Transient alteration of consciousness is a major clinical challenge of neurology. Evaluating transient impairment of consciousness is critical to diagnose epileptic seizures, syncope, parasomnias, organic encephalopathies, and psychogenic nonepileptic seizures (PNESs). Primary care physicians, general hospitalists, and emergency physicians struggle with these diagnoses in all but their most straightforward presentations, and typically are baffled when faced with the coexistence of multiple causes of episodic alteration of consciousness within 1 patient. The consulting neurologist is best equipped among physicians to direct the diagnostic process in determining the causes of transiently impaired consciousness, whether these are amenable to neurologic therapeutics or are best treated by a cardiologist, psychiatrist, or neurosurgeon. The neurologist must enter the patient care arena with extensive knowledge of the causes of transiently impaired consciousness, which this review presumes. Consciousness is often considered by its form (state of arousal) and its content (subjectively appreciated information processing of the moment), and with many limitations both form and content are objectively evaluated based on the individual's behavior. The patient's subjective appreciation of impaired consciousness is in itself likely to be impaired, but sometimes provides useful clues to the cause of impairment. This review aims to assist the general neurologist in honing strategies and techniques to maximize recovery of information from the patient and witnesses of impaired consciousness events, and to push beyond wishful mindsets of the patient, family members, and even the referring physicians, who may bias their information toward diagnoses that are unlikely to recur, are easily treated, or are nonpsychiatric. The increased sensitivity and specificity of prolonged electrophysiologic tests are associated with greater cost and relative inconvenience, and this review also aims to assist the consulting neurologist in mastering the indications and applications of results for each test.
Neurology clinic evaluation of ambulatory patients
Ambulatory patients who are referred to an adult or pediatric neurologist for consultation in the outpatient clinic are almost never unconscious during the visit. Occasionally a patient loses consciousness in the waiting room or the examination room; this occurs with high seizure frequency or easily triggered seizures, as in children with absence seizures, but in adults seems more often to occur with PNESs. In general the neurologist does not expect to witness brief events with impaired consciousness, but only to receive information about them from the patient and lay witnesses of the events.
Transient amnesia for environmental stimuli is the core feature of unconscious events that the patient may describe to the neurologist. Subjectively appreciated diminution of attentiveness and mental capacities (“confusion”) may be described following impaired consciousness. Complete unresponsiveness (or markedly diminished responsiveness) is the core feature of unconscious (or impaired) events that event witnesses should be able to clearly describe to the neurologist, often with observations of motor behaviors. Even attentive adults who witnessed an event with unconsciousness usually have a limited and often distorted verbal description of the patient's behaviors during a brief and unexpected attack.1
We found that 4 types of repetitive movements or postural changes could be consistently described by lay witnesses of unconscious events in concordance with behaviors evident during video-EEG recording of the habitual events,2
as described in .
Global motor phenomena in differential diagnosis of events with impaired consciousness
Our strategy in obtaining history from a witness of one or more impaired consciousness events is first to establish that the patient was unresponsive, then to obtain unstructured recollections (“What did you see?”), followed by specific questions that are general (“Did he fall?” “Did her entire body jerk or twitch?” “Did he simply stare, or did his face or body move?”), then progressively more specific and adapted to the types of ictal semiologies and postictal dysfunctions. Many patients with epileptic or nonepileptic seizures have more than one type of seizure. Primary care physicians frequently miss important information about these events, and may make assumptions about the nature of the events that further obscure the diagnosis. The majority of time during an initial visit with these patients often is devoted to collecting and reviewing information of the ictal semiology, including auras and prodromes (if any) and postictal states, from the patient and event witnesses. The history of these events is the cornerstone of optimal neurologic diagnosis.
The general condition of the patient before unconscious events and of medical, surgical, and psychiatric history frequently reveals significant risk factors for epilepsies or epileptic syndromes, or in other cases for provoked epileptic seizures in the absence of epilepsy (which are always grand mal seizures, never absences or complex partial seizures).
Syncope is a common cause of sudden alteration of consciousness, typically preceded by lightheadedness and rarely lasting longer than a minute.3
Syncopal myoclonus and urinary incontinence can resemble epileptic seizures. Rhythmic jerking preceded by rigidity or posturing is more consistent with seizures. In syncope, atypical presyncopal symptoms and post-event confusion may mimic epileptic seizures. Neurologists should consider PNESs when drop attacks are reported to be quite frequent and to involve confusion. Causes of syncope are summarized in . Early referral of recurrent syncope for cardiologic consultation is important due to a number of etiologies that constitute life-threatening cardiac conditions.
In many cases the nature of recurring unconsciousness is clearly due to medication-resistant epilepsy.4
It must be accepted that one-third of epilepsy patients cannot have seizures controlled with standard medications. When a clear epilepsy diagnosis has not led to full control of seizures, without adverse effects of medications, rather than to repeatedly check EEGs and other tests in hopes of finding a more tractable diagnosis, the general neurologist would do best to refer the patient to a specialized epilepsy center.
Psychogenic events do not have pathognomonic history, interictal EEG, or imaging findings.5
While numerous risk factors for PNESs can be found on mental health history and on neuropsychological testing, none of these excludes the possibility of epilepsy or organic nonepileptic events. Recording actual behavioral events with video-EEG monitoring is generally required for confident diagnosis of PNESs, with subsequent psychiatric consultation.
Parasomnias also can generate behaviors that are indistinguishable from those of nocturnal epileptic seizures. Nocturnal events must be recorded with video-EEG or polysomnography for definitive diagnosis.6
When unconscious events are recurrent and refractory to therapies for the presumed diagnosis, the consulting neurologist can consider several electrophysiologic studies, as reviewed below. In addition to the hematologic and chemical tests, the urine toxicology screen should not be neglected (to detect illicit substances and also prescribed psychoactive agents not reported by the patient).
Consultative evaluation of emergency department and hospitalized patients
In emergency department (ED) and hospitalized patients, impaired consciousness events may have been witnessed by medical or nursing personnel. In addition to information derived from lay witnesses of earlier events, these patients benefit from a higher quality of behavioral observation to guide diagnosis. On the other hand, the deck is stacked against the neurologist who is asked to see a patient who now is awake and at baseline mentation, but has no memory of the period during which he or she was “found down” or “confused” at home or in public, without witnesses of the behaviors at onset of impaired consciousness. When the patient and companions are able to describe earlier such episodes, and no circumstantial evidence or medical history such as previously diagnosed epilepsy or cardiac arrhythmia are reported, the emergency evaluation will depend on metabolic and toxin screens, a head CT, an EKG, and perhaps later a brief routine EEG. Most of these tests are relatively insensitive to intermittent disturbances of cerebral and cardiac electrophysiology and to subtle cerebral lesions that are unassociated with neurologic deficits on examination during waking consciousness.
The neurologist must consider any established condition that can be a risk factor for epileptic seizures or for organic or psychogenic nonepileptic events, while suspiciously looking beyond these diagnoses to avoid missing additional causes or correlates of transiently impaired consciousness.
Transient global amnesia does not cause impaired consciousness. The patient is often brought to the ED for “impaired consciousness,” but the observers usually can clearly describe the patient's ongoing verbal and nonverbal responsiveness throughout the several hours of retrograde-anterograde amnesia. Etiologies include ischemia in the mesial temporal lobe, migrainous phenomena, epileptic seizures, and psychogenic amnesia.7
Thus, the laboratory evaluation often is similar to that of staring spells and automatistic events of impaired consciousness.
Intensive care unit evaluation of critically ill patients
A special problem of critically ill patients is sudden onset of altered consciousness that persists longer than is typical for most seizures, but may be due to a reversible cause of stupor-coma. The differential diagnosis of alteration of consciousness in the acute inpatient setting is somewhat narrower if the symptoms are transient. A thorough clinical evaluation is necessary, with special emphasis on neurologic etiologies, review of laboratory values, and vital signs.8
summarizes some of the more common etiologies.
Evaluation for potentially reversible acute severe encephalopathies
Strokes and TIAs occur in critically ill patients, triggered by prothrombotic states, inflammation, infections, cardiac arrhythmias, paradoxical emboli, and various other causes. Rarely, decreased level of consciousness is the only manifestation of a TIA, but if the deficits are transient, diminished responsiveness might be the only sign noted and reported by the intensive care unit (ICU) staff.
Nonconvulsive status epilepticus (NCSE) is increasingly recognized as a cause of altered consciousness in the critically ill. In one series, 19% of all ICU patients with unexplained coma had subclinical seizures on continuous EEG monitoring.9
While NCSE may present as a waxing-waning twilight state with intermittent automatisms, all too often the level of responsiveness is invariant over hours, and automatisms are absent.10
An EEG is absolutely necessary in order to exclude NCSE. While a routine EEG is long enough to detect the recurring partial electrographic seizures or spike-wave discharges of NCSE, a longer EEG recording is necessary to exclude acute repetitive seizures as a condition that also can cause persisting but reversible stupor. In the absence of NCSE and acute repetitive seizures, the EEG data measure the severity of organic encephalopathies of all types.
Critically ill patients often have had prior events of brief unconsciousness, and contacting lay and professional witnesses of these events is as important as in outpatient and ED evaluations. General factors at play include acute hepatic or renal failure, infections, and fever. Hypertensive encephalopathy can be quite transient.11
Postoperative patients in stupor-coma have detailed records of intraoperative blood pressure monitoring, but periods of hypotension during transfers between postanesthesia units and ICUs may be poorly documented. There is often alteration of drug absorption, distribution with changes in intravascular volumes and protein binding, drug metabolism and elimination, and drug-drug interactions.12
A thorough review of drugs is needed.
An important and common phenomenon of alteration of alertness in the ICU, often fluctuating, is delirium. Recent data have established a high prevalence of delirium in the ICU, up to 80% in high-risk medical ICU patients.13
Delirium is an independent predictor of increased mortality and worse outcome. Recognized risk factors for delirium in the ICU include sepsis, fever, exposure to sedatives and opiates, disruption of sleep, preexisting dementia, older age, and acute sensory deprivation (such as missing eyeglasses and hearing aid).
Special studies for evaluation of transient loss of awareness
Outpatient EEG and ambulatory EEG monitoring
Interictal epileptiform activities (“spikes”) are valuable signs of epileptic excitability, which sometimes are recorded on brief, routine scalp EEG.14
In most epilepsies interictal spikes can be intermittently infrequent, such that prolonged EEG recordings often detect spikes missed with routine EEGs. Ambulatory EEG can detect interictal spikes and electrographic seizures that occur frequently.15
The absence of audio-video recording renders ambulatory EEG inadequate for diagnosis of PNESs and most organic nonepileptic events, however, as the lay witness reports of event behaviors are generally unreliable for detailed behavioral analysis. Recently, ambulatory video-EEG has become available. This will be expected to greatly expand definitive diagnosis of nonepileptic events in outpatients. Tapering of antiseizure medications to induce seizures for diagnostic purposes is widely practiced on inpatient epilepsy monitoring units, but is not considered safe for patients to perform at home.
Inpatient video-EEG monitoring
Diagnostic video-EEG monitoring is a powerful tool in evaluating episodically impaired consciousness.16
The EEG data can fully distinguish epileptic seizure discharges (with phenomena specific to partial-onset vs generalized-onset epileptic seizures), pathologic generalized slowing and depression of electrocerebral activities (of syncope and organic encephalopathies), onset of unconscious behaviors during sleep (specific to REM-onset vs slow-wave sleep onset), and absence of change from waking or drowsy baseline EEG activities (in PNESs). The synchronized audio-video recording of behaviors is highly informative, and even when myogenic and kinesigenic artifacts obscure the EEG recording the recorded behavioral data frequently support definitive seizure diagnosis. The most significant diagnostic limitation of inpatient video-EEG is failure to record the habitual seizures, when baseline seizure frequency is low or when an individual patient has less frequent seizures under hospital conditions that reduce the usual physical activities or psychological stresses. Ambulatory video-EEG may address the latter issue, although the procedural organization required to keep the patient on camera at home often precludes videorecording of events. Medication tapering during inpatient video-EEG often accelerates diagnosis, as well as offering the opportunity to alter medication regimens and simultaneously screen for early adverse effects.
Only habitual events should be considered for the overall seizure diagnosis. Some patients have events that have not previously occurred during the hospital stay. Hypnosis and other forms of intense or directive behavioral event induction are not reliable for seizure diagnosis. If strong suggestion induces a nonhabitual, nonepileptic event, this may merely indicate that the patient is highly suggestible, but this does not exclude a diagnosis of epilepsy.
Electrocardiographic tilt-table testing
The EKG tilt-table test, or head-up tilt test, can diagnose the cause of syncope when hypotension or cardiac arrhythmia occurs in response to a maintained upright posture while strapped on the table.17
With EKG, blood pressure, and sometimes transcranial Doppler and other monitoring in place, the patient is held upright and if necessary given sublingual nitroglycerin or another inducing agent to provoke syncope or near-syncope. The laboratory cardiologist will be able to administer specific therapy if the blood pressure does not rapidly normalize with resumption of the flat position. A positive result may lead directly to cardiac pacemaker placement depending on the diagnosis, so neurologists usually arrange for brain MRI prior to tilt-table testing when both tests are indicated.
Electrocardiographic loop monitor
Implantable loop records are small loop recorders that are implanted in the left chest subcutaneously.18
They do not require any transvenous leads. They can record for up to 40 minutes of a single lead EKG, battery lasting up to 3 years. Syncopal events can be marked by patients via an external device. Several studies have shown that in recurrent syncope, these devices diagnosed arrhythmias as the cause of syncope. Some of these devices are MRI compatible as they lack a transvenous lead, although data can be erased by the magnetic field.19