Since an enormous variety of conditions can cause episodes of transiently disturbed consciousness or function, the major component of clinical management consists of separating out the various causes. Determining the nature of events can be very challenging, particularly when the event has been unwitnessed or when the patient is an inadequate historian otherwise.
Obtaining a clear account of the nature of the attack is the most important single component of the assessment, ascertaining precisely the circumstances of the event, any warning that was suffered, the duration of the attack, exactly what occurred during the event, nature and speed of recovery as well as whether there were any focal or lateralizing signs after the event. There is no substitute for a detailed history of the attack from both the patient and any eyewitnesses, as well as obtaining a detailed account of the circumstances of the event. For instance, what was the patient engaged in the day and evening prior to the event? Was there sleep deprivation or other medical problems? Was there unusual stress or anxiety  ? Obtaining an eyewitness account is crucial and the telephone is an invaluable device in this regard, particularly now mobile telephones are so widely owned. Dramatic disparity is often noted between the eyewitness's and patient's stories. Whereas the patient may recall a simple fall or brief loss of consciousness, an eyewitness may provide a detailed account of generalized convulsion with postictal confusion, tongue biting and so on. Patients are often amnestic for the circumstances of the event. In other situations they may deliberately try to conceal it. Clinical examination may provide useful information, but is most often non-contributory. Supportive investigations including EEGs and structural imaging may provide additional evidence for the diagnosis, but interpreting all these factors requires the greatest clinical skill and judgement. EEG is most often normal interictally in the adult patient with epilepsy; conversely, around 20% of people have minor and irrelevant abnormalities on interic-tal traces which are frequently misinterpreted as confirming a diagnosis of epilepsy [4,5],
Events that occur under particular circumstances should always raise the suspicion of syncope. There are obvious causes, such as venesection, painful surgical procedures and watching unpleasant movies, but mechanical causes such as cough, urination or defaeca-tion may also provoke syncope; the difference between syncope and epilepsy is detailed below. Episodes of loss of consciousness occurring with postural change are more likely to be syncopal. Shock, fright or extreme emotion may precipitate syncope also but are frequently recognized to be non-epileptic events. Other physical alterations such as change in head position, rolling in bed, looking up at a high shelf or bench precipitating an attack would suggest a vestibular basis. Relationship to eating might establish a hypogly-caemic basis.
Events that occur during sleep, even if only some of the time, are almost always epileptiform. Sleep disorders and other movement abnormalities might occasionally be confused but non-epileptic events never occur during sleep, though some patients maintain they were asleep when they occurred; this can be difficult to resolve without video-EEG monitoring , Seizures are sometimes linked to particular phases of the menstrual cycle, and whilst once interpreted as a functional element, this is very common in women with organic episodes and should always be taken seriously.
Episodes that occur when under emotional stress, when the patient is experiencing difficult circumstances, particularly in the cognitively impaired might relate to behavioural problems rather than seizure activity. However, the distinction is sometimes difficult particularly when it is suggested that the behavioural alterations are a feature of the postictal state.
The symptoms in the immediate moments prior to the event are diagnostically critical. Those who describe focal neurological symptoms, such as clonic jerks, olfactory or gustatory hallucinations, rising epigastric aura, intense déjà vu or similar phenomena are much more likely to be having seizures. However, some symptoms can be fairly non-specific such as light headedness and dizziness. True vertigo is rarely a feature of epileptic attacks, but it is not always easy to distinguish vertigo from brief seizures. A visual aura can be epileptiform but most often is migrainous. If a typical account of shimmering scotomatous deficit evolving over some minutes with or without a headache following and possibly associated with other neurological symptoms are described, then migraine becomes a strong possibility.
The duration of attacks is probably the best single guide when considering the nature of turns. Epileptic events are almost always seconds to minutes in duration. Migrainous neurological symptoms are usually 15-20min in duration; the subsequent headache can last for hours but may occasionally be absent. With epileptic events there is often some warning and gradual build-up to maximal deficit whereas with ischaemic vascular episodes the onset is abrupt and deficit typically maximal at the outset with gradual resolution. Since consciousness is usually unimpaired in focal cerebrovascular events involving the hemispheres, altered consciousness during attacks of this type is also more suggestive of an epileptic aetiology. Generalized tonic-clonic seizures typically last 40-90 s but occasionally are longer. Reports of attacks lasting hours, whether considered to be complex partial events or generalized tonic-clonic attacks, should always raise the suspicion of non-epileptic episodes. Whilst status epilepticus both convulsive and non-convulsive are certainly possibilities, it is a relatively uncommon event amongst people with chronic seizures. After an event, rapid recovery, perhaps with sweatiness or nausea and vomiting, are more typical of syncope. Partial and generalized tonic-clonic seizures are usually followed by a period of confusion. Occasionally there is marked alteration in mood and behaviour postictally; less often a true psychosis occurs postictally which, though typically self-limiting, sometimes dominates the presentation.
Activity during the event often helps clarify the nature of the attack. If absences are typical with abrupt cessation of activity and prompt resumption of activity at the end of the few-second-long episode, then the diagnosis is usually clear. Classical complex partial seizures with a warning followed by loss of contact, oral and manual automatisms and postictal confusion, sometimes with lat-eralizing signs noted during or after the event, are obviously clear-cut. Generalized convulsive activity can be more difficult to distinguish from syncope. With generalized tonic-clonic seizures people may or may not have a warning, the event usually lasts less than a minute or two, tongue biting and incontinence are common and there is often marked confusion postictally. The total absence of confusion after a generalized convulsive event should immediately raise the suspicion that the event was not epileptic. After the event and confusion has settled patients may strenuously deny that anything occurred, certainly that consciousness was impaired. They sometimes become convinced that others around them, most particularly their family, are conspiring to make a diagnosis of epilepsy in these circumstances. Particularly in the elderly, this denial sometimes takes on delusional proportions.
During a seizure well-organized motor activity is uncommon, though automatisms can sometimes be preservative and simple activities are continued, although in an incomplete and sometimes clumsy manner. The purposeless nature of motor activity during the events usually draws the attention of those around the patient. The description of normal performance of complex activity such as driving a car or riding a bicycle suggests the attacks are non-epileptic. More often these partial events are truly simply partial in nature, and patients are able to continue normal activity. Partial seizures of temporal lobe origin are usually associated with altered consciousness, at least to some degree, though this is often not perceived by the patient. There are rare accounts of patients suffering generalized tonic-clonic convulsions and being able to recall events around them after the episode. This rare phenomena is usually related to generalized motor convulsive activity resulting from lesions in the frontal or parietal cortex where consciousness can sometimes be preserved despite the bilateral symmetry of the motor activity
Similarly, seizures of extratemporal origin, particularly those originating in the frontal lobes, sometimes have bizarre features that may be similar to non-epileptic events. Furthermore, video-EEG monitoring with scalp electrodes can be unremarkable during these events, obscuring the issue diagnostically. Helpful clues are the stereotypic nature of attacks that often cluster, and that may occur during sleep. If unusual events occur in association with a structural cerebral pathology the diagnosis is usually clear. Great caution must be exercised diagnosing non-epileptic events in patients with bizarre clinical events that have a structural pathology demonstrated on MRI, particularly if it is extratemporal in location.
Prolonged 'absences', typically occurring whilst driving, are a common reason for referral to the epilepsy clinic. The patient describes driving or walking some distance, and then finding themselves at their destination (or just missing it), and not able to recall how they got there. If they have made the trip without difficulty, arrived at their destination and there is no sign of damage to the vehicle, it is highly unlikely such activity occurred during a seizure. These patients—and the referring doctors —are typically very anxious about the event (in contrast to many patients who have had complex partial events whilst driving!). It can be difficult to provide satisfactory reassurance that this is a benign phenomenon experienced to some degree by very many people.
Neurological examination is rarely helpful in patients who present with episodic disorders. Stigmata of a phakomatosis, the finding of a significant hemi-atrophy, lateralized weakness or reflex change and, of course, transiently lateralizing signs immediately postictally can be very useful. Acutely after a seizure perhaps the most useful physical sign is the observation of petechiae over the upper trunk and face particularly sometimes a quite striking phenomenon but usually subtle. Tongue bites and evidence of incontinence might be present if the patient is seen early enough. Although most tongue bites are lateral after epileptic seizures, the tip of the tongue and occasionally even the lips or cheeks can be bitten. Injuries such as fractures and bruising are not so helpful, often occurring through loss of consciousness with syncope for example. Shoulder dislocation, particularly posterior dislocation, and crush-fractured vertebrae are highly suggestive that a seizure has occurred, and are never seen in syncope or non-epileptic events. Back pain or radicular pain postevent should always be investigated with X-rays of the region; these injuries are often not diagnosed correctly and can lead to significant problems in returning to normal activity. Tests for vestibular abnormalities might be performed and sometimes provoke attacks.
Cardiac examination might disclose features to suggest an alternative aetiology for episodic disorders. Cardiac bruits, valvular heart disease, cardiomegaly or postural hypotension, tics and other movement abnormalities might be detected during the physical examination.
Occasionally patients have seizures whilst being examined. Most often these episodes are non-epileptic. Hyperventilation might be induced deliberately having informed the patient of your purpose, but other floridly non-epileptic attacks are sometimes brought on by simple tests, such as deep tendon reflexes, fundoscopy or suggestion. One needs to exercise great caution interpreting such events but most often they provide strong primary evidence as to the true nature of the episodes. Vulnerable patients with epilepsy may be easily induced to have non-epileptic events in some circumstances, particularly if they believe the organic nature of events is being questioned. There is considerable pressure to 'perform' for some, whether during the examination and history or video-EEG monitoring. Thus, the use of suggestion and other provocative procedures should only be performed in special circumstances ,
Laboratory tests, such as biochemistry and haematological screens, add little to the diagnosis of epilepsy. Occasionally a primary metabolic disturbance such as hyponatraemia is found, but this almost always occurs in a specific clinical setting and in the context of other recognized metabolic abnormalities. Elevation in creatine kinase (CK) and white blood cell count might transiently occur after a seizure , Serum prolactin levels rise transiently after seizures, reaching a peak at a bout 15 min after the event and returning to normal after around an hour. Obtaining a blood prolactin level can be useful in the diagnosis of events of uncertain type, provided it is done close enough to the episode. Prolactin levels are elevated following generalized convulsions in about 90% of cases, following complex partial seizures in probably only about 50% and not elevated following simple partial episodes. There is some uncertainty as to how prolactin changes might be interpreted in other settings, such as syncope and migraine. Also, numerous medications and other pathological conditions can cause changes in prolactin levels, although generally these do not cause transient fluctuations like seizures do [10,11]. Although in principal serum prolactin ought to be a useful test, it is difficult to implement because of the time scale and the fact that most seizures do not occur in circumstances where obtaining an acute sample is possible. At times though, serum prolactin estimation provides useful supportive information. It is not appropriately used as the primary diagnostic modality ,
Other tests that can be useful include structural imaging, CT scan or MRI. Visualization of a focal cerebral pathology involving the cortex may provide useful supportive evidence for a diagnosis of epilepsy, but finding a structural pathology does not prove attacks are epileptiform. Conversely, not finding a structural pathology does not exclude a diagnosis of epilepsy, even if the symptomatology is focal. The sensitivity of MRI scans particularly with quantitative measures is now so great that it is uncommon in focal seizures of long standing not to find a relevant abnormality. However, in some patients abnormalities are never demonstrated perhaps because they are too small or subtle, or do not exist. The aetiology of these seizure types is often unknown and many appear to have a relatively good prognosis.
Functional imaging tests such as SPECT and PET are more appropriately used in conjunction with video-EEG monitoring or as part of surgical work-up in specialty epilepsy units. They are rarely helpful in a diagnostic setting.
EEGs and video-EEG monitoring are extremely useful tests that need careful interpretation. Unfortunately EEGs show an enormous range of minor abnormalities, benign variants, artefactual change and other confusing features that are often misinterpreted as evidence that there is a cerebral disturbance of some sort [4,6,8,12], Whilst EEG can provide confirmation of precisely the type of epilepsy, and occasionally the location of a structural pathology, more often it leads to erroneous diagnosis of epilepsy when minor changes are misinterpreted. The EEG should never be substituted for a good clinical history; EEG changes, even if epileptiform should be interpreted cautiously. There is a very strong case to be made for not doing studies like EEG if the primary diagnosis is non-epileptic, provided there are strong clinical grounds for an alternative diagnosis. Video-EEG monitoring is as close to a gold standard as is available. Actually capturing events, witnessing directly the physical accompaniment of the attacks and observing the EEG changes which occur with this, often allows a specific diagnosis or the exclusion of epilepsy. However, simple partial events, extratemporal episodes, particularly from the frontal lobes even if associated with altered consciousness, are sometimes not associated with changes on the EEG. On the other hand, generalized convulsions always, and most complex partial events usually, show diagnostic EEG change. Simple partial events, particularly those involving sensorimotor cortex, are very often normal, even if the seizure activity is continuous.
Repeated observations over time also help make the correct diagnosis. Clinicians often feel obliged to arrive at the correct diagnosis immediately and at first consultation in episodes where alteration in consciousness has occurred. Whilst there are good reasons for this, and obviously serious causes need to be excluded rapidly, when the diagnosis is unclear, it is much better to leave the diagnosis open, because an erroneous diagnosis of epilepsy has serious implications for the patient. The concern with unexplained episodes of altered consciousness generally relates to safety during driving and perhaps in the workplace, and these activities might need to be restricted if the nature of episodes is uncertain but this will depend on the specific circumstances of the patient, the frequency of attacks and their character. Even if these do need to be restricted to some degree, this is a much better precaution than the so-called 'therapeutic trial' of anticonvulsant that often gives rise to uncertain and confusing results, sometimes leading to the de facto diagnosis of epilepsy. Much more harm is done through the incorrect diagnosis of epilepsy than keeping an open mind and reviewing the situation when more information is to hand, after implementing appropriate safety precautions.
Was this article helpful?