INTRODUCTION — Psychogenic nonepileptic seizures (PNES) are nonepileptic events resembling seizures or syncopal attacks. The etiology, epidemiology, clinical features, and diagnosis of PNES are discussed in this review. Management and prognosis are reviewed separately. (See "Psychogenic nonepileptic seizures: Management and prognosis".)
Other nonepileptic paroxysmal disorders are discussed elsewhere. (See "Nonepileptic paroxysmal disorders in adolescents and adults".)
DEFINITION AND TERMINOLOGY — Psychogenic nonepileptic seizures (PNES) can be regarded as involuntary experiential and behavioral responses to internal or external triggers [1]. As behaviors, they are not accompanied by pathological changes in brain physiology (such as epileptiform electroencephalography [EEG] activity or cerebral oligemia). They have been thought of as clinically resembling epileptic seizures, but some resemble syncope.
There is an unfortunate multiplicity of different names for this condition. PNES is probably the most widely used term, but alternatives such as psychogenic seizures, dissociative seizures, functional seizures, psychogenic nonepileptic attacks (PNEA), and nonepileptic seizures are also used.
ETIOLOGY — The etiology and pathogenesis of PNES as a disorder are unclear. Most evidence suggests a primarily psychological or behavioral mechanism; brain imaging evidence hints at biological factors that may at least predispose to PNES. However, most factors thought to be important in etiology (eg, sexual abuse and other forms of early trauma) are nonetheless only present in a proportion of patients [2-9], suggesting that a number of alternative pathways might lead to PNES.
One proposed model considers two pathways [10]. In the traumatic pathway, there is a proposed "cutoff" phenomenon in response to intrusion of intolerable memories [11]. In the developmental pathway, PNES are derived from maladaptive coping in the face of complex life tasks and milestones along the patient's continuum of psychosocial maturation. Dysfunctional family relationships are reported by many PNES patients and/or family members [5,12-14]. A repressive style of coping may develop in response, characterized by avoidance of emotional difficulties, focusing conscious attention on external physical symptoms [15]. Another model integrates the strikingly heterogeneous nature of the PNES population into a pathway that includes the potential for multiple social, psychological, and physical factors to play a role [16,17].
Functional imaging studies have documented a range of changes in brain activity or functional connectivity, alterations in emotion control, executive, attentional, sensorimotor, and default mode networks [18-20]. These studies have been small, and the effects have been found at group level (ie, not in all individuals), so it remains unclear how these findings relate to the pathogenesis of PNES themselves or to factors that might predispose to PNES or other population effects.
Psychosocial stressors that have been reported to precipitate the emergence of PNES as a disorder include bereavement; pregnancy; ongoing physical, verbal, or sexual abuse; lawsuits; job pressure; financial difficulties; impending divorce; socioeconomic deprivation; domestic conflicts; and assault [5,21-25]. In children, separation anxiety, school avoidance, and parental discord or divorce may also play a role [26,27]. Sometimes a recent event has seemed minor but may serve as the "last straw" in a series of events or may serve as a reminder of more significant remote events [28]. Cultural variation and seizure modeling (where the clinical semiology of seizures is influenced by exposure to seizures in others, directly, in video or written material) may also play a role [5,29,30].
EPIDEMIOLOGY AND POPULATION CHARACTERISTICS
●Incidence – From population-based studies, including only video-electroencephalography (EEG) confirmed cases, the estimated incidence of PNES is up to 4.90 per 100,000 persons per year [31-33] One of these studies included non-video-EEG confirmed cases, which increased the incidence to 6.17 per 100,000 persons per year [32].
●Prevalence – The prevalence of PNES is more difficult to measure. Indirect estimation from non-population data has suggested the broad range of 2 to 33 per 100,000 persons [34]. One population study from Norway found a prevalence of 10.6/100,000 [35]. It is known that some patients with PNES continue to have events after diagnosis but cease to present to medical services for them [36], and estimates or population data to quantify this are lacking. Among patients referred to outpatient epilepsy centers, 5 to 25 percent have PNES, while 25 to 40 percent of patients evaluated in epilepsy monitoring units for intractable seizures are diagnosed with PNES [33,37]. Two large studies have found that PNES are more prevalent in lower socioeconomic groups [38,39].
PNES appear to occur across a wide range of racial and cultural backgrounds, though comparative epidemiological studies are lacking.
●Age at onset – PNES most commonly present in the third decade of life [2,40-42]. However, all age groups can be affected, including children, teenagers, and older adults [43-47]. In one study, patients with learning disabilities had a relatively younger age of onset compared with those with a history of physical or psychosocial trauma [3]. (See "Nonepileptic paroxysmal disorders in children" and "Seizures and epilepsy in older adults: Treatment and prognosis".)
●Overlap with conversion disorders – The population of patients with PNES shares many characteristics with that of patients with conversion disorders in general, and the two coexist in many patients. PNES are grouped under Conversion Disorder in DSM-5 [48] (See "Conversion disorder in adults: Terminology, diagnosis, and differential diagnosis" and "Conversion disorder in adults: Clinical features, assessment, and comorbidity".).
Unlike conversion disorders, factitious disorder and malingering imply conscious production of signs and symptoms. They are rarely associated with PNES. (See "Factitious disorder imposed on self (Munchausen syndrome)".)
●Female predominance – PNES has a female predominance, with larger series converging around 75 to 80 percent [3,22,41,42,49]. The sex distribution of patients with PNES is more evenly distributed in specific subpopulations, such as children, individuals with intellectual disability, and older adults [22,38,46,50], possibly reflecting different mixes of etiologic factors.
●Comorbid conditions – Patients with PNES have a high burden of psychiatric disease, epilepsy, developmental disabilities, recurrent medical evaluations for unexplained somatic complaints, and other conditions, as discussed below. (See 'Common comorbidities' below.)
CLINICAL MANIFESTATIONS
Precipitants and circumstances
●Witnesses present – Most episodes of PNES occur in front of witnesses [46,51,52]. In one study, the occurrence of an episode in the doctor's waiting or examination room was estimated to have a 75 percent predictive value for PNES [51]. Similarly, in a study of patients undergoing EEG monitoring, episodes that occurred at the time of electrode placement were found to be PNES, not epileptic seizures [53]. In a study of psychological intervention, PNES occurred four times as commonly during sessions as epileptic seizures [54].
●Relationship to sleep – At the time of taking a history, PNES are reported to arise some of the time from sleep in up to 50 percent of patients, and relatives may confirm they have witnessed this [55-57]. When recorded in a monitoring unit, PNES do not occur during sleep [53,58] but have been noted to arise quickly after awakening [56]. A history of events that arise exclusively from sleep suggests epilepsy or sleep disorder (eg, rapid eye movement [REM] behavioral disorder) [57]. (See "Sleep-related epilepsy syndromes", section on 'Clinical features' and 'Video-EEG monitoring' below.)
●Stress – While it is intuitive that PNES might be associated with stressful situations, witness responders may be more likely to report this than patients [59]. Stress is also commonly cited as a seizure precipitant in patients with epilepsy [60,61]. As an item of initial history, this is a relatively poor discriminator.
●Menstrual cycle – Increased seizure frequency during the perimenstrual time period suggests epileptic seizures. In one series, perimenstrual exacerbation was associated with 13 of 27 patients with epileptic seizures versus 1 of 38 patients with PNES [62]. (See "Initial treatment of epilepsy in adults", section on 'Catamenial epilepsy'.)
●Pregnancy – PNES may present during pregnancy [24,63,64], presumably on the basis that pregnancy may be stressful.
●Emergence from general anesthesia – PNES are known to occur on emergence from general anesthesia [65-67], possibly due to disinhibiting effects of anaesthetic agents.
Patient report of the events — Some patients are reluctant to describe their experiences of PNES in detail [68,69], so clinical information may be sparse. However, an aura is reported by some patients (25 to 60 percent) and may be a more common symptom than in epilepsy [5,7,51]. One study of 100 patients [59] found that the experiential content of the aura was variable, with an overlap in symptomatology between the two conditions. Nonetheless, items such as déjà vu, ictal tongue biting, burns, and frothing from the mouth were never experienced by the majority of PNES patients. Another study [70] of 354 patients with epilepsy and PNES found that symptoms of panic attacks or hyperventilation (palpitation, sweating, dyspnea, choking feeling, chest discomfort, dizziness, derealization, fear of dying, paraesthesia) were more frequent in patients with PNES.
Eyewitness report of the events — To the eyewitness, PNES can take a variety of forms, but two broad forms, convulsive PNES and swoon PNES, account for the great majority [5,40,41,45,52,71-73].
Convulsive PNES — This form might be mistaken for a tonic-clonic seizure (table 1A). It consists of a fall if the patient is upright, with movements of all four limbs, often trunk and head. The movements are most often tremors, which can be low frequency-high amplitude (often called alternating movements), or high frequency-low amplitude. "Thrashing" movements do occur but are less common. There may be back arching and side-to-side low frequency high amplitude tremors of the head. This broad form accounts for the majority of PNES.
Swoon PNES — This form might be mistaken for a syncopal episode, as it consists of a fall to the ground or "slump" to one side if sitting, with little or no movement (table 1B). The eyes are usually closed, and the patient is unresponsive. This type accounts for a significant minority of PNES [40,73]. When events of this type are prolonged, they are almost always PNES [72]. If the event is directly witnessed, then the pulse should immediately be checked, as cardiac syncope or arrest may look the same. It should be ascertained that an eyewitness account extends from before the beginning until after the end of the event, as many epileptic seizures are followed by a period where the patient lies still.
Intermediate PNES — Some PNES are intermediate, in the sense that there is a fall to the ground with variable tremor. Some patients may have both types at different times or may have a swoon that evolves into a convulsive PNES.
Absence and myoclonic presentations — Absence and myoclonic presentations are uncommon in adults and appear to occur most often in teenagers. There is little information in the literature regarding clinical or other features. Absence type PNES look clinically like petit mal absence, with a short (seconds) arrest of activity and responsiveness and may be similarly frequent. Unlike epileptic absence seizures, they usually do not interrupt the patient during speaking or other activities. Myoclonic PNES present as jerking movements with onset in the teens, the common differential diagnosis being juvenile myoclonic epilepsy (JME) prior to the first seizure. The onset of the jerks is often abrupt, and there is no morning predominance (as in JME). Once seen on video, the movements, while sudden and brief, are often more complicated than a simple myoclonic jerk and may involve trunk and head.
Other behaviors — Other behaviors may occur during PNES, and some are useful in distinguishing them from epileptic seizures (table 1A). One systematic review found that long duration, fluctuating course, asynchronous movements, pelvic thrusting, side-to-side head or body movements, ictal eye closure, ictal crying, memory recall, and absent postictal confusion were the most reliable signs in distinguishing PNES from epileptic seizure; occurrence out of true sleep favored epileptic seizure (though this can only be confirmed during EEG monitoring) [58]. However, no single semiologic feature is sufficiently sensitive or specific for PNES to be useful as a sole indicator [49]. In addition, patients and eyewitnesses can report clinical features that are discrepant from each other and from review of video-recorded events [59].
●Motor activity – A variety of convulsive-like motor activities can occur in PNES. While motor manifestations of an epileptic seizure usually take the form of brief tonic posturing or a synchronized convulsion with a defined progression of motor activity, movements in PNES are more often tremors (see 'Convulsive PNES' above), synchronous or alternating, and may wax and wane over the course of the ictus [37,40,58,74]. Specific movements such as writhing, thrashing, pelvic thrusts, opisthotonus (arched back), and jactitation (rolling from side to side, side to side head movements) suggest PNES, but these are not always present, particularly in children [40,46,52,72,74,75]. Consistent lateralization of motor activity other than tonic posture can be seen in PNES.
Many patients have PNES with consistent semiology and little variation [55,58,73,76] (ie, the events are stereotyped), so this characteristic is of little or no discriminatory value, especially as seen through the filter of an eyewitness account. (See "Focal epilepsy: Causes and clinical features", section on 'Frontal lobe epilepsy' and "Sleep-related epilepsy syndromes", section on 'Sleep-related hypermotor epilepsy (previously NFLE)'.)
●Tongue-biting and self-injury – Classic symptoms of epileptic seizures such as tongue-biting, incontinence, and self-injury are more common in epileptic seizures, but they can be reported in a third or more of patients with PNES [2,7,40,41,58,71,77,78]. A tongue bitten on the side (versus the tip) and severe tongue-biting (with laceration) are more specific for epileptic seizure [78,79]. Seizure-related thermal burn injuries are also highly specific for epileptic seizures [80], whereas friction burns suggest PNES.
●Level of awareness – Incomplete loss of consciousness during the episode, suggested either by partial responsiveness to stimuli or by later recall of events during segments of ictal unresponsiveness, supports PNES [40,81,82]. However, the patient account of an episode cannot distinguish unconsciousness from amnesia, and the eyewitness account cannot distinguish unconsciousness from unresponsiveness. The presence of an alpha rhythm on the EEG (ie, neurophysiologic evidence of wakefulness) during an episode in which the patient is clinically altered or amnestic strongly supports a diagnosis of PNES (versus epileptic seizure). (See 'Video-EEG monitoring' below.)
●Vocalizations – Ictal features of emotional overlay, such as weeping, stuttering, and vocalizations with affective content, are uncommon in epileptic seizures and suggest PNES [46,83-86]. When vocalization occurs in an epileptic seizure, it is usually isolated to the seizure onset and is frequently characterized by an inarticulate cry or fragmented guttural utterance, caused by a tonic diaphragm forcing air against tonic or clonic vocal cords [87]. By contrast, the ictal vocalization of PNES may exhibit affective content and not only occur at seizure onset, but also persist or even intensify through the course of the seizure and may persist into or arise during the postseizure period.
●Auras – A seizure aura is frequently reported in PNES (25 to 60 percent) and may be a more common symptom than in epilepsy [5,7,52]. (See 'Patient report of the events' above.)
●Autonomic signs – Autonomic manifestations during an ictus (eg, tachycardia, cyanosis) suggest epileptic seizure, and their absence, particularly during a major convulsion, suggests PNES [74,86,88].
●Eye closure – Eyes are usually open during the ictus of a convulsive epileptic seizure [58,75,79,81,89]. Video EEG studies of the sensitivity and specificity of eye closure as a sign for PNES have yielded variable results [74,81,90] according to the definition of eye closure and, in these series, observer reports of ictal eye closure were not reliable. If the patient resists an attempt on the part of the examiner to open the eye(s), then that suggests PNES.
●Duration – While the ictus of an epileptic seizure is typically very brief, often less than one minute, PNES are rarely less than one minute long and are usually much longer [39-41]. Prolonged episodes (ie, psychogenic status epilepticus) are not rare. In one series, 78 percent of patients with PNES reported at least one event longer than 30 minutes, and half of these patients had recurrent episodes of psychogenic status epilepticus [91].
●Frequency – Patients with PNES generally report a higher seizure frequency than patients with epilepsy [92]. Suspicion for PNES is raised by daily convulsive events, despite a normal exam and alert appearance during history-taking [91]. Demonstration of an apparently explosive attack burden from the moment of the first event and onward can also be characteristic of PNES [93].
Postictal features
●Return to baseline – Rapid alerting and reorientation are common after PNES but uncommon with epileptic seizures, except for certain seizure types, such as absence or frontal lobe seizures [52,55,58,81,94]. However, in some case series, half or more of patients with PNES exhibited postictal confusion or drowsiness [5,52].
●Respiratory changes – The postictal period after a generalized tonic-clonic seizure was found in one cohort to be characterized by a breathing pattern of deep and prolonged inspiratory and expiratory phases (stertorous breathing pattern), compared with shallow, rapid respirations in patients after a PNES [58,75,89]. Epileptic seizures arising from the frontal lobe, however, were associated with a postictal breathing pattern similar to PNES. Another case series described a stertorous postictal breathing pattern after 6 of 27 epileptic seizures, but none of the 15 observed PNES [74].
A key caution is that neurobehavioral manifestations during the postictal phase of epileptic seizures can resemble the ictal features of PNES in some patients [95]. During history-taking, it is important to clarify that the witness has observed the onset of the episode.
COMMON COMORBIDITIES
Concurrent or past epilepsy — Estimates of the prevalence of concurrent epilepsy among patients with PNES vary widely [3,5,7,22,96-102]. The true figure probably lies in the 5 to 22 percent range, and large studies suggest that the prevalence of epilepsy is approximately 10 percent for the general PNES population [3,22]. Several factors make it difficult to know what the true figure is, including lack of population data, differing diagnostic criteria used to determine the coexistence of epilepsy, a tendency to see making a dual diagnosis as a "safe" option, and the tendency for large centers to see difficult cases.
The typical clinical course of adults with both diagnoses is that the onset of epilepsy predates the onset of PNES by approximately 10 years [92]. In such cases, the patient's own history of epilepsy may be the model for their PNES. In one case series of patients with video-electroencephalography (EEG) documentation of both PNES and epileptic seizures, most patients (18 of 20) had notably different clinical semiology for PNES compared with their epileptic seizures [103]. However, another series found that patients with PNES and probable temporal lobe epilepsy had symptoms that appeared to be more characteristic of temporal lobe seizures, possibly because of learned behavior from their own seizures [104], though in some cases that may raise the question of how they know what that behavior is.
A few patients (2 to 9 percent) with epilepsy develop PNES after epilepsy surgery [105-107]. These typically emerge soon after surgery. This phenomenon may be due in part to a psychological dependence on seizures, although timing is similar to that of other psychiatric complications of epilepsy surgery.
Developmental disabilities — PNES is well described in patients with neurologic developmental disabilities [25,45,108]. The prevalence of mixed PNES with epilepsy can be up to 30 percent in this population [50]. Among these patients, distinguishing PNES from epileptic seizures and other repetitive, stereotyped behaviors is a particular challenge. In one study, individuals with PNES and learning disabilities were more likely to have circumstantial triggering of events and more prolonged events than individuals without learning disabilities [3].
Psychopathology — While patients with PNES have a higher burden of psychiatric disease and symptoms than control groups with epilepsy, these conditions are also common in patients with epilepsy and are not specific for PNES [2,6,39,40,42,109-114]. The presence of these conditions is often not known at the time of presentation with PNES and may instead be revealed later as part of the evaluation when sufficient rapport has been established. (See "Psychogenic nonepileptic seizures: Management and prognosis", section on 'Neuropsychological testing and psychiatric evaluation'.)
Psychiatric conditions associated with PNES include [23,71,110,111,115-119]:
●Depression
●Anxiety
●Somatic symptom and related disorders
●Post-traumatic stress disorder
●Dissociative disorders
●Personality disorders, especially borderline personality disorder, but also narcissistic, histrionic, and antisocial personalities
It is common for patients to have more than one of these conditions [110]. (See 'Etiology' above.)
By contrast, depression with major psychosis and schizophrenia are relatively uncommon in patients with PNES [45].
Unexplained somatic complaints — A history of recurrent medical evaluations for other unexplained somatic complaints is common. In one series, 30 percent of PNES patients had a sufficient burden of gastrointestinal complaints, pain, fatigue, and other complaints to meet criteria for an undifferentiated somatoform disorder [26]. In another study that included over 250 patients with either PNES or epilepsy who underwent video-EEG monitoring over a two-year period, patients with PNES were significantly more likely than those with epilepsy (66 versus 27 percent) to have one or more of the following diagnoses: fibromyalgia; chronic fatigue syndrome (CFS), also known as myalgic encephalomyelitis/chronic fatigue syndrome (ME/CFS); chronic pain syndrome; headaches; irritable bowel syndrome; asthma; and gastroesophageal reflux disease [27].
Others — An increased risk of prior physical insults has been described in patients with PNES, including antecedent minor head trauma in 30 to 44 percent of patients [6,26,67,96,120,121], health-related complications [3], and intracranial surgical procedures [67]. However, recall bias and other psychological factors may contribute to these apparent associations.
DIAGNOSTIC EVALUATION
DSM criteria — Whereas conversion disorder was previously approached as a diagnosis of exclusion, the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5) stipulates "rule-in" of a conversion disorder diagnosis based on inclusion of clinical findings that are incongruent to known anatomy, physiology, or diseases (eg, resistance to attempts by the examiner to open the eye) [48]. The previous requirement to exclude feigning was abandoned, since this can be difficult to detect without surveillance or forensic evaluation. DSM-5 no longer requires the presence of psychologic factors to precede or exacerbate conversion symptoms, as such factors may not always be immediately apparent from the history. This is in good accord with the practical demands of the diagnosis of PNES. (See "Conversion disorder in adults: Terminology, diagnosis, and differential diagnosis".)
When to suspect the diagnosis — Factors that should prompt examination of a diagnosis of PNES are summarized in the table (table 2). Evidence from retrospective studies suggests that the diagnosis of PNES is delayed by an average of seven years from the time of onset and may be delayed by up to 10 to 15 years in some cases [2,77,122]. It is therefore important to consider PNES when evaluating patients with paroxysmal events, as missing this diagnosis can be consequential. Recognizing PNES can be challenging even for experienced observers. Nonetheless, clues that raise suspicion for this diagnosis are often apparent from the clinical history. Differences between tonic-clonic seizures and convulsive PNES, and between syncope and swoon PNES, are summarized in the tables (table 1A-B) and discussed above. (See 'Eyewitness report of the events' above.)
●Importance of history – As with other paroxysmal events, diagnosis is highly dependent on history, from the patient and from eyewitnesses, and can be challenging.
Patients are usually seen in the interictal state when there are no helpful clinical signs (neurologic examination findings in the interictal period do not contribute to the diagnosis of PNES itself) and the tests we can bring to bear have significant limitations. We therefore rely heavily on history from patients and from eyewitnesses to the events. Tests that detect disordered physiology (or lack of it) during the events (eg, video electroencephalography [EEG]) can be highly sensitive and specific, but are resource-demanding and are only useful if an event can be induced or if events are occurring often enough to catch by chance.
●Poor response to antiseizure medications – Many patients with PNES have seizures for years prior to diagnosis, and most have been treated unsuccessfully with antiseizure medications [41,49,77]. A failure to make even small improvements in seizure frequency despite good trials of antiseizure medications suggests the diagnosis of PNES. However, PNES may respond partially to treatment with antiseizure medications [123,124], and paradoxical worsening of PNES with antiseizure medication treatment may be associated with PNES. For these reasons, response to treatment with antiseizure medication should be viewed with caution as a diagnostic criterion.
Patients who present with prolonged PNES are often treated with drug protocols for status epilepticus and fail to respond [125-127]. (See 'Consequences of missed or delayed diagnosis' below.)
Referral — Once the diagnosis of PNES is suspected, then referral to a neurologist (better still, an epilepsy specialist neurologist) is indicated. If EEG is available, then a standard EEG recording with video may well induce an event and should be done at an early stage. Relatives should be encouraged to record events using a phone camera. These measures can usefully shorten the diagnostic process.
Differential diagnosis — The differential diagnosis of PNES is wide (table 3), though in the majority of cases it narrows down to relatively few diagnostic distinctions, involving syncope, epileptic seizure, and PNES (table 1A and table 1B). The primary differential diagnosis of PNES (table 3) has traditionally been thought to be "epileptic seizures" [97]. However, the main differential diagnosis of "swoon" types PNES is syncope (vasovagal or cardiac), as epileptic seizures do not usually present with "fall down lie still" events.
Therefore, for the majority of patients, the initial differential diagnosis is more precisely expressed as follows:
●Tonic-clonic seizure versus convulsive PNES (table 1A)
●Syncope (vasovagal or cardiac) versus swoon PNES (table 1B)
●Absence seizure versus absence PNES
The differential diagnosis also includes several less common conditions:
●Stereotyped behaviors – In patients with learning disabilities, stereotyped behaviors should be in the initial differential diagnosis [24].
●Hypermotor frontal lobe seizures – Hypermotor frontal lobe seizures are rare relative to PNES but may be mistaken for them, as they are associated with high amplitude proximal automatic (thrashing) movements that are variably eupraxic and adapted to the physical environment [37,55,58,94,128-131]. The clinical distinction is aided by the following characteristics of hypermotor frontal lobe seizures that are not usual in PNES:
•They are brief (seconds to tens of seconds)
•They often occur exclusively during sleep
•They may include unilateral tonic posturing
•They may be accompanied by "bizarre" facial expressions
In one video-EEG study, tonic contraction of the upper extremities in abduction was a feature of 90 percent of 63 supplementary motor seizures and occurred in none of 111 PNES [55]. Where attacks occur exclusively during sleep, the main differential diagnosis with epilepsy is more often sleep disorder than PNES. (See "Focal epilepsy: Causes and clinical features", section on 'Frontal lobe epilepsy'.)
●Other nonepileptic paroxysmal disorders – Other nonepileptic paroxysmal disorders may be mistaken for either epileptic seizures or for PNES [97]. These vary by age group and include sleep disorders, movement disorders, migraine, and syncope (table 3 and table 1B). These are discussed in detail separately. (See "Approach to abnormal movements and behaviors during sleep" and "Nonepileptic paroxysmal disorders in adolescents and adults" and "Evaluation and management of the first seizure in adults".)
Confirming the diagnosis — The best confirmation of a diagnosis of PNES is achieved by recording events that are confirmed by eyewitnesses as typical, using video EEG monitoring with electrocardiogram (ECG). This should always be the aim. Aside from their diagnostic value, video-EEG recordings of events are useful in confirming the seizure semiology with witnesses and in conferring the ability to communicate the diagnosis positively and with conviction. (See 'Video-EEG monitoring' below and "Psychogenic nonepileptic seizures: Management and prognosis", section on 'Explaining the diagnosis'.)
Often, an initial standard EEG and video recording with hyperventilation, photic stimulation and simple verbal suggestion can achieve this aim without the need for hospital admission, and this is usually our first step. If no attack is captured, then we undertake video EEG monitoring in hospital. Where there are clinical indicators of comorbid epilepsy, potentially causative lesions on imaging, or when the patient is on antiseizure medications that have to be withdrawn for monitoring, we admit to hospital for monitoring as the initial step. Serum markers for seizures can be useful in emergency department and other situations in which neurology expertise is not immediately at hand. We do not use them in our ward, preferring to obtain video footage that can be shown to an epilepsy specialist [132].
All patients are imaged, preferably with MRI, if for no other reason than that it can be difficult to convince patients that the diagnosis is not a medical one without it.
In some situations, such as when events take place only at school or in other specific situations, then we would initially use ambulatory EEG monitoring in preference, despite its disadvantages. (See 'Other monitoring techniques' below.)
Video-EEG monitoring — Video-EEG monitoring combines extended EEG monitoring with time-locked video acquisition that allows for analysis of clinical and electrographic features during a captured event. A single ECG channel should be monitored at the same time. The yield of monitoring is dependent on patient selection but is generally good; 73 to 96 percent of patients will have typical PNES within the first 48 hours of recording [39,53,133,134].
If an event is clinically compatible with PNES, then the diagnosis is best confirmed by the observation of a normal awake EEG before, during, and after, the same applying (sinus tachycardia aside) to the ECG. The event then has to be identified as the only type of event that has occurred (the word "typical" is often used in this context, but "only" is more exact). This is done by careful comparison of the recorded event with descriptions of past events, preferably in addition to showing the video recording to family, caregivers, and others.
Clinical context is crucial. Alpha rhythm is the neurophysiologic correlate of alertness and is not expected to be present during any event that includes lack of responsiveness. Syncope (cardiac or vasovagal) is seldom recorded in the monitoring unit, but cerebral hypoperfusion coincides with a progression from theta slowing, to delta slowing, to flattening, depending on severity [135], as well as bradycardia.
Discordant findings (eg, lack of responsiveness with concurrent alpha rhythm on EEG) are supportive of PNES.
To make this judgement, it is obvious that the EEG waveforms during the event (or at least a useful part of it) should be discernable despite muscle and movement artefact. Rarely, brief tonic seizures may occur, with flattening of the EEG the only observable change. Thus, video-EEG interpretation can require care, skill, and sometimes recordings of multiple events to identify changes with confidence.
There are two uncommon but well-known scenarios in which epileptic seizures may pass without EEG change. The first is easily identified. Focal aware (simple partial) seizures (eg, twitching of a finger, or a subjective experience) often involve only a small (and deep) volume of cortex, and may not be accompanied by detectable surface EEG change [136]. Second, hypermotor frontal lobe seizures may occur with the EEG either entirely obscured by movement and muscle artefact, or normal (see "Video and ambulatory EEG monitoring in the diagnosis of seizures and epilepsy"). Clinical features suggestive of hypermotor frontal lobe seizures are discussed elsewhere in the topic. (See 'Differential diagnosis' above.)
When the history suggests more than one independent event type, each type should be recorded on video-EEG and identified. Event types that are not documented on video-EEG should be diagnosed with a more cautious level of certainty (table 4) [137].
One study of inter-rater reliability of video-EEG found only moderate agreement (kappa = 0.57) for the diagnosis of PNES [138]. Limitations of this study that differ from "real-life" use of this test include a forced choice paradigm (epileptic versus psychogenic nonepileptic versus physiologic nonepileptic seizures) that was based on the analysis of a single episode with no access to other clinical data.
Event induction — Where video and ECG are also recorded, standard EEG (including hyperventilation and photic stimulation) may be supplemented with the use of suggestion or nocebo designed to induce a PNES and across studies, approximately 70 to 90 percent of patients with PNES will have a typical event [24,49,52,139-142]. This procedure has attracted some controversy, especially in which a nocebo such as saline injection is used. The use of nocebo appears to have declined, but there is now widespread acceptance of simple verbal suggestion in combination with hyperventilation and photic stimulation. Our preference is to carry out initial standard EEG with hyperventilation, photic stimulation, and simple verbal suggestion. We do not do this in the monitoring unit if the patient has already had an unsuccessful attempt at induction, as we have found the yield to be very low. We do not use nocebo. The diagnostic yield of simple verbal suggestion may be as good as with nocebo in any event [52,139,140,143] and does not appear to be impaired by informing the patient that a psychogenic seizure may be recorded [140,143].
The success rate of induction may be higher among patients with certain clinical characteristics, including convulsive presentation, uncommon cognitive and affective self-reported symptoms, absence of prior induction exposure, and a history of previous events in medical settings [140,142].
Note that epileptic seizures may also occur with induction (unsurprisingly in the cases of hyperventilation and photic stimulation) [97,100,133,144-147]. A patient with epilepsy may experience new and nonepileptic symptoms under these circumstances, but the features are usually atypical of their usual seizures and hence have limited clinical relevance [148]. As with video recordings obtained conventionally, all recorded events should be carefully compared with descriptions of previous attacks and shown to relatives and appropriate others.
Routine (interictal) EEG — When using routine EEG in patients suspected of having PNES, it is helpful to remember some salient points:
●The sensitivity of routine EEG for epilepsy has been estimated at between 20 and 55 percent. Therefore, even at the upper end of this range, a negative result does not usefully exclude epilepsy.
●If a routine EEG recording captures epileptiform abnormalities, then this can usefully confirm a clinical diagnosis of epilepsy. However, it does not exclude PNES.
●In up to 30 percent of patients with PNES, routine EEG with hyperventilation and photic stimulation will provoke a typical event [139], usefully confirming the diagnosis of PNES.
●The sensitivity of routine EEG (if no clinical event is captured) for PNES is zero.
Interictal EEG abnormalities are also seen in some patients with PNES only [77,103], though these are usually non-epileptiform and of limited diagnostic significance for epilepsy. If clearly epileptiform abnormalities are seen in a patient otherwise felt to have PNES alone, then re-examination of the diagnosis (of both PNES and past or present epilepsy) is indicated. Abnormal interictal EEG is more common if there is a past history of epilepsy, learning disability, or other underlying neurologic injury or disease. (See "Electroencephalography (EEG) in the diagnosis of seizures and epilepsy".)
Other monitoring techniques — Where no episode is recorded or video EEG is not available, outpatient ambulatory EEG, ideally with concomitant video, may be helpful, as may video recording by phone camera. For video recordings, the absence of concomitant EEG is a significant limitation [74], as is the fact that the beginning of the event is not usually captured. However, observation of the clinical event on video can provide a key criterion to support higher levels of diagnostic certainty in the ILAE staged approach to the diagnosis of PNES (table 4) [68]. (See 'Postictal features' above and 'Levels of diagnostic certainty' below.)
Serum testing — Certain laboratory studies can help differentiate convulsive PNES from generalized epileptic seizure. However, these tests come with caveats, lack sensitivity and specificity, and should not be carried out in lieu of a good description of the witnessed ictus or a mobile phone video recording (that can be shown to an experienced observer).
●Prolactin – The pooled sensitivity of an elevated prolactin (twice the baseline level) from several studies is 60 percent for generalized tonic-clonic seizures and 46 percent for focal seizures with impaired awareness [149]. The sensitivity is lower for focal seizures with retained awareness as well as for frontal lobe seizures [150].
The timing of measuring prolactin is crucial. In typical cases, prolactin levels peak 15 to 20 minutes after the seizure and return to baseline levels within an hour [149]. Considering the known circadian fluctuation of serum prolactin [151], the optimal baseline prolactin level for comparison purposes should be drawn at approximately the same time on the next day after the initial postictal serum prolactin measurement and at least six hours after the last seizure. Prolactin elevations have not been well characterized in the setting of repeated seizures or status epilepticus and have unclear utility in such settings.
Prolactin levels are unlikely to rise after a PNES, but elevations can occur after syncope [149,152]. However, a nonelevated prolactin level does not imply PNES. Conditions associated with hyperprolactinemia may confound the results, even when paired baseline and postictal prolactin measurements are available. These conditions include pregnancy/lactation, prolactinomas, primary hypothyroidism, and certain drugs (eg, dopamine antagonists) [149]. (See "Causes of hyperprolactinemia".)
●Other serum markers – Other laboratories that may help distinguish PNES from epileptic seizures and other physiologic events include creatine phosphokinase (CPK), cortisol, white blood cell count, lactate dehydrogenase, partial pressure of carbon dioxide (pCO2), ammonia, and neuron-specific enolase [153-156]. CPK levels in particular are often elevated after generalized tonic-clonic seizures but not after partial seizures. The later rise and prolonged elevation of CPK, up to 24 hours postictally, make this test somewhat more useful in the outpatient setting. However, a defined threshold level for abnormality, sensitivity, and specificity remains to be determined for CPK, as for other serum markers [157,158].
Neuroimaging — We image the brain in all patients with suspected PNES, usually with magnetic resonance imaging (MRI). A neuroimaging study, preferably MRI, should be obtained detect a potential structural cause for epileptic seizures. A normal MRI scan is almost always required when explaining the diagnosis to patients and relatives in any event. However, there is no direct role of structural imaging in the diagnosis of PNES. Several series report abnormal brain MRI in 10 to 38 percent of patients with PNES [5,41,77,121], while many patients with epileptic seizures have normal brain MRI.
There is no established role of functional imaging in the evaluation of suspected PNES. Abnormalities on interictal single-photon emission computed tomography (SPECT) have been reported in both PNES and epilepsy. A change in focal perfusion from an ictal or postictal SPECT compared with an interictal SPECT has been found to distinguish epileptic seizures from PNES where the patient has both [159-161]. While SPECT may be useful in principle when movement-related muscle artifact obscures EEG interpretation, it is practically difficult to carry out, and the sensitivity and specificity are uncertain. We do not use this test.
Levels of diagnostic certainty — In practice, ideal diagnostic information is often not achieved. Events may not occur during monitoring, there may be no eyewitness account, or video-EEG may simply not be available.
Recognizing these limitations, the Nonepileptic Seizure Task Force of the International League Against Epilepsy (ILAE) delineates how PNES can be diagnosed with varying levels of certainty according to the combination of three elements (table 4) [68]:
●Event description by patient and/or witness
●Clinician observation of seizures as an in-person witness or by video recordings
●Available ictal and interictal EEG data
Earlier recognition of PNES under this graded approach may promote earlier exploration of potential psychologic underpinnings and treatment options. In patients whose clinical data meet lower levels of diagnostic certainty, iterative assessments over time by the treating neurologist remain vital, as emergence of diagnostic data might refute an initial diagnosis.
CONSEQUENCES OF MISSED OR DELAYED DIAGNOSIS — Patients with PNES may suffer inappropriate long-term treatment with antiseizure medications [49]. Prolonged episodes, psychogenic status epilepticus in particular, are often treated with toxic antiseizure medication doses, intubation, and iatrogenically induced coma [49,64,125,162-165]. Serious adverse events and death may result [126,165]. There are costs and burdens to the individual and to the health care systems from recurrent visits to the emergency department and hospitalizations for uncontrolled, unrecognized PNES [166,167]. They also have the opportunity cost of delaying appropriate psychologic treatment, something that has been associated with worse outcome [21].
SUMMARY AND RECOMMENDATIONS
●Psychogenic nonepileptic seizures (PNES) are events thought to have mainly psychologic origins. They clinically mimic epileptic seizures or syncope but are not associated with abnormal neuronal activity or reduced perfusion to the brain. (See 'Definition and terminology' above and 'Etiology' above.)
●PNES has a female predominance. (See 'Epidemiology and population characteristics' above.)
●PNES include a variety of clinical manifestations, some of which are suggestive, although not independently diagnostic, in distinguishing PNES from other differential diagnoses; these are listed in the table for convulsive PNES and swoon PNES, the two most common forms (table 1A-B). (See 'Clinical manifestations' above.)
●Patients with PNES have a high burden of psychiatric disease, epilepsy, developmental disabilities, recurrent medical evaluations for unexplained somatic complaints, and other conditions. (See 'Common comorbidities' above.)
●The diagnosis of PNES is generally established by video-electroencephalography (EEG) monitoring, in which captured clinical events are examined in conjunction with EEG activity. Most patients with PNES will have an event within two days of monitoring. Other tests (interictal EEG, neuroimaging, neuropsychological tests, and laboratory studies) are used primarily to investigate alternative etiologies and are not diagnostic of PNES. (See 'Diagnostic evaluation' above.)
●For the majority of patients, the differential diagnosis of PNES includes tonic-clonic seizure, syncope, and absence seizure. Hypermotor frontal lobe seizures, which are rare, can have unusual clinical features and can be confused with PNES, especially when there is no correlate on the surface EEG. Features that suggest that the events are frontal lobe seizures include short duration, highly stereotyped features, and occurrence during physiologic sleep. (See 'Differential diagnosis' above.)
●The management and prognosis of PNES are reviewed separately. (See "Psychogenic nonepileptic seizures: Management and prognosis".)
ACKNOWLEDGMENTS — The UpToDate editorial staff acknowledges David K Chen, MD, who contributed to earlier versions of this topic review.
The editorial staff at UpToDate would also like to acknowledge Alan Ettinger, MD, MBA, who contributed to an earlier version of this topic review.
