Prognostication in comatose survivors of cardiac arrest

Introduction: Hypoxic-ischemic encephalopathy (HIE) is the leading cause of death in comatose patients after cardiac arrest resuscitation. Poor neurological outcome is defined as death from neurological cause, persistent vegetative state, or severe neurological disability which is predicted in these patients by assessing the severity of HIE. Background: The most commonly used indicators of severe HIE include a bilateral absence of corneal and pupillary reflexes, bilateral absence of N2O waves of short-latency somatosensory evoked potentials, high blood concentrations of neuron-specific enolase, unfavorable patterns on electroencephalogram, and signs of diffuse HIE on computed tomography or magnetic resonance imaging of the brain. Current guidelines recommend performing prognostication no earlier than 72 hours after return of spontaneous circulation in all comatose patients with an absent or extensor motor response to pain, after having excluded confounders such as residual sedation that may interfere with clinical examination. A multimodal approach combining multiple prognostication tests are recommended so that the risk of a false prediction can be minimized. Materials: Neuroprognostication is vital and yet continues to be one of the most controversial topics in post-resuscitation care. Specifically, concerning HIE, the 2006 practice parameters of the American Academy of Neurology provide specific recommendations for the prognostication of neurologic outcomes for cardiac arrest survivors not treated with therapeutic hypothermia (TH). To date, there is no adequate paradigm for prognostication in HIE treated with TH. Clinical examination including the presence or absence of brainstem reflexes, motor responses and absence of myoclonus were traditionally used to predict a favorable prognosis. Electrophysiologic testing in the form of somatosensory evoked potentials (SSEP), the serum biomarker neuron-specific enolase (NSE), as well as neuroimaging, have been employed as additional tests to attempt to improve the predictive accuracy of neuroprognostication. However, what limited certainty these tests and parameters provided has become even more questionable in the setting of therapeutic hypothermia. The use of sedatives and analgesics adds a degree of uncertainty given unpredictable drug effects on patients’ neurologic status. EEG, SSEP are the most common electrophysiological modalities utilized in neuroprognostication. EEG has been evaluated in the prognostication of cardiac arrest survivors and has also led to some essential clinical discoveries. The 2006 American Academy of Neurology (AAN) practice parameters assign EEG a false-positive rate (FPR) of 3% with a CI of 0.9; making it the least predictive method to determine neurologic outcomes. Abend et al., pooled four existing studies on EEG in cardiac arrest (CA) patients who had undergone therapeutic hypothermia and found that 29% of these patients had acute electrographic non-convulsive status epilepticus (NCSE). Conclusion: There is no good evidence from well-designed studies to support substantial accuracy of early prognostication ( < 72 hours post-arrest) in cardiac arrest survivors treated with therapeutic hypothermia. Given our lack of understanding of how therapeutic hypothermia improves outcomes, as well as its effects on emergence from the coma and its well-described effects in altering drug metabolism and clearance, it is prudent to be more conservative in approaching prognostication. Patients should be observed for a minimum of 72 hours post-arrest. However, 5-7 or more days of observation may be necessary to fully account for the effects of therapeutic hypothermia.