- Email: [email protected]
How reliable is ictal duration to differentiate psychogenic nonepileptic seizures from epileptic seizures?
Epilepsy & Behavior 66 (2017) 127–131
Contents lists available at ScienceDirect
Epilepsy & Behavior journal homepage: www.elsevier.com/locate/yebeh
How reliable is ictal duration to differentiate psychogenic nonepileptic seizures from epileptic seizures? Udaya Seneviratne a,b,⁎, Erica Minato a, Eldho Paul c,d a
Department of Neuroscience, Monash Medical Centre, Melbourne, Australia Department of Medicine, School of Clinical Sciences at Monash Health, Monash University, Melbourne, Australia Monash Centre for Health Research and Implementation, School of Public Health and Preventive Medicine, Monash University, Melbourne, Australia d Department of Clinical Haematology, Alfred Hospital, Melbourne, Australia b c
a r t i c l e
i n f o
Article history: Received 12 August 2016 Revised 19 October 2016 Accepted 19 October 2016 Available online xxxx Keywords: Epilepsy monitoring Seizure semiology Psychogenic nonepileptic seizure Receiver operating characteristics curve Status epilepticus Diagnostic pitfalls
a b s t r a c t We sought to investigate (1) differences in ictal duration between psychogenic nonepileptic seizures (PNES) and epileptic seizures (ES), (2) the odds of being PNES when seizures last ≥5 min, and (3) the value of ictal duration as a diagnostic test to differentiate PNES from ES. We retrospectively reviewed video-EEG recordings and tabulated ictal durations of all PNES and ES. We estimated the mean ictal durations of PNES and ES using linear mixed models. The odds of being PNES when seizures last ≥5 min were estimated using logistic regression. We used receiver operating characteristics (ROC) curves to study the overall diagnostic accuracy of ictal duration in differentiating PNES from ES. We studied 441 ES and 341 PNES recorded from 138 patients. The mean ictal duration of PNES (148.7 s, 95% CI: 115.2–191.8) was significantly longer (p b 0.001) than that of ES (47.7 s, 95% CI: 37.6–60.6). The odds of being PNES was about 24 times higher (Odds ratio: 23.8, 95% CI: 7.9–71.3) when the ictal duration was ≥5 min. The ROC curve yielded an area under the curve of 0.80 (95% CI 0.73–0.88). Youden's index identified 123.5 s as the optimal threshold to diagnose PNES with 65% sensitivity and 93% specificity. Our results indicate that ictal duration is a useful test to raise suspicion of PNES. When a seizure lasts ≥5 min, it is 24 times more likely to be PNES with the potential risk of misdiagnosis as status epilepticus. © 2016 Elsevier Inc. All rights reserved.
1. Introduction Differentiating psychogenic nonepileptic seizures (PNES) from epileptic seizures (ES) is a major diagnostic challenge faced by clinicians and the rate of misdiagnosis is as high as 20–30% [1,2]. The misdiagnosis of PNES as ES often leads to unnecessary interventions and treatment with antiepileptic drugs resulting in adverse outcomes including death [3,4]. Prolonged PNES mimicking status epilepticus (“pseudostatus epilepticus”) is in particular a challenging condition to diagnose [5]. The gold standard diagnostic test for seizures is videoelectroencephalographic (VEEG) monitoring. However, it is an expensive investigation with limited availability. Therefore, researchers have evaluated the use of semiology without EEG in differentiating PNES from ES. The results indicate variable diagnostic accuracy depending on the experience of the observer [6,7].
⁎ Corresponding author at: Department of Neuroscience, Monash Medical Centre, Clayton, Melbourne, VIC 3168, Australia. E-mail addresses: [email protected] (U. Seneviratne), [email protected] (E. Minato), [email protected] (E. Paul).
http://dx.doi.org/10.1016/j.yebeh.2016.10.024 1525-5050/© 2016 Elsevier Inc. All rights reserved.
One major drawback in using most semiological features is the lack of quantifiable objective measurements. Being an objective measure, ictal (seizure) duration is a unique semiological sign. Several studies have found that ictal duration is significantly longer in PNES compared with ES (Table 1) [8–14]. However, only one study presents diagnostic accuracy at a single threshold [15], and none of the studies, as summarized in Table 1, provides information on the overall diagnostic accuracy of the ictal duration as a test. Seizures lasting ≥5 min are diagnosed as status epilepticus and treated aggressively. Hence, there is a potential risk for PNES to be misdiagnosed as status epilepticus because PNES tend to last longer than ES. Against this backdrop, we sought to investigate three research questions in relation to ictal duration; (1) How does ictal duration of PNES differ from that of ES? (2) When seizures last ≥5 min, what are the odds of them being PNES, which can potentially be misdiagnosed as status epilepticus? (3) How reliable is ictal duration as a diagnostic test to differentiate PNES from ES? We hypothesized that ictal duration is a reliable test to differentiate PNES from ES, and when a seizure lasts ≥5 min the odds are higher for being PNES with the potential risk of misdiagnosis as status epilepticus.
128
U. Seneviratne et al. / Epilepsy & Behavior 66 (2017) 127–131
Table 1 Summary of previous studies on the diagnostic value of ictal duration to differentiate ES from PNES. Reference
Mean age (SD)
N subjects (N seizures)
ES
PNES
ES
PNES
ES
PNES
ES
PNES
Azar (2008)
28.5 (9.9)
36.2 (13.3)
16 (24)
NS
NS
PNES longer than ES (p = 0.003)
37.1 (12.3) 33.2 (8.2) NS NS 25 25 NS NS Total cohort-35.2 (12.8) 26.4 (14) 38.2 (12) Median 39 Median 36
65 (24) 43 (19) 40.2 (17.8) NS (all seizures b2 m) 119 (50) 28.2 (7.79) 83.3 (50.5) 51 (30) NS
140 (62)
Brown (1991) Henry (1998) Jedrzejczak (1999) Kanner (1990) Pierelli (1989) Saygi (1992) Syed (2011)
GTCS 15 (23) FLHS 9 (21) 25 (NS) 133 (70) 55 (74) 12 (63) 12 (208) 11 (63) 23 (84)
2547.3 (4096.2) 438 1427 (4035) 173.1 (158) 724.5 (1557.5) 176 (166) NS
NS NS NS 26.6 60 NS NS
NS NS NS 132.3 240 NS NS
PNES longer than ES (p = 0.004) ES b 2 m, PNES ≥ 2 m (p b 0.001) PNES longer than ES (p b 0.001) PNES longer than ES (p = 0.01) PNES longer than ES (p b 0.0001) PNES longer than ES (p b 0.05) Seizure N2 min diagnoses PNES with sensitivity 67% & specificity 48%
23 (NS) 24 (28) 55 (221) 44 (111) 15 (87) 12 (29) 12 (36)
Seizure duration mean in seconds (SD)
Seizure duration median in seconds
Conclusions
ES = epileptic seizure; PNES = psychogenic nonepileptic seizure; N = total number; NS = not specified; SD = standard deviation; GTCS = generalized tonic-clonic seizure; FLHS = frontal lobe hypermotor seizure.
2. Methods We retrospectively reviewed all consecutive VEEG recordings of patients who underwent monitoring at the epilepsy monitoring unit (EMU) of Monash Medical Centre, Melbourne, Australia from May 2005 to June 2015. We only included adults (≥ 18 years) and selected studies which captured PNES and ES in isolation or combination. Events with subjective symptoms without obvious semiological features were excluded. Similarly, electrographic epileptic seizures without clinical semiology were excluded from the analysis. The VEEG data were acquired using the Compumedics digital EEG system (Compumedics Ltd., Melbourne, Australia) with the international 10–20 system of electrode placement. Antiepileptic drug (AED) tapering and sleep deprivation were routine practices in the EMU. We did not use other seizure provocation techniques such as hyperventilation, intermittent photic stimulation, and placebo injections. We collated clinical and demographic data from medical records. We reviewed all seizures captured on VEEG during the study period. The final diagnosis of PNES or ES was based on the consensus opinion of at least two epileptologists after reviewing clinical information, investigation results, and VEEG findings, including semiology. The diagnosis had been established following the VEEG monitoring, in the multidisciplinary meeting, prior to the current study. We considered this consensus diagnosis as the “gold standard” for our study. The reader is referred to Seneviratne et al. for a more detailed account [16]. For the current study, two investigators, an epileptologist (US) and an EEG technologist (EM), studied each seizure video carefully, in synchrony with the EEG, to measure the ictal duration. We measured ictal duration from the onset of first observable behavioral change to the offset of clinical semiology, based on the consensus between the two raters. Psychogenic nonepileptic seizures were classified based on the semiology as described previously [16]. We estimated the mean ictal duration using linear mixed models adjusting for repeat measures. As the ictal duration had a positively skewed distribution, logarithmic transformation was applied prior to the analysis with the results reported as geometric means and 95% confidence intervals (CI). The risk of being PNES when a seizure lasts 5 min or more was estimated using logistic regression adjusting for repeat measures with the results reported as an odds ratio and the corresponding 95% CI. We performed receiver operating characteristic (ROC) curve analysis to study the overall diagnostic accuracy of ictal duration for diagnosing PNES. The receiver operating characteristic curve of ictal duration was constructed by plotting sensitivity against 1-specificity at a range of thresholds [17]. We used the area under the curve (AUC) as the measure of overall diagnostic accuracy [18]. The AUC was interpreted as follows; 0.5, differentiation of PNES from ES no better
than chance; 0.6–0.69, poor differentiation; 0.7–0.79 fair differentiation; 0.8–0.89, good differentiation; and 0.9–1, outstanding differentiation [18]. When a patient had more than one seizure recorded, we used the mean ictal duration of the subject as our measure to construct the ROC curve. When both ES and PNES were captured from the same subject, we calculated means separately for the two types of events. The optimal cut-off point for the ictal duration to differentiate PNES from ES was determined using Youden's index [19]. The data analyses were performed with IBM SPSS (version 21) statistical software package (IBM Corporation, New York, USA) and SAS software version 9.4 (SAS Institute, Cary, NC, USA). A p value b0.05 indicated statistical significance. The study protocol was approved by the Human Research Ethics Committees of Monash Health.
3. Results We studied a total of 782 seizures (ES,441; PNES,341) from 138 patients consisting of 72 (52.2%) females and 66 (47.8%) males with the mean age (±SD) of 43 ± 16.6 years (range, 18–91). A higher proportion of females was seen in the PNES group compared with the ES group (71% vs 37%). Mean ages of ES and PNES cohorts were comparable (44.2 ± 17.8 & 41.7 ± 15.3 respectively). Epileptic seizures alone were captured from 73 (52.9%) patients, whereas 62 (44.9%) had only PNES. Both ES and PNES were recorded from three patients (2.2%). In the “ES alone” group, 62 (84.9%) and 11 (15.1%) had focal and generalized epilepsy syndromes, respectively. Different ES types in the cohort are summarized in Table 2. Table 3 highlights semiologic subtypes of PNES and their corresponding durations. The mean duration of VEEG monitoring was 4.5 ± 1.5 days. In the cohort, 93% of ES lasted b 2 min compared with 48% in PNES. Furthermore, 21.4% of PNES were 5 min or longer compared with 1.1% in ES (Table 4 & Fig. 1 A). In the mixed model analysis, the geometric means of the ictal duration of ES and PNES were 47.7 (95% CI: 37.6–60.6) and 148.7 (95% CI: 115.2–191.8) seconds, respectively. PNES was about three times longer than ES and the difference was statistically significant (p b 0.001). The analysis further revealed that the odds of being PNES were about 24 times greater (odds ratio: 23.8, 95% CI: 7.9–71.3) when the ictal duration was 5 min or more compared to those with b5 min of ictal duration. The ROC curve revealed an AUC of 0.80 (95% CI 0.73–0.88) indicating a good overall diagnostic accuracy of the test (Fig. 1B). Table 5 summarizes sensitivities, specificities, and predictive values at different cut-off values of ictal duration for PNES diagnosis. The diagnostic specificity for PNES increases and the sensitivity decreases with increasing ictal
U. Seneviratne et al. / Epilepsy & Behavior 66 (2017) 127–131 Table 2 Epileptic seizure types in the cohort.
129
Table 4 Comparison of ictal duration between epileptic and psychogenic nonepileptic seizures.
Seizure type
Frequency (%)
Complex partial Generalized tonic Absence Simple partial Secondarily generalized tonic-clonic Myoclonic Generalized tonic-clonic Atonic
287 (65.1) 61 (13.8) 35 (7.9) 17 (3.9) 15 (3.4) 14 (3.2) 7 (1.6) 5 (1.1)
duration. Youden's index identified the optimal cut-off point as 123.5 s. At this threshold, the sensitivity was 65% and the specificity was 93%. 4. Discussion We have studied the differences in ictal duration between PNES and ES in a large cohort of patients assessed in the EMU of a tertiary hospital. Our study reaffirms that the mean ictal duration of PNES is longer than ES. Additionally, we report three novel findings: (1) when seizures last ≥5 min, the odds of being PNES are 24 times higher, (2) as indicated by the AUC of the ROC curve, ictal duration is a useful diagnostic test to differentiate PNES from ES, and (3) ictal duration longer than 123.5 s is the optimal threshold value to diagnose PNES with 65% sensitivity and 93% specificity. These results have important implications for routine clinical practice. The difficulties in differentiating PNES from ES have been well described [20]. The use of various semiological features in the diagnosis has been studied by many researchers, but the diagnostic accuracy of semiological signs is variable [21]. The lack of a quantitative measure in some signs and the variability in accuracy among different observers are the main limitations in the use of semiology [6]. Previous studies have shown that long ictal duration, fluctuating course, asynchronous movements, pelvic thrusting, side-to-side movements, ictal eye closure, ictal crying, and memory recall are helpful signs to diagnose PNES with variable sensitivities and specificities [21]. Ictal duration is a quantifiable sign that can be measured without VEEG facilities. Hence, it could potentially be a very useful sign if the diagnostic accuracy is validated. Our literature search yielded nine studies evaluating the use of ictal duration to differentiate PNES from ES and eight of them found PNES to be significantly longer than ES, similar to our results (Table 1). However, only one study reported a measure of diagnostic accuracy [15]. In this study, 45 semiological signs were analyzed among 120 seizures. Seizures lasting N2 min were found to be diagnostic of PNES with 67% sensitivity and 48% specificity [15]. With this methodology, only one diagnostic threshold (2 min) is reported. As ictal duration is a continuous measure, ROC curve analysis is a more robust method to find the overall diagnostic accuracy of the sign as well as different sensitivities and specificities at multiple thresholds [17]. Our ROC curve analysis yielded an AUC of 0.8, indicating that ictal duration is a good test to distinguish PNES from ES. Furthermore, we report diagnostic sensitivities, specificities, and predictive values at multiple cut-off points of ictal duration. It should be noted that with increasing ictal duration, diagnostic specificity for PNES increases and sensitivity decreases. Hence, if a higher threshold is selected, the false
Table 3 Psychogenic nonepileptic seizure types with corresponding ictal durations. Seizure type
Number (%)
Median duration (inter quartile range) (seconds)
Complex motor Rhythmic motor Dialeptic Mixed Hypermotor
127 (37.2) 102 (29.9) 59 (17.3) 48 (14.1) 5 (1.5)
129 (64.5 to 259.5) 128.5 (54.5 to 312.2) 198 (30 to 333) 213 (135.7 to 489.7) 194 (119 to 240.5)
Ictal duration Epileptic seizures (seconds)
≤59 60–119 120–179 180–239 240–299 ≥300 Total
Psychogenic nonepileptic seizures
Frequency %
Cumulative % Frequency %
Cumulative %
287 124 19 5 1 5 441
65.1 93.2 97.5 98.6 98.9 100 100
28.7 48.1 63.3 70.7 78.6 100 100
65.1 28.1 4.3 1.1 0.2 1.1 100
98 66 52 25 27 73 341
28.7 19.4 15.2 7.3 7.9 21.4 100
negative rate is likely to be higher. Therefore we emphasize that the diagnosis of PNES should not be entirely based on a single sign. Rather, the ictal duration can be used as a diagnostic sign to raise suspicion of PNES in order to plan further investigations such as video-EEG monitoring. Epileptic seizures lasting ≥ 5 min are defined as status epilepticus [22]. This is a serious neurological emergency requiring urgent interventions such as intravenous antiepileptic medications and sometimes intubation with monitoring in the intensive care unit. Our study reveals a potential pitfall here. We found that seizures lasting ≥5 min were 24 times more likely to be PNES. When a patient presents with a prolonged seizure, quick decisions have to be made based on the clinical grounds including semiology. The epileptology expertise and emergency EEGs are not often available in this situation. Hence, there is a high likelihood for PNES to be misdiagnosed as status epilepticus in the emergency setting. This danger is highlighted in the literature. In the intensive care unit of a tertiary hospital, 23% of cases with status epilepticus were later confirmed as “pseudostatus epilepticus” [23]. Another study reported that 27% of patients with PNES had intensive care admissions with pseudostatus epilepticus [24]. Misdiagnosis of PNES as status epilepticus leads to unnecessary interventions such as intubation, intravenous benzodiazepines, and central venous catheters, resulting in complications including respiratory depression, and even death [4, 23,25]. Hence, we emphasize the clinical importance of this potential pitfall. Our study has some limitations. The retrospective study design is a major limitation. This study was conducted in the epilepsy monitoring unit at a tertiary center, thereby introducing a bias. Patients with refractory epilepsy were overrepresented in the cohort and their findings may not be directly extrapolated to all patients presenting with seizures. Similarly, the proportion of complex partial seizures was much higher than other seizure types, most probably reflecting the sampling bias. Antiepileptic drug withdrawal, as well as sleep deprivation, are routine practices in the EMU. Additionally, the policy of our EMU is to terminate seizures with intravenous benzodiazepines when a convulsive seizure lasts N 5 min and a non-convulsive seizure lasts N10 min. Hence, ictal duration recorded in the EMU may be different from that of naturallyoccurring seizures. It should also be noted that the two observers in our study were not blinded to the diagnosis. Our retrospective methodology does not reflect how seizures are assessed in the clinical setting. This is a limitation. However, reviewing videos by two experienced observers is likely to provide a more accurate measure of the ictal duration. We also would like to highlight potential pitfalls in the use of ictal duration. Defining semiological seizure onset and offset may be difficult for an observer without epilepsy training. Time estimates may also be wrong, and in a previous study, we demonstrated that observers tended to underestimate ictal duration [26]. Misdiagnosing status epilepticus as PNES based on ictal duration may cause unwarranted morbidity and mortality if emergency treatment is withheld. We defined seizure onset as the first observed change in behavior and offset as the return to the baseline with the termination of behavioral changes. This distinction could be subjective and observer dependent, particularly without an accompanying EEG, in the real-life
130
U. Seneviratne et al. / Epilepsy & Behavior 66 (2017) 127–131
Fig. 1. A: Histogram illustrating ictal (seizure) duration of epileptic and psychogenic nonepileptic seizures. B: The receiver operating characteristics curve of ictal duration. (ES = epileptic seizure; PNES = psychogenic nonepileptic seizure).
setting. Further, seizure termination may merge into the postictal symptoms and the distinction could be challenging without a synchronous EEG recording. We excluded auras from the study; hence our results cannot be extrapolated to interpret the ictal duration of auras. In our study, we only analyzed the seizure duration measured by two trained observers. Therefore, the results cannot be directly extrapolated to other settings such as outpatient clinics and lay people's estimates of ictal duration. However, there are indirect ways to measure the ictal duration in the outpatient setting. Family members may bring videos of seizures captured on smartphones and video cameras. Home videos are useful in the evaluation of seizures [27]. Doctors can potentially use those home videos to measure ictal duration. Diagnosing PNES is a challenging task [28]. Semiology is useful, but it should not be used as the sole criterion. Though we have focused on a single semiological feature in our study, the diagnosis should not be based on a single sign. We emphasize that PNES should be diagnosed based on multiple criteria, including history, investigations, and semiology as recommended by the International League Against Epilepsy Task Force [29]. Our study shows that ictal duration is a useful test in the EMU setting when the duration is assessed by experienced observers. To be more meaningful clinically, our hypothesis needs to be tested prospectively, in the real-life setting in hospitals, involving patients presenting with seizures.
5. Conclusions In summary, our study demonstrates that ictal duration is significantly longer in PNES. Hence, prolonged seizures are more likely to be PNES than ES and can potentially be misdiagnosed as status epilepticus. The results also indicate that ictal duration can be used as a good test to differentiate PNES from ES and we provide diagnostic sensitivities, specificities, and predictive values at different thresholds. Given the complex nature of this disorder, we suggest ictal duration be used as the initial guide to planning further evaluation in order to arrive at the final diagnosis. Disclosure Associate Professor Seneviratne has received travel and speaker honoraria from UCB Pharma. Mr. Paul and Miss Minato do not report any disclosures.
Funding This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
References Table 5 Sensitivity, specificity, and positive and negative predictive values for different thresholds of the ictal duration for the PNES diagnosis. Ictal duration (seconds)
Sensitivity (%)
Specificity (%)
PPV (%)
NPV (%)
≥30 ≥60 ≥90 ≥120 ≥150 ≥180 ≥210 ≥240 ≥270 ≥300
92.3 81.5 75.4 64.6 56.9 43.1 33.8 29.2 26.2 26.2
19.2 42.5 69.9 90.4 94.5 97.3 97.3 98.6 100.0 100.0
50.4 55.8 69.0 85.7 90.2 93.3 91.7 95.0 100.0 100.0
73.7 72.1 76.1 74.2 71.1 65.7 62.3 61.0 60.3 60.3
PPV: positive predictive value, NPV: negative predictive value, PNES: psychogenic nonepileptic seizure.
[1] Scheepers B, Clough P, Pickles C. The misdiagnosis of epilepsy: findings of a population study. Seizure-European Journal of Epilepsy 1998;7:403–6. [2] Smith D, Defalla BA, Chadwick DW. The misdiagnosis of epilepsy and the management of refractory epilepsy in a specialist clinic. QJM 1999;92:15–23. [3] Howell SJ, Owen L, Chadwick DW. Pseudostatus epilepticus. Q J Med 1989;71: 507–19. [4] Reuber M, Baker GA, Gill R, Smith DF, Chadwick DW. Failure to recognize psychogenic nonepileptic seizures may cause death. Neurology 2004;62:834–5. [5] Dworetzky BA, Bubrick EJ, Szaflarski JP. Nonepileptic psychogenic status: markedly prolonged psychogenic nonepileptic seizures. Epilepsy Behav 2010;19:65–8. [6] Seneviratne U, Rajendran D, Brusco M, Phan TG. How good are we at diagnosing seizures based on semiology? Epilepsia 2012;53:e63–6. [7] Erba G, Giussani G, Juersivich A, Magaudda A, Chiesa V, Lagana A, et al. The semiology of psychogenic nonepileptic seizures revisited: can video alone predict the diagnosis? Preliminary data from a prospective feasibility study. Epilepsia 2016; 57:777–85. [8] Azar NJ, Tayah TF, Wang L, Song Y, Abou-Khalil BW. Postictal breathing pattern distinguishes epileptic from nonepileptic convulsive seizures. Epilepsia 2008;49:132–7. [9] Brown MC, Levin BE, Ramsay RE, Katz DA, Duchowny MS. Characteristics of patients with nonepileptic seizures. J Epilepsy 1991;4:225–9.
U. Seneviratne et al. / Epilepsy & Behavior 66 (2017) 127–131 [10] 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;39:175–82. [11] Jedrzejczak J, Owczarek K, Majkowski J. Psychogenic pseudoepileptic seizures: clinical and electroencephalogram (EEG) video-tape recordings. Eur J Neurol 1999;6:473–9. [12] Kanner AM, Morris HH, Luders H, Dinner DS, Wyllie E, Medendorp SV, et al. Supplementary motor seizures mimicking pseudoseizures: some clinical differences. Neurology 1990;40:1404–7. [13] 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;30:513–23. [14] Saygi S, Katz A, Marks DA, Spencer SS. Frontal lobe partial seizures and psychogenic seizures: comparison of clinical and ictal characteristics. Neurology 1992;42:1274–7. [15] Syed TU, LaFrance Jr WC, Kahriman ES, Hasan SN, Rajasekaran V, Gulati D, et al. Can semiology predict psychogenic nonepileptic seizures? A prospective study. Ann Neurol 2011;69:997–1004. [16] Seneviratne U, Reutens D, D'Souza W. Stereotypy of psychogenic nonepileptic seizures: insights from video-EEG monitoring. Epilepsia 2010;51:1159–68. [17] Mallett S, Halligan S, Thompson M, Collins GS, Altman DG. Interpreting diagnostic accuracy studies for patient care. BMJ 2012;345, e3999. [18] Hosmer D, Lemeshow S. Applied logistic regression. 2nd edition. New Jersey: John Wiley & Sons; 2000. [19] Youden WJ. Index for rating diagnostic tests. Cancer 1950;3:32–5. [20] Dickinson P, Looper KJ. Psychogenic nonepileptic seizures: a current overview. Epilepsia 2012;53:1679–89.
131
[21] 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;81:719–25. [22] Trinka E, Cock H, Hesdorffer D, Rossetti AO, Scheffer IE, Shinnar S, et al. A definition and classification of status epilepticus-report of the ILAE Task Force on Classification of Status Epilepticus. Epilepsia 2015;56:1515–23. [23] Walker MC, Howard RS, Smith SJ, Miller DH, Shorvon SD, Hirsch NP. Diagnosis and treatment of status epilepticus on a neurological intensive care unit. QJM 1996;89: 913–20. [24] Reuber M, Pukrop R, Mitchell AJ, Bauer J, Elger CE. Clinical significance of recurrent psychogenic nonepileptic seizure status. J Neurol 2003;250:1355–62. [25] Holtkamp M, Othman J, Buchheim K, Meierkord H. Diagnosis of psychogenic nonepileptic status epilepticus in the emergency setting. Neurology 2006;66: 1727–9. [26] Seneviratne U, Ma H, Phan TG. How do doctors in training react to seizures? Epilepsy Behav 2016;54:104–9. [27] Samuel M, Duncan JS. Use of the hand held video camcorder in the evaluation of seizures. J Neurol Neurosurg Psychiatry 1994;57:1417–8. [28] Gates JR, Ramani V, Whalen S, Loewenson R. Ictal characteristics of pseudoseizures. Arch Neurol 1985;42:1183–7. [29] LaFrance Jr WC, Baker GA, Duncan R, Goldstein LH, Reuber M. Minimum requirements for the diagnosis of psychogenic nonepileptic seizures: a staged approach: a report from the International League Against Epilepsy Nonepileptic Seizures Task Force. Epilepsia 2013;54:2005–18.
Contents lists available at ScienceDirect
Epilepsy & Behavior journal homepage: www.elsevier.com/locate/yebeh
How reliable is ictal duration to differentiate psychogenic nonepileptic seizures from epileptic seizures? Udaya Seneviratne a,b,⁎, Erica Minato a, Eldho Paul c,d a
Department of Neuroscience, Monash Medical Centre, Melbourne, Australia Department of Medicine, School of Clinical Sciences at Monash Health, Monash University, Melbourne, Australia Monash Centre for Health Research and Implementation, School of Public Health and Preventive Medicine, Monash University, Melbourne, Australia d Department of Clinical Haematology, Alfred Hospital, Melbourne, Australia b c
a r t i c l e
i n f o
Article history: Received 12 August 2016 Revised 19 October 2016 Accepted 19 October 2016 Available online xxxx Keywords: Epilepsy monitoring Seizure semiology Psychogenic nonepileptic seizure Receiver operating characteristics curve Status epilepticus Diagnostic pitfalls
a b s t r a c t We sought to investigate (1) differences in ictal duration between psychogenic nonepileptic seizures (PNES) and epileptic seizures (ES), (2) the odds of being PNES when seizures last ≥5 min, and (3) the value of ictal duration as a diagnostic test to differentiate PNES from ES. We retrospectively reviewed video-EEG recordings and tabulated ictal durations of all PNES and ES. We estimated the mean ictal durations of PNES and ES using linear mixed models. The odds of being PNES when seizures last ≥5 min were estimated using logistic regression. We used receiver operating characteristics (ROC) curves to study the overall diagnostic accuracy of ictal duration in differentiating PNES from ES. We studied 441 ES and 341 PNES recorded from 138 patients. The mean ictal duration of PNES (148.7 s, 95% CI: 115.2–191.8) was significantly longer (p b 0.001) than that of ES (47.7 s, 95% CI: 37.6–60.6). The odds of being PNES was about 24 times higher (Odds ratio: 23.8, 95% CI: 7.9–71.3) when the ictal duration was ≥5 min. The ROC curve yielded an area under the curve of 0.80 (95% CI 0.73–0.88). Youden's index identified 123.5 s as the optimal threshold to diagnose PNES with 65% sensitivity and 93% specificity. Our results indicate that ictal duration is a useful test to raise suspicion of PNES. When a seizure lasts ≥5 min, it is 24 times more likely to be PNES with the potential risk of misdiagnosis as status epilepticus. © 2016 Elsevier Inc. All rights reserved.
1. Introduction Differentiating psychogenic nonepileptic seizures (PNES) from epileptic seizures (ES) is a major diagnostic challenge faced by clinicians and the rate of misdiagnosis is as high as 20–30% [1,2]. The misdiagnosis of PNES as ES often leads to unnecessary interventions and treatment with antiepileptic drugs resulting in adverse outcomes including death [3,4]. Prolonged PNES mimicking status epilepticus (“pseudostatus epilepticus”) is in particular a challenging condition to diagnose [5]. The gold standard diagnostic test for seizures is videoelectroencephalographic (VEEG) monitoring. However, it is an expensive investigation with limited availability. Therefore, researchers have evaluated the use of semiology without EEG in differentiating PNES from ES. The results indicate variable diagnostic accuracy depending on the experience of the observer [6,7].
⁎ Corresponding author at: Department of Neuroscience, Monash Medical Centre, Clayton, Melbourne, VIC 3168, Australia. E-mail addresses: [email protected] (U. Seneviratne), [email protected] (E. Minato), [email protected] (E. Paul).
http://dx.doi.org/10.1016/j.yebeh.2016.10.024 1525-5050/© 2016 Elsevier Inc. All rights reserved.
One major drawback in using most semiological features is the lack of quantifiable objective measurements. Being an objective measure, ictal (seizure) duration is a unique semiological sign. Several studies have found that ictal duration is significantly longer in PNES compared with ES (Table 1) [8–14]. However, only one study presents diagnostic accuracy at a single threshold [15], and none of the studies, as summarized in Table 1, provides information on the overall diagnostic accuracy of the ictal duration as a test. Seizures lasting ≥5 min are diagnosed as status epilepticus and treated aggressively. Hence, there is a potential risk for PNES to be misdiagnosed as status epilepticus because PNES tend to last longer than ES. Against this backdrop, we sought to investigate three research questions in relation to ictal duration; (1) How does ictal duration of PNES differ from that of ES? (2) When seizures last ≥5 min, what are the odds of them being PNES, which can potentially be misdiagnosed as status epilepticus? (3) How reliable is ictal duration as a diagnostic test to differentiate PNES from ES? We hypothesized that ictal duration is a reliable test to differentiate PNES from ES, and when a seizure lasts ≥5 min the odds are higher for being PNES with the potential risk of misdiagnosis as status epilepticus.
128
U. Seneviratne et al. / Epilepsy & Behavior 66 (2017) 127–131
Table 1 Summary of previous studies on the diagnostic value of ictal duration to differentiate ES from PNES. Reference
Mean age (SD)
N subjects (N seizures)
ES
PNES
ES
PNES
ES
PNES
ES
PNES
Azar (2008)
28.5 (9.9)
36.2 (13.3)
16 (24)
NS
NS
PNES longer than ES (p = 0.003)
37.1 (12.3) 33.2 (8.2) NS NS 25 25 NS NS Total cohort-35.2 (12.8) 26.4 (14) 38.2 (12) Median 39 Median 36
65 (24) 43 (19) 40.2 (17.8) NS (all seizures b2 m) 119 (50) 28.2 (7.79) 83.3 (50.5) 51 (30) NS
140 (62)
Brown (1991) Henry (1998) Jedrzejczak (1999) Kanner (1990) Pierelli (1989) Saygi (1992) Syed (2011)
GTCS 15 (23) FLHS 9 (21) 25 (NS) 133 (70) 55 (74) 12 (63) 12 (208) 11 (63) 23 (84)
2547.3 (4096.2) 438 1427 (4035) 173.1 (158) 724.5 (1557.5) 176 (166) NS
NS NS NS 26.6 60 NS NS
NS NS NS 132.3 240 NS NS
PNES longer than ES (p = 0.004) ES b 2 m, PNES ≥ 2 m (p b 0.001) PNES longer than ES (p b 0.001) PNES longer than ES (p = 0.01) PNES longer than ES (p b 0.0001) PNES longer than ES (p b 0.05) Seizure N2 min diagnoses PNES with sensitivity 67% & specificity 48%
23 (NS) 24 (28) 55 (221) 44 (111) 15 (87) 12 (29) 12 (36)
Seizure duration mean in seconds (SD)
Seizure duration median in seconds
Conclusions
ES = epileptic seizure; PNES = psychogenic nonepileptic seizure; N = total number; NS = not specified; SD = standard deviation; GTCS = generalized tonic-clonic seizure; FLHS = frontal lobe hypermotor seizure.
2. Methods We retrospectively reviewed all consecutive VEEG recordings of patients who underwent monitoring at the epilepsy monitoring unit (EMU) of Monash Medical Centre, Melbourne, Australia from May 2005 to June 2015. We only included adults (≥ 18 years) and selected studies which captured PNES and ES in isolation or combination. Events with subjective symptoms without obvious semiological features were excluded. Similarly, electrographic epileptic seizures without clinical semiology were excluded from the analysis. The VEEG data were acquired using the Compumedics digital EEG system (Compumedics Ltd., Melbourne, Australia) with the international 10–20 system of electrode placement. Antiepileptic drug (AED) tapering and sleep deprivation were routine practices in the EMU. We did not use other seizure provocation techniques such as hyperventilation, intermittent photic stimulation, and placebo injections. We collated clinical and demographic data from medical records. We reviewed all seizures captured on VEEG during the study period. The final diagnosis of PNES or ES was based on the consensus opinion of at least two epileptologists after reviewing clinical information, investigation results, and VEEG findings, including semiology. The diagnosis had been established following the VEEG monitoring, in the multidisciplinary meeting, prior to the current study. We considered this consensus diagnosis as the “gold standard” for our study. The reader is referred to Seneviratne et al. for a more detailed account [16]. For the current study, two investigators, an epileptologist (US) and an EEG technologist (EM), studied each seizure video carefully, in synchrony with the EEG, to measure the ictal duration. We measured ictal duration from the onset of first observable behavioral change to the offset of clinical semiology, based on the consensus between the two raters. Psychogenic nonepileptic seizures were classified based on the semiology as described previously [16]. We estimated the mean ictal duration using linear mixed models adjusting for repeat measures. As the ictal duration had a positively skewed distribution, logarithmic transformation was applied prior to the analysis with the results reported as geometric means and 95% confidence intervals (CI). The risk of being PNES when a seizure lasts 5 min or more was estimated using logistic regression adjusting for repeat measures with the results reported as an odds ratio and the corresponding 95% CI. We performed receiver operating characteristic (ROC) curve analysis to study the overall diagnostic accuracy of ictal duration for diagnosing PNES. The receiver operating characteristic curve of ictal duration was constructed by plotting sensitivity against 1-specificity at a range of thresholds [17]. We used the area under the curve (AUC) as the measure of overall diagnostic accuracy [18]. The AUC was interpreted as follows; 0.5, differentiation of PNES from ES no better
than chance; 0.6–0.69, poor differentiation; 0.7–0.79 fair differentiation; 0.8–0.89, good differentiation; and 0.9–1, outstanding differentiation [18]. When a patient had more than one seizure recorded, we used the mean ictal duration of the subject as our measure to construct the ROC curve. When both ES and PNES were captured from the same subject, we calculated means separately for the two types of events. The optimal cut-off point for the ictal duration to differentiate PNES from ES was determined using Youden's index [19]. The data analyses were performed with IBM SPSS (version 21) statistical software package (IBM Corporation, New York, USA) and SAS software version 9.4 (SAS Institute, Cary, NC, USA). A p value b0.05 indicated statistical significance. The study protocol was approved by the Human Research Ethics Committees of Monash Health.
3. Results We studied a total of 782 seizures (ES,441; PNES,341) from 138 patients consisting of 72 (52.2%) females and 66 (47.8%) males with the mean age (±SD) of 43 ± 16.6 years (range, 18–91). A higher proportion of females was seen in the PNES group compared with the ES group (71% vs 37%). Mean ages of ES and PNES cohorts were comparable (44.2 ± 17.8 & 41.7 ± 15.3 respectively). Epileptic seizures alone were captured from 73 (52.9%) patients, whereas 62 (44.9%) had only PNES. Both ES and PNES were recorded from three patients (2.2%). In the “ES alone” group, 62 (84.9%) and 11 (15.1%) had focal and generalized epilepsy syndromes, respectively. Different ES types in the cohort are summarized in Table 2. Table 3 highlights semiologic subtypes of PNES and their corresponding durations. The mean duration of VEEG monitoring was 4.5 ± 1.5 days. In the cohort, 93% of ES lasted b 2 min compared with 48% in PNES. Furthermore, 21.4% of PNES were 5 min or longer compared with 1.1% in ES (Table 4 & Fig. 1 A). In the mixed model analysis, the geometric means of the ictal duration of ES and PNES were 47.7 (95% CI: 37.6–60.6) and 148.7 (95% CI: 115.2–191.8) seconds, respectively. PNES was about three times longer than ES and the difference was statistically significant (p b 0.001). The analysis further revealed that the odds of being PNES were about 24 times greater (odds ratio: 23.8, 95% CI: 7.9–71.3) when the ictal duration was 5 min or more compared to those with b5 min of ictal duration. The ROC curve revealed an AUC of 0.80 (95% CI 0.73–0.88) indicating a good overall diagnostic accuracy of the test (Fig. 1B). Table 5 summarizes sensitivities, specificities, and predictive values at different cut-off values of ictal duration for PNES diagnosis. The diagnostic specificity for PNES increases and the sensitivity decreases with increasing ictal
U. Seneviratne et al. / Epilepsy & Behavior 66 (2017) 127–131 Table 2 Epileptic seizure types in the cohort.
129
Table 4 Comparison of ictal duration between epileptic and psychogenic nonepileptic seizures.
Seizure type
Frequency (%)
Complex partial Generalized tonic Absence Simple partial Secondarily generalized tonic-clonic Myoclonic Generalized tonic-clonic Atonic
287 (65.1) 61 (13.8) 35 (7.9) 17 (3.9) 15 (3.4) 14 (3.2) 7 (1.6) 5 (1.1)
duration. Youden's index identified the optimal cut-off point as 123.5 s. At this threshold, the sensitivity was 65% and the specificity was 93%. 4. Discussion We have studied the differences in ictal duration between PNES and ES in a large cohort of patients assessed in the EMU of a tertiary hospital. Our study reaffirms that the mean ictal duration of PNES is longer than ES. Additionally, we report three novel findings: (1) when seizures last ≥5 min, the odds of being PNES are 24 times higher, (2) as indicated by the AUC of the ROC curve, ictal duration is a useful diagnostic test to differentiate PNES from ES, and (3) ictal duration longer than 123.5 s is the optimal threshold value to diagnose PNES with 65% sensitivity and 93% specificity. These results have important implications for routine clinical practice. The difficulties in differentiating PNES from ES have been well described [20]. The use of various semiological features in the diagnosis has been studied by many researchers, but the diagnostic accuracy of semiological signs is variable [21]. The lack of a quantitative measure in some signs and the variability in accuracy among different observers are the main limitations in the use of semiology [6]. Previous studies have shown that long ictal duration, fluctuating course, asynchronous movements, pelvic thrusting, side-to-side movements, ictal eye closure, ictal crying, and memory recall are helpful signs to diagnose PNES with variable sensitivities and specificities [21]. Ictal duration is a quantifiable sign that can be measured without VEEG facilities. Hence, it could potentially be a very useful sign if the diagnostic accuracy is validated. Our literature search yielded nine studies evaluating the use of ictal duration to differentiate PNES from ES and eight of them found PNES to be significantly longer than ES, similar to our results (Table 1). However, only one study reported a measure of diagnostic accuracy [15]. In this study, 45 semiological signs were analyzed among 120 seizures. Seizures lasting N2 min were found to be diagnostic of PNES with 67% sensitivity and 48% specificity [15]. With this methodology, only one diagnostic threshold (2 min) is reported. As ictal duration is a continuous measure, ROC curve analysis is a more robust method to find the overall diagnostic accuracy of the sign as well as different sensitivities and specificities at multiple thresholds [17]. Our ROC curve analysis yielded an AUC of 0.8, indicating that ictal duration is a good test to distinguish PNES from ES. Furthermore, we report diagnostic sensitivities, specificities, and predictive values at multiple cut-off points of ictal duration. It should be noted that with increasing ictal duration, diagnostic specificity for PNES increases and sensitivity decreases. Hence, if a higher threshold is selected, the false
Table 3 Psychogenic nonepileptic seizure types with corresponding ictal durations. Seizure type
Number (%)
Median duration (inter quartile range) (seconds)
Complex motor Rhythmic motor Dialeptic Mixed Hypermotor
127 (37.2) 102 (29.9) 59 (17.3) 48 (14.1) 5 (1.5)
129 (64.5 to 259.5) 128.5 (54.5 to 312.2) 198 (30 to 333) 213 (135.7 to 489.7) 194 (119 to 240.5)
Ictal duration Epileptic seizures (seconds)
≤59 60–119 120–179 180–239 240–299 ≥300 Total
Psychogenic nonepileptic seizures
Frequency %
Cumulative % Frequency %
Cumulative %
287 124 19 5 1 5 441
65.1 93.2 97.5 98.6 98.9 100 100
28.7 48.1 63.3 70.7 78.6 100 100
65.1 28.1 4.3 1.1 0.2 1.1 100
98 66 52 25 27 73 341
28.7 19.4 15.2 7.3 7.9 21.4 100
negative rate is likely to be higher. Therefore we emphasize that the diagnosis of PNES should not be entirely based on a single sign. Rather, the ictal duration can be used as a diagnostic sign to raise suspicion of PNES in order to plan further investigations such as video-EEG monitoring. Epileptic seizures lasting ≥ 5 min are defined as status epilepticus [22]. This is a serious neurological emergency requiring urgent interventions such as intravenous antiepileptic medications and sometimes intubation with monitoring in the intensive care unit. Our study reveals a potential pitfall here. We found that seizures lasting ≥5 min were 24 times more likely to be PNES. When a patient presents with a prolonged seizure, quick decisions have to be made based on the clinical grounds including semiology. The epileptology expertise and emergency EEGs are not often available in this situation. Hence, there is a high likelihood for PNES to be misdiagnosed as status epilepticus in the emergency setting. This danger is highlighted in the literature. In the intensive care unit of a tertiary hospital, 23% of cases with status epilepticus were later confirmed as “pseudostatus epilepticus” [23]. Another study reported that 27% of patients with PNES had intensive care admissions with pseudostatus epilepticus [24]. Misdiagnosis of PNES as status epilepticus leads to unnecessary interventions such as intubation, intravenous benzodiazepines, and central venous catheters, resulting in complications including respiratory depression, and even death [4, 23,25]. Hence, we emphasize the clinical importance of this potential pitfall. Our study has some limitations. The retrospective study design is a major limitation. This study was conducted in the epilepsy monitoring unit at a tertiary center, thereby introducing a bias. Patients with refractory epilepsy were overrepresented in the cohort and their findings may not be directly extrapolated to all patients presenting with seizures. Similarly, the proportion of complex partial seizures was much higher than other seizure types, most probably reflecting the sampling bias. Antiepileptic drug withdrawal, as well as sleep deprivation, are routine practices in the EMU. Additionally, the policy of our EMU is to terminate seizures with intravenous benzodiazepines when a convulsive seizure lasts N 5 min and a non-convulsive seizure lasts N10 min. Hence, ictal duration recorded in the EMU may be different from that of naturallyoccurring seizures. It should also be noted that the two observers in our study were not blinded to the diagnosis. Our retrospective methodology does not reflect how seizures are assessed in the clinical setting. This is a limitation. However, reviewing videos by two experienced observers is likely to provide a more accurate measure of the ictal duration. We also would like to highlight potential pitfalls in the use of ictal duration. Defining semiological seizure onset and offset may be difficult for an observer without epilepsy training. Time estimates may also be wrong, and in a previous study, we demonstrated that observers tended to underestimate ictal duration [26]. Misdiagnosing status epilepticus as PNES based on ictal duration may cause unwarranted morbidity and mortality if emergency treatment is withheld. We defined seizure onset as the first observed change in behavior and offset as the return to the baseline with the termination of behavioral changes. This distinction could be subjective and observer dependent, particularly without an accompanying EEG, in the real-life
130
U. Seneviratne et al. / Epilepsy & Behavior 66 (2017) 127–131
Fig. 1. A: Histogram illustrating ictal (seizure) duration of epileptic and psychogenic nonepileptic seizures. B: The receiver operating characteristics curve of ictal duration. (ES = epileptic seizure; PNES = psychogenic nonepileptic seizure).
setting. Further, seizure termination may merge into the postictal symptoms and the distinction could be challenging without a synchronous EEG recording. We excluded auras from the study; hence our results cannot be extrapolated to interpret the ictal duration of auras. In our study, we only analyzed the seizure duration measured by two trained observers. Therefore, the results cannot be directly extrapolated to other settings such as outpatient clinics and lay people's estimates of ictal duration. However, there are indirect ways to measure the ictal duration in the outpatient setting. Family members may bring videos of seizures captured on smartphones and video cameras. Home videos are useful in the evaluation of seizures [27]. Doctors can potentially use those home videos to measure ictal duration. Diagnosing PNES is a challenging task [28]. Semiology is useful, but it should not be used as the sole criterion. Though we have focused on a single semiological feature in our study, the diagnosis should not be based on a single sign. We emphasize that PNES should be diagnosed based on multiple criteria, including history, investigations, and semiology as recommended by the International League Against Epilepsy Task Force [29]. Our study shows that ictal duration is a useful test in the EMU setting when the duration is assessed by experienced observers. To be more meaningful clinically, our hypothesis needs to be tested prospectively, in the real-life setting in hospitals, involving patients presenting with seizures.
5. Conclusions In summary, our study demonstrates that ictal duration is significantly longer in PNES. Hence, prolonged seizures are more likely to be PNES than ES and can potentially be misdiagnosed as status epilepticus. The results also indicate that ictal duration can be used as a good test to differentiate PNES from ES and we provide diagnostic sensitivities, specificities, and predictive values at different thresholds. Given the complex nature of this disorder, we suggest ictal duration be used as the initial guide to planning further evaluation in order to arrive at the final diagnosis. Disclosure Associate Professor Seneviratne has received travel and speaker honoraria from UCB Pharma. Mr. Paul and Miss Minato do not report any disclosures.
Funding This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
References Table 5 Sensitivity, specificity, and positive and negative predictive values for different thresholds of the ictal duration for the PNES diagnosis. Ictal duration (seconds)
Sensitivity (%)
Specificity (%)
PPV (%)
NPV (%)
≥30 ≥60 ≥90 ≥120 ≥150 ≥180 ≥210 ≥240 ≥270 ≥300
92.3 81.5 75.4 64.6 56.9 43.1 33.8 29.2 26.2 26.2
19.2 42.5 69.9 90.4 94.5 97.3 97.3 98.6 100.0 100.0
50.4 55.8 69.0 85.7 90.2 93.3 91.7 95.0 100.0 100.0
73.7 72.1 76.1 74.2 71.1 65.7 62.3 61.0 60.3 60.3
PPV: positive predictive value, NPV: negative predictive value, PNES: psychogenic nonepileptic seizure.
[1] Scheepers B, Clough P, Pickles C. The misdiagnosis of epilepsy: findings of a population study. Seizure-European Journal of Epilepsy 1998;7:403–6. [2] Smith D, Defalla BA, Chadwick DW. The misdiagnosis of epilepsy and the management of refractory epilepsy in a specialist clinic. QJM 1999;92:15–23. [3] Howell SJ, Owen L, Chadwick DW. Pseudostatus epilepticus. Q J Med 1989;71: 507–19. [4] Reuber M, Baker GA, Gill R, Smith DF, Chadwick DW. Failure to recognize psychogenic nonepileptic seizures may cause death. Neurology 2004;62:834–5. [5] Dworetzky BA, Bubrick EJ, Szaflarski JP. Nonepileptic psychogenic status: markedly prolonged psychogenic nonepileptic seizures. Epilepsy Behav 2010;19:65–8. [6] Seneviratne U, Rajendran D, Brusco M, Phan TG. How good are we at diagnosing seizures based on semiology? Epilepsia 2012;53:e63–6. [7] Erba G, Giussani G, Juersivich A, Magaudda A, Chiesa V, Lagana A, et al. The semiology of psychogenic nonepileptic seizures revisited: can video alone predict the diagnosis? Preliminary data from a prospective feasibility study. Epilepsia 2016; 57:777–85. [8] Azar NJ, Tayah TF, Wang L, Song Y, Abou-Khalil BW. Postictal breathing pattern distinguishes epileptic from nonepileptic convulsive seizures. Epilepsia 2008;49:132–7. [9] Brown MC, Levin BE, Ramsay RE, Katz DA, Duchowny MS. Characteristics of patients with nonepileptic seizures. J Epilepsy 1991;4:225–9.
U. Seneviratne et al. / Epilepsy & Behavior 66 (2017) 127–131 [10] 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;39:175–82. [11] Jedrzejczak J, Owczarek K, Majkowski J. Psychogenic pseudoepileptic seizures: clinical and electroencephalogram (EEG) video-tape recordings. Eur J Neurol 1999;6:473–9. [12] Kanner AM, Morris HH, Luders H, Dinner DS, Wyllie E, Medendorp SV, et al. Supplementary motor seizures mimicking pseudoseizures: some clinical differences. Neurology 1990;40:1404–7. [13] 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;30:513–23. [14] Saygi S, Katz A, Marks DA, Spencer SS. Frontal lobe partial seizures and psychogenic seizures: comparison of clinical and ictal characteristics. Neurology 1992;42:1274–7. [15] Syed TU, LaFrance Jr WC, Kahriman ES, Hasan SN, Rajasekaran V, Gulati D, et al. Can semiology predict psychogenic nonepileptic seizures? A prospective study. Ann Neurol 2011;69:997–1004. [16] Seneviratne U, Reutens D, D'Souza W. Stereotypy of psychogenic nonepileptic seizures: insights from video-EEG monitoring. Epilepsia 2010;51:1159–68. [17] Mallett S, Halligan S, Thompson M, Collins GS, Altman DG. Interpreting diagnostic accuracy studies for patient care. BMJ 2012;345, e3999. [18] Hosmer D, Lemeshow S. Applied logistic regression. 2nd edition. New Jersey: John Wiley & Sons; 2000. [19] Youden WJ. Index for rating diagnostic tests. Cancer 1950;3:32–5. [20] Dickinson P, Looper KJ. Psychogenic nonepileptic seizures: a current overview. Epilepsia 2012;53:1679–89.
131
[21] 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;81:719–25. [22] Trinka E, Cock H, Hesdorffer D, Rossetti AO, Scheffer IE, Shinnar S, et al. A definition and classification of status epilepticus-report of the ILAE Task Force on Classification of Status Epilepticus. Epilepsia 2015;56:1515–23. [23] Walker MC, Howard RS, Smith SJ, Miller DH, Shorvon SD, Hirsch NP. Diagnosis and treatment of status epilepticus on a neurological intensive care unit. QJM 1996;89: 913–20. [24] Reuber M, Pukrop R, Mitchell AJ, Bauer J, Elger CE. Clinical significance of recurrent psychogenic nonepileptic seizure status. J Neurol 2003;250:1355–62. [25] Holtkamp M, Othman J, Buchheim K, Meierkord H. Diagnosis of psychogenic nonepileptic status epilepticus in the emergency setting. Neurology 2006;66: 1727–9. [26] Seneviratne U, Ma H, Phan TG. How do doctors in training react to seizures? Epilepsy Behav 2016;54:104–9. [27] Samuel M, Duncan JS. Use of the hand held video camcorder in the evaluation of seizures. J Neurol Neurosurg Psychiatry 1994;57:1417–8. [28] Gates JR, Ramani V, Whalen S, Loewenson R. Ictal characteristics of pseudoseizures. Arch Neurol 1985;42:1183–7. [29] LaFrance Jr WC, Baker GA, Duncan R, Goldstein LH, Reuber M. Minimum requirements for the diagnosis of psychogenic nonepileptic seizures: a staged approach: a report from the International League Against Epilepsy Nonepileptic Seizures Task Force. Epilepsia 2013;54:2005–18.