Seizure 45 (2017) 142–150

Contents lists available at ScienceDirect

Seizure
journal homepage: www.elsevier.com/locate/yseiz

Psychological interventions for psychogenic non-epileptic seizures:

A meta-analysis
Perri Carlson* , Kathryn Nicholson Perry
Australian College of Applied Psychology, Level 11, 255 Elizabeth Street, Sydney, NSW 2000, Australia

A R T I C L E I N F O A B S T R A C T

Article history: Purpose: The aim of this meta-analysis is to evaluate and synthesize the available evidence from the
Received 27 September 2016 previous 20 years regarding the utility of psychological interventions in the management of psychogenic
Received in revised form 7 December 2016 non-epileptic seizures (PNES).
Accepted 11 December 2016
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
Keywords: outcome measurement. Studies were required to assess one or more types of psychological intervention
Psychogenic seizures
for the treatment of PNES in adults. Data from 13 eligible studies was pooled to examine the effectiveness
Non-epileptic seizures
PNES
of psychological interventions in treating PNES on two primary outcomes: seizure reduction of 50% or
Pseudoseizures 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.
© 2016 British Epilepsy Association. Published by Elsevier Ltd. All rights reserved.

1. Introduction marginalized between neurology and psychiatry, with neither

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 [1,2]. Despite the growing
amount of research contributing to our understanding of PNES and
its causes, there is little evidence available about successful
treatments [3,4].
Prognosis for people with PNES is poor [5]. Diagnosis is often
focused on the exclusion of epilepsy and consequently, PNES
becomes a non-disease [6]. People with PNES tend to be
Abbreviation: PNES, psychogenic non-epileptic seizures.
* Corresponding author. Permanent address: Central West Neurology &
Neurosurgery, 93 Byng Street, Orange, NSW 2800, Australia.
E-mail addresses: perri@cwnn.com.au (P. Carlson),
Kathryn.NicholsonPerry@acap.edu.au (K. Nicholson Perry).
profession taking ownership of patient care [7]. As such, many
patients are not referred to or do not engage with mental health
services [3,4,6]. Once a diagnosis of PNES is made, anti-convulsant
therapy is typically ceased and treatment options are unclear and
rarely pursued [1,7]. Stigma often surrounds a diagnosis of PNES,
fueled by poor understanding, education or support for the
condition [1,6]. Research also tells us that, without treatment, the
majority of people with PNES continue to have seizures and many
experience a worsening of symptoms [8,9,5].
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 [10,4,11]. Consequently, PNES repre-
sent a serious problem for clinicians in developing and imple-
menting evidence-based psychological interventions and there is
currently little in the way of quality evidence which can inform
clinical treatment decisions [12]. 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

http://dx.doi.org/10.1016/j.seizure.2016.12.007
1059-1311/© 2016 British Epilepsy Association. Published by Elsevier Ltd. All rights reserved.
 P. Carlson, K. Nicholson Perry / Seizure 45 (2017) 142–150
common being cognitive behavioral therapy (CBT), psychoanalyti-
cal 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 [12]. Most are small in scale and
conducted in hospital or medical facilities, reflective of real life
clinical treatment settings [13,14,9,15]. These studies are inter-
spersed with a handful of small scale and pilot randomized
controlled trials (RCTs) [16,17,18] 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
[3,12,4].
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 [19],
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 [10,4]. 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 [20].
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 heteroge-
neous group of patients and presentations [10,4,11,15]. 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 [21]. Naturalistic studies inform clinicians, researchers
and other health professionals about treatments, as they would be
performed in practice, without exclusions and controlled con-
ditions [22].
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 [23]. 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
[22]. 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
[22]. 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
[21]. 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 [23].
143
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 [24].
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 interven-
tions (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
144 P. Carlson, K. Nicholson Perry / Seizure 45 (2017) 142–150
Table 1
Search strategy in MEDLINE via OvidSP database for identifying studies for the treatment of psychogenic non-epileptic seizures in adults.
MEDLINE via OvidSP Search
Search terms (MeSH, medical subject headings)
NEAD OR PNES OR pseudoseizure*
conversion disorder* OR
dissociative OR
functional OR hysterical seizure*, functional epilepsy,
nonepileptic AND seizure* AND therapy OR psychotherapy OR CBT
2016 and in accordance with Preferred Reporting Items for
Systematic Reviews and Meta-Analyses (PRISMA) statement for
reporting meta-analysis [25].
2.4. Data collection process
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 character-
istics (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 [26]. 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
Search options Found
Limiters 67
Published date: 1996–2016
Language: English
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.
3.2. Study characteristics
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 [14,28] 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.
 P. Carlson, K. Nicholson Perry / Seizure 45 (2017) 142–150 145

Idenficaon
Records identified through database Additional records identified
searching through other sources
(n = 164) (n = 3)

Records after duplicates removed

(n = 157)
Screening

Titles screened Records excluded

(n = 157) (n = 136)

Full-text articles excluded

Eligibility

(n = 5)
Full-text articles assessed
- Case study (1)
for eligibility
- Outcome measures did
(n = 21)
not meet eligibility (2)
- Participants did not meet
eligibility (2)

Studies included in
qualitative synthesis Full-text articles excluded,
(n = 16) Insufficient data (n = 3)
Included

Studies included in
quantitative synthesis
(meta-analysis)
(n = 13)

Fig. 1. Flow diagram of study selection.

A variety of interventions were included in the final meta-
analyses. They included 3 CBT interventions [27,28,18], 4 psycho-
dynamic treatments [13,29,9,31], 1 paradoxical intention therapy
[16], 1 mindfulness-based intervention [14], 2 psychoeducational
interventions [15,32] and 2 eclectic interventions [30,11]. The
majority of interventions ran between 10 and 24 weeks (range 3–
52 weeks) and took place in tertiary hospital settings.
Two of the included studies were RCT’s and the remainder were
observational, pre-post designs without control groups. Of the
RCT’s one study [18] featured four randomized groups, CBT-based

Table 2
Descriptive characteristics of studies meeting inclusion criteria (N = 13).

Study Design N Recruitment Gender Mean Confirmatory Concurrent

age (vEEG) epilepsy
Ataoglu et al. [16] Turkey RCT 30 Hospital ED 29F 1M 23 No (EEG only) No
Barry et al. [13] USA Before-after, non-controlled study 7 Hospital EMU 7F 45 Yes Yes
Baslet et al. [14] USA Before, after, non-controlled case 6 University Medical Centre 6F 39.6 Yes Yes
series
de Oliveira Santos et al. [29] Before-after, non-controlled study 37 Hospital Epilepsy Group 29F 8M 32 Yes Yes
Brazil
Goldstein et al. [27] UK Before-after, non-controlled study 16 Hospital Neuropsychiatry 14F 2M 34.9 Yes No
Unit
Kuyk et al. [30] Netherlands Before-after, non-controlled study 22 Hospital EMU 17F 5M 30.6 Yes No
LaFrance et al. [28] USA Before-after, non-controlled study 21 Hospital Neuropsychiatry 17F 4M 36 Yes Yes
Clinic
LaFrance et al. [18] Multicentre pilot RCT 34 Hospitals 7F 2M 37.9 Yes No
USA (9 CBT
only)
Mayor et al. [31] Before-after, non-controlled study 108 Hospitals 33F 45 Yes No
UK 14M
Mayor et al. [32] Multicentre before-after, non- 13 3 Neuroscience centres 24F 5M 37 Yes No
UK controlled study
Metin et al. [9] Turkey Before-after, non-controlled study 9 Unreported 8F 1M 22.5 Yes No
Rusch et al. [11] USA Before-after, non-controlled study 33 Medical College EMU 25F 8M 33.8 Yes Yes
Zaroff et al. [15] USA Before-after, non-controlled study 10 University Medical Centre 6F 4M 35.7 Yes No
EMU

Note: RCT = randomized controlled trial; ED = emergency department; EMU = epilepsy monitoring unit.
146 P. Carlson, K. Nicholson Perry / Seizure 45 (2017) 142–150

Table 3
Interventions and outcome characteristics of the included studies (N = 11).

Study
Ataoglu et al. [16] Turkey
Barry et al. [13] USA
Baslet et al. [14] USA
de Oliveira Santos
et al. [29] Brazil
Goldstein et al. [27] UK
Kuyk et al. [30] Netherlands
LaFrance et al. [28] USA
LaFrance et al. [18] USA
Metin et al. [6] Turkey
Mayor et al. [31] UK
Mayor et al. [32] UK
Rusch et al. [11] USA
Intervention and duration
Individual PIT (2/day for
3 weeks)
Group psychodynamic therapy
(1 week for 32 weeks)
Individual mindfulness-based
psychotherapy (12  1/week or
fortnightly)
Individual psychoanalysis
(1  50 min/week for 12 months)
Individual CBT (12  1/week or
fortnightly)
Eclectic psychotherapy
(individual, group and family)
treatment duration unreported
Individual CBT (12  1/week)
CBT-only: individual CBT
(60 min/week  12)
TAU;
CBT w/Sertraline;
Sertraline-only
Group psychoanalytic and
behavioural therapy
(12  90 min/week)
Psychodynamic IPT
(19  50 min/week or fortnight)
Manualized psychoeducation
(4  60 min/week)
6 separate forms of
psychotherapy
Treatment duration unreported
1. Identification of symptoms
with cognitive therapy and
exposure
2. Intensive psychotherapy
3. Insight-orientated psycho-
therapy
4. CBT
5. Exposure-based therapy
6. Behavioural management
strategies
Outcome measures
SF
HRSA
BDI, GSI, SC-90, SF, MINI,
SCID-D,
SIMS, HHC, HAS, BDI-II, SF,
medication
SF
Social loss
Emotional loss
Professional loss
Welfare payments
SF, IPQ, MHLC, HAS, WSAS,
fear questionnaire
SF, STAI, BAI, SF-36, DISQ,
UCL
SF, MHDS, DTA, BIS, BDI-II,
SC-90, GAF, DIS, OHS, CGI,
QOLIE-31, FAD, LIFE-RIFT
SF, MHDS, DTA, BIS, BDI-II,
SC-90, GAF, DIS, OHS, CGI,
QOLIE-31, FAD, LIFE-RIFT
SF, BDI, DES, STAI, SF-36,
TAS-20
SF, SF-36, CORE-OM, PHQ-
15
SF, SF-36, symptom
attribution, FAI, WSAS,
health service utilisation
SF
Time points
Baseline and post-
treatment
Baseline, 16 weeks and
post-treatment
Baseline, 6th session,
post-treatment
Baseline and post-
treatment
Baseline, post-treatment
and 6 month follow up
Baseline, discharge and at
6 month follow up
Baseline, 1 week, 4 weeks,
post-treatment and at 4, 8,
12 month follow up
Pre-treatment baseline,
treatment initiation
(week 2), midpoint (week
8), and post-treatment
(week 16), follow up
Baseline, 1 month, post-
treatment and at 3, 4, 6,
9 and 12 month follow up
Within 1–3 weeks of
treatment initiation, 12–
61 month follow up
Baseline and follow up
surveys
Baseline, at discharge and
at 6 month follow up.
Reported at <12 months
and >12 months
Intervention outcome (seizure
frequency)
14/15 report seizure freedom
4/7 report seizure freedom
3/6 report seizure freedom
5/6 report 50% seizure
reduction
11/37 report seizure freedom
4/16 report seizure freedom
13/16 report 50% seizure
reduction
6/22 report seizure freedom
15/22 report 50% seizure
reduction
11/17 report seizure freedom
16/17 report 50% seizure
reduction
3/9 report seizure freedom
5/9 report 50% seizure
reduction
6/9 report seizure freedom
9/9 report 50% seizure
reduction
12/47 report seizure freedom
4/13 report seizure freedom
7/13 report 50% seizure
reduction
21/26 report seizure freedom
25/26 report 50% seizure
reduction

Mean number of sessions = 9.5

Zaroff et al. [15] USA Group psychoeducational SF, CISS, DTS, CES, STAXI-2, Baseline and post- 3/4 report seizure freedom
program 3 groups (10  1h/ QOLIE-31 treatment 3/4 report 50% seizure
week) reduction

Note. BAI = Beck Anxiety Inventory; BDI = Beck Depression Inventory; BDI-II = Beck Depression Inventory-II; BIS = Barrett Impulsivity Scale; CES = Curious Experiences Survey;
CGI = Clinical Global Impressions; CISS = Coping Inventory for Stressful Situations; DIS = Dissociative Experiences Scale; DISQ = Dissociation Questionnaire; DTS = Davidson
Trauma Scale; FAD = Family Assessment Device; GAF = Global Assessment of Functioning; GSI = Global Severity Index; HAS = Hospital Anxiety and Depression Scale;
HRSA = Hamilton Rating Scale for Anxiety; IPQ = Illness Perception Questionnaire; LIFE-RIFT = Longitudinal Interval Follow-up Evaluation-Range of Impaired Functioning Tool;
MINI = Mini International Neuropsychiatric Interview; MHLC = Multi-dimensional Health Locus of Control; MHDS = Modified Hamilton Depression Scale; OHS = Oxford
Handicap Scale; QoLiE-31 = Quality of Life in Epilepsy Inventory-31; SC-90 = Symptom Checklist-90; SCID-D = Structured Clinical Interview for Diagnosis of Dissociative
Disorders; SF = Seizure frequency; STAI = State-trait Anxiety Inventory, STAXI-2 = State-trait Anxiety Expression Inventory-2; SF-36 = The Short Form (36) Health Survey; TAS-
20 = Toronto Alexithymia Scale; UCL = Utrecht Coping List; WSAS = Work and Social Adjustment Scale.

psychotherapy (N = 9), CBT-based psychotherapy combined with
anti-depressant treatment (SSRI’s; selective serotonin-reuptake
inhibitors) (N = 10), SSRI treatment alone (N = 9) and treatment as
usual (N = 10) [18]. The other RCT [16] compared paradoxical
intervention therapy (N = 15) with diazepam (N = 15).
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
 P. Carlson, K. Nicholson Perry / Seizure 45 (2017) 142–150
Fig. 2. Risk of bias summary based on the NHLBI quality assessment tools.
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.
3.4. Synthesis of results
The included studies were combined to examine the effec-
tiveness of psychological interventions in treating PNES on the
147
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.
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.
3.5. Risk of bias across studies
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 experi-
ence a reduction in seizures of 50% or more.
These results make for a markedly more positive prognosis
than reported in the literature to date [3,8,5] 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 [1,3,33] and there is only a
limited amount of quality data relating to the prevalence of those
who experience spontaneous seizure reduction [3]. 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 [5]. 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.
McKenzie et al. [8] 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
148 P. Carlson, K. Nicholson Perry / Seizure 45 (2017) 142–150

Study Prev (95% CI) % Weight

Ataoglu et al 2003 0.93 ( 0.74, 1.00) 8.02
Barry et al 2008 0.57 ( 0.19, 0.92) 6.38
Baslet et al 2015 0.50 ( 0.11, 0.89) 6.01
de Oliveira Santos et al 2014 0.30 ( 0.16, 0.46) 9.34
Goldstein et al 2004 0.25 ( 0.06, 0.50) 8.14
Kuyk et al 2008 0.27 ( 0.10, 0.48) 8.67
LaFrance et al 2009 0.65 ( 0.40, 0.86) 8.25
LaFrance 2014 0.33 ( 0.06, 0.68) 6.96
Mayor et al 2010 0.26 ( 0.14, 0.39) 9.57
Mayor et al 2013 0.31 ( 0.08, 0.59) 7.74
Metin et al 2013 0.67 ( 0.32, 0.94) 6.96
Rusch et al 2001 0.81 ( 0.63, 0.94) 8.91
Zaroff et al 2004 0.25 ( 0.00, 0.79) 5.05

Overall 0.47 ( 0.33, 0.62) 100.00

Q=54.38, p=0.00, I2=78%

0 1
Prev

Fig. 3. Forest plot of prevalence, 95% confidence intervals, and % weights for studies measuring seizure freedom.

Study Prev (95% CI) % Weight

Ataoglu et al 2003 0.93 ( 0.74, 1.00) 11.0
Baslet et al 2015 0.83 ( 0.41, 1.00) 7.1
Goldstein et al 2004 0.81 ( 0.58, 0.97) 11.2
Kuyk et al 2008 0.68 ( 0.47, 0.86) 12.5
LaFrance et al 2009 0.94 ( 0.76, 1.00) 11.5
LaFrance 2014 0.56 ( 0.22, 0.87) 8.8
Mayor et al 2013 0.54 ( 0.26, 0.80) 10.4
Metin et al 2013 1.00 ( 0.82, 1.00) 8.8
Rusch et al 2001 0.96 ( 0.84, 1.00) 13.1
Zaroff et al 2004 0.75 ( 0.21, 1.00) 5.6

Overall 0.82 ( 0.70, 0.91) 100.0

Q=21.54, p=0.01, I2=58%

0.2 0.4 0.6 0.8 1

Prevalence

Fig. 4. Forest plot of prevalence, 95% confidence intervals, and % weights for studies measuring seizure reduction of 50% or more.

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 [8]. In LaFrance et al. RCT, outcomes were measured
at16 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 [18]. 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 [21]. 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 [34]. 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 [29].
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 [14,28].
Other studies [13,11] made no such accommodations, making this
influence on treatment outcomes difficult to quantify.
 P. Carlson, K. Nicholson Perry / Seizure 45 (2017) 142–150 149

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 [11].
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 [35]. 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 [1], and yet very few people are referred
for treatment [3,33]. 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 [3].
To the authors’ knowledge, this is the first meta-analysis to
investigate the prevalence of seizure reduction following psycho-
logical 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 [10,11]. Therefore, therapies must aim to
accommodate a diverse range of psychological histories, interper-
sonal problems and range of functioning if they are to be successful
[11].
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.
Appendix A.
Funnel plot of selected studies

0.14
0.16
0.18
0.2
0.22
0.24
0.26
0.28
Standard error

0.3
0.32
0.34
0.36
0.38
0.4
0.42
0.44
0.46
0.48
0.5
0.52
1 2

Arcsin Prevalence
150 P. Carlson, K. Nicholson Perry / Seizure 45 (2017) 142–150

Appendix B. [11] *Rusch MD, Morris GL, Allen L, Lathrop L. Psychological treatment of
nonepileptic events. Epilepsy Behav 2001;2(3):277–83.
[12] Martlew J, Pulma J, Marson AG. Psychological and behavioural treatments for
adults with non-epileptic attack disorder. Cochrane Database Syst Rev 2014(2).
Database search strategy and results [13] *Barry J, Wittenberg D, Bullock K, Michaels J, Classen C, Fisher R. Group therapy
for patients with psychogenic nonepileptic seizures: a pilot study. Epilepsy
Behav 2008;13(4):624–9.
MEDLINE via OvidSP [14] *Baslet G, Dworetzky B, Perez DL, Oser M. Treatment of psychogenic
Search terms (MeSH, medical subject headings) Search options Found nonepileptic seizures: updated review and findings from a mindfulness-
based intervention case series. Clin EEG Neurosci 2015;46(1):54–64.
NEAD OR PNES OR pseudoseizure* Limiters 67
[15] *Zaroff CM, Myers L, Barr WB, Luciano D, Devinsky O. Group psychoeducation
conversion disorder* OR Published date: as treatment for psychological nonepileptic seizures. Epilepsy Behav 2004;5
1996–2016 (4):587–92.
dissociative OR Language: [16] *Ataoglu A, Ozcetin A, Icmeli C, Ozbulut O. Paradoxical therapy in conversion
english reaction. J Korean Med Sci 2003;18(4):581–4.
functional OR hysterical seizure*, functional Human studies [17] Goldstein L, Chalder T, Chigwedere C, Khondoker M, Moriarty J, Toone B, et al.
epilepsy, Cognitive-behavioral therapy for psychogenic nonepileptic seizures: a pilot
nonepileptic AND seizure* AND therapy OR RCT. Neurology 2010;74(24):1986–94.
psychotherapy OR treatment OR CBT [18] *LaFrance C, Baird G, Barry J, Blum A, Webb A, Keitner G, et al. Multicenter pilot
treatment trial for psychogenic nonepileptic seizures: a randomized clinical
trial. JAMA Psychiatry 2014;71(9):997–1005.
[19] Spring B. Evidence-based practice in clinical psychology: what is it, why it
matters; what you need to know. J Clin Psychol 2007;63(7):611–31.
[20] Reuber M, Burness C, Howlett S, Brazier J, Grunewald R. Tailored
PsychINFO psychotherapy for patients with functional neurological symptoms: a pilot
study. J Psychosom Res 2007;63(6):625–32.
[21] Stroup DF, Berlin JA, Morton SC, Olkin I, Williamson GD, Rennie D, et al. Meta-
Search terms (MeSH, subject headings) Search options Found analysis of observational studies in epidemiology: a proposal for reporting.
NEAD OR PNES OR pseudoseizure* Limiters 97 JAMA 2000;283(15):2008–12.
conversion disorder* OR Published date: [22] Fitzpatrick-Lewis D, Ciliska D, Thomas H. The methods for the synthesis of
1996–2016 studies without control groups. Hamilton, ON: National Collaborating Centre
for Methods and Tools; 2009. http;//www.nccmt.ca/pubs/non-RCT2_EN.pdf.
dissociative OR Language:
[23] Borenstein M, Hedges L, Higgins J, Rothstein HR. Introduction to meta-analysis.
english
Chichester, United Kingdom: Wiley, Ltd.; 2009.
functional OR hysterical seizure*, functional Human studies
[24] Williams A, Eccleston C, Morley S. Psychological therapies for the management
epilepsy, of chronic pain (excluding headache) in adults. Cochrane Database Syst Rev
nonepileptic AND seizure* AND therapy OR 201211:.
psychotherapy OR treatment OR CBT [25] Liberati A, Altman DG, Tetzlaff J, Mulrow C, Gotzsche PC, Ioannidis JP, et al. The
PRISMA statement for reporting systematic reviews and meta-analyses of
studies that evaluate healthcare interventions: explanation and elaboration.
Br Med J 2009;33(9):b2700.
[26] National Heart, Lung, and Blood Institute. Study quality assessment tools.
NHLBI; 2014 Available at: www.nhlbi.nih.gov/health-pro/guidelines/in-devel-
References1 op/cardiovascular-risk-reduction/tools/. (Accessed 25 June 2016).
[27] *Goldstein LH, Deale AC, Mitchell-O’Malley SJ, Toone BK, Mellers JD. An
[1] Benbadis SRMD. Nonepileptic behavioral disorders: diagnosis and treatment. evaluation of cognitive behavioral therapy as a treatment for dissociative
Continuum Lifelong Learn Neurol 2013;19(3):715–29. seizures: a pilot study. Cogn Behav Neurol 2004;17(1):41–9.
[2] Karakis I, Montouris GD, Piperidou C, Luciano MS, Meador KJ, Cole AJ. Patient [28] *LaFrance Jr. WC, Miller IW, Ryan CE, Blum AS, Solomon DA, Kelley JE, et al.
and caregiver quality of life in psychogenic non-epileptic seizures compared to Cognitive behavioral therapy for psychogenic nonepileptic seizures. Epilepsy
epileptic seizures. Seizure 2014;23(1):47–54. Behav 2009;14(4):591–6, doi:http://dx.doi.org/10.1016/j.yebeh.2009.02.016.
[3] Bodde NM, Brooks JL, Baker GA, Boon PA, Hendriksen JG, Mulder OG, et al. [29] *de Oliveira Santos N, Benute GR, Santiago A, Marchiori PE, Lucia MC.
Psychogenic non-epileptic seizures—definition, etiology, treatment and Psychogenic non-epileptic seizures and psychoanalytical treatment: results.
prognostic issues: a critical review. Seizure 2009;18(8):543–53. Rev Assoc Med Bras 2014;60(6):577–84.
[4] Reuber M, Mayor R. Recent progress in the understanding and treatment of [30] *Kuyk J, Siffels MC, Bakvis P, Swinkels WA. Psychological treatment of patients
nonepileptic seizures. Curr Opin Psychiatry 2012;25(3):244–50. with psychogenic non-epileptic seizures: an outcome study. Seizure 2008;17
[5] Reuber M, Pukrop R, Bauer J, Helmstaedter C, Tessendorf N, Elger CE. Outcome (7):595–603.
in psychogenic nonepileptic seizures: 1 to 10-year follow-up in 164 patients. [31] *Mayor R, Howlett S, Grunewald R, Reuber M. Long-term outcome of brief
Ann Neurol 2003;53(3):305–11. augmented psychodynamic interpersonal therapy for psychogenic non-
[6] Smith BJ. Closing the major Gap in PNES research: finding a home for a epileptic seizures: seizure control and health care utilization. Epilepsia
borderland disorder. Epilepsy Curr 2014;14(2):63–7. 2010;51(7):1169–76.
[7] LaFrance Jr. CW, Baker GA, Duncan R, Goldstein LH, Reuber M. Minimum [32] *Mayor R, Brown RJ, Cock H, House A, Howlett S, Smith P, et al. A feasibility
requirements for the diagnosis of psychogenic nonepileptic seizures: a staged study of a brief psycho-educational intervention for psychogenic nonepileptic
approach: a report from the International League Against Epilepsy Non- seizures. Seizure 2013;22(9):760–5.
epileptic Seizures Task Force. Epilepsia 2013;54(11):2005–18. [33] Fink P, Steen Hansen M, Søndergaard L. Somatoform disorders among first-
[8] McKenzie P, Oto M, Russell A, Pelosi A, Duncan R. Early outcomes and time referrals to a neurology service. Psychosomatics 2005;46:540–8.
predictors in 260 patients with psychogenic nonepileptic attacks. Neurology [34] Higgins JPT, Green S, editors. Cochrane handbook for systematic reviews of
2010;74(1):64–9. interventions, Version 5.1.0 [updated March 2011]. The Cochrane Collabora-
[9] *Metin SZ, Ozmen M, Metin B, Talasman S, Yeni SN, Ozkara C. Treatment with tion; 2011. . Available from: www.handbook.cochrane.org.
group psychotherapy for chronic psychogenic nonepileptic seizures. Epilepsy [35] Norcross JC. The therapeutic relationship. In: Duncan BL, Miller SD, Wampold
Behav 2013;28(1):91–4. BE, Hubble MA, editors. The heart and soul of change: delivering what works in
[10] Brown RJ, Reuber M. Towards an integrative theory of psychogenic non- therapy. 2nd ed. Washington, DC, US: American Psychological Association;
epileptic seizures (PNES). Clin Psychol Rev 2016;47:55–70. 2010. p. 113–41.

1
References marked with an asterisk indicate studies included in the meta-
analysis.