Neuroprognostication After Cardiac Arrest: What the 72-Hour Wait Is Actually Measuring

ICU clinical · Updated 2026 · Educational content — always follow your unit's protocol and providers' orders

The hardest conversation in the post-arrest ICU is not the code — it is day two, when the family asks "is he going to wake up?" and the honest answer is "we are not allowed to know yet." Neuroprognostication — the structured process of estimating neurologic recovery after resuscitation — is deliberately slow, deliberately multimodal, and deliberately pessimism-resistant. Understanding why makes you the nurse who can hold that family steady through the wait. This continues our post-arrest series: TTM for ICU nurses and TTM after cardiac arrest.

Why nothing gets called early

Two reasons, and they drive everything else. First, sedation and hypothermia wreck the exam: propofol, benzodiazepines, opioids, and neuromuscular blockade linger — longer still with hypothermia-slowed clearance and renal/hepatic injury — so an unresponsive patient at 24 hours may simply be a sedated patient (see the propofol guide for context on sedation clearance). Second, and more important, the self-fulfilling prophecy: if teams predict poor outcome early and withdraw life-sustaining therapy, the prediction "proves" itself — and some of those patients would have recovered. Modern guidance pushes prognostication to at least 72 hours after return to normothermia, off confounding sedation, precisely to break that loop. The waiting is not indecision. It is the safety mechanism.

The multimodal principle

No single test is allowed to condemn a patient. The framework looks for concordance — multiple independent modalities all pointing the same direction — before anyone speaks about a poor prognosis, because every individual test has false positives:

The nurse's actual role in prognostication

1. Protect the exam. Documented sedation interruptions, accurate last-dose times for every sedative and paralytic, and flagging renal/hepatic dysfunction that delays clearance — the neurology consult is only as good as this record.
2. Chart the trajectory, not the moment. "Localizing this morning, withdrawal only tonight" is exactly the kind of serial observation that changes team-level interpretation. Your q1–q4h neuro checks are the densest data stream in the workup.
3. Guard against the prophecy at the bedside. Hallway phrases like "he's gone" in earshot of family — or charting editorial conclusions — anchor decisions before the data exists. Describe findings; let the multimodal process conclude.
4. Catch the confounders: hypoglycemia, severe metabolic derangement, non-convulsive seizures, residual paralytic (request train-of-four if blockade was used) — each can mimic devastation.
5. Translate the wait for families. The most useful sentence in this ICU week: "The brain needs time to show us what it can do, and the tests are designed so we never give up on someone too early." It is accurate, kind, and buys the process its 72+ hours.

When the news is bad — and when it is not

When concordant multimodal findings support a poor prognosis, the pathway shifts to goals-of-care conversations, and the nurse's job becomes family support, dignity, and — where consistent with the patient's wishes — organ donation referral per protocol (a referral, in most systems, that is legally the OPO's conversation to have, not the bedside team's). But remember the other branch: a meaningful fraction of post-arrest patients wake late — days after sedation stops — which is exactly why the framework refuses to hurry. Some of the best days in critical care are day-6 thumbs-ups from patients whose day-2 exam looked hopeless.