G Model

YSEIZ 2859 No. of Pages 4

Seizure xxx (2016) xxx–xxx

Contents lists available at ScienceDirect

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

The misdiagnosis of epilepsy: Appraising risks and managing

uncertainty
Maria (Meritxell) Oto
Scottish Epilepsy Centre, 20 St Kenneth Drive, Glasgow G51 4QD, United Kingdom

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

Article history: Purpose: To present evidence from the literature on the rates, underlying causes and consequences of the
Received 5 September 2016 misdiagnosis of epilepsy and place these meaningfully within a practical framework of risk appraisal and
Received in revised form 4 November 2016 managed diagnostic uncertainty towards informing a clinical practice that might make misdiagnosis less
Accepted 30 November 2016
likely.
Available online xxx
Method: Narrative review.
Results: Misdiagnosis of epilepsy remains common and the consequences for the individual significant.
Keywords:
Evidence and critical appraisal are presented as regards the absolute level of risk associated with the false
Misdiagnosis
Epilepsy
positive diagnosis epilepsy, and reasons as to why those risks need to be appraised against the risks
Risk appraisal associated to false negative diagnosis.
Diagnostic uncertainty Conclusions: Diagnostic error is not entirely avoidable and a degree of uncertainty, and perforce risk, is
intrinsic to the diagnostic process of epilepsy.
The risks of a false negative diagnosis of epilepsy must be appraised against the also significant risks of a
false positive diagnosis.
Crown Copyright © 2016 Published by Elsevier Ltd on behalf of British Epilepsy Association. All rights
reserved.

1. Introduction misdiagnosis [4]. As such, as well as presenting evidence from the

Over the last 20 years there have been significant advances in
epilepsy research in terms of the identification of underlying
causes and mechanisms and the development of more tolerable
and efficacious treatments. However, none of these advances are
practically meaningful without an accurate diagnosis. Epilepsy is
still overwhelmingly a clinical diagnosis and rates of misdiagnosed
epilepsy remain stubbornly high.
Consensus statements and guidelines consistently recommend
early availability of and referral to specialist epilepsy services as a
way of addressing this issue [1,2]. However the diagnosis of
epilepsy can be challenging even for experienced clinicians [3]. The
difficulty arises not so much from a greater or lesser ability to
recognise epilepsy, but in the particular problems in assessment of
risk and management of uncertainty specific to situations where
epilepsy may be a possibility but a final diagnosis has to await
further confirmation. The perceived risk of not treating, even in a
circumstance where the probability of epilepsy is low, mitigates
against circumspection and encourages practice that results in
E-mail address: meritxelloto@hotmail.com (M.(. Oto).
literature on the rates, underlying causes and consequences of the
misdiagnosis of epilepsy, the intention of this paper is to place
these meaningfully within a practical framework of risk appraisal
and managed diagnostic uncertainty towards informing a clinical
practice that might make misdiagnosis less likely.
2. Misdiagnosis
2.1. Prevalence of misdiagnosis
Reported misdiagnosis rates vary substantially with estimates
ranging between 2% and 71%. This wide variation reflects the
heterogeneity across studies in terms of setting, inclusion of
patients with refractory epilepsy, diagnostic criteria, diagnostic
methods and the experience of the referring clinician [5–15]. With
the exception of the studies below, such studies are likely to be
confounded, and their populations too highly selected, to derive a
‘base rate’ of misdiagnosis that would inform practitioners
addressing the needs of patients presenting following an
apparently first seizure, or the majority of those diagnosed with
epilepsy who are never deemed ‘treatment refractory’.

http://dx.doi.org/10.1016/j.seizure.2016.11.029
1059-1311/Crown Copyright © 2016 Published by Elsevier Ltd on behalf of British Epilepsy Association. All rights reserved.

Please cite this article in press as: M.M. Oto, The misdiagnosis of epilepsy: Appraising risks and managing uncertainty, Seizure: Eur J Epilepsy
(2016), http://dx.doi.org/10.1016/j.seizure.2016.11.029
G Model
YSEIZ 2859 No. of Pages 4
2 M.M. Oto / Seizure xxx (2016) xxx–xxx
By contrast two purposely designed studies addressed the
prevalence of misdiagnosis within the community. In the UK an
early study assessing the prevalence of epilepsy within the
community, reviewed the diagnosed of 214 patients from seven
general practice surgeries. Following review by an epilepsy
specialist, alternative causes for the attack disorder were found
in 49 (23%) mostly a cardiovascular cause or psychogenic non-
epileptic seizures (PNES) [8]. Another UK based study identified
275 patients with epilepsy treated with antiepileptic drugs (AED)
from 26 general practices. All patients were reviewed by two
experienced epilepsy specialists who concluded that the diagnosis
of epilepsy was in doubt in 16.3% of patients [9].
Thus the misdiagnosis rate in unselected patients with a
diagnosis of epilepsy may be in the region of 20%.
Higher rates of misdiagnosis are found in adults or children
with apparently treatment refractory epilepsy referred to second-
ary care in and outside the UK [5]. When Smith et al.
retrospectively reviewed the diagnosis of relatively unselected
patients referred to an epilepsy clinic, the overall misdiagnosis rate
amongst the 184 patients with the diagnosis of epilepsy and
treated with antiepileptic drugs was 26.1% (46/184). In this study
the most common conditions to be mistaken for epilepsy were
PNES and syncope and more than half of the patients were on
medication [10]. A further study, specifically excluding patients
with suspected PNES, still found that of 74 patients referred to an
epilepsy clinic, half of whom were deemed to have refractory
epilepsy, 41.9% had an alternative, most commonly syncopal or
cardiovascular, cause of their symptoms [11]. A retrospective
survey of children admitted to a tertiary centre in Denmark found
that 30% of the children referred with a definite diagnosis, did not
have epilepsy [12].
Even more selected case series from secondary care that would
be expected to include more unusual presentations generally
confirm syncope followed by PNES as the most common conditions
underlying misdiagnosis [5]. The ‘typical’ scenario of epilepsy
misdiagnosis conforms to a relatively prosaic narrative of short
lived and in all probability benign ‘collapses’ whose resolution may
be mistaken for a response to AED. The possibility exists that a
substantial minority of patients with apparently remitted epilepsy
on treatment do not have epilepsy.
Finally it is important to point out that although higher
proportions of misdiagnosis have been found amongst non-
experts, figures of over 20% have also been reported amongst
patients under the care of specialists and referred on to tertiary
centres [13,14].
2.2. Consequences of misdiagnosis
In essence, a misdiagnosis of epilepsy carries with it all the
secondary handicaps and limitations of a diagnosis of epilepsy in
terms of stigma and social marginalisation, lifestyle limitation,
employment and driving restrictions, and the side effects and
potential teratogenic effects of AEDs [3,5,15]. It seems reasonable
to speculate that clinician’s insufficient understanding of the
profound implications of a diagnosis of epilepsy in and of itself
contributes to an over-readiness to make the diagnosis.
Once established, the diagnosis of epilepsy is not readily
challenged or reviewed even amongst specialists. Patients
eventually correctly diagnosed as having PNES will on average
have acquired their misdiagnosis, and its consequences, 7–10 years
previously [16].
Some populations court particular risk from misdiagnosis;
specifically patients with unrecognised cardiogenic syncope or
patients suffering from PNES.
A small minority of patients with apparent syncope will
transpire to have an underlying liability to serious arrhythmia,
Please cite this article in press as: M.M. Oto, The misdiagnosis of epilepsy: Appraising risks and managing uncertainty, Seizure: Eur J Epilepsy
(2016), http://dx.doi.org/10.1016/j.seizure.2016.11.029
often but not always associated with ECG changes and if untreated
associated with a high mortality rate [17,18]. Speculative prescrip-
tion of AED in this circumstance has potentially disastrous results.
Psychiatric morbidity in PNES populations is substantial and
worsens the prognosis. A misdiagnosis of epilepsy not only
misattributes the primary psychological nature of the attacks but
also prevents appropriate treatment of the substantial associated
psychiatric morbidity [16].
Particular risks of iatrogenic harm are incurred by PNES
patients presenting with prolonged attacks misattributed to
apparent status when this leads to inappropriate use of high
doses of intravenous medication or even admission to the Intensive
Care Unit and intubation, with all the morbidity that this entails
[19].
The economic consequences of misdiagnosis are also signifi-
cant; figures form NICE guidelines estimated the direct total
national medical costs between 164 and 188 million pounds [1]. As
well as the costs of an erroneous diagnosis, “undiagnosing”
epilepsy is also costly since reversing a diagnosis is at times more
complicated and patients may require video EEG monitoring or
inpatient admission for a diagnostic withdrawal of medication
[10].
2.3. Reasons for the misdiagnosis of epilepsy
As in medicine more generally there are no short cuts to an
accurate clinical diagnosis. Epilepsy misdiagnosis however, of all
medical missteps, seems to occur within a particular matrix of
factors that discourage circumspection and encourage immediate
diagnosis on a basis of an inadequate history, traditional but
unreliable ‘red flags’, over-interpretation or misuse of medical
investigations (mainly EEG), and the inaccurate perception that the
immediate clinical course will be grave if intervention is delayed.
The single most important factor in epilepsy misdiagnosis is the
failure to appreciate the importance of a thorough and expert
clinical history and its corroboration by a witness description [10].
Rather than diagnostic insight, ‘expertise’ in this circumstance
applies more to perseverance in seeking a history and witness
description as well as reservation of judgement when these are
unavailable.
To complicate matters, epileptic seizures can manifest in many
ways and although there are constellations of features that would
alert the clinician to the possibility of the diagnosis one way or
another, there is no single pathognomonic semiological feature
that would in isolation absolutely endorse a diagnosis of epilepsy
or non-epilepsy. Unfortunately many of the often rehearsed ‘red
flags’ of clinical tradition (self-injury, attacks arising from apparent
sleep, urinary incontinence) have been shown to have little or no
discriminant value, and for the most part are actively misleading if
taken in isolation [16].
Laboratory investigations are of limited value in epilepsy
diagnosis at the level of the individual patient and in the context of
practical decision making. None has sufficient sensitivity or
specificity to confirm or rule out a diagnosis.
Interictal EEG is a valuable test in the further investigation of an
established diagnosis of epilepsy, however it has little role in
diagnosis per se. Overreliance on and misinterpretation of routine
EEGs has been found to be a contributory factor in the misdiagnosis
of epilepsy [10].
Interpreting EEG reports can be as challenging as interpreting
the EEG itself. Non-specific EEG abnormalities are not uncommon
in the general population and more frequently observed in
populations at higher risk of manifesting non-epileptic attacks,
especially those prescribed some types of psychotropic medica-
tion. To the inexperienced a subsequent report of ‘non-specific
focal slowing’, might be sufficient to consolidate suspicion into
G Model
YSEIZ 2859 No. of Pages 4
M.M. Oto / Seizure xxx (2016) xxx–xxx
diagnosis [21,22]. A study investigating EEG abnormalities in
patients with PNES found that over 50% of patients had had one or
more EEGs reported as abnormal [23].
More complex and time consuming investigations, such as
simultaneous video EEG recording, have an undeniable role as a
diagnostic aid. However their reliability also depends on an
accurate clinical history and its usefulness is limited to situations
where there is a reasonable probability of capturing a typical
attack.
The role of neuroimaging in diagnosing epilepsy has clear
limitations. An abnormal brain scan does not automatically
confirm the diagnosis of epilepsy. MRI scans in particular can also
be a source of error, as non-specific abnormalities or incidental
findings may be misinterpreted as the cause for the attack disorder
[22].
3. Uncertainty
3.1. Clinician’s perception of risk and management of diagnostic
uncertainty
Epilepsy is associated with an increased risk of morbidity and
mortality. However there are also significant risks attached to
misdiagnosing epilepsy and these need equal consideration.
Unfortunately, over and above concerns at the lack of access to
trained opinion, reported rates of misdiagnosis still suggest that
clinicians perceive missing a diagnosis of epilepsy and not treating
seizures as being riskier than “erring on the side of caution” and
treating despite insufficient clinical information and the risks of
exposure to unnecessary AED.
Although probably informed by a range of concerns, unwar-
ranted treatment of a poorly sustained or uncertain diagnosis of
epilepsy seems to have as its basis a perception of risks that can
only be addressed by prompt AED treatment and whose course will
be one of inevitable deterioration, increased morbidity and even
mortality otherwise.
As already mentioned, mortality rates amongst people with
epilepsy are higher than in the general population. However, most
of the increased mortality is attributable to the underlying
comorbidities, and only a small percentage relates to the direct
consequences of seizures [24]. Early treatment with AED reduces
early recurrence of seizures but does not change the long term
outcome and has no influence on mortality [25,26]. It has been
argued that there is a need to review the risks and benefits of
routine prescription of AED for newly diagnosed patients in the
light of a finding that spontaneous remission of the epilepsies may
occur in up to 30% of patients [27,28].
Therapeutic trials of AED are undertaken on occasion to resolve
uncertainty. In fact most paroxysmal attack disorders will follow a
course far too variable to distinguish a response, and as the
majority are benign and self-limiting an apparent but spurious
response to AED is more likely than not. A lack of response to AED is
often listed as a diagnostic pointer to PNES but a significant
proportion of patients with PNES, given the near universal
psychiatric comorbidity, would be expected to demonstrate the
substantial placebo response common to all non-psychotic
psychiatric disorders [16].
Far from intending to minimise the significant risks
associated with epilepsy, the points raised in this section
attempt to contextualise these risks within the population of
patients presenting for the first time with paroxysmal attacks of
uncertain cause and at most risk of misdiagnosis and all that
this entails.
Please cite this article in press as: M.M. Oto, The misdiagnosis of epilepsy: Appraising risks and managing uncertainty, Seizure: Eur J Epilepsy
(2016), http://dx.doi.org/10.1016/j.seizure.2016.11.029
3
3.2. Managing uncertainty
An accurate diagnosis of epilepsy is often possible following a
detailed history and a good eyewitness account. However there are
situations where absent or vague eye witness descriptions or
unusual clinical presentations of attacks make reaching a definite
diagnosis challenging or impossible at least in the short term. As
already mentioned in this circumstance diagnostic tests may not
be helpful and potentially misleading.
As doctors we are trained to accurately diagnose disease, adopt
clear diagnostic labels and offer appropriate treatment to patients.
In general clinicians feel uncomfortable with diagnostic uncer-
tainty. Committing to a definite diagnosis eliminates this
uncertainty but from that moment determines the future
management of the attack disorder and forecloses the consider-
ation of an alternative diagnosis.
Delaying the diagnosis of epilepsy until one is absolutely certain
is clearly not possible. Even when a diagnosis is made by a panel of
experts the false positive rate is around 5% [6]. Ultimately,
misdiagnosis is not merely difficult to avoid; it is inherent. For this
reason clinicians should always be mindful that diagnoses may be
wrong and adopt a practice where diagnoses are routinely
questioned and reviewed.
As a reflection of a greater ability to recognise limitations in the
clinical information and incorporate an awareness of the risk of
misdiagnosis, experts are more likely to admit diagnostic
uncertainty than non specialists. It has been suggested that
adopting the label of “possible epilepsy” or “unclassified paroxys-
mal event” as a positive diagnostic category could reduce the rate
of misdiagnosis by quasi-operationally incorporating diagnostic
uncertainty as a legitimate option and avoiding precipitate
diagnosis based on incomplete information [15,29].
4. Conclusion
Misdiagnosis of epilepsy remains common and the consequen-
ces for the individual significant [5]. Improved services, training,
and in particular early access to specialist assessment, as
recommended by expert authorities, are unexceptionably likely
to lead to a reduction in misdiagnosis as part of a general
improvement in epilepsy care. However other factors have to be
considered. Epilepsy by its nature presents suddenly and
dramatically, is initially assessed outwith specialist services, and
even within specialist settings is associated with a singularly
complex matrix of perceived risk and diagnostic uncertainty [15].
These factors inevitably militate against a considered and accurate
diagnosis in a substantial number of cases.
In this article I have attempted to present what is known on
epilepsy misdiagnosis within the context of risk assessment,
acknowledged uncertainties and practical decision making that
impacts directly on patient care. Often, rather than the ability to
clinically recognise epilepsy, the decision making errors that lead
to misdiagnosis are more about poor management of the
possibility of epilepsy, even when the pre-test probability is
acknowledged to be low, because of a misperception of the dangers
of diagnostic delay and an exaggerated perception of the benefits
of proceeding on an uncertain diagnosis.
Acknowledging that uncertainty and diagnostic error cannot be
completely eliminated, encouraging a clinical stance where
deferring a diagnosis is a legitimate option, and routine review
and reconsideration of apparently secure diagnoses, would make
the process safer and reduce the chance of erroneous and harmful
misdiagnosis.
G Model
YSEIZ 2859 No. of Pages 4
4 M.M. Oto / Seizure xxx (2016) xxx–xxx
Conflict of interest statement
None.
References
[1] NICE. The epilepsies: diagnosis and management of the epilepsies in children
and young people in primary and secondary care. National Institute for Clinical
Excellence (NICE) Clinical Guideline 20; 2004.
[2] SIGN. Diagnosis and management of epilepsies in children and young people.
Scottish Intercollegiate Guidelines Network (SIGN); 2015.
[3] Chadwick D, Smith D. The misdiagnosis of epilepsy: the rate of misdiagnosis
and wide treatment choices are arguments for specialist care of epilepsy.
(Editorials). Br Med J 2002;324(March (7336)):495–7.
[4] Van Donselaar CA, Stroink H, Arts WF. How confident are we of the diagnosis of
epilepsy? Epilepsia 2006;47(October (s1)):9–13.
[5] Xu Y, Nguyen D, Mohamed A, Carcel C, Li Q, Kutlubaev MA, et al. Frequency of a
false positive diagnosis of epilepsy: a systematic review of observational
studies. Seizure 2016;(October):167–74.
[6] Stroink H, Van Donselaar CA, Geerts AT, Peters AC, Brouwer OF, Arts WF. The
accuracy of the diagnosis of paroxysmal events in children. Neurology 2003;60
(March (6)):979–82.
[7] Josephson CB, Rahey S, Sadler RM. Neurocardiogenic syncope: frequency and
consequences of its misdiagnosis as epilepsy. Can J Neurol Sci 2007;34(May
(02)):221–4.
[8] Scheepers B, Clough P, Pickles C. The misdiagnosis of epilepsy: findings of a
population study. Seizure 1998;7(October (5)):403–6.
[9] Leach JP, Lauder R, Nicolson A, Smith DF. Epilepsy in the UK: misdiagnosis,
mistreatment, and undertreatment?: the Wrexham area epilepsy project.
Seizure 2005;14(October (7)):514–20.
[10] Smith D, Defalla BA, Chadwick DW. The misdiagnosis of epilepsy and the
management of refractory epilepsy in a specialist clinic. QJM 1999;92(January
(1)):15–23.
[11] Zaidi A, Clough P, Cooper P, Scheepers B, Fitzpatrick AP. Misdiagnosis of
epilepsy: many seizure-like attacks have a cardiovascular cause. J Am Coll
Cardiol 2000;36(July (1)):181–4.
[12] Uldall P, Alving J, Hansen LK, Kibaek M, Buchholt J. The misdiagnosis of
epilepsy in children admitted to a tertiary epilepsy centre with paroxysmal
events. Arch Dis Child 2006;91(March (3)):219–21.
[13] Alsaadi TM, Thieman C, Shatzel A, Farias S. Video-EEG telemetry can be a
crucial tool for neurologists experienced in epilepsy when diagnosing seizure
disorders. Seizure 2004;13(January (1)):32–4.
Please cite this article in press as: M.M. Oto, The misdiagnosis of epilepsy: Appraising risks and managing uncertainty, Seizure: Eur J Epilepsy
(2016), http://dx.doi.org/10.1016/j.seizure.2016.11.029
[14] Benbadis SR, O'Neill E, Tatum WO, Heriaud L. Outcome of prolonged video-EEG
monitoring at a typical referral epilepsy center. Epilepsia 2004;45(September
(9)):1150–3.
[15] Chowdhury FA, Nashef L, Elwes RD. Misdiagnosis in epilepsy: a review and
recognition of diagnostic uncertainty. Eur J Neurol 2008;15(October
(10)):1034–42.
[16] Oto M, Reuber M. Psychogenic non-epileptic seizures: aetiology, diagnosis and
management. Adv Psychiatr Treat 2014;20(January (1)):13–22.
[17] Kapoor WN, Karpf M, Wieand S, Peterson JR, Levey GS. A prospective
evaluation and follow-up of patients with syncope. New Eng J Med 1983;309
(July (4)):197–204.
[18] Linzer M, Grubb BP, Ho S, Ramakrishnan L, Bromfield E, Estes NM.
Cardiovascular causes of loss of consciousness in patients with presumed
epilepsy: a cause of the increased sudden death rate in people with epilepsy?
Am J Med 1994;96(February (2)):146–54.
[19] Walker MC, Howard RS, Smith SJ, Miller DH, Shorvon SD, Hirsch NP. Diagnosis
and treatment of status epilepticus on a neurological intensive care unit. QJM
1996;89(December (12)):913–20.
[21] Benbadis SR, Tatum WO. Over-interpretation of EEGs and misdiagnosis of
epilepsy. J Clin Neurophysiol 2003;20(February (1)):42–4.
[22] Fattouch J, Di Bonaventura C, Strano S, Vanacore N, Manfredi M, Prencipe M,
et al. Over-interpretation of electroclinical and neuroimaging findings
insyncopes misdiagnosed as epileptic seizures. Epileptic Disord 2007;9(May
(2)):170–3.
[23] Reuber M, Fernández G, Bauer J, Singh DD, Elger CE. Interictal EEG
abnormalities in patients with psychogenic nonepileptic seizures. Epilepsia
2002;43(September (9)):1013–20.
[24] Hitiris N, Mohanraj R, Norrie J, Brodie MJ. Mortality in epilepsy. Epilepsy Behav
2007;10(May (3)):363–76.
[25] Marson A, Jacoby A, Johnson A, Kim L, Gamble C, Chadwick D, et al. Immediate
versus deferred antiepileptic drug treatment for early epilepsy and single
seizures: a randomised controlled trial. The Lancet 2005;365(June
(9476)):2007–13.
[26] Lhatoo SD, Johnson AL, Goodridge DM, MacDonald BK, Sander JW, Shorvon SD.
Mortality in epilepsy in the first 11 to 14 years after diagnosis: multivariate
analysis of a long-term, prospective, population-based cohort. Annals Neurol
2001;49(March (3)):336–44.
[27] Kwan P, Sander JW. The natural history of epilepsy: an epidemiological view. J
Neurol Neurosurg Psychiatr 2004;75(October (10)):1376–81.
[28] McIntosh AM, Berkovic SF. Treatment of new-onset epilepsy: seizures beget
discussion. Lancet 2005;365(June (9476)):1985–6.
[29] Beach R, Reading R. The importance of acknowledging clinical uncertainty in
the diagnosis of epilepsy and non-epileptic events. Arch Dis Child 2005;90
(December (12)):1219–22.