Which Two Statements Accurately Reflect the Recommendations for Post–Cardiac Arrest Neuroprognostication

The two statements that accurately reflect the recommendations for post–cardiac arrest neuroprognostication are: (1) Before any prognostic testing is interpreted, the patient must be rewarmed after targeted temperature management and confounders such as sedation, neuromuscular blockade, and metabolic derangement must be excluded; and (2) no single test or finding should be used alone—prognostication requires a multimodal strategy combining serial clinical examination, electroencephalography, somatosensory evoked potentials, neuroimaging, and blood biomarkers, interpreted by a multidisciplinary team. These recommendations come from the American Heart Association and European Resuscitation Council and apply to comatose survivors of cardiac arrest.

The core principle: multimodal, delayed, confounder-free

Neuroprognostication after cardiac arrest is a formal recommendation, not a single lab value. The two correct statements above are not arbitrary—they anchor a body of guidance built on a hard-won lesson from the early 2000s. Early protocols often declared poor outcomes using one or two signs (an absent pupil reflex, a low Glasgow score, a burst-suppression pattern on EEG) within 24–48 hours of arrest. Those protocols produced false positives: patients labeled unsalvageable who later woke up, especially after targeted temperature management (TTM) altered the pharmacokinetics of sedatives and prolonged drug clearance.

The mechanism behind "multimodal" is statistical, not ceremonial. Each individual prognostic test has imperfect sensitivity and specificity. Combining independent tests multiplies the chance of a false-positive poor outcome downward, provided the tests probe different injury pathways. A clinical exam probes brainstem and cortical integration. SSEPs probe the dorsal column–medial lemniscus pathway. EEG probes cortical network activity. Neuron-specific enolase probes axonal injury. MRI probes structural damage. When five different instruments agree on severe injury, the probability of a good outcome drops steeply. When they disagree, the disagreement itself is information—it says "wait." The principle is that a single test can be wrong for a mechanical reason (hypothermia, drug effect, artifact), but five unrelated tests are unlikely to be wrong in the same direction.

Why timing and rewarming are the other accurate statement

The mechanism for the timing requirement is straightforward physiology plus pharmacology. TTM at 32–36 °C slows hepatic metabolism of sedatives, so midazolam and fentanyl clear far more slowly during and after cooling. A patient who appears deeply comatose and unresponsive during cooling may be comatose partly because of drug, not injury. Rebound hyperthermia after rewarming can confound the exam in the opposite direction. Electrophysiology changes with temperature: burst suppression and EEG suppression are more common at colder temperatures, and SSEP latencies shift. If you interpret these findings while the patient is still cold or metabolically unstable, you are measuring the thermometer as much as the brain.

The recommendation therefore specifies that prognostication should be performed in the normothermic patient (target 36.5–37.5 °C core temperature), generally no earlier than 72 hours after return of spontaneous circulation (ROSC), and only after excluding confounders: sedative washout, neuromuscular blockade reversal, and correction of hypothermia, hyperthermia, hypoxia, hypotension, and severe metabolic derangement. The 72-hour threshold is a floor, not a deadline. Many protocols extend to 96–120 hours or more when the clinical picture is ambiguous. The plain-language version: you cannot judge a brain until the patient's physiology is stable and the drugs have worn off.

Comparison table: prognostic tests and what each contributes

The table is not a menu to pick from. Guidance recommends using tests in combination, and the 2023–2024-era documents emphasize that SSEP absence of the N20 response, when present bilaterally, is among the more specific indicators of poor outcome—but specificity only helps decision-making when combined with other modalities and a stable exam. A good outcome is still technically possible in a small fraction of patients with a single adverse finding, which is exactly why the multimodal rule exists.

The counterintuitive detail: high specificity is not a license

The most common professional caveat is this: high specificity is not permission to withdraw care on its own. A test with 95% specificity for poor outcome means that 5% of patients with that finding might have survived with reasonable function. When you apply such a test to thousands of patients, the absolute number of misclassified survivors is not trivial. This is why the guidelines phrase recommendations around a multimodal composite rather than a single threshold, and why they emphasize a multidisciplinary team—neurology, neurophysiology, critical care, palliative care—rather than a single consultant's sign-off.

There is a second counterintuitive point about family communication. Prognostication is not a verdict; it is a probability distribution, and the recommendations treat structured family communication as part of the process. The team should disclose uncertainty honestly, describe the trajectory and time course, and avoid definitive statements about outcome that outrun the data. A family told "he will not recover" based on one EEG pattern has been told something the evidence cannot support.

Quantified context: thresholds that require reconciliation

Several widely used thresholds appear across guidance, and they must be reconciled rather than applied mechanically:

Bilaterally absent pupillary light reflex at ≥72 hours: strongly associated with poor outcome, but confounded by sedation and hypothermia. Bilaterally absent N20 on SSEP at ≥72 hours: highly specific for poor neurologic outcome; the exact figure cited varies by cohort. Neuron-specific enolase: thresholds differ by assay; values are assay-dependent and cannot be pooled across manufacturers. Guidance uses assay-specific cutoffs rather than a universal number. Malignant EEG patterns (burst suppression, generalized suppression, status epilepticus): associated with poor outcome but not sufficient alone because of drug and temperature effects. MRI: extensive cortical and deep gray matter injury on diffusion-weighted imaging correlates with poor outcome; extent and location matter, and sedation/transport logistics matter.

The reconciliation rule: each test informs, none decides. A threshold is a data point that changes the probability; the probability changes management; management is a team decision. The plain-language translation: no single number, no single scan, no single reflex is a decision.

Practical actions: what a care team actually does

Rewarm to normothermia (36.5–37.5 °C) and confirm it before interpreting any prognostic test. Confirm adequate sedation washout and neuromuscular blockade reversal; document last sedative dose and timing. Perform serial exams rather than a single snapshot, starting no earlier than 24 hours and formally at ≥72 hours post-ROSC. Obtain continuous or serial EEG early, and interpret malignant patterns in the context of drugs and temperature. Obtain SSEP, biomarker, and imaging as adjuncts, in a protocolized sequence rather than opportunistically. Present the composite to a multidisciplinary team and communicate uncertainty to the family in plain language.

Red flags: interpreting a single EEG pattern drawn during sedation; withdrawing life-sustaining therapy before rewarming; using a neuron-specific enolase value across mismatched assays; treating an MRI performed during transport instability as definitive; declaring a poor outcome at 24 hours.

Timeline table: the neuroprognostication sequence

Misconceptions and limitations

The most persistent misconception is that a single absent reflex or a single EEG pattern equals a poor outcome. The second is that earlier equals better—more aggressive early prognostication produces more false positives, not more accuracy. The third is that neuroprognostication is purely a neurologist's task; guidelines explicitly place it in a multidisciplinary framework.

Limitations are real. Much of the evidence base derives from cohorts treated under specific TTM protocols, and outcomes may differ with changes in post-arrest care, targeted temperature targets, and sedation regimens. Biomarker thresholds are assay-dependent. Imaging interpretation requires expertise that may not be available around the clock. The 72-hour threshold is a floor, not a guarantee that the picture is clear. And communication of uncertainty is difficult to standardize, which makes it the least evidence-based part of an otherwise protocolized process.

For the "which two statements" question, the answer is the multimodal, confounder-free assessment and the delayed, post-rewarming, multidisciplinary interpretation. Everything else in this article exists to explain why those two statements are the ones that survived scrutiny.

This article is general educational information about clinical guidance, not a diagnosis, treatment plan, or recommendation for any individual patient. Neuroprognostication decisions require a qualified local clinical team applying current protocols to the specific patient.