies in goals of care decisions presents a major challenge. Accuracy of prognostication testing is limited especially early on, but detailed EEG analysis may help avoid premature withdrawal of life support.

Objective To investigate the prognostic value of a more standardized approach to stratifying electroencephalographical (EEG) patterns and spectral types for patients after cardiac arrest.

Design/Methods A prospectively enrolled cohort of patients with delayed recovery of consciousness after cardiac arrest was stratified into 5 independent EEG patterns and 4 power spectral types. Cerebral performance category (CPC) at discharge provided the primary outcome measure.

Results Out of 72 total consecutive patients, 40 had continuous EEG background without significant epileptiform activity (EEG pattern I), and 7 exhibited organized, reactive alpha rhythm activity (spectral type 0) by day 3. Six patients had CPC 1–2 by discharge, all of whom had EEG pattern I, and 5 of the 6 exhibited spectral type 0 by day 3 (the 6th patient reached type 0 by day 4). Two patients with spectral type 0 by day 3 fully regained consciousness but were scored CPC 3 due to baseline stroke-related deficits. Most patients with EEG pattern I still had poor outcomes (19 with CPC 3, 15 with CPC 4–5). Including spectral analysis, sensitivity for predicting good outcomes (CPC 1–2) was 83.3% (35.9–99.6%), and specificity was 97.0% (89.5–99.6%). If the 2 patients with baseline deficits were counted as good outcomes, the accuracy would improve. Standard prognostication testing all yielded 100% specificity for poor outcomes but low sensitivity, with magnetic resonance imaging being the most sensitive at 54.1% (36.9–70.5%).

Conclusions Qualitative EEG analysis can be useful for prognostication but suffers from poor sensitivity like other modalities. Using spectral analysis can further improve the diagnostic accuracy of EEG.

Authors/Disclosures
Kurt Y. Qing, MD, PhD (New York Presbyterian Hospital, Weill Cornell Medical Center) PRESENTER	Dr. Qing has nothing to disclose.
Peter B. Forgacs, MD	Dr. Forgacs has received personal compensation for serving as an employee of OVID Therapeutics Inc. Dr. Forgacs has received stock or an ownership interest from OVID Therapeutics Inc.
Nicholas D. Schiff, MD	Dr. Schiff has received publishing royalties from a publication relating to health care. Dr. Schiff has received personal compensation in the range of $500-$4,999 for serving as a study section member with NIH. Dr. Schiff has a non-compensated relationship as a Guidelines Panel Member with AAN that is relevant to AAN interests or activities.

��ɫ��

��ɫ��