Prognostication After Cardiac Arrest in a Patient With Acute Kidney Injury
A 66-year-old man with witnessed ventricular fibrillation arrest, bystander CPR, and ROSC after 22 minutes. STEMI treated with PCI. He received targeted temperature management at 33°C for 24 hours, rewarmed slowly, and is now at 96 hours post-arrest. Sedation (propofol and fentanyl) was stopped 60 hours ago; no neuromuscular blockade for 72 hours. Temperature 36.6°C, MAP 78 mmHg on norepinephrine 4 mcg/min, PaO2 110 on FiO2 0.35, lactate normal. Exam: pupils 3 mm and nonreactive bilaterally, corneal reflexes absent, no motor response to painful stimulus, no myoclonus. Creatinine 2.6 mg/dL (baseline 1.0), NSE 42 mcg/L, EEG shows continuous generalized suppression with brief bursts, SSEP not yet returned. The family asks, "Is he brain dead? Can we stop?"
1. Can you make a definitive poor-neurologic-outcome prediction now? Identify what is confounded and what you would still obtain.
Read the worked answer
Not yet - the examination is supportive but not yet definitive because two important confounders are in play.
What is confounded:
- Renal failure invalidates NSE as a stand-alone marker. NSE is cleared renally; a value of 42 mcg/L with creatinine 2.6 cannot be interpreted against the usual threshold. It is a data point, not the verdict.
- The EEG pattern is concerning but not sufficient alone. Continuous suppression with brief bursts can be seen with sedative washout, diffuse hypoxic injury, or hypothermia. It supports a poor prognosis; it does not establish it.
- Timing. Guideline-based multimodal prognostication should occur no earlier than 72 hours after return of spontaneous circulation, in a normothermic, sedation-free patient without metabolic confounders. He is at 96 hours - timing is adequate, but the metabolic confounder is not resolved.
What to obtain:
- Bilateral median nerve SSEP (absent N20 bilaterally has the lowest false-positive rate of any single test).
- MRI with DWI at ≥72 h (diffuse cortical restriction carries a poor prognosis).
- Repeat NSE trend (rising versus falling), and a formal EEG read for reactivity.
Then integrate all modalities, require concordance, and hold the conversation with that framework.
2. The family asks a direct question: "Is he brain dead?" How do you answer, and how do you structure the conversation?
Read the worked answer
Separate two things for them: 'brain dead' is a legal determination with a specific procedure; 'what kind of recovery is possible' is a prognostic question. He is not yet declared brain dead.
How to structure it:
- Answer the question asked, plainly. "No - brain death has not been determined. That requires a formal set of tests, and we haven't completed them." Do not let a family member leave believing a legal status has been assigned.
- Explain what brain death testing requires in concrete terms: a known cause compatible with the diagnosis, normothermia, no sedatives or neuromuscular blockers, no severe metabolic derangement, and then a clinical exam plus an apnea test (or ancillary studies if the apnea test cannot be done safely).
- Address the sedative and metabolic confounders explicitly - why you cannot do an apnea test while he is on pressors with a creatinine of 2.6, and why a definitive exam requires the drugs to be fully cleared.
- Give honest prognostic framing. Nonreactive pupils, absent corneal reflexes, and no motor response at 96 hours with a suppressed EEG make a meaningful neurologic recovery very unlikely, while emphasizing that the final determination rests on the pending tests and a repeated exam.
- Elicit their values and what they fear. Ask what he would consider unacceptable, and offer a defined time frame for reconvening (e.g., after SSEP and MRI in 24-48 h) rather than an open-ended wait.
Document the conversation, the questions asked, and the plan.
3. The SSEP returns with absent bilateral N20 responses and MRI shows diffuse cortical restricted diffusion. The team wants to proceed with brain death testing. What are the practical obstacles in this patient, and how do you handle them?
Read the worked answer
The obstacles are hemodynamic and metabolic, not conceptual; address them before declaring the test invalid.
Preconditions before clinical brain death testing:
- Drug clearance. Confirm no propofol, fentanyl, or neuromuscular blocker effect. Propofol and fentanyl clear over hours, but renal failure and prolonged infusion lengthen that - use the medication record and, if uncertain, supportive evidence rather than assumption.
- Normothermia and no metabolic derangement to the extent possible. Severe hypotension, profound acidosis, or extreme hypernatremia can mimic or mask brain death criteria; his creatinine 2.6 alone is not disqualifying if other parameters are optimized, but it must be acknowledged.
- Hemodynamics. MAP 78 on norepinephrine 4 mcg/min is workable but the apnea test will drop preload and pressure. Preoxygenate to PaO2 >200 mmHg, keep the circuit disconnected at the airway, and set a clear threshold (e.g., SBP <90 mmHg) at which you abandon the test and use an ancillary study instead.
- Apnea test mechanics. Baseline PaCO2 ~40, disconnect, observe for respiratory effort, and target a rise to ≥60 mmHg (or ≥20 mmHg above baseline). If the patient becomes unstable before that endpoint, abort and obtain an ancillary study (CT angiography or radionuclide perfusion) - an aborted test is not a negative test.
- Ancillary studies are not a substitute for clinical criteria unless they are needed for safety; document why.
Handling the process properly matters as much as the conclusion - this family is watching how the determination is made.