To be included in this systematic review, the following criteria were considered: (1) studies published in English between 2006 (year of the first publication on the topic) (Owen et al., 2006) and 2021, (2) peer-reviewed observational study (i.e., cross-sectional, longitudinal, retrospective or prospective) and reviews (reviews, systematic reviews, meta-analyses), (3) human subjects aged > 18 years old, (4) diagnosed as being in a VS/UWS or MCS (based on a standardized assessment scale such as the Coma Recovery Scale–Revised), and (5) evaluated using neuroimaging and/or electrophysiological active paradigms (defined as participants instructed to mentally perform a task). All etiologies and all clinical settings were included.

This systematic review was performed in accordance with the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guideline (Liberati et al., 2009). Search terms were generated in consultation with the Powell library at the University of California Los Angeles (Supplementary material). An electronic search of published studies was performed on both PubMed and Google Scholar in January 2021. The titles and abstracts of all articles in the search were divided among members (RF, EN, RL, NL, MF, LT, A-MM, SD, VV, KI, BE, OG, MZ, FB, AT, and AE) of the International Brain Injury Association (IBIA) Disorders of Consciousness Special Interest Group (DoC SIG) and then screened. Additional articles were manually searched by cross-referencing using the “cited by” function as well as by reviewing the reference section of the selected papers. Relevant articles from this initial screening were then gathered by CS. The same IBIA DoC SIG members were then asked to screen selected manuscripts and, if relevant, extract the following information: 1) are there specific name(s) used to designate the target clinical entity?, 2) if yes, which name(s)?, 3) is there any rationale for using such name mentioned in the text?

Descriptive statistical analyses (i.e., measures of frequency such as count, percent) and qualitative analyses were used in this study using Microsoft Excel Spreadsheet Software.

The original article from Owen et al. (2006) described the phenomenon without providing a name (“…some noncommunicative patients, including those diagnosed as vegetative, minimally conscious, or locked in, may be able to use their residual cognitive capabilities to communicate their thoughts to those around them by modulating their own neural activity”). However, in an article published the year after, Owen and coworkers used, for the first time, the term “covert awareness” (Owen et al., 2007). The name was nevertheless only mentioned in the title and no explicit definition or explanation for using this term was provided in the text. Since then, other terms including “covert” (the most frequent being “covert consciousness” and “covert cognition”) have been used interchangeably to designate the presence of awareness, consciousness, top-down cognitive processing, volition, command-following or communication using task-based fMRI or EEG paradigms, in the absence of any behavioral signs of consciousness. An explicit rationale to use “covert awareness” or any of the “covert” related terms were not found in the articles published by the same authors and included in our search. The use of these terms seems therefore primarily descriptive.

In 2009, Giacino and coworkers first introduced the term “functional LIS” observing that “… commonly held notions about brain–behavior relationships should be revisited in this patient population…they clearly illustrate the wide discrepancy that may exist between observable behavior and the underlying neurophysiologic processes believed to support cognitive processing. Such findings also force us to consider the unsettling possibility that cognitive function may be at least partially preserved in this case, but lack a mode of expression as the consequence of severely dysfunctional sensory and motor systems. In a sense, these findings may reflect a “functional” LIS…” (Giacino et al., 2009). In 2011, Bruno and coworkers more formally suggested the use of such term arguing that it emphasizes the dissociation between these patients' motor dysfunctions and their preserved higher cognitive functions as shown by functional imaging techniques (Bruno et al., 2011). In 2012, Laureys and Schiff specified that “…this designation should be reserved, however, for patients who show consistent and reliable communication using non-speech and non-gestural communication through direct brain signaling” (Laureys and Schiff, 2012).

The last publication emphasizes the importance of demonstrating communication in order to use such a term (and not only command-following and/or higher cognitive functions as Giacino and Bruno refers to). It nevertheless also complicates the use of such diagnosis since it requires showing consistent communication. This implies serial assessments using neuroimaging or electrophysiology, which, currently, might be practically complicated for clinicians. Moreover, considering the uncertainty about the full extent of residual cognitive function in such patients, using the term “LIS” which is characterized by relatively preserved cognition has been criticized (Owen, 2015; Schiff, 2015). Another criticism addresses the underlying neuropathology since LIS patients typically demonstrate specific lesions to the ventral pons which might not be the case in this population (Owen, 2015). Formisano and colleagues have been arguing in favor of using “functional” LIS as it underlines the distinction from “classical” or “complete” LIS and underlines the presence of a potential functional disconnection syndrome, as it is the case in diffuse axonal injury with associated supratentorial lesions (Formisano et al., 2013).

In 2014, Gosseries and coworkers criticized the use of “functional LIS” since the LIS is not a DoC, and since it could be confusing as well as a misnomer to use this term. Instead, the authors suggested that the use of non-behavioral MCS (or MCS^*) is a more clinically accurate alternative as it is more descriptive and more consistent with other MCS terminology (i.e., MCS+ and MCS-) (Gosseries et al., 2014). In cases of dissociation between behavioral and neuroimaging testing, “…especially in the case of patients who are diagnosed as being UWS by bedside testing but then diagnosed MCS with neuroimaging techniques,” the use of MCS^* might therefore be a better description of the phenomenon (Gosseries et al., 2014). According to the authors, the term MCS^* can designate both VS/UWS patients who respond to active paradigms using neuroimaging and VS/UWS patients who demonstrate a brain activity at rest that is more compatible with MCS patients. One could wonder if distinct names should be given to each case scenario since they might reflect different levels of cognitive functioning and therefore different types of patients. Such a diagnosis (MCS^*) would also only apply to VS/UWS that respond to active paradigms since the authors mention that “…If patients in MCS- show command following during ancillary testing, they could be diagnosed as in MCS+^*.” Nevertheless, based on our search, such term (MCS+^*) has not been used afterwards in the literature. Moreover, the term non-behavioral MCS has been criticized since it might not adequately describe the patients' residual cognitive abilities, because responses to complex active paradigms require preserved attention, language comprehension, and working memory which is currently assumed as being severely altered in MCS (Owen, 2015).

Following the publication of Fernández-Espejo et al. (2015), who showed structural disconnection between thalamus and primary motor cortex (potentially leading to a deficit in motor expression) in one “covertly aware” patient (Fernández-Espejo et al., 2015), Schiff and coworkers wrote an editorial on these results and suggested the use of the term “Cognitive motor dissociation” (CMD) to account for “… the sharp dissociation of a retained but unrecognized (covert) cognitive capacity in some severely brain-injured patients with non-purposeful or absent behavioral responses” (as per the author, including patients in VS/UWS, MCS- or complete LIS who would demonstrate such response) (Schiff, 2015). This term constitutes a neutral description of the phenomenon and aims at being a general umbrella under which specific types of responses could be subcategorized (e.g., functional LIS, MCS^*). However, some clarifications are needed regarding its definition since “non-purposeful or absent behavioral responses” can paradoxically not include MCS-. The inclusion of complete LIS might actually lead such a diagnosis to be applied to different types of pathologies and not to be specific to our population (such as amyotrophic lateral sclerosis or patients with severe Guillain Barrè syndrome which may present CMD in an advanced stage of the disease) (Zasler et al., 2019).

More recently, Edlow and coworkers also suggested the use of higher-order cortex motor dissociation (HMD), which they define as “… functional MRI and EEG responses within association cortex (e.g., Wernicke's area) during passive language or music stimuli despite absence of behavioral evidence of language” (Edlow et al., 2017). In this case scenario, patients who did not show any behavioral signs of receptive language but did show a relevant high-order cortical activation of the language network using a passive paradigm (i.e., passive exposure to a stimulus without instructing the patient to perform a task) might be in a HMD. The authors specified that this term should not be used for designating covert awareness (and so is distinct from previous terms such as CMD) but to label dissociation between neuroimaging and behavioral findings detected early on in the recovery process (e.g., in the intensive care unit).

Even though this systematic review leads us to consider a more confined list of names for diagnosing patients with covert awareness, it is clear that each term currently conveys confusion in terms of 1) the population it addresses (i.e., VS/UWS, MCS-, cLIS) and/or 2) the profile of responses it covers (i.e., command-following, communication and/or resting brain activity). Indeed, MCS^* seems to only include VS/UWS patients (Gosseries et al., 2014), while there is no clear mention of what bedside DoC diagnosis would functional LIS include (Giacino et al., 2009; Bruno et al., 2011; Laureys and Schiff, 2012). Additionally, functional LIS might designate patients either who only show “consistent communication” (based on Laureys and Schiff, 2012) or “higher cognitive functions” (based on Giacino et al., 2009; Bruno et al., 2011), while MCS^* might designate VS/UWS patients who respond to an active paradigm or who show resting state brain activity reflecting consciousness (Gosseries et al., 2014). A last dimension that might confuse clinicians is the method used to detect the phenomenon. Indeed, some authors mention only neuroimaging while others also include electrophysiology (Gosseries et al., 2014; Edlow et al., 2017). It is most likely that the method used would not lead to distinct terms but that should be clarified in the future. It should also be decided in the future whether covert awareness (as well as other similar terms such as covert consciousness and covert cognition) should have its own clear definition and should be used as a diagnosis. Beyond the need for clarity that was highlighted for each of these names, there is a need to decide whether one or several terms should be used for diagnosis or whether one umbrella term with associated subcategories should be used to fit specific patients' profile (e.g., higher cognition function vs. communications). Of course, this study has several limitations. Only two search engines were used for our systematic review (Pubmed and Google Scholar). Our systematic review was also limited to peer-reviewed publications and therefore excluded book chapters and dissertations. Only manuscripts in English were included which might have excluded names used in other languages. Finally, we focused our search on active paradigms and did therefore not include resting state paradigms or techniques such as Positron Emission Tomography (PET) or Transcranial Magnetic Stimulation (TMS or TMS-EEG), that might have been performed in that population. Future review of the literature could be more exhaustive and might lead to additional terms that were overlooked in this study.