If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
A meta-analysis evaluating prevalence of seizure reduction in psychological treatments for PNES.
•
82% of those completing psychotherapy report a reduction in seizures of ≥50%.
•
Results suggest psychotherapy yields greater rates of seizure reduction and freedom than TAU.
Abstract
Purpose
The aim of this meta-analysis is to evaluate and synthesize the available evidence from the previous 20 years regarding the utility of psychological interventions in the management of psychogenic non-epileptic seizures (PNES).
Method
Studies were retrieved from MEDLINE via OvidSP and PsychINFO. Selection criteria included controlled and before-after non-controlled studies including case series, using seizure frequency as an outcome measurement. Studies were required to assess one or more types of psychological intervention for the treatment of PNES in adults. Data from 13 eligible studies was pooled to examine the effectiveness of psychological interventions in treating PNES on two primary outcomes: seizure reduction of 50% or more and seizure freedom. A meta-analysis was conducted with data extracted from 228 participants with PNES.
Results
Interventions reviewed in the analysis included CBT, psychodynamic therapy, paradoxical intention therapy, mindfulness and psychoeducation and eclectic interventions. Meta-analysis synthesized data from 13 studies with a total of 228 participants with PNES, of varied gender and age. Results showed 47% of people with PNES are seizure free upon completion of a psychological intervention. Additional meta-analysis synthesized data from 10 studies with a total of 137 participants with PNES. This analysis found 82% of people with PNES who complete psychological treatment experience a reduction in seizures of at least 50%.
Conclusion
The studies identified for this analysis were diverse in nature and quality. The findings highlight the potential for psychological interventions as a favorable alternative to the current lack of treatment options offered to people with PNES.
Psychogenic non-epileptic seizures (PNES) have a debilitating impact on quality of life. This may involve psychological, social, financial and physical consequences including the inability to work, drive or carry out everyday tasks [
]. Despite the growing amount of research contributing to our understanding of PNES and its causes, there is little evidence available about successful treatments [
Minimum requirements for the diagnosis of psychogenic nonepileptic seizures: a staged approach: a report from the International League Against Epilepsy Nonepileptic Seizures Task Force.
Minimum requirements for the diagnosis of psychogenic nonepileptic seizures: a staged approach: a report from the International League Against Epilepsy Nonepileptic Seizures Task Force.
]. Research also tells us that, without treatment, the majority of people with PNES continue to have seizures and many experience a worsening of symptoms [
Whilst PNES is a condition defined by physical manifestations, it is understood to be psychological in nature with a wide variety of aetiological factors involved [
]. Consequently, PNES represent a serious problem for clinicians in developing and implementing evidence-based psychological interventions and there is currently little in the way of quality evidence which can inform clinical treatment decisions [
]. The body of research indicates that psychological interventions for PNES are in the early stages of development. These encompass a number of approaches, the most common being cognitive behavioral therapy (CBT), psychoanalytical and psychoeducational therapies.
The majority of the research into psychological interventions for PNES is comprised of observational studies, involving pre-post studies without control groups [
] providing a promising start in the pursuit of high quality research into PNES interventions. On their own however, they are insufficient in number to allow clinicians to draw conclusions about broader treatment recommendations [
There are several reasons for the limited number of high quality studies in this field. RCT’s are typically performed in highly controlled environments where extraneous variables can be controlled. The majority of RCT’s do not allow for the presence of co-morbid disorders, common among people with PNES [
], leaving these people unrepresented in the research. RCT’s also commonly require a single standardized treatment which is difficult to develop for such a group as diverse as those with PNES which can be a symptom of various affective and psychiatric factors [
]. Furthermore, RCT to waiting list or treatment as usual (TAU) is unattractive, and can be deemed unethical for patients who are unwell, when similar treatments are available outside of the research setting [
Overall, the individual research studies are suggestive of favorable outcomes in terms of reducing seizures for those who complete psychological treatment. However, as a whole, the literature is laborious to interpret as studies are often published in a variety of different medical, psychological or psychiatric journals, use different methodologies, and are presented in such a way as to make them difficult to compare with one another. As it stands, the evidence is indicative of both the current state of clinical interventions for PNES as they occur in practice, and reflective of the populations they aim to treat. It is also representative of the diversity of approaches required in addressing such a heterogeneous group of patients and presentations [
]. The observational designs utilized by the majority of researchers in this field, whilst of limited methodological quality, have the capacity to evaluate treatment outcomes in people with multiple problems, complex or atypical presentations in real life clinical settings [
]. Naturalistic studies inform clinicians, researchers and other health professionals about treatments, as they would be performed in practice, without exclusions and controlled conditions [
To date, there are no meta-analytical reviews of psychological interventions for PNES. This absence may be ascribed to the lack of RCT’s, the customary design used for a meta-analytical review and synthesis [
]. However, when considering the high social, psychological and financial costs associated with PNES, there is an imperative to utilize the current body of research to its full extent [
]. Additionally, considering the complex nature of PNES, combined with the difficulty and high cost of RCT’s, it is unlikely there will be a sufficient number of RCT’s conducted in the near future for this type of meta-analysis to be performed. Meanwhile, uncontrolled and naturalistic treatment evaluations in clinical service-settings provide valuable information in their own right [
]. Increasingly, as in other areas of health research, the combination of large amounts of observational literature and the pressure for timely, accurate clinical information compels researchers to utilize observational studies using meta-analysis [
]. Combining this diffuse body of research will also enable this information to be more readily accessible, and therefore, help educate clinicians of current evidence-based treatments for PNES [
The aim of this systematic review and meta-analysis is to evaluate and examine the available evidence from the previous 20 years regarding the effectiveness of psychological interventions in the management of PNES. Using meta-analysis, this study predicts that psychological interventions for PNES will be shown to be associated with both seizure freedom and reductions in seizure frequency of 50% or more.
2. Method
2.1 Protocol
A review protocol for this study was developed in December 2015 and is available upon request from the author (PC).
2.2 Eligibility criteria
Eligible studies were required to be published electronically in peer reviewed science journals in the English language between 1996 and 2016. PNES was defined as the experience of non-epileptic seizures of psychological origin as diagnosed by a neurologist, psychologist or psychiatrist and confirmed by electroencephalogram (EEG) or video-EEG (vEEG). Given the important differences between PNES in adults and children, studies required the inclusion of participants aged 16 years and older (>50% of the participants are ≥16 years). Due to the limited amount of research in this area, the search was open to all prospective human studies, including controlled and before-after non-controlled studies including case series. Qualitative single case studies, and retrospective studies were excluded, as were review articles and conference abstracts.
Studies were included if they evaluated the effectiveness of at least one psychological intervention undertaken to lessen the frequency of PNES. An intervention was considered to be psychological in nature if it was based on a psychological theory or model specifically designed to alter psychological processes thought to underlie or significantly contribute to pain, distress, and disability [
Additional selection criteria included using seizure frequency as an outcome measure. Regardless of the method of reporting, this criterion was included in the qualitative synthesis. Studies selected as eligible for quantitative synthesis were required to provide sufficient information on the primary outcome of seizure frequency so as to enable the calculation of either seizure reduction and/or seizure freedom rates. Studies were excluded if they examined the effectiveness of non-psychological interventions (i.e. medication) or focused on psychological interventions that targeted other outcomes (i.e. employment status, cost efficacy).
2.3 Search and selection strategy
In order to decide which studies to include in the analyses, an extensive literature search was conducted utilising two online academic databases, MEDLINE via OvidSP and PsychINFO (see Appendix A). To do this, a search strategy was developed using a wide-ranging pool of MeSH/thesaurus terms tailored to each database (see electronic search strategy for MEDLINE via OvisSP database in Table 1). The search was conducted by the author (PC) and included records from 1996 to July 2016. If the article title indicated relevance then the abstract was read. The complete article was read if the abstract indicated the article met the inclusion criteria. Following this, reference lists from selected studies were examined for additional relevant papers. The authors consulted in the event of any queries and discrepancies were resolved by discussion. This search was conducted on 3 June 2016 and in accordance with Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement for reporting meta-analysis [
The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate healthcare interventions: explanation and elaboration.
One reviewer (PC) completed the data extraction process using an electronic data extraction sheet based on the Cochrane Public Health Group Data Extraction and Assessment Template. A pilot form was developed and tested using two randomly selected studies and the final form adjusted accordingly to define study characteristics (i.e. author, date, sample size), sample characteristics (i.e. mean age and range, definition of PNES), intervention characteristics (i.e. intervention type, delivery method, duration, relevant theoretical basis), outcome characteristics (i.e. outcome measures), and trial characteristics (i.e. inclusion and exclusion criteria, recruitment, setting). A cross-check of the data was completed by the reviewer (PC) with additional evaluation by random selection conducted by author (KNP). This author (KNP) was also consulted in the event of any queries or discrepancies.
When published articles did not present sufficient statistical data (information was missing or incomplete) from which to calculate the meta-analysis, authors were contacted. Three studies were identified as providing insufficient data for meta-analysis and were contacted. Two authors declined to provide data and one did not respond.
2.5 Quality and risk of bias
Selected studies were assessed for methodological quality using (a) the quality assessment of controlled intervention studies tool and (b) the quality assessment tool for before-after (pre-post) studies with no control group, both developed by the National Heart, Lung, and Blood Institute [
]. These tools are based on quality assessment methods and other tools developed by researchers in the Agency for Healthcare Research and Quality, evidence-based practice centers, and the Cochrane Collaboration. The tools are designed to critically appraise the internal validity of these specific study designs. Each tool includes items for evaluating potential flaws in study methods or implementation, including sources of bias (e.g., patient selection, performance, attrition, and detection), confounding, study power, the strength of causality in the association between interventions and outcomes, and other factors. The tools consist of 14 (controlled tool) or 12 (pre-post tool) items, each rated as yes, no, other (cannot determine, not reported or not applicable). A final quality rating (good, fair or poor) was assigned for each study and reasons for a rating of poor was noted.
2.6 Statistical analysis
Meta-analysis software MetaXL (http://www.epigear.com/index_files/metaxl.html) was used to conduct all statistical analyses in Microsoft Excel. Due to the high heterogeneity of seizure frequency found in PNES populations, the majority of studies reported data which was not normally distributed. As such, mean seizure frequencies were unable to be compared. An alternative analysis pooling single proportions was used to calculate a meta-analysis of prevalence.
This study utilized a random effects model to determine the percentage of people who experienced a reduction in seizure frequency of 50% or more. A second meta-analysis of prevalence was conducted to establish the percentage of those who experience seizure freedom following intervention. Only studies providing sufficient data to estimate either the percentage of people reporting seizure reduction of 50% or more, or seizure freedom following intervention, were included in the analyses. If studies provided outcome results for several time points, post values were defined as the first score available, closest to the time of intervention completion, in order to increase comparability. Participants who did not complete the intervention, or reported seizure freedom at baseline, were excluded from the analysis.
3. Results
3.1 Study selection
Fig. 1 depicts the study selection process and results of each review step. A search of two databases yielded a total of 164 citations. Of these, 7 were eliminated as they were duplicates. Titles of 157 studies were screened with 89 selected for review at the abstract level. The full text of 21 studies were reviewed and assessed for eligibility. Five of these did not meet the inclusion criteria and were eliminated during extraction. The remaining 16 studies were considered eligible for qualitative synthesis. A further 3 studies were considered to have insufficient data for analysis. Requests for this data were declined or unreturned. In conclusion, data from 13 studies with a total sample of 346 participants with PNES was collected. Excluding those participants who did not complete interventions, completed alternative interventions or reported seizure freedom at baseline, a total sample of 228 participants was extracted and incorporated in the meta-analyses.
Table 2 provides a summary of descriptive characteristics of the 13 included studies. Studies were published between 2001 and 2014. Study participants (N = 346) were primarily female (85.5%) aged between 16 and 60 years. Twelve of 13 studies utilized vEEG monitoring prior to recruitment with all studies utilizing EEG in their recruitment of participants. Of the 13 studies included in the final analyses, 8 excluded participants with a current diagnosis of epilepsy or where vEEG revealed the potential for equivocal diagnosis. Two of those [
] who included participants with confirmed epilepsy, reportedly did so when seizures were clearly defined and could be identified as either PNES, or epileptic in nature. Table 3 provides a summary of interventions, measures and outcomes for each of the included studies.
Long-term outcome of brief augmented psychodynamic interpersonal therapy for psychogenic nonepileptic seizures: seizure control and health care utilization.
Long-term outcome of brief augmented psychodynamic interpersonal therapy for psychogenic nonepileptic seizures: seizure control and health care utilization.
Long-term outcome of brief augmented psychodynamic interpersonal therapy for psychogenic nonepileptic seizures: seizure control and health care utilization.
Studies relied on a mixture of self-report and clinician-rated measures. There was some uniformity across studies in the use of secondary outcome measures. Studies measured a wide range of factors including employment, welfare payments, disability, medication changes, coping styles, anger expression and other indicators of functioning and psychopathology.
3.3 Evaluation of study quality
A total of 13 studies met inclusion criteria for statistical meta-analysis. Fig. 2 outlines quality ratings for each of the eligible studies. Two of the 13 studies used a randomized controlled trial (RCT) design while the remaining 11 were before-after non-controlled studies. All studies provided empirical and theoretical rationale for their objectives, however, methodological and statistical detail varied greatly. This included incomplete data relating to participant recruitment, primary and secondary outcome measures, intervention application and duration, and other sources of bias.
Fig. 2Risk of bias summary based on the NHLBI quality assessment tools.
The included studies were combined to examine the effectiveness of psychological interventions in treating PNES on the two primary outcomes examined: seizure reduction of 50% or more and seizure freedom. Results from 13 individual studies were pooled to obtain the prevalence of those reporting a seizure freedom following intervention. Results of a Chi-squared test for homogeneity showed a moderately high level of heterogeneity (I2 = 78%, Q = 54.38, p = <0.01). A random-effects model meta-analysis was used to calculate prevalence rates. Results are displayed in Fig. 3. Prevalence calculations were possible for all 13 studies. Overall, results showed that 47% (95% CI 0.33–0.62) of participants experienced seizure freedom following intervention.
Fig. 3Forest plot of prevalence, 95% confidence intervals, and % weights for studies measuring seizure freedom.
A second meta-analysis was conducted evaluating seizure freedom. Ten studies provided sufficient data for this analysis. Results from these studies were pooled to obtain the estimate of those reporting a reduction in seizure frequency of 50% or more. Prior to analysis a Chi-squared test for homogeneity was conducted. Results showed a moderate level of heterogeneity (I2 = 58%, Q = 21.54, p = <0.05). Following this, a random-effects model meta-analysis was used to calculate prevalence rates (see Fig. 4). Prevalence calculations were possible for all 10 studies. Overall, results showed that 82% (95% CI 0.70–0.91) of participants experienced a reduction in seizure frequency of 50% or more following intervention.
Fig. 4Forest plot of prevalence, 95% confidence intervals, and % weights for studies measuring seizure reduction of 50% or more.
In order to evaluate sources of bias, specifically publication bias, funnel plots of standard error using prevalence rates were examined. Points were not symmetrically dispersed (see Appendix B), indicating possible publication bias. In relation to blinding, due to the nature of psychological interventions under review, this was not possible.
4. Discussion
The aim of the current study was to evaluate the effectiveness of psychological interventions in the management of PNES by measuring the prevalence of both seizure freedom and seizure reduction of 50% or more. Meta-analytic methods were utilized to synthesize and review the existing literature. The first meta-analysis synthesized data from 13 studies with a total of 228 participants with PNES, of varied gender and age. The analysis revealed 47% of people with this condition are seizure free upon completion of a psychological intervention. Results from the second meta-analysis synthesized data from 10 studies with a total of 137 participants with PNES. The results indicate that 82% of people with PNES who complete psychological treatment experience a reduction in seizures of 50% or more.
These results make for a markedly more positive prognosis than reported in the literature to date [
] among those with PNES who do not receive psychological treatment. According to the literature, the overwhelming majority of people living with this condition remain untreated [
]. In 2003, Reuber et al. reported that without treatment, 29% of patients stated they were seizure-free more than four years following a diagnosis of PNES [
]. Therefore, for those with PNES who do not receive treatment, the vast majority will not improve. Although this study evaluated outcomes in the longer-term, more recent studies have measured short-term outcomes and found similar results.
] conducted a retrospective cohort study in 260 PNES patients in the UK. Participants were surveyed approximately 6–12 months after diagnosis and had received no treatments for PNES. The study found that without intervention, only 23% of patients with PNES reported a spontaneous reduction in seizure frequency of 50% or more. Potentially of greater concern was their finding that for those still experiencing seizures, frequency had increased by more than 50% from that recorded at baseline [
]. In LaFrance et al. RCT, outcomes were measured at 16 weeks. The study found 1 one out of 7 participants (14%) in the TAU group experienced a reduction in seizures of 50% or more, with 2 reporting an increase in PNES frequency [
]. Therefore, when compared with the existing evidence, the results of our meta-analyses indicate that psychological interventions for PNES may yield greater rates of seizure reduction (82%) and seizure freedom (47%) compared with those who do not receive psychotherapy (14–23%).
4.1 Limitations
There were a number of limitations of this study. The review utilized immediate treatment outcomes and did not include follow up data. As the effects of interventions tend to decay over time, the results may be interpreted as reflecting the maximal treatment effect currently achievable. It is therefore, difficult to generalize results in the long-term and unclear how consistent outcomes of seizure frequency remain over the years.
Whilst every effort was made to obtain relevant data from the study authors, some publications lacked information, which may have affected that study’s inclusion in the final analysis. Additionally, acknowledgment should also be made as to the bias involved whenever there is one primary reviewer responsible for identifying and selecting relevant studies for inclusion.
Selective publication, usually based on positive findings, also represents a threat to the validity of this study, as it does of all meta-analyses of observational studies [
]. In order to counter this effect, strict quality criteria were applied during the data extraction process. Regardless, analysis revealed the results may be subject to possible publication bias.
Other important considerations associated with the large proportion of observational studies in this analysis should also be noted. Caution should be applied when interpreting the results of non-controlled studies as they are unable to account for confounding variables [
Higgins J.P.T. Green S. Cochrane handbook for systematic reviews of interventions, Version 5.1.0 [updated March 2011]. The Cochrane Collaboration,
2011
]. The lack of control groups in a study design means these studies cannot prove beyond doubt that rates of seizure reduction or seizure freedom, are due to their intervention and not explained by phenomena such as placebo effects, duration or severity of symptoms, gender, or other unaccounted for factors [
Whilst some studies in the analysis attempted to limit common confounding variables such as concomitant epilepsy, others did not. Some studies, where participants held a concomitant diagnosis of epilepsy, did so on the basis that they were able to clearly differentiate between epileptic seizures and PNES [
] made no such accommodations, making this influence on treatment outcomes difficult to quantify.
4.2 Clinical and research implications
This analysis did little to highlight which type of therapies might be more beneficial in managing PNES than others. However, a number of the studies where prevalence rates were estimated to be ≥0.82 for seizure reduction of 50% or more, flexible treatment approaches were utilized. These treatments, whilst different, all demonstrate their ability to meet the needs of a heterogeneous patient population. Flexible treatment approaches are important, not merely because PNES is such a heterogeneous condition. Psychiatrists and psychologists often use clinical judgment when determining which methods may best suit particular patients or address the relevant maintaining factors underlying PNES [
]. This style also allows for collaboration between the clinician and the patient, typical in therapy and in developing a therapeutic alliance thought essential to successful treatment outcomes [
in: Duncan B.L. Miller S.D. Wampold B.E. Hubble M.A. The heart and soul of change: delivering what works in therapy. 2nd ed. American Psychological Association,
Washington, DC, US2010: 113-141
]. It may therefore be advantageous for future research to focus on identifying particular associated factors, underlying mechanisms or population groups for whom particular treatment interventions are most effective. Additionally, the results of the present study indicate that clinicians adopting a client-centered or flexible approach to treating PNES may achieve more beneficial outcomes for their patients than those utilizing a manualised methodology.
5. Conclusion
A number of brief, psychotherapeutic interventions for PNES have been reported over the past 20 years, though appropriately powered, controlled, effectiveness studies are still lacking. However, this should not be seen as justification for the current lack of treatment options provided to patients presenting with this condition. PNES is one of the most common medically unexplained neurological symptoms [
]. More must be done to educate clinicians of current treatment options and support patients in understanding the diagnosis. For this to happen, current treatments must be viewed in the context of the alternative, that is; no treatment, or worse; intermittent emergency treatment for epilepsy where the risk of iatrogenic complications is high [
To the authors’ knowledge, this is the first meta-analysis to investigate the prevalence of seizure reduction following psychological interventions for PNES. The results of the analyses indicate that psychological interventions for PNES may yield greater rates of seizure reduction and seizure freedom compared to those who do not receive psychotherapy. Results also suggest the multiple factors associated with PNES may require the adoption of a flexible methodology in the treatment of PNES. It is not yet known whether particular types of psychopathology are associated with particular manifestations of PNES [
]. Therefore, therapies must aim to accommodate a diverse range of psychological histories, interpersonal problems and range of functioning if they are to be successful [
In conclusion, the published studies identified for this analysis were diverse in nature and quality. However, the findings highlight the potential for psychological interventions as a favorable alternative to the current lack of treatment options offered to people with PNES. They also demonstrate the need for the future exploration of a wide variety of treatment approaches in this area with improved methodological designs.
Conflict of interest statement
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Minimum requirements for the diagnosis of psychogenic nonepileptic seizures: a staged approach: a report from the International League Against Epilepsy Nonepileptic Seizures Task Force.
The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate healthcare interventions: explanation and elaboration.
Long-term outcome of brief augmented psychodynamic interpersonal therapy for psychogenic nonepileptic seizures: seizure control and health care utilization.
Higgins J.P.T. Green S. Cochrane handbook for systematic reviews of interventions, Version 5.1.0 [updated March 2011]. The Cochrane Collaboration,
2011 (Available from:)
in: Duncan B.L. Miller S.D. Wampold B.E. Hubble M.A. The heart and soul of change: delivering what works in therapy. 2nd ed. American Psychological Association,
Washington, DC, US2010: 113-141
30322-3&id=gr1.jpg){kind=link}
30322-3&id=gr2.jpg){kind=link}
30322-3&id=gr3.jpg){kind=link}
30322-3&id=gr4.jpg){kind=link}
30322-3&id=fx1.jpg){kind=link}