Ethical issues of nosology in disorders of consciousness

1 Introduction

Consciousness is a unique and ineffable phenomenon with two clinical dimensions: arousal (wakefulness) and awareness of self and environment. In recent decades, consensus has emerged around three principal syndromic categories of severe disorders of consciousness (DoC): coma, characterized by unarousable unresponsiveness; the vegetative state, also called the unresponsive wakefulness syndrome (VS/UWS), in which patients have wakefulness without awareness; and the minimally conscious state (MCS), in which patients have wakefulness but limited or inconsistent awareness (Giacino et al., 2014). Some authors have subdivided these general categories, for example, into MCS+ and MCS-, based on the presence of language response (Bruno et al., 2011). Others have advocated for terms spanning existing categories, such as cognitive-motor dissociation, to describe patients who appear unresponsive but possess covert consciousness (Schiff, 2015). A few non-DoC conditions may be mistaken for a DoC by an unwary examiner: the locked-in syndrome in which patients cannot motorically demonstrate their intact awareness because of profound paralysis; and brain death, the irreversible cessation of all brain functions. Less severe DoC include delirium and metabolic/toxic encephalopathies. The history and evolution of DoC nosology has been documented elsewhere (Bernat and Fins, 2023).

During the last two decades, technological innovations, clinical experience, and improved understanding of brain networks have raised challenges to the traditional nosology of severe DoC. First was the critical discovery that in some unresponsive patients diagnosed as VS/UWS by clinical examination, assessments of brain activity using functional magnetic resonance imaging (fMRI), ¹⁸F-fluorodeoxyglucose positron emission tomography (FDG-PET), or quantitative electroencephalography (qEEG) revealed evidence of awareness and, in some, the ability to communicate by thought patterns (Monti et al., 2010). Yet, there is no consensus on how to best categorize such patients with covert consciousness who fulfill the clinical criteria for VS/UWS, yet clearly retain awareness by functional testing.

Second, a mismatch has arisen between the current definitions of severe DoC and medical practitioners’ ability to diagnose these disorders. This dissociation inevitably results from limitations in an examiner’s ability to correctly ascertain the internal experience of conscious awareness, and from the fact that levels of consciousness comprise a continuum that cannot be neatly divided into discrete categories (Fischer and Truog, 2017). This mismatch raises several ethical consequences that threaten to undermine public trust in clinicians. Finally, syndromic categories of severe DoC are diagnosed without accounting for the longitudinal course of an illness or its underlying pathophysiology. The imprecision inherent to the current nosologic system hampers communication with family, research on these disorders, and prognostic accuracy. Here we analyze the ethical issues raised by inadequate nosology and conclude with a strategy to amend DoC nosology that mitigates these challenges.

2 Ethical issues associated with diagnosing DoC: The problem of covert consciousness

The traditional nosology of severe DoC relies on findings of neurological examination. However, the bedside clinical examination for awareness is utterly dependent on the patient’s motoric behavioral responses because another person’s conscious awareness only can be inferred (Bernat, 2017). The failure to respond to stimuli does not necessarily point to a DoC. A patient may lack the ability to respond to a stimulus because of factors unrelated to consciousness, such as paralysis, impaired vision or hearing, aphasia, or neglect. Even well-validated scales, such as the Coma Recovery Scale–Revised (CRS-R), rely solely on the careful observation of the patient’s motor responses to standardized stimuli (Giacino et al., 2014), which are subject to many confounding factors (Pincherle et al., 2021).

Evidence that has been accumulated over the past two decades shows that a significant minority of patients who appear unresponsive when observed for motor activity will demonstrate evidence of responsiveness based on patterns of volitional brain activity when tested with ancillary tests that directly target neural activity, rather than overt motor behaviors (Giacino et al., 2018). For example, using FDG-PET or fMRI in one cohort, investigators found up to 30% of patients with clinical VS/UWS had neural activity evidence of covert consciousness (Stender et al., 2014). Other investigators reported a 15% prevalence of covert consciousness (Kondziella et al., 2015).

How to clinically classify such patients remains an unresolved question. Validated diagnostic methods, based on motor responses during beside examination, classify patients without discernible behavioral responses as VS/UWS. Yet the definition and the philosophical underpinnings of current nosology are based in consciousness, so a patient who reliably demonstrates some degree of awareness is undoubtedly conscious.

As technology progresses, it is inevitable that new techniques will be able to interrogate brain processes not detectable on a motor-response-based bedside examination. Some authors have advocated for the creation of a unique category to describe clinically unresponsive patients with covert consciousness called cognitive-motor dissociation (Schiff, 2015). However, this categorization has been criticized because all patients with awareness should receive the same treatment, whether they exhibit motor responses or only equivalent neural responses (Bayne et al., 2017). In other words, a disturbance of consciousness is the fundamental feature defining a DoC –not evidence of the patient’s behavioral response. The demonstration of a neural response, despite the absence of a motor response, is sufficient grounds to establish awareness.

Additionally, adding an intermediate category between VS/UWS and MCS to describe patients with covert consciousness introduces ethical and practical challenges. First, fMRI, FDG-PET, and qEEG, and the technical expertise to perform these protocols, are not widely available, so the universal application of these techniques is currently impractical. Should all patients who appears to have VS/UWS be transferred to centers with these capabilities? Such a change in practice would carry substantial costs and may not be defensible from a distributive justice standpoint.

Even if that challenge could be overcome, others remain. Although covert consciousness can be detected in some patients who appear to be vegetative or in coma, current paradigms that use qEEG, fMRI and FDG-PET are imperfectly sensitive. Many patients with clinical behavioral evidence of responsiveness lack demonstrable fMRI or EEG brain responsiveness to the paradigms used to detect covert consciousness (Edlow et al., 2017). If these techniques produce false-negative determinations (e.g., failure to detect consciousness when it is present), they cannot be used to reliably exclude covert consciousness. Thus, both bedside motor responsiveness and ancillary neural testing paradigms can produce false-negative results by failing to detect consciousness, even when consciousness is preserved.

Further, as technology advances, additional techniques to probe awareness and responsiveness undoubtedly will emerge. The inexorable march of technological progress logically means that clinicians not using the most advanced diagnostic tools will forever be at risk of underdiagnosing patients with covert consciousness. Current fMRI, FDG-PET, and qEEG studies constitute only the leading edge of technologies that can probe for neural responsiveness and covert consciousness.

Thus, no technique can perfectly identify responsiveness, let alone conscious awareness. The bedside examination can overlook patients with covert consciousness, and ancillary testing techniques can fail to identify consciousness detected by bedside examination. Both methods likely miss responsiveness that may be detectable by a future, as yet undiscovered, technology. So how can clinicians be confident in their ability to accurately classify DoC?

How to optimally classify a patient with covert consciousness is not simply a semantic question. Family members and medical practitioners will make treatment decisions based on their perception of a patient’s awareness. For example, a patient with awareness ethically requires treatment with anxiolytics and pain medications, which could worsen prognosis, while a truly unconscious patient presumably cannot feel pain or anxiety. Also, decisions to continue life-sustaining therapy hinge on accurate prognostication. Patients with VS/UWS generally have poorer prognosis for improvement than patients with MCS. Patients who appear to have UWS, but actually possess covert consciousness, have a better prognosis than those without covert consciousness (Edlow et al., 2021; Claassen et al., 2019).

Detecting covert consciousness is not solely a matter of distinguishing between VS/UWS and MCS. Covert, subclinical responsiveness has also been found in patients thought to be in coma according to a bedside behavioral assessment (Claassen et al., 2019). Although it remains unclear how many patients who appear to be comatose have evidence of covert consciousness, this finding challenges the generally held belief that wakefulness is a necessary precondition to awareness. This finding raises two additional questions. First, should DoC be reclassified based on two axes –arousal and awareness –rather than on a continuum? Second, is it defensible to diagnose DoC, such as coma, solely on bedside examination findings? If treatment decisions, such as withdrawal of life-sustaining therapy, hinge on diagnosis, a false-positive determination (i.e., diagnosing coma in a patient with covert consciousness) is unacceptable. However, it remains uncertain how best to minimize false-positive determinations. Should all patients with severe DoC be required to undergo ancillary testing?

3 Ethical issues associated with diagnosing DoC: Mismatch between diagnostic methods and current syndromic definitions

The dissociation between the definition of DoC syndromes (based on consciousness), and the diagnosis of these syndromes (based on observable responses) raises several challenges. First, the lay public relies on the medical profession for guidance. When a physician proclaims to a patient’s family that their loved one is in VS/UWS, the family should be able to trust that physician’s statement. Yet, in 15–33% of cases, the physician cannot judge the patient’s responsiveness –let alone awareness or consciousness –merely by bedside motor examination. And the false-positive rate of VS/UWS diagnosis continues to be 30–40% (Schnakers et al., 2009; Wang et al., 2020). The mismatch between diagnostic practice and current DoC nosology based on consciousness threatens public trust in physicians, their diagnostic and prognostic abilities, and the foundation of DoC nosology.

Two broad strategies could better align the definitions of DoC with their diagnoses. One solution is to change the diagnostic methods. Because it is impossible to ever directly perceive or experience another person’s conscious awareness, the best clinicians can do is to probe for responsiveness and assume awareness based on the patient’s response to stimulation. Current diagnostic methods are limited in sensitivity. Therefore, to properly align criteria and preserve current nosology, diagnostic tests that can better discriminate responsiveness must be developed.

The second strategy to better align DoC definitions and diagnoses is to alter the definitions. One solution would be to refine categories to capture the fact that only responsiveness is being assessed, not awareness. In such a nosologic system, coma and VS/UWS would be reclassified as unresponsive syndromes. They could be further refined to account for the method used to interrogate responsiveness. For example, a patient could be “unresponsive by motor examination,” clarifying that the lack of responsiveness is based on motor response, or “unresponsive by fMRI, FDG-PET, or qEEG.” Severe DoC categories could be further refined by adding qualifiers to account for the possibly low sensitivity of current diagnostic tests that measure responsiveness. For example, the VS/UWS label could be retained, but renamed or qualified as “wakeful, with responsiveness not yet observed.”

This revised nosology would have the benefit of improving the fidelity of communication but remain limited by its imprecise nature. Surrogate decision-makers are ultimately concerned with whether their loved one is conscious, not whether the examining physicians have observed responsiveness. Nevertheless, such a revised nosology is more intellectually honest than the current system, which does not account for the mismatch between definition and diagnostic practice. Later, we offer a proposal for a revised DoC nosology to address these deficiencies.

4 Ethical issues associated with prognostication in DoC: Consequences of diagnostic inaccuracy

Current nosology is limited to several broad discrete categories. But, this classification erects definitional boundaries between syndromes that are absent in biology because consciousness is a continuum with varying depths to responsiveness and arousal (Fischer and Truog, 2017). Further, each patient’s illness pursues a different clinical course before arriving at a DoC diagnosis. Syndromic diagnoses are useful only if they can help guide better medical care. For example, making a syndromic diagnosis should allow for better treatment, better research, more accurate prognostication, or some other action that is validly based on the diagnosis. When a syndrome is too broad and encompasses heterogeneous conditions, it fails in these goals.

By ignoring the mechanism of brain injury, current DoC nosology fails in the task of accurate prognostication. Decades ago, clinicians recognized that patients with PVS/UWS from traumatic brain injury generally had a better prognosis than patients with PVS/UWS due to neuronal hypoxic-ischemic injury (Multi-Society Task Force on PVS, 1994). Nevertheless, studies on prognosis of patients with DoC can conflate mechanisms of injury because they are not part of the diagnosis. It therefore becomes difficult for treating physicians to know how to guide treatment discussions with surrogate decision-makers, given the dearth of information that incorporates both the severity of the DoC and the mechanism of injury (Bernat, 2017).

Many published studies of prognosis in DoC use syndromic diagnoses as categories. Even when studies separate traumatic and non-traumatic mechanisms of injury, there is too much heterogeneity within the non-traumatic category to yield useful prognostic information. Indeed, outcome studies of both VS/UWS and MCS are contaminated by the heterogeneity of patients within syndromic diagnostic categories. Patients diagnosed as VS/UWS or MCS because they share similar examination features often have completely different underlying pathologies (e.g., trauma, hypoxic-ischemic neuronal injury, stroke, meningoencephalitis), distribution of brain lesions, ages, and co-morbidities. It is obvious that prognosis must be individualized for each patient and not simply pronounced based on their membership in a diagnostic clinical syndrome.

Studies of outcomes in series of patients with traumatic brain injuries more accurately depict their prognosis than studies on outcomes in syndromes of VS/UWS or MCS (Bernat, 2016). Paradoxically, current DoC categories and practices could even undermine ethical decision-making if physicians and caregivers fallaciously overestimated poor prognosis based merely on the label of impaired consciousness. Studies have shown that patients who are perceived to have a poorer prognosis often receive less aggressive medical care (Murray et al., 1993). Current labels for severe DoC connote poor prognosis, and these labels are sometimes used as a justification by physicians to advocate a decision not to treat.

The “self-fulfilling prophecy” describes a situation in which patients with a perceived poor prognosis undergo premature de-escalation or termination of treatment which assures their demise (Bernat, 2004;Wilkinson, 2009). Prognostic studies of patients with DoC are plagued by this bias, which has severely limited research into the natural history of these disorders. Early withdrawal of life-sustaining therapy because of the presence of a DoC recently was found to be a key factor accounting for the poor prognosis among patients with large intracerebral hemorrhages (Alkhachroum et al., 2021).

A more sophisticated nosology of DoC could include the mechanism of injury and temporal course. Current nosology is based on a cross-sectional assessment of consciousness but does not describe the clinical course of the injury or illness. Patients with DoC often fluctuate across categorical boundaries. A revised DoC nosology that incorporates the mechanism of injury and its temporal trend would yield more actionable treatment information, allow physicians to prognosticate more accurately, and provide greater homogeneity for research studies.

To highlight the shortcomings of current DoC nosology, let us compare it to that of the clinical syndrome of dementia. The current DSM-5 recharacterized dementia as “major neurocognitive disorder.” This descriptive term correctly identifies a patient with a decline in cognitive abilities that is severe enough to interfere with independence in daily life. Such a term can be helpful for certain goals, such as coding or billing, where the cause and trajectory of the neurocognitive disorder is unimportant. But this broad term is obviously inadequate for prognostication, communication, or research. Is the “major neurocognitive disorder” due to a static process or a neurodegenerative one? Is there an underlying autoimmune encephalopathy or vitamin deficiency which may be reversible? Has the patient been stable for 15 years after the last brain infarction, or is it rapidly progressive prion disease? Is it early dementia or end-stage dementia? It is common to add modifiers such as severity or time-course when discussing dementia (e.g., chronic, rapidly progressive) or to append information about the underlying pathology (e.g., Lewy body dementia). Factors such as time course and underlying pathology obviously influence prognosis and research and are vital to communication.

Similar to calling dementia “major neurocognitive disorder,” severe DoC labels such as MCS can be useful for certain general applications, such as communicating with an insurance company to justify prolonged services for a family meeting about goals of care. But such a broad category is less useful and may be counterproductive for other purposes such as research, communication with families or other clinicians, and prognostication. Given the complexity of brain disorders, one cannot capture the breadth of DoC or the uniqueness of an individual patient in a few broad syndromic diagnoses, just as major neurocognitive disorders cannot describe the breadth of dementias. An improved DoC nosology should incorporate underlying mechanisms and other important clinical factors into the diagnostic categories.

Treatment ethical issues also are related to nosology and classification. Irrespective of the confidence clinicians may have that a DoC patient is utterly unaware, clinicians should assume the patient retains awareness when speaking about the patient’s condition in the patient’s presence and in reaching decisions on palliative treatment. Many experienced clinicians talk to DoC patients and try to reassure them even though the clinicians suspect the patients are unaware (La Puma, 1988). Failing to detect awareness is a serious clinical error but a worse one is talking about a patient as if he were unconscious in the patient’s presence when the patient is aware. Similarly, it is good medical practice to err on the side of providing palliative treatment for all DoC patients in case they retain a capacity for suffering (Fins and Bernat, 2018).

5 Proposal for a revised DoC nosology

Current DoC nosology has several shortcomings: 1) it fails to account for covert consciousness; 2) it is categorized by the level of consciousness but diagnosed based on motoric behavioral responses; and 3) it is limited by broad syndromic definitions that do not account for pathophysiologic mechanism or temporal course. These shortcomings are not merely semantic, because serious defects in nosology can undermine prognostication, treatment, research, and communication with surrogates and others.

A revised DoC nosology should address these shortcomings, and it should also acknowledge the inherent limitations that arise when a clinician tries to assess the awareness and psychological interiority of another person. A revised nosology could be based on consciousness, but it must rely on descriptors of responsiveness to remain intellectually honest. A successfully revised nosology should also include the mechanism of brain injury, which ultimately influences prognosis. We believe a revised nosology should also acknowledge the indisputable fact that covert consciousness could still be present, even when testing protocols fail to detect neural responsiveness.

Our revised nosology incorporates three major changes. First, similar to the current system, syndromes should be described based on wakefulness and responsiveness.

Second, in addition to the syndromic diagnoses, an appended modifier should describe how the syndrome was diagnosed. Examples include “arousable unresponsiveness by motor exam” or “arousable unresponsiveness by PET/fMRI/EEG.” The former name communicates a purely clinical diagnosis in which no ancillary tests were used, and both names admit the possibility that responsiveness might be detected with future technological developments.

Third, a further modifier should be added to describe the mechanism of injury. Examples of diagnoses with this revised nosology would include “unarousable responsiveness by MRI and EEG, from traumatic brain injury,” or “unarousable, responsive by EEG, from encephalitis.” It would also be possible to add modifiers in cases where fMRI or FDG-PET is not available, such as “UUS, not yet tested by fMRI or PET, from anoxic injury.” Finally, a temporal modifier could be added to describe the course, as worsening, stable, or improving.

Such a revised nosologic system would provide features that the current nosology system lacks. It would allow clinicians to tailor discussions with family members regarding prognosis and appropriate treatment. It would permit research into the natural history of DoC and allow studies of individual conditions to be less contaminated by the inclusion of patients with unrelated conditions that have similar clinical findings. This system also mitigates the ethical challenges posed by the mismatch between DoC definition and DoC diagnosis, by renaming categories based on responsiveness, not consciousness.

The nosology of clinical syndromes requires refinement into dimensions of greater specificity because new technologies have opened vistas of insight that had been previously closed. In much the same way that specific DNA testing has transformed neurogenetic disease classification and nosology from phenotype-based to genotype-based, the availability of innovative methods of assessing conscious awareness must change DoC classification and nosology. The effort to refine nosology provides an opportunity to increase specificity for the purposes of treatment decisions, prognostic accuracy, communication clarity, and research diagnostic uniformity. The new nosology paradigm is forward-thinking because it allows future technological developments to be incorporated. Both DoC patient care and research could benefit from the new nosology paradigm.

Conflict of interest

The authors report no conflicts of interest related to this manuscript.