Cohort Study: One in Four Patients Demonstrates Covert Cognition Despite Behavioral Unresponsiveness

Background: Cognitive motor dissociation (CMD) refers to the presence of specific neuroimaging or electrophysiological responses to commands in patients otherwise incapable of voluntary behavioral output. Detecting CMD is clinically relevant because its underdiagnosis may lead to premature decisions regarding goals of care, life-sustaining treatment, and rehabilitation efforts. Although several single-center studies have suggested that CMD may exist in 10–20% of patients with disorders of consciousness, larger multinational data were lacking, particularly using both functional magnetic resonance imaging (fMRI) and electroencephalography (EEG).

Objective: To determine how often CMD occurs in a large, multinational cohort of adults with impaired consciousness and to evaluate the clinical variables potentially associated with this phenomenon.

Methods: This prospective cohort study included 353 adults with disorders of consciousness recruited from six international centers between 2006 and 2023. Enrolled participants had at least one behavioral assessment using the Coma Recovery Scale–Revised (CRS-R) and underwent task-based fMRI, EEG, or both. Sites utilized validated analytic pipelines and automated data processing to minimize false positives. Participants were divided into two groups: those without observable responses to verbal commands (coma, vegetative state, or minimally conscious state–minus) and those with observable responses (minimally conscious state–plus or emerged). CMD was defined as the absence of any observable behavioral response to commands, combined with a positive command-following signal on fMRI or EEG.

Results: Among 241 participants with no overt command-following, 25% showed CMD through either fMRI alone, EEG alone, or both. CMD was more common in younger patients, those assessed later after injury, and those with traumatic brain injury. Interestingly, in 112 participants who did exhibit command-following on bedside exams, only 38% demonstrated confirmatory responses on fMRI or EEG. These findings support the notion that the tasks used for neuroimaging and electrophysiological assessments may require more sustained cognitive engagement than typical bedside evaluations.

Conclusions: CMD was identified in about one in four patients who lacked behavioral command-following. Combining fMRI with EEG likely increases detection rates compared to either modality alone. The results highlight the need for increased awareness of covert cognitive activity in this population, given potential ramifications for prognosis, family counseling, and clinical care.

Implications for Practice: Clinicians should consider the possibility of CMD in patients who appear unresponsive at the bedside. When feasible, employing both fMRI and EEG might reveal hidden cognitive capacities that can guide patient-centered decisions, encourage targeted therapies, and allow healthcare teams to respect potential consciousness and autonomy. However, such technologies remain limited to specialized centers.

Study Strengths and Limitations: Strengths include a diverse sample from multiple international sites and the integration of two complementary neurodiagnostic techniques. Limitations involve heterogeneous recruitment practices, variations in local data acquisition methods, and potential selection biases toward patients who survived until advanced testing was available. Additionally, the absence of standardized paradigms across sites reduced consistency of results.

Future Research: Further large-scale investigations should standardize fMRI and EEG protocols and determine whether earlier and more consistent identification of CMD affects functional outcomes. Efforts to refine and validate automated analytic pipelines could facilitate widespread adoption of these techniques in routine clinical settings.

Reference: Bodien YG, Allanson J, Cardone P, et al. Cognitive Motor Dissociation in Disorders of Consciousness. New England Journal of Medicine. 2024;391:598-608. DOI: http://doi.org/10.1056/NEJMoa2400645