Prognosis of Coma After Cardiac Arrest

Introduction to Prognostication

The American Heart Association recommends that prognostication of coma after cardiac arrest should not be attempted before 72 hours after return of spontaneous circulation, as the prognosis is poor overall, with many patients dying, remaining permanently unresponsive, or unable to perform independent activities 1, 2

No clinical neurological signs or diagnostic studies reliably predict poor outcome during the first 24 hours after return of spontaneous circulation (ROSC) 1, 3
Neurologic prognosis remains uncertain before at least 72 hours after ROSC 4
Decisions to limit or withdraw life-sustaining treatment should never rely on a single prognostication element 4, 5

Absence of both pupillary light and corneal reflexes at 72 hours predicts poor outcome with high reliability (false-positive rate [FPR] 0%, 95% CI 0-9%) 1, 6, 7
Bilateral absence of N20 cortical response to median nerve somatosensory evoked potentials (SSEPs) after 24 hours predicts poor outcome with high reliability (FPR 0%, 95% CI 0-3%) 1, 3
Glasgow Coma Scale (GCS) score ≤5 at 72 hours (FPR 0%, 95% CI 0-6%) 1, 7
Motor response with GCS motor score ≤2 at 72 hours has a FPR of 5% (95% CI 2-9%) 7

EEG findings within 24 hours after ROSC are not reliable predictors of outcome 1, 3
In normothermic patients without confounders, the following EEG patterns are associated with poor outcome (FPR 3%, 95% CI 0.9-11%): generalized suppression to <20 μV, burst-suppression pattern with generalized epileptic activity, diffuse periodic complexes on a flat background 3
One week after cardiac arrest, specific EEG findings may be useful for predicting poor outcomes 3

MRI and CT are the most studied neuroimaging modalities 3
Extensive cortical and subcortical lesions on MRI are associated with poor neurological outcome 3
Despite potential, neuroimaging has not yet been proven as an independently accurate modality for prediction of outcome 3

Wait at least 72 hours after ROSC before making definitive prognostic assessments 1, 4
Ensure absence of confounding factors (sedatives, hypotension, hypothermia, neuromuscular blockade, seizures, hypoxemia) 1, 3
Perform thorough neurological examination focusing on brainstem reflexes 1
Obtain electrophysiological studies (EEG and SSEPs) 3
Consider neuroimaging (MRI or CT) 3
Interpret biomarkers cautiously, especially in patients treated with hypothermia 7

Making prognostic decisions too early (<72 hours after ROSC) 1, 4
Relying on a single prognostic element rather than a multimodal approach 4, 5
Using biomarkers as sole predictors of outcome 7
Not considering confounding factors that may affect clinical examination findings 1, 3