Chapter 24. Alterations in Mental Status

The human brain, then, is the most complicated organization of matter that we know.

           Isaac Asimov

The complexity of the brain renders its normal functioning—especially the production of consciousness—uniquely vulnerable to acute metabolic derangements and structural deformation. As a perpetual glucose and oxygen glutton, the brain is extremely intolerant of sudden changes in energy homeostasis and in vivo neurons begin to die after only minutes of fuel deprivation. Likewise, the diffuse circuitry responsible for consciousness in the brain makes anatomic insults involving both cerebral hemispheres and the brainstem reticular activating system necessary and sufficient to perturb mental status. Regardless of etiology, altered mental status (AMS) or brain failure frequently prolongs hospital length of stay and worsens the prognosis of patients in the critical care setting. Rapid diagnosis is necessary to differentiate imminently life-threatening brain failure from more benign, reversible forms. As an amalgam of evidence-based practice and our clinical experience, this chapter will focus on the diagnostic and management challenges of AMS in the intensive care unit (ICU).

AMS is an impairment of consciousness, which is comprised of arousal and awareness.1 Arousal refers to general brain wakefulness, while awareness defines whether the individual has knowledge of his or her own existence and surroundings. Awareness demands a certain degree of arousal but it may be dissociated as is best exemplified by the persistent vegetative state (PVS)—awake patients without clinically demonstrable self-awareness.2

Mental status forms the core of any neurologic examination. Caregivers at all levels should be trained to abandon labeling a patient “unresponsive” in favor of more descriptive categories based on the physical examination: lethargy, obtundation, stupor, and coma (Table 24-1).3,4 Lethargic patients manifest decreased alertness but retain awareness of their environment. Obtunded patients require a stimulus to rouse and follow simple commands but have lost awareness of their immediate surroundings. Stuporous patients do not follow commands and require a continuous painful stimulus to exhibit signs of arousal. Finally, comatose patients exhibit no awareness and no significant arousal response to even painful stimuli. Coma results from bilateral cerebral hemisphere impairment or dysfunction of the reticular activating system in the brainstem; unilateral hemispheric disease (such as a middle cerebral artery stroke) does not typically lead to coma unless there is associated midline shift and resultant contralateral hemispheric dysfunction. Although these categories are useful to help qualitatively describe the level of depressed consciousness in a patient, the lack of standardized definitions for these terms makes them prone to misuse and variable interpretation.

The Glasgow Coma Scale (GCS) remains one of the most important quantitative barometers for mental status (Table 24-2). Although originally designed by neurosurgeons in 1974 to classify patients with traumatic brain injury (TBI), the GCS has become a common language among acute care providers and has stood the test of time due to its ease of use, minimal interobserver variability, and prognostic capability.5–7 GCS is no longer solely a TBI tool, and its predictive value has been demonstrated in other diagnoses such as intracerebral hemorrhage, subarachnoid hemorrhage, intracranial subdural hematoma, ischemic stroke, Alzheimer's dementia, and poisoning.8–13

The GCS is composed of a motor, verbal, and eye opening score; however, some studies have suggested that the most useful (because it can be performed on intubated patients) and predictive component of the GCS is the motor score.14,15 Moreover, besides its limited utility in intubated patients (in which the verbal score is replaced by “T”), the GCS has often been criticized for failing to include brainstem reflexes.16 Nevertheless, the GCS remains the reigning worldwide consciousness scale and continues to facilitate both clinical research and decision making.

The motor component of the GCS deserves particular attention because it contains the most information and typically requires the most practice and effort to confidently extract from the physical examination. To meet the criteria of following commands, we recommend that the patient must show two fingers or wiggle his or her thumb to a verbal command. A common pitfall occurs in aphasic patients, who may mimic the examiner and therefore make it seem as though commands are being followed. Similarly, we recommend strict criteria for determining localizing, wherein the patient exhibits both cranial and caudal localization of the stimulus. Finally, withdrawal should be recognized as being a fairly complex and nonstereotypical movement of an extremity away from a painful stimulus; it must be distinguished from the more simple, posturing flexion and extension movements.

Although there are myriad causes of AMS in the ICU, the basic dichotomy in brain failure is whether it has been anatomically or metabolically engendered. Clinical acumen gives the physician a general gestalt as to what type of cognitive reserve one can expect in a patient. Cognitive reserve is a function of patient age, baseline brain function, brain volume, comorbidities, and perhaps duration of stay in an ICU. For example, one may expect that a urinary tract infection might cause obtundation in an 82-year-old woman in the ICU but would be reluctant to attribute such a cause to obtundation in a 30-year-old patient. History provided by the patient's family can aid in the diagnosis of AMS by both helping gauge cognitive reserve (e.g., by identifying baseline brain function and whether there are signs of an underlying dementia) and identifying substance abuse disorders, most notably alcohol dependence. As a good rule of thumb, anatomic causes of AMS (intracerebral hemorrhage, ischemic stroke, subarachnoid hemorrhage, cerebral venous sinus thrombosis, vasospasm, hydrocephalus) tend to have a more rapid onset, tend to cause a greater decrement in GCS, and are usually more immediately life-threatening than many of the “bad humor” encephalopathies such as fever, hyponatremia, or septic encephalopathy.

Pupil size and light reactivity and motor exam symmetry are typically preserved in metabolic encephalopathy. Even in the presence of nondepolarizing neuromuscular blockade when motor exam has been lost, pupil reactivity is typically preserved.17 Dilation of a pupil with loss of light reactivity heralds third cranial nerve compression with very few exceptions; rarely seizure may cause this same response.18 Prospective studies have confirmed that both anisocoria and light reflex loss have high positive predictive values for structural coma.19 Extremely small symmetric and reactive pupils may be caused by a pontine lesion. Moreover, metabolic encephalopathy may amplify or accentuate a baseline motor asymmetry (e.g., an old hemiparesis from stroke may become more pronounced in the presence of hyponatremia) but rarely manifests with a motor exam asymmetry. Recalling the brain's appetite for glucose, hypoglycemia is one cause of metabolic encephalopathy and rapid-onset coma that may cause a focal neurologic deficit along with AMS; glucose determination must accompany any acute evaluation of AMS since hypoglycemic episodes in diabetics have not uncommonly been misdiagnosed as strokes.20,21

The prevalence of etiologies for AMS in the ICU undoubtedly depends on the type of unit. Multiple studies have shown that AMS prolongs hospital stay and is an independent risk factor for mortality in the ICU.22–26 In one study of 1,758 patients admitted to a medical ICU for a non-neurologic reason, metabolic encephalopathy was found to be the primary cause of AMS followed by seizures; among the metabolic encephalopathies, septic encephalopathy was the leading culprit followed by hepatic and renal encephalopathies.27

In another study involving patients in a medical ICU, Isensee et al reported that metabolic encephalopathy was the most frequent cause of AMS and those patients with AMS had more than double the mortality rate than was seen in patients without AMS.28

Septic encephalopathy is a leading cause of AMS in the critical care setting; however, the pathogenesis for brain failure in sepsis remains unknown. As a neurophysiologic indicator of brain failure, abnormal evoked potentials were found in 84% of septic ICU patients in one study.29 Recent studies using mouse models of lipopolysaccharide-induced encephalopathy have suggested that the cytokine tumor necrosis factor (TNF) plays an important role in initiating and maintaining an inflammatory state in the brain.30 Brain microabscesses, abnormal amino acid metabolism, changes in brain neurotransmitter concentrations, reduced cerebral blood flow and oxygen utilization, and weakening of the blood–brain barrier with resultant cerebral edema have all been implicated in the pathogenesis of septic encephalopathy.31–34 Cerebrospinal fluid analysis is normal or with mild protein elevation, and electroencephalogram (EEG), the most sensitive test for septic encephalopathy, usually demonstrates a pattern consistent with metabolic encephalopathy with diffuse slow waves (predominantly delta waves) often with a triphasic pattern. Rarely, septic encephalopathy may manifest with focal neurologic deficits in addition to AMS.35 Regardless of the causes, sepsis in the ICU causes acute AMS and may often leave patients with long-term cognitive morbidity. Animal models are helping to pave the way toward therapies that might counter the pathophysiologic mechanisms behind brain failure in sepsis; however, at the present time, our only defense is early identification and treatment of infection.

There are little epidemiologic data on seizures in critical care; however, even one seizure in an adult ICU patient may double mortality.25 While the management of isolated partial or generalized seizures and status epilepticus is generally well known among intensivists, much less attention has been granted to subclinical seizures and nonconvulsive status epilepticus (NCSE). There is no current international EEG-based definition or classification scheme for NCSE. NCSE has been defined as AMS associated with continuous epileptiform changes on EEG in the absence of motor signs.36 Unfortunately, the diagnosis of NCSE is further hampered by a lack of a pathognomonic EEG pattern. NCSE may be focal or generalized based on EEG that most commonly shows spike-and-wave or polyspike-and-wave discharges (usually with frequency <2–3 Hz); some have divided the EEG patterns in NCSE into five categories: continuous focal spike and wave, continuous generalized spike and wave, continuous generalized sharp and wave, continuous focal sharp and wave, and continuous periodic lateralized epileptiform discharges.37 The prevalence of NCSE in patients with coma has been estimated to be as high as 3–8%; this is likely underestimated because the more commonly utilized routine EEG (20–30 minutes) is less sensitive for convulsive and nonconvulsive epileptic activity than continuous EEG.38 The diagnosis of NSCE is therefore often missed, and this has been a frequent argument among proponents of continuous EEG monitoring in the ICU; NCSE, independent of etiology, worsens patient morbidity and mortality.39,40 Current treatment for NCSE involves prescribing benzodiazepines (especially lorazepam and midazolam), antiepileptics (phenytoin, fosphenytoin, valproate, and phenobarbital), and intravenous anesthetics such as propofol with a goal of suppressing EEG abnormalities.41

Locked-in syndrome (LIS) is the constellation of anarthria, quadriplegia, and horizontal gaze paresis caused most commonly by an ischemic insult to the ventral pons42 (Figure 24-1). Patients with LIS can communicate solely with blinking and vertical eye movements. LIS is not a disorder of consciousness but can easily be misconstrued as coma since there is a near complete deprivation of voluntary activity. Especially in patients with brainstem pathology, physicians must remember to assess for vertical eye motion prior to declaring a patient comatose. While still a grave diagnosis, LIS needs to be identified and distinguished from coma because there may be a counterintuitive potential for good recovery with supportive care and aggressive rehabilitation.43

Brought to the public eye by the case of Terry Schiavo, the PVS is a form of AMS in which there is preserved alertness, sleep–wake cycles, and autonomic control but absent awareness and only reflexive movements.44 If such patients exhibit some nonreflexive movements as well, they are classified as being in a minimally conscious state (MCS). The PVS and MCS may develop as patients recover from coma. Head trauma with diffuse axonal injury and anoxia are the major brain insults that engender the PVS. Neuroimaging tools such as functional magnetic resonance imaging (MRI) have reinvigorated ethical debates concerning the PVS by demonstrating that these patients may retain some component of conscious awareness.45 Nevertheless, the PVS remains a devastating diagnosis and reversibility is rare. Emerging technologies such as deep brain stimulation offer the potential of augmenting consciousness in these patients.46

In vitro studies have shown that central nervous system neurons can tolerate between 20 and 60 minutes of complete ischemic anoxia without irreversible injury.47 However, the injury in vivo is much more severe and occurs in much less time. Immediately after cessation of circulation to the brain, the cerebral vessels dilate in response to the local environmental factors and increased Paco2. Because the brain has no stores of glucose, cellular metabolism quickly ceases. Loss of nutrients and hypoxia cause the most sensitive structures to lose cellular integrity. This results in leakage from capillaries, edema, and disruption of cells and leads to release of lysosomes, proteases, and other damaging compounds into the surrounding tissues.48 This in turn results in clogged microcirculation, stasis, and a vicious circle of increasing damage backing up into the macrocirculation. If this process is allowed to continue for a variable length of time and blood flow is then reestablished, the increased pressure gradient in the damaged area tends to disrupt the architecture, much like the sudden bursting of the Hoover Dam might do to downstream communities. The result is a progressive postresuscitative hypoperfusion state, in which blood flow is decreased to below the 20% level within 90 minutes postreperfusion and remains at this low level for up to 18 hours.49,50

Two theories have been offered to explain these phenomena: (1) Massive calcium (Ca2+) overloading of the cells may be the initial stage of irreversible damage.51 Normally, the extracellular Ca2+ level is high and the intracellular level is low. The damage of the cell membrane from hypoxia and loss of nutrient flow allows the gradient to shift and Ca2+ to enter the cell, causing interference with enzymes, DNA, RNA, mitochondria, and energy production cycles. Infusion of high levels of Ca2+ into precapillary arterioles causes vasospasm and a vicious cycle of decreased flow, more depletion of oxygen and nutrients, and so on. (2) During ischemia, oxygen-free radicals may be created by abnormal metabolism. These free radicals attack DNA, RNA, and mitochondria, resulting in irreversible damage.52

As it pertains to brain death in the United States, philosophical ruminations on the meaning of life and death are fortuitously muffled by state law.53 Brain death is a legal death that is relatively resistant to interpretation.54 The brain death exam basically interrogates the brainstem at a body temperature greater than 32°C and in the absence of drugs that can suppress the central nervous system or the neuromuscular junction. It is a diagnosis of what is, not what might be.55 Simply put, brain death equals legal death. Once objective criteria are met on physical examination, a physician can pronounce a patient brain dead; some states require more than one physician to make this pronouncement. The family is informed that the patient has died, life support is withdrawn, and a death certificate is completed. Overall, the criteria behind brain death are similar worldwide.

A brief summary of a typical brain death protocol is presented here:56

• The cause of injury must be known. There must be clear evidence of an acute, catastrophic, irreversible brain injury. This is extremely important. There must be clear, objective evidence of brain injury on computed tomography (CT) or MRI of the brain that is compatible with the physical exam. A physical exam compatible with brain death is by itself insufficient, as numerous tabloid media sources routinely proclaim when supposed brain dead patients unexpectedly awaken.

Reversible conditions that may confuse the clinical diagnosis of brain death must be excluded, including:

• Hypothermia; body temperature must be greater than 32°C
• Drug intoxication or inadvertent neuromuscular blockade
• Hypoperfusion and shock

The physical examination:

• No response to verbal or visual command
• No spontaneous ventilation
• No spontaneous musculoskeletal movements to pain
• Pupils fixed and nonreactive
• No oculocephalic reflex
• Negative oculovestibular reflex
• No corneal reflex
• No gag or cough reflex

The apnea test: This test should be the last test and should be conducted after two clinical exams (separated by the mandatory observation period) have confirmed the absence of brainstem function. The patient is disconnected from the ventilator while oxygenation of the lungs continues passively. By calculation (Paco2 rises 4 Torr in the first minute and 3 Torr every minute thereafter), the patient is allowed to build up to a Paco2 of 60 Torr or more without becoming hypoxic. If there is no respiratory effort, the test is considered confirmatory.57

The EEG: An EEG is not necessary for confirmation of brain death, as small artifacts may confuse the issue. If ordered, the EEG should show electrocerebral silence for at least 30 minutes and must conform to established criteria for brain death.58

When the cause of death cannot be determined without absolute accuracy, consider cerebral angiography. The absence of intracranial arterial circulation as demonstrated by four-vessel angiography confirms brain death.59

Normally, two separate examinations are done, one by a neurologist or neurosurgeon and the other by a critical care specialist or anesthesiologist with critical care experience. If, after this extensive clinical examination, the patient shows no sign of neurologic function and the cause of the injury is known, the patient can be pronounced dead (by neurologic criteria) and a death certificate is completed with the time of death noted as the time the protocol was completed.