Skip to main content
NCBI home page
Neurology: Clinical Practice logoLink to Neurology: Clinical Practice
. 2022 Jun;12(3):234–247. doi: 10.1212/CPJ.0000000000001170

Using Semiology to Classify Epileptic Seizures vs Psychogenic Nonepileptic Seizures

A Meta-analysis

Subramanian Muthusamy 1, Udaya Seneviratne 1, Catherine Ding 1, Thanh G Phan 1
PMCID: PMC9208424  PMID: 35747545

Abstract

Background and Objectives

A misdiagnosis of psychogenic nonepileptic seizures (PNES) and epileptic seizures (ES) is common. In the absence of the diagnostic gold standard (video EEG), clinicians rely on semiology and clinical assessment. However, questions regarding the diagnostic accuracy of different signs remain. This meta-analysis aimed to evaluate the diagnostic accuracy of semiology in PNES and ES.

Methods

We systematically searched PubMed, PsycInfo, and Medline for original research publications published before 8 February 2021 with no restriction on search dates to identify studies that compared semiology in ES and PNES in epilepsy monitoring units. Non-English publications, review articles, studies reporting on only PNES or ES, and studies limited to patients with developmental delay were excluded. Study characteristics and proportions of event groups and patient groups demonstrating signs were extracted from each article. A bivariate analysis was conducted, and data were pooled in a random effects model for meta-analysis. The I2 statistic was calculated to assess statistical heterogeneity. The revised Quality Assessment of Diagnostic Accuracy Studies tool was used to assess the risk of bias in included studies. The positive likelihood ratio (PLR) and negative likelihood ratio (NLR) were calculated. A PLR >10 or an NLR <0.1 largely affected the posttest probability of a diagnosis (ES or PNES), whereas a PLR between 5 and 10 or an NLR between 0.1 and 0.2 moderately affected the posttest probability of a diagnosis (ES or PNES).

Results

The meta-analysis included 14 studies comprising 800 patients with ES and 452 patients with PNES. For PNES, ictal eye closure (PLR 40.5 95% confidence interval [CI] 16.2–101.3; I2 = 0, from 3 studies) and asynchronous limb movements (PLR 10.2; 95% CI 2.8–37.7; I2 = 0, from 3 studies) reached a PLR threshold >5. No single sign reached a PLR threshold >5 for ES.

Discussion

While all signs require an interpretation in the overall clinical context, the presence of ictal eye closure and asynchronous limb movements are reliable discriminative signs for PNES.

graphic file with name NEURCLINPRACT2021070273FFU1.jpg

Psychogenic nonepileptic seizures (PNES) resemble epileptic seizures (ES) but are not associated with an underlying electrophysiologic abnormality.1 Although PNES is a common neuropsychiatric condition with an estimated prevalence up to 33 per 100,000 individuals and accounts for one quarter of patients admitted to epilepsy monitoring units, a misdiagnosis as an ES is common.2,3 Patients with PNES experience long diagnostic delays, receive inappropriate treatments, incur increased health-care utilization costs, and have a standardized mortality rate 2.5 times that of the general population.4-6 A clinical assessment by an expert with video EEG monitoring (VEM) of typical events is the diagnostic gold standard.7 In the absence of this expertise, clinicians make the diagnosis on the basis of clinical assessment and seizure semiology in routine clinical practice (based on patient or eyewitness account, direct visualization, or seizure videos when available).

A recent study8 provided class II evidence demonstrating the high accuracy of smartphone videos of seizures as an adjunct to the clinical assessment of patients presenting with seizures, but questions remain as to what signs best discriminate between PNES and ES in the absence of VEM. Certain signs occur more frequently in ES or PNES. For example, studies have consistently demonstrated that ictal eye closure or prolonged seizure duration are more frequently observed in PNES than ES.9,10 One previous meta-analysis reviewed the diagnostic utility of ictal eye closure in PNES and reported a pooled specificity of 0.895 and a positive likelihood ratio (PLR) of 5.5, but this study combined aggregate data reported at the patient-group and event-group levels.9 This has not been replicated for other clinical signs.11-13 Against this backdrop, we sought to evaluate the diagnostic accuracy of different signs in differentiating ES from PNES.

Methods

The primary research question of this meta-analysis was to evaluate the diagnostic accuracy of different signs in differentiating ES from PNES. This meta-analysis was not prospectively registered. It was conducted in keeping with the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) guidelines.14

Search Strategy and Study Eligibility

We systematically searched PubMed, PsycInfo, and Medline for original research publications published before 8 February 2021 with no restriction on search dates. The search strategy included keywords or Medical Subject Heading terms relating to epileptic seizures, psychogenic nonepileptic seizures, and different clinical signs. These signs included the following: eye closure, eyelid fluttering, gaze deviation, head version, axial posturing, back arching, pelvic thrusting, asynchronous movements, temporal profile, emotions, limb postures, and vocalization. A detailed description of the search strategy is outlined in eAppendix 1 (links.lww.com/CPJ/A334). The reference lists of relevant studies (included publications and review articles) were screened to identify additional publications missed by the electronic search.

Inclusion criteria were as follows: (1) adult or pediatric patients; (2) prospective or retrospective study design; (3) studies comparing 1 or more signs in ES and PNES; and (4) the diagnosis of ES and PNES and the signs confirmed using VEM. Exclusion criteria were as follows: non-English publications, review articles, and studies reporting on only PNES or only ES. Studies limited to patients with intellectual disability (ID) were summarized and discussed separately because they represent an important but distinct subgroup of patients with PNES (as highlighted by ILAE PNES Taskforce15), and differentiating nonseizure stereotyped behaviors from PNES and ES can be challenging in this subgroup.16 A targeted systematic literature search (using PubMed and MEDLINE) using terms synonymous with intellectual disability and psychogenic nonepileptic seizures was performed to identify relevant studies. The reference lists of these relevant studies were screened to identify additional suitable publications.

Data Extraction and Quality Assessment

The results from the systematic search were imported into the Covidence platform (covidence.org), and duplicates were removed. Two authors independently reviewed the abstracts and titles of publications for eligibility. Subsequently, 2 authors reviewed the full text of publications for inclusion in the study. Any disagreements were resolved by consensus.

Data collected from the studies included the following: study author(s), year of publication, number of patients and events, and proportion of patients or events displaying a particular sign. The proportion of patient groups or event groups demonstrating a particular sign was extracted into two-by-two tables diagnosing ES or PNES (true positive, true negative, false positive, and false negative). When multiple publications based on the same cohort of patients were encountered, the study with the most relevant information and the highest number of subjects was selected. When required, authors were contacted by electronic mail for any additional data.

A quality assessment of publications was conducted by one of the authors (C.D.). The revised Quality Assessment of Diagnostic Accuracy Studies was used to complete the risk of bias assessment.17 This tool assesses the risk of bias and applicability for studies of diagnostic accuracy against the following domains: patient selection, index test, reference test, and flow and timing of the test. Each domain is given one of the following rating: a low risk, an unclear risk, or a high risk.

Assignment of Level of Evidence

The American Academy of Neurology criteria for rating diagnostic studies was used by the authors to jointly assign a class of evidence rating for each publication.18

Statistical Analysis

A bivariate analysis was conducted using the R statistical package (R Foundation for Statistical Computing, Vienna, Austria).19,20 Sensitivity, specificity, PLR, and negative likelihood ratio (NLR) with corresponding 95% confidence intervals (CIs) were determined for each sign in diagnosing ES or PNES. Data were pooled in a random effects (DerSimonian and Laird) method for a meta-analysis.

Estimates of PLR and NLR were calculated. The PLR (ratio of sensitivity to 1−specificity) indicates the likelihood of PNES compared with absence of PNES when a particular sign (e.g., eye closure) is present in patients. The NLR (ratio of 1-sensitivity to specificity) indicates the likelihood of PNES compared with absence of PNES when a sign is absent in patients. A PLR >10 or an NLR <0.1 largely affects the posttest probability of a diagnosis (ES or PNES) whereas a PLR between 5 and 10 or an NLR between 0.1 and 0.2 moderately affects the posttest probability of a diagnosis (ES or PNES).21 The results were visualized on forest plots of likelihood ratios.

Statistical heterogeneity among different studies was calculated using the I2 statistic.22 An I2 value of 0%–25% is considered to be low, whereas an I2 value of 75%–100% indicates high heterogeneity.22 In the presence of high heterogeneity, a meta-regression was performed (assessing the sample size effects, study type, country of publication, and year of publication). Funnel plots were used to test for publication bias and identify outliers.23 Sensitivity analyses were performed using the leave-one-out analysis.

Standard Protocol Approvals, Registrations, and Patient Consents

Because this article did not directly involve human subjects, while using only data from published articles, an institutional review board approval was not required.

Data Availability

Data pertaining to this study will be shared with qualified investigators on request.

Results

Study Characteristics and Patient Demographics

The PRISMA flowchart summarizing the study selection process is shown in Figure 1. Thirty-six publications met the inclusion criteria. Ten studies did not provide data that could be extracted into 2 × 2 tables. Twelve studies either reported particular sign(s) that were not reported in any other study or reported data as proportion of event groups. The remaining 14 studies were included in the meta-analysis.24-38 These were all retrospective studies. There were 800 ES patients and 452 PNES patients. The proportion of female patients was 38% in the ES group and 74% in the PNES group. The mean patient age in the ES groups ranged between 10.7 and 35.2 years. The mean patient age in the PNES groups ranged between 8.1 and 40.1. Table 1 summarizes the characteristics of the studies included in the meta-analysis.

Figure 1. PRISMA Flowchart Summarizing the Study Selection Process.

Figure 1

Open in a new tab

PRISMA = Preferred Reporting Items for Systematic Reviews and Meta-analyses.

Table 1.

Summary of Studies Included in This Meta-analysis

graphic file with name NEURCLINPRACT2021070273TT1.jpg

Open in a new tab

Quality Assessment

The risk of bias for patient selection was high or unclear in 4 studies (29%). The risk of bias for the index test and reference test was unclear in 4 (29%) and 2 (14%) studies, respectively. The risk of bias in flow and timing was low in all included studies. eTable 1 (links.lww.com/CPJ/A334) summarizes the risk of bias and concerns regarding applicability for all studies included in the meta-analysis.

A Meta-analysis of Seizure Semiology

Signs that could be meta-analyzed included the following: eye closure, eyelid blinking or fluttering, pelvic thrusting, side-to-side head shaking, ictal weeping, asynchronous limb movements, body rocking, vocalization, automatisms, ictal version, hyperactive movements of the lower limbs, the presence of a postictal phase, and postictal confusion. Signs that could not be meta-analyzed included the following: ictal hyperventilation, opisthotonus, brief or prolonged seizure duration (<2 or >5 minutes), thrashing or writhing, ictal stuttering, hand postures (claw hand, fist hand, flaccid hand, index finger pointing, and fanning posture), fluctuating course, postictal whispering, postictal blinking or head shaking, postictal looking around, postictal asking “what happened?”, ictal eye opening, eye deviation, ictal cry, ictal rhythmic grunting, discernible onset, discernible offset, stereotypy, postictal loud breathing, postictal nose rubbing, postictal agitation, and postictal amnesia. The signs that could not be meta-analyzed were due to one of the following reasons: (1) only 1 study published on a particular sign (e.g., index finger pointing in ES and PNES) and (2) only aggregate data at the event-group level rather than patient-group level available for analysis.

Although all studies provided aggregate data, some reported the occurrence of 1 or more signs as a proportion of event groups, whereas others provided these data as a proportion of patient groups. For example, some studies reported the proportion of patients in the ES and PNES groups in whom ictal eye closure was observed, whereas other studies reported the proportion of ES events or PNES events in which eye closure was observed. These could not be pooled together for the statistical analysis. Proportions of patients per group were preferentially used to minimize the risk of underestimating or overestimating the true effect size.

For PNES, signs that reached a PLR threshold of >5 were ictal eye closure (PLR 40.5 95% CI 16.2–101.3; I2 = 0, from 3 studies), asynchronous limb movements (PLR 10.2; 95% CI 2.8–37.7; I2 = 0, from 3 studies), and ictal weeping (PLR 8.2; 95% CI 1.9–35.5; I2 = 0, from 4 studies). For PNES, the absence of only 1 sign showed a trend toward an NLR threshold of <0.2, ictal eye closure (NLR 0.2; 95% CI 0.1–0.5; I2 = 0, based on 3 studies). These findings are summarized in Table 2. Figure 2 shows forest plots of positive likelihood ratios for clinical signs in the diagnosis of PNES. Figure 3 shows forest plots of positive likelihood ratios for clinical signs in the diagnosis of ES. For ES, the only sign that reached the PLR threshold of >5 was ictal head version (PLR 5.0; 95% CI 0.8–30.8; I2 = 0, from 2 studies). The absence of no single sign reached an NLR threshold of <0.2 for ES (Table 3).

Table 2.

Summary of Diagnostic Measures of Selected Semiologic Features for Predicting Psychogenic Nonepileptic Seizures

graphic file with name NEURCLINPRACT2021070273TT2.jpg

Open in a new tab

Figure 2. Forest Plots of Positive Likelihood Ratios for (A) Ictal Eye Closure, (B) Ictal Weeping, (C) Asynchronous Limb Movements, and (D) Pelvic Thrusting Predicting Psychogenic Nonepileptic Seizures.

Figure 2

Open in a new tab

FN = number of false negatives; FP = number of false positives; PLR = positive likelihood ratio; TN = number of true negatives; TP = number of true positives.

Figure 3. Forest Plot of Positive Likelihood Ratios for (A) Automatisms and (B) Ictal Head Version Predicting Epileptic Seizures.

Figure 3

Open in a new tab

FN = number of false negatives; FP = number of false positives; PLR = positive likelihood ratio; TN = number of true negatives; TP = number of true positives.

Table 3.

Summary of Diagnostic Measures of Selected Semiologic Features for Predicting Epileptic Seizures

graphic file with name NEURCLINPRACT2021070273TT3.jpg

Open in a new tab

The leave-one-out analysis was performed for asynchronous limb movements, ictal weeping, and ictal head version, given the presence of outliers. In the case of asynchronous limb movements, there were 2 outliers. After excluding one of these outliers25 from the statistical analysis, the pooled PLR was still greater than 5 (PLR: 8.9; 95% CI 1.5–52.0; I2 = 0). Excluding the other outlier33 from the statistical analysis produced a similar result (PLR 7.7; 95% CI 1.7–34.8; I2 = 0%). After the exclusion of the outlier studies from the statistical analysis for ictal weeping38 (PLR 4.9; 95% CI 0.9–27.4; I2 = 0%) and ictal head version30 (PLR 2.6; 95% CI 0.5–14.0), the PLR threshold of >5 was no longer reached (although ictal weeping showed a trend toward a PLR >5), reflecting the disproportionate influence of these studies on the pooled effect size.

Statistical heterogeneity was low (Tables 2 and 3), and a meta-regression was therefore not performed. Funnel plots demonstrating the effect sizes of studies against measures of precision are shown in eFigure 1 (links.lww.com/CPJ/A334). Although these studies lay within the funnel, given the low number of studies for each sign, the interpretation was somewhat limited.

No other sign reached a PLR threshold >5 for predicting either ES or PNES. eTables 2 and 3 (links.lww.com/CPJ/A334) summarize the measures of diagnostic accuracy for signs that could not be meta-analyzed (individual studies or studies reporting proportions at the event-group level).

Level of Evidence

Of the 14 studies included in the meta-analysis, 4 studies26,34,37,38 were given a Class II rating and 10 studies24,25,27-33,36 a Class III rating.

PNES and Intellectual Disability

The targeted literature search and screening of study titles, abstracts, and the references lists of relevant publications identified 17 publications. After a full-text review, we excluded 3 review articles15,16,39 and 8 original publications,40-47 which did not provide descriptive data on clinical phenomenology of nonepileptic seizures or described nonseizure behavioral episodes.

Of the remaining 6 studies, 2 evaluated the role of clinical signs in differentiating PNES from ES. The largest study48 published to date included 124 institutionalized patients with ID referred for VEM to evaluate behaviors suggestive of epilepsy. Patients in the PNES-only group (n = 50, 40.3%) demonstrated several clinical signs or behaviors that were not observed in patients in the ES-only group (n = 20, 16.1%): most commonly, hyperventilation (14%) and rocking (12%). The frequency of clinical signs such as eye closure (ES 5% vs PNES 4%), repetitive eye blinks (ES 8% vs PNES 12%), automatisms (ES 15% vs PNES 16%), and ictal vocalization (ES 5% vs PNES 12%) were not statistically significantly (p > 0.05) different between the ES and PNES groups. A residential VEM study49 composed 24 patients with epilepsy, 12 with PNES, and 14 with a mixed diagnosis of ES/PNES described the utility of peri-ictal responsiveness to social cues in differentiating PNES from ES. For example, PNES was more likely to be preceded by attempts to alert others (OR 12.4; 95% CI 3.2–47.7), show altered intensity in the presence of others (OR 199.4; 95% CI 12.0–3309.9), and display postictal behaviors affected by the presence of others (OR 91.1; 95% CI 17.2–482.1) (irrespective of whether the patient was diagnosed with comorbid ID).

Four studies evaluated the clinical characteristics of PNES in those with and without a comorbid diagnosis of ID. One of these studies50 looked at patients with comorbid PNES and ES and reported tonic-like, tonic-clonic–like, and absence-like seizures as the most common phenotypes of PNES. In this cohort of patients, patients' PNES shared characteristics with patients' ES in 80% of cases. In another study51 of 288 patients with PNES, similar proportions of patients in the ID and non-ID groups experienced events characterized as convulsive (ID group 68%; non-ID group 60%) or swoon (ID group 20%; non-ID group 22%). However, attacks that were classified as blanks occurred only in the non-ID group. A third study52 reported on 7 patients with mild ID and epilepsy with comorbid PNES. One of these patients experienced PNES that was similar to their ES. Other reported semiology in this group included arc en circle, akinetic, myoclonic jerks, and generalized convulsions. PNES with movement sequences composed of rocking, shaking, shuddering, arm-waving, falling, and staring episodes were observed in patients with and without ID in the fourth study.53 These studies are summarized in eTable 4 (links.lww.com/CPJ/A334).

It was not possible to meta-analyze this subgroup data because of 2 major reasons: (1) heterogenous or vaguely described clinical signs (e.g., tonic-like, blanks, and swoon) evaluated across different studies; (2) clinical signs (e.g., eye closure in ES vs PNES) presented in a 2 x 2 table in only one of these studies.48

Discussion

There are several important findings from this meta-analysis of more than 800 ES patients and 452 PNES patients: (1) in the absence of VEM, a careful study of semiology can help clinicians differentiate PNES from ES; (2) the only signs with a moderate to large effect on the posttest probability of PNES are ictal eye closure and asynchronous limb movements; (3) the absence of ictal eye closure shows a trend toward being a useful sign for ruling out PNES; however, the specified threshold (NLR <0.2) was not reached; (4) no single sign had a moderate or large effect on the posttest probability for diagnosing ES; and (5) no single sign was observed exclusively in either PNES or ES.

Studies have consistently reported that ictal eye closure is more frequently observed during PNES than ES. However, authors in these studies looked at different aspects of ictal eye closure. Whereas some studies describe eye condition as open or closed with no further characterization, other studies were more specific in their definition. Specific parameters evaluated by authors include eye condition at seizure onset,54 eye condition at peak of symptoms,26 percentage of eye closure during the course of the seizure,55 forceful eye closure, and eye closure by other means including covering with hand.34 Typically, PNES is associated with forceful eye closure, resistance to eye opening, eye closure at peak of symptoms, covering of eyes during the seizure, and eye closure for a greater proportion of the duration of the seizure (1 study reported that eye closure >55% of seizure duration yielded the highest positive predictive value for PNES).26,34,35,56 The exception to this rule was patients with PNES with purely sensory symptoms or agitated behaviors, who demonstrated ictal eye closure in only 11.1% and 6.9% of seizures, respectively, in 1 study.35 On the contrary, ES is more frequently associated with eye opening or widening at seizure onset and during the seizure. Indeed, 1 study reported that eye opening or widening had sensitivity of 100% and specificity of 84% for predicting ES.54 A detailed evaluation of the eye condition at the beginning and course of the seizure (in patients without isolated sensory symptoms or agitated behaviors) can therefore be a very useful sign for differentiating PNES from ES.

However, patient or observer reports of this sign may be unreliable. One study noted that self-reports (sensitivity: 0.535, specificity: 0.507) or observer (sensitivity: 0.419, specificity: 0.478) reports of ictal eye closure were unreliable at predicting PNES when compared with VEM.55 Therefore, the clinician should not rely on patient or observer accounts alone. Studies have demonstrated how the availability of a seizure video can increase the diagnostic accuracy in addition to a standard history and physical examination alone.8 Given the increasing availability of smartphones with cameras, patients or their caregivers should be encouraged to record typical events, and seizure videos should form a routine part of the clinical assessment for these patients.8

Other eye signs that have been studied in the literature include repetitive eye blinking, eyelid fluttering, and eye deviation. The literature does not support eyelid fluttering25,33 or repetitive eye blinking26,29,36 as a reliable discriminator between PNES or ES (Table 2). Forceful, rapid bilateral eye blinking or eyelid fluttering maybe observed in seizures arising from the occipital lobe.57 Intracerebral recordings have also demonstrated that ictal blinking can be associated with occipitotemporal, mesial temporal, and insular-opercular epilepsies.58 Eyelid blinking was most commonly bilateral; however, it can occasionally be ipsilateral or contralateral to the side of the epileptogenic focus.58 Repetitive eyelid blinking or fluttering is also observed in PNES, and studies have not demonstrated a statistically significant difference in the frequency of these signs in patients with PNES or ES.26,33 On the contrary, sustained ictal head version26,30 (or eye deviation)26 were both features that were more frequently observed in ES compared with PNES. Based on the results of this meta-analysis, we would not recommend the use of repetitive (unilateral or bilateral) eye blinking or eyelid fluttering, in isolation, to differentiate ES from PNES. However, the finding of eye deviation26 (when ascertained by a trained clinician) should prompt the clinician to consider a diagnosis of ES.

The other useful sign for clinicians to potentially rule in a diagnosis of PNES is asynchronous limb movements. This sign was more frequently observed in PNES compared with that in ES across several studies, making it a reliable discriminator of PNES and ES.25,26,33,56 However, it is worth noting that this sign is not exclusive to PNES. Asynchronous limb movements may be observed during epileptic automatisms and hyperkinetic seizures.24,56 Indeed, 1 study noted that there was no statistically significant difference in the frequency of asynchronous limb movements in patients with generalized convulsive nonepileptic seizures compared with patients with frontal lobe hypermotor seizures.56 The sign seems to be most useful when differentiating convulsive ES from convulsive PNES.

Similar to asynchronous limb movements, ictal weeping was consistently observed more frequently in PNES compared with ES (despite not reaching the PLR threshold of >5).26,29,33,38 It can also rarely be observed in ES. In 1 study, the frequency of epileptic ictal crying was 0.1%.59 In the same study, 4 of 5 ES patients with ictal weeping were diagnosed with frontal lobe epilepsy.59 We would suggest that both signs (asynchronous limb movements and ictal weeping) should prompt clinicians to consider PNES as the more likely diagnosis.

This meta-analysis suggested that the presence of signs such as pelvic thrusting, side-to-side head shaking, or body rocking, while useful, should not be used in isolation to make a diagnosis of PNES and ES. These signs were observed in both PNES and ES (particularly, hyperkinetic seizures; Table 2). In the case of pelvic thrusting or side-to-side head shaking, several studies observed the lack of statistically significant difference in the frequency of these signs in PNES compared with frontal lobe hypermotor seizures.24,37,56 Of note, 1 study noted that the nature of pelvic thrusting may be a useful discriminator of PNES from ES, with only forward pelvic thrusting being seen in the PNES cohort, whereas only retropelvic thrusting was observed in the ES cohort.25 However, the sample size of this particular study was small, and this requires further evaluation. The presence of certain signs that can be observed in both PNES and ES (particularly frontal lobe or hyperkinetic seizures) highlighted 2 points—first, these signs need to be interpreted in the overall clinical context; and second, it is important to characterize the movements in detail because differences within a particular sign may potentially have a discriminative value.

Although not a major focus of this study, studies have consistently demonstrated that PNES have a higher mean duration compared with ES.10,24,27,56 Prolonged seizure duration (seizure duration longer than 5 minutes) should alert the clinician to this diagnostic possibility.60 Seizure duration is ascertained very precisely in these studies using objective clinical or electrophysiologic criteria by trained epileptologists. Timing seizures in the real world by nonexpert clinicians, let alone a layperson, is likely to be less precise. Indeed, 1 study demonstrated that doctors in training tend to underestimate the seizure duration.61 In some cases, patients or their caregivers provide an estimate of the seizure duration. Therefore, an accurate timing of seizure duration (using a stopwatch or other device) is critical.

Furthermore, patients with convulsive ES longer than 5 minutes are diagnosed with status epilepticus—a neurologic emergency with potentially high morbidity and mortality requiring emergent management with intravenous antiseizure medications and, in refractory cases, intravenous anesthetic therapy and intubation. In the absence of an expert clinical assessment or EEG, there is a potential for PNES lasting longer than 5 minutes to be mistaken for and treated as convulsive status epilepticus. This is a well-recognized phenomenon (termed nonepileptic psychogenic status) and may occur in up to 25% of patients with PNES presenting to the hospital with prolonged seizures.62 These patients may receive an inappropriate treatment including antiseizure medication, anesthetic agents, or even intubation, exposing them to the potential morbidity and mortality associated with these interventions. Psychogenic nonepileptic status is also associated with greater health-care utilization costs.5 In settings where an immediate neurologist assessment or EEG is not readily available, seizure videos and telehealth platforms may provide a practical alternative to ensure that patients are correctly diagnosed and appropriately managed.8

Given that no single sign is exclusive to either PNES or ES and there is insufficient evidence to support many signs, in isolation, as a reliable discriminator of PNES from ES (Tables 2 and 3, eTables 2 and 3, links.lww.com/CPJ/A334), the clinician should not rely on semiology alone. A detailed neurologic assessment (a history taking and a physical examination) with an accurate determination of seizure semiology (using patient and eyewitness accounts and high-quality seizure videos where available) is critical in the assessment of these patients.8,63 Given that this meta-analysis is a synthesis of published aggregate data, it was not possible to evaluate how different combinations of clinical signs (e.g., eye closure and asynchronous limb jerking) may assist clinicians to differentiate PNES from ES.

The ILAE PNES taskforce acknowledges PNES in individuals with ID as an important and distinct subgroup.15 This group may represent an etiologically distinct group with a greater emphasis on the environment or social interactions in its explanation and treatment.49 The evidence for female preponderance is less clear, and there is a trend toward longer diagnostic delay (10.7 vs 6.9 years).51 Rates of nonepileptic psychogenic status were also reported to be more common in this subgroup.16 In addition, there were lower reported rates of antecedent sexual, physical, or emotional trauma than those reported in the general population. Clinical manifestations were heterogenous, and no single sign was believed to be pathognomic. In patients with a mixed diagnosis of ES, PNES, and ID, PNES may share characteristics with patients' ES.50 The distinction between behavioral episodes, PNES, and ES remains an ongoing diagnostic challenge. The utility of clinical signs such as eye closure in reliably differentiating PNES from ES was not established based on our study.48 Video EEG remains an important tool that can help clinicians distinguish between these diagnoses.

This study has several limitations. The clinical signs were not specifically defined in all publications (e.g., ictal version was defined in 1 study30 but not another26), and most did not use blinded assessors to ascertain these clinical signs. We tried to mitigate the effect of this by only looking at studies that report on patients assessed in epilepsy monitoring units by epileptologists (because they are best trained to accurately ascertain and describe these signs). While this ensures that the diagnoses and clinical signs are accurately ascertained, it is unclear how these findings translate to the real-world experience. Given the paucity of publications on the use of specific clinical signs to differentiate PNES from ES in patients with ID, we were unable to extrapolate the results of this meta-analysis to this patient population. Furthermore, we were unable to perform subgroup analyses of differences within a particular clinical sign (e.g., forceful eye closure and covering eyes) because there were not enough studies published on each subtype to do so.

The accuracy of nonexperts in ascertaining and describing these signs (in the absence of video EEG) is not established in the literature. At the very least, good-quality seizure videos may represent more reliable clinical data compared with patient or observer accounts alone,55 but this will need to be evaluated further. Following on from the abovementioned point, the findings of this study may reflect (at least partly) tertiary center bias. Patients admitted for VEM may represent a more complex cohort of patients (for example, patients with drug refractory epilepsy, and unclear diagnosis). They may undergo medication withdrawal, sleep deprivation, or other provocation techniques to bring on seizures that individuals in the community would not be exposed to. As such, the seizures experienced by these patients may not necessarily be representative of the typical case in the community. Third, only seizures with clinically observable signs were included in most of these studies. This meta-analysis does not provide supportive evidence for clinicians trying to differentiate dialeptic PNES or PNES aura from ES.64,65 Fourth, the diagnosis of ES or PNES is not based on a single sign alone, and there is a clearly defined approach to making this diagnosis.63 Future studies should assess how combinations of clinical signs can help clinicians differentiate PNES from ES. Fifth, despite a large number of signs that have been reported in the literature, most could not be meta-analyzed because of a lack of studies. For example, only 1 study (with a small sample size of 20 ES and 20 PNES patients) evaluated the discriminative ability of stereotypy and sequence of signs in differentiating PNES from ES.66 Future studies should evaluate the discriminative ability of sequence or a combination of signs further. Sixth, many of the studies included in this meta-analysis have small sample sizes, which was reflected in the wide CIs that were observed in our calculated measures of diagnostic accuracy.

In the absence of VEM, a careful study of semiology (where possible using seizure videos) with an understanding of the diagnostic accuracy and limitations of different signs is an important aspect of clinical assessment of patients presenting with seizures. The diagnostic accuracy of different signs varies, and no single sign is exclusively observed in PNES or ES. While all signs require interpretation in the overall clinical context and may help clinicians in the diagnostic process, the most reliable signs for diagnosing PNES are ictal eye closure and asynchronous limb movements based on this meta-analysis. In the case of ES, we did not find similarly reliable discriminative signs. In addition to the presence or absence of particular signs (such as ictal eye closure), differences within a particular sign (e.g., forceful eye closure, covering eyes) can also provide valuable information for clinicians trying to differentiate PNES from ES. Given that these signs are rarely interpreted in isolation, future studies should evaluate how the sequence or combination of signs may influence the clinician's diagnostic process.

TAKE-HOME POINTS

  • → A careful study of semiology can help clinicians differentiate PNES from ES.

  • → Based on our meta-analysis, the only signs with a moderate to large impact on the posttest probability of PNES are ictal eye closure and asynchronous limb movements.

  • → No single sign was observed exclusively in either PNES or ES.

  • → These signs should be interpreted in the overall clinical context rather than in isolation.

  • → Future studies should look at how combinations of signs may assist clinicians in differentiating PNES from ES.

Appendix. Authors

Appendix.

Open in a new tab

Study Funding

No targeted funding reported.

Disclosure

The authors report no disclosures relevant to the manuscript. Full disclosure form information provided by the authors is available with the full text of this article at Neurology.org/cp.

References

  • 1.Perez DL, LaFrance WC Jr. Nonepileptic seizures: an updated review. CNS Spectr. 2016;21(3):239-246. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Asadi-Pooya AA, Sperling MR. Epidemiology of psychogenic nonepileptic seizures. Epilepsy Behav. 2015;46:60-65. [DOI] [PubMed] [Google Scholar]
  • 3.Salinsky M, Spencer D, Boudreau E, Ferguson F. Psychogenic nonepileptic seizures in US veterans. Neurology. 2011;77(10):945-950. [DOI] [PubMed] [Google Scholar]
  • 4.Kerr WT, Janio EA, Le JM, et al. Diagnostic delay in psychogenic seizures and the association with anti-seizure medication trials. Seizure. 2016;40:123-126. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Seneviratne U, Low ZM, Low ZX, et al. Medical health care utilization cost of patients presenting with psychogenic nonepileptic seizures. Epilepsia. 2019;60(2):349-357. [DOI] [PubMed] [Google Scholar]
  • 6.Nightscales R, McCartney L, Auvrez C, et al. Mortality in patients with psychogenic nonepileptic seizures. Neurology. 2020;95(6):e643. [DOI] [PubMed] [Google Scholar]
  • 7.Duncan R. Psychogenic nonepileptic seizures: diagnosis and initial management. Expert Rev Neurother. 2010;10(12):1803-1809. [DOI] [PubMed] [Google Scholar]
  • 8.Tatum WO, Hirsch LJ, Gelfand MA, et al. Assessment of the predictive value of outpatient smartphone videos for diagnosis of epileptic seizures. JAMA Neurol. 2020;77(5):593-600. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Brigo F, Ausserer H, Nardone R, Tezzon F, Manganotti P, Bongiovanni LG. Clinical utility of ictal eyes closure in the differential diagnosis between epileptic seizures and psychogenic events. Epilepsy Res. 2013;104(1-2):1-10. [DOI] [PubMed] [Google Scholar]
  • 10.Seneviratne U, Minato E, Paul E. How reliable is ictal duration to differentiate psychogenic nonepileptic seizures from epileptic seizures? Epilepsy Behav. 2017;66:127-131. [DOI] [PubMed] [Google Scholar]
  • 11.Avbersek A, Sisodiya S. Does the primary literature provide support for clinical signs used to distinguish psychogenic nonepileptic seizures from epileptic seizures? J Neurol Neurosurg Psychiatry. 2010;81(7):719-725. [DOI] [PubMed] [Google Scholar]
  • 12.Devinsky O, Gazzola D, LaFrance WC Jr. Differentiating between nonepileptic and epileptic seizures. Nat Rev Neurol. 2011;7(4):210-220. [DOI] [PubMed] [Google Scholar]
  • 13.Gilmour GS, MacIsaac R, Subotic A, Wiebe S, Josephson CB. Diagnostic accuracy of clinical signs and symptoms for psychogenic nonepileptic attacks versus epileptic seizures: a systematic review and meta-analysis. Epilepsy Behav. 2021;121:108030. [DOI] [PubMed] [Google Scholar]
  • 14.Moher D, Liberati A, Tetzlaff J, Altman DG. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. PLoS Med. 2009;6(7):e1000097. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Kanemoto K, LaFrance WC Jr, Duncan R, et al. PNES around the world: where we are now and how we can close the diagnosis and treatment gaps-an ILAE PNES Task Force report. Epilepsia Open. 2017;2(3):307-316. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Rawlings GH, Novakova B, Beail N, Reuber M. What do we know about non-epileptic seizures in adults with intellectual disability: a narrative review. Seizure. 2021;91:437-446. [DOI] [PubMed] [Google Scholar]
  • 17.Whiting PF, Rutjes AW, Westwood ME, et al. QUADAS-2: a revised tool for the quality assessment of diagnostic accuracy studies. Ann Intern Med. 2011;155(8):529-536. [DOI] [PubMed] [Google Scholar]
  • 18.Gronseth GS, Cox J, Gloss D, et al. ; on behalf of the Guideline Development, Dissemination, and Implementation Subcommittee of the American Academy of Neurology. Clinical Practice Guideline Process Manual, 2017 ed. American Academy of Neurology, 2017. [Google Scholar]
  • 19.Reitsma JB, Glas AS, Rutjes AW, Scholten RJ, Bossuyt PM, Zwinderman AH. Bivariate analysis of sensitivity and specificity produces informative summary measures in diagnostic reviews. J Clin Epidemiol. 2005;58(10):982-990. [DOI] [PubMed] [Google Scholar]
  • 20.Meta-Analysis of Diagnostic Accuracy [computer program]. 2020. [Google Scholar]
  • 21.Jaeschke R, Guyatt GH, Sackett DL. Users' guides to the medical literature. III. How to use an article about a diagnostic test. B. What are the results and will they help me in caring for my patients? The Evidence-Based Medicine Working Group. Jama. 1994;271(9):703-707. [DOI] [PubMed] [Google Scholar]
  • 22.Higgins JPT, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analyses. BMJ. 2003;327(7414):557-560. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Sterne JAC, Sutton AJ, Ioannidis JPA, et al. Recommendations for examining and interpreting funnel plot asymmetry in meta-analyses of randomised controlled trials. BMJ. 2011;343:d4002. [DOI] [PubMed] [Google Scholar]
  • 24.Saygi S, Katz A, Marks DA, Spencer SS. Frontal lobe partial seizures and psychogenic seizures: comparison of clinical and ictal characteristics. Neurology. 1992;42(7):1274-1277. [DOI] [PubMed] [Google Scholar]
  • 25.Gates JR, Ramani V, Whalen S, Loewenson R. Ictal characteristics of pseudoseizures. Arch Neurol. 1985;42(12):1183-1187. [DOI] [PubMed] [Google Scholar]
  • 26.Chen DK, Graber KD, Anderson CT, Fisher RS. Sensitivity and specificity of video alone versus electroencephalography alone for the diagnosis of partial seizures. Epilepsy Behav. 2008;13(1):115-118. [DOI] [PubMed] [Google Scholar]
  • 27.Henry TR, Drury I. Ictal behaviors during nonepileptic seizures differ in patients with temporal lobe interictal epileptiform EEG activity and patients without interictal epileptiform EEG abnormalities. Epilepsia. 1998;39(2):175-182. [DOI] [PubMed] [Google Scholar]
  • 28.Carmant L, Kramer U, Holmes GL, Mikati MA, Riviello JJ, Helmers SL. Differential diagnosis of staring spells in children: a video-EEG study. Pediatr Neurol. 1996;14(3):199-202. [DOI] [PubMed] [Google Scholar]
  • 29.Devinsky O, Sanchez-Villaseñor F, Vazquez B, Kothari M, Alper K, Luciano D. Clinical profile of patients with epileptic and nonepileptic seizures. Neurology. 1996;46(6):1530-1533. [DOI] [PubMed] [Google Scholar]
  • 30.Gröppel G, Pataraia E, Olbrich A, et al. Clinical symptoms in psychogenic seizures. Wien Klin Wochenschr. 1999;111(12):469-475. [PubMed] [Google Scholar]
  • 31.Holmes GL, Sackellares JC, McKiernan J, Ragland M, Dreifuss FE. Evaluation of childhood pseudoseizures using EEG telemetry and video tape monitoring. J Pediatr. 1980;97(4):554-558. [DOI] [PubMed] [Google Scholar]
  • 32.Slater JD, Brown MC, Jacobs W, Ramsay RE. Induction of pseudoseizures with intravenous saline placebo. Epilepsia. 1995;36(6):580-585. [DOI] [PubMed] [Google Scholar]
  • 33.De Paola L, Terra VC, Silvado CE, et al. Improving first responders' psychogenic nonepileptic seizures diagnosis accuracy: development and validation of a 6-item bedside diagnostic tool. Epilepsy Behav. 2016;54:40-46. [DOI] [PubMed] [Google Scholar]
  • 34.Chung SS, Gerber P, Kirlin KA. Ictal eye closure is a reliable indicator for psychogenic nonepileptic seizures. Neurology. 2006;66(11):1730-1731. [DOI] [PubMed] [Google Scholar]
  • 35.DeToledo JC, Ramsay RE. Patterns of involvement of facial muscles during epileptic and nonepileptic events: review of 654 events. Neurology. 1996;47(3):621-625. [DOI] [PubMed] [Google Scholar]
  • 36.Pierelli F, Chatrian GE, Erdly WW, Swanson PD. Long-term EEG-video-audio monitoring: detection of partial epileptic seizures and psychogenic episodes by 24-hour EEG record review. Epilepsia. 1989;30(5):513-523. [DOI] [PubMed] [Google Scholar]
  • 37.Geyer JD, Payne TA, Drury I. The value of pelvic thrusting in the diagnosis of seizures and pseudoseizures. Neurology. 2000;54(1):227-229. [DOI] [PubMed] [Google Scholar]
  • 38.Walczak TS, Bogolioubov A. Weeping during psychogenic nonepileptic seizures. Epilepsia. 1996;37(2):208-210. [DOI] [PubMed] [Google Scholar]
  • 39.Chapman M, Iddon P, Atkinson K, et al. The misdiagnosis of epilepsy in people with intellectual disabilities: a systematic review. Seizure. 2011;20(2):101-106. [DOI] [PubMed] [Google Scholar]
  • 40.Desai P, Talwar D. Nonepileptic events in normal and neurologically handicapped children: a video-EEG study. Pediatr Neurol. 1992;8(2):127-129. [DOI] [PubMed] [Google Scholar]
  • 41.Donat JF, Wright FS. Episodic symptoms mistaken for seizures in the neurologically impaired child. Neurology. 1990;40(1):156-157. [DOI] [PubMed] [Google Scholar]
  • 42.Kotagal P, Costa M, Wyllie E, Wolgamuth B. Paroxysmal nonepileptic events in children and adolescents. Pediatrics. 2002;110(4):e46. [DOI] [PubMed] [Google Scholar]
  • 43.Uldall P, Alving J, Hansen LK, Kibaek M, Buchholt J. The misdiagnosis of epilepsy in children admitted to a tertiary epilepsy centre with paroxysmal events. Arch Dis Child. 2006;91(3):219-221. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.Bye AM, Kok DJ, Ferenschild FT, Vles JS. Paroxysmal non-epileptic events in children: a retrospective study over a period of 10 years. J Paediatr Child Health. 2000;36(3):244-248. [DOI] [PubMed] [Google Scholar]
  • 45.Kanemoto K, Goji H, Tadokoro Y, Kato E, Oshima T. Psychogenic non-epileptic seizure in patients with intellectual disability with special focus on choice of therapeutic intervention. Seizure. 2017;45:2-6. [DOI] [PubMed] [Google Scholar]
  • 46.Ueda R, Shimizu-Motohashi Y, Sugai K, et al. Seizure imitators monitored using video-EEG in children with intellectual disabilities. Epilepsy Behav. 2018;84:122-126. [DOI] [PubMed] [Google Scholar]
  • 47.Glaze DG, Schultz RJ, Frost JD. Rett syndrome: characterization of seizures versus non-seizures. Electroencephalogr Clin Neurophysiol. 1998;106(1):79-83. [DOI] [PubMed] [Google Scholar]
  • 48.Neill JC, Alvarez N. Differential diagnosis of epileptic versus pseudoepileptic seizures in developmentally disabled persons. Appl Res Ment Retard. 1986;7(3):285-298. [DOI] [PubMed] [Google Scholar]
  • 49.Wardrope A, Wong S, McLaughlan J, Wolfe M, Oto M, Reuber M. Peri-ictal responsiveness to the social environment is greater in psychogenic nonepileptic than epileptic seizures. Epilepsia. 2020;61(4):758-765. [DOI] [PubMed] [Google Scholar]
  • 50.van Ool JS, Haenen AI, Snoeijen-Schouwenaars FM, et al. Psychogenic nonepileptic seizures in adults with epilepsy and intellectual disability: a neglected area. Seizure. 2018;59:67-71. [DOI] [PubMed] [Google Scholar]
  • 51.Duncan R, Oto M. Psychogenic nonepileptic seizures in patients with learning disability: comparison with patients with no learning disability. Epilepsy Behav. 2008;12(1):183-186. [DOI] [PubMed] [Google Scholar]
  • 52.Magaudda A, Gugliotta SC, Tallarico R, Buccheri T, Alfa R, Laganà A. Identification of three distinct groups of patients with both epilepsy and psychogenic nonepileptic seizures. Epilepsy Behav. 2011;22(2):318-323. [DOI] [PubMed] [Google Scholar]
  • 53.Duchowny MS, Resnick TJ, Deray MJ, Alvarez LA. Video EEG diagnosis of repetitive behavior in early childhood and its relationship to seizures. Pediatr Neurol. 1988;4(3):162-164. [DOI] [PubMed] [Google Scholar]
  • 54.Syed TU, LaFrance WC Jr, Kahriman ES, et al. Can semiology predict psychogenic nonepileptic seizures? A prospective study. Ann Neurol. 2011;69(6):997-1004. [DOI] [PubMed] [Google Scholar]
  • 55.Syed TU, Arozullah AM, Suciu GP, et al. Do observer and self-reports of ictal eye closure predict psychogenic nonepileptic seizures? Epilepsia. 2008;49(5):898-904. [DOI] [PubMed] [Google Scholar]
  • 56.Azar NJ, Tayah TF, Wang L, Song Y, Abou-Khalil BW. Postictal breathing pattern distinguishes epileptic from nonepileptic convulsive seizures. Epilepsia. 2008;49(1):132-137. [DOI] [PubMed] [Google Scholar]
  • 57.Williamson PD, Thadani VM, Darcey TM, Spencer DD, Spencer SS, Mattson RH. Occipital lobe epilepsy: clinical characteristics, seizure spread patterns, and results of surgery. Ann Neurol. 1992;31(1):3-13. [DOI] [PubMed] [Google Scholar]
  • 58.Lagarde SD, Dirani MD, Trebuchon AP, et al. Ictal blinking in focal seizures: insights from SEEG recordings. Seizure. 2020;81:21-28. [DOI] [PubMed] [Google Scholar]
  • 59.Asadi-Pooya AA, Wyeth D, Sperling MR. Ictal crying. Epilepsy Behav. 2016;59:1-3. [DOI] [PubMed] [Google Scholar]
  • 60.Anis S, Fahoum F, Korczyn AD, et al. Atypical duration of epileptic and psychogenic nonepileptic events. Epilepsy Behav. 2020;111:107145. [DOI] [PubMed] [Google Scholar]
  • 61.Seneviratne U, Ma H, Phan TG. How do doctors in training react to seizures? Epilepsy Behav. 2016;54:104-109. [DOI] [PubMed] [Google Scholar]
  • 62.Reuber M, Pukrop R, Mitchell AJ, Bauer J, Elger CE. Clinical significance of recurrent psychogenic nonepileptic seizure status. J Neurol. 2003;250(11):1355-1362. [DOI] [PubMed] [Google Scholar]
  • 63.LaFrance WC Jr, Baker GA, Duncan R, Goldstein LH, Reuber M. Minimum requirements for the diagnosis of psychogenic nonepileptic seizures: a staged approach. Epilepsia. 2013;54(11):2005-2018. [DOI] [PubMed] [Google Scholar]
  • 64.Fisher RS, Cross JH, French JA, et al. Operational classification of seizure types by the International League Against Epilepsy: Position Paper of the ILAE Commission for Classification and Terminology. Epilepsia. 2017;58(4):522-530. [DOI] [PubMed] [Google Scholar]
  • 65.Seneviratne U, Reutens D, D'Souza W. Stereotypy of psychogenic nonepileptic seizures: insights from video-EEG monitoring. Epilepsia. 2010;51(7):1159-1168. [DOI] [PubMed] [Google Scholar]
  • 66.Vogrig A, Hsiang JC, Ng J, Rolnick J, Cheng J, Parvizi J. A systematic study of stereotypy in epileptic seizures versus psychogenic seizure-like events. Epilepsy Behav. 2019;90:172-177. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Data Availability Statement

Data pertaining to this study will be shared with qualified investigators on request.

Articles from Neurology: Clinical Practice are provided here courtesy of American Academy of Neurology

RESOURCES