FULL-LENGTH ORIGINAL RESEARCH
Diagnostic practice of psychogenic nonepileptic seizures (PNES) in the pediatric setting *Bianca T. Wichaidit, †John R. Østergaard, and *‡Charlotte U. Rask Epilepsia, **(*):1–8, 2014 doi: 10.1111/epi.12881
SUMMARY
Bianca T. Wichaidit is a postgraduate student at the Medical School of Aarhus University, Denmark.
Objective: No formal guidelines for diagnosing psychogenic nonepileptic seizures (PNES) in children exist, and little is known about the clinical practice of diagnosing PNES in the pediatric setting. We therefore performed a national survey as a first step to document pediatricians’ current diagnostic practice for PNES. Methods: A questionnaire was distributed to all pediatricians (n = 64) working in the field of neuropediatrics and/or social pediatrics in the Danish hospital setting to uncover their use of terminology and of the International Classification of Diseases, 10th Revision (ICD-10) codes as well as their clinical diagnostic approach to pediatric PNES. The questionnaire included questions on 18 history and 24 paroxysmal event characteristics. Results: The response rate was 95% (61/64). There was no consensus on which terminology and diagnostic codes to use. Five history characteristics (psychosocial stressors/ trauma, sexual abuse, paroxysmal events typically occur in stressful situations, no effect of antiepileptic drugs, and physical abuse) and six paroxysmal event characteristics (resisted eyelid opening, avoidance/guarding behavior, paroxysmal events occur in the presence of others, closed eyes, rarely injury related to paroxysmal event, and absence of postictal change) were agreed to be very predictive of PNES by at least 50% of the pediatricians. Supplementary diagnostic tests such as blood chemistry measurements (e.g., blood glucose or acute phase reactants; i.e., white blood cell count and Creactive protein) and electrocardiography were inconsistently used. Only 49% of the respondents reported to use video-electroencephalography (VEEG) frequently as part of their diagnostic procedure. Significance: To our knowledge, this is the first national survey that offers a systematic insight into the diagnostic practices for children with PNES in the hospital setting. The results demonstrate a need for clinical guidelines to improve and systematize the diagnostic approach for PNES in children. KEY WORDS: Children, Psychogenic nonepileptic seizures, Diagnosis, Survey.
Children can present with paroxysmal events characterized by observable abrupt and involuntary changes in consciousness or behavior that may resemble those seen in
epilepsy, but which are without associated electrophysiologic changes as measured by electroencephalography (EEG). These events are often referred to as psychogenic nonepileptic seizures (PNES), functional seizures, or pseudoseizures, and have been shown to be associated with underlying psychosocial stressors such as school-related difficulties and family/interpersonal conflicts, and in rarer cases more severe trauma such as sexual abuse.1 Although deemed as a potentially problematic term,2,3 “PNES” is used in the present paper, as it is still the most widely used term in the pediatric literature. Pediatric PNES is often confused with epilepsy, and misdiagnosis and diagnostic delay for several years are
Accepted October 29, 2014. *The Research Clinic for Functional Disorders and Psychosomatics, Aarhus University Hospital, Aarhus, Denmark; †The Department of Pediatrics, Aarhus University Hospital, Aarhus, Denmark; and ‡Child and Adolescent Psychiatric Center, Aarhus University Hospital, Aarhus, Denmark Address correspondence to Bianca Taaning Wichaidit, Peder Skrams Gade 31, 1. tv., 8200 Aarhus N, Denmark. E-mail: biancawichaidit@gmail. com Wiley Periodicals, Inc. © 2014 International League Against Epilepsy
1
2 B. T. Wichaidit et al. commonly reported problems.4,5 Misdiagnosis leads to unnecessary investigations, invasive procedures, and improper treatment with antiepileptic drugs (AEDs) that may have harmful side effects, as well as considerable emotional and/ or financial costs for the child, the family, and the health care system.1 A narrow focus on biomedical investigations and treatment in the diagnostic workup of PNES furthermore hinders a biopsychosocial approach that might prove more helpful. The literature advocates a multifaceted diagnostic approach including a joint assessment of history, semiology of the event, video-EEG (VEEG), and a psychiatric assessment.5–8 VEEG includes video recording of a habitual event and concomitant ictal EEG, and is considered to be the gold standard for diagnosing PNES.5–7 Thus VEEG without abnormal EEG correlates is highly suggestive of a diagnosis of PNES in the presence of confirmatory psychiatric diagnosis of conversion disorder. The diagnostic process may also include supplementary diagnostic tests such as magnetic resonance imaging (MRI), electrocardiography (ECG), and blood tests to rule out other differential diagnoses (e.g., brain tumors, heart disease, and endocrinologic conditions such as hypoglycemia).5 Currently there is a paucity of research into the extent to which the recommended approach to diagnosing PNES is actually applied. Questionnaire surveys have been conducted among clinicians working primarily in the adult setting, but they have focused mainly on treatment aspects rather than the diagnostic process.9,10 The aim of the present study was therefore to conduct the first systematic national survey to quantify and describe the diagnostic approaches used by clinicians in the pediatric hospital setting. Although this was primarily set out to be an exploratory study, we a priori expected to find a certain variation in diagnostic approach and that the use of VEEG and supplementary diagnostic tests could be influenced by the clinician’s degree of clinical experience and specialization.
Methods Study design and study population The study was conducted as a questionnaire survey with a cross-sectional design. The target population was all Danish pediatricians working in the field of neuropediatrics and/or social pediatrics, that is, the subgroup of clinicians who diagnose and manage the majority of children with PNES in Denmark (DK). To reach all pediatricians caring for this patient group, the chief consultants of the pediatric departments across DK were asked to identify the pediatricians in charge of PNES patients. In total, 64 eligible pediatricians were identified and invited to participate in the survey. Questionnaire design A literature review revealed no existing questionnaire suitable for the present survey. We therefore designed a Danish-language questionnaire de novo. The full questionEpilepsia, **(*):1–8, 2014 doi: 10.1111/epi.12881
naire can be seen online in the Supporting Information Data S1. Some items were inspired by existing questionnaires addressing the management of adult PNES and related pediatric conditions (functional disorders/somatization, conversion disorder),10–15 but the majority of items were developed de novo based on a systematic review of the pediatric literature including reported history and semiology characteristics of PNES. To evaluate the clinical relevance of the overall content, especially regarding items on history characteristics and event characteristics, and the flow and format of questions, a panel of three Danish experts (one in child and adolescent psychiatry, one in neuropediatrics, and one in adult psychiatry with expertise in functional disorders) reviewed the questionnaire. Subsequently, two English experts in child and adolescent psychiatry reviewed a preliminary draft version translated into English (for details, see Acknowledgments). The expert panel pointed out questions that overlapped or lacked clinical relevance in a pediatric population (e.g., items regarding induction methods such as hypnosis or placebo saline infusion, which are not commonly used in pediatric settings). These questions were subsequently modified or excluded with guidance by suggestions from the expert panel and after thorough discussions in the study group. After these modifications, a preliminary pilot version in paper format was distributed to 25 residents in pediatrics. Based on these pilot results, the last modifications were made. Further details of the questionnaire design are shown in Figure S1. The final questionnaire was implemented as both a paper and electronic version using the online software SurveyXact (www.surveyxact.com) and consisted of 98 items designed to cover the following domains: (1) Demographic data about respondents; (2) use of terminology; (3) use of the International Classification of Diseases 10th Revision (ICD-10) codes; (4) diagnostic approach, for example, emphasis on history characteristics and event semiology and use of supplementary diagnostic tests such as VEEG, magnetic resonance imaging (MRI) and blood tests in the diagnostic process; (5) treatment approach, for example, use of different treatment modalities such as psychoeducation and psychological therapy, and referral to other services such as child and adolescent psychiatrists; and (6) the respondents’ “Attitudes and beliefs” about PNES in general, for example, questions about whether the respondents believe that children with PNES are faking, whether the available treatment options are sufficient, and whether the physicians worry about overlooking organic disease. Response format was “check what applies” format for lists and categorical (5-point response scale) for closed-ended questions. Thus, two different 5-point response scales addressed the diagnostic approaches: scale 1, which consisted of “always,” “often,” “sometimes,” “rarely,” and “never”; and scale 2, which consisted of “a whole lot,” “a lot,” “some,” “a little,” and “not at all.” One open-ended
3 Diagnostic Practice of PNES in Children question addressed the terminology used by the respondents. The full questionnaire can be seen online, Supporting Information. Data collection Approval from the Danish Data Protection Agency was obtained before initiation of data collection (journal number: 1-16-02-19-13), and the study was presented to The Central Denmark Region Committees on Health Research Ethics. Invitations to participate were sent by letter. The invitation letter contained project information, the questionnaire, a return envelope with paid postage, and information on how to complete the survey electronically using the Internet (using SurveyXact). Reminders were sent by e-mail and by letter to nonresponders up to a maximum of three times. Data preparation Respondents who chose to reply electronically entered their responses directly into the SurveyXact database. Responses received in hard copy were entered manually into the same database by the first author (BTW). The first author furthermore went through all received questionnaires, and in case of unclear or missing responses, the respondent in question was approached by e-mail and asked to clarify or answer missing items. A few uncertain responses could not be determined and were therefore considered missing in the final analyses. The full data set was exported from SurveyXact to STATA (www.stata.com) as a CSV file, i.e., in a delimited data format (Microsoft Excel, Redmond, WA, U.S.A.). Analyses For the present study, only data from domains 1 to 4 in the questionnaire were included in the analyses, whereas data related to domains 5 and 6 will be presented elsewhere. Frequencies and percentages are reported for categorical variables and mean and standard deviation (SD) for continuous variables. To assess whether subspecialization or the level of experience influenced the use of VEEG, we performed various subanalyses, comparing the proportions of respondents that answered “often or always” from different subgroups (for details see Table 3). Similar subanalyses were done to compare the same subgroups of respondents that reported to use the specific ICD-10 code for conversion disorder (F44.5). Using a Wald test (z-test), we compared the subgroups with different subspecializations: (1) respondents working in the field of social pediatrics versus respondents working in the field of neuropediatrics, (2) respondents working in the field of social pediatrics versus those working in both the neuropediatric and social pediatric fields, and (3) respondents working in the field of neuropediatrics versus those working in both the neuropediatric and social pediatric fields. We used chi-square tests to examine the effect of clinical experi-
ence on diagnostic practices by comparing (1) clinically experienced physicians who had seen >15 children with PNES during their career with less experienced clinicians who had seen ≤15 children with PNES during their career; (2) consultants (experienced specialized doctors) versus nonconsultants (residents or staff doctors); (3) pediatricians with and without expert neuropediatric training; and (4) those working at a secondary versus tertiary hospital. Based on clinical reasoning, the responses from the 5-point scales addressing the items on diagnostic approaches were categorized into two groups for both the comparative and descriptive analyses: scale 1: “often/always” versus “sometimes/ rarely/never,” and scale 2: “a lot/a whole lot” versus “some/ a little,” and “not at all.” In addition, the use of other supplementary diagnostic tests was compared between the above-mentioned subgroups. We dichotomized the responses to each of the six items concerning use of ECG, MRI, computed tomography (CT) of the cerebrum, blood glucose test, serum prolactin level test, and urine screening for inborn errors of metabolism, meaning that the answer “never,” “rarely,” or “sometimes” would give a score of zero, and the answer “often” or “always” would give a score of one. We then added up the scores for each respondent giving a total score between zero (if the respondent had answered “never/rarely/sometimes” to all the items) and six (if the respondent had answered “often/always” to all the items). This total score variable was then compared between the subgroups by the means of Mann-Whitney tests to assess whether one or more subgroup would be more likely to perform supplementary diagnostic tests than others. All analyses were performed using STATA version 12 for Windows. The statistical tests presented are two-tailed, and a significance level of p < 0.05 was used with Bonferroni correction for multiple comparisons.
Results In total, a response rate of 95% (61/64) was obtained. The questionnaire was 100% completed by 95% (58/61) of the respondents; the remaining three completed at least 80%. Table 1 shows descriptive data on the respondents. Use of terminology and diagnostic codes When communicating with children with events not related to epilepsy and their families, the terms most frequently stated to be the most appropriate to use were functional seizures (34%) and PNES (25%). Overall, 17 different diagnostic ICD-10 codes were used for pediatric PNES (Fig. 1). In total, 69% stated that they used more than one ICD-10 code, and two respondents stated that they used five different codes. The two most widely used codes were “F 91.8” (Other conduct disorders) and “R 56.8” (Other and unspecified convulsions), both of which were used by 54% of the respondents. The specific psychiatEpilepsia, **(*):1–8, 2014 doi: 10.1111/epi.12881
4 B. T. Wichaidit et al. Table 1. Demographic characteristics of the whole study sample (n = 61) based on items in domain 1 in the questionnaire Demographic data Mean age, years (SD) Sex, N (%) Male Female Hospital, N (%) Secondary level Tertiary level Othera Position, N (%): Pediatric resident Pediatric staff doctor Pediatric consultant Respondents working in the, N (%):b Neuropediatric field alone Social pediatric field alone Both fields Full-time occupation in one or both fields, N (%):b Yes No Number of children with PNES seen during career, N (%) 1–5 6–15 16–30 >30
Study sample (n = 61) 52 (8.68) 21 (34) 40 (66) 28 (46) 29 (47) 4 (7) 3 (5) 13 (21) 45 (74) 41 (68) 10 (17) 9 (15)
47 (78) 13 (22)
3 (5) 13 (21) 17 (28) 28 (46)
a Other: Highly specialized epilepsy hospital (Filadelfia) (n = 2), “emergency hospital” (n = 1), “combined secondary- and tertiary level” (n = 1). b One respondent missing for this category.
ric diagnosis for a conversion disorder in ICD-10, that is, F44.5, was used by 31%, and its use was not statistically significantly related to length of experience and/or type of subspecialty when the analyses were adjusted for multiple comparisons (p > 0.01 for all subanalyses). We used a conservative two-tailed Bonferroni criterion of p < 0.007 to adjust for seven multiple comparisons. Diagnostic setting Most of the respondents (76%) stated that the diagnostic assessment of a child suspected of having PNES took place both in the inpatient and outpatient setting as opposed to only in the outpatient or only in the inpatient setting. A pediatric emergency room visit without further follow-up was reported “never” to happen by 70%, “rarely” by 25%, and “sometimes” by 5%. History and semiology of the event The respondents were asked to indicate to which degree they judged each of the 18 history characteristics and 24 seizure-like characteristics commonly encountered in the literature to support a diagnosis of PNES. The five history characteristics that the respondents emphasized the most (>50% ticked the option indicating that this particular Epilepsia, **(*):1–8, 2014 doi: 10.1111/epi.12881
Figure 1. Responses to the following item in the questionnaire (domain 3): “Which of the following diagnostic ICD-10 codes do you use?” F 91.8, Other conduct disorders; F 44.5, Dissociative convulsions; F 98.9, Unspecified behavioral and emotional disorders with onset usually occurring in childhood and adolescence; Other* F 43.8, Other reactions to severe stress; F 43.9, Reaction to severe stress unspecified; F 44.4, Dissociative motor disorders; F 45.0, Somatization disorder; F 45.9, Somatoform disorder unspecified; R 56.8, Other and unspecified convulsions; R 25.2, Cramp and spasm; R 55.9, Syncope and collapse; R 25.8, Other an unspecified abnormal involuntary movements; R 29.8, Other and unspecified symptoms and signs involving the nervous and musculoskeletal system; Z 03.3, Observation for suspected nervous system disorder; Z 03.2, Observation for suspected mental and behavioral disorders; Other** Z 03.8, Observation for other suspected diseases and conditions; G 40.9, Epilepsy unspecified. Epilepsia ILAE
characteristic supported PNES “a lot” or “a whole lot”) were the following: “psychosocial stressors/trauma,” “sexual abuse,” “the paroxysmal events typically occur in stressful situations,” “no effect of antiepileptic drugs (AEDs),” and “physical abuse” (Fig. 2). The six event characteristics that the respondents emphasized the most were as follows: “resisted eyelid opening,” “avoidance/guarding behavior,” “the paroxysmal events occur in the presence of others,” “closed eyes,” “rarely injury related to the paroxysmal events,” and “absence of postictal change” (Fig. 3). Only two of the event characteristics (i.e., resisting eyelid opening and avoidance/guarding behavior during the paroxysmal event) and none of the history characteristics were judged to be predictive of PNES by a substantial part of the respondents (70% or more). Use of EEG and supplementary diagnostic tests Routine EEG and 24-h EEG recordings were stated by the respondents to be performed “often” or “always” by 57/ 61 (93%) and 18/61 (30%), respectively, and VEEG “often” or “always” by 30/61 (49%) (Table 2).
5 Diagnostic Practice of PNES in Children
Figure 2. Percentage of respondents who answered “a lot” or “a whole lot” to the following item in the questionnaire (domain 4): “How much do you think the following history characteristics suggest a diagnosis of functional seizures?” (The item wording has been abbreviated for conciseness): 1, Psychosocial stressors; 2, Sexual abuse; 3, Occur in stressful situations; 4, No effect of antiepileptic drugs; 5, Physical abuse; 6, Functional symptoms in family; 7, Other functional symptoms; 8, Psychiatric comorbidity; 9, Functional symptoms among acquaintances; 10, Frequent health care contacts; 11, Rarely nocturnal events; 12, Several event types (no stereotypy); 13, Epilepsy in the family; 14, Epilepsy among acquaintances; 15, High event frequency; 16, Developmental disorders; 17, Somatic comorbidity; 18, Prior head trauma. Epilepsia ILAE
The majority of the respondents replied that they would “often or always” perform ECG (64%) and blood glucose test (66%), and that they would “rarely, never, or only sometimes” perform MRI (70%), CT of the cerebrum (98%), blood tests for renal function (56%), acute phase reactants (white blood cell count, C-reactive protein) (72%), hematology (62%), liver enzyme count (67%), serum prolactin level test (92%), or urine screening for inborn errors of metabolism (98%). Comparative analyses showed that a statistically significantly larger proportion of consultants and respondents working at a tertiary hospital “often” or “always” performed VEEG, and that respondents working in the social pediatric field compared to respondents working in the neuropediatric field performed significantly more supplementary diagnostic tests (Table 3).
Discussion This national survey demonstrates disagreement among pediatricians on use of terminology and diagnostic approach for pediatric PNES. The findings correspond to results reported in similar studies on adult PNES10 and pediatric benign chronic pain, another condition dominated by medically unexplained or functional symptoms.16 Several studies, including a survey of American parents of children with PNES,2 have addressed patients’ and their relatives’ views on terms used for various medically unexplained symptoms.2,3,17,18 It was found that “functional” is perceived as one of the least offensive or pejorative terms,
Figure 3. Percentage of respondents who answered “a lot” or “a whole lot” to the following item in the questionnaire (domain 4): “How much do you think the following characteristics of a paroxysmal event suggest a diagnosis of functional seizures?” (The item wording has been abbreviated for conciseness): 1, Resist eyelid opening; 2, Avoidance/guarding behavior; 3, Occurrence in the presence of others; 4, Closed eyes; 5, Rarely injury; 6, No postictal change; 7, Intelligible speech; 8, Pelvic thrusting; 9, Long duration; 10, Response when addressed; 11, Side-to-side head movements; 12, Fluctuating pattern during episode; 13, Asymmetrical movements; 14, Fluctuating intensity; 15, Pupil light reflex intact; 16, Preserved consciousness; 17, Crying in the middle or end of the event; 18, No tongue bite; 19, Gradual onset with slow escalation in intensity; 20, No urinary incontinence; 21, No cyanosis; 22, No fecal incontinence; 23, Normal plantar response postictal; 24, Rapid, shallow, irregular, quiet breathing pattern postictal. Epilepsia ILAE
Table 2. Distribution of responses (n = 61) to the following question: “How often do you perform the following EEG investigations on a child suspected of having PNES?” (domain 4 in the questionnaire) Standard EEG, N (%) 24-h EEG, N (%) VEEG, N (%)
Always
Often
Sometimes
Rarely
Never
32 (52) 2 (3) 10 (16)
25 (41) 16 (26) 20 (33)
3 (5) 26 (43) 21 (34)
1 (2) 11 (18) 6 (10)
0 6 (10) 4 (7)
whereas “psychogenic” is perceived as one of the most offensive ones. Thus, because the choice of terminology is important to patients and their families, the conflicting responses from our survey point toward a need for agreement on which terms to use in the clinical setting. Considerable inconsistencies in coding practices for PNES were also reported, which could mirror different conceptualizations of the disorder. Some pediatricians may for example perceive PNES as a syncope-like disorder, whereas other pediatricians may perceive it as a convulsive disorder and use diagnostic codes accordingly (i.e., “R 55.9 Syncope and collapse” and “R 56.8 other and unspecified convulsions,” respectively). These potential inconsistencies in conceptualization may aggravate the commonly described confusion about the diagnosis among patients and their families. Future studies should therefore explore this subject further and whether the confusion could be based on a lack of training of pediatricians on how to assess conversion disorder in children. The inconsistencies may also constitute Epilepsia, **(*):1–8, 2014 doi: 10.1111/epi.12881
6 B. T. Wichaidit et al. Table 3. Comparison of subgroups with different degrees of clinical experience and specialization with regard to frequent use of VEEG and number of supplementary diagnostic tests (i.e., MRI, CT, ECG, blood tests, urine screening for inborn errors of metabolism) in pediatric patients with PNES Performance of VEEG (‘Often’ or ‘Always’)
Number of supplementary tests
Proportion (95% CI)
RR (95% CI)
Test statistics v2(1), p-value
Consultants (n = 45) Nonconsultants (n = 16)
0.6 (0.4–0.7) 0.2 (0.0–0.4)
3.2 (1.1–9.1)
8.04, 0.005*
2.0 (1.0–4.5) 4.0 (0.3–6.0)
0.46, 0.65
61
Neuropediatric expert training (n = 26) No neuropediatric expert training(=34)
0.7 (0.4–0.8) 0.4 (0.2–0.6)
1.7 (1.0–2.8)
4.34, 0.04
2.0 (0.0–4.0) 3.0 (1.8–6.3)
2.49, 0.01
60a
Working at a tertiary hospital (n = 29) Working at a secondary hospital (n = 28)
0.7 (0.5–0.8) 0.3 (0.1–0.5)
2.3 (1.2–4.4)
7.80, 0.005*
3.0 (1.0–6.0) 2.0 (1.0–6.0)
0.17, 0.87
57b
0.5 (0.4–0.7) 0.4 (0.2–0.6) Proportion (95% CI)
1.4 (0.7–2.8)
1.18, 0.28
3.0 (1.0–5.5) 2.5 (1.3–6.0)
0.37, 0.72
61
Subgroups
Seen >15 children with PNES (n = 45) Seen ≤15 children with PNES (n = 16)
Respondents working in thec Neuropediatric field (n = 41) Social pediatric field (n = 10)
RR (95% CI)
Median (25/75 percentiles)
Test statistics MW: Z, p-value
No. of obs
Test statistics BR: Z, p-value
Test statistics MW: Z, p-value
0.6 (0.5–0.8)a 0.2 (0.1–0.7)b
3.0 (0.9–10.8)
1.7, 0.08
2.0 (0.0–4.5)a 6.0 (3.8–7.3)b
3.03, 0.003*
51
Both fields (n = 9) Social pediatric field
0.3 (0.1–0.8)c 0.2 (0.1–0.7)b
1.6 (0.4–7.8)
0.7, 0.52
2.0 (1.0–2.5)c 6.0 (3.8–7.3)b
0.26, 0.010
19
Neuropediatric field Both fields
0.6 (0.5–0.8)a 0.3 (0.1–0.8)c
1.8 (0.7–4.8)
1.2, 0.22
2.0 (0.0–4.5)a 2.0 (1.0–2.5)c
0.26, 0.80
50
RR, relative risk; BR, binomial regression, MW, Mann-Whitney. a One missing on this item. b Four responses could not be categorized as either working in secondary or tertiary hospital and was therefore considered as missing. c One respondent could not be categorized as working in either the neuropediatric or social pediatric field, or in both fields (n = 60 in all these subanalyses). * Statistically significant using a conservative two-tailed Bonferroni criterion of p < 0.007 to adjust for seven multiple comparisons.
a major obstacle for inter- and intradisciplinary communication as well as for performing systematic research in this area. In addition, it is of interest that one of the two most frequently reported ICD-10 codes was the psychiatric code “F 91.8” (Other conduct disorders), even though the most correct psychiatric diagnosis would be conversion disorder (i.e., in the context of ICD-10, “F 44.5”). The Danish Pediatric Society recommended at one point that “F 91.8” should be the diagnostic code for children with PNES and added the term “pseudoseizure” to the code. This probably explains why so many Danish pediatricians in the present study reported to use this specific code for PNES. The use of “F 44.5” was not found to be related to length of experience and/or type of subspecialty, and the inappropriate use of psychiatric ICD-10 codes may therefore reflect that, in general, confusion arises when psychiatric codes are being applied in a general hospital setting. The diagnostic workup of children with PNES was mainly reported to take place in pediatric departments, and no patients were seen solely at pediatric emergency rooms without further follow-up, which indicates that diagnosing PNES often requires a longer period of assessment.6 Overall, there seemed to be a lack of agreement concerning which history and event characteristics could be suggestive of a PNES diagnosis, that is, to which degree the Epilepsia, **(*):1–8, 2014 doi: 10.1111/epi.12881
clinicians thought that the listed characteristics supported a diagnosis of PNES. This may reflect that no single sign can be considered pathognomonic for the diagnosis,19,20 and that clusters of event characteristics may be more useful in the differentiation between PNES and epileptic seizures.21 An additional explanation, as suggested by Alessi et al.,22 could be lack of awareness among pediatricians that the classic signs observed in adults with PNES are not common in children. An overall variation in the use of the gold standard investigation VEEG1,7,8,19,23 was documented. The higher use of VEEG by consultants and doctors working in a tertiary hospital suggests that pediatricians with longer clinical experience are more aware of the importance of VEEG, but also that there may be potential differences in VEEG accessibility in the secondary and tertiary settings. One could also speculate whether respondents who stated that they performed VEEG less frequently may be of the opinion that it is possible to reach a correct diagnosis of PNES without VEEG. However, interictal EEG has a number of limitations, and eyewitness descriptions alone can be misleading and unreliable.6,7,19,20 The findings therefore suggest a need for increased awareness of the importance of VEEG among clinicians, but also that the test needs to be made more widely available in the clinical setting.
7 Diagnostic Practice of PNES in Children Determination of the serum prolactin level is not recommended for diagnostic purposes, since it is not always elevated following epileptic seizures and may be affected by other medical conditions or medications.23 Consistent with this, serum prolactin levels were rarely determined by the respondents. However, other supplementary diagnostic tests showed to be used in an inconsistent manner, and respondents working in the social pediatric field were significantly more likely to perform additional tests compared to respondents working in the neuropediatric field. Again, this seems to support that lack of clinical experience and/or expertise will influence the individual pediatrician’s diagnostic workup. Altogether, this calls for consensus and guidelines with regard to the use of supplementary tests as well. To the best of our knowledge, the present study is the first attempt to survey a representative population of clinicians’ diagnostic practices for pediatric PNES in the hospital setting. The main strengths of the study are the very high participation rate, that the majority of the respondents were specialists in pediatrics, and that they all had experience with PNES. Because eligible respondents were identified by the chief consultants of all pediatric departments in DK, we ensured that all or close to all doctors working within this field in the hospital setting were approached. However, some limitations of the study should be considered. First, given that this is a national survey from DK, the results may not be generalizable to clinical settings in other countries. Nevertheless, the results correspond well with those of the few other studies concerning similar topics that have been performed in other European countries.10,16 To our knowledge, children with PNES in other countries are also mainly diagnosed by doctors in the field of neuropediatrics, as found in our study.1,5,7 It therefore seems likely that the overall clinical picture of this survey could be the same elsewhere. Second, the questionnaire could have been further validated. Yet, the pilot testing confirmed the content validity of the questionnaire. The high response rate on all the items indicates that it was easily understood, and to address the constructs pertinent to PNES, the questionnaire was designed by review of the literature and evaluated by experts in the field. Although it might have been more appropriate to pilot test the questionnaire on specialists, we chose to use trainees in order not to reduce the number of eligible participants for the actual survey. Third, it should be mentioned that the terms “functional seizures” and “PNES” were used in the questionnaire, and it therefore cannot be ruled out that some respondents were biased to pick one of those as their preferred term. Finally, the comparative analyses were based on small numbers, reducing the power and performance of more sophisticated analyses (e.g., such as the effect of being both a consultant and working at a tertiary hospital on the use of VEEG and supplementary tests). In addition, there may have been considerable overlap between some of the subgroups we defined so that the
significant findings identified the same respondents. Regardless of these limitations, the results still indicate a lack of overall consensus among pediatricians on the use of VEEG and supplementary tests in this patient group. In conclusion, this survey strongly supports the need for clinical guidelines that should be developed by specialists in the field and designed to create consensus on diagnostic practice. The first step could be to focus on the use of terminology and diagnostic codes. The formulation of a framework for which diagnostic approach to use at this point is also not beyond reach, although larger and more standardized studies are needed to establish evidence-based guidelines. Future studies should include relevant control groups for comparison and use standardized questionnaires to obtain systematic data on history. VEEG recordings should be reviewed by blinded, experienced epileptologists to compare PNES and epilepsy with regard to event semiology and to identify clusters of characteristics that may be particularly predictive for pediatric PNES. This could potentially lead to an enhanced diagnostic accuracy, which ultimately would lead to a better outcome for children with PNES.
Acknowledgments The authors thank Elena Garralda, Professor emeritus, MD, PhD, Academic Unit of Child and Adolescent Psychiatry, Imperial College, London, UK; Cornelius Ani, Consultant and Honorary Senior Lecturer, MD, Academic Unit of Child and Adolescent Psychiatry, Imperial College, London, UK; and Per Fink, Professor, MD, PhD, DMSc, The Research Clinic For Functional Disorders and Psychosomatics, Aarhus University Hospital, Denmark, for their helpful comments on the preliminary questionnaire. We also thank Jens Søndergaard Jensen, MSc, The Research Clinic for Functional Disorders, Aarhus University Hospital, Denmark, for statistical advice. Furthermore, we thank all the pediatricians who participated in the study. “The Danish Council for Independent Research | Medical Sciences,” “Central Denmark Region – the Foundation for Improvement of Psychiatric Research at the Child and Adolescent Psychiatric Center,” “Beatrice Surovell Haskell Fund for Child Mental Health Research of Copenhagen,” and “The Møller Foundation – A.P. Møller and wife Chastine McKinneyMøller” funded the project.
Disclosure None of the authors have any conflict of interest to disclose. We confirm that we have read the journal’s position on issues involved in ethical publication and affirm that this report is consistent with those guidelines.
References 1. Reilly C, Menlove L, Fenton V, et al. Psychogenic nonepileptic seizures in children: a review. Epilepsia 2013;54:1715–1724. 2. Morgan LA, Dvorchik I, Williams KL, et al. Parental ranking of terms describing nonepileptic events. Pediatr Neurol 2013;48:378–382. 3. Stone J, Campbell K, Sharma N, et al. What should we call pseudoseizures? The patient’s perspective. Seizure 2003;12:568–572. 4. Dhiman V, Sinha S, Rawat VS, et al. Children with psychogenic nonepileptic seizures (PNES): a detailed semiologic analysis and modified new classification. Brain Dev 2013;36:287–293. 5. Plioplys S, Asato MR, Bursch B, et al. Multidisciplinary management of pediatric nonepileptic seizures. J Am Acad Child Adolesc Psychiatry 2007;46:1491–1495. Epilepsia, **(*):1–8, 2014 doi: 10.1111/epi.12881
8 B. T. Wichaidit et al. 6. Lafrance WC Jr, Baker GA, Duncan R, et al. 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– 2018. 7. Patel H, Dunn DW, Austin JK, et al. Psychogenic nonepileptic seizures (pseudoseizures). Pediatr Rev 2011;32:e66–e72. 8. Wood BL, Haque S, Weinstock A, et al. Pediatric stress-related seizures: conceptualization, evaluation, and treatment of nonepileptic seizures in children and adolescents. Curr Opin Pediatr 2004;16:523– 531. 9. LaFrance WC Jr, Rusch MD, Machan JT. What is “treatment as usual” for nonepileptic seizures? Epilepsy Behav 2008;12:388–394. 10. Mayor R, Smith PE, Reuber M. Management of patients with nonepileptic attack disorder in the United Kingdom: a survey of health care professionals. Epilepsy Behav 2011;21:402–406. 11. Sahaya K, Dholakia SA, Lardizabal D, et al. Opinion survey of health care providers towards psychogenic non epileptic seizures. Clin Neurol Neurosurg 2012;114:1304–1307. 12. Shneker BF, Elliott JO. Primary care and emergency physician attitudes and beliefs related to patients with psychogenic nonepileptic spells. Epilepsy Behav 2008;13:243–247. 13. Ani C, Reading R, Lynn R, et al. Incidence and 12-month outcome of non-transient childhood conversion disorder in the U.K. and Ireland. Br J Psychiatry 2013;202:413–418. 14. Whitehead K, Kandler R, Reuber M. Patients’ and neurologists’ perception of epilepsy and psychogenic nonepileptic seizures. Epilepsia 2013;54:708–717. 15. Furness P, Glazebrook C, Tay J, et al. Medically unexplained physical symptoms in children: exploring hospital staff perceptions. Clin Child Psychol Psychiatry 2009;14:575–587. 16. Konijnenberg AY, De Graeff-Meeder ER, Kimpen JL, et al. Children with unexplained chronic pain: do pediatricians agree regarding the
Epilepsia, **(*):1–8, 2014 doi: 10.1111/epi.12881
17. 18. 19. 20. 21. 22. 23.
diagnostic approach and presumed primary cause? Pediatrics 2004;114:1220–1226. Edwards MJ, Stone J, Lang AE. From psychogenic movement disorder to functional movement disorder: it’s time to change the name. Mov Disord 2013;29:849–852. Stone J, Wojcik W, Durrance D, et al. What should we say to patients with symptoms unexplained by disease? The “number needed to offend”. BMJ 2002;325:1449–1450. Devinsky O, Gazzola D, LaFrance WC Jr. Differentiating between nonepileptic and epileptic seizures. Nat Rev Neurol 2011;7:210–220. Syed TU, LaFrance WC Jr, Kahriman ES, et al. Can semiology predict psychogenic nonepileptic seizures? A prospective study. Ann Neurol 2011;69:997–1004. Hubsch C, Baumann C, Hingray C, et al. Clinical classification of psychogenic non-epileptic seizures based on video-EEG analysis and automatic clustering. J Neurol Neurosurg Psychiatry 2011;82:955–960. Alessi R, Vincentiis S, Rzezak P, et al. Semiology of psychogenic nonepileptic seizures: age-related differences. Epilepsy Behav 2013;27:292–295. Plioplys S, Siddarth P, Asato MR, et al. Clinicians’ views on antiepileptic medication management in nonepileptic seizures. J Child Neurol 2013;29:746–750.
Supporting Information Additional Supporting Information may be found in the online version of this article: Data S1. Questionnaire regarding children with functional seizures/psychogenic nonepileptic seizures. Figure S1. The process of developing the questionnaire.
Diagnostic practice of psychogenic nonepileptic seizures (PNES) in the pediatric setting *Bianca T. Wichaidit, †John R. Østergaard, and *‡Charlotte U. Rask Epilepsia, **(*):1–8, 2014 doi: 10.1111/epi.12881
SUMMARY
Bianca T. Wichaidit is a postgraduate student at the Medical School of Aarhus University, Denmark.
Objective: No formal guidelines for diagnosing psychogenic nonepileptic seizures (PNES) in children exist, and little is known about the clinical practice of diagnosing PNES in the pediatric setting. We therefore performed a national survey as a first step to document pediatricians’ current diagnostic practice for PNES. Methods: A questionnaire was distributed to all pediatricians (n = 64) working in the field of neuropediatrics and/or social pediatrics in the Danish hospital setting to uncover their use of terminology and of the International Classification of Diseases, 10th Revision (ICD-10) codes as well as their clinical diagnostic approach to pediatric PNES. The questionnaire included questions on 18 history and 24 paroxysmal event characteristics. Results: The response rate was 95% (61/64). There was no consensus on which terminology and diagnostic codes to use. Five history characteristics (psychosocial stressors/ trauma, sexual abuse, paroxysmal events typically occur in stressful situations, no effect of antiepileptic drugs, and physical abuse) and six paroxysmal event characteristics (resisted eyelid opening, avoidance/guarding behavior, paroxysmal events occur in the presence of others, closed eyes, rarely injury related to paroxysmal event, and absence of postictal change) were agreed to be very predictive of PNES by at least 50% of the pediatricians. Supplementary diagnostic tests such as blood chemistry measurements (e.g., blood glucose or acute phase reactants; i.e., white blood cell count and Creactive protein) and electrocardiography were inconsistently used. Only 49% of the respondents reported to use video-electroencephalography (VEEG) frequently as part of their diagnostic procedure. Significance: To our knowledge, this is the first national survey that offers a systematic insight into the diagnostic practices for children with PNES in the hospital setting. The results demonstrate a need for clinical guidelines to improve and systematize the diagnostic approach for PNES in children. KEY WORDS: Children, Psychogenic nonepileptic seizures, Diagnosis, Survey.
Children can present with paroxysmal events characterized by observable abrupt and involuntary changes in consciousness or behavior that may resemble those seen in
epilepsy, but which are without associated electrophysiologic changes as measured by electroencephalography (EEG). These events are often referred to as psychogenic nonepileptic seizures (PNES), functional seizures, or pseudoseizures, and have been shown to be associated with underlying psychosocial stressors such as school-related difficulties and family/interpersonal conflicts, and in rarer cases more severe trauma such as sexual abuse.1 Although deemed as a potentially problematic term,2,3 “PNES” is used in the present paper, as it is still the most widely used term in the pediatric literature. Pediatric PNES is often confused with epilepsy, and misdiagnosis and diagnostic delay for several years are
Accepted October 29, 2014. *The Research Clinic for Functional Disorders and Psychosomatics, Aarhus University Hospital, Aarhus, Denmark; †The Department of Pediatrics, Aarhus University Hospital, Aarhus, Denmark; and ‡Child and Adolescent Psychiatric Center, Aarhus University Hospital, Aarhus, Denmark Address correspondence to Bianca Taaning Wichaidit, Peder Skrams Gade 31, 1. tv., 8200 Aarhus N, Denmark. E-mail: biancawichaidit@gmail. com Wiley Periodicals, Inc. © 2014 International League Against Epilepsy
1
2 B. T. Wichaidit et al. commonly reported problems.4,5 Misdiagnosis leads to unnecessary investigations, invasive procedures, and improper treatment with antiepileptic drugs (AEDs) that may have harmful side effects, as well as considerable emotional and/ or financial costs for the child, the family, and the health care system.1 A narrow focus on biomedical investigations and treatment in the diagnostic workup of PNES furthermore hinders a biopsychosocial approach that might prove more helpful. The literature advocates a multifaceted diagnostic approach including a joint assessment of history, semiology of the event, video-EEG (VEEG), and a psychiatric assessment.5–8 VEEG includes video recording of a habitual event and concomitant ictal EEG, and is considered to be the gold standard for diagnosing PNES.5–7 Thus VEEG without abnormal EEG correlates is highly suggestive of a diagnosis of PNES in the presence of confirmatory psychiatric diagnosis of conversion disorder. The diagnostic process may also include supplementary diagnostic tests such as magnetic resonance imaging (MRI), electrocardiography (ECG), and blood tests to rule out other differential diagnoses (e.g., brain tumors, heart disease, and endocrinologic conditions such as hypoglycemia).5 Currently there is a paucity of research into the extent to which the recommended approach to diagnosing PNES is actually applied. Questionnaire surveys have been conducted among clinicians working primarily in the adult setting, but they have focused mainly on treatment aspects rather than the diagnostic process.9,10 The aim of the present study was therefore to conduct the first systematic national survey to quantify and describe the diagnostic approaches used by clinicians in the pediatric hospital setting. Although this was primarily set out to be an exploratory study, we a priori expected to find a certain variation in diagnostic approach and that the use of VEEG and supplementary diagnostic tests could be influenced by the clinician’s degree of clinical experience and specialization.
Methods Study design and study population The study was conducted as a questionnaire survey with a cross-sectional design. The target population was all Danish pediatricians working in the field of neuropediatrics and/or social pediatrics, that is, the subgroup of clinicians who diagnose and manage the majority of children with PNES in Denmark (DK). To reach all pediatricians caring for this patient group, the chief consultants of the pediatric departments across DK were asked to identify the pediatricians in charge of PNES patients. In total, 64 eligible pediatricians were identified and invited to participate in the survey. Questionnaire design A literature review revealed no existing questionnaire suitable for the present survey. We therefore designed a Danish-language questionnaire de novo. The full questionEpilepsia, **(*):1–8, 2014 doi: 10.1111/epi.12881
naire can be seen online in the Supporting Information Data S1. Some items were inspired by existing questionnaires addressing the management of adult PNES and related pediatric conditions (functional disorders/somatization, conversion disorder),10–15 but the majority of items were developed de novo based on a systematic review of the pediatric literature including reported history and semiology characteristics of PNES. To evaluate the clinical relevance of the overall content, especially regarding items on history characteristics and event characteristics, and the flow and format of questions, a panel of three Danish experts (one in child and adolescent psychiatry, one in neuropediatrics, and one in adult psychiatry with expertise in functional disorders) reviewed the questionnaire. Subsequently, two English experts in child and adolescent psychiatry reviewed a preliminary draft version translated into English (for details, see Acknowledgments). The expert panel pointed out questions that overlapped or lacked clinical relevance in a pediatric population (e.g., items regarding induction methods such as hypnosis or placebo saline infusion, which are not commonly used in pediatric settings). These questions were subsequently modified or excluded with guidance by suggestions from the expert panel and after thorough discussions in the study group. After these modifications, a preliminary pilot version in paper format was distributed to 25 residents in pediatrics. Based on these pilot results, the last modifications were made. Further details of the questionnaire design are shown in Figure S1. The final questionnaire was implemented as both a paper and electronic version using the online software SurveyXact (www.surveyxact.com) and consisted of 98 items designed to cover the following domains: (1) Demographic data about respondents; (2) use of terminology; (3) use of the International Classification of Diseases 10th Revision (ICD-10) codes; (4) diagnostic approach, for example, emphasis on history characteristics and event semiology and use of supplementary diagnostic tests such as VEEG, magnetic resonance imaging (MRI) and blood tests in the diagnostic process; (5) treatment approach, for example, use of different treatment modalities such as psychoeducation and psychological therapy, and referral to other services such as child and adolescent psychiatrists; and (6) the respondents’ “Attitudes and beliefs” about PNES in general, for example, questions about whether the respondents believe that children with PNES are faking, whether the available treatment options are sufficient, and whether the physicians worry about overlooking organic disease. Response format was “check what applies” format for lists and categorical (5-point response scale) for closed-ended questions. Thus, two different 5-point response scales addressed the diagnostic approaches: scale 1, which consisted of “always,” “often,” “sometimes,” “rarely,” and “never”; and scale 2, which consisted of “a whole lot,” “a lot,” “some,” “a little,” and “not at all.” One open-ended
3 Diagnostic Practice of PNES in Children question addressed the terminology used by the respondents. The full questionnaire can be seen online, Supporting Information. Data collection Approval from the Danish Data Protection Agency was obtained before initiation of data collection (journal number: 1-16-02-19-13), and the study was presented to The Central Denmark Region Committees on Health Research Ethics. Invitations to participate were sent by letter. The invitation letter contained project information, the questionnaire, a return envelope with paid postage, and information on how to complete the survey electronically using the Internet (using SurveyXact). Reminders were sent by e-mail and by letter to nonresponders up to a maximum of three times. Data preparation Respondents who chose to reply electronically entered their responses directly into the SurveyXact database. Responses received in hard copy were entered manually into the same database by the first author (BTW). The first author furthermore went through all received questionnaires, and in case of unclear or missing responses, the respondent in question was approached by e-mail and asked to clarify or answer missing items. A few uncertain responses could not be determined and were therefore considered missing in the final analyses. The full data set was exported from SurveyXact to STATA (www.stata.com) as a CSV file, i.e., in a delimited data format (Microsoft Excel, Redmond, WA, U.S.A.). Analyses For the present study, only data from domains 1 to 4 in the questionnaire were included in the analyses, whereas data related to domains 5 and 6 will be presented elsewhere. Frequencies and percentages are reported for categorical variables and mean and standard deviation (SD) for continuous variables. To assess whether subspecialization or the level of experience influenced the use of VEEG, we performed various subanalyses, comparing the proportions of respondents that answered “often or always” from different subgroups (for details see Table 3). Similar subanalyses were done to compare the same subgroups of respondents that reported to use the specific ICD-10 code for conversion disorder (F44.5). Using a Wald test (z-test), we compared the subgroups with different subspecializations: (1) respondents working in the field of social pediatrics versus respondents working in the field of neuropediatrics, (2) respondents working in the field of social pediatrics versus those working in both the neuropediatric and social pediatric fields, and (3) respondents working in the field of neuropediatrics versus those working in both the neuropediatric and social pediatric fields. We used chi-square tests to examine the effect of clinical experi-
ence on diagnostic practices by comparing (1) clinically experienced physicians who had seen >15 children with PNES during their career with less experienced clinicians who had seen ≤15 children with PNES during their career; (2) consultants (experienced specialized doctors) versus nonconsultants (residents or staff doctors); (3) pediatricians with and without expert neuropediatric training; and (4) those working at a secondary versus tertiary hospital. Based on clinical reasoning, the responses from the 5-point scales addressing the items on diagnostic approaches were categorized into two groups for both the comparative and descriptive analyses: scale 1: “often/always” versus “sometimes/ rarely/never,” and scale 2: “a lot/a whole lot” versus “some/ a little,” and “not at all.” In addition, the use of other supplementary diagnostic tests was compared between the above-mentioned subgroups. We dichotomized the responses to each of the six items concerning use of ECG, MRI, computed tomography (CT) of the cerebrum, blood glucose test, serum prolactin level test, and urine screening for inborn errors of metabolism, meaning that the answer “never,” “rarely,” or “sometimes” would give a score of zero, and the answer “often” or “always” would give a score of one. We then added up the scores for each respondent giving a total score between zero (if the respondent had answered “never/rarely/sometimes” to all the items) and six (if the respondent had answered “often/always” to all the items). This total score variable was then compared between the subgroups by the means of Mann-Whitney tests to assess whether one or more subgroup would be more likely to perform supplementary diagnostic tests than others. All analyses were performed using STATA version 12 for Windows. The statistical tests presented are two-tailed, and a significance level of p < 0.05 was used with Bonferroni correction for multiple comparisons.
Results In total, a response rate of 95% (61/64) was obtained. The questionnaire was 100% completed by 95% (58/61) of the respondents; the remaining three completed at least 80%. Table 1 shows descriptive data on the respondents. Use of terminology and diagnostic codes When communicating with children with events not related to epilepsy and their families, the terms most frequently stated to be the most appropriate to use were functional seizures (34%) and PNES (25%). Overall, 17 different diagnostic ICD-10 codes were used for pediatric PNES (Fig. 1). In total, 69% stated that they used more than one ICD-10 code, and two respondents stated that they used five different codes. The two most widely used codes were “F 91.8” (Other conduct disorders) and “R 56.8” (Other and unspecified convulsions), both of which were used by 54% of the respondents. The specific psychiatEpilepsia, **(*):1–8, 2014 doi: 10.1111/epi.12881
4 B. T. Wichaidit et al. Table 1. Demographic characteristics of the whole study sample (n = 61) based on items in domain 1 in the questionnaire Demographic data Mean age, years (SD) Sex, N (%) Male Female Hospital, N (%) Secondary level Tertiary level Othera Position, N (%): Pediatric resident Pediatric staff doctor Pediatric consultant Respondents working in the, N (%):b Neuropediatric field alone Social pediatric field alone Both fields Full-time occupation in one or both fields, N (%):b Yes No Number of children with PNES seen during career, N (%) 1–5 6–15 16–30 >30
Study sample (n = 61) 52 (8.68) 21 (34) 40 (66) 28 (46) 29 (47) 4 (7) 3 (5) 13 (21) 45 (74) 41 (68) 10 (17) 9 (15)
47 (78) 13 (22)
3 (5) 13 (21) 17 (28) 28 (46)
a Other: Highly specialized epilepsy hospital (Filadelfia) (n = 2), “emergency hospital” (n = 1), “combined secondary- and tertiary level” (n = 1). b One respondent missing for this category.
ric diagnosis for a conversion disorder in ICD-10, that is, F44.5, was used by 31%, and its use was not statistically significantly related to length of experience and/or type of subspecialty when the analyses were adjusted for multiple comparisons (p > 0.01 for all subanalyses). We used a conservative two-tailed Bonferroni criterion of p < 0.007 to adjust for seven multiple comparisons. Diagnostic setting Most of the respondents (76%) stated that the diagnostic assessment of a child suspected of having PNES took place both in the inpatient and outpatient setting as opposed to only in the outpatient or only in the inpatient setting. A pediatric emergency room visit without further follow-up was reported “never” to happen by 70%, “rarely” by 25%, and “sometimes” by 5%. History and semiology of the event The respondents were asked to indicate to which degree they judged each of the 18 history characteristics and 24 seizure-like characteristics commonly encountered in the literature to support a diagnosis of PNES. The five history characteristics that the respondents emphasized the most (>50% ticked the option indicating that this particular Epilepsia, **(*):1–8, 2014 doi: 10.1111/epi.12881
Figure 1. Responses to the following item in the questionnaire (domain 3): “Which of the following diagnostic ICD-10 codes do you use?” F 91.8, Other conduct disorders; F 44.5, Dissociative convulsions; F 98.9, Unspecified behavioral and emotional disorders with onset usually occurring in childhood and adolescence; Other* F 43.8, Other reactions to severe stress; F 43.9, Reaction to severe stress unspecified; F 44.4, Dissociative motor disorders; F 45.0, Somatization disorder; F 45.9, Somatoform disorder unspecified; R 56.8, Other and unspecified convulsions; R 25.2, Cramp and spasm; R 55.9, Syncope and collapse; R 25.8, Other an unspecified abnormal involuntary movements; R 29.8, Other and unspecified symptoms and signs involving the nervous and musculoskeletal system; Z 03.3, Observation for suspected nervous system disorder; Z 03.2, Observation for suspected mental and behavioral disorders; Other** Z 03.8, Observation for other suspected diseases and conditions; G 40.9, Epilepsy unspecified. Epilepsia ILAE
characteristic supported PNES “a lot” or “a whole lot”) were the following: “psychosocial stressors/trauma,” “sexual abuse,” “the paroxysmal events typically occur in stressful situations,” “no effect of antiepileptic drugs (AEDs),” and “physical abuse” (Fig. 2). The six event characteristics that the respondents emphasized the most were as follows: “resisted eyelid opening,” “avoidance/guarding behavior,” “the paroxysmal events occur in the presence of others,” “closed eyes,” “rarely injury related to the paroxysmal events,” and “absence of postictal change” (Fig. 3). Only two of the event characteristics (i.e., resisting eyelid opening and avoidance/guarding behavior during the paroxysmal event) and none of the history characteristics were judged to be predictive of PNES by a substantial part of the respondents (70% or more). Use of EEG and supplementary diagnostic tests Routine EEG and 24-h EEG recordings were stated by the respondents to be performed “often” or “always” by 57/ 61 (93%) and 18/61 (30%), respectively, and VEEG “often” or “always” by 30/61 (49%) (Table 2).
5 Diagnostic Practice of PNES in Children
Figure 2. Percentage of respondents who answered “a lot” or “a whole lot” to the following item in the questionnaire (domain 4): “How much do you think the following history characteristics suggest a diagnosis of functional seizures?” (The item wording has been abbreviated for conciseness): 1, Psychosocial stressors; 2, Sexual abuse; 3, Occur in stressful situations; 4, No effect of antiepileptic drugs; 5, Physical abuse; 6, Functional symptoms in family; 7, Other functional symptoms; 8, Psychiatric comorbidity; 9, Functional symptoms among acquaintances; 10, Frequent health care contacts; 11, Rarely nocturnal events; 12, Several event types (no stereotypy); 13, Epilepsy in the family; 14, Epilepsy among acquaintances; 15, High event frequency; 16, Developmental disorders; 17, Somatic comorbidity; 18, Prior head trauma. Epilepsia ILAE
The majority of the respondents replied that they would “often or always” perform ECG (64%) and blood glucose test (66%), and that they would “rarely, never, or only sometimes” perform MRI (70%), CT of the cerebrum (98%), blood tests for renal function (56%), acute phase reactants (white blood cell count, C-reactive protein) (72%), hematology (62%), liver enzyme count (67%), serum prolactin level test (92%), or urine screening for inborn errors of metabolism (98%). Comparative analyses showed that a statistically significantly larger proportion of consultants and respondents working at a tertiary hospital “often” or “always” performed VEEG, and that respondents working in the social pediatric field compared to respondents working in the neuropediatric field performed significantly more supplementary diagnostic tests (Table 3).
Discussion This national survey demonstrates disagreement among pediatricians on use of terminology and diagnostic approach for pediatric PNES. The findings correspond to results reported in similar studies on adult PNES10 and pediatric benign chronic pain, another condition dominated by medically unexplained or functional symptoms.16 Several studies, including a survey of American parents of children with PNES,2 have addressed patients’ and their relatives’ views on terms used for various medically unexplained symptoms.2,3,17,18 It was found that “functional” is perceived as one of the least offensive or pejorative terms,
Figure 3. Percentage of respondents who answered “a lot” or “a whole lot” to the following item in the questionnaire (domain 4): “How much do you think the following characteristics of a paroxysmal event suggest a diagnosis of functional seizures?” (The item wording has been abbreviated for conciseness): 1, Resist eyelid opening; 2, Avoidance/guarding behavior; 3, Occurrence in the presence of others; 4, Closed eyes; 5, Rarely injury; 6, No postictal change; 7, Intelligible speech; 8, Pelvic thrusting; 9, Long duration; 10, Response when addressed; 11, Side-to-side head movements; 12, Fluctuating pattern during episode; 13, Asymmetrical movements; 14, Fluctuating intensity; 15, Pupil light reflex intact; 16, Preserved consciousness; 17, Crying in the middle or end of the event; 18, No tongue bite; 19, Gradual onset with slow escalation in intensity; 20, No urinary incontinence; 21, No cyanosis; 22, No fecal incontinence; 23, Normal plantar response postictal; 24, Rapid, shallow, irregular, quiet breathing pattern postictal. Epilepsia ILAE
Table 2. Distribution of responses (n = 61) to the following question: “How often do you perform the following EEG investigations on a child suspected of having PNES?” (domain 4 in the questionnaire) Standard EEG, N (%) 24-h EEG, N (%) VEEG, N (%)
Always
Often
Sometimes
Rarely
Never
32 (52) 2 (3) 10 (16)
25 (41) 16 (26) 20 (33)
3 (5) 26 (43) 21 (34)
1 (2) 11 (18) 6 (10)
0 6 (10) 4 (7)
whereas “psychogenic” is perceived as one of the most offensive ones. Thus, because the choice of terminology is important to patients and their families, the conflicting responses from our survey point toward a need for agreement on which terms to use in the clinical setting. Considerable inconsistencies in coding practices for PNES were also reported, which could mirror different conceptualizations of the disorder. Some pediatricians may for example perceive PNES as a syncope-like disorder, whereas other pediatricians may perceive it as a convulsive disorder and use diagnostic codes accordingly (i.e., “R 55.9 Syncope and collapse” and “R 56.8 other and unspecified convulsions,” respectively). These potential inconsistencies in conceptualization may aggravate the commonly described confusion about the diagnosis among patients and their families. Future studies should therefore explore this subject further and whether the confusion could be based on a lack of training of pediatricians on how to assess conversion disorder in children. The inconsistencies may also constitute Epilepsia, **(*):1–8, 2014 doi: 10.1111/epi.12881
6 B. T. Wichaidit et al. Table 3. Comparison of subgroups with different degrees of clinical experience and specialization with regard to frequent use of VEEG and number of supplementary diagnostic tests (i.e., MRI, CT, ECG, blood tests, urine screening for inborn errors of metabolism) in pediatric patients with PNES Performance of VEEG (‘Often’ or ‘Always’)
Number of supplementary tests
Proportion (95% CI)
RR (95% CI)
Test statistics v2(1), p-value
Consultants (n = 45) Nonconsultants (n = 16)
0.6 (0.4–0.7) 0.2 (0.0–0.4)
3.2 (1.1–9.1)
8.04, 0.005*
2.0 (1.0–4.5) 4.0 (0.3–6.0)
0.46, 0.65
61
Neuropediatric expert training (n = 26) No neuropediatric expert training(=34)
0.7 (0.4–0.8) 0.4 (0.2–0.6)
1.7 (1.0–2.8)
4.34, 0.04
2.0 (0.0–4.0) 3.0 (1.8–6.3)
2.49, 0.01
60a
Working at a tertiary hospital (n = 29) Working at a secondary hospital (n = 28)
0.7 (0.5–0.8) 0.3 (0.1–0.5)
2.3 (1.2–4.4)
7.80, 0.005*
3.0 (1.0–6.0) 2.0 (1.0–6.0)
0.17, 0.87
57b
0.5 (0.4–0.7) 0.4 (0.2–0.6) Proportion (95% CI)
1.4 (0.7–2.8)
1.18, 0.28
3.0 (1.0–5.5) 2.5 (1.3–6.0)
0.37, 0.72
61
Subgroups
Seen >15 children with PNES (n = 45) Seen ≤15 children with PNES (n = 16)
Respondents working in thec Neuropediatric field (n = 41) Social pediatric field (n = 10)
RR (95% CI)
Median (25/75 percentiles)
Test statistics MW: Z, p-value
No. of obs
Test statistics BR: Z, p-value
Test statistics MW: Z, p-value
0.6 (0.5–0.8)a 0.2 (0.1–0.7)b
3.0 (0.9–10.8)
1.7, 0.08
2.0 (0.0–4.5)a 6.0 (3.8–7.3)b
3.03, 0.003*
51
Both fields (n = 9) Social pediatric field
0.3 (0.1–0.8)c 0.2 (0.1–0.7)b
1.6 (0.4–7.8)
0.7, 0.52
2.0 (1.0–2.5)c 6.0 (3.8–7.3)b
0.26, 0.010
19
Neuropediatric field Both fields
0.6 (0.5–0.8)a 0.3 (0.1–0.8)c
1.8 (0.7–4.8)
1.2, 0.22
2.0 (0.0–4.5)a 2.0 (1.0–2.5)c
0.26, 0.80
50
RR, relative risk; BR, binomial regression, MW, Mann-Whitney. a One missing on this item. b Four responses could not be categorized as either working in secondary or tertiary hospital and was therefore considered as missing. c One respondent could not be categorized as working in either the neuropediatric or social pediatric field, or in both fields (n = 60 in all these subanalyses). * Statistically significant using a conservative two-tailed Bonferroni criterion of p < 0.007 to adjust for seven multiple comparisons.
a major obstacle for inter- and intradisciplinary communication as well as for performing systematic research in this area. In addition, it is of interest that one of the two most frequently reported ICD-10 codes was the psychiatric code “F 91.8” (Other conduct disorders), even though the most correct psychiatric diagnosis would be conversion disorder (i.e., in the context of ICD-10, “F 44.5”). The Danish Pediatric Society recommended at one point that “F 91.8” should be the diagnostic code for children with PNES and added the term “pseudoseizure” to the code. This probably explains why so many Danish pediatricians in the present study reported to use this specific code for PNES. The use of “F 44.5” was not found to be related to length of experience and/or type of subspecialty, and the inappropriate use of psychiatric ICD-10 codes may therefore reflect that, in general, confusion arises when psychiatric codes are being applied in a general hospital setting. The diagnostic workup of children with PNES was mainly reported to take place in pediatric departments, and no patients were seen solely at pediatric emergency rooms without further follow-up, which indicates that diagnosing PNES often requires a longer period of assessment.6 Overall, there seemed to be a lack of agreement concerning which history and event characteristics could be suggestive of a PNES diagnosis, that is, to which degree the Epilepsia, **(*):1–8, 2014 doi: 10.1111/epi.12881
clinicians thought that the listed characteristics supported a diagnosis of PNES. This may reflect that no single sign can be considered pathognomonic for the diagnosis,19,20 and that clusters of event characteristics may be more useful in the differentiation between PNES and epileptic seizures.21 An additional explanation, as suggested by Alessi et al.,22 could be lack of awareness among pediatricians that the classic signs observed in adults with PNES are not common in children. An overall variation in the use of the gold standard investigation VEEG1,7,8,19,23 was documented. The higher use of VEEG by consultants and doctors working in a tertiary hospital suggests that pediatricians with longer clinical experience are more aware of the importance of VEEG, but also that there may be potential differences in VEEG accessibility in the secondary and tertiary settings. One could also speculate whether respondents who stated that they performed VEEG less frequently may be of the opinion that it is possible to reach a correct diagnosis of PNES without VEEG. However, interictal EEG has a number of limitations, and eyewitness descriptions alone can be misleading and unreliable.6,7,19,20 The findings therefore suggest a need for increased awareness of the importance of VEEG among clinicians, but also that the test needs to be made more widely available in the clinical setting.
7 Diagnostic Practice of PNES in Children Determination of the serum prolactin level is not recommended for diagnostic purposes, since it is not always elevated following epileptic seizures and may be affected by other medical conditions or medications.23 Consistent with this, serum prolactin levels were rarely determined by the respondents. However, other supplementary diagnostic tests showed to be used in an inconsistent manner, and respondents working in the social pediatric field were significantly more likely to perform additional tests compared to respondents working in the neuropediatric field. Again, this seems to support that lack of clinical experience and/or expertise will influence the individual pediatrician’s diagnostic workup. Altogether, this calls for consensus and guidelines with regard to the use of supplementary tests as well. To the best of our knowledge, the present study is the first attempt to survey a representative population of clinicians’ diagnostic practices for pediatric PNES in the hospital setting. The main strengths of the study are the very high participation rate, that the majority of the respondents were specialists in pediatrics, and that they all had experience with PNES. Because eligible respondents were identified by the chief consultants of all pediatric departments in DK, we ensured that all or close to all doctors working within this field in the hospital setting were approached. However, some limitations of the study should be considered. First, given that this is a national survey from DK, the results may not be generalizable to clinical settings in other countries. Nevertheless, the results correspond well with those of the few other studies concerning similar topics that have been performed in other European countries.10,16 To our knowledge, children with PNES in other countries are also mainly diagnosed by doctors in the field of neuropediatrics, as found in our study.1,5,7 It therefore seems likely that the overall clinical picture of this survey could be the same elsewhere. Second, the questionnaire could have been further validated. Yet, the pilot testing confirmed the content validity of the questionnaire. The high response rate on all the items indicates that it was easily understood, and to address the constructs pertinent to PNES, the questionnaire was designed by review of the literature and evaluated by experts in the field. Although it might have been more appropriate to pilot test the questionnaire on specialists, we chose to use trainees in order not to reduce the number of eligible participants for the actual survey. Third, it should be mentioned that the terms “functional seizures” and “PNES” were used in the questionnaire, and it therefore cannot be ruled out that some respondents were biased to pick one of those as their preferred term. Finally, the comparative analyses were based on small numbers, reducing the power and performance of more sophisticated analyses (e.g., such as the effect of being both a consultant and working at a tertiary hospital on the use of VEEG and supplementary tests). In addition, there may have been considerable overlap between some of the subgroups we defined so that the
significant findings identified the same respondents. Regardless of these limitations, the results still indicate a lack of overall consensus among pediatricians on the use of VEEG and supplementary tests in this patient group. In conclusion, this survey strongly supports the need for clinical guidelines that should be developed by specialists in the field and designed to create consensus on diagnostic practice. The first step could be to focus on the use of terminology and diagnostic codes. The formulation of a framework for which diagnostic approach to use at this point is also not beyond reach, although larger and more standardized studies are needed to establish evidence-based guidelines. Future studies should include relevant control groups for comparison and use standardized questionnaires to obtain systematic data on history. VEEG recordings should be reviewed by blinded, experienced epileptologists to compare PNES and epilepsy with regard to event semiology and to identify clusters of characteristics that may be particularly predictive for pediatric PNES. This could potentially lead to an enhanced diagnostic accuracy, which ultimately would lead to a better outcome for children with PNES.
Acknowledgments The authors thank Elena Garralda, Professor emeritus, MD, PhD, Academic Unit of Child and Adolescent Psychiatry, Imperial College, London, UK; Cornelius Ani, Consultant and Honorary Senior Lecturer, MD, Academic Unit of Child and Adolescent Psychiatry, Imperial College, London, UK; and Per Fink, Professor, MD, PhD, DMSc, The Research Clinic For Functional Disorders and Psychosomatics, Aarhus University Hospital, Denmark, for their helpful comments on the preliminary questionnaire. We also thank Jens Søndergaard Jensen, MSc, The Research Clinic for Functional Disorders, Aarhus University Hospital, Denmark, for statistical advice. Furthermore, we thank all the pediatricians who participated in the study. “The Danish Council for Independent Research | Medical Sciences,” “Central Denmark Region – the Foundation for Improvement of Psychiatric Research at the Child and Adolescent Psychiatric Center,” “Beatrice Surovell Haskell Fund for Child Mental Health Research of Copenhagen,” and “The Møller Foundation – A.P. Møller and wife Chastine McKinneyMøller” funded the project.
Disclosure None of the authors have any conflict of interest to disclose. We confirm that we have read the journal’s position on issues involved in ethical publication and affirm that this report is consistent with those guidelines.
References 1. Reilly C, Menlove L, Fenton V, et al. Psychogenic nonepileptic seizures in children: a review. Epilepsia 2013;54:1715–1724. 2. Morgan LA, Dvorchik I, Williams KL, et al. Parental ranking of terms describing nonepileptic events. Pediatr Neurol 2013;48:378–382. 3. Stone J, Campbell K, Sharma N, et al. What should we call pseudoseizures? The patient’s perspective. Seizure 2003;12:568–572. 4. Dhiman V, Sinha S, Rawat VS, et al. Children with psychogenic nonepileptic seizures (PNES): a detailed semiologic analysis and modified new classification. Brain Dev 2013;36:287–293. 5. Plioplys S, Asato MR, Bursch B, et al. Multidisciplinary management of pediatric nonepileptic seizures. J Am Acad Child Adolesc Psychiatry 2007;46:1491–1495. Epilepsia, **(*):1–8, 2014 doi: 10.1111/epi.12881
8 B. T. Wichaidit et al. 6. Lafrance WC Jr, Baker GA, Duncan R, et al. 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– 2018. 7. Patel H, Dunn DW, Austin JK, et al. Psychogenic nonepileptic seizures (pseudoseizures). Pediatr Rev 2011;32:e66–e72. 8. Wood BL, Haque S, Weinstock A, et al. Pediatric stress-related seizures: conceptualization, evaluation, and treatment of nonepileptic seizures in children and adolescents. Curr Opin Pediatr 2004;16:523– 531. 9. LaFrance WC Jr, Rusch MD, Machan JT. What is “treatment as usual” for nonepileptic seizures? Epilepsy Behav 2008;12:388–394. 10. Mayor R, Smith PE, Reuber M. Management of patients with nonepileptic attack disorder in the United Kingdom: a survey of health care professionals. Epilepsy Behav 2011;21:402–406. 11. Sahaya K, Dholakia SA, Lardizabal D, et al. Opinion survey of health care providers towards psychogenic non epileptic seizures. Clin Neurol Neurosurg 2012;114:1304–1307. 12. Shneker BF, Elliott JO. Primary care and emergency physician attitudes and beliefs related to patients with psychogenic nonepileptic spells. Epilepsy Behav 2008;13:243–247. 13. Ani C, Reading R, Lynn R, et al. Incidence and 12-month outcome of non-transient childhood conversion disorder in the U.K. and Ireland. Br J Psychiatry 2013;202:413–418. 14. Whitehead K, Kandler R, Reuber M. Patients’ and neurologists’ perception of epilepsy and psychogenic nonepileptic seizures. Epilepsia 2013;54:708–717. 15. Furness P, Glazebrook C, Tay J, et al. Medically unexplained physical symptoms in children: exploring hospital staff perceptions. Clin Child Psychol Psychiatry 2009;14:575–587. 16. Konijnenberg AY, De Graeff-Meeder ER, Kimpen JL, et al. Children with unexplained chronic pain: do pediatricians agree regarding the
Epilepsia, **(*):1–8, 2014 doi: 10.1111/epi.12881
17. 18. 19. 20. 21. 22. 23.
diagnostic approach and presumed primary cause? Pediatrics 2004;114:1220–1226. Edwards MJ, Stone J, Lang AE. From psychogenic movement disorder to functional movement disorder: it’s time to change the name. Mov Disord 2013;29:849–852. Stone J, Wojcik W, Durrance D, et al. What should we say to patients with symptoms unexplained by disease? The “number needed to offend”. BMJ 2002;325:1449–1450. Devinsky O, Gazzola D, LaFrance WC Jr. Differentiating between nonepileptic and epileptic seizures. Nat Rev Neurol 2011;7:210–220. Syed TU, LaFrance WC Jr, Kahriman ES, et al. Can semiology predict psychogenic nonepileptic seizures? A prospective study. Ann Neurol 2011;69:997–1004. Hubsch C, Baumann C, Hingray C, et al. Clinical classification of psychogenic non-epileptic seizures based on video-EEG analysis and automatic clustering. J Neurol Neurosurg Psychiatry 2011;82:955–960. Alessi R, Vincentiis S, Rzezak P, et al. Semiology of psychogenic nonepileptic seizures: age-related differences. Epilepsy Behav 2013;27:292–295. Plioplys S, Siddarth P, Asato MR, et al. Clinicians’ views on antiepileptic medication management in nonepileptic seizures. J Child Neurol 2013;29:746–750.
Supporting Information Additional Supporting Information may be found in the online version of this article: Data S1. Questionnaire regarding children with functional seizures/psychogenic nonepileptic seizures. Figure S1. The process of developing the questionnaire.
