Electroencephalographic (EEG) Video Monitoring: Health Technology Assessment Reports, 1990 Number 4

Foreword

The Office of Health Technology Assessment (OHTA) evaluates the safety and effectiveness of new or unestablished medical technologies that are being considered for coverage under federally financed health programs. These assessments are performed at the request of the Health Care Financing Administration (HCFA) and the Office of Civilian Health and Medical Program of the Uniformed Services (OCHAMPUS). They are the basis for recommendations to these programs regarding coverage policy decisions.

Questions about coverage for certain health care technologies are directed to HCFA and OCHAMPUS by such interested parties as insurers, manufacturers, contractors, and practitioners. Those questions of a medical, scientific, or technical nature are formally referred to OHTA for assessment.

OHTA's assessment process includes a comprehensive review of the medical literature and emphasizes broad and open participation from within and outside the Federal Government. A range of expert advice is obtained by widely publicizing the plans for conducting the assessment through publication of an announcement in the Federal Register and solicitation of input from Federal agencies, medical specialty societies, insurers, and manufacturers. The involvement of these experts helps assure inclusion of the experienced and varying viewpoints needed to round out the data derived from individual scientific studies in the medical literature.

After OHTA receives information from experts and the scientific literature, the results are analyzed and synthesized into an assessment report. Each report represents a detailed analysis of the safety, clinical effectiveness, and uses of new or unestablished medical technologies considered for coverage. These Health Technology Assessment Reports form the basis for the Public Health Service recommendations to HCFA and OCHAMPUS and are disseminated widely. Individual reports are available to the public once a coverage decision regarding the subject technology has been made.

OHTA is part of the Agency for Health Care Policy and Research (AHCPR), Public Health Service, Department of Health and Human Services.

• Thomas V. Holohan, M.D.

• Director

• Office of Health Technology Assessment

• J. Jarrett Clinton, M.D.

• Acting Administrator,

• AHCPR Assistant Surgeon General

• Copies may be obtained at no charge from:

• Division of Information and Publications

• AHCPR

• 5600 Fishers Lane

• Parklawn Building, Room 18-12

• Rockville, MD 20857; (301) 443-4100

Introduction

Electroencephalographic (EEG*) video monitoring is a technique which provides simultaneous documentation of the behavioral and electroencephalographic manifestations of seizures. Electroencephalographic monitoring, a common diagnostic procedure that records brain wave patterns, has been routinely used in the diagnosis and classification of epilepsy and allied seizure disorders for many years. Investigators and clinicians have reported that EEG video monitoring results in significantly expanded EEG diagnostic capability.

The standard or routine EEG examination is performed in a clinician's office or laboratory environment and usually lasts from 20 minutes to 1 hour. The examination is usually undertaken in quiet, relaxed circumstances with the patient seated or lying comfortably with eyes closed. When the routine part of the examination has been completed, recording continues while activation procedures such as sleep, sleep deprivation, light stimulation, and hyperventilation are undertaken in an attempt to provoke abnormalities.(1)

The ultimate goal of the routine EEG examination is to record and identify abnormal (epileptiform) electrical discharges to assist in making the clinical diagnosis. The abnormal discharges may occur as isolated events (interictal or between seizures) or in association with clinical symptoms (ictal or during a seizure). An ictal event (seizure) seldom occurs during the routine EEG, which usually lasts less than an hour. Therefore, when an EEG abnormality is recorded, it is typically interictal.

The presence of interictal epileptiform discharges in the EEG of a patient with suspected epilepsy does not establish the diagnosis beyond doubt, since approximately 10 percent of the normal population will have some abnormal discharge in their EEG. The diagnosis of epilepsy remains a clinical, not an electroencephalographic, diagnosis. However, epileptic phenomena can in most cases be diagnosed when the interictal EEG is combined with a history, a neurologic examination, and an accurate description of the seizures by the patient and/ or others.(2,3,4) In some patients, the distinction between epileptic seizures and nonepileptic seizures may be difficult to make. According to Gumnit, the term epileptic seizure refers to a sudden change in the electrical activity of the brain, usually accompanied by changes in behavior.(5) Nonepileptic seizures are sudden changes of behavior which are not caused by an independent sudden change in the electrical activity of the brain. Nonepileptic seizures can be divided into psychogenic seizures and physiological events not involving the central nervous system, such as syncope and cardiac standstill. Gumnit uses the term psychogenic seizures to include malingering, sudden episodic events related to conversion reactions, misinterpretation of normal physiological stimuli, and behavioral changes secondary to psychological conflict. Gumnit recommends the use of the term psychogenic seizure rather than the term pseudoseizure, which is often used by authors to describe nonepileptic seizures.

Ambulatory cassette EEG monitoring was developed as an extension of the routine EEG examination. The new technology allows a patient to wear an entire monitoring system throughout the day. This outpatient technique relies on the capability of a portable cassette recorder to continuously record brain wave patterns during a 24-hour period. Investigators and clinicians have reported that ambulatory cassette EEG monitoring for 24 hours or longer significantly expands EEG diagnostic capability.(6).

In some difficult cases of suspected epilepsy, more precise information may be required for an accurate diagnosis and effective treatment. In these situations, EEG video monitoring performed in the hospital or in epilepsy centers has been used to help make the differential diagnosis of epileptic versus nonepileptic seizure. For epileptic patients whose seizures have remained uncontrolled, the technique has been used to classify the epileptic seizure for proper diagnosis and treatment. Moreover, EEG video monitoring has been used to establish the diagnosis and focality of an epileptic disorder for surgical intervention.(3)

In EEG video monitoring at least 10 EEG channels record a patient's brain-wave patterns. A simultaneous video picture of the patient is displayed and antiepileptic drug levels are monitored, if indicated. The EEG is displayed on the video screen along with face and whole-body views of the patient, thus permitting the correlation of electrical with clinical manifestations of the seizure. The primary significance of EEG video monitoring lies in its potential for improving the diagnosis and treatment of patients with intractable seizures (patients with unsatisfactory seizure control). This report assesses the safety and clinical effectiveness of this procedure in the diagnosis and treatment of seizure disorders.

Background

One significant finding in the search for a treatment of epilepsy was the discovery that clinical (epileptic) seizures are accompanied by abnormal electrical discharges in the brain. In the majority of cases, these discharges can be identified by measuring the changes in the electrical field on the surface of the scalp. Because the pattern of discharges can be diagnostic, the use of electrodes attached to the scalp to evaluate electrical discharges in the brain has become routine.(7) The clinical use of the EEG to study convulsive disorders depends not so much on the occurrence of EEG changes during seizures, but rather on transient or continuous deviations in interseizure records.(8) The interictal EEG is abnormal in many patients and exhibits features that help to establish a diagnosis.

An EEG may be defined operationally as a graphic representation of the varying differences in electrical potential between two points on the surface of the head, plotted against time. Electrodes are placed on various parts of the head and the potential difference between any two is amplified and recorded on continuously moving paper. In standard EEG examinations, 8, 10, 12, or 16 pairs of electrodes are used for simultaneous recording on multichannel instruments.(8) Current practice is to employ scalp electrodes on the patient's head using the international 10-20 system. Multiple channel recordings provide spatial analysis of the EEG and several simultaneous readings of an abnormal event that aid in distinguishing the abnormal event from an artifact, or "normal "

Routine EEG examinations are frequently normal between seizure episodes. Some authors claim that from 30 to 70 percent of patients, depending on the type of epilepsy, show abnormal EEGs during routine recordings.(7,9) Increased detection of epileptiform abnormalities can be achieved by repeating routine EEG examinations and by using techniques to stimulate electrical responses. These techniques result in EEG studies that reveal abnormalities in the majority of seizure patients.

Some patients with suspected epilepsy fail to demonstrate ictal or interictal abnormalities even after repeated routine recordings. The ability of the ambulatory cassette monitor to record EEG for prolonged periods of time has been shown to increase the likelihood of obtaining an abnormal EEG during a seizure. It also provides reliable data for evaluating some patients with suspected syncope, transient ischemic attacks, psychogenic seizures, and poorly defined seizure disorders. An ambulatory EEG is indicated when a patient has episodes suggestive of epilepsy and when the history, examination, and routine EEG do not resolve the diagnostic uncertainties.(6).

Studies of the treatment of intractable seizures indicate that more complex cases may require data obtained with the intensive neurodiagnostic technique of EEG video monitoring. In these cases the distinction between epileptic seizure disorders and other episodic problems that may not be epileptic in nature are difficult to make. For example, generalized tonic-clonic and complex partial seizures are types of epileptic seizures often confused with psychogenic attacks. EEG video monitoring at hospitals and specialized epilepsy centers has been shown, in recent years, to be capable of resolving the confusion. Before a patient is selected for EEG video monitoring, it is necessary to evaluate the adequacy of previous diagnostic studies and drug therapy.

Bowden and coworkers (1975) found that the use of prolonged EEG recordings doubled the number of epileptic patients showing abnormal discharges.(10) The utility of continuous EEG recording has been also enhanced by the simultaneous use of closed-circuit television to provide a video recording of behavior.(11). Because the technique requires both a high degree of sophistication in equipment and extensive use of manpower, only a few centers have offered this type of intensive monitoring. Monitoring areas may be dedicated recording suites with trained technologists in attendance. Some settings simulate a home environment with a living room, a dining area, and toilet facilities. Seizure alarm buttons may be available to the patients, visiting relatives, and the staff to signal the occurrence of events that might represent epileptic and nonepileptic seizures. Camera and microphone arrangements allow nearly continuous observation of the patient who engages in activities such as conversation, television viewing, games, sleeping, and eating. However, the cameras are set for fixed locations, hampering free movement and requiring the patient to sit or lie in selected spots. Self-tracking video monitoring systems are now being introduced which will allow the cameras to follow the patient.

Prolonged EEG recordings with telemetry give the patient greater freedom. The patient may be connected to the EEG machine by means of a cable or by radiotransmitter. This approach permits the recording of patients under less stringent conditions; however, the use of telemetry increases the potential for recording artifacts.(12). Telemetry can be used effectively with simultaneous video recordings of the patient, with visual observation of the patient, or alone.

Portable equipment that can be used for bedside monitoring is also available. Gumnit, however, questions the quality of data collected with portable equipment in an uncontrolled environment without the constant attendance of a trained technologist.(5) EEG video monitoring, lasting from hours to weeks, usually entails continuous, prolonged EEG recordings with simultaneous closed-circuit television monitoring of the patient. The two must be closely synchronized, or the information is not useful.(5) Both the EEG and television pictures of the patient are recorded on video tape. According to Gumnit it is best if a split screen technique is used to produce a close up of the face as well as a full view of the body. He found that most institutions specializing in treating patients with intractable seizures or epilepsy surgery would benefit from having the ability to superimpose EEG upon a full figure and facial closeup. However, this method has the disadvantage of limiting the number of EEG channels that can be clearly displayed. The alternative is to record the two sets of data separately and provide a common and easily readable time code which appears on both the video screen and the EEG paper record. Regardless of the approach, the tape can be replayed to study a seizure. Data reduction methods including computer programs developed to recognize and selectively record electrographic seizures are an important advance in prolonged EEG video monitoring.

At present, epileptic seizures are classified by their clinical manifestations. Precise anatomic and pathophysiologic correlates of specific ictal behaviors are still largely unknown. According to Engel, classification plays an essential role in the diagnosis and management of epileptic disorders.(13) Diagnosis of a specific syndrome may have important therapeutic and prognostic implications. Because the choice of an antiepileptic drug is determined by the seizure type, the differential diagnosis is essential for determining appropriate drug therapy. Before seizures are deemed intractable it is necessary to be certain that the correct drugs have been used in the correct amounts.

Investigators and clinicians have reported that EEG video monitoring results in significantly expanded EEG diagnostic capability and is clinically effective and necessary in some patients. Many studies have been published reporting the usefulness of this technique, and medical specialty groups have provided recommendations for its use.

Rationale

The principal goal of long-term EEG video monitoring in patients with seizures is to provide simultaneous documentation of the clinical manifestations correlated to the EEG changes during the patient's ictal events or seizures. Proponents of EEG video monitoring believe that this technique is clinically useful for those patients for whom a diagnosis could not be made on the basis of a neurological examination, routine EEG studies, and ambulatory cassette EEG monitoring. They believe that this technique, once a research tool, has proved valuable in the management of difficult diagnostic and therapeutic problems. Video coverage of the patient's behavior changes correlated with the EEG provides important information otherwise unavailable.

According to proponents, if a patient has a disorder which is not due to epilepsy, the establishment of the correct diagnosis by EEG video monitoring can avoid the continued use of an unnecessary medication and direct treatment towards a more appropriate form of care. Furthermore, if the technique documents epilepsy different from that diagnosed originally, a subsequent change in the medical regimen might result in improved seizure control. Proponents point out that if seizures cannot be controlled medically, EEG video monitoring can be integrated with other testing modalities in the evaluation of patients with intractable epilepsy considered for surgery.

Review of Available Literature

In 1980 Holmes et al used prolonged simultaneous EEG telemetry and video recordings (TEEG-VR) to study patients with frequent seizures refractory to standard medical treatment.(14) Using TEEG-VR the investigators studied the clinical characteristics of pseudoseizures. The pseudoseizures were defined as paroxysmal clinical events resembling epileptic seizures but not accompanied by preictal, ictal, or postictal changes on the EEG. Clinical characteristics of pseudoseizures and epileptic seizures documented by TEEG-VR were used to create useful differentiating criteria. According to Holmes et al several significant differences between pseudoseizures and epileptic seizures were noted. They observed combativeness, crying, yelling, and vulgar language only with pseudoseizures. The findings and outcomes of this and similar studies of pseudoseizures are summarized in Table 1.

Holmes et al found that differentiating pseudoseizures from epileptic seizures was difficult. Pseudoseizures often resemble epileptic seizures, especially in the case of partial seizures. In patients having epileptic seizures, pseudoseizures may cause the patient to report an increase in seizure frequency and lead the physician to change anticonvulsant therapy. Moreover, when pseudoseizures occur in a nonepileptic patient, the physician may mistakenly initiate anticonvulsant therapy. Although Holmes and associates believe that the combined application of EEG and video recording is a powerful diagnostic tool, they note that the occasional patient may still require depth electrode studies to establish the diagnosis. It is generally recognized that seizures which arise from foci deep in temporal lobe structures do not produce changes in a surface EEG.(15)

Porter et al followed patients with intractable epilepsy for up to 57 months in order to determine the long-term benefits of patient management based on information derived from the simultaneous recording of seizures with video tape and prolonged telemetered EEG.(16) The patients received video EEG monitoring for periods of 3 to 30 weeks (mean, 12 weeks). At followup patient outcomes were measured by improved seizure control, a reduction in the number of medications, and improved social adjustment. The results of this and other video monitoring studies of intractable seizures are presented in Table 2.

Penry and Porter estimated that less than 5 percent of all epileptic patients might require intensive monitoring.(17) They recommended intensive monitoring for the patient who has recurring seizures over a period of years, has failed to respond adequately to antiepileptic drugs singly or in combination, and has failed to respond to conventional outpatient treatment in addition to one or more hospitalizations.

Video EEG monitoring was used by Gulick et al to study the ictal clinical phenomena of pseudoseizures.(18) In 1982, they reported their analysis of ictal phenomena associated with 71 pseudoseizures recorded in 27 patients. Ictal phenomena included response to verbal stimuli, motor activity, and semipurposeful behavior. They found that individual phenomena and ictal patterns for pseudoseizures may simulate all of the major phenomena and patterns of epileptic seizures. With repeated playback of each episode, they found that many of the behavioral changes recorded in pseudoseizures resembled changes seen in epileptic seizures, but rarely did the complete pattern of observable phenomena closely resemble an epileptic seizure. However, pseudoseizures without motor activity were more difficult to differentiate from epileptic seizures than those with motor activity. The authors concluded that pseudoseizures include a broad spectrum of phenomena and may superficially simulate any of the major types of epileptic seizures.

At the Diagnostic, Treatment and Rehabilitation Unit of the University of Virginia Comprehensive Epilepsy Program, Sutula et al evaluated the efficacy of intensive monitoring in a series of patients with refractory seizures.(19) Many of the patients were receiving two or more antiepileptic drugs. Establishing a correct diagnosis, reducing seizure frequency, and reducing both the use of drugs and drug toxicity were the criteria used to measure the success of the program. The use of the technique resulted in a significant reduction in drug therapy received by each patient. Concomitant with this reduction, improved diagnostic accuracy and more appropriate drug therapy reduced the frequency of seizures in many patients (Tables 1 and 2). Although Sutula et al concluded the intensive monitoring was useful, they questioned whether a patient would function more effectively and independently in social, vocational, and educational spheres as a result of changes in diagnosis and treatment. While the average length of stay in the epilepsy unit for patients in this study was 8 weeks, the National Association of Epilepsy Centers has suggested that an average length of stay for an effective evaluation can be accomplished in 10 to 16 days.

Based on a 1981 study of 181 consecutive EEG and video monitoring records (of 139 patients), Binnie et al concluded that intensive monitoring provides useful information in the evaluation of patients with proven or suspected epilepsy.(20) They found that the most common indication for the use of EEG video monitoring was to distinguished epileptic from nonepileptic seizures. Other common indications for use of the technique included determination of seizure frequency and localization of an epileptic focus in patients considered for surgery. Information regarding seizure frequency is used to guide therapy, and the authors found this information, particularly with absence seizures, valuable for assessing therapy. Absence seizures (petit mal) are notable for their brevity and the paucity of motor activity.

In 1982, Desai et al at the Clinical Epilepsy Section, National Institute of Neurological and Communicative Disordes and Stroke (NINCDS) examined the value of intensive monitoring in diagnosing psychogenic seizures.(21) Two types of epileptic seizures most likely to be confused with psychogenic attacks are generalized tonic-clonic and complex partial seizures. These seizure types have a relatively limited range of expression, and the investigators believe a diagnosis-by-exclusion can be obtained by studying data collected with video monitoring. A diagnosis of epileptic or psychogenic seizure is made by comparing the characteristics of the video EEG recordings of the questionable seizure with those of known seizure types. The investigators believe the correct diagnosis of a seizure depends on the number of episodes and the experience of the investigator. Desai et al found that most epileptic seizures have similar history and findings, although some differ markedly from the usual pattern.

In the Desai study EEG video monitoring included determinations of plasma antiepileptic drug levels. Because of the relationship between epileptic seizures and the medication regimen, EEG changes associated with changes in regimen can be valuable in the differential diagnosis of psychogenic seizures. According to Desai et al, the withdrawal of medication is advantageous in a monitoring setting where an increase in seizure frequency allows more attacks to be recorded and increases the likelihood a correct diagnosis will be made, regardless of the seizure type.

Reports in the literature indicate that in some of the intensive monitoring studies, patients were tested for as many as 8 to 20 weeks.(16,19) In contrast, Luther et al used a 3-hour EEG video monitoring strategy in 30 patients with poorly controlled seizures to study medically intractable attacks.(22) In cases in which the nonepileptic seizure did not occur spontaneously, a seizure was induced by sequential activation techniques that included verbal suggestion, photic stimulation, hyperventilation, or intravenous saline. The investigators found that nonepileptic seizures could be induced easily during the monitoring session and that most nonepileptic seizure in this study population mimicked generalized seizures. There was no alteration of the background EEG prior to, during, or following the seizure. Data from this study do not support an earlier assumption that nonepileptic seizures occur primarily in patients with epilepsy. According to the investigators, EEG video monitoring combined with activation techniques and a detailed history helps establish the diagnosis of nonepileptic seizure. Table 3 provides a summary of outpatient EEG video monitoring studies.

A similar study of successful daytime (outpatient) EEG video monitoring over 6-8 hours, in conjunction with provocative measures, was conducted by Rowan et al.(23) They also found that daytime EEG video monitoring combined with activation techniques seemed to be an effective technique for identifying the type of seizure and for differentiating paroxysmal events of uncertain origin (Table 3. There is a growing interest in the use of outpatient EEG video monitoring over 3-12 hours for diagnostic purposes.

According to Perry et al, physicians can make a correct diagnosis in the vast majority of seizure cases on evidence provided by a routine interictal EEG and a reliable description of the patient's seizure.(3) For the patient with seemingly intractable epilepsy they found determining what treatment is most appropriate requires more precise information on the type of seizure, including the clinical and EEG pattern, the duration, and the frequency. Results from this retrospective study of patients with intractable epilepsy who had received both routine EEG studies and intensive monitoring are also reported in Table 1.

More recent studies tend to support the view that EEG video monitoring provides information which improves diagnosis. Eglie et al described clinical results from studies at the Swiss Epilepsy Center where patients are admitted with seizures of unknown etiology and with epileptic seizures unresponsive to drugs.(24) The benefits of EEG video monitoring were determined by studying the number of patients that subsequently became seizure-free or experienced a reduction in the frequency of their seizures (Tables 1 and 2).

Roberts and Fitch used EEG (telemetry) video monitoring at the National Hospital, Queen Square, London.(25) Monitoring of 102 patients involved a single 6-hour daytime or overnight session. The monitored interval depended upon the time of day when the risk of seizure was greatest. To increase the chances of recording an episode, patients were selected for monitoring based on the frequency of seizures. In one group patients were monitored in order to try to differentiate between epilepsy, cardiac arrhythmias, psychogenic attacks, and other episodic events. For those cases in which no events were recorded, the authors found that a normal sleep EEG was a useful negative finding in patients suspected of psychogenic episodes. A second group of patients was monitored to clarify the type of epileptic seizure or the site of an epileptic focus. Some of the patients in this latter group were potential candidates for surgical treatment; however, the use of eight EEG channels limited the ability to precisely localize the site of epileptic activity. Other patients who were monitored had conditions such as primary generalized epilepsy and sleep disorders. The experience at the National Hospital is presented in Tables 1 and 3.

The experiences of investigators at other specialized epilepsy centers have also been reported in the literature. As a followup to the study of EEG video monitoring reported in 1981, (20) Binnie and coworkers evaluated changes in patient management and outcome for 272 patients monitored by video EEG and followed for 12 months.(26) The EEG video monitoring provided clinically relevant information (i.e., occurrence of recorded seizures, focus identification) in 79 percent of the sessions. Information obtained during 65 percent of the sessions affected patient management (change in treatment regimen). Over 25 percent of the sessions provided information that led to long-term therapeutic benefits (e.g., reduced hospitalizations, return to school or work). Because focus identification is essential for neurosurgical intervention, monitoring was often prolonged and sleep deprivation and antiepileptic drug withdrawal were used to increase seizure frequency in these patients.

In a study reported by Willmer and Brunet, many patients had previous multiple hospital admissions for both assessment and treatment of seizures.(27) Results of patient monitoring by radiotelemetry EEG with simultaneous video recordings for up to 4 days are provided in Tables 1 and 2. According to the investigators, the high percentage of inconclusive postmonitoring diagnoses (31) could be reduced by either repeating and extending monitoring sessions or by a more aggressive use of activation procedures such as anticonvulsant medication withdrawal.

Ramani used verbal suggestions to enhance the possibility of provoking an attack during 4 to 6 hours of recordings or repeated sessions and anticonvulsant drug withdrawal for patients who failed to present in "ictal" recording.(28) Additionally, he recommended drug withdrawal even after successful recordings of a psychogenic seizure to avoid overlooking real seizures that may coexist. Some programs maintain the patient on at least one drug that is effective for this seizure to prevent an increase in generalized seizures during drug withdrawal (convulsive status epilepticus). Studies by Walsh and others, Wada, and Lee and coworkers contained findings supportive of EEG video monitoring.(29,30,31) In these studies, the authors found that prolonged EEG monitoring with simultaneous video taping provided information that helped identify nonepileptic attacks and differentiate them from epileptic seizures. (The studies of Walsh et al and Wada are summarized in the tables.)

Lee et al presented data from 644 patients who underwent EEG video monitoring at three different epilepsy centers during a 15-year time span for periods of 3 hours to 35 days.(31) According to Lee et al, video recordings of behavioral changes without associated EEG correlates helped differentiate nonepileptic seizures from primary generalized and secondary generalized epileptic seizures. Many of the characteristics of secondary generalized epilepsies were similar to those of primary generalized seizures, and EEG video monitoring distinguished behavioral and electrographic patterns. The investigators used the technique to assess drug or surgical treatments and to detect early recurrence of seizures after antiepileptic drug withdrawal. Examples of difficult diagnostic problems were presented wherein EEG video monitoring provided useful information that altered patient management and improved patient outcome.

In a 1986 report, Williamson described the role of EEG video monitoring in classifying and diagnosing complex partial seizures.(32) Complex partial seizures can present with a variety of clinical manifestations, and these seizures may resemble psychogenic attacks which can lead to diagnostic error. Williamson concluded that video recordings of patients' behavioral changes, with and without associated EEG correlates, helped differentiate complex partial seizures from psychogenic attacks. In addition to the EEG video monitoring. Williamson recommended that a trained observer examine the patients during seizures because important autonomic functions are usually not apparent in the video recordings. The observer can also obtain information on responsiveness, language function, ability to read, ability to recognize simple objects, motor function, visual fields, and orientation of patients.

A number of reports have examined indications for surgical treatment of epilepsy.(33,34,35,36). The National Institutes of Health Consensus Development Conference draft statement on surgery for epilepsy recommends that before a patient is considered for surgery, an evaluation should determine that cardiogenic syncope, pseudoseizures, and other nonseizure states that can closely mimic epileptic attacks have been excluded and epilepsy is present.(33) The evaluation should also determine the epileptic seizure type and According to the report, EEG video monitoring is widely used in the evaluation of complex seizures to exclude pseudoseizures and to establish clinical focality.

In a 1987 report concerning indications for surgical treatment of epilepsy, Andermann identified operative criteria as the failure of a thorough trial of antiepileptic medication and demonstration of focal origin of the attacks in an area of the brain that can be removed without producing significant neurological deficit.(34) According to Andermann, an accurate electroclinical assessment provided by video monitoring synchronized with EEG recordings is an essential prerequisite for considering a surgical approach to the seizure problem. A successful outcome of surgical treatment depends upon accurate localization of the brain tissue responsible for generating the seizures. EEG video monitoring is used to localize interictal as well as ictal epileptiform abnormalities. This approach is often facilitated by the withdrawal of the patient's antiepileptic medication.

According to Engel, EEG video monitoring provides the most accurate approach to analyze ictal behavior.(35) With video monitoring, seizures can be shown to relatives and friends for positive identification of the habitual episodes in question. Engel found that EEG video monitoring was used for some or all patients considered for surgery by almost 90 percent of the 44 epilepsy centers surveyed.

In the evaluation of patients for epilepsy surgery, intensive EEG video monitoring helps rule out those patients for whom nonepileptic (i.e., physiologic or psychogenic) seizures account for their intractable paroxysmal episodes. In other patients, use of intensive monitoring increases the probability of recording the characteristic ictal event that accounts for their medical intractability. According to Gates, 85 percent of patients with intractable seizures had episodes recorded by EEG within 2 weeks and 100 percent of presurgical candidates within 1 month following judicious reduction of their medication.(36)

To determine if the epileptic seizure of concern is of unifocal origin, Gates recommended synchronized EEG video monitoring of 8-10 characteristic seizures. For a definitive surgical localization, he suggested at least 12 and preferably 16 or more channels of surface/sphenoidal recordings to decrease the likelihood of missing possible multiple epileptogenic foci. Acceptance of surface/sphenoidal EEG ictal localization as adequate for the complete presurgical evaluation is controversial. In some cases, intracranial recording by means of epidural, subdural, or intracerebral electrodes may be appropriate for a more precise epileptogenic localization.(33,35)

Discussion

Epilepsy is a clinical diagnosis which in most cases can be characterized with a standard clinical history, examination and EEG. Standard EEG has the advantages of expanded spatial sampling, often utilizing 16 or more recording channels and a variety of montages that characterize abnormalities. In some cases it has not been possible to confirm or support a diagnosis of epilepsy or confirm or support the differential diagnosis of cardiac or psychogenic episodes from epilepsy using the standard EEG. While the importance of the standard EEG cannot be overemphasized, the limitation of its usefulness has always been its short monitoring time due to patient restrictions. The value of standard recordings is limited in certain situations both by their brevity and by the artificial nature of the EEG laboratory, which in some patients may inhibit seizure activity.(38,39). Perry et al obtained a 7-percent ictal yield using routine EEG examinations and a 70-percent ictal yield in the same patients using EEG video monitoring and withdrawal of anticonvulsant drugs.(3) The 70-percent ictal yield includes multiple repetitions of initially negative results.(23)

The review of published literature indicates that the clinical effectiveness and utility of EEG video monitoring has not been tested in randomized controlled trials or multicenter clinical trials. Most studies that have assessed the outcome of EEG video monitoring have been case series that have emphasized seizure elimination or reduction in frequency, but failed to provide information regarding the technique's diagnostic accuracy. While the reduction in seizure frequency is an important outcome measurement, reports of changes in seizure frequency need to be quantitated and related to clinical improvement. In addition, patients should be followed long enough to accurately determine the effectiveness of the technique. Only a few of the studies provided followup information. Nevertheless, published data derived primarily from single-institution studies suggest that the judicious use of EEG video monitoring has proven to be valuable in the diagnosis and management of selected patients with difficult cases of known or suspected epilepsy (intractable seizures) and related disorders. Because astute clinical judgment is often sufficient in the classification and differential diagnosis of epilepsy, Penry and Porter estimate that only 5 percent of patients with epilepsy considered to have intractable seizures will require EEG video monitoring.(17) Other estimates range from 20 to 30 percent.(33)

The differential diagnosis of epileptic from nonepileptic attacks is a particularly difficult problem in a significant minority of seizure patients. These patients are poorly controlled with drugs and have often been treated with increasing dosages and types of antiepileptic medications; differentiating true epileptic from nonepileptic attacks may be particularly difficult in this group. It has been reported that from 42 to 90 percent of psychogenic seizure patients have no evidence of coexisting epilepsy, and a reasonable diagnosis of psychogenic seizure can be made on the basis of history and routine EEG recordings in about 50 percent of cases. However, in the remaining patients diagnosis may be quite difficult.(28) Such cases of intractable seizure disorders have been studied in detail by intensive video monitoring techniques and many investigators have reported EEG video monitoring especially useful for differentiating nonepileptic seizures from true epileptic attacks; the diagnosis of nonepileptic seizure is improved by the observation of seizure-like spells in the absence of simultaneous ictal EEG abnormalities and may help to establish the presence of other events such as syncope, cardiac arrhythmias, transient ischemic attacks, narcolepsy or sleep disorders, and other behavioral disorders that may be confused with epilepsy (21,31,32). (see also Table 1). The benefits ascribed to the use of EEG video recording in these circumstances include reliable and accurate differentiation of true seizures from psychogenic seizures and the concomitant discontinuation of unnecessary and potentially toxic drugs. Investigators reported that the use of this technology has been instrumental in decreasing medication dosage, the number of anticonvulsant drugs prescribed, and, with counseling, decreased frequency of nonepileptic seizure episodes.(14,16,19,24)

The use of this technology in epileptic patients with intractable seizures may also allow the physician to more accurately classify the type(s) of epileptic seizure, which is essential for prescribing appropriate therapy.(40) Using this technique, Sutula et al were able to revise the seizure classification in almost half of the 40 patients studied, and Egli et al to determine the correct seizure classification in 180 of 215 epileptic patients.(19,24) There have been diagnostic classification changes, unrecognized seizure types identified, and improved seizure control resulting from video EEG monitoring studies of patients with refractory epileptic seizures (Table 2). According to Gloor, the proper diagnostic assessment of epileptic seizures requires knowledge of the relationship between the clinical manifestations of a seizure and the concomitant electrophysiological signals recorded in the EEG.(41)

Several investigators emphasized the use of EEG video monitoring to assess drug and surgical treatment results(31,40). ; they have successfully used the technique to quantitate epileptic attacks before, during, and after trials with anticonvulsant drugs and after surgical intervention to determine whether the intractable seizure persists or disappears following treatment. The technique was found to have been particularly helpful in deciding if a given drug combination is effective.

The role of EEG video monitoring in the surgical evaluation of seizure patients has been addressed by many investigators and most recently in a draft statement by the National Institutes of Health Consensus Development Conference on Surgery for Epilepsy.(33,34,35,36,42) Due to a decrease in surgical morbidity and mortality as well as improved techniques, the number of seizure patients considered suitable candidates for surgery has recently increased.(33) Investigators considering surgical treatment require anatomic localization of the epileptogenic foci most responsible for the patient's ictal symptoms. Video monitoring synchronized to the EEG recording is performed in order to provide an accurate analysis of the ictal behavior, as well as an accurate localization of the seizure's origin.(35) Information provided by the monitoring helps determine the suitability of surgical treatment and perhaps the specific surgical procedure for which the patient might be a candidate. EEG video monitoring is only part of the recommended presurgical assessment.

As indicated by certain studies cited in this report, EEG video monitoring has been successfully performed on an inpatient basis(3,16,19). or as an outpatient procedure (Table 3). Evaluations that can be completed in 3 to 12 hours of monitoring (diurnal) in an EEG laboratory suite are considered outpatient. Patients with frequent (at least three per week) intractable minor seizures and patients who are being evaluated for the efficacy of anticonvulsant drugs may be tested on an outpatient basis.(43,44) The use of activation techniques to increase the frequency of attacks has contributed to the reported success of EEG video monitoring in the outpatient setting.

According to some investigators, certain clinical problems may require prolonged EEG video monitoring in an inpatient setting. Inpatient EEG video monitoring has been recommended for patients with seizures, such as partial seizures, which are not frequent enough to be recorded in a 3-to 12-hour session; for seizures occurring only at night during induction of sleep; for seizures provoked on awakening; for seizures, such as frequent generalized motor or prolonged complex partial seizures associated with medical risk; for seizures provoked by drug withdrawal; and for seizures where intracranial electrode implantation is indicated.(42,43,44) Using activation procedures in conjunction with inpatient monitoring, Willmer and Brunet were able to evaluate most cases of complicated seizure disorders in less than 1 week.(27) Inpatient monitoring allowed the controlled tapering of anticonvulsant medications and supervision of sleep deprivation.

Some published reports have maintained that EEG video monitoring is costly.(14,19,45) It entails the simultaneous and prolonged recording of clinical seizures with video tape and EEG recordings and requires physician supervision, as well as the presence of other trained personnel to deal with the medical problems that might accompany a seizure. This is particularly true if depth electrodes have been implanted or if anticonvulsant drugs have been withdrawn in order to increase the yield of ictal events.(43) Costs for the technique will depend on the length and number of hospital stays required to record a sufficient number of seizures. Monitoring requires that multiple events be recorded until the clinician is satisfied that the diagnosis has been made. Although it may take a few hours to a month of monitoring to capture sufficient ictal events permitting an accurate diagnosis, the average length of stay is approximately 10-16 days. Rowan recommends daytime, 6-to 8-hour monitoring in conjunction with activation procedures. With no hospital admissions and less recording time, yet providing a similar yield to inpatient monitoring in capturing events, outpatient monitoring, according to Rowan, is cost effective in diagnosing difficult cases of epilepsy.(23) In contrast, inpatient stays for patients without reliable diagnoses who require repeated and lengthy hospitalizations due to inadequate control of their seizures are quite costly.

The American Electroencephalographic Society has published guidelines for long-term neurodiagnostic monitoring in epilepsy.(46) For the purpose of the guidelines, long-term monitoring is defined as the simultaneous recording of EEG (cable, telemetry, cassette) and behavior (self, observer, video) over extended periods of time. According to the guidelines, long-term monitoring permits an accurate diagnosis, classification, and characterization of seizures, as well as quantification of seizure activity and localization of epileptogenic lesions. The guidelines provide an outline of recommended clinical indications for long-term monitoring as well as qualifications for personnel, equipment, and procedures.

Summary

EEG video monitoring is the simultaneous documentation of the clinical and electroencephalographic manifestations of seizures. Using this type of monitoring, the EEG can be displayed on a video screen with face and whole-body views of the patient. This technique allows correlation of the electrical and clinical manifestations of a seizure. Frequent monitoring of plasma antiepileptic drug levels often accompany the recordings. Patient monitoring areas are usually dedicated recording suites in hospitals or specialized epilepsy centers; such suites simulate a home environment and have trained personnel in attendance.

Epilepsy is a clinical diagnosis in which most cases can be characterized with the standard EEG. There are some cases in which it has not been possible to confirm or support a diagnosis of epilepsy or the differential diagnosis of cardiac or psychogenic episodes from epilepsy with the standard EEG. An ambulatory cassette monitor increases the likelihood of recording an EEG during a seizure and provides reliable data in the evaluation of some patients with suspected syncope, transient ischemic attack, psychogenic seizures, and poorly defined seizure disorders. Studies of the treatment of intractable seizures (unsatisfactory control of seizure) indicate that more complex cases may require the data obtained with the intensive neurodiagnostic technique of EEG video monitoring. Estimates of the percentage of epileptic patients (with intractable seizures) requiring EEG video monitoring range from 5 to 30 percent. Before seizures are deemed intractable it is necessary to be certain that the correct drugs have been used in the correct amounts; therefore prior to considering a patient a candidate for EEG video monitoring it should be determined that there has been a reasonable trial of the appropriate antiepileptic drug with adequate monitoring of compliance and the effects of treatment.

The review of the published literature suggests that EEG video monitoring may have clinical utility in the evaluation and treatment of patients with intractable seizures or seizure-like conditions when conventional diagnostic methods have proven inadequate. This includes differentiating nonepileptic seizures from true epileptic seizures. The video recording permits repeated viewing of the clinical sequence and comparisons made to recordings of known seizures.

For epileptic patients with intractable seizures, the use of EEG video monitoring has been emphasized to provide correct seizure classification and improved patient management in addition to its role in the surgical evaluation of seizure patients. Bizarre behaviors that are not easily recognized as seizures can be appropriately identified by a simultaneous epileptiform discharge of EEG. Conversely, video tape evidence of classic behavioral manifestations of a seizure may be sufficient to diagnose epilepsy even in the absence of a clearly defined epileptiform EEG abnormality.

Based on the studies cited in this report, EEG video monitoring may be viewed as both an outpatient and inpatient procedure. Information provided by this technique has improved patient outcome by permitting accurate diagnoses and modified therapy. The technique has expanded the number of patients viewed as candidates for surgery.