In conditions of prognostic uncertainty, neurointensivists and families often struggle to make decisions for unresponsive patients with brain injuries. However, a new study demonstrates that after brain injury electrophysiological assessment (electroencephalography [EEG]) has potential prognostic value. 1 This could have profound implications for clinically and ethically consequential decisions, such as withdrawing life-sustaining treatment or managing potentially life-threatening events such as sepsis. Here, we outline 2 interrelated ethical considerations for neurointensivists who use these methods when communicating prognostic information to patients' families.
Discovering Covert Consciousness in the Intensive Care UnitA 2019 study of 104 seemingly unconscious patients with brain injuries cared for in a neuro-intensive care unit (ICU) demonstrated that 16 patients responded to the command to open and close their hand. 1 Researchers did not observe this hand movement at the bedside. Instead, they witnessed EEG evidence of motor cortex activation.A machine learning algorithm identified latent evidence of brain responsiveness from multiple EEG test results. Six days after injury, half of the 16 patients with responsive brain activity recovered the ability to behaviorally follow commands, and follow-up showed that 7 patients recovered functional independence. Compared with patients who showed no evidence of motor cortex activation and recovered (12 of 84 [14%]), those with brain responsiveness recovered faster and with greater odds of a positive outcome.These results are similar to those of previous studies reporting brain activity in unresponsive patients with brain injuries. 2 Indeed, a significant proportion of unresponsive patients, estimated to be at least 15%, are able to modulate their brain activity to command. 3 These patients are correctly regarded as unconscious according to clinical criteria, yet their brain activity shows that they are aware. Such patients are variously described as having cognitive-motor dissociation or as being covertly conscious. Their consciousness is hidden until we discover it by measuring their brain activity.These findings have important implications for neurocritical care. The novel combination of machine learning with EEG can improve prognostic accuracy. The feasibility of these methods also suggests that they are likely to be used. In the days following brain injury, EEG is more practical for ICU patients compared with other costly and complex neuroimaging methods. 4 Despite the promise of these methods, little is known about how families will use test results to make VIEWPOINT