The Patient Who Knew Too Much

I am a neurologist. My entire career has been built on the premise that consciousness is a product of brain function — that who we are, what we perceive, what we remember, and what we feel are electrochemical processes occurring within the three pounds of tissue inside our skulls. I still believe this, broadly. But one patient has made me less certain than I used to be.

His name was Thomas, and he was sixty-one years old when he suffered a cardiac arrest in our hospital's cardiac care unit. He had been admitted for unstable angina and was awaiting a catheterization when his heart stopped. The code team responded within thirty seconds. CPR was initiated immediately. But his initial rhythm was asystole — a flatline — and he remained in asystole despite multiple rounds of epinephrine and atropine.

Because he was already on a cardiac monitor, we had an excellent timeline. His cardiac arrest began at 14:32. Advanced cardiac life support continued for eighteen minutes. During that entire period, there was no spontaneous cardiac output — his circulation was maintained entirely by chest compressions. And because we had placed EEG leads for seizure monitoring earlier that day (he had a history of occasional absence seizures), we had a continuous EEG recording throughout the code. The EEG was isoelectric — flat — from approximately thirty seconds after arrest until the return of spontaneous circulation at 14:50.

An isoelectric EEG means there is no detectable electrical activity in the cerebral cortex. No brainwaves. No neuronal firing. By every measure we have, Thomas's brain was not functioning during those eighteen minutes. He was not unconscious — he was, in neurological terms, not present.

When he woke up the following day, he told his wife about a conversation he had overheard between me and the cardiology fellow while we were reviewing his chart at the nurses' station — a conversation that took place approximately twenty feet from his bed, during the code, while his EEG was flat.

"What's his EF?" the fellow had asked me (meaning Thomas's ejection fraction, a measure of cardiac function).

"Twenty-five percent," I had answered. "If he makes it, he's going to need an ICD."

Then I had said to the nurse, "Call his wife. She should be here."

Thomas repeated all three sentences back to me verbatim. He described the fellow's voice — the specific accent, which was Eastern European. He described where I had been standing. He described the nurse picking up the phone to call his wife. All of this happened while his brain was, by every measurable parameter, non-functional.

I have spent my career studying the neural correlates of consciousness. I have published papers on the default mode network, on the role of the thalamus in awareness, on the mechanisms of anesthesia. I am not a novice in this field. And I have no explanation for what Thomas experienced. The temporal lobe hypothesis — that NDEs are caused by seizure-like activity in the dying brain — cannot account for veridical perception during an isoelectric EEG. There is no seizure activity to speak of. The anoxia hypothesis — that the experience is a hallucination produced by an oxygen-starved brain — cannot account for accurate, verifiable recall of conversations that occurred during that anoxia.

I am not saying that Thomas's experience proves the existence of an afterlife or a soul or consciousness independent of the brain. I am saying that it challenges the reductive materialist model of consciousness in ways that science has not yet resolved. And I am saying that as someone who has spent his life studying the brain, I am now more comfortable with the phrase "I don't know" than I used to be.