In hospital and emergency neurology, the clinical analysis of unresponsive and comatose patients becomes a practical necessity. There is always an urgent need to determine the underlying disease and the direction in which it is evolving in order to protect the brain against more serious or irreversible damage. When called upon, the physician must therefore be prepared to implement a rapid, systematic investigation of the comatose patient and prompt therapeutic and diagnostic action that allows little time for deliberate, leisurely investigation.

Some idea of the dimensions of the problem of coma can be obtained from published statistics. Eighty years ago, in two large municipal hospitals, it was estimated that 3 percent of all admissions to the emergency wards were for diseases that had caused coma. Alcoholism, cerebral trauma, and cerebrovascular diseases were the most common, accounting for 82 percent of the comatose patients admitted to the Boston City Hospital (Solomon and Aring). Epilepsy, drug intoxication, diabetes, and severe infections were the other major causes for admission. It is perhaps surprising to learn that more contemporary figures from large city hospitals are much the same; they emphasize that the common conditions underlying coma are relatively invariant in general medical practice. For example, in the series collected by Plum and Posner (Table 17-1), only 25 percent proved to have cerebrovascular disease, and in only 6 percent was coma the consequence of trauma. Indeed, all intracranial masses and their secondary effects—such as tumors, abscesses, hemorrhages, and infarcts—made up less than one-third of the coma-producing diseases. A majority was the result of exogenous (drug overdose) and endogenous (metabolic) intoxications and hypoxia. Subarachnoid hemorrhage, meningitis, and encephalitis accounted for another 5 percent of the total. Thus, intoxication, stroke, and cranial trauma stand as the “big three” of coma-producing conditions. Equally common in some series, although obvious and usually transient, is the coma that follows seizures or resuscitation from cardiac arrest.

The terms consciousness, confusion, stupor, unconsciousness, and coma have been endowed with so many different meanings that it is almost impossible to avoid ambiguity in their usage. They are not strictly medical terms but are also literary, philosophic, and psychologic ones. The word consciousness is the most difficult of all. William James remarked that everyone knows what consciousness is until he attempts to define it. To the psychologist, consciousness denotes a state of continuous awareness of one’s self and environment. Knowledge of self includes all “feelings, attitudes and emotions, impulses, volitions, and the active or striving aspects of conduct”; in short, a near continuous self-awareness of a person’s mental functioning, particularly of cognitive processes and their relation to memories and experience. These can be judged only by the individual’s verbal account of his introspections and indirectly, by his actions.

Physicians, being more practical and objective for the most part, give greater credence to the patient’s behavior and reactions to overt stimuli than to what the patient says. For this reason, they use the term consciousness in its broadest operational meaning—namely, the state of awareness of self and environment, and responsiveness to external stimulation and inner need. This narrow definition has an advantage in that unconsciousness has the opposite meaning: a state of unawareness of self and environment or a suspension of those mental activities by which people are made aware of themselves and their environment, coupled always with a diminished responsiveness to environmental stimuli.

Arousal, or the level of consciousness, refers to the appearance of being awake as displayed by the facial muscles, eye opening, fixity of gaze, and body posture, i.e., wakefulness. A clear distinction is made in medicine between the level of consciousness and the content of consciousness, the latter reflecting the quality and coherence of thought and behavior. For neurological purposes, the loss of normal arousal is by far the more important and dramatic aspect of disordered consciousness and the one identified by laypersons and physicians as being the central feature of coma.

Much more could be said about the history of our ideas concerning consciousness, and the theoretical problems with regard to its definition. There has been an ongoing polemic among philosophers of mind as to whether it will ever be possible to understand mind and consciousness in terms of reductionist physical entities, such as cellular and molecular neural systems. Although it serves little practical purpose to review these subjects in detail here, we note that contemporary investigations indicate that one constructive approach is to define the neurobiologic correlates of those elements of consciousness that are subject to observation by behavioral, electrical, and particularly imaging methods. Importantly, these controversies are informed in neurology by analyses of unusual neurologic disorders, such as those that disturb perception and consciousness of perception (phantom limb, “blindsight,” etc.). The interested reader is referred to the discussions of consciousness by Crick and Koch, Plum and Posner, Young, and Zeman listed in the references.

The following definitions are of service to clinicians and provide a convenient terminology for describing the states of awareness and responsiveness of patients.

Normal Consciousness

This is the condition of the normal person when awake. In this state the individual is fully responsive to a thought or perception and indicates by his behavior and speech the same awareness of self and environment as that of the examiner. There is attention to, and interaction with, the immediate surroundings. This normal state may fluctuate during the day from one of keen alertness or deep concentration with a marked constriction of the field of attention to one of mild general inattentiveness, but even in the latter circumstances, the normal individual can be brought immediately to a state of full alertness and function.

Confusion

The term confusion lacks precision, but in general it denotes an inability to think with customary speed, clarity, and coherence. Almost all states of confusion are marked by some degree of inattentiveness and disorientation. In this condition the patient does not take into account all elements of his immediate environment. This state also implies a degree of imperceptiveness and distractibility, referred to traditionally as “clouding of the sensorium.” Here, one difficulty is to define thinking, a term that refers variably to problem solving or to coherence of ideas. Confusion results most often from a process that influences the brain globally, such as a toxic or metabolic disturbance or a dementia. In addition, any condition that causes drowsiness or stupor, including the natural state that comes from sleep deprivation, results in some degradation of mental performance and the emergence of inattentiveness and a state of confusion. In this way, confusion, which exists along the axis of content of consciousness, is linked to alertness and the level of consciousness.

A confusional state can also accompany focal cerebral disease in various locations, particularly in the right hemisphere, or result from disorders that disturb mainly language, memory, or visuospatial orientation, but a distinction is made between these isolated disruptions in mental function and the global confusional state. They represent special states that are analyzed differently, matters discussed further in Chaps. 20 and 23.

The mildest degree of confusion may be so slight that it can be overlooked unless the examiner searches for deviations from the patient’s normal behavior and ability to carry on a coherent conversation. The patient may even be roughly oriented as to time and place, with only occasional irrelevant remarks betraying a lack of clarity and slowness of thinking. Their responses are inconsistent, attention span is reduced, and they are unable to stay on one topic, together suggesting a fundamental flaw in attention. Usually, they are disoriented and distractible, at the mercy of every stimulus. Sequences of movement may reveal impersistence and poor planning.

Severely confused and inattentive persons are unable to do more than carry out the simplest commands, and these only inconsistently and in brief sequence. Speech may be limited to a few words or phrases; or the opposite pertains—namely, some confused individuals are voluble. They give the appearance of being unaware of much that goes on around them, are disoriented in time and place, do not grasp their immediate situation or the predicament of their own confusion, and may misidentify people or objects. These illusions may lead to fear or agitation. Occasionally, hallucinatory, illusionary, or delusional experiences impart a psychotic cast to the clinical picture, obscuring the deficit in attention.

Many events that involve the confused patient leave no trace in memory; in fact, the capacity to recall events of the past hours or days is one of the most delicate tests of mental clarity. Another is the use of “working memory,” which requires the temporary storage of the solution of one task for use in the next. A deficit in working memory, which is such a common feature of the confusional states, can be demonstrated by tests of serial subtraction, and the spelling of words (or repeating a phone number) forward and then backward. Careful analysis will show these defects to be tied to inattention and impaired perception or registration of information rather than to a fault in retentive memory. These phenomena that betray inattention are the central features of most confusional states. As already stated, the observed behavior of a confused person transcends inattention alone. It may incorporate elements of clouded interpretation of internal and external experience, and an inability to integrate and attach symbolic meaning to experience (apperception). The degree of confusion often varies from one time of day to another. It tends to be least pronounced in the morning and increases as the day wears on, peaking in the early evening hours (“sundowning”) when the patient is fatigued, and environmental cues are not as clear.

In some medical writings, particularly in the psychiatric literature, the terms delirium and confusion are used interchangeably, the former connoting nothing more than a nondescript confusional state. However, in the syndrome of delirium tremens (observed most often but not exclusively in alcoholics), the vivid hallucinations; extreme agitation; trembling, startling easily, and convulsion; and the signs of overactivity of the autonomic nervous system suggest to us that the term delirium should be retained for this type of highly distinctive confusional syndrome (elaborated in Chap. 20).

As commented earlier in the discussion of the term “confusion,” a relationship between the level of consciousness and disordered thinking or, content of consciousness, is evident as patients pass through states of inattention, drowsiness, confusion, stupor, and coma.

Drowsiness and Stupor

In these states, mental, speech, and physical activity are reduced. Drowsiness denotes an inability to sustain a wakeful state without the application of external stimuli. Furthermore, in distinction to stupor discussed later, alertness is sustained spontaneously for at least some brief period, without the further neccessity of stimuli. As a rule, some degree of inattentiveness and mild confusion are coupled with drowsiness, both improving with arousal. The patient still shifts positions somewhat naturally and without prompting. The lids droop; there may be snoring, the jaw and limb muscles are slack, and the limbs are relaxed. This state is indistinguishable from light sleep, sometimes with, slow arousal elicited by speaking to the patient or applying a tactile stimulus.

Stupor describes a state in which the patient can be roused only by vigorous and repeated stimuli and in which arousal cannot be sustained without repeated stimulation. Responses to spoken commands are either absent, curtailed, or slow and inadequate. Restless or stereotyped motor activity is common, and there is a reduction or elimination of the natural shifting of body positions. When left unstimulated, these patients quickly drift back into a deep sleep-like state. The eyes move outward and upward, a feature that is shared with sleep (see further on). Tendon and plantar reflexes, and the breathing pattern may or may not be altered, depending on how the underlying disease has affected the nervous system. In psychiatry, the term stupor has been used in a second sense—to denote an uncommon condition in which the perception of sensory stimuli is presumably normal but activity is suspended and motor activity is profoundly diminished (catatonia, or catatonic stupor).

However, these states, including coma, exist in a continuum, and an alternative practical method of making distinctions between them was given by Fisher, who suggested that a verbal command is required to overcome drowsiness whereas a noxious stimulus is required to overcome stupor. This allows for further gradations in the level of consciousness based on the intensity of stimulation that is necessary to produce arousal. Also encompassed in this continuum is the observation that stuporous and drowsy patients may not always be aroused to a fully awake state.

Coma

The patient who is incapable of being aroused by external stimuli or inner need, is in a state of coma. There are variations in the degree of coma, and the findings and signs depend on the underlying cause of the disorder. In its deepest stages, no meaningful or purposeful reaction of any kind is obtainable and corneal, pupillary, pharyngeal responses are diminished. In lighter stages, sometimes referred to by the ambiguous terms semicoma or obtundation, most of the above reflexes can be elicited, and the plantar reflexes may be either flexor or extensor (Babinski sign). As mentioned, the depth of coma and stupor may be gauged by the response to externally applied stimuli and is most useful in assessing the direction in which the disease is evolving, particularly when compared in serial examinations.

Relationship of Sleep to Coma

Persons in sleep give little evidence of being aware of themselves or their environment; in this respect, they are unconscious. Sleep shares a number of other features with the pathologic states of drowsiness, stupor, and coma. These include yawning, closure of the eyelids, cessation of blinking and reduction in swallowing, upward deviation or divergence or roving movements of the eyes, loss of muscular tone, decrease or loss of tendon reflexes, and even the presence of Babinski signs and irregular respirations, sometimes Cheyne-Stokes in type. Upon being awakened from deep sleep, a normal person may be confused for a few moments, as every physician knows from personal experience. Nevertheless, sleeping persons may still respond to unaccustomed stimuli and are capable of some mental activity in the form of dreams that leave traces of memory, thus differing from stupor or coma. The most important difference, of course, is that persons in sleep, when stimulated, can be roused to normal and persistent consciousness. There are important physiologic differences as well. Cerebral oxygen uptake does not decrease during sleep, as it usually does in coma. Recordable electrical activity—electroencephalographic (EEG) and cerebral evoked responses—and spontaneous motor activity differ in the two states, as indicated later in this chapter and in Chap. 19. The anatomic and physiologic bases for these differences are only partly known.

Vegetative State

With increasing refinements in the treatment of severe systemic diseases and cerebral injury, larger numbers of patients, who formerly would have died, have survived for indefinite periods without regaining any meaningful mental function. For the first week or two after the cerebral injury, these patients are in a state of deep coma. Then they begin to open their eyes, at first in response to painful stimuli, and later spontaneously and for increasingly prolonged periods. The patient may blink in response to threat or to light and intermittently the eyes move from side to side, seemingly following objects or fixating momentarily on the physician or a family member and giving the erroneous impression of recognition. Respiration may quicken in response to stimulation, and certain automatisms—such as swallowing, bruxism, grimacing, grunting, and moaning—may be observed (Zeman). However, the patient remains unresponsive and, for the most part, unconscious, does not speak, and shows no signs of awareness of the environment or inner need; motor activity is limited to primitive postural and reflex movements of the limbs. There is loss of sphincter control. There may be arousal or wakefulness in alternating cycles as reflected in partial eye opening, but the patient regains neither awareness nor purposeful behavior. These features define the vegetative state. One sign of the vegetative state is a lack of consistent visual following of objects; brief observation of ocular movements is subject to misinterpretation, and repeated examinations are required. These perspectives have been altered by the findings of some conscious activity that can be detected by functional imaging in relation to certain commands and verbal cues such as the individual’s name as detailed below.

This state is characterized by one of a number of EEG abnormalities. There may be predominantly low-amplitude delta-frequency background activity, burst suppression, widespread alpha and theta activity, an alpha coma pattern, and sleep spindles, all of which have been described in this syndrome, as summarized by Hansotia (see Chap. 2). One important feature is a lack of—or minimal change in—the background EEG activity during and after stimulating the patient.

In the initial days and weeks, this syndrome of unconscious awakening has been referred to as the vegetative state and, if lasting 3 months after nontraumatic and 12 months after traumatic injury, the syndrome has been termed the persistent vegetative state (PVS; Jennett and Plum). These terms have gained wide acceptance and apply to this clinical appearance whatever the underlying cause.

The most common pathologic bases of this state are diffuse cerebral injury as a result of closed head trauma, widespread necrosis of the cortex after cardiac arrest, and thalamic necrosis from a number of causes. Most often, the prominent pathologic changes are in the thalamic and subthalamic nuclei, as in the widely known Quinlan case (Kinney et al) rather than solely in the cortex; this holds for postanoxic as well as traumatic cases. A review by J.H. Adams and colleagues found these thalamic changes, but attributed them to secondary degeneration from white matter and cortical lesions. However, in several of our cases the thalamic damage stood almost alone as the cause of persistent “awake coma.” In traumatic cases, the pathologic findings are often of diffuse subcortical white matter degeneration (described as diffuse axonal injury), prominent thalamic degeneration, and ischemic damage in the cortex. Taken together, these anatomic findings suggest the concept that PVS is a state in which the cortex is either diffusely injured or effectively disconnected and isolated from the thalamus, or the thalamic nuclei are destroyed. In either the traumatic or anoxic types of PVS, atrophy of the cerebral white matter may lead to ventricular enlargement and thinning of the corpus callosum.

The vegetative state or the minimally conscious state described further on, may also be the terminal phase of progressive cortical degenerative processes such as Alzheimer and Creutzfeldt-Jakob disease (where the pathologic changes may include the thalamus).

In all these clinical states, the profound and widespread dysfunction of the cerebrum is reflected by extreme reductions in cerebral blood flow and metabolism, measured with positron emission tomography (PET) and other techniques. On the basis of PET studies in a patient with carbon monoxide poisoning, Laureys and colleagues observed that the main difference between the vegetative state and the later recovered state was the degree of hypometabolism in the parietal lobe association areas. Anatomic changes in this same cortical region have been implicated in the transition from minimally conscious to a more awake state. The finding in this PET study that noxious somatosensory stimulation fails to activate the association cortices is consistent with the concept that large regions of cortex are isolated from thalamic input or that the critical parietal interpretive areas are isolated from the rest of the cortex.

Of practical value is the observation that the CT and MRI may show progressive and profound cerebral atrophy in cases of vegetative state. In the absence of this atrophy after several months or more, it may be unwise to offer a pessimistic prognosis. One patient with clinical features of the traumatic vegetative state but lacking cerebral atrophy on imaging studies regained normal cognitive ability after a year, although he remained paralyzed (R. Cranford, personal communication).

These observations notwithstanding, there is little doubt that the neuroanatomic and neurophysiologic basis of the vegetative state will prove to be complex or at least separable into categories defined by the locus of brain damage. In particular, a striking observation has been made by Owen and colleagues in a 23-year-old woman who had been vegetative for 5 months after a head injury (thus not strictly speaking a “persistent” vegetative state). They observed localized cortical activity in the middle and superior temporal gyri in response to the presentation of spoken sentences that was comparable to the brain activity in normal individuals. Di and colleagues have similarly demonstrated brain activation only to the patient’s own name and not to other names in vegetative patients. These data suggest that some forms of mental processing can go on during a vegetative state but it is not clear if this situation is representative nor does it provide information about self-awareness, a requisite for consciousness. The most compelling demonstration of cognitive processing in vegetative and minimally conscious patients has been provided by Monti and colleagues as displayed by functional MRI. Five of their 54 patients, all with traumatic brain injury but none after anoxic ischemic damage, could on command, willfully modulate focal brain activity by playing tennis (frontal lobe activation) or mentally navigating a familiar place such as their home (temporal lobe activation). In one patient, this activity could be used as a means of communication.

At a minimum, these demonstrations emphasize the care that must be taken in establishing diagnoses of PVS and minimally conscious states. Whether these findings with functional imaging simply reflect preserved islands of function in severe brain injury that were not examinable clinically or whether they require an entire rethinking of the neurologic examination that determines the state of consciousness cannot yet be stated (see editorial by Ropper, 2010).

An additional observation of some consequence is the finding of purported axonal growth over time in a patient with traumatic brain injury who had been in a minimally conscious state (see below) for 19 years and then began to speak and comprehend, while remaining virtually quadriplegic. Voss and colleagues, using sophisticated MRI diffusion tensor imaging, have shown axonal sprouting in the posterior parietal and midline cerebellar regions. They compared the results of tensor imaging to a patient who had been in a minimally conscious state for 6 years without improvement and to 20 normal individuals. Their findings are subject to several interpretations, but axonal growth in the parietal lobes offers a potential explanation for the few instances in which recovery from severe injury does occur. When combined with the findings of Laureys and colleagues, a case can be made for the posterior parietal regions as necessary for integrated consciousness. This further raise the possibility that certain islands of limited awareness may be dissociated from global brain function.

Additional terms that have been used to describe this syndrome of preserved autonomic and respiratory function without cognition include apallic syndrome and neocortical death. A position paper has codified the features of the PVS and suggests dropping a number of related ambiguous terms, although some, such as akinetic mutism, discussed further on, have a more specific neurologic meaning and are still useful (see Multi-Society Task Force on PVS).

It is difficult to predict which comatose patients will later fall permanently into the vegetative or minimally conscious categories (see Chap. 40). Plum and Posner reported that of 45 patients with signs of the vegetative state at 1 week after onset, 13 had awakened and 5 of these had satisfactory outcomes. After being vegetative for close to 2 weeks, only 1 recovered to a level of moderate disability; after 2 weeks, the prognosis was uniformly poor. Larger studies by Higashi and colleagues have given similar results. As a rough guide to prognosis specifically in head injury, Braakman and colleagues found that among a large group of comatose patients, 59 percent regained consciousness within 6 h, but of those in a vegetative state at 3 months, none became independent. At no time before 3 or 6 months was it possible to distinguish patients who would remain in a vegetative state from those who would die. Further comments regarding recovery are made in the next section on the minimally conscious state.

A study by the Multi-Society Task Force on PVS concluded that the outcome from a vegetative state is better in traumatic as compared to nontraumatic cases. J.H. Adams and coworkers have proposed that this reflects differences in the state of thalamic neurons in the two situations. They suggested that after acute hypoxia, neurons subjected to ischemic necrosis are liable to be permanently lost; by contrast, in trauma, the loss of thalamic neurons is more frequently secondary to transsynaptic degeneration following diffuse axonal injury, allowing a greater potential for recovery. Many of these ideas are speculative.

Minimally Conscious State

The vegetative state blends into a less severe but still profound dementia that has been termed the “minimally conscious state,” wherein the patient is capable of some rudimentary behavior such as following a simple command, gesturing, or producing single words or brief phrases, always in an inconsistent way from one examination to another (see Giacino et al). Here, there is preservation of the ability to carry out basic motor behaviors that demonstrate a degree of awareness, at least at some times. The minimally conscious state is found as either a transitional or permanent condition and is sometimes difficult to separate from akinetic mutism discussed further on. Any notion of such a patient’s self-awareness is purely conjectural, but there may be an impressive array of behaviors and activation of associative cortex that suggest at least some relationship to processing of external information beyond a rudimentary level (see discussion by Bernat). The causes and pathologic changes underlying the minimally conscious state are identical to those of the vegetative state, including the frequent finding of thalamic and multiple cerebral lesions, and the distinction between them is one of degree.

It is useful to maintain a critical view of reports of remarkable recuperation after months or years of prolonged coma or the vegetative state. When the details of such cases become known, it is evident that recovery might reasonably have been expected. There are, however, numerous reported instances of partial recovery in patients—particularly children and young adults—who display vegetative features for several weeks or, as Andrews and Childs and Mercer describe, even several months after injury. Such observations cast doubt on unqualified claims of success with certain therapies, such as sensory stimulation. Nevertheless, the occurrence of very late recovery in adults must be acknowledged (see Andrews; Higashi et al; and Rosenberg et al, 1977) and a relation of awakening to the recovery of connections to the parietal lobes has already been mentioned. Cases of improvement from the “minimally conscious state” are more plausible than those from the vegetative state. More recent reports, for example by Estraneo and colleagues and by Luaté and coworkers, may be more instructive but still not entirely directive. In contrast to the notion that late recovery is exceptional, the first case series of 50 consecutive patients in a PVS for a year, 10 showed late improvement an average of 2 years later, but all were severely impaired. In the second series, none of the 12 vegetative patients improved at 5 years but 13 of 39 MCS cases emerged to consciousness with severe disability.Of course, the assignation of a poor prognosis by the application of these terms to an individual patient often leads to the withdrawal of care, and the self-fulfilling poor prognosis. This is a much discussed problem that has not been satisfactorily but it emphasizes that simply labeling patients with certain diagnoses has implications for accurately assessing the natural history of some diseases.

Among the interesting recent therapeutic observations, one observation has come from Schiff and colleagues, who were able to improve function by stimulating the medial (interlaminar) thalamic nuclei through implanted electrodes in a patient who had been initially vegetative and made a natural transition to a minimally conscious state after traumatic brain injury. Longer periods of eye opening and increased responses to execute commands, such as bringing a cup to his mouth, were observed, including, for the first time since his injury, intelligible verbalization. The authors point out that this individual had preserved language cortex and connections between thalamus and cortex. Whether this remarkable result is generalizable is not known.

It cannot go without comment that the degree of disability that families find acceptable varies greatly and leads to difficult decisions regarding the continuation of medical care. The knowledgeable, sympathetic, and flexible physician is in the best position to offer perspective and guide these matters as discussed at the end of this chapter.

Locked‐in Syndrome

The states of coma described above and the vegetative state must be distinguished from a syndrome in which there is little or no disturbance of consciousness, but only an inability of the patient to respond adequately. The latter is referred to as the locked-in syndrome or the deefferented state. The term pseudocoma as a synonym for this state is best avoided, because it is used by some physicians to connote the unconsciousness of the hysteric or malingerer, the dissociative state, or catatonia. The locked-in syndrome is most often caused by a lesion of the ventral pons (basis pontis) as a result of occlusion of the basilar artery. Such an infarction spares both the somatosensory pathways, and the ascending neuronal systems responsible for arousal and wakefulness, as well as certain midbrain elements that allow the eyelids to be raised in wakefulness; the lesion essentially interrupts the corticobulbar and corticospinal pathways, depriving the patient of speech and the capacity to respond in any way except by vertical gaze and blinking. Severe motor neuropathy (e.g., Guillain-Barré syndrome), pontine myelinolysis, or periodic paralysis may have a similar effect.

Akinetic Mutism

One could logically refer to the locked-in state as akinetic mutism insofar as the patient is akinetic (motionless) and mute, but this is not the sense in which the term was originally used by Cairns and colleagues, who described a patient who appeared to be awake but was unresponsive (actually their patient was able to answer in whispered monosyllables). Following each of several drainings of a third ventricular cyst, the patient would become aware and responsive but would have no memory for any of the events that had taken place when she was in the akinetic mute state. This state of apparent vigilance in an imperceptive and unresponsive patient has been referred to by French authors as coma vigile, but the same term has been applied to the vegetative state.

The term akinetic mutism has been applied to yet another group of patients who are silent and inert as a result of bilateral lesions usually of the anterior parts of the frontal lobes, leaving intact the motor and sensory pathways; the patient is profoundly apathetic, lacking to an extreme degree the psychic drive or impulse to action (abulia). However, the abulic patient, unlike Cairns’ patient, registers most of what is happening about him and is intensely stimulated, may speak normally, relating events observed in the recent and distant past.

Catatonia

The patient with catatonia appears unresponsive, in a state that simulates stupor, light coma, or akinetic mutism. There are no signs of structural brain disease, such as pupillary or reflex abnormalities. Oculocephalic responses are preserved, as in the awake state—i.e., the eyes move concurrently as the head is turned. There is usually resistance to eye opening, and some patients display a waxy flexibility of passive limb movement that gives the examiner a feeling of bending a wax rod (flexibilitas cerea); there is also the retention for a long period of seemingly uncomfortable limb postures (catalepsy). Peculiar motor mannerisms or repetitive motions, seen in a number of these patients, may give the impression of seizures; choreiform jerking has also been reported, but the latter sign should also suggest the possibility of seizure activity. The EEG shows normal posterior alpha activity that is attenuated by stimulation. Catatonia is discussed further in Chaps. 20 and 53.

Because there is considerable imprecision in the use of terms by which various states of reduced consciousness are designated, the physician would be better advised to supplement designations such as coma and akinetic mutism by simple descriptions indicating whether the patient appears awake or asleep, drowsy or alert, aware or unaware of his surroundings, and responsive or unresponsive to a variety of stimuli. This requires that the patient be observed more frequently or over a longer period than the several minutes usually devoted to this portion of the neurologic examination. The aforementioned findings of apparent limited responsiveness reflected with functional imaging only further emphasizes the care with which these clinical diagnoses should be determined.

In the late 1950s, European neurologists called attention to a state of coma in which the brain was irreversibly damaged and had ceased to function, but pulmonary and cardiac function could still be maintained by artificial means. Mollaret and Goulon referred to this condition as coma dépassé (a state beyond coma). A Harvard Medical School committee, in 1968, called it brain death and established a set of clinical criteria by which it could be recognized (Beecher et al). R.D. Adams, who was a member of the committee, defined the state as one of complete unresponsiveness to all modes of stimulation, arrest of respiration, and absence of all EEG activity for 24 h. The concept that a person is dead if the brain is dead and that death of the brain may precede the cessation of cardiac function has posed a number of important ethical, legal, and social problems, as well as medical ones. All aspects of brain death have since been the subject of close study by several professional committees, which for the most part have confirmed the 1968 guidelines for determining that the brain is dead. The American Academy of Neurology published guidelines on this subject in 1995 and affirmed them with some refinements in 2010. The monograph by Wijdicks is a thorough modern source on the subject of brain death and also addresses the subject from an international perspective.

The philosophical underpinnings of the equating of brain death to death, giving it the same status as cessation of cardiorespiratory death, a utilitarian approach, are complex.

The ethical and moral dimensions of brain death are complex and subject to differing interpretations in various societies, religions, and cultures. One justification for equating brain death with somatic death is the general inevitability of cardiorespiratory failure in patients who fulfill the standard criteria. This tenet has exceptions, the most striking of which is a well-studied case of 20-year survival in a boy who had meningitis reported by Reptinger and colleagues, and other cases of long survival have been described with varying degrees of documentation. These have been collected by Shewmon who makes the further point that the arguments equating brain death with death on the basis of the brain’s role in creating “somatic unity” are weakened by the existence of such cases as well as by delivery of live babies from brain-dead mothers. In the end, these philosophical concerns matter but the operational state called brain death serves both patients and society well and is compatible with most of the world’s religions.

The central considerations in the diagnosis of brain death are (1) absence of all cerebral functions; (2) absence of all brainstem functions, including spontaneous respiration; and (3) irreversibility of the state. Following from the last of these criteria, it is necessary to demonstrate an irrefutable cause of the underlying catastrophic brain damage (e.g., trauma, cardiac arrest, cerebral hemorrhage) and to exclude reversible causes such as drug overdose and extreme hypothermia.

In the diagnosis of brain death, the absence of cerebral function is demonstrated by the presence of deep coma and total lack of spontaneous movement and of motor and vocal responses to all visual, auditory, and cutaneous stimulation. Spinal reflexes may persist, and the toes often flex slowly in response to plantar stimulation; but a well-developed Babinski sign is unusual in our experience (although its presence does not exclude the diagnosis of brain death). Extensor or flexor posturing is seen from time to time as a transitional phenomenon just before or after brain death becomes evident, and the status of these movements in the diagnosis is ambiguous, but the physician should proceed cautiously in declaring a patient dead in the presence of posturing and should consider conducting the examination again at a later time.

The absence of brainstem function is judged by the loss of spontaneous eye movements, midposition of the eyes, and lack of response to oculocephalic and caloric (oculovestibular) testing; the presence of dilated or midposition fixed pupils (not smaller than 3 mm); paralysis of bulbar musculature (no facial movement or gag, cough, corneal, or sucking reflexes); an absence of motor and autonomic responses to noxious stimuli; and absence of respiratory movements. The clinical findings should show complete absence of brain function, not an approximation that might be reflected, e.g., by small or poorly reactive pupils, slight eye deviation with oculovestibular stimulation, posturing of the limbs, as mentioned earlier.

As a demonstration of destruction of the medulla, it has become customary to perform an “apnea test” to demonstrate unresponsiveness of the medullary centers to a high carbon dioxide tension. This test is conducted by first employing preoxygenation for several minutes with high inspired oxygen tension, the purpose of which is to displace nitrogen from the alveoli and create a reservoir of oxygen that will diffuse along a gradient into the pulmonary circulation. The patient can then be disconnected from the respirator for several minutes during which time 100 percent oxygen is being delivered by cannula or ventilator that has its pumping mechanism turned off; this allows the arterial PCO2 to rise to 50 to 60 mm Hg (typically, CO2 rises approximately 2.5 mm Hg per minute at normal body temperature—slower if the patient is hypothermic). The hypercarbia serves both as a stimulus to breathing and a confirmation that spontaneous ventilation has failed. If no breathing is observed and examination of the blood gases shows that an adequate level of PCO2 has been attained, the presence of this component of brain death is corroborated. Several sets of formal criteria have chosen a level of CO2 of 60 mm Hg (7.98 kPa [kilopascals]) as adequate to stimulate the medulla, even under circumstances in which it has been badly damaged. In our experience, patients who have a severely damaged brainstem but nonetheless breathe, have done so at a PCO2 well below 50 mm Hg, but there are exceptions in which higher levels were required as a stimulus.

The risks of apnea testing are minimal, as discussed in the American Academy of Neurology’s 2010 document, but hypotension, hypoxemia, cardiac arrhythmias, and barotrauma may occasionally occur. In patients who cannot tolerate the test for more than a brief period, initially raising the CO2 rapidly by insufflation of this gas has been suggested, but this approach has not been studied extensively. Delivering oxygen during the test with a low tidal volume and a ventilator rate of 1 to 2 breaths per minute or by continuous positive airway pressure may ameliorate hypoxia and resultant hypotension to allow apnea to continue for a period long enough to reach the target PCO2, but this technique has also not been adequately studied.

Most, but not all, brain-dead patients have diabetes insipidus. The absence of this syndrome in some cases reflects the imprecision of clinical signs in detecting a total loss of brain function. Other ancillary bedside tests may be conducted. Among the ones we use from time to time is the inability to produce tachycardia in response to the injection of atropine; this reflects the loss of innervation of the heart by the medullary vagal neurons.

The authors have observed a number of dramatic spontaneous movements when severe hypoxia is attained upon terminal disconnection from the ventilator for several minutes. These include opisthotonos with chest expansion that simulates a breath, elevation of the arms and crossing them in front of the chest or neck (termed the Lazarus sign by Ropper, 1984), head-turning, shoulder-shrugging, and variants of these posturing-like movements. For this reason the advice that the family not be in attendance immediately after mechanical ventilation has been discontinued.

The EEG provides confirmation of cerebral death, and many institutions prefer this corroboration by the demonstration of electrocerebral silence (“flat” or, more accurately, isoelectric EEG, shown first by Schwab). However, most U.S. institutions do not require an EEG for the confirmation of death. Electrocerebral silence is considered to be present if there is no electrical potential of more than 2 mV during a 30-min recording except for artifacts created by the ventilator, electrocardiograph, and surrounding electrical devices; the absence of these artifacts suggests a technical problem with the recording.

There are cases on record in which a patient with an isoelectric EEG has had preserved brainstem reflexes so that cerebral unresponsiveness and a flat EEG do not alone signify brain death; both may also occur and may be reversible in states of profound hypothermia or intoxication with sedative-hypnotic drugs and immediately following cardiac arrest. Therefore, it has been recommended that the diagnosis of brain death not be entertained until several hours have passed from the time of initial observation. If the examination is performed at least approximately 6 h after the precipitating event, and there is prima facie evidence of overwhelming brain injury from trauma, or massive cerebral hemorrhage (the most common conditions causing brain death), there is no need for serial testing. If cardiac arrest was the antecedent event, or the cause of neurologic damage is unclear, or drug or alcohol intoxication could reasonably have played a role in suppressing the brainstem reflexes, it is advisable to wait about 24 h before repeating the testing and pronouncing the patient dead. Toxicologic screening of the serum or urine is requisite in the latter circumstances.

The impact of any requirement to perform a second brain death examination at some interval such as 6 hours has been studied by Lustbader and coworkers. Their extensive survey in New York State, where a second examination had been recommended by a panel, was instructive; of 1,311 adult and pediatric cases, none who were found to be brain dead regained brainstem function on a second test that was performed about 18 hours later. However, 12 percent had cardiac arrest and in others, consent for organ donation was withheld during the time between examinations. Several authoritative authors have argued against a second test on this basis.

Because evoked potentials show variable abnormalities in brain-dead patients, they are not of primary value in the diagnosis. Some centers use nuclide brain scanning or cerebral angiography to demonstrate an absence of blood flow to the brain, equating this with brain death; but there are technical pitfalls in the use of these methods, and it is preferable to establish the diagnosis of death primarily on clinical grounds. The specificity of radionuclide scanning is close to 100 percent but there is a self-referential aspect to this statement as the clinical diagnosis has been used as a gold standard. An additional problem arises in the observation that the sensitivity may be only about 75 percent (Joffe et al). Others take the view that demonstrating absent cerebral blood flow equates with brain death. False-negative tests are possible if a small amount of filling of the intracranial vertebral arteries or nuclide uptake in the inferior cerebellum is revealed. The same can be said for transcranial Doppler sonography, which in brain death shows a to-and-fro, pendelfluss blood-flow pattern in the basal vessels.

The main difficulties that arise in relation to brain death are not the technical ones, but those involving the sensitive conversations with the family of the patient and, to some extent, with other medical professionals. These tasks often fall to the neurologist. It is best not to embark on clinical or EEG testing for brain death unless there is a clear intention on the part of the physician to remove the ventilator or follow through with organ donation at the end of the process. The nature of testing for brain death and its intended outcome should be explained to the family in plain language. The family’s desires regarding organ transplantation should be sought after adequate time has passed for them to absorb the shock of the circumstances. Neurologists must, of course, resist pressures from diverse sources that might lead them to the premature designation of a declaration of brain death. To avoid the appearance of conflict of motivations, most centers have a separate team, often from an organ bank, to address the issues of organ transplantation after brain death has been established. The complex matter of a family’s desire to maintain ventilation and other medical support in a brain-dead relative is best addressed with kind consideration and counseling by clergy, ethics (“optimal care”) committees, and hospital staff so as to avoid confrontation. Time often allows such situations to be defused.