Psychogenic Seizures

"You'd better ask the doctors here about my illness, sir. Ask them whether my fit was real or not.''--Smerdyakov to Ivan in The Brothers Karamazov ^[1]

A subgroup of the patients seen for the treatment of seizures are patients who do not have epilepsy. These may constitute some 20% of patients at epilepsy referral centers. ^[2] Assuming ^[3,4] that epilepsy centers primarily serve patients with intractable seizures (but not only or all such patients), that patients with intractable seizures are about one-third of all patients with epilepsy, and that the point prevalence of epilepsy in the United States is 0.6%, there may be several hundred thousand patients with pseudoseizures in the United States, and analogous numbers of patients elsewhere. Pseudoseizures are frequently misdiagnosed, perhaps because the manifestations of the disorder are variable. Patients can continue having episodes for years. Some suggest that the longer the episodes continue to occur, the less likely they are to come under control. ^[5] For this reason it is important to know the characteristics of this disorder, so that appropriate and early treatment can occur. The literature discusses the relative advantages or disadvantages of various terms, ^[2,6,7] but no one term is entirely satisfactory. "Psychogenic seizure'' and "pseudoseizure'' are the most commonly used and they will be used somewhat interchangeably in this review.

Psychogenic seizures are episodes of altered movement, emotion, sensation, or experience, similar to those due to epilepsy, but which have purely emotional causes. King et al. ^[8] assessed the accuracy in differentiating episodes of epilepsy from pseudoseizures by three groups. They found that admitting physicians correctly determined episodes as due to epilepsy in 14 of 17 cases. Hospital monitoring personnel correctly diagnosed these in 8 of 10 cases, and a neurologist viewing the videotape without knowledge of the EEG findings did so in 37 of 52 cases. Psychogenic seizures were correctly diagnosed in 8 of 16, 13 of 16, and 63 of 86 cases. These findings underscore the difficulties inherent in conclusions about either pseudoseizures or the epilepsies when these conclusions are based upon clinical observation alone, difficulties inherent in the writings of decades past. Because of these limitations, recent studies have almost always considered both behavioral observations and EEG. Nonetheless, this review includes the ideas of clinicians such as Charcot and Gowers that have helped shape our understanding of psychogenic seizures and are part of the inheritance of all neurologists.

Clinical characteristics of pseudoseizures.

The onset of individual episodes is often gradual, at times beginning with stress ^[9-12] or with auditory ^[10] or visual ^[13] stimuli. One might expect that pseudoseizures only would occur when the patient is awake, so an episode during sleep could not be a psychogenic seizure. However, EEGs during "sleep onset'' pseudoseizures ^[14,15] show that the patient actually is awake before the episode begins. A variety of symptoms at onset can at times be confused with an aura due to epilepsy. Symptoms reported include palpitations, malaise, choking, numbness, peripheral sensory disturbances, pain, odors or tastes, and visual hallucinations or distortions Table 1. ^[6]

Motor phenomena include tonic movements and tonic posturing, repetitive motor movements, and absence of movement, including apparent loss of tone. Increased muscle tone can include opisthotonos, ^[13,16,17] including the classic "are en circle'' described by Charcot. ^[18,19] One study described tonic or opisthotonic postures in 13 of 27 patients, ^[20] another, 11 of 25 patients with forward pelvic thrusting and 11 of 25 with whole-body rigidity. ^[21] Repetitive motor movements can consist of jerking of either the arms, or legs, or both, and can be regular or irregular and coordinated or not. ^[6] One of the just-mentioned studies ^[20] described jerking in 10 of 27, flailing or thrashing in 5 of 27, and "rhythmical `coordinated''' movements in 6 of 27 patients. The other ^[21] described out-of-phase clonic activity in 14 of 25 patients in the arm and in 14 of 25 in the legs. This study reported that in-phase clonic activity occurred in 5 of 25 patients in the arms and 4 of 25 in the legs. ^[21] Other motor phenomena include side-to-side head movements in 9 of 25 ^[21] and in 7 of 47 ^[22] patients, and trembling in 14 of 27 ^[20] and 11 of 43 patients. ^[12] Meierkord et al. ^[23] reported prominent motor features in 73 of 110 patients and emphasized that the sequence and rhythmicity of motor activities during attacks differed from those observed during generalized tonic-clonic seizures due to epilepsy. In particular "the evolution of the clonic jerks from fast small-amplitude to slower large-amplitude movements ... and the rapid contraction and slow relaxation of the true epileptic clonic jerks are not usually seen in pseudoseizures.'' Leis et al. ^[22] said a continuum of movements occurred, so that absolute categorizations were impossible, and emphasized that their patients most commonly were unresponsive and moved little or not at all. They could not evaluate every sign in every patient, but reported that little or no movement occurred in the upper extremities in 22 of 47 and in the lower extremities in 31 of 40. There were no pelvic movements in 43 of 47, no general body rigidity in 44 of 47, no head movements in 30 of 47, and no eye movements in 37 of 45 patients. Perhaps in keeping with these findings, atonia was observed in 6 of 22 seizures and 2 of 22 patients in two other studies. ^[10,20]

The eyes may deviate upward during a psychogenic seizure, and there may be rapid tremorous movements of the lids. The patient may resist eye-opening. ^[10,20-22]

Upgoing toes can be found on plantar stimulation of a patient experiencing a psychogenic seizure ^[3,24]; healthy adults can learn ^[25] to dorsiflex their toes voluntarily in response to this maneuver (personal communication).

Decreased responsiveness is common in psychogenic seizures (20 of 27, ^[20] 82 of 110, ^[23] and 23 of 25 ^[21] patients), but neither this nor the patient subsequently denying recollection of the episode indicates that the patient was in fact unconscious, if the baseline awake EEG does not change during the period of unresponsiveness.

Apparently semipurposeful or purposeful behaviors occurred in 14 of 27 patients in one study ^[20] and in 48 of 110 in another. ^[23] Patients can demonstrate symptoms reminiscent of partial seizures, such as swallowing, mouthing, chewing, licking, or smacking of the lips (5 of 18 ^[26] and 19 of 27 ^[20] patients). Metrick et al. ^[27] reporting on children, commented that psychogenic seizures, but not those due to epilepsy, "almost always could be interrupted by either verbal and/or physical intervention.'' There can be apparently confused behaviors such as looking about, picking up or moving objects, and undressing, which can appear either directed or nondirected. ^[6] Violence can occur, including biting, striking, kicking, slapping, pushing, obscene gesturing, clawing at the face or clothes, and directed rage. ^[6] In one study injury occurred in 19 of 110 patients. ^[23] Another reported that seizure movements were sometimes so forceful that patients vaulted the bed rails. ^[10] Peguero et al. ^[28] interviewed 73 patients with psychogenic seizures by telephone. Twenty-nine reported injuries, including lacerations, fractures, and bruises. Thirty-two reported biting their tongues. Injury during seizures was more common among pseudoseizure patients who had attempted suicide than among those who had not (14 of 23 versus 15 of 50 patients).

The combination of widespread motor movements with subsequent ability to recall the details of the episode might be thought sufficient to diagnose a pseudoseizure. However, both focal and widespread unilateral motor seizures due to epilepsy can occur without loss of consciousness. Also, supplementary motor seizures characteristically involve both sides of the body, but patients often maintain awareness during the episode. ^[29]

A variety of vocalizations can occur, including crying, yelling, screaming, sobbing, or dramatic, tragic, obscene, or mystical vocalizations. ^[6] One study ^[30] described weeping during or immediately after a psychogenic seizure episode in 10 patients, found this in only 1 of 200 patients with epilepsy, and noted that Penfield and Jasper ^[31] described this in only 1 of 220 patients undergoing cortical stimulation.

A variety of autonomic changes can accompany psychogenic seizures. Respiratory changes include choking or coughing (with the skin being cyanotic, pallid, dusky, ^[6] or flushed) and signs or symptoms of hyperventilation syndrome, including lightheadedness; acral paresthesias; palpitations, tachycardia, dyspnea, or chest pain; gastrointestinal distress; headache; dysphoria or panic; fatigue or weakness; and muscle cramps, spasms, or tetany. ^[19,32] Gowers ^[19] described laryngospasm, large pupils, and apparently decreased sensitivity of the conjunctivae during psychogenic seizures. Large pupils may occur due to increased, stress-related, sympathetic discharge but the light responses should be intact, if a sufficiently bright light is utilized. Testing of the cornea with a cotton swab could help differentiate psychogenic seizures from epilepsy, ^[22] but if a patient is willing to sustain self-injury, a patient also might not react to irritation of the conjunctivae.

Either urinary or fecal incontinence, or both, occurred in 84 of 412 patients. ^[6,23,28] One study ^[28] reported fecal incontinence in 5 of 32 patients with urinary incontinence, but in 0 of 41 patients without this. Both Charcot and Gowers felt that these would not occur during psychogenic seizures, whereas Freud argued that they could, if consistent with the patient's underlying psychopathology. ^[6] In this sense, incontinence is one more symptom of self-destructive behavior.

One study found the mean duration of psychogenic seizures to be 108 seconds (range 3 seconds to 13 minutes). ^[33] A second reported 1 of 47 patients had a seizure of less than 10 seconds, 6 of 47 10 seconds to 1 minute, 5 of 47 1 to 2 minutes, and 35 of 47 patients a seizure of greater than two minutes duration. ^[22] In a third, 5 of 110 patients had episodes lasting less than 1 minute, and 8 of 110 patients had episodes lasting more than 30 minutes. ^[23] A fourth found that generalized tonic-clonic episodes due to epilepsy had mean duration of 70 seconds and median duration of 69 seconds (range 50 to 92 seconds); psychogenic episodes resembling these had mean duration of 134 seconds and median duration of 75 seconds (range 20 to 805 seconds). ^[21] Therefore, psychogenic seizures tend, but do not have, to be longer than those due to epilepsy. Episodes can terminate as abruptly as in epilepsy, but may also terminate relatively gradually. ^[10,20] In one series, 11 of 30 patients reported postictal confusion or somnolence, but behaviors suggesting these were observed in only three during video EEG monitoring. ^[10]

An important, and in at least some practices surprisingly common, ^[3,7] subgroup of patients comprises those with psychogenic status or pseudostatus. Patients can present with a variety of symptoms, including all of those listed above. Specific patients may exhibit limb movements ranging from those reminiscent of generalized tonic-clonic seizures, ^[3] to "slow, subtle writhing or in-phase limb movements,'' ^[22] to apparently semipurposeful or automatic movements reminiscent of the automatisms of complex partial seizures, to unresponsiveness with lack of movement. Patients may allow themselves to be intubated. ^[3,34,35] Complications due to emergency interventions for pseudostatus have included respiratory arrest, septicemia, pneumonia, urinary tract infection, cellulitis, and foot drop, these complications ensuing following sedative medication, venous access, and parenteral injections. ^[3]

Patients with psychogenic seizures have been from 4 to 77 years of age. ^[6,15] One study ^[12] reported 43 children with psychogenic seizures beginning at age 5 to 16 years, another ^[23] 9 patients who had onsets of their episodes after age 50. About three-fourths of reported patients are women (734 women and 250 men in 21 articles ^{[3,6,14,15,22,23,27,28,30,34-45]}).

Diagnosis.

Certain features are helpful in diagnosing psychogenic seizures, but none are absolute. For example, gradual seizure onset is common during psychogenic seizures and unusual in patients with epilepsy. However, some patients with epilepsy report a sense of dysphoria or altered mood for a considerable time, even many hours, prior to the actual seizure. Movements variously described as quivering, flailing, or thrashing are common during psychogenic seizures, but not during episodes due to epilepsy. Nonsynchronous movements are more common during pseudoseizures than during episodes of epilepsy, as are side-to-side movements of the head and directed reaching. However, frontal lobe seizures can manifest a variety of unusual behaviors, including nonsynchronous movements, pedaling, kicking or thrashing movements, vocalizations, and sexual automatisms ^[23,34,46]; moreover these patients may develop complex partial status. ^[46] Lancman et al. ^[12] reported that seven children had video EEG repeated several years later. The episodes recorded years later showed a greater resemblance to epilepsy in 5 of 7, suggesting that a learning or practice effect may be present, at least in some patients.

Some authors have classified psychogenic seizures by their resemblances to various types of epilepsy, for example, those like complex partial seizures versus those like generalized tonic-clonic seizures. However, Leis et al. ^[22] found that pseudoseizures were distinctly different in their categories from episodes of epilepsy. Their patients were most commonly unresponsive with little or no movement. Meierkord et al. ^[23] subdivided their patients into those with collapse or limpness (37 of 110) and those with prominent positive motor features (73 of 110). During episodes of collapse, the patient might manifest a "dramatic `swoon''' and lie motionless. During episodes with prominent motor features, the progression, pattern, and rhythmicity differed from what occurs in epilepsy. They report that 82 of 110 showed decreased responsiveness and 48 of 110 showed apparently purposeful behavior. Therefore, some patients manifested both kinds of behaviors and, in summary, there are at least two constellations of signs. One pertains to responsiveness and reactivity. The other pertains to movement per se.

Patients with pseudoseizures can have organic cerebral pathology. ^[27] Also, events may not be due to epilepsy but not be due to psychogenic seizures either. A variety of physiologic events, including movement disorders, migraine, gastroesophageal reflux, parasomnias, ^[7,27] and syncopal disorders may be misinterpreted as being due to epilepsy. ^[6,47,48]

Laboratory testing.

The combination of a normal EEG with an episode manifested by loss of responsiveness is virtually diagnostic of a psychogenic seizure, especially if the patient subsequently indicates that occurrences during the episode cannot be recalled. Nonetheless, there are some important caveats. First, an EEG can be normal during simple partial seizures (i.e., seizures not manifested by loss of consciousness), for example, from the frontal lobe. ^[23,46,49] The use of additional scalp electrodes, especially in combination with anticonvulsant withdrawal, ^[27] can reduce the number of false negatives during ictal recordings. ^[50] Second, it is the ictal, not the postictal, EEG pattern that is the most important. Third, a patient could have either focal or generalized slowing due to an independent cause, such as a diffuse encephalopathy or a cortical lesion, without this finding necessarily being indicative of the presence of epilepsy. ^[6] Fourth, generalized epileptiform bursts can occur during drug, especially barbiturate, withdrawal regardless of whether the patient has epilepsy; only definite epileptiform patterns can be taken as evidence consistent with epilepsy; artifacts or normal-variant patterns can be confused with epileptiform discharges; and definite epileptiform discharges can on rare occasions occur in patients without epilepsy, or in asymptomatic relatives of patients with epilepsy. ^[6,51] Interictal epileptiform discharges were reported in 2 of 13 ^[3] and 3 of 5 patients ^[27] with psychogenic seizures but no history of epilepsy and in 6 of 30 patients with both. ^[10] Therefore, interictal epileptiform discharges may be incidental, or may help support a diagnosis of epilepsy, but they don't exclude a diagnosis of pseudoseizures. Fifth, syncope might cause loss of consciousness, falling, and (at least brief) shaking of the limbs. The EEG might show rhythmic slow-wave discharges or EEG "flattening'' during the episode. This is an EEG abnormality, but a nonepileptiform abnormality. Sixth, a prolactin level obtained within 15 to 30 minutes after the episode can be helpful. A several-fold increase relative to baseline is evidence consistent with a diagnosis of epilepsy. Both false-positive and false-negative results can occur, however, and elevation is most common after generalized tonic-clonic or temporal lobe seizures, but not after frontal lobe or other simple partial seizures. ^[6,52-57] Finally, it is important to insure that the episode recorded in the laboratory is representative of those that occur spontaneously at home.

Seizure induction protocols.

Because episodes do not necessarily occur spontaneously in the laboratory, there has been wide interest in induction procedures. Methods utilized have included compression of body parts, photic stimulation, verbal suggestion, placement of a tuning fork or moistened patches on the skin, and intravenous administration of saline or other placebos. ^[6,42,45,58] When performing an induction, one should be sure to do this with sensitivity to the emotional needs, including the need for a sense of dignity, of the patient. Induction procedures can allow patients the opportunity of having an episode without being coercive, dismissive, or misleading. Inductions are similar to maneuvers performed during the standard neurologic examination, utilized to help determine whether a particular symptom or sign is, or is not, of neurologic origin. However, not every patient with pseudoseizures will have an episode in response to suggestion and an episode of epilepsy might occur in the course of an induction procedure. ^[45,59]

About 70% of patients in three recent series had seizures that were inducible using either saline infusion ^[45,58] or an alcohol-soaked patch. ^[42] Walczak et al., ^[45] however, emphasized that two patients with epilepsy had their typical episodes induced by this procedure, and three patients had atypical episodes. Slater et al. ^[58] described six patients who had episodes that they judged due neither to pseudoseizures or epilepsy (motor tic, syncope due to mitral valve prolapse, neurocardiogenic syncope, narcolepsy, panic attacks, and episodic dyscontrol syndrome causing rage attacks). Bazil et al. ^[60] evaluated 52 consecutive inpatients or outpatients followed as part of a more general epilepsy practice (although one at an epilepsy center). Nineteen had clinically typical episodes during induction, 12 atypical episodes, and 21 no episodes. The authors commented that pseudoseizures therefore may be common in a general seizure population, but some patients might have mimicked their usual episodes purely because of the circumstances of the testing. Therefore one should pay careful attention to the clinical circumstances of the test and to the clinical details of the induced episodes. Additional evaluations are important in many cases.

Several studies have found that only a subgroup (perhaps 10 to 30%) of patients with psychogenic seizures also have epilepsy. ^[6,23,59,61] On the other hand, another ^[34] found the coexistence to be more common, occurring in about 60% of adults and 20% of children with pseudoseizures. The exact percentages might well vary with specific clinical populations and, perhaps, with diagnostic criteria. Whatever the exact numbers, the clinical implications remain the same: any given individual patient might have two diagnoses rather than one, but when one makes a diagnosis, for example, the diagnosis of pseudoseizures, one should not automatically infer the presence of a second diagnosis, for example, that of epilepsy. Rather, one should consider each diagnosis specifically and independently within the context of the specific clinical situation.

Walczak et al. ^[45] discussed whether an induction procedure might violate patient-physician trust and diminish a patient's willingness to receive psychiatric intervention. Following induction, they told all patients definitively that they had used saline, in a supportive and nonaccusatory manner. All patients agreed to psychotherapeutic intervention following explanation of their diagnoses. Lancman et al. ^[42] similarly commented that an induction did not appear to result in subsequent difficulties with the doctor-patient relationship. The concept of beneficence as applied to medical ethics ^[62] dictates that the physician should not only respect and avoid harming a patient, but also should make positive contributions toward that patient's well-being and toward the removal of possible harmful conditions, maximizing the benefits and minimizing the risks of a particular treatment. An induction procedure can make a contribution to this by avoiding the potential hazards of an inappropriate diagnosis with the attendant prescription of inappropriate and possibly hazardous treatments. It also can increase the likelihood of appropriate treatment for the condition that the patient actually has. ^[6]

Etiology.

A psychogenic seizure is, almost by definition, a conversion symptom, and can frequently co-occur along with other symptoms. ^[63,64] However, this could occur for a large number of reasons, including reactions to a specific event, expressions of dependency or attention-seeking needs, or manifestations of a wish to exert greater control. ^[6] One should distinguish between the use of the term "conversion'' in the sense of a symptom occurring in response to an emotion and in the more specific sense used in psychiatric diagnosis. Just as the manifestations of the psychogenic seizure episodes differ from one patient to the next, so the underlying personality profiles differ. Also, since some patients with epilepsy report their seizures can be precipitated by emotional changes, the presence of a psychiatric diagnosis is not sufficient to infer the presence of psychogenic seizures.

Patients with pseudoseizures frequently report coming from chaotic families. ^[27,43] In addition, a number of papers indicate that psychogenic seizures may occur in response to sexual or physical abuse. ^[6,41] Because physical and sexual abuse may be more common in the general population than previously thought, one must be cautious about inferring that a specific association is present unless it actually is demonstrated. ^[65] However, one study ^[41] compared a group of patients with pseudoseizures (71 with only pseudoseizures, 14 with pseudoseizures plus complex partial seizures) with a group of 40 patients with complex partial seizures alone. This study found that, among 85 patients with pseudoseizures, there was a 24% incidence of sexual and 15.5% incidence of physical abuse histories versus 7.1% and 2.5% in the complex partial seizures group. The increased frequency continued to be present when correcting for the composition of the two groups (about threequarters of the pseudoseizure patients were women, versus about half of the patients with complex partial seizures alone, and women are more likely to report sexual abuse).

Depression is another common finding among patients with psychogenic seizures. ^[43] Suicide attempts have been reported ^[3,6]; one recent study ^[28] concluded that tongue biting and other self-injury, as well as incontinence, during pseudoseizure episodes were additive in predicting a higher likelihood of suicide attempts. The occurrence of suicide attempts in this population also raises the possibility of a disorder of impulse control, as may occur in patients with borderline personality disorder. Patients with borderline personality disorder may be depressed and also may have disturbances in body image such as might be manifested by conversion symptoms. ^[6]

Other diagnoses in patients with pseudoseizures include obsessive-compulsive personality, inadequate personality, schizophrenia, and Munchausen syndrome. ^[2,6,34,66] Alper et al. ^[67] reported 92 consecutive patients with nonepileptic seizures, evaluated over a 2-year period by a single psychiatrist using a standard ^[68] interview protocol. Sixty-six were diagnosed to have conversion disorder, five somatoform disorder. The remainder had: anxiety disorders (six panic disorder, two posttraumatic stress disorder, one anxiety disorder); psychotic disorders (three psychosis, two schizophrenia, one schizoaffective disorder, and one bipolar disorder); and other disorders (two attention deficit disorder, one tic disorder, one depersonalization disorder, one malingering). Are some patients malingering? The writer Axel Munthe reported that many of Charcot's patients at the Salpetriere were consciously imitating events they saw occurring among their fellow patients with epilepsy ^[69] and Gowers commented that consciously feigned episodes could occur among "professional beggars ... soldiers ... and criminals.'' ^[19] It can be impossible, however, to know with certainty whether malingering is occurring in any individual case unless, like Smerdyakov, ^[1] the patient tells you so.

Gates and Erdahl ^[70] developed a classification scheme based upon observations of behaviors in 34 patients, and delineation of five subgroups: an inadequate coping mechanism group, a misinterpretation group, a psychotic group, a highlighting group, and an emotional conflict group. The five patients who were not coping adequately were of borderline intellectual capacity and their pseudoseizures seemed an attempt to manage their environment. The misinter-pretation group comprised four patients who had misperceived normal body sensations, often then had been given anticonvulsants, and then had developed toxic drug effects that exacerbated the original experiences. Two psychotic patients had developed behaviors misjudged as epileptic seizures. Two were patients who had epilepsy but "tended to highlight or embellish'' the manifestations of the seizures. Twenty-one were patients whose episodes seemed examples of conversion reactions in the original sense, that is physical expressions of emotional conflicts. They explored how they could classify patients using DSM-III-R ^[68] methodology and suggested several groups. The first included somatization disorder, conversion disorder, hypochondriasis, and undifferentiated somatization disorder. The second included panic disorder, the third psychotic disorder, the fourth factitious disorder and malingering, and the fifth reinforced behavior pattern. The last group did not easily fit into DSM-III-R; many patients in this group had mild or greater mental retardation. Since different patients are likely to respond to different forms of intervention, classifications such as those just described are important. They should derive from structured interviews using standardized diagnostic criteria, can help in identifying patients who do not easily fit into standard categories, and should be the first step in devising appropriate and individualized treatment plans.

A number of groups have investigated whether objective testing instruments could be used to distinguish patients with pseudoseizures from those with epilepsy. Using the Minnesota Multiphasic Personality Inventory (MMPI), Wilkus et al. ^[71] suggested three separate rules for identifying patients with psychogenic seizures. These were: (1) a "conversion V'' pattern, with the Hy (hypochondriasis) and Hs (hysteria) scales elevated and higher than the D (depression) scale; (2) a higher Pd (psychopathic deviate) scale score; and (3) a higher Sc (schizophrenia) scale score. They were able to classify correctly 80 to 90% of patients using these rules. Henrichs et al. ^[72] were, but Vanderzant et al. ^[44] were not, able to confirm the suggested algorithm.

Wilkus and Dodrill ^[37] found that they previously ^[71] had studied two kinds of patients, those whose psychogenic seizures had either a prominent affectual component or minor motor activity and those whose episodes had a minor affectual component or prominent motor activity. They pointed out that Vanderzant et al. ^[44] studied only patients with motor episodes. Wilkus and Dodrill could not distinguish among patients with prominent motor features either, but they could distinguish among patients with prominent affectual components during their seizures. They therefore suggested all of the studies just described might be relatively consistent with one another and the MMPI criteria proposed might help in identifying specific patient subgroups. In keeping with this, one subsequent study ^[35] described an "almost universal conversion pattern,'' upon MMPI assessment of patients with psychogenic status epilepticus. Another ^[39] found the MMPI "conversion V'' profile occurred more often in patients with psychogenic seizures. The distinctions they found accurately categorized 80% of patients with epilepsy and 60% of patients with psychogenic seizures. These authors also found an elevation on the depression or schizophrenia scale in some patients. ^[39] A third study ^[73] found that stringent use of the "conversion V'' profile resulted in false-positive rates as low as 2% in patients with intractable epilepsy who were candidates for surgery.

Some authors have found that patients with psychogenic seizures have a relative increase in cognitive difficulties, ^[35,73,74] perhaps due to medication effects, ^[35] but this has not been found by others. ^{[6,27,37,58,71,75]} One group ^[39] commented that, although many patients with psychogenic seizures had scores well within normal limits, most had at least some evidence of impairment. From a treatment perspective, the critical initial step is determining whether a given patient has pseudoseizures. If the patient does, cognitive and emotional assessments can help in determining why they might be occurring in that patient, whether cognitive difficulties are present, and what treatment plan might be best.

A recent study ^[76] used the California Verbal Learning Test, a test of word memorization, recall, and recognition. Comparing 12 patients with left temporal lobe epilepsy, 11 with right temporal lobe epilepsy, and 18 with pseudoseizures, the study found a "negative response bias,'' a reduced likelihood of false-positive responses during the word recognition phase of the test, among patients with pseudoseizures. The authors of this study were able to obtain a specificity of 91% and sensitivity of 61% for the diagnosis of pseudoseizures using this method and speculated that the findings might reflect denial mechanisms present in this group of patients. They cautioned, however, that this measure was not an independent or unique diagnostic sign for this disorder.

Treatment.

However the diagnosis is made, that diagnosis should be linked with a multidisciplinary treatment plan. ^{[6,44,58,77-79]} The first step in treatment is a tactful presentation of the diagnosis to the patient, together with an initial exploration of possible reasons for the occurrence of the episodes and a recommendation for appropriate psychotherapeutic intervention. ^{[2,5-7,80,81]} In some patients, a single counseling session that includes appropriate presentation of the diagnosis is apparently sufficient, since a subgroup of patients who have had no other treatment are found symptom-free at follow-up. ^[5,23,59] More often, however, patients benefit from additional treatment. One study ^[59] reported that 6 of 10 who received subsequent additional psychotherapeutic intervention became seizure-free versus 2 of 9 who did not. In a second, larger study, ^[23] 25 of 43 who received additional psychotherapeutic intervention became free of their episodes, but only 3 of 27 did not. Patients who were women, led independent lives, and who did not also have epilepsy were more likely to be seizure-free at follow up. ^[23] In another study, of 50 patients, ^[82] recent onset of seizures and normal psychiatric evaluation were associated with better outcome; there were (nonsignificant) trends in this direction for patients who had nonconvulsive psychogenic seizures and who were employed. A study of somatization disorders in general ^[83] found that subsequent health care charges declined by 53% in a group of 38 patients who received psychotherapeutic treatment; there was no overall change, although there were wide fluctuations in degree of improvement, in an equal number of similar patients who did not receive such treatment. Perhaps because of a shorter duration of symptoms, ^[82] Wyllie et al. ^[5] found that 13 of 18 pediatric patients (8 to 18 years old) were episode-free at follow-up, compared with 5 of 20 adults (25 to 56 years old).

A variety of techniques have been advocated for treating patients with psychogenic seizures, including behavioral modification approaches, hypnosis, milieu-related techniques, individual counseling, hypnosis, pharmacotherapy, and family counseling. ^[77-79] Family counseling may be particularly important, at least in selected cases, given the data that the family milieu frequently is disrupted in patients with this disorder. ^[12,43,71] On the other hand, patients with Munchausen syndrome are less likely to have a response to treatment. ^[6,34] The neurologist should keep in mind that it is not always a simple matter for an appropriate strategy to be devised by the mental health team, particularly when patients do not fall easily into a specific diagnostic category. Moreover, there often are only rudimentary data allowing the treating team to link a specific presentation of psychogenic seizures, plus a specific underlying personality makeup, to a specific treatment and from there to a likely outcome. It therefore would be worthwhile for neurologists and mental health professionals to work together toward developing, and assessing the efficacy of, treatment strategies devised for specific groups of patients ^[70] who present with these behaviors, and to determine what information might help in deciding on a treatment plan. Because of the potential morbidity of the disorder, and its implications regarding the overall functioning of the patient, diagnosis and treatment should occur as early and rapidly as possible.

Acknowledgments

I thank Jason Brandt, PhD, Barry Gordon, MD, and Paul Mullin, MD for helpful discussions regarding this manuscript.