To study resting EEG spectral features related to cortico-thalamic integrity in relation to outcomes in a prospective cohort of patients post-cardiac arrest.

In comatose post-cardiac arrest patients, prognostication of neurological recovery is often challenging. Previous retrospective studies demonstrated that frequency spectral features of background EEG activity can provide early insight into the integrity of cortico-thalamic circuitry and may correlate with recovery of consciousness after cardiac arrest. We aim to validate these findings in an independent, prospective cohort, as well as explore the predictive value of quantification of spectral features.

Thirty-four patients were prospectively enrolled. EEG segments from morning and night (~9AM and 9PM) up to 3 consecutive days were analyzed in a blinded manner. EEG data were sorted into the following categories: (O) - malignant (i.e. epileptiform), (A) - features suggesting loss of cortico-thalamic integrity, (B) - features suggesting impaired cortico-thalamic integrity, and (D) - normal background activity. Outcome was cerebral performance categories (CPC) at time of discharge.

Two patients had good outcomes (CPC = 2), and 14 had moderate outcomes (CPC = 3). Twenty had poor outcomes (CPC = 4 or 5), with 4 having cardiac death, 1 brain death, 1 in coma, and 14 passed after withdrawal of care. Of the EEG data, 5 patients were in category O, 14 in A, 10 in B, and 5 in D. Results from Kruskal-Wallis test suggest that there is statistically significant evidence that outcomes and EEG categories are dependent (p = 0.0276). Multiple comparison yielded only significant difference in outcomes between categories O and D, with a trending effect in the other groups.

EEG spectral features correlate with better outcomes. With more data and with detailed quantification of spectral features, differences in outcome in relation to EEG categories may be more apparent.