Книги по МРТ КТ на английском языке / PLUM AND POSNER S DIAGNOSIS OF STUPOR AND COM-1

confounding factors. This procedure also has the advantage, when positive, of establishing a structural cause of brain death (i.e., absence of blood flow to the brain). In cases where the original cause of cerebral injury is not known, the absence of blood flow provides the crucial information necessary to declare brain death with certainty.

Physiologically, two events may produce failure of the cerebral circulation. First, a sudden and massive increase in ICP (e.g., during subarachnoid hemorrhage) may cause it to rise to the level of arterial perfusion pressure at which point cerebral circulation ceases. The second, and probably more common, occurrence is a progressive loss of blood flow that accompanies death of the brain. As the dead tissue becomes edematous, the local tissue pressure exceeds capillary perfusion pressure, resulting in stasis of blood flow, further edema, and further vascular stasis. If the respiratory and cardiovascular systems are kept functioning for many hours or days after brain circulation has ceased, the brain undergoes autolysis at body temperature, resulting in a soft and necrotic organ at autopsy referred to by pathologists as a ‘‘respirator brain.’’21

Demonstration of the failure of intracerebral filling at the level of entry of the carotid and vertebral arteries indicates brain death. Recently, magnetic resonance angiography (MRA) has been reported for diagnosis of brain death, but this technique is less reliable, as MRA often fails to demonstrate slow flow. Additional MRI criteria for brain death include loss of the subarachnoid spaces, slow flow in the intracavernous and cervical internal carotid arteries, loss of flow void in both small and large intracranial arteries and venous sinuses, and loss of gray-white matter distinction on T1-weighted images, but ‘‘supranormal’’ distinction on T2weighted images.22 However, until additional data are available on the reliability of these indicators for determining brain death, the presence of complete cessation of brain function on examination, or complete loss of blood flow, must remain the gold standards for diagnosis. In addition, the rarity of ventilators that are compatible with MR scanners continues to limit the availability of this mode of diagnosis.

Bilateral insonation of the intracranial arteries using a portable 2-MHz pulsed Doppler device (transcranial Doppler ultrasonograph

[TCD]) is now a widely used confirmatory test for brain death.23 The middle cerebral arteries are insonated on both sides through the temporal bone above the zygomatic arch (of note, up to 10% of patients may not have temporal insonation windows, limiting use of this method) and the vertebral arteries or basilar artery through a suboccipital transcranial window. Two types of abnormalities have been correlated with brain death: (1) an absence of diastolic or reverberating flow, indicating the loss of arterial contractive force, and (2) the appearance of small systolic peaks early in systole, indicative of high vascular resistance. Both abnormalities are associated with significant elevations of ICP. The technique is limited by the requirement of skill in the operation of the equipment and has a potentially high error rate for missing blood flow because of incorrect placement of the transducer. Recent studies report a sensitivity of 77% and a specificity of 100% of diagnosing brain death if both the middle cerebral arteries and the basilar artery were insonated; sensitivity improved with increasing time of evaluation following initial clinical diagnosis.24

Cerebral scintigraphy measures the failure of uptake of the radioisotope nuclide technetium (Tc) 99m hexametazime in brain parenchyma. This technique has shown good correlation with cerebral angiography. The test can be done at the bedside using a portable gamma camera after injection of isotope, which should be used within 30 minutes after its reconstitution. A static image of 500,000 counts obtained at several time points is recommended (taken immediately, 30 to 60 minutes after injection, and at 2 hours past injection time25). A recent prospective study using 99m Tc-hex- amethyl-propylamineoxime (HMPAO) single photon emission tomography (SPECT) in 50 comatose and brain-dead patients to examine cerebral perfusion found the characteristic ‘‘empty skull’’ image indicating arrest of cerebral perfusion in 45 of 47 brain-dead patients.26 The bedside nuclide brain scan test is probably the best adjunct test to confirm the diagnosis in unclear cases. It is inexpensive, can be done without moving a patient on a ventilator, and is extremely reliable when it shows an empty skull (see Figure 8–1). This test can be considered a gold standard for use in difficult cases.

ELECTROENCEPHALOGRAPHY/ EVOKED POTENTIAL MEASUREMENTS

The EEG has little place in the determination of brain death, except perhaps in those rare cases where other clinical evidence is equivocal. An isoelectric EEG, often termed electrocerebral inactivity by electroencephalographers that lasts for a period of 6 to 12 hours in a patient who is not hypothermic and has not ingested or been given depressant drugs, identifies forebrain death (because the EEG does not demonstrate brainstem activity, it can be isoelectric in patients with brainstem reflexes who are clearly not brain dead). Silverman and associates reported on a survey of 2,650 isoelectric EEGs that lasted up to 24 hours.27 Only three patients in this group, each in coma caused by overdose of central nervous system depressant drugs, recovered cerebral function. However, Heckmann and colleagues28 have reported a patient with an isoelectric EEG following cardiac arrest who showed residual brainstem function, including spontaneous breathing and SPECT evidence of cerebral blood flow, for 7 weeks prior to death.

Electrical interference makes artifact-free EEG or evoked potential records exceedingly difficult to obtain in the intensive care setting. Moreover, technical recording errors can simulate electrocerebral activity as well as electrocerebral inactivity and several ostensibly isoelectric tracings must be discarded because

of faulty technique. A national cooperative group has published technical requirements necessary to establish electrocerebral silence (Table 8–5), and has produced an atlas illustrating potential problems of interpretation of the EEG in coma.29 It should be noted that the EEG is not infallible, even with anoxicischemic injury. Cerebral activity may be absent on the EEG for up to several hours following cardiac arrest, only to return later.30 A prolonged vegetative existence is occasionally possible in such cases despite the presence of an initially silent EEG. After depressive drug poisoning, total loss of cerebral hemispheric function and electrocerebral silence have been observed for as long as 50 hours with full clinical recovery.

Physicians have appropriately raised questions as to whether a few fragments of cerebral electrical activity mean anything when they arise from a body that has totally lost all capacity for the brain to regulate internal and external homeostasis. Death is a process in which different organs and parts of organs lose their living properties at widely varying rates. Death of the brain occurs when the organ irreversibly loses its capacity to maintain the vital integrative functions regulated by the vegetative and consciousness-mediating centers of the brainstem. Not surprisingly, the time when the state of brain death is reached often precedes the final demise of small collections of electrically generating cells in the cerebral hemispheres,

Table 8–5 Electroencephalographic Recording for Diagnosing

Cerebral Death

1.A minimum of eight scalp electrodes and ear reference electrodes

2.Interelectrode impedances under 10,000 ohms, but over 100 ohms

3.Test of integrity of recording system by deliberate creation of electrode artifact by manipulation

4.Interelectrode distances of at least 10 cm

5.No activity with a sensitivity increased to at least 2 mV/mm for 30 minutes with inclusion of appropriate calibrations

6.The use of 0.3- or 0.4-second time constants during part of the recording

7.Recording with an electrocardiogram and other monitoring devices, such as a pair of electrodes on the dorsum of the right hand, to detect extracerebral responses

8.Tests for reactivity to pain, loud noises, or light

9.Recording by a qualified technician

10.Repeat record if doubt about electrocerebral silence (ECS)

11.Telephonic transmitted electroencephalograms are not appropriate for determination of ECS

From Bennett et al.29

as evidenced by the observation that 20% of 56 patients meeting other clinical criteria for brain death had residual EEG activity lasting up to 168 hours.31 Thus, EEG examinations may pick up a few patients with brainstem death who have not yet progressed to full brain death. Given the extremely poor prognosis of such individuals, using EEG as a criterion for prolonging the period of futile life support is not a service to them.

Diagnosis of Brain Death in Profound Anesthesia or Coma of Undetermined Etiology

It must be repeatedly emphasized that patients with very deep but reversible anesthesia due to sedative drug ingestion can give the clinical appearance of brain death, and even can have an electrically silent EEG. Furthermore, recovery in such instances has been observed even when the EEG showed no physiologic activity for as

long as 50 hours. Given such evidence, when and how is one to decide in such cases that anesthesia has slipped into death and further cardiopulmonary support is futile? Unfortunately, few empirical data provide an answer to the question, particularly if faced with the complex problem of a patient with a coma of undetermined origin. In such cases, the combination of a prolonged period of observation (more than 24 hours), loss of cerebral perfusion, and exclusion of other potential confounds is required.32 It is important to test drug levels and follow the patient until the drug is eliminated. A general guideline proposed for known intoxications is the following: an observation period greater than four times the half-life of the pharmacologic agent should be used.4 Of course, the presence of unmeasured metabolites, potentiation by additional medications, and impaired renal or hepatic clearance are likely to complicate individual evaluations.

Pitfalls in the Diagnosis

of Brain Death

Potential pitfalls accompany the diagnosis of brain death, particularly when coma occurs in hospitalized patients or those who have been chronically ill. Almost none of these will lead to serious error in diagnosis if the examining physician is aware of them and attends to them when examining individual patients who are considered brain dead. In fact, there are no reported cases of ‘‘recovery’’ from correctly diagnosed brain death.

With meticulous efforts, other organs (e.g.

heart, kidney, etc.) can be sustained, but usually only for hours or days.33,34 Prolonged survival of peripheral organs is quite rare,35,36 so

much so that in the few reported cases, one must question whether the clinical criteria were correctly met. Conversely, there are several reported cases of recovery from ‘‘cardiac’’ death,37 the Lazarus phenomenon (not to be confused with Lazarus sign, a spinal reflex [see page 334]). A number of case reports describe patients with clinical and electrocardiographic cardiac arrest who, after failed attempts at resuscitation, are pronounced dead, only to be discovered to be alive later, sometimes in the mortuary.38 Some of these pitfalls are outlined in Table 8–6.

Table 8–6 Some Pitfalls in the

Diagnosis of Brain Death

Adapted from Wijdicks.4

In comatose patients, pupillary fixation does not always mean absence of brainstem function. In rare instances, the pupils may have been fixed by pre-existing ocular or neurologic disease. More commonly, particularly in a patient who has suffered cardiac arrest, atropine has been injected during the resuscitation process and pupils are widely dilated; fixed pupils may result without indicating the absence of brainstem function. Neuromuscular blocking agents also can produce pupillary fixation, although in these instances the pupils are usually midposition or small rather than widely dilated.

Similarly, the absence of vestibulo-ocular responses does not necessarily indicate absence of brainstem vestibular function. Like pupillary responses, vestibulo-ocular reflexes may be absent if the end organ is either poisoned or damaged. For example, traumatic injury producing basal fractures of the petrous bone may cause unilateral loss of caloric response. Some otherwise neurologically normal patients suffer labyrinthine dysfunction from peripheral disease that predates the onset of coma. Other patients with chronic illnesses have suffered ototoxicity from a variety of drugs, including antibiotics such as gentamicin. In these patients, vestibulo-ocular responses may be absent even

though other brainstem processes are still functioning. Finally, a variety of drugs, including sedatives, anticholinergics, anticonvulsants, chemotherapeutic agents, and tricyclic antidepressants, may suppress vestibular and/or oculomotor function to the point where oculovestibular reflexes disappear.

Pitfalls in the diagnosis of apnea in comatose patients maintained on respirators have been discussed above.

The absence of motor activity also does not guarantee loss of brainstem function. Neuromuscular blockers are often used early in the course of artificial respiration when the patient is resisting the respirator; if suspected brain death subsequently occurs, there may still be enough circulating neuromuscular blocking agent to produce absence of motor function when the examination is carried out. One report has described the simulation of brain death by excessive sensitivity to succinylcholine39; in this case the presence of activity in the EEG established cerebral viability. If neuromuscular blockade has been recently withdrawn, guidelines require that a peripheral nerve stimulator be used to demonstrate transmission (e.g., a train of four stimulation pulses produces four thumb twitches).

Therapeutic overdoses of sedative drugs to treat anoxia or seizures likewise may abolish reflexes and motor responses to noxious stimuli. At least two reports document formal brain death examinations in reversible intoxications with tricyclic antidepressant and barbiturate agents.40,41

There are pitfalls in using the EEG as an ancillary technique in the diagnosis of cerebral death. Isoelectric EEGs with subsequent recovery have been reported with sedative drug overdoses, after anoxia, during hypothermia, following cerebral trauma, and after encephalitis, especially cases of diffuse acute disseminated encephalomyelitis.5

REFERENCES

1.Mollaret P, Goulon M. [The depassed coma (preliminary memoir).]. Rev Neurol (Paris) 1959; 101, 3–15.

2.A definition of irreversible coma. Report of the Ad Hoc Committee of the Harvard Medical School to Examine the Definition of Brain Death. JAMA 1968; 205, 337–340.

3.Uniform Laws Annotated 320 Uniform Determination of Death Act. 1990.

4.Wijdicks EF. The diagnosis of brain death. N Engl J Med 2001; 344, 1215–1221.

5.An appraisal of the criteria of cerebral death. A summary statement. A collaborative study. JAMA 1977; 237, 982–986.

6.Wijdicks EF. Determining brain death in adults. Neurology 1995; 45, 1003–1011.

7.Bates D, Caronna JJ, Cartlidge NE, et al. A prospective study of nontraumatic coma: methods and results in 310 patients. Ann Neurol 1977; 2, 211–220.

8.Jorgensen EO. Spinal man after brain death. The unilateral extension-pronation reflex of the upper limb as an indication of brain death. Acta Neurochir (Wien) 1973; 28, 259–273.

9.President’s Commission for the Study of Ethical Problems in Medicine and Biomedical Behavioral Research Defining Death: Medical, Legal and Ethical Issues in the Determination of Death. 1981.

10.Shlugman D, Parulekar M, Elston JS, et al. Abnormal pupillary activity in a brainstem-dead patient. Br J Anaesth 2001; 86, 717–720.

11.Wijdicks EFM. Temporomandibular joint compression in coma. Neurology 1996; 46, 1774–1774.

12.Christie JM, O’Lenic TD, Cane RD. Head turning in brain death. J Clin Anesth 1996; 8, 141–143.

13.Hanna JP, Frank JI. Automatic stepping in the pontomedullary stage of central herniation. Neurology 1995; 45, 985–986.

14.de Freitas GR, Andre C. Absence of the Babinski sign in brain death: a prospective study of 144 cases. J Neurol 2005; 252, 106–107.

15.Martı´-Fa`bregas J, Lo´pez-Navidad A, Caballero F, et al. Decerebrate-like posturing with mechanical ventilation in brain death. Neurology 2000; 54, 224–227.

16.Ropper AH. Unusual spontaneous movements in brain-dead patients. Neurology 1984; 34, 1089–1092.

17.Saposnik G, Bueri JA, Maurin˜o J, et al. Spontaneous and reflex movements in brain death. Neurology 2000; 54, 221–223.

18.Saposnik G, Maurino J, Saizar R, et al. Spontaneous and reflex movements in 107 patients with brain death. Am J Med 2005; 118(3), 311–314.

19.McNair NL, Meador KJ. The undulating toe flexion sign in brain death. Mov Disord 1992; 7, 345–347.

20.Schafer JA, Caronna JJ. Duration of apnea needed to confirm brain death. Neurology 1978; 28, 661–666.

21.Walker AE, Diamond EL, Moseley J. The neuropathological findings in irreversible coma. A critique of the ‘‘respirator.’’ J Neuropathol Exp Neurol 1975; 34, 295–323.

22.Lee DH, Nathanson JA, Fox AJ, et al. Magnetic resonance imaging of brain death. Can Assoc Radiol J 1995; 46, 174–178.

23.Sloan MA, Alexandrov AV, Tegeler CH, et al. Assessment: transcranial Doppler ultrasonography: report of the Therapeutics and Technology Assessment Subcommittee of the American Academy of Neurology. Neurology 2004; 62, 1468–1481.

24.Kuo JR, Chen CF, Chio CC, et al. Time dependent validity in the diagnosis of brain death using transcra-

nial Doppler sonography. J Neurol Neurosurg Psychiatry 2006; 77, 646–649.

25.Bonetti MG, Ciritella P, Valle G, et al. 99mTc HMPAO brain perfusion SPECT in brain death. Neuroradiology 1995; 37, 365–369.

26.Facco E, Zucchetta P, Munari M, et al. 99mTcHMPAO SPECT in the diagnosis of brain death. Intensive Care Med 1998; 24, 911–917.

27.Silverman D, Masland RL, Saunders MG, et al. Irreversible coma associated with electrocerebral silence. Neurology 1970; 20, 525–533.

28.Heckmann JG, Lang CJ, Pfau M, et al. Electrocerebral silence with preserved but reduced cortical brain perfusion. Eur J Emerg Med 2003; 10, 241–243.

29.Bennett DR, Hughes JR, Korein J. Atlas of Electroencephalography in Coma and Cerebral Death. EEG at the Bedside or in the Intensive Care Unit. San Diego: Raven Press, 1976.

30.Jorgensen EO. Clinical note. EEG without detectable cortical activity and cranial nerve areflexia as parameters of brain death. Electroencephalogr Clin Neurophysiol 1974; 36, 70–75.

31.Grigg MM, Kelly MA, Celesia GG, et al. Electroencephalographic activity after brain death. Arch Neurol 1987; 44, 948–954.

32.Practice parameters for determining brain death in adults (summary statement). The Quality Standards Subcommittee of the American Academy of Neurology. Neurology 1995; 45, 1012–1014.

33.Yoshioka T, Sugimoto H, Uenishi M, et al. Prolonged hemodynamic maintenance by the combined administration of vasopressin and epinephrine in brain death: a clinical study. Neurosurgery 1986; 18, 565–567.

34.Hung TP, Chen ST. Prognosis of deeply comatose patients on ventilators. J Neurol Neurosurg Psychiatry 1995; 58, 75–80.

35.Shewmon DA. Chronic ‘‘brain death’’: meta-analysis and conceptual consequences. Neurology 1998; 51, 1538–1545.

36.Repertinger S, Fitzgibbons WP, Omojola MF, et al. Long survival following bacterial meningitis-associated brain destruction. J Child Neurol 2006; 21, 591–595.

37.Maleck WH, Piper SN, Triem J, et al. Unexpected return of spontaneous circulation after cessation of resuscitation (Lazarus phenomenon). Resuscitation 1998; 39, 125–128.

38.Mullie A, Miranda D. A premature referral to the mortuary. Cerebral recovery with barbiturate therapy. Acta Anaesthesiol Belg 1979; 30, 145–148.

39.Tyson RN. Simulation of cerebral death by succinylcholine sensitivity. Arch Neurol 1974; 30, 409–411.

40.Grattan-Smith PJ, Butt W. Suppression of brainstem reflexes in barbiturate coma. Arch Dis Child 1993; 69, 151–152.

41.Yang KL, Dantzker DR. Reversible brain death. A manifestation of amitriptyline overdose. Chest 1991; 99, 1037–1038.

42.Laureys S, Owen AM, Schiff ND. Brain function in coma, vegetative state, and related disorders. Lancet Neurol 2004; 3, 537–546.

INTRODUCTION

PROGNOSIS IN COMA

PROGNOSIS BY DISEASE STATE

Traumatic Brain Injury

Nontraumatic Coma

Vascular Disease

Central Nervous System Infection

Acute Disseminated Encephalomyelitis

Hepatic Coma

Depressant Drug Poisoning

VEGETATIVE STATE

Clinical, Imaging, and Electrodiagnostic

Correlates of Prognosis in the

Vegetative State

MINIMALLY CONSCIOUS STATE

Late Recoveries From the Minimally

Conscious State

LOCKED-IN STATE

MECHANISMS UNDERLYING OUTCOMES OF SEVERE BRAIN INJURY: NEUROIMAGING STUDIES AND CONCEPTUAL FRAMEWORKS

FUNCTIONAL IMAGING OF VEGETATIVE

STATE AND MINIMALLY

CONSCIOUS STATE

Atypical Behavioral Features in the Persistent

Vegetative State

Neuroimaging of Isolated Cortical Responses

in Persistent Vegetative State Patients

POTENTIAL MECHANISMS UNDERLYING RESIDUAL FUNCTIONAL CAPACITY IN SEVERELY DISABLED PATIENTS

Variations of Structural Substrates

Underlying Severe Disability

The Potential Role of the Metabolic ‘‘Baseline’’ in Recovery of Cognitive Function

The Potential Role of Regionally Selective Injuries Producing Widespread Effects on Brain Function

ETHICS OF CLINICAL DECISION MAKING

AND COMMUNICATION WITH

SURROGATES (J.J. FINS)

Surrogate Decision Making, Perceptions,

and Needs

Professional Obligations and Diagnostic

Discernment

Time-Delimited Prognostication and Evolving

Brain States: Framing the Conversation

Family Dynamics and Philosophic

Considerations

INTRODUCTION

It is much more difficult to predict the outcome for patients with severe brain damage than to make the usually straightforward diagnosis of brain death. Brain death is a single biologic state with an unequivocal future, while severe brain injuries span a wide range of outcomes (Figure 9–1) depending on a number of variables that include not only the degree of neurologic injury, but also the presence and severity of medical complications. Scientific, philosophic, and emotional uncertainties that attend predictions of outcome from brain damage can intimidate even the most experienced physicians. Nevertheless, the problem must be faced; physicians are frequently called upon to treat patients with severe degrees of neurologic dysfunction. To do the job responsibly, the physician must organize available information to anticipate as accurately as possible the likelihood that the patient will either recover or remain permanently disabled. The physician’s role as a translator of

medical knowledge is essential in counseling families who must make the ultimate decisions concerning the care of an unconscious patient. The financial and emotional costs of caring for those left hopelessly damaged can exhaust both family and medical staff. Physicians must attempt to reduce those burdens, while at the same time retaining an unwavering commitment to do everything possible to treat those who can benefit.

In the 26 years since the publication of the third edition of Stupor and Coma, several groups of neurologists and neurosurgeons have initiated studies to identify and quantify early clinical,neurophysiologic, radiologic,andbiochemical indicants that might predict outcome in comatose patients. These studies have identified the etiology of injury, the clinical depth of coma, and the length of time that a patient remains comatose as the most critical factors. Additional important factors include the age of the patient, the neurologic findings, and concurrent medical complications (particularly the complications of increased intracranial pressure [ICP] and hypoxia in the setting of traumatic in-

juries). Several limitations, as discussed below, place stringent demands on physicians to carefully consider all available historical details and the reliability of clinical and laboratory evaluations in their consideration of prognosis for an individual patient.

Prospective studies of prognosis in adults and children indicate that within a few hours or days after the onset of coma, neurologic signs and electrophysiologic markers in many patients differentiate, with a high degree of probability, the extremes of no improvement or good recovery. Unfortunately, radiologic and biochemical indicators have generally provided less accurate predictions of outcome, with some exceptions discussed below. Accurate prognostication improves over time, but it is still unclear how early one can make accurate predictions within different diagnostic categories (e.g., vegetative state [VS] vs. minimally conscious state [MCS]). The first section of this chapter details what we now know about prognosis, emphasizing broad outcome categories and shortterm outcomes rather than outcomes beyond a year or longer, although we recognize that rarely, even severely brain-injured patients may improve after many years (see page 371). We use the scheme in Table 9–1 to assess the reliability of the data presented in this section.

The second section addresses mechanisms that may underlie recovery, or lack thereof, from coma. Severe cognitive disabilities can arise from at least two fairly different anatomic injuries: (1) extensive, relatively uniform diffuse axonal injury or hypoxic-ischemic damage causing widespread neuronal death and (2) focal cerebral injuries causing functional al-

Table 9–1 Levels of Evidence

Level I Data from randomized trials with low false-positive (alpha) and low false-negative (beta) errors

Level II Data from randomized trials with high false-positive (alpha) or high false-negative (beta) errors

Level III Data from nonrandomized concurrent cohort studies

Level IV Data from nonrandomized cohort studies using historical controls

Level V Data from anecdotal case series

Modified from Broderick et al.1

teration of integrative systems in the upper brainstem and thalamus. New studies suggest that physiologic correlates of brain function in some severely disabled patients with relatively intact cerebral structures may ultimately lead to identification of residual cerebral capacities.

Figure 9–1 shows a conceptual organization of functional outcomes following severe brain injuries and indicates that a very wide range of cognitive capacities may be present in the setting of impaired motor function, including normal cognition in the locked-in state (LIS). In this section we discuss advances in neuroimaging aimed at uncovering the biologic distinctions that underlie VS, persistent vegetative state (PVS), MCS, and related enduring disorders of consciousness following coma. Despite the relatively small number of studies in this area to date, functional imaging has added to our general understanding of pathophysiologic mechanisms in VS. Ongoing work in MCS patients suggests that significant physiologic differences in brain function will generally separate these categories.

The third section addresses important ethical considerations in dealing with comatose patients and their families and caregivers.

PROGNOSIS IN COMA

Coma has a grave prognosis. For the two most carefully studied etiologies of coma, traumatic brain injury and cardiopulmonary arrest, mortality ranges from 40% to 50% and 54% to 88%,2 respectively. These statistics have actually improved since the last edition of Stupor and Coma, because of better acute management both in the field and in intensive care. Beyond mortality statistics, very few studies of prognosis in coma have looked at large numbers of patients for careful evaluation of outcomes other than survival or death. These indicate that patients comatose from traumatic brain injury have a significantly better prognosis than patients with anoxic injuries. For example, of 1,000 trauma patients in coma for at least 6 hours, 39% recovered independent function at 6 months,3 whereas only 16% of 500 patients suffering nontraumatic coma made similar recoveries at 1 year.4

Statistics such as the above, however, are too coarse to guide individual patient management.

That step requires clinical judgment combined with accurate knowledge of the medical literature, as applied specifically to the patient’s history and awareness of common diagnostic pitfalls. This section reviews efforts to predict outcome from coma for different etiologies. The reader will find that the literature continues to provide little specific information about the kind of outcome enjoyed or suffered by patients.5 As a result, except where specified, descriptions of recovery from coma often connote little more than survival and fail to tell one about the social, vocational, or emotional outcome (i.e., the human qualities) of the life that followed.

The Glasgow Outcome Scale (GOS; Table 9–2) originates from a study of outcome following nontraumatic coma in 500 patients. The definitions attempted to identify fairly precisely what was meant by each grade of outcome. Only a small number of outcomes were chosen in the hope that sufficient numbers of patients would fall into each class to allow statistical analysis, but that important differences in medical and social recovery would not be excessively blurred. A shortcoming of this classification is that the category of severe disability (3) is too broad in that it includes all patients who cannot

Table 9–2 Glasgow Outcome

Scale (GOS)

Severe disability (3) Patients who regain at least some cognitive function but depend on others for daily support

Vegetative state (2) Patients who awaken but give no sign of cognitive awareness

No recovery (1) Patients who remain in coma until death

From Jennett and Bond,5 with permission.

function independently, from those minimally conscious to those almost independent. There still exists a need for further subdivision and consideration of outcomes in the severely disabled group, as discussed below. An important limitation in evaluating the prognostic data in the literature is that some studies conflate death, VS, and severely disabled but conscious outcomes of coma survivors. For example, when using the prognostic data provided below, care should be taken to distinguish indicators of death from those indicating outcomes including severe disability, which remains a very broad category. Moreover, many outcome studies do not provide sufficient follow-up of subjects to assess outcomes of permanent VS. To allow comparisons across studies, this chapter indicates the GOS cutoff score used in each report below and does not categorize outcomes as ‘‘good,’’ ‘‘bad,’’ ‘‘favorable,’’ or ‘‘unfavorable.’’

Another fundamental issue in determining a prognosis for any individual patient is the etiology of injury. It must be recognized that the overwhelming weight of medical knowledge for prognosis in coma falls into two large categories: traumatic brain injury (TBI) and anoxicischemicencephalopathy(AIE).Unfortunately, there are many additional etiologies that can produce coma and related disorders of consciousness, and it is often not possible simply to place an individual patient with another etiology into the context of TBI or AIE. Where possible, information specific to other etiologies is provided below, but the physician should recognize this general limitation when formulating a prognosis for a comatose patient who has not suffered a traumatic brain injury or cardiac arrest.

PROGNOSIS BY DISEASE STATE

Traumatic Brain Injury

More effort has been directed at trying to predict outcome from TBI than from any other cause of coma. This emphasis reflects the high prevalence of TBI (estimated at 1.5 to 2.0 million persons per year in the United States6), the young age of most patients (peak 15 to 24 years old), and the enormous financial, social, and emotional impact of the illness that may persist for decades. Coma arising from TBI has

a better prognosis than nontraumatic coma, possibly because patients are usually younger and the pathophysiology differs from other types of coma. Recovery after prolonged traumatic coma is well described and, unlike nontraumatic causes, unconsciousness for 1 month does not necessarily preclude significant recovery. Severe head injury causing 6 hours or more of coma still carries a 40% probability of recovering to a level of moderate disability or better.7 A comprehensive literature review by the Brain Trauma Foundation in 20008 organizes evidence-based data for early prognostic signs in TBI; class I prognostic evidence is listed in Table 9–3.

The Glasgow Coma Scale (GCS) score has at least a 70% positive predictive value (PPV) for an outcome less than 4 on the GOS, if evaluations done after cardiopulmonary resuscitation were performed after sedative and paralytic agents had been metabolized. Gennarelli and colleagues9 found a progressive increase in mortality for patients with descending GCS scores in the 3 to 8 range in 46,977 headinjured patients. Two studies provide class I evidence for the predictive value of the GCS. Narayan and associates10 prospectively studied

Table 9–3 Class I Evidence for Early Prognosis in Coma Due to Traumatic Brain Injury

I.Glasgow Coma Scale (see Chapter 1): worsening outcome grades in continuous stepwise manner with lower GCS score

II.Age: 70% positive predictive value (PPV) with increasingly worse outcome in continuous and stepwise manner associated with

increasing age

III.Absent pupillary responses: 70% PPV of an outcome <4 on the GOS

IV. Hypotension/hypoxia: systolic blood pressure <90 mm Hg has a 67% PPV for an outcome <4 on the GOS outcome,

and 79% PPV when combined with evidence of hypoxia

V.Computed tomography imaging abnormalities: 70% PPV of an outcome <4 on the GOS with initial abnormalities including compression, effacement, or blood within

the basal cisterns, or extensive traumatic subarachnoid hemorrhage

Developed from the Brain Trauma Foundation Management and Prognosis of Severe Traumatic Brain Injury.8

133 patients of all age ranges and found that 62% of patients with a GCS of 3 to 5, when examined either in the emergency room or on admission to an intensive care unit, at later evaluation had a GOS of 1 (Table 9–1). Braakman and colleagues11 prospectively studied 305 patients and correlated GOS level 1 outcomes in 100% of patients with a GCS of 3, 80% with a GCS of 4, and 68% with a GCS of 5. The several studies examined in the Brain Trauma Foundation review support a survival rate of 20% for patients with the lowest GCS scores and an outcome above the level of severe disability (GOS 4 or 5) in 8% to 10% of the patients, limiting the use of the GCS alone for prognosis.

MOTOR FINDINGS

A reasonably good indication of outcome can be obtained by testing motor responses to noxious stimulation.12,13 Abnormal flexor (decorticate), abnormal extensor (decerebrate), or predominantly flaccid responses in patients with severe head injury denote an outcome of less than 4 on the GOS in every reported series. By 6 hours, motor responses no better than abnormal flexor were associated with a mortality of 63%, while abnormal extensor or flaccid responses predicted an 83% mortality.7 Unfortunately, the European Brain Injury Consortium found that the motor score of the GCS was untestable in 28% of 1,005 patients at the time of admission to a neurosurgery service, and that the full GCS score could not be assessed in 44% patients due to prehospital medications and management with intubation.12 Testing of the motor response by application of nail-bed or supraorbital pressure is considered most reliable but may be complicated by

tissue injury (e.g., periorbital swelling or quadriplegia).14,15

AGE

Advanced age unfavorably influences outcome in traumatic coma. Paradoxically, elderly patients may require a much longer recovery time, so it is risky to predict ultimate recovery early in the course. Of 600 patients with severe head injury causing coma, 56% of those younger than age 20 recovered to a GOS of 4 or 5. This number fell to 39% between age 20 and 59 years and to only 5% among those older than