EEG of the Dying Brain
Kulsatree Praditukrit1, Shwetank Singh2, Neel Fotedar1
1University Hospitals Cleveland Medical Center, 2Case Western Reserve University
Objective:

To report EEG findings in dying human brain.

Background:

The neurophysiological signature of the dying human brain is not well understood. Recently, a single study identified increase in gamma power on quantitative-EEG (qEEG) after withdrawal of the ventilatory support and before death. 

Design/Methods:

We identified seven terminally ill patients who passed away while undergoing continuous scalp video-electroencephalography (vEEG) recording. We performed a retrospective analysis of the raw EEG and qEEG blinded to when the ventilatory support was withdrawn. We classified the EEG into three phases- continuous, discontinuous (burst suppression) and terminal (background suppression). All patients had continuous EKG monitoring as well. 

Results:

Four out of seven patients were noted to have all three phases on EEG from when the EEG was connected to when the patient was declared dead or lost the EKG signal, whichever appeared first. Three patients were in phase II at onset.

Phase I: This phase was characterized by continuous theta or delta activity with or without generalized periodic discharges. The duration varied widely from 11 to 21 hours for four patients. 

Phase II: This phase was marked by onset of generalized burst-suppression pattern with a duration of three minutes to as long as 22 hours (median five hours). The duration of the suppressed segments increased progressively from 1-2 seconds initially to up to 30 seconds in the later part.

Phase III: This was the terminal phase characterized by generalized background suppression (amplitude<10uV) from its onset until either the declaration of death or the loss of ECG signals. The median duration was 31 minutes.

Conclusions:

The EEG of a dying brain is characterized by three distinct phases of continuous theta-delta activity, discontinuous burst-suppression pattern, and terminal background suppression. We did not observe any increase of faster frequencies in our data.

10.1212/WNL.0000000000206704