NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

Panayiotopoulos CP. The Epilepsies: Seizures, Syndromes and Management. Oxfordshire (UK): Bladon Medical Publishing; 2005.

Cover of The Epilepsies

The Epilepsies: Seizures, Syndromes and Management.

Show details

Chapter 2Optimal Use of the EEG in the Diagnosis and Management of Epilepsies

Clinical note

The EEG,1–5 which is entirely harmless and relatively inexpensive, is the most important investigation in the diagnosis and management of epilepsies providing that it is properly performed by experienced technicians and carefully studied and interpreted in the context of a well-described clinical setting by experienced physicians.
Although not a substitute for a clinical examination, EEG is an integral part of the diagnostic process in epilepsies and this should not be underrated.
More than one-half of children and adults currently referred for a routine EEG are suspected of suffering from or do suffer from epilepsies. The EEG is indispensable in the correct syndromic diagnosis of these patients.

Epilepsies are usually easy to diagnose. However, as with any other medical condition, they are sometimes difficult and challenging. I use the EEG as an integral part of the diagnostic process. In this sense, there is more than enough justification for performing an EEG after the first seizure or in patients suspected of having epilepsy. The EEG may be the only means of an incontrovertible syndromic diagnosis. That the patient may not need treatment6 is not a convincing argument against such a practice. The prime aim in medicine is the diagnosis, which determines the prognosis and management strategies.

The role of the EEG is to help the physician to establish an accurate diagnosis. In most conditions (infantile spasms, myoclonic epilepsies, idiopathic generalised epilepsy (IGE), symptomatic generalised epilepsy, temporal lobe epilepsy (TLE), Landau–Kleffner syndrome, benign childhood focal seizures and photosensitive and other reflex epilepsies) the EEG may specifically confirm or may specifically direct towards such a diagnosis if this is clinically uncertain. In other situations it may not be helpful with normal rhythms or some non-specific diffuse or paroxysmal slow activity. These cases may need an EEG during sleep, on awakening or both and again it may not reveal specific changes in approximately 10% of patients. However, even a normal EEG in an untreated patient may be useful as it may exclude some of the above conditions where EEG abnormalities are expected to be high, such as in IGE with absence seizures.

An EEG in chronic epilepsies or treated patients may be uninformative and misleading. Obtaining previous medical and EEG reports is essential.

A request for an EEG should describe the clinical problem well but, because this is rarely the reality, the EEG technologist should also obtain the relevant clinical information.

The Value of Routine Inter-Ictal or Ictal Extracranial EEG in Epilepsies Should Neither Be Overrated nor Undervalued

EEG in epilepsies is overrated by some and undervalued by others. The truth is in between.

Reasons Why EEG Should Not Be Undervalued

  • An EEG is the only available investigation for recording and evaluating the paroxysmal discharges of cerebral neurons causing seizures. The appropriate evaluation of patients with epileptic disorders is often impossible without an EEG. In the majority of cases the clinical diagnosis is concordant with the EEG findings. However, it is often with the help of an EEG that the correct diagnosis is established, particularly if the clinical information is inadequate or misleading (Figure 2.1). On other occasions, the clinical data are sounder than the EEG findings, particularly if these are non-specific or in chronic cases of treated epilepsies.
  • The seizure and epileptic syndrome classifications are based on combined clinico-EEG manifestations. Epileptic syndromes, the most important advance of recent epileptology, were mainly identified because of their EEG manifestations.
  • Focal and generalised epilepsies are often difficult to differentiate without an EEG even by the most experienced epileptologists (Figure 2.1).
  • It is the EEG which will often document beyond any doubt that the ‘daydreaming’ of a child is due to absence seizures (Figure 2.2), that long-lasting episodes of behavioural changes are due to non-convulsive status epilepticus, that the ‘eyelid ticks’ are due to eyelid myoclonia with photosensitivity, that the clumsiness on awakening is due to myoclonic jerks and that periodic bedwetting is due to nocturnal seizures.
  • The EEG in neonatal seizures is the most powerful investigative tool (p. 93).
Figure 2.1. Use and misuse of EEGs. Two girls with similar EEG abnormalities of generalised spike/multiple spike and slow wave discharges either spontaneous or elicited by intermittent photic stimulation.

Figure 2.1

Use and misuse of EEGs. Two girls with similar EEG abnormalities of generalised spike/multiple spike and slow wave discharges either spontaneous or elicited by intermittent photic stimulation. Top: proper use of EEGs. This (more...)

Figure 2.2. The Significance of EEG in the Diagnosis of Epilepsies; EEG of Four Patients with Epileptic Seizures.

Figure 2.2

The Significance of EEG in the Diagnosis of Epilepsies; EEG of Four Patients with Epileptic Seizures. Top: Ictal EEG of a 4-year-old boy who had frequent brief episodes of 'panic' without impairment of consciousness or convulsive (more...)

Reasons Why EEG Should Not Be Overvalued

  • The EEG may be oversensitive in conditions such as the benign childhood seizure susceptibility syndrome and sightless in others such as frontal epilepsies or often TLEs. On rare occasions even ictal events may be undetected in a surface EEG (some frontal seizures are a typical example of this). Patients with mainly focal epilepsies may have a series of normal EEGs and the EEG localisation is not always concordant with ictal intracranial recordings. More than 40% of patients with epileptic disorders may have one normal inter-ictal EEG although this percentage falls dramatically to 8% with series of EEGs and appropriate activating procedures, particularly sleep.7
  • The frequency of seizures is not proportional to the EEG paroxysmal ‘epileptogenic’ discharges. Severely ‘epileptogenic’ EEGs may be recorded from patients with infrequent or controlled clinical seizures and vice versa. The EEG abnormalities do not reflect the severity of the epileptic disorder.3
  • More than 10% of normal people may have non-specific EEG abnormalities and approximately 1% may have ‘epileptiform paroxysmal activity’ without seizures.4 The prevalence of these abnormalities is higher in children, with 2–4% having functional spike discharges.3
  • Paroxysmal epileptiform activity is high in patients with non-epileptic, neurological or medical disorders or with neurological deficits. For example, children with congenital visual deficits frequently have occipital spikes and patients with migraine have a high incidence of sharp paroxysmal activity and other abnormalities.3,8

Sources of Error in EEG

Even the most reliable investigative tools in medicine cannot escape severe errors because of poor technical quality (equipment, personnel or both), interpretation by poorly qualified physicians or both. A competent report should not only describe the EEG abnormality accurately, but also provide its significance and meaning in accordance with a well-described clinical setting.3

Failing to achieve this leads to severe errors and erroneous criticism such as ‘a routine inter-ictal EEG is one of the most abused investigations in clinical medicine and is unquestionably responsible for great human suffering’.9 Anything in medicine, clinical or laboratory, may be harmful if misinterpreted. Raising standards, not abandoning the service, is the proper response (Figure 2.1).10

That a patient with a brain tumour may not have clinical signs does not invalidate the clinical examination and the same is true for the EEG.

The main cause of concern and suffering is that physicians, including a few epilepsy authorities, have misunderstood the EEG, its value and its limitations.

Providing that the EEG is technically correct, the following are in my opinion the most important sources of error listed in order of significance.

  1. The single most significant source of error is that the EEG is often interpreted out of the clinical context. There are two reasons for this. Firstly, the referring physician provides inadequate information regarding the events (‘patient with loss of consciousness or grand mal seizures’, ‘black-outs: epilepsy?’ and ‘unexplained aggressiveness: TLE?’) and often fails to mention other medical conditions or drugs that may significantly affect the EEG and its interpretation. Secondly, the reporting clinical neurophysiologist prefers a convenient but rather unhelpful and uncommitted abbreviation of the factual report (‘normal EEG’, ‘an abnormal EEG with generalised discharges of frontal origin’ and so forth). In St Thomas’ Hospital, I am in the advantageous position to be the referring and reporting physician and this practice may expand to other clinics for epilepsies. Further, our EEG technologists are well trained in obtaining the missing clinical information (see page 38).11
  2. Non-specific EEG abnormalities are overemphasised without suggesting means of clarifying their significance, with a sleep-deprived EEG for example or after obtaining more clinical data. Episodic focal slow waves are non-specific and may occur in a normal person or a patient with migraine, with mild cerebrovascular disease or even cerebral tumours. They may be of lateralising significance even if they are infrequent and of small amplitude in a patient with a well-established clinical history of temporal lobe seizures.12
  3. An EEG during hyperventilation, drowsiness and sleep may produce significant changes, which are often difficult to interpret even by experienced neurophysiologists. EEGs in babies and children are even more complex and demanding. Non-epileptic episodic transients such as benign epileptiform transients of sleep, six and 14 positive spikes per second and rhythmic mid-temporal discharges may often be misinterpreted as evidence of ‘epilepsy’.4
  4. Previous EEG records and results are lost, destroyed or not sought. EEGs recorded at the initial stages of the disease and particularly before treatment are significant not only at the first medical presentation of the patients, but also in the reevaluation of those with long-standing epilepsy. These patients are mainly referred for treatment modifications because they are free of seizures and still on medication, they had a recent convulsion after a long seizure-free period, their seizures are not controlled with long-standing medication, they have adverse reactions to their medication or it is appropriate to change to new anti-epileptic drugs (AEDs), anticipated pregnancy in women and so forth. The answer in all these cases is difficult. It requires a thorough clinical evaluation, review of previous medical and EEG records and the establishment of the appropriate epileptic syndrome. It does not depend on the findings of a new EEG, which may be misleading; for example, it may be normal in a patient with IGE who is on sodium valproate or may show focal slow wave paroxysms in patients with well-documented generalised spike and wave discharges (GSWDs) in an early EEG. However, on other occasions a recent EEG may prompt towards the correct advice documenting mild epileptic seizures such as myoclonic jerks or absences in a patient ‘free of seizures’ or of ‘continuing focal seizures’ inadequately treated with carbamazepine.
  5. Alteration of the EEG by drugs (such as neuroleptics or anti-epileptics) or coexisting medical conditions (cerebrovascular disease, electrolyte disturbances or a previous head injury).

Activating Procedures

Activating procedures are intended to improve the EEG diagnostic yield by inducing or enhancing epileptogenic paroxysms. Hyperventilation and intermittent photic stimulation are routinely applied in awake stage EEGs. Drowsiness, sleep and awakening are also very important activating procedures.


Hyperventilation often induces EEG changes such as diffuse and paroxysmal slow activity in normal people and particularly in young persons who overbreathe well. These changes do not last for more than 30 s after cessation of hyperventilation and they should not be confused with abnormal epileptogenic disturbances, which are also activated by hyperventilation.

Current practice in EEG departments is usually to ask the patient to hyperventilate for 3 min. Patients are instructed to breathe deeply rather than quickly at a rate of 20 deep breaths/min.13,14 This will cause an air exchange of 20–50 l/min in adults and a drop in pCO2 of 4–7ml%.15 Young children are encouraged to hyperventilate by asking them to blow on a brightly coloured pinwheel or a balloon. Infants often hyperventilate while sobbing.

Breath Counting for the Detection of Cognitive Impairment during GSWDs of 3–4 Hz

Breath counting is the most effective and practical means of reliably testing transient cognitive impairment during GSWDs induced by hyperventilation.16,17 Breath counting should be used routinely during overbreathing in patients with GSWDs. Physicians and EEG technologists cannot appreciate its value prior to using and experiencing the benefits of this technique in their clinical practice.

A GSWD of 3–4 Hz, which is the electrical accompaniment of typical absences, is nearly invariably (more than 90% of untreated patients) induced or enhanced by hyperventilation. In current routine practice, if a GSWD occurs the technician is expected to detect possible associated ictal clinical symptoms such as eye opening, staring, cessation of overbreathing, myoclonic jerks, abnormal eye movements, automatisms and so forth. A verbal test stimulus (a phrase, number or rhyme) during the discharge, which the patient is asked to recall, is used for assessing cognitive impairment. Thus, cognitive impairment is examined in an all or none fashion or recall or not recall fashion. Marked impairment of consciousness is unlikely to be missed even in routine non-video EEG recordings. However, in less severe cases, where impairment of consciousness is mild, the patient often successfully recalls the verbal test stimulus given by the technician during the discharge (Figure 1.2). This results in misinterpretation of the electrical event as a larval or subclinical event. Further, in brief discharges the verbal test stimulus is often given at the end or after the discharge has terminated. It is practically impossible to make an accurate judgement on this without appropriate video–EEG recordings.

Commentary on the Guidelines for the Use of EEG Methodology in the Diagnosis of Epilepsy

The recently published Guidelines for the Use of EEG Methodology in the Diagnosis of Epilepsy by the International League Against Epilepsy (ILAE) Commission on European Affairs: Subcommission on European Guidelines21 is of doubtful significance. Firstly, it emphasises aspects relevant to the old paper EEG recordings paying little attention to digital EEG. Secondly, it recommends all-night sleep deprivation for adults and recording during sleep only. Thirdly, no emphasis is placed on the significance and means of detecting clinical events during EEG discharges. Finally, it does not provide a contribution on the significance of video–EEG recording in epilepsies.

It is well documented that, with appropriate psychological testing, brief generalised spike or polyspike and slow wave discharges of 3–4 Hz are often associated with momentary impairment of cognitive function, even when absences are not clinically apparent.18–20 However, these psychological tests are in general not user-friendly and therefore unsuitable for routine clinical practice, which is the reason that they are not used in clinical EEG departments.

Breath counting is a simple modification of the routine EEG technique of hyperventilation, which is easily performed by any patient who can count irrespective of age and intelligence. The technician demonstrates the technique to the patient. The patient counts each deep breath at its expiration phase loudly and consecutively.

Breath counting allows accurate identification of even mild transient cognitive impairment during GSWDs.17 This is manifested as slurring of the speech, cessation, delay, hesitation, errors in counting with repetitions and counting out of sequence (Figures 1.2 and 2.12.3). Some patients may take two consecutive breaths during or immediately after the GSWD, but they count only the last of the two breaths.

Figure 2.3. The Significance of Recording the EEG in the Awakening Stage.

Figure 2.3

The Significance of Recording the EEG in the Awakening Stage. From a video–EEG recording of a 23-year-old woman with IGE and absence seizures. She was thought to be free of seizures. A long EEG the previous day, (more...)

Clinical note

Breath counting is powerful in detecting transient cognitive abnormalities because of simultaneously testing attention, concentration, memory, sequential precision and language function. The patient’s performance during breath counting acts as its own control.

Intermittent Photic Stimulation

Intermittent photic stimulation is significant for the detection of photosensitive patients (as detailed in Chapter 13. Photoparoxysmal discharges, which are often generated in the occipital regions, indicate a genetically determined photosensitivity and may occur in more than 1% of healthy subjects.

Other Forms of Appropriate Activation of Reflex Seizures

Other forms of appropriate activation should be used and they are as much fascinating as rewarding in patients with reflex seizures such as reading, pattern, musicogenic, proprioceptive and noogenic epilepsy. Their detection is of significance with regard to diagnosis and management. The avoidance of precipitating factors may be all that is needed in certain patients with reflex seizures.

Drowsiness, Sleep and Awakening

Drowsiness, sleep and awakening are important to study with an EEG in patients with epileptic disorders, particularly in those who produce a normal routine awake stage EEG or in those where their seizures are consistently associated with these physiological stages. However, drowsiness and sleep are associated with dramatic physiological EEG changes, which may imitate epileptogenic paroxysms. Their interpretation should be left only to highly experienced clinical neurophysiologists, otherwise significant errors are inevitable. In sleep stage EEGs, recording should always continue to include the awakening stage. It is well known that seizures and EEG paroxysms may only occur at this stage in certain epileptic syndromes such as IGE.

Important note

Partial instead of all-night sleep deprivation should be preferred.

The recording should continue to include the awakening stage, which is an uncommon practice in most EEG laboratories.

Sleep stage EEG recording is routinely applied for the investigation of patients with a suspected or established diagnosis of epileptic seizures. This is mainly performed because of (1) an alert EEG being normal or equivocal, (2) suspected nocturnal epileptic seizures or (3) possible activation by sleep that may be important in diagnosis, such as in benign childhood focal seizures or suspected electrical status epilepticus during sleep.

The routine practice in most EEG departments is to ask for an all-night sleep deprivation EEG. The EEG is performed the next morning with the aim of obtaining a good sleep stage EEG. As soon as this is achieved the recording is stopped, the patient is awakened and allowed to go home. This practice surprisingly is also recommended in ILAE guidelines.21 However, all-night sleep deprivation is inconvenient for the patient and often for the whole family and may induce seizures in susceptible individuals, particularly after leaving the EEG department in the awakening period. Patients with juvenile myoclonic epilepsy (JME) are particularly vulnerable because seizures mainly occur on awakening after sleep deprivation.

Drug-induced sleep is also applied in some departments as a substitute for all-night sleep deprivation. Usually, quinalbarbitone, chloral hydrate and, more recently, melatonin is given to the patient prior to or during electrode placement, with the aim of obtaining a sleep stage EEG after which the same procedure as above is followed. There is no requirement for an EEG on awakening. However, drugs that induce sleep may interfere with normal patterns and patients may find it difficult to be sufficiently alert for the rest of the day. They may also have a seizure on awakening as above.

My practice is to perform a sleep EEG that is as close as possible to the natural state and habits of the patient and thereby achieve best results with minimal discomfort and minimal risk to the patient. This is achieved with partial sleep deprivation, which is a practical, more natural, less disturbing and equally rewarding approach.22 The EEG is recorded during the awake, sleep and awakening stages.

In order to achieve satisfactory results, we ask patients to go to sleep 1–2 h later and wake 1–2 h earlier than their routine practice. However, we do not apply this rigidly, but try to adjust it to each patient’s sleep habits. Some patients find it easy to go to sleep whereas others find it difficult. We tell them that our aim is to obtain a natural sleep stage EEG and ask them to go to bed later than usual and wake up very early in the morning and remain awake until their appointment time, which we organise at 13:30 h for adults and at 11:30 h for children. The patient is asked to lie down and relax in a darkened and quiet recording room after application of the EEG electrodes. We allow 30–60 min sleep recording depending on the depth of sleep and EEG abnormalities. Subsequently, the patient is awakened and when alerted sat up and the EEG continues, also including hyperventilation and intermittent photic stimulation. This last phase on awakening lasts for approximately 15 min also depending on EEG interictal and ictal abnormalities. The whole procedure is performed with video–EEG recording, particularly for patients in whom minor or major seizures are expected on clinical or previous EEG grounds.

We have audited this method and found that 93.3% of the patients reached stage II–III and occasionally stage IV of sleep and in all patients EEG recording successfully continued on awakening. The other 6.7% failed to sleep although in some of them stage I was achieved.

In our practice, we have numerous examples of patients with clinical, mainly brief, minor seizures documented only in the awakening state, while their sleep stage video–EEG recording was normal or showed clinically asymptomatic generalised discharges.

The EEG Recording Should Be Tailored to the Specific Circumstances of the Individual Patient

  • The technician should be alerted to apply the appropriate stimulus when reading or other forms of reflex epilepsy are clinically suspected.
  • Patients with IGE may have a normal or non-specifically abnormal routine EEG. In these patients an EEG after partial sleep deprivation with video–EEG recording during sleep and on awakening frequently reveals clinical and EEG ictal events.

The same applies to patients with nocturnal seizures who may have a normal EEG while awake.

  • Women with catamenial seizures should have an EEG during their vulnerable periods if EEGs at other times are inconclusive.

The Role of EEG Technologists

The principal responsibility of EEG technologists is a competent EEG recording and a factual report. However, their role should be more than this, when considering the following.

  • Currently approximately 70% of EEG referrals are for epileptic disorders.
  • Referrals commonly come from general paediatricians/general physicians who may not be familiar with the syndromic diagnosis of epilepsies.
  • Information in the request form is usually inadequate.
  • An EEG technologist spends 15–20 min preparing the patient for the recording, which may be valuably used for obtaining information about such things as minor seizures, precipitating factors and the circadian distribution and other aspects of the particular individual. The interpretation of the EEG depends on a patient’s clinical history, which often is poor or missing.
  • Currently non-medical health professionals such as nurses are rightly involved in the management of epilepsies.

A well-qualified EEG technician is expected and should be trained to have a thorough knowledge of seizures and epileptic syndromes.11

In my department the EEG technologists often provide me with the correct syndromic diagnosis of our patients based on such a dual approach. Even interpretation of a normal EEG may be significantly different according to clinical information (see the illustrative cases on page 40).

Digital EEG

Digital EEG is the paperless recording of an EEG using computer-based instrumentation. The data are stored on electronic media, such as magnetic drives or optical disks and displayed on a monitor. Digital EEG recording has many advantages compared to analogue and paper EEGs, including retrospective reformatting (without the need for recording new data), storage, automatic event detection, quantification and networking capabilities.26–30 Reading of the EEG record with user-selected montages, filters, vertical scaling (gain/sensitivity) and horizontal scaling (time resolution or compression) allows for more accurate interpretation. Digital EEG replaces the need to warehouse or microfilm paper records, enables optional additional EEG signal processing and allows for electronic exchange of EEGs. Producing EEG figures is also notoriously difficult from paper EEGs particularly if blue ink was used. Most of the figures in this book were easily reproduced from digital EEGs.

Video–EEG Recording Should Be Made Routine Practice

Video–EEG recording should be mandatory in the evaluation of patients suspected of having seizures because it is the only means of reaching an incontrovertible diagnosis if clinical events occur during the recording. These may incidentally occur during the EEG or be predictably recorded based on their circadian distribution and the precipitating factors. Video–EEG machines are relatively inexpensive today with advances in digital compression and storage technology. Cost can be reduced to a minimum by using a commercially available camcorder synchronised with the EEG.

An EEG discharge is of great diagnostic and management significance if it is associated with clinical manifestations. However, these symptoms may be minor and escape recognition in routine EEGs without video recordings (Figures 1.2 and 2.12.3). Video–EEG recordings are particularly important in the identification and categorisation of absences, which are easily elicited by hyperventilation, myoclonic jerks or focal seizures which may be inconspicuous and psychogenic or other non-epilepic seizures (Figure 1.1) particularly those of the hyperventilation syndrome.31

Seizures or other paroxysmal events may occur at any stage during the EEG. Therefore, it is advisable to start and continue video recording during the whole EEG procedure. Vasovagal attacks often occur during EEG electrode placement. Psychogenic or fraudulent non-epileptic seizures often happen at the end of an EEG while removing the electrodes, particularly when the patient is told that the EEG is normal. Other types of non-epileptic seizures such as paroxysmal kinesiogenic choreoathetosis may also be captured with video–EEG recording and prompt the correct diagnosis.

An EEG Report Should Be Helpful and Committed: It Should Not Be an Abbreviated Factual Report

One of the most important sources of error in EEGs is that the reporting clinical neurophysiologist prefers a convenient but rather unhelpful and uncommitted approach, which is often an abbreviation of the factual technical report: ‘normal EEG’, ‘an abnormal EEG with generalised discharges of frontal origin’, ‘there is an active spike in the occipital regions’, ‘focal episodic left temporal slowing without genuine epileptiform activity’ and so forth. This is inadequate, often uninformative and sometimes misleading. The receiving physician is often unfamiliar with these EEG terms and their significance. My approach is to provide as much information as possible supplementing the traditional conclusion with an opinion and often a comment, which improves the EEG contribution.

Illustrative Cases with a Normal Routine EEG Where the Clinical Information Requires a New Appropriately Tailored EEG

  • The normal routine EEG of a teenager with a single GTCS on awakening after significant sleep deprivation, fatigue and unaccustomed alcohol consumption.
    Patient note

    Conclusion: the EEG is normal but, because the patient’s seizure occurred on awakening after significant precipitating factors, we have arranged an EEG after partial sleep deprivation.

    The EEG was normal again, but after awakening there were brief (1–2 s), asymptomatic, GSWDs of 3–4 Hz during hyperventilation. These were consistent with the clinical impression of a low threshold to IGE. The patient was advised regarding precipitating factors and no drug treatment was given.

  • The normal routine EEG of an 8-year-old child with GTCS during sleep.
    Patient note

    Conclusion: the routine awake EEG is normal, but because this is a child with a nocturnal convulsive seizure, a sleep EEG is indicated.

    The sleep EEG showed centro-temporal spikes thus documenting the diagnosis of Rolandic epilepsy and securing an excellent prognosis.

  • The normal ictal and inter-ictal EEG of a 20-year-old man referred for frequent brief clusters of bizarre movements, which ‘sounds like pseudo-seizures’.
    Patient note

    Conclusion: the routine awake EEG is normal. However, in view of the history that the paroxysmal events mainly occurred during sleep, the technician allowed time for the patient to go to sleep, during which several of his habitual attacks were recorded with video–EEG recording. These were typical hypermotor epileptic seizures thus documenting the diagnosis of mesial frontal lobe epilepsy (supplementary sensory motor epilepsy).

  • The normal EEG of a 50-year-old man with ‘GTCS from age 12 years, but on remission for 10 years. Stop anti-epileptic medication?’
    Patient note

    Conclusion: the EEG is within normal limits. However, stopping medication is not recommended because, according to the information provided to the EEG technologist, the patient continues to have brief seizures. These consist of unilateral multicoloured and spherical visual hallucinations lasting for a few seconds to a minute often progressing to deviation of the eyes and head. They are identical to those occurring prior to his GTCS. The patient suffers from occipital epilepsy with visual seizures and secondarily GTCS, a situation frequently misdiagnosed as migra-lepsy.

    Despite this report medication was discontinued. Two months later the patient had GTCS at work and he lost his job.

Illustrative Cases with Abnormal EEGS Showing That the Reporting Physician May Make a Significant Contribution to the Correct Diagnosis and Management

  • An EEG with brief generalised discharges of spikes and waves in a 30-year-old man referred because of a first GTCS.
    Patient note

    Conclusion: the EEG is of good organisation with a well-formed alpha rhythm. It is abnormal because of brief generalised discharges of small spikes and waves of 3–4 Hz, which are facilitated by hyperventilation. These are not associated with any ictal clinical manifestations tested with video–EEG recording and breath counting.

    Opinion: the EEG abnormality indicates a low threshold to IGE. This is consistent with the clinical information that the recent GTCS occurred in the morning after sleep deprivation and alcohol consumption. The patient is not aware of absences or myoclonic jerks. There is a remote possibility that these abnormalities are due to frontal lesions32 or subependymal heterotopia (a distinct neuronal migration disorder associated with epilepsy).33 This may indicate the need for high-resolution MRI though I expect it to be normal.

    Comment: this patient may not need any drug treatment, but he should be advised regarding precipitating factors.

    MRI was normal and the patient did not have any other seizures in the next 10 years of follow-up.

  • A 35-year-old woman was referred because of prolonged confusional premenstrual episodes.
    Patient note

    Conclusion: the EEG is of good organisation with a well-formed alpha rhythm. It is suspiciously but not definitely abnormal because of a brief and inconspicuous generalised burst of larval spikes and theta waves.

    Opinion: the EEG abnormality is mild and not conclusive. We have organised an EEG during her vulnerable premenstrual period because the confusional episodes may be non-convulsive status epilepticus.

    This was performed and showed definite and frequent generalised discharges of spikes/multiple spikes and slow waves of 3–4 Hz associated with impairment of consciousness and eyelid flickering. No further confusional episodes occurred after treatment with valproate.

  • A 17-year-old man was referred because of a ‘single episode of loss of consciousness and convulsions. Epilepsy?’
    Patient note

    Conclusion: the EEG is of good organisation with a well-formed alpha rhythm. It is abnormal because of brief runs of monomorphic theta waves around the left anterior temporal regions.

    Opinion: the EEG abnormality is mild but definite, although it does not show conventional epileptogenic features. However, because it is strictly unilateral a high-resolution MRI is indicated. This is also mandated because, according to the information gathered by our EEG technologist, the recent convulsive episode was preceded by an ascending epigastric sensation and fear, which had also occurred in isolation several times over the previous 2 years. These raise the possibility of hippocampal epilepsy.

    MRI confirmed left hippocampal sclerosis and a sleep stage EEG showed a clear-cut sharp and slow wave focus in the left anterior temporal electrode.

  • An EEG with occipital spikes from a 6-year-old child referred because of ‘a prolonged episode of loss of consciousness with convulsions’.
    Patient note

    Conclusion: the EEG is of good organisation with a well-formed alpha rhythm, which is often interrupted by clusters of high-amplitude bi-occipital sharp and slow waves.

    Opinion: the EEG abnormality of occipital spikes is often associated with benign seizures in this age group. From the clinical description, this child may suffer from ‘Panayiotopoulos syndrome’, whereby seizures are often solitary or infrequent as detailed in the attached paper (I enclose a brief report if the referring physician is not aware of the condition). However, occipital paroxysms may also occur in 1% of normal children and even more frequently in children with congenital visual abnormalities (strabismus and amblyopia) and other conditions with or without seizures.

    Comment: this EEG should be interpreted in accordance with the clinical manifestations in this child.* In particular was the event nocturnal or diurnal? What were the symptoms that preceded the convulsions and what was their duration? Did he have autonomic disturbances, vomiting or eye deviation? Is this a normal child with normal vision and development? Please, let me know, as treatment may not be needed.

  • A 19-year-old student was referred because of ‘infrequent GTCS from age 13 but now in remission for 2 years. Stop medication?’
    Patient note

    Conclusion: the resting EEG is within normal limits, but hyperventilation elicited high-amplitude generalised spike and slow wave discharges of 3–4 Hz with a duration of 3–5 s. These were associated with significant errors during breath counting as documented with video–EEG recording.

    Opinion: the EEG documents that the patient suffers from IGE with active mild absence seizures.

    Comment: according to the information provided by the patient, she also occasionally has mild myoclonic jerks on awakening particularly during examination periods. These indicate that she suffers from JME and treatment should continue for many years to come.

  • A 5-year-old child was referred for an EEG because of ‘learning difficulties and absence seizures’.
    Patient note

    Conclusion: this is a long video–EEG recording during an awake period and naturally occurring sleep stages I–IV. The awake stage EEG shows frequent high-amplitude sharp and slow wave complexes mainly around the right central and right posterior parietal electrode. Frequently, this occurs simultaneously with a left-sided similar sharp and slow wave complex, but this is always of higher amplitude on the right. In sleep, these discharges appear continuously as in electrical status epilepticus.

    Opinion: this is a very abnormal EEG, which raises the possibility of Landau–Kleffner syndrome. There is not the slightest evidence of absence seizures.

    On the basis of this EEG the child was appropriately assessed and found to suffer from Landau–Kleffner syndrome.

There are numerous similar examples that this type of communication between electroencephalographers and clinicians is essential for a better diagnosis and management of patients with epilepsies. The problems become even more complicated and demanding in the interpretation of EEGs from patients referred for possible epileptic seizures who also suffer from co-morbid conditions such as migraine, psychiatric diseases, cerebrovascular insufficiency and so forth and who may also be on various medications. In these cases it is often important to admit that ‘the EEG, although abnormal, may be misleading in view of the migraine and psychiatric or previous head injuries of the patient. The EEG abnormality cannot be taken as evidence or not of epilepsy’.

The Significance of the EEG after the First Afebrile Seizure

Clinical note

Routine EEG recording is a standard recommendation of the diagnostic evaluation of a child after a first afebrile seizure.34 This comes from the Quality Standards Subcommittee of the American Academy of Neurology based on analysis of evidence.34 However, in the UK an EEG is not recommended after the first afebrile seizure, a practice that may have significant adverse implications in the correct diagnosis and management (see Panayiotopoulos10 and associated commentary6).

The First Seizure

Most of the epilepsies manifest with primarily or secondarily GTCS, which may herald the onset or occur long after the beginning of the disease. Studies on the prognosis and treatment of the ‘first seizure’ mainly refer to GTCS, although this may not be the first seizure in the patient’s life.35 Myoclonic jerks, absences and focal seizures are less dramatic but more important than GTCS for diagnosis. In one study, 74% of patients with newly identified unprovoked seizures had experienced multiple seizure episodes prior to their first medical contact.36

The recurrence rate after a first convulsive seizure varies from 27 to 81%, thereby reflecting significant differences in selection, treatment and methodological criteria.3,10

An abnormal EEG and particularly GSWDs have been reported as a consistent predictor of recurrence in all36–40 but one study which was in adults41 (see Panayiotopoulos3,10 for reviews). In a meta-analysis of 16 publications on the risk of recurrence after a first fit, seizure aetiology and EEG findings were the stronger predictors of recurrence.37 This was confirmed in another study of 407 children with a first unprovoked afebrile seizure.40 In idiopathic and cryptogenic seizures the EEG was the most important predictor of outcome with 52% risk of recurrence at 2 years in those with an abnormal EEG versus 28% in those with a normal EEG.40 The EEG showed specific abnormalities of focal spikes or GSWDs in 32.5% of 268 children after their first idiopathic seizure.39

Numerous studies of homogeneous patients have established markedly different remission rates and prognosis depending on the syndromic classification. For example, consider the excellent prognosis in febrile seizures, benign childhood focal seizures such as Rolandic or Panayiotopoulos syndrome, the lifelong liability to seizures or their worsening in JME, mesial TLE with hippocampal sclerosis, epileptic encephalopathies and so forth.

Why an EEG after the First Afebrile Seizure?

A convulsive seizure is a dramatic event in a child’s life and that of their family.42 As in all other fields of medicine, they are entitled to a diagnosis, prognosis and management, which is specific and precise.3,35 Aiming to this goal, an EEG after the first seizure is essential.34

That an epileptiform EEG is associated with a two to three times higher risk for recurrence than a normal EEG is well established.3,10,36–40 However, the most significant reasons for having an EEG after a single afebrile convulsion are fourfold.

Firstly, it is possible to recognise children with the features of specific epileptic syndromes. Between 10% and 40% of children with benign childhood focal seizures may not have had more than a single fit, thus depriving them of a precise diagnosis and prognosis under the current practice, in some counties, of not having an EEG after the first seizure. On other occasions, symptomatic epilepsies may be established requiring early attention.

Secondly, minor seizures such as absences, myoclonic jerks or focal fits may be recorded, which have immense diagnostic and treatment implications.

Thirdly, the EEG is essential in establishing seizure-precipitating factors such as video games or television, thus leading to early and appropriate advice.

Fourthly, an EEG in an untreated stage of an epileptic syndrome is imperative. This is most likely to happen if the EEG is requested after the first seizure. Many paediatricians would be reluctant to withhold treatment after a second or more seizures which are expected to occur in one-quarter of children within 3 months after their first fit. Requesting an EEG at that stage may be too late, considering that EEG waiting lists may sometimes be lengthy. Masking or altering the EEG with AEDs may prevent a correct seizure or syndrome diagnosis. This in turn will be detrimental for management, which may be long term, and expensive in terms of medication, which is often seizure specific.


Haas LF. Hans Berger (1873–1941), Richard Caton (1842–1926), and electroencephalography. J Neurol Neurosurg Psychiatry. 2003;74:9. [PMC free article: PMC1738204] [PubMed: 12486257]
Binnie CD, Stefan H. Modern electroencephalography: its role in epilepsy management. Clin Neurophysiol. 1999;110:1671–97. [PubMed: 10574283]
Panayiotopoulos CP. Benign Childhood Partial Seizures and Related Epileptic Syndromes. London: John Libbey & Company Ltd; 1999.
Niedermeyer E, Lopes da Silva F. Electroencephalography. Basic Principles, Clinical Applications, and Related Fields. 4. Baltimore: Williams & Wilkins; 1999.
Blume WT. Current trends in electroencephalography. Curr. Opin Neurol. 2001;14:193–7. [PubMed: 11262735]
Cross J. Significance of the EEG after the first afebrile seizure: commentary. Arch Dis Child. 1998;78:576–7. [PMC free article: PMC1717611] [PubMed: 9713020]
Binnie CD, Prior PF. Electroencephalography. J Neurol Neurosurg Psychiatr. 1994;57:1308–19. [PMC free article: PMC1073178] [PubMed: 7964803]
Sand T. Electroencephalography in migraine: a review with focus on quantitative electroencephalography and the migraine vs. epilepsy relationship. Cephalalgia. 2003;23(Suppl 1):5–11. [PubMed: 12699455]
Chadwick D. Diagnosis of epilepsy. Lancet. 1990;336:291–5. [PubMed: 1973981]
Panayiotopoulos CP. Significance of the EEG after the first afebrile seizure. Arch Dis Child. 1998;78:575–6. [PMC free article: PMC1717611] [PubMed: 9713020]
Sanders S, Rowlinson S, Manidakis I, Ferrie CD, Koutroumanidis M. The contribution of the EEG technologists in the diagnosis of Panayiotopoulos syndrome (susceptibility to early onset benign childhood autonomic seizures) Seizure. 2004;13:567–73. [PubMed: 15519916]
Koutroumanidis M, Binnie CD, Elwes RD, et al. Interictal regional slow activity in temporal lobe epilepsy correlates with lateral temporal hypometabolism as imaged with 18FDG PET: neurophysiological and metabolic implications. J Neurol Neurosurg Psychiatry. 1998;65:170–6. [PMC free article: PMC2170184] [PubMed: 9703166]
American Electroencephalographic Society. Guideline one: minimum technical requirements for performing clinical electroencephalography. J Clin Neurophysiol. 1994;11:2–5. [PubMed: 8195422]
American Electroencephalographic Society. Guideline two: minimum technical standards for pediatric electroencephalography. J Clin Neurophysiol. 1994;11:6–9. [PubMed: 8195427]
Takahashi T. Activation methods. In: Niedermeyer E, Lopes da Silva F, editors. Electroencephalography. Basic Principles, Clinical Applications, and Related Fields. Baltimore, MD: Williams & Wilkins; 1999. pp. 261–84.
Panayiotopoulos CP, Baker A, Grunewald RA, Rowlinson S. Breath counting during 3 Hz generalized spike and wave discharges. J Electrophysiol Technol. 1993;19:15–23.
Giannakodimos S, Ferrie CD, Panayiotopoulos CP. Qualitative and quantitative abnormalities of breath counting during brief generalized 3 Hz spike and slow wave ‘subclinical’ discharges. Clin Electroencephalogr. 1995;26:200–3. [PubMed: 8575099]
Aarts JH, Binnie CD, Smit AM, Wilkins AJ. Selective cognitive impairment during focal and generalized epileptiform EEG activity. Brain. 1984;107:293–308. [PubMed: 6421454]
Provinciali L, Signorino M, Censori B, Ceravolo G, Del Pesce M. Recognition impairment correlated with bisynchronous epileptic discharges. Epilepsia. 1991;32:684–9. [PubMed: 1915177]
Mirsky AF, Duncan CC, Levav LM. Neuropsychological and psychophysiological aspects of absence epilepsy. In: Duncan JS, Panayiotopoulos CP, editors. Typical Absences and Related Epileptic Syndromes. London: Churchill Communications Europe; 1995. pp. 112–21.
Flink R, Pedersen B, Guekht AB, et al. Guidelines for the use of EEG methodology in the diagnosis of epilepsy. International League Against Epilepsy: commission report. Commission on European Affairs: Subcommission on European Guidelines. Acta Neurol. Scand. 2002;106:1–7. [PubMed: 12067321]
Peraita-Adrados R, Gutierrez-Solana L, Ruiz-Falco ML, Garcia-Penas JJ. Nap polygraphic recordings after partial sleep deprivation in patients with suspected epileptic seizures. Neurophysiol Clin. 2001;31:34–9. [PubMed: 11281068]
Panayiotopoulos CP. Panayiotopoulos Syndrome: A Common and Benign Childhood Epileptic Syndrome. London: John Libbey & Company; 2002.
Panayiotopoulos CP, Agathonikou A, Sharoqi IA, Parker AP. Vigabatrin aggravates absences and absence status. Neurology. 1997;49:1467. [PubMed: 9371946]
Panayiotopoulos CP. Efficacy of lamotrigine monotherapy. Epilepsia. 2000;41:357–9. [PubMed: 10714412]
American Electroencephalographic Society. Guideline fourteen: guidelines for recording clinical EEG on digital media. J Clin Neurophysiol. 1994;11:114–15. [PubMed: 8195415]
Levy SR, Berg AT, Testa FM, Novotny EJ, Chiappa KH. Comparison of digital and conventional EEG interpretation. J Clin Neurophysiol. 1998;15:476–80. [PubMed: 9881918]
Blum DE. Computer-based electroencephalography: technical basics, basis for new applications, and potential pitfalls. Electroencephalogr Clin Neurophysiol. 1998;106:118–26. [PubMed: 9741772]
Nuwer MR, Comi G, Emerson R, et al. IFCN standards for digital recording of clinical EEG. The International Federation of Clinical Neurophysiology Electroencephalogr. Clin Neurophysiol Suppl. 1999;52:11–14. [PubMed: 10590972]
Scherg M, Ille N, Bornfleth H, Berg P. Advanced tools for digital EEG review: virtual source montages, whole-head mapping, correlation, and phase analysis. J Clin Neurophysiol. 2002;19:91–112. [PubMed: 11997721]
Ferrie CD, Agathonikou A, Panayiotopoulos CP. Electroencephalography and video–electroencephalography in the classification of childhood epilepsy syndromes. J R Soc Med. 1998;91:251–9. [PMC free article: PMC1296700] [PubMed: 9764078]
Ferrie CD, Giannakodimos S, Robinson RO, Panayiotopoulos CP. Symptomatic typical absence seizures. In: Duncan JS, Panayiotopoulos CP, editors. Typical Absences and Related Epileptic Syndromes. London: Churchill Communications Europe; 1995. pp. 241–52.
Raymond AA, Fish DR, Stevens JM, Sisodiya SM, Alsanjari N, Shorvon SD. Subependymal heterotopia: a distinct neuronal migration disorder associated with epilepsy. J Neurol Neurosurg Psychiatry. 1994;57:1195–202. [PMC free article: PMC485486] [PubMed: 7931380]
Hirtz D, Ashwal S, Berg A, et al. Practice parameter: evaluating a first nonfebrile seizure in children: report of the quality standards subcommittee of the American Academy of Neurology, The Child Neurology Society, and The American Epilepsy Society. Neurology. 2000;55:616–23. [PubMed: 10980722]
King MA, Newton MR, Jackson GD, Berkovic SF. Epileptology of the first-seizure presentation: a clinical, electroencephalographic, and magnetic imaging study of 300 consecutive patients. Lancet. 1998;352:1007–11. [PubMed: 9759742]
Hauser WA, Rich SS, Annegers JF, Anderson VE. Seizure recurrence after a 1st unprovoked seizure: an extended follow-up. Neurology. 1990;40:1163–70. [PubMed: 2381523]
Berg AT, Shinnar S. The risk of seizure recurrence following a first unprovoked seizure: a quantitative review. Neurology. 1991;41:965–72. [PubMed: 2067659]
Hart YM, Sander JW, Johnson AL, Shorvon SD. National General Practice Study of Epilepsy: recurrence after a first seizure [see comments] Lancet. 1990;336:1271–4. [PubMed: 1978114]
Shinnar S, Kang H, Berg AT, Goldensohn ES, Hauser WA, Moshe SL. EEG abnormalities in children with a first unprovoked seizure. Epilepsia. 1994;35:471–6. [PubMed: 8026390]
Shinnar S, Berg AT, Moshe SL, et al. The risk of seizure recurrence after a first unprovoked afebrile seizure in childhood: an extended follow-up. Pediatrics. 1996;98:216–25. [PubMed: 8692621]
Hopkins A, Garman A, Clarke C. The first seizure in adult life. Value of clinical features, electroencephalography, and computerised tomographic scanning in prediction of seizure recurrence. Lancet. 1988;1:721–6. [PubMed: 2895259]
Hoekelman RA. A pediatrician’s view. The first seizure – a terrifying event [editorial] Pediatr Ann. 1991;20:9–10. [PubMed: 1900359]



Note: The EEG technologist should have obtained this information (p. 38).

Copyright © 2005, Bladon Medical Publishing, an imprint of Springer Science+Business Media.
Bookshelf ID: NBK2601