Evidence review: Diagnosis of epilepsies

National Guideline Centre (UK)

NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

Evidence review: Diagnosis of epilepsies

Epilepsies in children, young people and adults: diagnosis and management

Evidence review 3

NICE Guideline, No. 217

Authors

National Guideline Centre (UK).

London: National Institute for Health and Care Excellence (NICE); 2022 Apr.

ISBN-13: 978-1-4731-4513-9

1. Diagnosis of epilepsy

1.1. Introduction

Epilepsy is diagnosed in people who have had two unprovoked seizures or in those who have had one seizure, but there are features to suggest a high risk of recurrence. Confirming and diagnosing epilepsy can be difficult and relies heavily on the description of seizures. Many different conditions can cause epilepsy, although often, an underlying cause is not identified. Conditions associated with epilepsy include brain infections, brain injury, brain malformations, metabolic disorders, stroke, dementia and underlying genetic abnormalities. This evidence review evaluates the accuracy of a range of diagnostic strategies to optimise diagnosis and assessment in people who may have epilepsy.

1.2. Review question: What is the most accurate approach for 1) diagnosis of epilepsy and 2) differentiation between types of epilepsy?

1.2.1. Summary of the protocol

For full details see the review protocol in Appendix A.

1.2.2. Methods and process

This evidence review was developed using the methods and process described in Developing NICE guidelines: the manual.¹³⁸ Methods specific to this review question are described in the review protocol in Appendix A.

1.2.3. Effectiveness evidence

1.2.3.1. Included studies

77 studies were included in this diagnostic accuracy review⁶^, ⁷^, ¹⁰^, ¹¹^, ¹⁶^, ²⁰^, ²⁵^, ²⁶^, ²⁸^, ³⁹^, ⁴³^, ⁵⁶^, ⁵⁸^, ⁶⁰^–⁶²^, ⁶⁴^, ⁶⁵^, ⁶⁸^, ⁶⁹^, ⁷³^–⁷⁵^, ⁸¹^, ⁸²^, ⁸⁴^, ⁸⁶^, ⁸⁷^, ⁹⁰^, ⁹²^, ⁹⁴^, ⁹⁶^, ⁹⁷^, ⁹⁹^, ¹⁰⁰^, ¹⁰²^, ¹⁰⁷^, ¹⁰⁹^, ¹¹¹^, ¹¹⁴^, ¹¹⁶^, ¹²⁴^, ¹²⁵^, ¹³¹^, ¹³²^, ¹³⁶^, ¹³⁷^, ¹⁴³^–¹⁴⁶^, ¹⁵⁸^–¹⁶¹^, ¹⁶³^, ¹⁶⁶^, ¹⁷¹^, ¹⁷⁶^, ¹⁷⁷^, ¹⁷⁹^–¹⁸¹^, ¹⁸⁴^, ¹⁸⁶^, ¹⁹¹^, ¹⁹³^, ¹⁹⁴^, ¹⁹⁶^, ¹⁹⁹^, ²⁰⁰^, ²⁰³^, ²⁰⁵^, ²⁰⁹^, ²¹³^, ²¹⁵^, ²¹⁶. The characteristics of these studies are summarised in Table 2, and evidence from these studies are summarised in the clinical evidence summaries (Table 3 to Table 16). Further details are available in the study selection flow chart in Appendix C.1, sensitivity and specificity forest plots and receiver operating characteristics (ROC) curves in Appendix E, and study evidence tables in Appendix D.

Analysis was stratified by the population requiring diagnostic attention: 1) children and adults with suspected epilepsy, or 2) children and adults with definite epilepsy, where uncertainty remains as to the type of epilepsy. The aim of most studies was not to differentiate between different types of epilepsy but to differentiate epilepsy from no epilepsy, and only two studies⁶⁴^, ¹³² fitted into the latter stratum. Some studies⁶^, ⁷^, ⁵⁸^, ⁶⁸^, ⁸²^, ⁸⁶^, ¹⁰⁰^, ¹¹⁴^, ¹²⁴^, ¹³⁶^, ¹⁵⁹^, ¹⁶³^, ¹⁸⁶^, ²⁰⁰^, ²⁰⁵ evaluated an index test in an epilepsy population that was restricted to a certain type (such as temporal lobe epilepsy). However, the findings from these were evaluated in the first stratum because the ability of the index test to differentiate between the specific type and no epilepsy was being assessed; that is, these studies were not differentiating between different types of epilepsy. The sub-types of epilepsy included status epilepticus (SE), non-convulsive status epilepticus (NCSE), temporal lobe epilepsy (TLE), frontal lobe epilepsy (FLE), partial epilepsy, focal epilepsy, generalised epilepsy, generalised genetic epilepsy, autoimmune epilepsy, and absence seizures. These categories overlap but reflected the classification systems of the included papers. The types of epilepsy are highlighted in the results tables where appropriate.

For each of the above strata, pre-hoc sub-grouping strategies (conditional on observed heterogeneity) were:

Age: <2, 2-11, 11-18, 18-55, >55
Learning disability / no learning disability
Head injury / no head injury
Gender
Type of epilepsy
Person carrying out the index tests

Sub-grouping was only considered for the two meta-analyses concerning interictal routine EEG and postictal stertorious breathing, as these were the only analyses where heterogeneity was evident. However, none of the protocol sub-grouping strategies were able to ‘explain’ heterogeneity (by yielding homogenous results within each sub-group) in either meta-analysis. Only 5 diagnostic meta-analyses were possible because at least 3 studies are required for a valid pooling of results, and for most index tests, only one or two studies were available.

Several studies did not recruit consecutively from the population under clinical suspicion but instead employed a case-control strategy where they recruited people with gold-standard confirmed epilepsy, as well as others with specific differential diagnoses that were also confirmed by a gold-standard method. In the majority of cases, the differential diagnosis was psychogenic non-epileptic seizures (PNES). These studies have been highlighted in the analysis because this approach has an important impact on the interpretation of specificity results. Specificity measures may have been affected because the propensity towards false positives may be associated with the characteristics of the non-epilepsy group. For example, a group of people with PNES may be more likely (or less likely) to yield false-positive results than a more random group of people who were initially suspected of epilepsy. However, the sensitivity of the index test will not be affected by this approach, as sensitivity will depend solely on the response of the group who have gold-standard confirmed epilepsy. It should also be mentioned that in some papers, the target condition for diagnosis was not epilepsy but PNES (for example, the paper expressed the accuracy for detecting PNES, rather than epilepsy). These studies were still included because it was possible to convert the results to those that would have been observed had epilepsy been the target condition. This was achieved in most cases by simply exchanging the sensitivity and specificity measures. However, this could only occur if the study was restricted to epilepsy and PNES. If the non-PNES group comprised groups additional to those with epilepsy, then it was not possible to extrapolate the sensitivity and specificity for the detection of epilepsy.

Gold standards varied between studies, but the protocol had allowed for a variety of approaches. For inclusion, a study needed to have a sufficient description of the gold standard to permit the assumption that it was the best method available to the researchers when doing the study. If a study gave no indication of the methods used to decide on the gold standard diagnosis, it was excluded.

For the purposes of decision-making, sensitivity and specificity were given equal priority. For a test to be able to be recommended as a diagnostic strategy, it would normally need to exceed 0.9 for both sensitivity and specificity, and values below 0.6 would be regarded as clinically useless. Poor sensitivity indicates that an unacceptably large number of patients with epilepsy would not be diagnosed as having epilepsy (false negatives), and might remain untreated. Poor specificity means that an unacceptable proportion of those without epilepsy would be misdiagnosed as having epilepsy (false positives), leading to unnecessary and potentially harmful treatments, as well as unwarranted anxiety.

Because of the large numbers of included studies and results, it was necessary to categorise the index tests in the results tables. This categorisation is arbitrary, is not based on a pre-defined system, and has no impact on the strength of results. The 12 categories of index test are: symptoms/signs/semiology; serum measures; ECG testing; Imaging tests; EEG tests; MEG/TMS tests; psychological measures; linguistic tests; EMG tests; accelerometer testing; clinical impression at admission based on a variety of data; and miscellaneous methods.

Finally, it is important to point out that this review question covers the 6 questions previously in the scope:

1.2.: Diagnostic accuracy of signs and symptoms
1.3.: What is the role of electrocardiograph (ECG) in distinguishing between seizures and non-seizure events after a first seizure or seizure like episode?
1.4.: What is the diagnostic accuracy of electroencephalogram (EEG) (including specific EEG techniques) in distinguishing between seizures and non-seizure events?
1.5.: What is the diagnostic accuracy of EEG (including specific EEG techniques) in identifying specific seizure types and epilepsy syndromes?
1.6.: What is the diagnostic accuracy of EEG (including specific EEG techniques) in assessing the likelihood of seizure recurrence after a first seizure

These questions were combined to ensure that we could capture testing strategies that combined elements from more than one of the original questions. For example, a testing strategy utilising signs and symptoms combined with EEG might not have fitted into either question 1.2 or 1.4. A combined question with a more open scope also allowed a greater range of index-test types to be included. Previously, using the 6 separate questions, the index test categories of imaging, magnetoencephalography, psychological tests, serum tests, EMG and accelerometer testing would not have been included, whereas they are now being considered in the review.

1.2.3.2. Excluded studies

Please see the excluded studies list in Appendix I.

1.2.4. Summary of clinical studies included in the evidence review

1.2.5. Quality assessment of clinical studies included in the evidence review

For measurement of imprecision, clinical decision thresholds were set at 0.90 [above which may be willing to recommend] and 0.60 [below which is clinically unhelpful (for both sensitivity and specificity).

STRATUM 1: Detection of any epilepsy (differentiation from no epilepsy)

STRATUM 2: Differentiation between specific types of epilepsy

See Appendix D for full evidence tables.

1.2.6. Economic evidence

1.2.6.1. Included studies

No health economic studies were included.

1.2.6.2. Excluded studies

No relevant health economic studies were excluded due to assessment of limited applicability or methodological limitations.

See also the health economic study selection flow chart in Appendix G.

1.2.7. Economic model

This area was not prioritised for new cost-effectiveness analysis.

1.2.8. Unit costs

Relevant unit costs are provided below to aid consideration of cost effectiveness. All unit costs sourced from NHS reference costs 2018-2019¹⁴⁰^REF. The unit costs included are EEG, ECG, MRI, CT, PET, SPECT and neurology appointments.

Other unit costs of relevance include blood tests (full blood count, liver function, glucose, and electrolytes) and venous blood gas (for accident and emergency admissions only). NHS reference costs list directly accessed pathology services unit costs as between £1 and £8.

1.3. Review question: What is the most clinically and cost-effective approach for diagnosis of epilepsies?

1.3.1. Summary of the protocol

For full details see the review protocol in 0.

1.3.2. Methods and process

This review is a review of trials that have compared health-related outcomes in people randomised to different diagnostic tests. Tests may differ in their influence on later health outcomes through stimulating a more or less appropriate treatment approach by virtue of their differing diagnostic accuracies. In addition, tests may influence outcomes such as quality of life through other effects unrelated to accuracy, such as patient comfort, duration of testing or length of time for results. Whilst accuracy is not measured directly in such randomised trials, the advantage of such studies is that they demonstrate clinical efficacy. In contrast a diagnostic accuracy study can only demonstrate the intrinsic diagnostic accuracy of the test and is unable to show how that accuracy affects health outcomes. However, such randomised trials are not commonly undertaken, and may provide equivocal results, and so a diagnostic accuracy review was also undertaken.

This evidence review was developed using the methods and process described in Developing NICE guidelines: the manual. Methods specific to this review question are described in the review protocol in appendix A and the methods document.

Declarations of interest were recorded according to NICE’s conflicts of interest policy.

1.3.3. Effectiveness evidence

1.3.3.1. Included studies

Two studies were included in the review.¹⁶⁵^, ²¹⁸ These are summarised in Table 2 below. Evidence from these studies is summarised in the clinical evidence summary in Table 3.

Both included studies comprised patients undergoing emergency care due to reduced consciousness. They may therefore lack some applicability to the target population of this review, who require a diagnostic work-up because they have a clinical history suggestive of epilepsy.

See also the study selection flow chart in Appendix C, study evidence tables in Appendix D, forest plots in Appendix E and GRADE tables in Appendix F.

1.3.3.2. Excluded studies

See the excluded studies list in Appendix K.

1.3.4. Summary of studies included in the effectiveness evidence

See Appendix D for full evidence tables.

1.3.5. Summary of the effectiveness evidence

See Appendix F for full GRADE tables

1.3.6. Economic evidence

1.3.6.1. Included studies

No health economic studies were included.

1.3.6.2. Excluded studies

No relevant health economic studies were excluded due to assessment of limited applicability or methodological limitations.

See also the health economic study selection flow chart in Appendix G.

1.3.7. Economic model

This area was not prioritised for new cost-effectiveness analysis.

1.3.8. Unit costs

Relevant unit costs are provided below to aid consideration of cost effectiveness. All unit costs sourced from NHS reference costs 2018-2019¹⁴⁰. The unit costs included are EEG, ECG, MRI, CT, PET, SPECT and neurology appointments.

1.4. Evidence statements

1.4.1. Effectiveness/Qualitative

None.

1.4.2. Economic

No relevant economic evaluations were identified.

1.5. The committee’s discussion of the evidence

1.5.1. The outcomes that matter most

1.5.1.1. Diagnostic accuracy review

For the diagnostic accuracy review the outcomes were sensitivity and specificity. The committee considered that both outcomes are important because the harms of reduced sensitivity and the harms of reduced specificity are similar in the context of epilepsy diagnosis. Reduced sensitivity means that some people who truly have epilepsy will not be successfully detected by the index test. These people will therefore remain undiagnosed and untreated, which can have serious consequences. Reduced specificity means that some people who truly do not have epilepsy will be misdiagnosed as having epilepsy. These people may receive unnecessary treatments, where possible harms are not ameliorated by benefits.

The committee agreed that ideally the thresholds for recommendation of index tests should be a sensitivity of 0.9 and a specificity of 0.9. Use of any test achieving this threshold would mean that no more than 10% of people with epilepsy would suffer a missed diagnosis (false negatives), and that no more than 10% of people without epilepsy would be misdiagnosed with epilepsy (false positives). Because it was thought that the harms of reduced specificity may be slightly less dangerous than the harms of reduced sensitivity, it was agreed some leeway might be made in cases where a test had specificity slightly below 0.9. However, it was agreed that sensitivity had to exceed 0.9 to allow recommendation.

1.5.1.2. RCT review

All outcomes (mortality, seizures, seizure frequency, time to withdrawal of treatment, quality of life and any adverse events) were considered critical and of equal priority for decision-making.

1.5.2. The quality of the evidence

1.5.2.1. Diagnostic accuracy review

Most of the evidence was graded as low or very low. The main reasons for this were a lack of blinding of index tests and gold standard tests, which may have caused detection bias. Imprecision of estimates also occurred frequently, partly due to the small sample sizes of some studies. Other studies also did not report 95% confidence intervals, or did not report raw data sufficiently clearly to allow calculation of 95% confidence intervals, which prevented assessment of precision for these studies. In addition, some studies used a ‘case-control’ approach. In such studies the overall sample were purposefully derived from one group of people who had epilepsy, and from another group who did not have epilepsy but instead had a specific differential diagnosis (such as psychogenic non epileptic seizures). This results in the non-epilepsy group in such studies being more homogeneous than would be expected in the protocol population, where participants were meant to be drawn consecutively from a more heterogeneous sample of people who were suspected of epilepsy. This reduced the representativeness of the population in such ‘case-control’ studies, and a downgrade for indirectness was therefore made.

1.5.2.2. RCT review

Evidence was graded as moderate to very low in both comparisons (continuous EEG versus routine EEG, and micro-EEG plus routine care versus routine care only). Risk of bias was related to a lack of reporting of allocation concealment in all outcomes across both comparisons. Imprecision varied between no serious imprecision and very serious imprecision across all outcomes in both comparisons, which fully explained the variability in overall grade observed.

1.5.3. Benefits and harms

The committee considered the evidence relating to the different types of index test used, in order to decide if any tests or strategies should be recommended. The index tests were divided into categories and discussed in turn, and the sections below relate to each discrete discussion. Discussion of the diagnostic accuracy and RCT evidence has been integrated where appropriate.

Discussion of benefits and harms in relation to the diagnostic accuracy evidence was simplified by the fact that the higher the sensitivity and specificity of an index test, the greater the benefits resulting from the index test achieving many true positive and true negative results, and the lower their harms resulting from index tests leading to fewer false positive and false negative results. As the committee were focussed on selecting tests where the sensitivity and specificity were very high, benefits were automatically optimised, and harms were automatically reduced. Discussion of benefits and harms in relation to RCT evidence is only discussed in the EEG section, as the two included RCTs were restricted to evaluating different methods of EEG.

Stratum 1: Differentiating between epilepsy and non-epilepsy

Semiology, signs and symptoms

Few semiological findings had adequate sensitivity and specificity to be considered for recommendation, but epileptologist observation of ‘eye opening or widening at onset of seizure’ and ‘eyes open during seizure’ during an in-hospital seizure video had excellent sensitivity and good specificity for differentiation between epilepsy and psychogenic non-epileptic seizures (PNES). However, these findings were not felt to be wholly relevant to the customary diagnostic situation, where in-hospital video-recordings of seizures would not normally be available. In a situation where hospital video recordings of seizures would be available, the gold standard method of video-EEG would normally be possible anyway, making such index tests unnecessary. Therefore, a recommendation specifically relating to using these semiological findings as individual diagnostic tests was not made.

The only sign or symptom-related finding with high accuracy was epileptologist history-taking and examination. Evidence from a high-powered study suggested that clinical diagnosis by an epileptologist, without ancillary assistance from any technological adjuncts such as EEG or imaging, was able to provide very good sensitivity and specificity for differentiating between epilepsy and any type of non-epilepsy in adults. In other words, these data suggested very small risks of a missed diagnosis and low risks of a misdiagnosis. The validity of this finding was enhanced by the fact that the gold standard for this study was video-EEG, which is regarded as the most valid method. These findings underlined the committee’s existing clinical view that patients should be referred to a specialist for diagnosis as soon as possible. Although the evidence was in adults, the recommendation was extended to children and young people on the basis that the committee did not think that the diagnostic accuracy of an expert clinical diagnosis would be affected by the patient’s age. Therefore, a recommendation was made that children, young people and adults should be referred to an expert clinician for assessment and diagnosis.

The committee also agreed that eye-witness reports of the seizure should be collected as a central part of the history taking by the expert. It was agreed that without witness-reports the history will lack information on essential features of a seizure than can increase the accuracy of a diagnosis. In addition, it was agreed that if video information is available, such as from mobile phones belonging to friends or family, this should also be used. It should be noted that the direct evidence relating to eye-witness reports and mobile phone video did not suggest either could be usefully used alone as an accurate diagnostic test, but the committee agreed that as part of the array of information collected in the history, they would enhance the accuracy of diagnosis by the expert clinician.

Serum measures

The committee considered the evidence for the use of serum measures, such as prolactin, lactate, anion gap, glial fibrillary astrocytic protein levels and ammonia, as post-ictal methods to diagnose epilepsy (differentiating between epilepsy and PNES). One study demonstrated that a paired prolactin test taken at 15 minutes and 2 hours after a seizure had high sensitivity for detection of generalised clonic tonic seizures, but the specificity indicated that 25% of people with no epilepsy might be mis-diagnosed by this test. Furthermore, the confidence intervals were wide, suggesting that the true result in the population might be much lower than that observed in the sample. Overall, the committee did not think that the sensitivity and specificity for any serum test were adequate, with unacceptable levels of harm likely to result from missed diagnoses or misdiagnoses. Therefore, no recommendations to use such tests were made..

ECG

In the one study examining this area, the ECG data were poorly reported, and it was unclear how the sensitivity and specificity had been evaluated. The committee were aware of existing guidance and practice relating to the use of ECG in investigation of people who have had episode of loss of consciousness. A 12-lead ECG is an accepted part of any initial evaluation of a patient with loss of consciousness to assess for underlying conduction abnormalities or abnormalities of QT interval or S and T waves. These might be important findings for diagnosis of a cardiac cause of loss of consciousness. A positive ECG increases the likelihood that there is a cardiac cause of a loss of consciousness and the NICE guideline provides guidance on red flag abnormalities that merit urgent assessment (Transient loss of consciousness (‘blackouts’) in over 16s, Clinical guideline [CG109]). An ECG will not rule in or rule out epilepsy, but the committee agreed with existing guidance and practice that ECG should be available alongside other tests and investigations to contribute to the overall information informing an accurate diagnosis made by an expert.

The committee also considered that non-epileptic seizure type events may be caused by metabolic disorders such as hypoglycaemia. Therefore, the committee also agreed, by consensus, that evaluation for metabolic disorders including hypoglycaemia should be included in the initial assessment.

Imaging tests

The diagnostic accuracy of MRI, CT, and single photon emission computed tomography (SPECT) were considered by the committee. 4T MRI and SPECT both demonstrated reasonable accuracy, but this did not reach the pre-hoc threshold set at 0.9 for sensitivity and close to 0.9 for specificity, and the uncertainty of estimates was high. Overall, none of the imaging devices were able to demonstrate sufficient sensitivity and specificity to assure the committee that the harms of false negatives and false positives would not be excessive. The committee therefore did not recommend any imaging modality for diagnostic purposes. However, the committee were aware of the importance of imaging in determining the presence of underlying structural causes of known epilepsy, and agreed that it was important to recommend that they continue to be used for that purpose.

EEG tests

The committee discussed the potential utility of EEG tests as an interictal test, allowing testing schedules that were not fully constrained by the timing of seizures. Routine interictal EEG, as well as ambulatory and provoked interictal EEG, demonstrated very good specificity alongside very poor sensitivity for detection of epilepsy. This indicated that routine EEG results could be useful for ‘ruling a patient in’ if epileptiform or other abnormalities were observed on the EEG trace, because the low specificity indicates that very few people without epilepsy will demonstrate such abnormalities. However, routine EEG cannot be used to ‘rule’ out epilepsy in a patient with a negative EEG, because a very large proportion of people with a true diagnosis of epilepsy do not show epileptiform abnormalities on a routine EEG.

Therefore, the committee agreed that routine EEG could be used to support a pre-existing clinical diagnosis of epilepsy, but should never be used to exclude a diagnosis. EEG could therefore not be usefully used as a solitary test, and the committee agreed it should never be requested unless reasonable certainty already existed that epilepsy was present.

The evidence suggested that some provoking manoeuvres such as hyperventilation might improve sensitivity. The committee therefore recommended that provoking manoeuvres could be applied during routine EEG when possible, but that the small risks of such manoeuvres (such as an induced seizure, with its associated risks) should be considered and relayed to the patients before testing. In addition, some evidence suggested that ambulatory EEG had better sensitivity than routine EEG, with specificity that was equal to routine EEG. This was supported by RCT evidence showing that ambulatory EEG picked up more seizures than routine EEG. The committee therefore recommended that ambulatory EEG could be used when possible or available. These recommendations concerning the addition of provoking manoeuvres and ambulatory methods were not made because it was thought that increased sensitivity would allow EEG to be used as an independent definitive test; in neither case did the evidence suggest that the elevated sensitivity would be high enough. However, in both cases the slight improvement in sensitivity permitted increased confidence that EEG findings could be even more appropriately used as one piece of supporting information in the overall diagnostic picture.

The timing of EEG was also discussed. No data were found relating to the association between time after seizure and diagnostic accuracy, but the consensus was that the earlier that EEG could be carried out, the higher the diagnostic accuracy. For this reason, a recommendation was made that EEG should be carried out as quickly as possible after the seizure, and the committee agreed this is ideally within 72 hours.

Evidence concerning the use of EEG synchrony measures was also discussed. It is believed that increased synchrony of cortical firing is a common feature of brain physiology in people with epilepsy. Therefore, although abnormalities of the interictal EEG trace may not be a sensitive indicator of epilepsy, measures of synchrony may be more useful. Some of the results in the literature appeared to support this idea, with two studies demonstrating excellent sensitivity and specificity for detection of partial epilepsy and temporal lobe epilepsy using this method. However, the confidence intervals around these estimates were wide, and the studies did not provide enough technical information to allow a full understanding of the exact nature of the test as it would be used clinically. The committee discussed how these testing methods are currently in the experimental stages and that they are not in general clinical use. Therefore, no recommendations in this area were made.

Finally, the committee discussed the particular limitations of EEG in detecting frontal lobe seizures due to anatomical barriers to electrode detection in the frontal lobe region. The committee also discussed how EEG may have some ability to differentiate between focal and generalised seizures. However due to the lack of direct evidence from the review and the greater importance of other topics, the committee agreed that these areas did not warrant recommendations.

Magnetoencephalography / Transcranial magnetic stimulation tests

Most of the evidence suggested that magnetoencephalography / transcranial magnetic stimulation tests had an inadequate combination of sensitivity and specificity. One study showed excellent sensitivity for paired pulse TMS with EEG immediately after hyperventilation, but specificity was low enough to yield an unacceptable number of misdiagnoses. Therefore, no recommendations were made in this area.

Psychological tests

Several psychological tests were considered, such as domains of the Personality Assessment Scale, or the Structured Interview of Malingered Symptomology. In all cases these were used to differentiate epilepsy from psychogenic non-epileptic seizures. However, the committee agreed that none of the measures had a sufficiently good combination of high sensitivity and high specificity to permit recommendations.

Linguistic tests

One study evaluated the diagnostic accuracy of linguistic analysis of a patient’s later description of seizure events. The sensitivity and specificity were reasonably high when measured by one experimental rater, but the confidence intervals were very wide, making it possible that the values were significantly below this. The other rater had far inferior sensitivity, with even wider confidence intervals. In addition, the reporting in the paper was unclear and it was not obvious whether the paper was reporting detection of epilepsy or detection of psychogenic non-epileptic seizures. Therefore, no recommendations were made in relation to this evidence.

Electromyography (EMG) and accelerometers

The committee discussed how EMG and accelerometers may be used to differentiate between epilepsy and PNES by detecting different patterns of motor unit activity or kinesiology during a seizure. Wrist accelerometers analysed with an automated algorithm proved to have good sensitivity and excellent specificity. Unfortunately, the data were based on sparse data, which resulted in wide confidence intervals. Therefore, the committee were unable to have sufficient confidence in the estimates to make a recommendation.

Initial diagnosis at admission

Three papers that utilised a variety of tests in order to make an initial diagnosis were considered by the committee. Two of the studies involved expert neurologists, and the tests included a history and available diagnostic testing without EEG. Both of these studies demonstrated very good sensitivity and good specificity, and the committee agreed that these findings confirmed those found in the semiology section suggesting that expert clinical diagnosis is highly accurate. This reinforced the decision to recommend initial referral to an expert for assessment.

Miscellaneous tests

Although most of the miscellaneous tests failed to have sufficient accuracy, the Epifinder, an artificial intelligence application which utilises pattern recognition to assist diagnosis, had good sensitivity and specificity. Unfortunately, the confidence intervals were too wide to permit sufficient certainty of results and so no recommendations were made..

Stratum 2: Differentiating between epilepsy sub-types

The committee discussed the evidence concerning differentiation between autoimmune epilepsy and other epilepsy, but none of the index tests evaluated were sufficiently accurate to warrant recommendation.

1.5.4. Cost effectiveness and resource use

No health economic studies were identified for this review question. Unit costs were presented to aid committee consideration of cost effectiveness.

The committee discussed the clinical evidence presented and noted that, adults, children and young people with new onset of seizures should be referred urgently for assessment of epilepsy. Initial assessment for epilepsy in current practice encompasses taking a detailed history of the persons seizures – including eyewitness accounts and video footage of these seizures if available – and conducting an ECG. Additional tests include neuroimaging and EEG. However, the committee noted an EEG should not be used to exclude a diagnosis of epilepsy.

The recommendations made by the committee ensure adults, children and young people with new onset of seizures are referred urgently for assessment of epilepsy by a specialist in epilepsy diagnosis and ensure the appropriate diagnostic tests to diagnose epilepsy are undertaken. A missed diagnosis of epilepsy can result in poor clinical outcomes for patients. Patients with missed diagnosis of epilepsy will unlikely be aware of the high risks associated with seizures for example, the risk of SUDEP and other related epilepsy accidents (e.g., drowning in the bath or being involved in a road traffic accident as a result of experiencing an unexpected seizure). For a non-drug refractory epilepsy population, SMRs for patients with epilepsy are highest in the first two years of an epilepsy diagnosis. Therefore, ensuring epilepsy patients are diagnosed and given appropriate advice as early as possible is imperative in reducing the risk of epilepsy mortality which is achieved by rendering patients’ seizure free on the appropriate ASMs. With a missed diagnosis of epilepsy patients who should be receiving ASMs will not be receiving these.

The committee noted that if an EEG is requested in current practice, this is not typically received by the patient within 72 hours (which is the ideal time frame recommended by the committee). In current practice an EEG would be carried out within 2-3 weeks. However, receiving an EEG within 72 hours once an EEG has been requested by a healthcare professional allows for more timely diagnosis of epilepsy.

The committee acknowledged that many epilepsy service centres are often limited by staff and equipment availability but noted the same number of people would be referred for an EEG – the EEG would just be undertaken at an earlier date. The committee however noted, that many epilepsy service centres will already be working at full capacity to maintain the current levels of service provision. The recommendation made by the committee states that, an EEG should be performed as soon as possible, stipulating that the ideal time frame is within 72 hours. Overall, the committee concluded that gradually decreasing the time frame for which people receive an EEG across epilepsy services would not result in a substantial resource impact. For epilepsy services already working at full capacity, in the short-term, additional resources may be required whilst neurophysiologists accommodate a change in practice. However, overall, once epilepsy services have adapted to offering EEGs for the diagnosis of epilepsy at a reduced time frame, epilepsy service centres will reach a new equilibrium for service provision, and no additional costs will be associated with this recommendation.

All other recommendations made are largely reflective of UK current practice. In current practice a small proportion of people will procced to sleep deprived EEG if routine EEG is normal due to a strong clinical suspicion of generalised epilepsy. Ambulatory EEG may be performed for people who present with an initial seizure but there is strong clinical suspicion that there have been previous undeclared of unrecognised events. In general, the majority of people who receive a routine EEG will not receive additional diagnostic EEG’s. However, these tests can provide useful information leading to better tailored health care.

Overall, the QALY gains associated with a correct diagnosis of epilepsy are highly likely to be cost effective. The recommendations made ensure people will receive a timely and appropriate diagnosis of epilepsy. Therefore, tailored health care plans will be implemented in the most feasible time frame possible, resulting in greater health outcomes for patients. As the committee made recommendations that were largely reflective of UK current practice, this recommendation is not expected to result in a significant resource impact.

1.5.5. Recommendations supported by this evidence review

This evidence review supports recommendations 1.2.1 – 1.2.10.

References

1.: Aass F, Kaada BR, Torp KH. The diagnostic and prognostic value of the initial electroencephalogram in children with convulsions. Acta Paediatrica. 1956; 45(4):335–342 [PubMed: 13326388]
2.: Ahdab R, Riachi N. Reexamining the added value of intermittent photic stimulation and hyperventilation in routine EEG practice. European Neurology. 2014; 71(1–2):93–98 [PubMed: 24335163]
3.: Al-Qudah AA, Abu-Sheik S, Tamimi AF. Diagnostic value of short duration outpatient video electroencephalographic monitoring. Pediatric Neurology. 1999; 21(3):622–625 [PubMed: 10513688]
4.: Alam-Eldeen MH, Hasan NMA. Assessment of the diagnostic reliability of brain CT and MRI in pediatric epilepsy patients. The Egyptian Journal of Radiology and Nuclear Medicine. 2015; 46(4):1129–1141
5.: Alapirtti T, Waris M, Fallah M, Soilu-Hanninen M, Makinen R, Kharazmi E et al. C-reactive protein and seizures in focal epilepsy: a video-electroencephalographic study. Epilepsia. 2012; 53(5):790–796 [PubMed: 22462619]
6.: Albadareen R, Gronseth G, Landazuri P, He J, Hammond N, Uysal U. Postictal ammonia as a biomarker for electrographic convulsive seizures: A prospective study. Epilepsia. 2016; 57(8):1221–1227 [PMC free article: PMC6631345] [PubMed: 27245120]
7.: Alving J. Serum prolactin levels are elevated also after pseudo-epileptic seizures. Seizure. 1998; 7(2):85–89 [PubMed: 9627196]
8.: An N, Zhao W, Liu Y, Yang X, Chen P. Elevated serum miR-106b and miR-146a in patients with focal and generalized epilepsy. Epilepsy Research. 2016; 127:311–316 [PubMed: 27694013]
9.: Angus-Leppan H. Diagnosing epilepsy in neurology clinics: a prospective study. Seizure. 2008; 17(5):431–436 [PubMed: 18282726]
10.: Arnold LM, Privitera MD. Psychopathology and trauma in epileptic and psychogenic seizure patients. Psychosomatics. 1996; 37(5):438–443 [PubMed: 8824123]
11.: Asadi-Pooya AA, Rabiei AH, Tinker J, Tracy J. Review of systems questionnaire helps differentiate psychogenic nonepileptic seizures from epilepsy. Journal of Clinical Neuroscience. 2016; 34:105–107 [PubMed: 27473020]
12.: Asano E, Pawlak C, Shah A, Shah J, Luat AF, Ahn-Ewing J et al. The diagnostic value of initial video-EEG monitoring in children--review of 1000 cases. Epilepsy Research. 2005; 66(1–3):129–135 [PubMed: 16157474]
13.: Ashrafi MR, Mohammadi M, Tafarroji J, Shabanian R, Salamati P, Zamani GR. Melatonin versus chloral hydrate for recording sleep EEG. European Journal of Paediatric Neurology. 2010; 14(3):235–238 [PubMed: 19616978]
14.: Aydin H, Oktay NA, Kizilgoz V, Altin E, Tatar IG, Hekimoglu B. Value of proton-MR-spectroscopy in the diagnosis of temporal lobe epilepsy; correlation of metabolite alterations with electroencephalography. Iranian Journal of Radiology. 2012; 9(1):1–11 [PMC free article: PMC3522336] [PubMed: 23329953]
15.: Azar NJ, Pitiyanuvath N, Vittal NB, Wang L, Shi Y, Abou-Khalil BW. A structured questionnaire predicts if convulsions are epileptic or nonepileptic. Epilepsy & Behavior. 2010; 19(3):462–466 [PubMed: 20926353]
16.: Azar NJ, Tayah TF, Wang L, Song Y, Abou-Khalil BW. Postictal breathing pattern distinguishes epileptic from nonepileptic convulsive seizures. Epilepsia. 2008; 49(1):132–137 [PubMed: 17651411]
17.: Barras P, Siclari F, Hügli O, Rossetti AO, Lamy O, Novy J. A potential role of hypophosphatemia for diagnosing convulsive seizures: A case-control study. Epilepsia. 2019; 60(8):1580–1585 [PubMed: 31211423]
18.: Barry JJ, Atzman O, Morrell MJ. Discriminating between epileptic and nonepileptic events: the utility of hypnotic seizure induction. Epilepsia. 2000; 41(1):81–84 [PubMed: 10643928]
19.: Batalha S, Dias AI. Diagnostic value of electroencephalography in the pediatric emergency department. Sinapse. 2010; 10(2):19–23
20.: Bayly J, Carino J, Petrovski S, Smit M, Fernando DA, Vinton A et al. Time-frequency mapping of the rhythmic limb movements distinguishes convulsive epileptic from psychogenic nonepileptic seizures. Epilepsia. 2013; 54(8):1402–1408 [PubMed: 23647194]
21.: Beghi M, Cornaggia I, Diotti S, Erba G, Harder G, Magaudda A et al. The semantics of epileptic and psychogenic nonepileptic seizures and their differential diagnosis. Epilepsy & Behavior. 2020; 111:107250 [PubMed: 32603809]
22.: Bell WL, Park YD, Thompson EA, Radtke RA. Ictal cognitive assessment of partial seizures and pseudoseizures. Archives of Neurology. 1998; 55(11):1456–1459 [PubMed: 9823830]
23.: Benbadis SR. A spell in the epilepsy clinic and a history of “chronic pain” or “fibromyalgia” independently predict a diagnosis of psychogenic seizures. Epilepsy & Behavior. 2005; 6(2):264–265 [PubMed: 15710315]
24.: Benbadis SR, Lancman ME, King LM, Swanson SJ. Preictal pseudosleep: a new finding in psychogenic seizures. Neurology. 1996; 47(1):63–67 [PubMed: 8710126]
25.: Benbadis SR, Wolgamuth BR, Goren H, Brener S, Fouad-Tarazi F. Value of tongue biting in the diagnosis of seizures. Archives of Internal Medicine. 1995; 155(21):2346–2349 [PubMed: 7487261]
26.: Benge JF, Wisdom NM, Collins RL, Franks R, Lemaire A, Chen DK. Diagnostic utility of the structured inventory of malingered symptomatology for identifying psychogenic non-epileptic events. Epilepsy & Behavior. 2012; 24(4):439–444 [PubMed: 22683287]
27.: Beniczky S, Polster T, Kjaer TW, Hjalgrim H. Detection of generalized tonic-clonic seizures by a wireless wrist accelerometer: a prospective, multicenter study. Epilepsia. 2013; 54(4):e58–61 [PubMed: 23398578]
28.: Bernardo D, Nariai H, Hussain SA, Sankar R, Salamon N, Krueger DA et al. Visual and semi-automatic non-invasive detection of interictal fast ripples: A potential biomarker of epilepsy in children with tuberous sclerosis complex. Clinical Neurophysiology. 2018; 129(7):1458–1466 [PubMed: 29673547]
29.: Bettini L, Croquelois A, Maeder-Ingvar M, Rossetti AO. Diagnostic yield of short-term video-EEG monitoring for epilepsy and PNESs: A European assessment. Epilepsy & Behavior. 2014; 39:55–58 [PubMed: 25200526]
30.: Bianchi E, Erba G, Beghi E, Giussani G. Self-reporting versus clinical scrutiny: the value of adding questionnaires to the routine evaluation of seizure disorders. An exploratory study on the differential diagnosis between epilepsy and psychogenic nonepileptic seizures. Epilepsy & Behavior. 2019; 90:191–196 [PubMed: 30578096]
31.: Biberon J, de Liege A, de Toffol B, Limousin N, El-Hage W, Florence AM et al. Differentiating PNES from epileptic seizures using conversational analysis on French patients: A prospective blinded study. Epilepsy & Behavior. 2020; 111:107239 [PubMed: 32599432]
32.: Bouma HK, Labos C, Gore GC, Wolfson C, Keezer MR. The diagnostic accuracy of routine electroencephalography after a first unprovoked seizure. European Journal of Neurology. 2016; 23(3):455–463 [PubMed: 26073548]
33.: Bozorg AM, Lacayo JC, Benbadis SR. The yield of routine outpatient electroencephalograms in the veteran population. Journal of Clinical Neurophysiology. 2010; 27(3):191–192 [PubMed: 20461015]
34.: Bozorg AMB, S.R. A simple scale to differentiate psychogenic nonepileptic attacks from epileptic seizures. Epilepsia. 2009; 50(Suppl 11):42–43
35.: Brenner JM, Kent P, Wojcik SM, Grant W. Rapid diagnosis of nonconvulsive status epilepticus using reduced-lead electroencephalography. The Western Journal of Emergency Medicine. 2015; 16(3):442–446 [PMC free article: PMC4427223] [PubMed: 25987926]
36.: Bronen RA, Fulbright RK, Spencer DD, Spencer SS, Kim JH, Lange RC et al. Refractory epilepsy: comparison of MR imaging, CT, and histopathologic findings in 117 patients. Radiology. 1996; 201(1):97–105 [PubMed: 8816528]
37.: Buttle SG, Lemyre B, Sell E, Redpath S, Bulusu S, Webster RJ et al. Combined conventional and amplitude-integrated EEG monitoring in neonates: A prospective study. Journal of Child Neurology. 2019; 34(6):313–320 [PubMed: 30761936]
38.: Chemmanam T, Radhakrishnan A, Sarma SP, Radhakrishnan K. A prospective study on the cost-effective utilization of long-term inpatient video-EEG monitoring in a developing country. Journal of Clinical Neurophysiology. 2009; 26(2):123–128 [PubMed: 19279502]
39.: Chen DK, Graber KD, Anderson CT, Fisher RS. Sensitivity and specificity of video alone versus electroencephalography alone for the diagnosis of partial seizures. Epilepsy & Behavior. 2008; 13(1):115–118 [PubMed: 18396110]
40.: Chen LS, Mitchell WG, Horton EJ, Snead OC, 3rd. Clinical utility of video-EEG monitoring. Pediatric Neurology. 1995; 12(3):220–224 [PubMed: 7619188]
41.: Chen T, Si Y, Chen D, Zhu L, Xu D, Chen S et al. The value of 24-hour video-EEG in evaluating recurrence risk following a first unprovoked seizure: A prospective study. Seizure. 2016; 40:46–51 [PubMed: 27344497]
42.: Chochoi M, Tyvaert L, Derambure P, Szurhaj W. Is long-term electroencephalogram more appropriate than standard electroencephalogram in the elderly? Clinical Neurophysiology. 2017; 128(1):270–274 [PubMed: 27843056]
43.: Choi YJ, Han MY, Lee EH. Children with transient loss of consciousness: Clinical characteristics and the effectiveness of diagnostic tests. Pediatrics and Neonatology. 2020; 61(6):584–591 [PubMed: 32680815]
44.: Chowdhury RN, Hasan AH, Rahman KM, Mondol BA, Deb SR, Mohammad QD. Interictal EEG changes in patients with seizure disorder: experience in Bangladesh. Springerplus. 2013; 2:27 [PMC free article: PMC3589623] [PubMed: 23482637]
45.: Cobb WA. The diagnostic value of the EEG in epileptic children. Proceedings of the Royal Society of Medicine. 1954; 47(10):846–850 [PubMed: 13215518]
46.: Collins S, Iansek R. A prospective study of the predictive value of electroencephalographic abnormalities for epileptic loss of consciousness. Clinical and Experimental Neurology. 1988; 25:103–108 [PubMed: 3267480]
47.: Colon AJ, Ossenblok P, Nieuwenhuis L, Stam KJ, Boon P. Use of routine MEG in the primary diagnostic process of epilepsy. Journal of Clinical Neurophysiology. 2009; 26(5):326–332 [PubMed: 19752741]
48.: Colon AJ, Ronner HE, Boon P, Ossenblok P. Evaluation of MEG vs EEG after sleep deprivation in epilepsy. Acta Neurologica Scandinavica. 2017; 135(2):247–251 [PubMed: 26957488]
49.: Cornaggia CM, Di Rosa G, Polita M, Magaudda A, Perin C, Beghi M. Conversation analysis in the differentiation of psychogenic nonepileptic and epileptic seizures in pediatric and adolescent settings. Epilepsy & Behavior. 2016; 62:231–238 [PubMed: 27494360]
50.: Cragar DE, Schmitt FA, Berry DT, Cibula JE, Dearth CM, Fakhoury TA. A comparison of MMPI-2 decision rules in the diagnosis of nonepileptic seizures. Journal of Clinical & Experimental Neuropsychology: Official Journal of the International Neuropsychological Society. 2003; 25(6):793–804 [PubMed: 13680457]
51.: Cuthill FM, Espie CA. Sensitivity and specificity of procedures for the differential diagnosis of epileptic and non-epileptic seizures: a systematic review. Seizure. 2005; 14(5):293–303 [PubMed: 15878291]
52.: Dash D, Sharma A, Yuvraj K, Renjith A, Mehta S, Vasantha PM et al. Can home video facilitate diagnosis of epilepsy type in a developing country? Epilepsy Research. 2016; 125:19–23 [PubMed: 27328162]
53.: De Paola L, Terra VC, Silvado CE, Teive HA, Palmini A, Valente KD et al. Improving first responders’ psychogenic nonepileptic seizures diagnosis accuracy: Development and validation of a 6-item bedside diagnostic tool. Epilepsy & Behavior. 2016; 54:40–46 [PubMed: 26645799]
54.: Deacon C, Wiebe S, Blume WT, McLachlan RS, Young GB, Matijevic S. Seizure identification by clinical description in temporal lobe epilepsy: how accurate are we? Neurology. 2003; 61(12):1686–1689 [PubMed: 14694030]
55.: del Barrio A, Jimenez-Huete A, Toledano R, Garcia-Morales I, Gil-Nagel A. Validity of the clinical and content scales of the Multiphasic Personality Inventory Minnesota 2 for the diagnosis of psychogenic non-epileptic seizures. Neurologia. 2016; 31(2):106–112 [PubMed: 24485649]
56.: Deli A, Huang YG, Toynbee M, Towle S, Adcock JE, Bajorek T et al. Distinguishing psychogenic nonepileptic, mixed, and epileptic seizures using systemic measures and reported experiences. Epilepsy & Behavior. 2021; 116:107684 [PubMed: 33545648]
57.: DeRoos ST, Chillag KL, Keeler M, Gilbert DL. Effects of sleep deprivation on the pediatric electroencephalogram. Pediatrics. 2009; 123(2):703–708 [PubMed: 19171641]
58.: Derry CP, Davey M, Johns M, Kron K, Glencross D, Marini C et al. Distinguishing sleep disorders from seizures: diagnosing bumps in the night. Archives of Neurology. 2006; 63(5):705–709 [PubMed: 16682539]
59.: Dhanuka AK, Jain BK, Daljit S, Maheshwari D. Juvenile myoclonic epilepsy: A clinical and sleep EEG study. Seizure. 2001; 10(5):374–378 [PubMed: 11488650]
60.: Dixit R, Popescu A, Bagić A, Ghearing G, Hendrickson R. Medical comorbidities in patients with psychogenic nonepileptic spells (PNES) referred for video-EEG monitoring. Epilepsy & Behavior. 2013; 28(2):137–140 [PubMed: 23747495]
61.: Dogan EA, Unal A, Unal A, Erdogan C. Clinical utility of serum lactate levels for differential diagnosis of generalized tonic-clonic seizures from psychogenic nonepileptic seizures and syncope. Epilepsy & Behavior. 2017; 75:13–17 [PubMed: 28806632]
62.: Douw L, de Groot M, van Dellen E, Heimans JJ, Ronner HE, Stam CJ et al. ‘Functional connectivity’ is a sensitive predictor of epilepsy diagnosis after the first seizure. PloS One. 2010; 5(5):e10839 [PMC free article: PMC2877105] [PubMed: 20520774]
63.: Du Pont-Thibodeau G, Sanchez SM, Jawad AF, Nadkarni VM, Berg RA, Abend NS et al. Seizure detection by critical care providers using amplitude-integrated electroencephalography and color density spectral array in pediatric cardiac arrest patients. Pediatric Critical Care Medicine. 2017; 18(4):363–369 [PMC free article: PMC5380542] [PubMed: 28234810]
64.: Dubey D, Singh J, Britton JW, Pittock SJ, Flanagan EP, Lennon VA et al. Predictive models in the diagnosis and treatment of autoimmune epilepsy. Epilepsia. 2017; 58(7):1181–1189 [PubMed: 28555833]
65.: Duez L, Beniczky S, Tankisi H, Hansen PO, Sidenius P, Sabers A et al. Added diagnostic value of magnetoencephalography (MEG) in patients suspected for epilepsy, where previous, extensive EEG workup was unrevealing. Clinical Neurophysiology. 2016; 127(10):3301–3305 [PubMed: 27573996]
66.: Dyken P, Rose S, Badaruddin R. Relative sensitivity of twelve hour telemetry, standard awake EEG and awake EEG with hyperventilation and photic stimulation in detection and management of petit mal (absence) epilepsy. Clinical Research. 1974; 22:95A
67.: Ebersole JS, Leroy RF. Evaluation of ambulatory cassette EEG monitoring: III. Diagnostic accuracy compared to intensive inpatient EEG monitoring. Neurology. 1983; 33(7):853–860 [PubMed: 6683370]
68.: Egawa S, Hifumi T, Nakamoto H, Kuroda Y, Kubota Y. Diagnostic reliability of headset-type continuous video EEG monitoring for detection of ICU patterns and NCSE in patients with altered mental status with unknown etiology. Neurocritical Care. 2020; 32(1):217–225 [PubMed: 31617115]
69.: Ehsan T, Fisher RS, Johns D, Lukas RJ, Blum D, Eskola J. Sensitivity and specificity of paired capillary prolactin measurement in diagnosis of seizures. Journal of Epilepsy. 1996; 9(2):101–105
70.: El-Kader AAA, Amer H, Hussein AAF, Mostafa S, El Gohary A, El-Fayoumy N. The implication of seizure semiology and video-EEG polysomnography in the diagnosis of frontal lobe epilepsy. Egyptian Journal of Neurology, Psychiatry and Neurosurgery. 2009; 46(2):395–408
71.: Elmer J, Coppler PJ, Solanki P, Westover MB, Struck AF, Baldwin ME et al. Sensitivity of continuous electroencephalography to detect ictal activity after cardiac arrest. JAMA Network Open. 2020; 3(4):e203751 [PMC free article: PMC7189220] [PubMed: 32343353]
72.: Elzawahry H, Do CS, Lin K, Benbadis SR. The diagnostic utility of the ictal cry. Epilepsy & Behavior. 2010; 18(3):306–307 [PubMed: 20627816]
73.: Erba G, Giussani G, Juersivich A, Magaudda A, Chiesa V, Lagana A et al. The semiology of psychogenic nonepileptic seizures revisited: Can video alone predict the diagnosis? Preliminary data from a prospective feasibility study. Epilepsia. 2016; 57(5):777–785 [PubMed: 26949106]
74.: Ettinger AB, Weisbrot DM, Nolan E, Devinsky O. Postictal symptoms help distinguish patients with epileptic seizures from those with non-epileptic seizures. Seizure. 1999; 8(3):149–151 [PubMed: 10356371]
75.: Ettinger ABC, P.K.; Jandorf, L.; Cabahug, C.J.; Oster, Z.H.; Atkins, H.L. Postictal SPECT in epileptic versus nonepileptic seizures. Journal of Epilepsy. 1998; 11(2):67–73
76.: Evans E, Koh S, Lerner J, Sankar R, Garg M. Accuracy of amplitude integrated EEG in a neonatal cohort. Archives of Disease in Childhood Fetal & Neonatal Edition. 2010; 95(3):F169–173 [PubMed: 20444809]
77.: Foley CM, Legido A, Miles DK, Grover WD. Diagnostic value of pediatric outpatient video-EEG. Pediatric Neurology. 1995; 12(2):120–124 [PubMed: 7779208]
78.: Fonseca Hernandez E, Olive Gadea M, Requena Ruiz M, Quintana M, Santamarina Perez E, Abraira Del Fresno L et al. Reliability of the early syndromic diagnosis in adults with new-onset epileptic seizures: A retrospective study of 116 patients attended in the emergency room. Seizure. 2018; 61:158–163 [PubMed: 30172139]
79.: Frenkel N, Friger M, Meledin I, Berger I, Marks K, Bassan H et al. Neonatal seizure recognition--comparative study of continuous-amplitude integrated EEG versus short conventional EEG recordings. Clinical Neurophysiology. 2011; 122(6):1091–1097 [PubMed: 21216190]
80.: Gates JR, Ramani V, Whalen S, Loewenson R. Ictal characteristics of pseudoseizures. Archives of Neurology. 1985; 42(12):1183–1187 [PubMed: 3933461]
81.: Geut I, Weenink S, Knottnerus ILH, van Putten M. Detecting interictal discharges in first seizure patients: ambulatory EEG or EEG after sleep deprivation? Seizure. 2017; 51:52–54 [PubMed: 28797915]
82.: Geyer JD, Payne TA, Drury I. The value of pelvic thrusting in the diagnosis of seizures and pseudoseizures. Neurology. 2000; 54(1):227–229 [PubMed: 10636155]
83.: Gilbert DL, Buncher CR. An EEG should not be obtained routinely after first unprovoked seizure in childhood. Neurology. 2000; 54(3):635–641 [PubMed: 10680796]
84.: Giorgi FS, Perini D, Maestri M, Guida M, Pizzanelli C, Caserta A et al. Usefulness of a simple sleep-deprived EEG protocol for epilepsy diagnosis in de novo subjects. Clinical Neurophysiology. 2013; 124(11):2101–2107 [PubMed: 23790524]
85.: Goenka A, Boro A, Yozawitz E. Comparative sensitivity of quantitative EEG (QEEG) spectrograms for detecting seizure subtypes. Seizure. 2018; 55:70–75 [PubMed: 29414138]
86.: Gonzalez-Cuevas M, Coscojuela P, Santamarina E, Pareto D, Quintana M, Sueiras M et al. Usefulness of brain perfusion CT in focal-onset status epilepticus. Epilepsia. 2019; 60(7):1317–1324 [PubMed: 31166616]
87.: Goselink RJM, van Dillen JJ, Aerts M, Arends J, van Asch C, van der Linden I et al. The difficulty of diagnosing NCSE in clinical practice; external validation of the Salzburg criteria. Epilepsia. 2019; 60(8):e88–e92 [PMC free article: PMC6852511] [PubMed: 31318040]
88.: Granados Sanchez AM, Orejuela Zapata JF. Diagnosis of mesial temporal sclerosis: sensitivity, specificity and predictive values of the quantitative analysis of magnetic resonance imaging. Neuroradiology Journal. 2018; 31(1):50–59 [PMC free article: PMC5789997] [PubMed: 28899220]
89.: Grau-Lopez L, Jimenez M, Ciurans J, Barambio S, Fumanal A, Becerra JL. Diagnostic yield of routine electroencephalography with concurrent video recording in detecting interictal epileptiform discharges in relation to reasons for request: A prospective study of 1,080 video-electroencephalograms. Journal of Clinical Neurophysiology. 2017; 34(5):434–437 [PubMed: 28520630]
90.: Hanrahan B, Ghearing G, Urban A, Plummer C, Pan J, Hendrickson R et al. Diagnostic accuracy of paroxysmal spells: Clinical history versus observation. Epilepsy & Behavior. 2018; 78:73–77 [PubMed: 29175694]
91.: Hauf M, Slotboom J, Nirkko A, von Bredow F, Ozdoba C, Wiest R. Cortical regional hyperperfusion in nonconvulsive status epilepticus measured by dynamic brain perfusion CT. AJNR: American Journal of Neuroradiology. 2009; 30(4):693–698 [PMC free article: PMC7051787] [PubMed: 19213823]
92.: Hendrickson R, Popescu A, Dixit R, Ghearing G, Bagic A. Panic attack symptoms differentiate patients with epilepsy from those with psychogenic nonepileptic spells (PNES). Epilepsy & Behavior. 2014; 37:210–214 [PubMed: 25084477]
93.: Hernandez-Ronquillo L, Thorpe L, Dash D, Hussein T, Hunter G, Waterhouse K et al. Diagnostic accuracy of the ambulatory EEG vs. routine EEG for first single unprovoked seizures and seizure recurrence: The DX-Seizure Study. Frontiers in Neurology. 2020; 11:223 [PMC free article: PMC7160330] [PubMed: 32328023]
94.: Hoefnagels WA, Padberg GW, Overweg J, Roos RA, van Dijk JG, Kamphuisen HA. Syncope or seizure? The diagnostic value of the EEG and hyperventilation test in transient loss of consciousness. Journal of Neurology, Neurosurgery and Psychiatry. 1991; 54(11):953–956 [PMC free article: PMC1014614] [PubMed: 1800665]
95.: Hong SJ, Kim H, Schrader D, Bernasconi N, Bernhardt BC, Bernasconi A. Automated detection of cortical dysplasia type II in MRI-negative epilepsy. Neurology. 2014; 83(1):48–55 [PMC free article: PMC4114179] [PubMed: 24898923]
96.: Huang LL, Wang YY, Liu LY, Tang HP, Zhang MN, Ma SF et al. Home videos as a cost-effective tool for the diagnosis of paroxysmal events in infants: Prospective study. JMIR MHealth and UHealth. 2019; 7(9):e11229 [PMC free article: PMC6746063] [PubMed: 31516128]
97.: Husain AM, Towne AR, Chen DK, Whitmire LE, Voyles SR, Cardenas DP. Differentiation of epileptic and psychogenic nonepileptic seizures using single-channel surface electromyography. Journal of Clinical Neurophysiology. 2020; 10.1097/WNP.0000000000000703 [PubMed: 32501944] [CrossRef]
98.: Izadyar S, Shah V, James B. Comparison of postictal semiology and behavior in psychogenic nonepileptic and epileptic seizures. Epilepsy & Behavior. 2018; 88:123–129 [PubMed: 30268021]
99.: Jackson A, Teo L, Seneviratne U. Challenges in the first seizure clinic for adult patients with epilepsy. Epileptic Disorders. 2016; 18(3):305–314 [PubMed: 27506513]
100.: Jaraba S, Reynes-Llompart G, Sala-Padro J, Veciana M, Miro J, Pedro J et al. Usefulness of HMPAO-SPECT in the diagnosis of nonconvulsive status epilepticus. Epilepsy & Behavior. 2019; 101(Pt B):106544 [PubMed: 31753769]
101.: Kadivar M, Moghadam EM, Shervin Badv R, Sangsari R, Saeedy M. A comparison of conventional electroencephalography with amplitude-integrated EEG in detection of neonatal seizures. Medical Devices Evidence and Research. 2019; 12:489–496 [PMC free article: PMC6911316] [PubMed: 31849541]
102.: Keezer MR, Simard-Tremblay E, Veilleux M. The diagnostic accuracy of prolonged ambulatory versus routine EEG. Clinical EEG & Neuroscience: Official Journal of the EEG & Clinical Neuroscience Society (ENCS). 2016; 47(2):157–161 [PubMed: 26376916]
103.: Kerr WT, Janio EA, Braesch CT, Le JM, Hori JM, Patel AB et al. Diagnostic implications of review-of-systems questionnaires to differentiate epileptic seizures from psychogenic seizures. Epilepsy & Behavior. 2017; 69:69–74 [PMC free article: PMC5423814] [PubMed: 28236725]
104.: Kerr WT, Janio EA, Braesch CT, Le JM, Hori JM, Patel AB et al. An objective score to identify psychogenic seizures based on age of onset and history. Epilepsy & Behavior. 2018; 80:75–83 [PMC free article: PMC5845850] [PubMed: 29414562]
105.: Kerr WT, Janio EA, Braesch CT, Le JM, Hori JM, Patel AB et al. Identifying psychogenic seizures through comorbidities and medication history. Epilepsia. 2017; 58(11):1852–1860 [PMC free article: PMC5669805] [PubMed: 28895657]
106.: Khamis H, Mohamed A, Simpson S, McEwan A. Detection of temporal lobe seizures and identification of lateralisation from audified EEG. Clinical Neurophysiology. 2012; 123(9):1714–1720 [PubMed: 22418593]
107.: Khan AY, Baade L, Ablah E, McNerney V, Golewale MH, Liow K. Can hypnosis differentiate epileptic from nonepileptic events in the video/EEG monitoring unit? Data from a pilot study. Epilepsy & Behavior. 2009; 15(3):314–317 [PubMed: 19362599]
108.: Khurana DS, Valencia I, Kruthiventi S, Gracely E, Melvin JJ, Legido A et al. Usefulness of ocular compression during electroencephalography in distinguishing breath-holding spells and syncope from epileptic seizures. Journal of Child Neurology. 2006; 21(10):907–910 [PubMed: 17005113]
109.: Kimiskidis VK, Tsimpiris A, Ryvlin P, Kalviainen R, Koutroumanidis M, Valentin A et al. TMS combined with EEG in genetic generalized epilepsy: A phase II diagnostic accuracy study. Clinical Neurophysiology. 2017; 128(2):367–381 [PubMed: 28007469]
110.: King MA, Newton MR, Jackson GD, Fitt GJ, Mitchell LA, Silvapulle MJ et al. Epileptology of the first-seizure presentation: A clinical, electroencephalographic, and magnetic resonance imaging study of 300 consecutive patients. Lancet. 1998; 352(9133):1007–1011 [PubMed: 9759742]
111.: Knox A, Arya R, Horn PS, Holland K. The diagnostic accuracy of video electroencephalography without event capture. Pediatric Neurology. 2018; 79:8–13 [PubMed: 29248327]
112.: Kolls BJ, Husain AM. Assessment of hairline EEG as a screening tool for nonconvulsive status epilepticus. Epilepsia. 2007; 48(5):959–965 [PubMed: 17433054]
113.: Koome M, Churilov L, Chen Z, Chen Z, Naylor J, Thevathasan A et al. Computed tomography perfusion as a diagnostic tool for seizures after ischemic stroke. Neuroradiology. 2016; 58(6):577–584 [PubMed: 26961195]
114.: Koren J, Herta J, Draschtak S, Potzl G, Furbass F, Hartmann M et al. Early epileptiform discharges and clinical signs predict nonconvulsive status epilepticus on continuous EEG. Neurocritical Care. 2018; 29(3):388–395 [PubMed: 29998425]
115.: Koster I, Ossenblok P, Brekelmans GJ, van der Linden I, Hillebrand A, Wijnen BF et al. Sensitivity of magnetoencephalography as a diagnostic tool for epilepsy: a prospective study. Epileptic Disorders. 2020; 22(3):264–272 [PubMed: 32554358]
116.: Kusmakar S, Karmakar C, Yan B, Muthuganapathy R, Kwan P, O’Brien TJ et al. Novel features for capturing temporal variations of rhythmic limb movement to distinguish convulsive epileptic and psychogenic nonepileptic seizures. Epilepsia. 2019; 60(1):165–174 [PubMed: 30536390]
117.: Kuyk J, Spinhoven P, van Dyck R. Hypnotic recall: a positive criterion in the differential diagnosis between epileptic and pseudoepileptic seizures. Epilepsia. 1999; 40(4):485–491 [PubMed: 10219276]
118.: Lalgudi Ganesan S, Stewart CP, Atenafu EG, Sharma R, Guerguerian AM, Hutchison JS et al. Seizure identification by critical care providers using quantitative electroencephalography. Critical Care Medicine. 2018; 46(12):e1105–e1111 [PubMed: 30188384]
119.: Lancman ME, Asconape JJ, Craven WJ, Howard G, Penry JK. Predictive value of induction of psychogenic seizures by suggestion. Annals of Neurology. 1994; 35(3):359–361 [PubMed: 8122889]
120.: Laroia N, Guillet R, Burchfiel J, McBride MC. EEG background as predictor of electrographic seizures in high-risk neonates. Epilepsia. 1998; 39(5):545–551 [PubMed: 9596208]
121.: Lawley A, Evans S, Manfredonia F, Cavanna AE. The role of outpatient ambulatory electroencephalography in the diagnosis and management of adults with epilepsy or nonepileptic attack disorder: A systematic literature review. Epilepsy & Behavior. 2015; 53:26–30 [PubMed: 26515156]
122.: Lawley A, Manfredonia F, Cavanna AE. Video-ambulatory EEG in a secondary care center: A retrospective evaluation of utility in the diagnosis of epileptic and nonepileptic seizures. Epilepsy & Behavior. 2016; 57(Pt A):137–140 [PubMed: 26949156]
123.: Lee JJ, Lee SK, Lee SY, Park KI, Kim DW, Lee DS et al. Frontal lobe epilepsy: clinical characteristics, surgical outcomes and diagnostic modalities. Seizure. 2008; 17(6):514–523 [PubMed: 18329907]
124.: Leitinger M, Trinka E, Gardella E, Rohracher A, Kalss G, Qerama E et al. Diagnostic accuracy of the Salzburg EEG criteria for non-convulsive status epilepticus: a retrospective study. Lancet Neurology. 2016; 15(10):1054–1062 [PubMed: 27571157]
125.: Li Y, Matzka L, Maranda L, Weber D. Anion gap can differentiate between psychogenic and epileptic seizures in the emergency setting. Epilepsia. 2017; 58(9):e132–e135 [PubMed: 28695610]
126.: Limotai C, Ingsathit A, Thadanipon K, Pattanaprateep O, Pattanateepapon A, Phanthumchinda K et al. Efficacy and economic evaluation of delivery of care with tele-continuous EEG in critically ill patients: a multicentre, randomised controlled trial (Tele-cRCT) study protocol. BMJ Open. 2020; 10(3):e033195 [PMC free article: PMC7059544] [PubMed: 32139485]
127.: Limotai C, Tasanaworapunya P, Thaipisuttikul I. Diagnostic performance of the electroencephalogram in the elderly manifesting with episodes of unresponsiveness. Clinical EEG & Neuroscience: Official Journal of the EEG & Clinical Neuroscience Society (ENCS). 2019; 50(3):180–187 [PubMed: 29788788]
128.: Liu LL, Hou XL, Zhang DD, Sun GY, Zhou CL, Jiang Y et al. Clinical manifestations and amplitude-integrated encephalogram in neonates with early-onset epileptic encephalopathy. Chinese Medical Journal. 2017; 130(23):2808–2815 [PMC free article: PMC5717859] [PubMed: 29176138]
129.: Liu Z, Wang Y, Liu X, Du Y, Tang Z, Wang K et al. Radiomics analysis allows for precise prediction of epilepsy in patients with low-grade gliomas. NeuroImage Clinical. 2018; 19:271–278 [PMC free article: PMC6051495] [PubMed: 30035021]
130.: Manez Miro JU, Diaz de Teran FJ, Alonso Singer P, Aguilar-Amat Prior MJ. Emergency electroencephalogram: Usefulness in the diagnosis of nonconvulsive status epilepticus by the on-call neurologist. Neurologia. 2018; 33(2):71–77 [PubMed: 27448521]
131.: Manni R, Terzaghi M, Repetto A. The FLEP scale in diagnosing nocturnal frontal lobe epilepsy, NREM and REM parasomnias: data from a tertiary sleep and epilepsy unit. Epilepsia. 2008; 49(9):1581–1585 [PubMed: 18410366]
132.: McGinty RN, Handel A, Moloney T, Ramesh A, Fower A, Torzillo E et al. Clinical features which predict neuronal surface autoantibodies in new-onset focal epilepsy: implications for immunotherapies. Journal of Neurology, Neurosurgery and Psychiatry. 2021; 92(3):291–294 [PMC free article: PMC7892387] [PubMed: 33219046]
133.: McGonigal A, Oto M, Russell AJ, Greene J, Duncan R. Outpatient video EEG recording in the diagnosis of non-epileptic seizures: a randomised controlled trial of simple suggestion techniques. Journal of Neurology, Neurosurgery and Psychiatry. 2002; 72(4):549–551 [PMC free article: PMC1737844] [PubMed: 11909925]
134.: McKenzie ED, Lim AS, Leung EC, Cole AJ, Lam AD, Eloyan A et al. Validation of a smartphone-based EEG among people with epilepsy: A prospective study. Scientific Reports. 2017; 7:45567 [PMC free article: PMC5377373] [PubMed: 28367974]
135.: Morales A, Bass NE, Verhulst SJ. Serum prolactin levels and neonatal seizures. Epilepsia. 1995; 36(4):349–354 [PubMed: 7607112]
136.: Mueller SG, Young K, Hartig M, Barakos J, Garcia P, Laxer KD. A two-level multimodality imaging Bayesian network approach for classification of partial epilepsy: Preliminary data. Neuroimage. 2013; 71:224–232 [PMC free article: PMC3619666] [PubMed: 23353601]
137.: Naganur VD, Kusmakar S, Chen Z, Palaniswami MS, Kwan P, O’Brien TJ. The utility of an automated and ambulatory device for detecting and differentiating epileptic and psychogenic non-epileptic seizures. Epilepsia Open. 2019; 4(2):309–317 [PMC free article: PMC6546070] [PubMed: 31168498]
138.: National Institute for Health and Care Excellence. Developing NICE guidelines: the manual [updated October 2020]. London. National Institute for Health and Care Excellence, 2014. Available from: http://www.nice.org.uk/article/PMG20/chapter/1%20Introduction%20and%20overview [PubMed: 26677490]
139.: Nguyen-Michel VH, Dinkelacker V, Solano O, Levy PP, Lambrecq V, Adam C et al. 4h versus 1h-nap-video-EEG monitoring in an Epileptology Unit. Clinical Neurophysiology. 2016; 127(9):3135–3139 [PubMed: 27472550]
140.: NHS England and NHS Improvement. 2018/19 National Cost Collection data. 2020. Available from: https://www.england.nhs.uk/national-cost-collection/#ncc1819 Last accessed: 18/06/2021.
141.: Nitzschke R, Muller J, Engelhardt R, Schmidt GN. Single-channel amplitude integrated EEG recording for the identification of epileptic seizures by nonexpert physicians in the adult acute care setting. Journal of Clinical Monitoring and Computing. 2011; 25(5):329–337 [PubMed: 22009108]
142.: Nitzschke R, Muller J, Maisch S, Schmidt GN. Single-channel electroencephalography of epileptic seizures in the out-of-hospital setting: an observational study. Emergency Medicine Journal. 2012; 29(7):536–543 [PubMed: 21636848]
143.: Noe KH, Grade M, Stonnington CM, Driver-Dunckley E, Locke DE. Confirming psychogenic nonepileptic seizures with video-EEG: sex matters. Epilepsy & Behavior. 2012; 23(3):220–223 [PubMed: 22341181]
144.: Okazaki EM, Yao R, Sirven JI, Crepeau AZ, Noe KH, Drazkowski JF et al. Usage of EpiFinder clinical decision support in the assessment of epilepsy. Epilepsy & Behavior. 2018; 82:140–143 [PubMed: 29625364]
145.: Oliva M, Pattison C, Carino J, Roten A, Matkovic Z, O’Brien TJ. The diagnostic value of oral lacerations and incontinence during convulsive “seizures”. Epilepsia. 2008; 49(6):962–967 [PubMed: 18325019]
146.: Ottman R, Barker-Cummings C, Leibson CL, Vasoli VM, Hauser WA, Buchhalter JR. Validation of a brief screening instrument for the ascertainment of epilepsy. Epilepsia. 2010; 51(2):191–197 [PMC free article: PMC2844922] [PubMed: 19694790]
147.: Ouyang CS, Yang RC, Chiang CT, Wu RC, Lin LC. EEG autoregressive modeling analysis: A diagnostic tool for patients with epilepsy without epileptiform discharges. Clinical Neurophysiology. 2020; 131(8):1902–1908 [PubMed: 32599273]
148.: Paldino MJ, Yang E, Jones JY, Mahmood N, Sher A, Zhang W et al. Comparison of the diagnostic accuracy of PET/MRI to PET/CT-acquired FDG brain exams for seizure focus detection: a prospective study. Pediatric Radiology. 2017; 47(11):1500–1507 [PubMed: 28512714]
149.: Papagno C, Montali L, Turner K, Frigerio A, Sirtori M, Zambrelli E et al. Differentiating PNES from epileptic seizures using conversational analysis. Epilepsy & Behavior. 2017; 76:46–50 [PubMed: 28927714]
150.: Patel AA, Ciccone O, Njau A, Shanungu S, Grollnek AK, Fredrick F et al. A pediatric epilepsy diagnostic tool for use in resource-limited settings: A pilot study. Epilepsy & Behavior. 2016; 59:57–61 [PubMed: 27088519]
151.: Pedersen M, Curwood EK, Vaughan DN, Omidvarnia AH, Jackson GD. Abnormal brain areas common to the focal epilepsies: Multivariate pattern analysis of fMRI. Brain Connectivity. 2016; 6(3):208–215 [PubMed: 26537783]
152.: Pensirikul AD, Beslow LA, Kessler SK, Sanchez SM, Topjian AA, Dlugos DJ et al. Density spectral array for seizure identification in critically ill children. Journal of Clinical Neurophysiology. 2013; 30(4):371–375 [PMC free article: PMC3743420] [PubMed: 23912575]
153.: Pollard JR, Eidelman O, Mueller GP, Dalgard CL, Crino PB, Anderson CT et al. TheTARC/sICAM5 ratio in patient plasma is a candidate biomarker for drug resistant epilepsy. Frontiers in Neurology. 2013; 3:181 [PMC free article: PMC3535822] [PubMed: 23293627]
154.: Rafiei SM. Usefulness of sleep-deprived EEG in the diagnosis of seizure disorders in children. Medical Journal of the Islamic Republic of Iran. 2004; 18(1):21–28
155.: Rakshasbhuvankar A, Rao S, Palumbo L, Ghosh S, Nagarajan L. Amplitude integrated electroencephalography compared with conventional video EEG for neonatal seizure detection: A diagnostic accuracy study. Journal of Child Neurology. 2017; 32(9):815–822 [PubMed: 28482764]
156.: Ramanujam B, Dash D, Tripathi M. Can home videos made on smartphones complement video-EEG in diagnosing psychogenic nonepileptic seizures? Seizure. 2018; 62:95–98 [PubMed: 30316048]
157.: Rasmussen NH, Kullberg B, Garre I, Lonborg L. Subclinical epileptic seizures in children. The diagnostic value of a short test programme during simultaneous EEG and video monitoring. Acta Paediatrica Scandinavica. 1987; 76(1):165–166 [PubMed: 3564996]
158.: Rawlings GH, Jamnadas-Khoda J, Broadhurst M, Grünewald RA, Howell SJ, Koepp M et al. Panic symptoms in transient loss of consciousness: Frequency and diagnostic value in psychogenic nonepileptic seizures, epilepsy and syncope. Seizure. 2017; 48:22–27 [PubMed: 28371670]
159.: Renzel R, Baumann CR, Poryazova R. EEG after sleep deprivation is a sensitive tool in the first diagnosis of idiopathic generalized but not focal epilepsy. Clinical Neurophysiology. 2016; 127(1):209–213 [PubMed: 26118491]
160.: Reuber M, Chen M, Jamnadas-Khoda J, Broadhurst M, Wall M, Grünewald RA et al. Value of patient-reported symptoms in the diagnosis of transient loss of consciousness. Neurology. 2016; 87(6):625–633 [PMC free article: PMC4977366] [PubMed: 27385741]
161.: Reuber M, Monzoni C, Sharrack B, Plug L. Using interactional and linguistic analysis to distinguish between epileptic and psychogenic nonepileptic seizures: a prospective, blinded multirater study. Epilepsy & Behavior. 2009; 16(1):139–144 [PubMed: 19674940]
162.: Robles L, Chiang S, Haneef Z. Review-of-systems questionnaire as a predictive tool for psychogenic nonepileptic seizures. Epilepsy & Behavior. 2015; 45:151–154 [PMC free article: PMC4424090] [PubMed: 25812935]
163.: Rosenow F, Wyllie E, Kotagal P, Mascha E, Wolgamuth BR, Hamer H. Staring spells in children: descriptive features distinguishing epileptic and nonepileptic events. Journal of Pediatrics. 1998; 133(5):660–663 [PubMed: 9821425]
164.: Rossetti AO, Schindler K, Alvarez V, Sutter R, Novy J, Oddo M et al. Does continuous video-EEG in patients with altered consciousness improve patient outcome? Current evidence and randomized controlled trial design. Journal of Clinical Neurophysiology. 2018; 35(5):359–364 [PubMed: 29533307]
165.: Rossetti AO, Schindler K, Sutter R, Ruegg S, Zubler F, Novy J et al. Continuous vs routine electroencephalogram in critically ill adults with altered consciousness and no recent seizure: A multicenter randomized clinical trial. JAMA Neurology. 2020; 77(10):1225–1232 [PMC free article: PMC7385681] [PubMed: 32716479]
166.: Rowberry T, Kanthimathinathan HK, George F, Notghi L, Gupta R, Bill P et al. Implementation and early evaluation of a quantitative electroencephalography program for seizure detection in the PICU. Pediatric Critical Care Medicine. 2020; 21(6):543–549 [PubMed: 32343109]
167.: Saeed M, Meghaji M, Al-Malky M, Al-Tubaity S. Interictal electroencephalography (EEG) and diagnosis of childhood epilepsy. Pakistan Paediatric Journal. 2010; 34(3):154–157
168.: Sargolzaei S, Cabrerizo M, Sargolzaei A, Noei S, Eddin A, Rajaei H et al. A probabilistic approach for pediatric epilepsy diagnosis using brain functional connectivity networks. BMC Bioinformatics. 2015; 16(Suppl 7):S9 [PMC free article: PMC4423569] [PubMed: 25953124]
169.: Satpute SC, D.; Franks, R. Diagnoses of neurobehavioral paroxysms in veterans of operation enduring freedom/ operation iraqi freedom (OEF/OIF) - Experiences from a VA epilepsy center. Epilepsy Currents. 2014; 14:168
170.: Schindler K, Wiest R, Kollar M, Donati F. Using simulated neuronal cell models for detection of epileptic seizures in foramen ovale and scalp EEG. Clinical Neurophysiology. 2001; 112(6):1006–1017 [PubMed: 11377259]
171.: Schmidt H, Woldman W, Goodfellow M, Chowdhury FA, Koutroumanidis M, Jewell S et al. A computational biomarker of idiopathic generalized epilepsy from resting state EEG. Epilepsia. 2016; 57(10):e200–e204 [PMC free article: PMC5082517] [PubMed: 27501083]
172.: Schoenenberger RA, Heim SM. Indication for computed tomography of the brain in patients with first uncomplicated generalised seizure. BMJ. 1994; 309(6960):986–989 [PMC free article: PMC2541265] [PubMed: 7950718]
173.: Schorner W, Meencke HJ, Felix R. Temporal-lobe epilepsy: comparison of CT and MR imaging. AJR American Journal of Roentgenology. 1987; 149(6):1231–1239 [PubMed: 3500615]
174.: Schramke CJ, Kay KA, Valeriano JP, Kelly KM. Using patient history to distinguish between patients with non-epileptic and patients with epileptic events. Epilepsy & Behavior. 2010; 19(3):478–482 [PubMed: 20850387]
175.: Schreiner A, Pohlmann-Eden B. Value of the early electroencephalogram after a first unprovoked seizure. Clinical Electroencephalography. 2003; 34(3):140–144 [PubMed: 14521275]
176.: Sen A, Scott C, Sisodiya SM. Stertorous breathing is a reliably identified sign that helps in the differentiation of epileptic from psychogenic non-epileptic convulsions: an audit. Epilepsy Research. 2007; 77(1):62–64 [PubMed: 17766088]
177.: Seneviratne U, Minato E, Paul E. How reliable is ictal duration to differentiate psychogenic nonepileptic seizures from epileptic seizures? Epilepsy & Behavior. 2017; 66:127–131 [PubMed: 28039841]
178.: Shah P, James S, Elayaraja S. EEG for children with complex febrile seizures. Cochrane Database of Systematic Reviews 2020, Issue 4. Art. No.: CD009196. DOI: 10.1002/14651858.CD009196.pub5. [PMC free article: PMC7142325] [PubMed: 32270497] [CrossRef]
179.: Sierra-Marcos A, Toledo M, Quintana M, Edo MC, Centeno M, Santamarina E et al. Diagnosis of epileptic syndrome after a new onset seizure and its correlation at long-term follow-up: longitudinal study of 131 patients from the emergency room. Epilepsy Research. 2011; 97(1–2):30–36 [PubMed: 21783344]
180.: Simani L, Elmi M, Asadollahi M. Serum GFAP level: A novel adjunctive diagnostic test in differentiate epileptic seizures from psychogenic attacks. Seizure. 2018; 61:41–44 [PubMed: 30077862]
181.: Slater JD, Brown MC, Jacobs W, Ramsay RE. Induction of pseudoseizures with intravenous saline placebo. Epilepsia. 1995; 36(6):580–585 [PubMed: 7555971]
182.: Slooter AJ, Vriens EM, Leijten FS, Spijkstra JJ, Girbes AR, van Huffelen AC et al. Seizure detection in adult ICU patients based on changes in EEG synchronization likelihood. Neurocritical Care. 2006; 5(3):186–192 [PubMed: 17290086]
183.: Stewart CP, Otsubo H, Ochi A, Sharma R, Hutchison JS, Hahn CD. Seizure identification in the ICU using quantitative EEG displays. Neurology. 2010; 75(17):1501–1508 [PMC free article: PMC2974462] [PubMed: 20861452]
184.: 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(6):979–982 [PubMed: 12654963]
185.: Sun J, Ma D, Lv Y. Detection of seizure patterns with multichannel amplitude-integrated EEG and the color density spectral array in the adult neurology intensive care unit. Medicine. 2018; 97(38):e12514 [PMC free article: PMC6160116] [PubMed: 30235767]
186.: Swartz BE, Brown C, Mandelkern MA, Khonsari A, Patell A, Thomas K et al. The use of 2-deoxy-2-[18F]fluoro-D-glucose (FDG-PET) positron emission tomography in the routine diagnosis of epilepsy. Molecular Imaging and Biology. 2002; 4(3):245–252 [PubMed: 14537129]
187.: Swingle N, Vuppala A, Datta P, Pedavally S, Swaminathan A, Kedar S et al. Limited-montage EEG as a tool for the detection of nonconvulsive seizures. Journal of Clinical Neurophysiology. 2020; 10.1097/WNP.0000000000000742 [PubMed: 32604191] [CrossRef]
188.: Swisher CB, White CR, Mace BE, Dombrowski KE, Husain AM, Kolls BJ et al. Diagnostic accuracy of electrographic seizure detection by neurophysiologists and non-neurophysiologists in the adult ICU using a panel of quantitative EEG trends. Journal of Clinical Neurophysiology. 2015; 32(4):324–330 [PubMed: 26241242]
189.: Syed TU, Arozullah AM, Loparo KL, Jamasebi R, Suciu GP, Griffin C et al. A self-administered screening instrument for psychogenic nonepileptic seizures. Neurology. 2009; 72(19):1646–1652 [PubMed: 19433737]
190.: Syed TU, Arozullah AM, Suciu GP, Toub J, Kim H, Dougherty ML et al. Do observer and self-reports of ictal eye closure predict psychogenic nonepileptic seizures? Epilepsia. 2008; 49(5):898–904 [PubMed: 18070093]
191.: Syed TU, LaFrance WC, Jr., Kahriman ES, Hasan SN, Rajasekaran V, Gulati D et al. Can semiology predict psychogenic nonepileptic seizures? A prospective study. Annals of Neurology. 2011; 69(6):997–1004 [PubMed: 21437930]
192.: Tafakhori A, Aghamollaii V, Modabbernia AH, Ghaffarpour M, Omrani HA, Harirchian MH et al. Evaluation of partial epilepsy in Iran: role of video-EEG, EEG, and MRI with epilepsy protocol. Iran J Neurol. 2011; 10(1–2):9–15 [PMC free article: PMC3829215] [PubMed: 24250836]
193.: Tatum WO, Hirsch LJ, Gelfand MA, Acton EK, LaFrance WC, Jr., Duckrow RB et al. Assessment of the predictive value of outpatient smartphone videos for diagnosis of epileptic seizures. JAMA Neurology. 2020; 77(5):593–600 [PMC free article: PMC6990754] [PubMed: 31961382]
194.: Tews W, Weise S, Syrbe S, Hirsch W, Viehweger A, Merkenschlager A et al. Is there a predictive value of EEG and MRI after a first afebrile seizure in children? Klinische Padiatrie. 2015; 227(2):84–88 [PubMed: 25419720]
195.: Thangavelu SK, Kasthuri N, Sundaram V, Aravind N, Bilakanti N. A stand-alone EEG monitoring system for remote diagnosis. Telemedicine Journal and e-Health. 2016; 22(4):310–316 [PubMed: 26447776]
196.: Thompson AW, Hantke N, Phatak V, Chaytor N. The Personality Assessment Inventory as a tool for diagnosing psychogenic nonepileptic seizures. Epilepsia. 2010; 51(1):161–164 [PMC free article: PMC2844915] [PubMed: 19490032]
197.: Titgemeyer Y, Surges R, Altenmuller DM, Fauser S, Kunze A, Lanz M et al. Can commercially available wearable EEG devices be used for diagnostic purposes? An explorative pilot study. Epilepsy & Behavior. 2020; 103(Pt A):106507 [PubMed: 31645318]
198.: Topjian AA, Fry M, Jawad AF, Herman ST, Nadkarni VM, Ichord R et al. Detection of electrographic seizures by critical care providers using color density spectral array after cardiac arrest is feasible. Pediatric Critical Care Medicine. 2015; 16(5):461–467 [PMC free article: PMC4456208] [PubMed: 25651050]
199.: Tyson BT, Baker S, Greenacre M, Kent KJ, Lichtenstein JD, Sabelli A et al. Differentiating epilepsy from psychogenic nonepileptic seizures using neuropsychological test data. Epilepsy & Behavior. 2018; 87:39–45 [PubMed: 30172082]
200.: van Diessen E, Otte WM, Braun KP, Stam CJ, Jansen FE. Improved diagnosis in children with partial epilepsy using a multivariable prediction model based on EEG network characteristics. PloS One. 2013; 8(4):e59764 [PMC free article: PMC3614973] [PubMed: 23565166]
201.: van Donselaar CA, Schimsheimer RJ, Geerts AT, Declerck AC. Value of the electroencephalogram in adult patients with untreated idiopathic first seizures. Archives of Neurology. 1992; 49(3):231–237 [PubMed: 1536624]
202.: Vanderzant CW, Giordani B, Berent S, Dreifuss FE, Sackellares JC. Personality of patients with pseudoseizures. Neurology. 1986; 36(5):664–668 [PubMed: 3703265]
203.: Varma AR, Moriarty J, Costa DC, Gaćinovic S, Schmitz EB, Ell PJ et al. HMPAO SPECT in non-epileptic seizures: preliminary results. Acta Neurologica Scandinavica. 1996; 94(2):88–92 [PubMed: 8891051]
204.: Velasco TR, Wichert-Ana L, Mathern GW, Araujo D, Walz R, Bianchin MM et al. Utility of ictal single photon emission computed tomography in mesial temporal lobe epilepsy with hippocampal atrophy: a randomized trial. Neurosurgery. 2011; 68(2):431–436 [PubMed: 21135733]
205.: Verhoeven T, Coito A, Plomp G, Thomschewski A, Pittau F, Trinka E et al. Automated diagnosis of temporal lobe epilepsy in the absence of interictal spikes. NeuroImage Clinical. 2018; 17:10–15 [PMC free article: PMC5842753] [PubMed: 29527470]
206.: Vespa PM, Olson DM, John S, Hobbs KS, Gururangan K, Nie K et al. Evaluating the clinical impact of rapid response electroencephalography: The DECIDE multicenter prospective observational clinical study. Critical Care Medicine. 2020; 48(9):1249–1257 [PMC free article: PMC7735649] [PubMed: 32618687]
207.: Vilyte G, Pretorius C. Personality traits, illness behaviors, and psychiatric comorbidity in individuals with psychogenic nonepileptic seizures (PNES), epilepsy, and other nonepileptic seizures (oNES): Differentiating between the conditions. Epilepsy & Behavior. 2019; 98(Pt A):210–219 [PubMed: 31382179]
208.: Von Oertzen J, Urbach H, Jungbluth S, Kurthen M, Reuber M, Fernandez G et al. Standard magnetic resonance imaging is inadequate for patients with refractory focal epilepsy. Journal of Neurology, Neurosurgery and Psychiatry. 2002; 73(6):643–647 [PMC free article: PMC1757366] [PubMed: 12438463]
209.: Vukmir RB. Does serum prolactin indicate the presence of seizure in the emergency department patient? Journal of Neurology. 2004; 251(6):736–739 [PubMed: 15311351]
210.: Wagner MT, Wymer JH, Topping KB, Pritchard PB. Use of the Personality Assessment Inventory as an efficacious and cost-effective diagnostic tool for nonepileptic seizures. Epilepsy & Behavior. 2005; 7(2):301–304 [PubMed: 16043418]
211.: Wang L, Long X, Aarts RM, van Dijk JP, Arends JBAM. EEG-based seizure detection in patients with intellectual disability: Which EEG and clinical factors are important? Biomedical Signal Processing and Control. 2019; 49:404–418
212.: Wardrope A, Newberry E, Reuber M. Diagnostic criteria to aid the differential diagnosis of patients presenting with transient loss of consciousness: A systematic review. Seizure. 2018; 61:139–148 [PubMed: 30145472]
213.: Watson P, Conroy A, Moran G, Duncan S. Retrospective study of sensitivity and specificity of EEG in the elderly compared with younger age groups. Epilepsy & Behavior. 2012; 25(3):408–411 [PubMed: 23110971]
214.: Weber AB, Albert DV, Yin H, Held TP, Patel AD. Diagnosis of electrical status epilepticus during slow-wave sleep with 100 seconds of sleep. Journal of Clinical Neurophysiology. 2017; 34(1):65–68 [PubMed: 28045858]
215.: Wilkus RJ, Dodrill CB, Thompson PM. Intensive EEG monitoring and psychological studies of patients with pseudoepileptic seizures. Epilepsia. 1984; 25(1):100–107 [PubMed: 6692785]
216.: Willert C, Spitzer C, Kusserow S, Runge U. Serum neuron-specific enolase, prolactin, and creatine kinase after epileptic and psychogenic non-epileptic seizures. Acta Neurologica Scandinavica. 2004; 109(5):318–323 [PubMed: 15080857]
217.: Yan P, Melman T, Yan S, Otgonsuren M, Grinspan Z. Evaluation of a novel median power spectrogram for seizure detection by non-neurophysiologists. Seizure. 2017; 50:109–117 [PubMed: 28732280]
218.: Zehtabchi S, Abdel Baki SG, Omurtag A, Sinert R, Chari G, Roodsari GS et al. Effect of microEEG on clinical management and outcomes of emergency department patients with altered mental status: a randomized controlled trial. Academic Emergency Medicine. 2014; 21(3):283–291 [PMC free article: PMC4047649] [PubMed: 24628753]
219.: Zibrandtsen IC, Kidmose P, Christensen CB, Kjaer TW. Ear-EEG detects ictal and interictal abnormalities in focal and generalized epilepsy - A comparison with scalp EEG monitoring. Clinical Neurophysiology. 2017; 128(12):2454–2461 [PubMed: 29096220]
220.: Zou R, Wang S, Zhu L, Wu L, Lin P, Li F et al. Calgary score and modified Calgary score in the differential diagnosis between neurally mediated syncope and epilepsy in children. Neurological Sciences. 2017; 38(1):143–149 [PubMed: 27747448]

Appendices

Appendix A. Review protocols

Appendix B. Literature search strategies

This literature search strategy was used for the following reviews:

What is the most accurate approach for 1) diagnosis of epilepsy, and 2) differentiation between types of epilepsy?
What is the most clinically and cost-effective approach for diagnosis of epilepsies?

The literature searches for this review are detailed below and complied with the methodology outlined in Developing NICE guidelines: the manual.¹³⁸

For more information, please see the Methodology review published as part of the accompanying documents for this guideline.

B.1. Clinical search literature search strategy

Download PDF (313K)

B.2. Health Economics literature search strategy

Download PDF (230K)

Appendix C. Clinical evidence selection

C.1. Flow chart of clinical study selection for the review of diagnostic accuracy

Download PDF (139K)

C.2. Flow chart of clinical study selection for the review of clinical efficacy of diagnostic strategies

Download PDF (110K)

Appendix D. Clinical evidence tables

D.1. Clinical evidence Diagnostic accuracy

Download PDF (1.3M)

Appendix E. Coupled sensitivity and specificity forest plots and sROC curves

E.1. Diagnostic accuracy

Download PDF (775K)

E.2. Diagnostic strategies

Download PDF (135K)

Appendix F. GRADE tables

Table 112. Clinical evidence profile: continuous EEG vs Routine EEG (PDF, 139K)

Table 113. Clinical evidence profile: micro EEG + routine care vs Routine care (PDF, 121K)

Appendix G. Health economic evidence selection

Download PDF (132K)

Appendix H. Health economic evidence tables

None.

Appendix I. Health economic model

No original economic modelling was undertaken for this review question.

Appendix J. QUADAS2 risk of bias assessment

Download PDF (257K)

Appendix K. Excluded studies

K.1. Excluded clinical studies

Download PDF (174K)

K.2. Excluded health economic studies

Download PDF (125K)

Tables

Table 1PICO characteristics of review question

Population	Inclusion: Strata: - Children and adults with suspected epilepsy. - Children and adults with epilepsy, where uncertainty remains as to the type of epilepsy Exclusion: New-born babies with acute symptomatic seizures
Target condition	Epilepsies, or type of epilepsy
Index test(s)	Any diagnostic strategies used in papers to detect 1) epilepsy, 2) type of epilepsy. These may include (for example) symptoms/signs, imaging, EEG, ECG, serum measures, either singly or in combination.
Reference standard(s)	Any gold standard used in the studies.
Outcomes	Diagnostic accuracy – sensitivity and specificity
Study design	Observational

Table 2Summary of studies included in the evidence review for detection of epilepsy

Study	Population	Index test(s)	Reference standard
Albadareen, 2016⁶	N=78; USA; Mean age 34.8 GCS (generalised convulsive seizure), 35.2 PNES-C (psychogenic nonepileptic seizures with convulsion, 40.1 FS (focal seizures); 57% female. Inclusion: Adult patients (≥18 years of age) admitted to the epilepsy monitoring unit for event characterization, seizure focus localization, or treatment optimization Exclusion: Factors known to be associated with hyperammonaemia: pre-existing liver disease/cirrhosis, current use of valproic acid or 5- fluorouracil, history of gastrointestinal bleeding, hematologic malignancies, and end-stage renal disease; no event during study. Non-epilepsy population: any suspected of epilepsy	baseline serum ammonia at cut-off >=80 micromol/L	VIDEO EEG
Alving, 1998⁷	N=58; Denmark; median age 28; 46/58 female Inclusion: People with diagnosed epilepsy or pseudo-epileptic seizures Exclusion: Uncertain diagnoses; insufficient seizure description; uncertainty about time elapsed from previous seizure to index seizure; neuroleptic drugs; pregnancy Non-epilepsy population: PNES	Postictal paired serum prolactin measurements at 3 different thresholds	Clinical and video EEG
Arnold, 1996¹⁰	N= 41; USA; mean age 34 years; 53.6% female Inclusion: Patients admitted to the inpatient 24-hour video/EEG monitoring unit for people with intractable seizures; aged >18 Exclusion: not reported Non-epilepsy population: PNES	Interviews to ascertain the following test data: Lifetime Axis I Current Axis I Current Axis II Trauma history	VIDEO EEG
Asadi-Pooya, 2016¹¹	N=60; mean age 28.6 years; 70% female Inclusion: Patients admitted to the Epilepsy Centre with a video-EEG confirmed diagnosis of epilepsy or PNES Exclusion: Patients with concomitant PNES and epilepsy Non-epilepsy population: PNES	Review of systems (ROS) questionnaire, which was in the medical records. This covered the following 10 systems, where each was graded as normal or abnormal: skin; head & ear, nose and throat (HENT); musculoskeletal; pulmonary; cardiovascular; gastrointestinal; genitourinary; hematologic; psychiatry; cognition and memory. The questionnaire was completed by the HCP according to the patient’s history. Scores were generated by any abnormality yielding a score of 1.	VIDEO EEG
Azar, 2008¹⁶	N=40; USA; mean age 34.4 years; 47.5% female Inclusion: Adult patients with epilepsy and generalised tonic-clonic seizures; patients with non-epileptic psychogenic seizures; people with hyper motor seizures from frontal lobe epilepsy Exclusion: Not reported Non-epilepsy population: PNES	Ictal and post ictal physical characteristics, recorded by video	VIDEO EEG
Bayly, 2013²⁰	N=35; Australia; mean age epilepsy/PNES: 33/38; 23/34 female Inclusion: Patients being offered video EEG for the diagnosis of seizure-like events; patients having a convulsive seizure (>10s, with rhythmic movements affecting at least 1 limb) detected by accelerometery during video EEG Exclusion: None reported Non-epilepsy population: PNES	Wrist accelerometer data	Consensus agreement based on clinical and EEG data
Benbadis, 1995²⁵	N=108; USA; mean age 43 years; 56% female Inclusion: All patients admitted to a Epilepsy Monitoring Unit for the diagnosis of spells or presurgical evaluation of epilepsy over a 6-month period. Patients selected whose episodes are characterised by bilateral motor phenomena, LOC, or both. Exclusion: Typical complex partial seizures, with altered awareness but no LOC Non-epilepsy population: syncope	Existence of tongue biting	VIDEO EEG
Benge, 2012²⁶	N=120; USA; Age and gender not reported Inclusion: Case files from patients at a large Veteran’s Affairs hospital’s continuous video-EEG long term monitoring (LTM) programme Exclusion: No SIMS data; missing LTM data; unclear LTM results Non-epilepsy population: PNES	SIMS questionnaire	VIDEO EEG
Bernardo, 2018²⁸	N=11; USA; mean age 21.3 months; 36% female. Inclusion: Infants with active medically refractive epilepsy related to tuberous sclerosis; all video EEGs recorded on Nihon Kohden systems; vEEG sampled at 3000Hz; vEEG recorded at 2 h or more from the most recent seizure; human visual identification of interictal scalp FR; at least 1 brain MRI previously obtained. Controls were children with no brain-related diagnoses including epilepsy, autism and developmental delay; underwent a normal overnight scalp vEEG for clinical reasons with normal results Exclusion: none reported Non-epilepsy population: healthy controls	Existence of interictal fast-ripple events	VIDEO EEG
Chen, 2008³⁹	N=43; USA; mean age 33.6; 29/43 female Inclusion: Patients had seizures with behavioural semiology suggestive of partial seizures, with or without secondary generalisation; EEGs from patients with epilepsy all showed recognisable changes though this was not known to blinded readers. Exclusion: Patients with known mixed epilepsy and PNES Non-epilepsy population: PNES	Ictal video evidence alone Ictal EEG evidence alone Selected ictal semiological features	Diagnosis of epilepsy or PNES was considered established by response to surgery, confirmation by invasive recording, response to psychiatric therapy, or surface video-EEG confirmation followed by serial observations for at least a year
Choi, 2020⁴³	N=160; South Korea; mean age 14.6 years; 59.4% female Inclusion: Under 18 years of age who had been admitted to the Department of Paediatrics or had visited the outpatient clinic or emergency department at Kyung Hee University Hospital (Seoul, South Korea) for TLOC between June 2013 and May 2018. Patients were initially identified who were assigned International Classification of Disease, 10th Revision (ICD-10) billing codes for “syncope and collapse” at the time of the first visit. The medical charts of patients with TLOC as the chief complaint were retrospectively analysed. Exclusion: Patients who had visited the hospital previously due to TLOC and were diagnosed with any disease; patients who had previously undergone any diagnostic tests; patients who had been diagnosed with acute systemic illness on visiting the hospital due to TLOC; patients who were taking medications that can lead to arrhythmia or orthostasis. Non-epilepsy population: any suspected of epilepsy	ECG Brain CT Brain MRI EEG Echocardiogram Head up tilt test	Clinical impression based on all data over prolonged follow up period.
Deli, 2021⁵⁶	N=69; mean age 36.2 years (PNES only); 59% female (PNES only) Inclusion: People with epilepsy or PNES admitted for V-EEG. Exclusion: None reported Non-epilepsy population: PNES	Reports of physical symptoms: Light headedness/dizziness Sensory disturbances/dysesthesias Hot flushes Palpitations	VIDEO EEG
Derry, 2006⁵⁸	N=62; Australia; mean age 27.9 years; 27.4% female Inclusion: Patients who had been referred to a sleep physician or neurologist with a history of nocturnal events of uncertain cause. Individuals with NFLE were eligible for the study if they had a history consistent with NFLE and at least 1 of the following: video-EEG monitoring with clinical or electrographic evidence of nocturnal frontal lobe seizures or a genetic mutation consistent with ADNFLE. Patients with parasomnias were recruited in 2 sub-groups. The first group consisted of subjects who were referred to a sleep clinic for diagnosis of their nocturnal events but in whom a definite diagnosis of “typical” parasomnia was made by the specialist without recourse to video-EEG monitoring. In this group, the diagnosis was made on the basis of the history independently by 3 clinicians (a consultant adult epileptologist, a consultant paediatric epileptologist, and a consultant sleep paediatrician), none of whom were involved in the validation of the FLEP scale. The second group comprised cases in which there was diagnostic uncertainty on the basis of the history alone and in which the diagnosis was established by video-EEG or PSG monitoring. These cases were designated “atypical” parasomnias. Exclusion: not reported Non-epilepsy population: arousal parasomnia and sleep disorder	FLEP scale	Expert interview and, when necessary, recording of events using video-EEG monitoring
Dixit, 2013⁶⁰	N= 280; USA; mean age not reported; 62.5% female Inclusion: People evaluated in EMU with video EEG Exclusion: Unclear diagnosis on vEEG; dual diagnosis of epilepsy/PNES; learning disability; first language not English Non-epilepsy population: PNES	Existence of >1 co-morbidities from medical records	VIDEO EEG
Dogan, 2017⁶¹	N=270; Turkey; age range 19-92; 42% female Inclusion: >=18 years; normal serum pH levels; final definitive diagnosis of generalised tonic-clonic seizures, psychogenic nonepileptic seizures or syncope. Needed to have CT/MRI, EEG and ECG data with observable clinical signs and symptoms. Exclusion: None reported Non-epilepsy population: psychogenic nonepileptic seizures and syncope	Serum lactate	Final definitive diagnosis of generalised tonic-clonic seizures, psychogenic nonepileptic seizures or syncope. Needed to have CT/MRI, EEG and ECG data with observable clinical signs and symptoms
Douw, 2010⁶²	N=161; Holland; mean age 52 years; 51% female Inclusion: 18 years old; evaluated with a standard EEG because of suspected epilepsy after a first possible seizure. Exclusion: not reported Non-epilepsy population: healthy controls	Degree of synchronisation of EEG in time domain, quantified by theta SL	Medical chart review was conducted for all patients to determine whether a clinical diagnosis of epilepsy was reached within a follow-up of one year.
Dubey, 2017⁶⁴	N= 387; USA; mean age 53/44 years; 47.7%/57.4% female Inclusion: Patients in whom autoimmune encephalopathy, autoimmune epilepsy or autoimmune dementia evaluations of serum, CSF, or both were requested; patients with ICD classification of epilepsy or recurrent seizures Exclusion: not reported Non-epilepsy population: any suspected of epilepsy	Antibody prevalence in epilepsy score (APE)	CNS-specific antibodies (neural antibody positive) in presence of confirmed diagnosis based on 2 unprovoked seizures at least 24hrs apart or one unprovoked seizure with additional clinical features suggesting a high probability of recurrence
Duez, 2016⁶⁵	N= 52; Denmark; median age 29 years; 69.2% female Inclusion: Paroxysmal clinical episodes, suggesting epileptic seizures; at least 3 normal EEG recordings, 2 of which included provocation methods of hyperventilation and photo stimulation and 1 of which was sleep-EEG Exclusion: not reported Non-epilepsy population: any suspected of epilepsy but with no interictal findings on provoked EEG	Magnetoencephalography	Diagnostic reference standard was inferred from the diagnosis obtained from the medical chart, after at least one year follow-up after MEG. This was based on all available clinical and para-clinical data for each patient, including description of witnessed seizures, home video recordings of seizures, neuroimaging, laboratory and neurophysiological data.
Egawa, 2020 #1740⁶⁸	N= 50; Japan median age 72 years; 34% female Inclusion: Altered Mental Status (AMS) with unknown aetiology Exclusion: Patients with consciousness recovered completely between HS-cv EEG and C-cEEG monitoring; if C-cEEG monitoring was not performed due to unavailability, or if the HS-cv EEG data were not clear enough due to artefact interruption. Those with do not attempt resuscitation (DNAR) declarations were also excluded, considering that earlier initiation of HS-cv EEG was not performed. Non-epilepsy population: any suspected of epilepsy	Headset-type continuous video EEG monitoring (HS-cv EEG monitoring).	Researchers performed definitive diagnosis of abnormal EEG patterns and NCSE by employing conventional continuous EEG [C-cEEG] monitoring with 21 collodion-type electrodes from the international 10–20 with video camera monitoring. All cEEG records were reviewed by at least two trained neurophysiologists or epileptologists. If any of the EEG findings were equivocal, consensus was used.
Ehsan, 1996⁶⁹	N= 50; USA; mean age 33 years; 60% female Inclusion: Patients admitted to epilepsy monitoring unit for video-EEG monitoring for a history of refractory seizures or non-epileptic events; first clinical event only analysed Exclusion: not reported Non-epilepsy population: any suspected of epilepsy	Paired capillary prolactin measures	VIDEO EEG or audio EEG
Erba, 2016⁷³	N= 21; Italy/USA; mean age >18 years; gender not reported Inclusion: Aged >18 years; admitted to epilepsy centre Exclusion: Lacked intellectual capacity to answer questionnaires Non-epilepsy population: any suspected of epilepsy	Video without EEG or other data	The GS diagnosis was that established by the clinical team after a comprehensive evaluation of the patient’s risk factors, comorbidities, psychosocial status, results of neurologic examination and neuroimaging, video semiology, EEG findings including purely electrical seizures, and the results of monitoring other physiologic parameters (ECG [electrocardiography], blood pressure, orthostatic testing, blood sugar, and so on) as appropriate.
Ettinger, 1998⁷⁵	N=22; USA; age range 10-46; 77.2% female Inclusion: Patients undergoing continuous video EEG monitoring on EMU; diagnostic testing carried out; episodes associated with impaired consciousness Exclusion: No altered awareness; pregnancy; use of neuroleptic agents; unobtainable PRL results; SPECT scans compromised by movement artefact; unacquired SPECT because of failure to inject radioisotope at correct time Non-epilepsy population: PNES	Postictal and interictal single photon emission computed tomography (SPECT).	VIDEO EEG
Ettinger, 1999⁷⁴	N=39; USA; mean age 41.4 years; 76.9% female Inclusion: Adult patients evaluated at the Epilepsy Management site between 1996-98; epilepsy patients were 1) focal with secondary generalisation, or 2) generalised tonic clonic; documented epilepsy on video-EEG for epilepsy group, and patients with episodes characterised by bilateral motor activity and altered responsiveness, but without video-EEG evidence of seizures or without significant post-ictal prolactin elevation Exclusion: Learning disability; mixed epileptic/NES; patients with interictal headaches Non-epilepsy population: PNES	Symptom questionnaire. The responses to the question, ‘what symptoms do you have after a seizure?’ were reviewed	VIDEO EEG
Geut, 2017⁸¹	N= 104; Holland; mean age 47 years; 35.6% female Inclusion: Patients with unprovoked focal or generalized seizures who were admitted to the Clinical Neurophysiology department. Unprovoked seizures were defined as convulsive episodes occurring in the absence of precipitating factors. This included seizures of unknown aetiology as well as seizures in relation to a demonstrated pre-existing brain lesion (remote symptomatic seizure). Patients were subsequently selected in whom the routine EEG (including hyperventilation and photic simulation) was normal or did not show convincing IEDs, and either a sdEEG or an aEEG was requested. Finally, both groups were matched for age and gender. Exclusion: Patients younger than 6 years, patients with known epilepsy and patients with provoked seizures. Non-epilepsy population: any suspected of epilepsy	Ambulatory EEG Sleep deprived EEG	The patients’ clinical record was evaluated for age, sex, first seizure, start of anti-epileptic drugs, MRI or CT results and whether or not diagnosis of epilepsy was made with a follow up of one year. The diagnosis of epilepsy was based on the new ILAE criteria published in 2014
Geyer, 2000⁸²	N= 261; USA; mean age 33.75 years; 39.8% female Inclusion: Patients with TLE, FLE, generalised epilepsy or PNES undergoing video EEG Exclusion: not reported Non-epilepsy population: PNES	Existence of ictal pelvic thrusts	VIDEO EEG
Giorgi, 2013⁸⁴	N=210; Italy; mean age 41 years; 45% female Inclusion: Sleep deprived EEG (SD EEG) requested as a prospective evaluation for suspected epileptic seizures; previous standard waking EEG not showing any interictal abnormalities (IIAs); not under antiepileptic drugs until at least date of SD EEG; previous 1.5T MRI; minimum 1 year follow up; final diagnosis performed in the centre and defined as ‘non-epilepsy’, ‘focal epilepsy’ or ‘generalised epilepsy’. Exclusion: juvenile myoclonic epilepsy Non-epilepsy population: any suspected of epilepsy	Sleep deprived EEG	Final diagnosis obtained after collegial discussion by epileptologists in the centre with at least 5 years’ experience in clinical epilepsy. Diagnosis confirmed based on recurrence of clear epileptic unprovoked seizures. Single seizures not included. Most patients also given video EEG or 24 hour dynamic EEGs. Clinical records also evaluated
Gonzalez-Cuevas, 2019⁸⁶	N= 29; Spain; mean age 64.75years; 48.3% female Inclusion: >=18 years old; PCT acquired immediately following diagnosis; clinical or EEG diagnosis of status epilepticus (SE) established in ER or hospitalisation Exclusion: Patients with delayed PCT acquisition; allergy to iodinated contrast material; other contraindications for PCT Non-epilepsy population: any suspected of epilepsy	Perfusion computed tomography	Diagnosis by ictal EEG and clinical semiology
Goselink, 2019⁸⁷	N= 187; Holland; age and gender not reported Inclusion: All consecutive EEG recordings from both adult and pediatric patients with a clinical suspicion of non-convulsive status epilepticus (NCSE); all consecutive EEG recordings without a clinical suspicion but with an abnormal EEG were included in the clinically ‘not suspected for NCSE’ group. Exclusion: Patients with technically insufficient EEG recordings and EEG recordings lasting <30 minutes Non-epilepsy population: any suspected of epilepsy	EEG review using SalzburgSalzburg criteria	Expert opinion of another four neurophysiologists who had access to all clinical information, including laboratory tests, imaging studies, response to treatment, follow-up and outcome, as well as all EEG recordings. The consensus view held as the final diagnosis.
Hanrahan, 2018⁹⁰	N=12; mean age 40.6 years; 33% female Inclusion: Patients admitted to the Epilepsy Monitoring Unit for ‘spell classification’ who had videos taken of their events during the evaluation Exclusion: not reported Non-epilepsy population: any suspected of epilepsy	Clinical history. Videos of the seizure event captured during EMU evaluation.	The paper describes EMU diagnosis as entailing video-EEG, clinical history and witnessed semiology. The reported EMU-confirmed diagnosis was considered final. The diagnosis was also described as ‘established’.
Hendrickson, 2014⁹²	N= 354; USA; mean age not reported; 64.4% female Inclusion: Patients undergoing vEEG monitoring; participated in either neuropsychological or psychological testing; interviewed for panic attack criteria Exclusion: Unclear diagnosis; episodes secondary to another primary disorder; diagnosis of both PNES and epilepsy Non-epilepsy population: PNES	Number of panic attack symptoms	VIDEO EEG
Hoefnagels, 1991⁹⁴	N= 119; USA; mean age not reported; 47% female Inclusion: All consecutive patients (> 15 years of age) referred to the neurological department because of one or more episodes of transient loss of consciousness. Transient loss of consciousness was defined as an episode of less than one hour with inability to maintain posture and to recall events during the episode. Exclusion: Patients with loss of consciousness due to trauma or subarachnoid haemorrhage and patients with pre-diagnosis of epilepsy. Non-epilepsy population: any suspected of epilepsy	Routine interictal EEG. If patient <65years, had an additional hyperventilation test (40 breaths per minute for 3 minutes. End tidal CO2 level had to be <2.5% after hyperventilation. Blood gases measured. Hyperventilation test considered negative if end tidal CO2 did not restore to >90% baseline value after 3 minutes recovery. Standard ECG given and assessed as normal or abnormal according to the QT-interval. Laboratory examination of serum sodium, potassium, calcium, phosphate, glucose, urea, ESR, liver function and FBC.	A definitive diagnosis of seizure was given by: movements during loss of consciousness and identified clonic movements from a range of movements imitated by the interviewer; if an eyewitness observed automatisms, such as chewing or lip smacking, during loss of consciousness; if the patient reported an unequivocal aura, such as a strange smell, preceding the event; if the patient felt confused immediately after the event (inability to recognise familiar persons or environment);if the patient had tongue biting. Unclear if needed just one of these or all of these to trigger a diagnosis.
Huang, 2019⁹⁶	N=12; China; mean age 16 months; gender unclear Inclusion: Infants with paroxysmal events that had been videoed; resolution was high enough to ensure facial features were visible; all possible body movements were recorded; sound in videos is clear, and excessive ventilation sounds can be distinguished. Exclusion: No consent from caregivers; video >1 minute long (may impair public playback) Non-epilepsy population: any suspected of epilepsy	Medical record only Medical record plus 1 minute video of event	All corresponding descriptions, home videos, and VEEG reports were presented to two senior epileptologists blind to the study purpose, and they made diagnoses accordingly
Husain, 2020⁹⁷	N=17; USA; mean age 49.1 years; 21.1% female Inclusion: Patients with a history of ES or PNES admitted to one of 3 EMUs for routine seizure characterisation Exclusion: Any patients on whom intracranial EEG monitoring was used Non-epilepsy population: PNES	sEMG classification of seizure events by expert review. Single channel surface EMG (sEMG) attached unilaterally on the belly of the biceps. Graphical user interface allowed expert review Automated sEMG classification. As above, but using an automated decision tool. This generated a ‘seizure score from 0-25 with a threshold of 8 or above (= epilepsy)	VIDEO EEG
Jackson, 2016⁹⁹	N=219; Australia; median age 45 years; 40% female Inclusion: Patients referred by the ED to the adult first seizure clinic at Monash medical centre Exclusion: not reported Non-epilepsy population: any suspected of epilepsy	ED initial assessment by ED doctors	Final diagnosis: Index test data, PLUS MRI brain scans and EEG data that had been collected after ED discharge, with decision made by study authors (epilepsy specialists).
Jaraba, 2019¹⁰⁰	N=55; Spain; mean age 62.1 years; 38.1% female Inclusion: All patients undergoing 99mTc-hexamethyl propyleneamine oxime [HMPAO] single photo emission computed tomography [SPECT] [HMPAO-SPECT] as part of their diagnostic workup in the centre; clinical suspicion of NCSE Exclusion: Patients with sub-optimal EEG recordings; patients with NCSE because of hypoxic-anoxic aetiology; no consensus on diagnosis; where EEG and HMPAO-SPECT were not done simultaneously Non-epilepsy population: any suspected of epilepsy	Ictal HMPAO SPECT scans (visual) Ictal HMPAO SPECT scans (quantitative) Ictal EEG using Salzburg criteria	Patients were classified as NCSE or non-NCSE following a consensus decision based on all clinical and paraclinical data, including EEG readings, laboratory data, therapeutic response, follow up and final outcome. Two clinicians evaluated these data independently blinded to HMPAPO-SPECT results. A third clinician was used to resolve conflicts.
Keezer, 2016¹⁰²	N=72; Canada; mean age 35 years; 61% female Inclusion: All patients undergoing a prolonged ambulatory EEG (paEEG); medical record at the MNI to allow expert to ascertain clinical diagnosis of epilepsy or not Exclusion: not reported Non-epilepsy population: any suspected of epilepsy	Routine EEG. Prolonged ambulatory EEG (paEEG).	One neurologist, a fellow of the Royal College of Physicians of Canada, reviewed medical records to identify those individuals with epilepsy. To minimize verification bias (i.e., constructing the reference standard with prior knowledge of the index test results), the assessor relied on the documented medical history and event semiology. Additional data collected were subject age, sex, epilepsy aetiology, the use of antiepileptic drug(s), and reason for referral by the treating physician
Khan, 2009¹⁰⁷	N=50; USA; mean age not reported; 57% female Inclusion: Patients being evaluated for a medically refractory seizure disorder; aged 18 or older; able to undergo hypnosis (able to hear and see) Exclusion: Pregnancy; learning disability; psychosis; under the influence of illicit substances Non-epilepsy population: any suspected of epilepsy	Patients underwent the Hypnotic Induction Profile	VIDEO EEG
Kimiskidis, 2017¹⁰⁹	N= 31; Greece; mean age 28 years; 54.8% female Inclusion: Patient group: Patients with GGE; passed TASS questionnaire except epilepsy-related questions; both clinical and EEG features consistent with GGE; at least 2 seizures and on AEDs Exclusion: Other CNS disorders; comorbid conditions; EEG evidence of focal abnormalities; slow spike and wave discharges or triphasic patterns; centrally acting drugs other than AEDs; past or present substance/ETOH abuse Non-epilepsy population: healthy controls	Paired pulsed transcranial magnetic stimulation	Diagnosis by 2 experienced epileptologists who reached consensus based on clinical and laboratory data.
Knox, 2018¹¹¹	N=340; USA; mean age 3.9 years; gender not reported Inclusion: First time vEEG without capturing a habitual event; at least 1 year of FU; on hospital database Exclusion: Neonates; diagnosis of epilepsy that predated the initial vEEG study by >1 month; no history of paroxysmal events Non-epilepsy population: any suspected of epilepsy	No event video EEG	Final definitive diagnosis based on full medical records and a minimum of 1 clinic visit in 1 year of follow up. Often unblinded to EEG results
Koren, 2018¹¹⁴	N=85; Austria; mean age 58.9 years; 51.8% female Inclusion: Neurological critical care patients with clinically suspected NCSE [unexplained deterioration or fluctuation of consciousness, subtle motor activity (persistent or fluctuating muscle twitching of the face or extremities, manual and oral automatisms) as well as pupillary and ocular movement abnormalities (nystagmus, hippus, mydriasis, or sustained eye deviation). Exclusion: not reported Non-epilepsy population: any suspected of epilepsy	Several early findings (first 30 minutes of EEG recordings) were tested: Early sporadic epileptiform discharges (SED) Early rhythmic and periodic EEG patterns of ‘ictal-interictal uncertainty’ (RPPIIIU) Early SED or RPPIIU Clinical signs of non-convulsive seizures (NCS) Early SED or RPPIIU and clinical signs of NCS Early SED, RPPIIU, or clinical signs of NCS	Critical care continuous EEG (for detection of NCSE). Used 21 electrodes according to the 10-20 system. Recordings performed as soon as possible following clinical suspicion of NCSE (all within 12 hours). EEG data classified according to the ACNS SCCET. Mean recording time was 72 (67) hours [range 5-388 hours]
Kusmakar, 2019¹¹⁶	N=79; Australia; mean age 31.6 years; 60% female Inclusion: Patients undergoing VIDEO EEG; history of events that mimicked generalised seizures or events characterised by the presence of bilateral convulsions Exclusion: Patients having intracranial monitoring or with a psychiatric disorder Non-epilepsy population: PNES	Wrist accelerometer	Decided by consensus between 2-6 epileptologists, where a decision was made based on clinical history, neuropsychiatric evaluation, neuroimaging, Video EEG for 3 days and observed seizure semiology
Leitinger, 2016¹²⁴	N= 120; Denmark/Austria; median age 65 years; 47% female Inclusion: Aged 4 months or older (if from tertiary centre); 18 years or older (if from the 2 secondary care centres); clinical suspicion of non-convulsive status epilepticus, having a history of decreased cognition/consciousness for at least 10 minutes. Exclusion: Participants with technically insufficient EEG recordings; EEG recordings lasting <20 minutes. Non-epilepsy population: any suspected of epilepsy	Routine EEG using Salzburg criteria	The reference standard was inferred from all clinical and para-clinical data, including EEG readings (but not the results of Salzburg criteria), laboratory data, neuroimaging data, therapeutic response, follow-up, and final outcome. For all patients and recordings, two authors evaluated these data independently, while blinded to the Salzburg criteria scorings
Li, 2017¹²⁵	N=54; USA; age and gender not reported Inclusion: ED discharge diagnosis of ‘generalised seizures’ or ‘generalised shaking episodes’; aged >=18 years; well documented spell onset within 24 hours of a basic metabolic panel drawn in the ED Exclusion: Other documented active medical problems that could cause acidosis and confound the analysis, such as sepsis, alcohol or medicine toxicity Non-epilepsy population: PNES	Anion gap	Abnormal interictal EEG showing epileptiform discharges, plus with a documented semiology of their event consistent with a generalised convulsive seizure. Subjects diagnosed as PNES if video EEG confirmed this.
Manni, 2008¹³¹	N= 71; Italy; mean age 54years; 15.5% female Inclusion: Patients with undefined (epileptic or parasomnia) nocturnal paroxysmal motor-behavioural episodes attending the Sleep Medicine and Epilepsy Unit (an outpatient facility) at the IRCCS “C. Mondino Institute of Neurology” Foundation in Pavia, Italy; final diagnosis of arousal parasomnias, NFLE or idiopathic RBD. Exclusion: not reported Non-epilepsy population: parasomnias or idiopathic RBD	FLEP scale	VIDEO EEG
McGinty, 2021¹³²	N= 219; UK; mean age 49 years; 49.8% female Inclusion: Consecutive adult patients with a diagnosis of new-onset focal epilepsy and their first seizure within the previous 12 months Exclusion: not reported Non-epilepsy population: any suspected of new onset focal epilepsy	ACE attention domain APE2 score	Detection of Neuronal surface-directed antibodies (NSAb)
Mueller, 2013¹³⁶	N=80; USA; mean age 35.9 years; 65% female Inclusion: Not reported, though all patients were reported to be seizure free for at least 24 hours before the MRI study. Exclusion: not reported Non-epilepsy population: healthy controls	4T MRI	Seizure semiology and prolonged ictal and interictal Video/EEG/Telemetry (VET)
Naganur, 2019¹³⁷	N=11; Australia; mean age (seizures/PNES) 20/24years; 58.3% female Inclusion: Patients admitted for VEM for the investigation of possible epilepsy were eligible for inclusion. Patients were eligible for inclusion if they experienced one of their typical clinical events of at least 20 seconds (s) in duration in which there was sustained, rhythmic or arrhythmic movements affecting at least one limb. This included patients with purely tonic or hyper motor movements. Exclusion: Patients experiencing solely non-convulsive seizures were excluded. Non-epilepsy population: PNES	Wrist accelerometer data	VIDEO EEG
Noe, 2012¹⁴³	N=439; USA; mean age 47.9 years; 64% female Inclusion: Patients admitted to EMU for spell classification Exclusion: Subjects with a known diagnosis of epilepsy admitted to EMU for pre-surgical evaluation, medication adjustment, status epilepticus, or seizure quantification. Non-epilepsy population: any suspected of epilepsy	Impression of the admitting epidemiologist, based on review of history, physical and available diagnostic testing as documented in the medical record prior to vEEG.	VIDEO EEG
Okazaki, 2018¹⁴⁴	N= 57; USA; mean age 42 years; 52.6% female Inclusion: People aged >18 admitted to having scalp continuous vEEG monitoring for episode classification Exclusion: People whose monitoring session was inconclusive because of the lack of recorded events Non-epilepsy population: any suspected of epilepsy	Epifinder application – a clinical decision support tool.	VIDEO EEG
Oliva, 2008¹⁴⁵	N=84; Australia; mean age 38.0 years; 50% female Inclusion: Patients admitted to Royal Melbourne Hospital for inpatient video monitoring, in whom at least 1 convulsive event was captured Exclusion: not reported Non-epilepsy population: any suspected of epilepsy	Existence of oral lacerations and incontinence. Information collected by medical scientists via direct questioning and examination of the patient after a convulsive event.	VIDEO EEG
Ottman, 2010¹⁴⁶	N=342; USA; mean age 54 years; 61% female Inclusion: All residents of the city of Rochester, MN, U.S.A., who were born in 1920 or later and had incidence of either epilepsy (two or more unprovoked seizures) or an isolated unprovoked seizure between 1935 and 1994. Exclusion: not reported Non-epilepsy population: healthy controls	General screening interview for epilepsy	A comprehensive review of the medical records of each case or control was carried out. Abstraction involved initial review by trained nurse abstractors followed by expert review by the study epileptologists and provided detailed information for the duration of each subject’s residence in the Rochester area, including all outpatient examinations, home and emergency room visits, hospitalization records, laboratory tests, and neurologic and other special examinations.
Rawlings, 2017¹⁵⁸	N= 293; UK; mean age 43.8 years; 73.0% female Inclusion: Patients with epilepsy or PNES supported by video EEG recordings of typical seizures involving TLOC identified from patient databases; patients with a diagnosis of recurrent syncope supported by pathophysiological evidence Exclusion: Patients unable to complete the questionnaire without help (learning disability) Non-epilepsy population: PNES or syncope	Panic measures. This was captured by the Paroxysmal Event Profile – this consists of 86 Likert style questions about symptoms, 7 of which were focussed on panic symptoms.	VIDEO EEG
Renzel, 2016¹⁵⁹	N= 237; Switzerland; mean age 38 years; 39.2% female Inclusion: Age >16; at least one routine EEG because of suspected epilepsy and been subsequently examined with an EEG SD (24 hours); full documentation of history, EEG and diagnosis available; no diagnosis made before SD EEG; no specific epileptiform changes in the EEG before SD-EEG; documented cerebral imaging via MRI within 2 years of EEG recordings Exclusion: Patients declined use of their data; no final diagnosis available; no adequate documentation of the medication taken; use of highly potent neuroleptic drugs Non-epilepsy population: any suspected of epilepsy	Sleep deprived EEG	Established after collegial discussion for each case by the study investigators according to the ILAE guidelines
Reuber, 2009¹⁶¹	N=20; UK; mean age 36.9 years; 65% female Inclusion: Refractory seizure disorders; referred for Video EEG; uncertainty between epilepsy and PNES; seizure captured by video; ictal EEG allowed unequivocal diagnosis of epilepsy or PNES Exclusion: Combined epilepsy and PNES; admitted for epilepsy surgery evaluation; non-fluent English; unable to complete self-report measures Non-epilepsy population: PNES	Linguistic aqnalysis	VIDEO EEG
Reuber, 2016¹⁶⁰	N=300; UK; mean age 43.5years; 73% female Inclusion: Patients with epilepsy or PNES supported by video EEG recordings of typical seizures involving TLOC identified from patient databases; patients with a diagnosis of recurrent syncope supported by pathophysiological evidence Exclusion: not reported Non-epilepsy population: PNES or syncope	Paroxysmal Event Profile Questionnaire – 86 items focussing on TLOC manifestations, plus 7 further questions related to demographic and clinical features.	VIDEO EEG
Rosenow, 1998¹⁶³	N=40; Germany; mean age 103.4 months; gender not reported Inclusion: Children presenting with a chief complaint of staring spells Exclusion: not reported Non-epilepsy population: any suspected of epilepsy	Symptom questionnaire.	VIDEO EEG
Rowberry, 2020¹⁶⁶	N=101; UK; median age 4 years; 47.5% female Inclusion: Patients under 18 years identified by PICU clinicians to be at risk of epileptic seizures and commenced on Quantitative EEG (qEEG) Exclusion: Patients with decompressive craniectomy and allergy to collodion glue Non-epilepsy population: any suspected of epilepsy	Quantitative EEG interpreted in real time by PICU clinicians	A clinical neurophysiologist retrospectively reviewed each qEEG recording to identify epilepsy seizures. The neurophysiologist had access to the same electrophysiology information available to the PICU clinicians. This included the raw EEG.
Schmidt, 2016¹⁷¹	N=68; UK; age 16-59 years; gender not reported Inclusion: IGE individuals were drug naïve Exclusion: not reported Non-epilepsy population: any suspected of epilepsy	Computational biomarker based on extent of synchrony between EEG channels and the normalised power spectrum from a short resting state interictal EEG	This was a ‘case-control’ design where 38 healthy controls and 30 people with a diagnosis of Idiopathic Generalised Epilepsy (IGE) were recruited. A diagnosis of epilepsy was confirmed in each IGE case by an experienced epilepsy specialist through observation of typical generalized spike-wave (GSW) activity on EEG either spontaneously or following hyperventilation or photic stimulation. For 10 of these people, the diagnosis was confirmed following an initial routine EEG. For the remaining 20, diagnosis was confirmed following sleep-deprived or longer-term EEG monitoring (including sleep). Similar healthy control EEG was collected at King’s College Hospital EEG department.
Sen, 2007¹⁷⁶	N = 36; UK; age and gender unclear Inclusion: Epilepsy or PNES Exclusion: not reported Non-epilepsy population: PNES	Existence of postictal stertorous breathing	Full use of all clinical data collected over 18 months
Seneviratne, 2017¹⁷⁷	N= 138; Australia; mean age 43 years; 52.2% female Inclusion: All patients undergoing monitoring at the EMU of Monash Medical Centre; adults aged >=18; diagnosed with PNES or ES Exclusion: Events with subjective symptoms or without obvious semiological features; electrographic epileptic seizures without clinical semiology Non-epilepsy population: PNES	Ictal duration	VIDEO EEG
Sierra-Marcos, 2011¹⁷⁹	N= 131; Spain; mean age 52.4years; 45% female Inclusion: Adult patients who consulted consecutively for a new onset seizure to the ER; stereotyped paroxysmal spell highly suggested an epileptic seizure Exclusion: Patients with previous seizures Non-epilepsy population: any suspected of epilepsy	Early EEG Follow up routine EEG Sleep deprived EEG CT	Full clinical, EEG, CT, video EEG AND 12 months follow up
Simani, 2018¹⁸⁰	N=82; Iran; mean age 30.9 years; 53.6% female Inclusion: Patients with a history of recurrent seizures, admitted to EMU for further evaluation; control group comprised healthy volunteers with no history of seizure. Exclusion: Patients with other medical, neurologic or psychiatric diseases, or history of recent head trauma; medications other than AEDs or psychoactive drugs Non-epilepsy population: any suspected of epilepsy	Post-seizure serum glial fibrillary astrocytic protein (GFAP) serum levels	VIDEO EEG
Slater, 1995¹⁸¹	N=49; USA; age and gender not reported Inclusion: Age >=18; patients admitted to EEG video telemetry unit. Exclusion: not reported Non-epilepsy population: PNES	Wilkus classification guideline: A patients has pseudo seizures if any of the following are true: a) hysteria or hypochondriasis score >=70 and one of the two highest points in the profile (disregarding the masculinity-femininity and social introversion scales, b) hysteria or hypochondriasis score >=80 and not necessarily among the two highest points, c) hysteria and hypochondriasis both >59 and both 10 points higher than the depression scale.	VIDEO EEG
Stroink, 2003¹⁸⁴	N= 760; Holland; ages 1 month to 16 years; gender not reported Inclusion: All children aged 1 month to 16 years referred by GP or paediatrician at participating hospital for a single seizure or suspected epilepsy Exclusion: Children with only neonatal, febrile or other acute symptomatic seizures; children referred from other hospitals for a second opinion Non-epilepsy population: any suspected of epilepsy	Clinical diagnosis: Attending paediatric neurologist completed an extensive questionnaire on description of events, including postictal signs, possible provoking factors, medical and family history. Standard EEG performed in each child. If no epileptiform discharges a recording after partial sleep deprivation was made, or in small children during a daytime nap.	Use of original data plus information gained over 5 years of follow up (if epilepsy originally diagnosed), 2 years of follow up (if single seizure) or 1 year of follow up (if no epilepsy diagnosis or single event at baseline).
Swartz, 2002¹⁸⁶	N=462; USA; age and gender not reported Inclusion: Patients referred to PET facility Exclusion: No seizures within 72 hours Non-epilepsy population: any suspected of epilepsy	Positron Emission Tomography with 2-deoxy-2[18F] fluro-D-glucose (FDG-PET)	VIDEO EEG
Syed, 2011¹⁹¹	N=35; USA; mean age 37.0 years; 60% female Inclusion: Seizure patients scheduled for vEEG; VEEG recorded epilepsy or PNES during stay Exclusion: not reported Non-epilepsy population: PNES	Epileptologist blinded and independent review of seizure videos in terms of the following semiological signs: 1) eye-opening or widening at onset of seizure, 2) abrupt onset, 3) post-ictal confusion/sleep Eye-witness accounts of seizure in terms of the following semiological signs: 1) eye-opening or widening at onset of seizure, 2) abrupt onset, 3) post-ictal confusion/sleep	VIDEO EEG
Tatum, 2020¹⁹³	N=44; USA; mean age 45.1 years; 70% female Inclusion: 18 years or older; voluntary consent; had completed a history assessment and physical examination; outpatients referred with events that could be epilepsy; submitted an outpatient smartphone video of their primary ictal event; underwent gold standard test of video-EEG; >95% of each survey completed by reviewers; had a final diagnosis Exclusion: <18 years; pregnant; incomplete or absent history/physical examination; no smartphone video; did not undergo gold standard; confirmed history of mixed epileptic and non-epileptic events; declined study participation; no informed consent Non-epilepsy population: any suspected of epilepsy	Patients provided a witness-generated outpatient smartphone video. History and physical examination done by 3 experts, lasting an average of 60 minutes	VIDEO EEG
Tews, 2015¹⁹⁴	N= 248; Germany; mean age 6.2 years; 45.2% female Inclusion: first afebrile seizure; aged 1 mo. to 18 yrs. not suffering from pre-existing neurological disorders Exclusion: situation-related or acute symptomatic seizures resulting from toxic, metabolic, infectious or traumatic reasons were excluded. Non-epilepsy population: any suspected of epilepsy	EEG MRI	Seizure recurrence at 48 months, with use of the International League Against Epilepsy definitions to clinically classify patients as having epilepsy
Thompson, 2010¹⁹⁶	N= 184; USA; mean age 37 years; 67.4% female Inclusion: Patients completing the Personality Assessment Inventory (PAI) and video EEG at the regional epilepsy centre. Exclusion: Not diagnosed by video EEG as either epilepsy or PNES Non-epilepsy population: PNES	Psychological indices PNES (Psychogenic nonepileptic seizures); threshold for PNES >=1 SOM-C (conversion); threshold for PNES >=70 SOM (somatic complaints); threshold for PNES >=70 SOM-S (somatisation); threshold for PNES >=70 DEP (Depression); threshold for PNES >=60 DEP-P (Depression-physiological); threshold for PNES >=70 ANX-P (Anxiety-Physiological); threshold for PNES >=60	VIDEO EEG
Tyson, 2018¹⁹⁹	N=105; USA; mean age 36.9 years; 54.3% female Inclusion: Patients with neuropsychological assessments, and data on psychometric testing Exclusion: not reported Non-epilepsy population: PNES	Multivariate model of psychometric testing, using 4 measures of cognitive ability – vocabulary, information, Boston naming test and letter fluency)	EEG evidence of ES, with neurological exam, seizure semiology and neuroradiological findings. Video EEG used to exclude PNES so likely that video EEG was used for all, although not directly stated.
van Diessen, 2013²⁰⁰	N=70; Holland; mean age 10 years; 31.4% female Inclusion: One or more suspected epileptic event(s) were eligible for our study. Children included who were eventually diagnosed with new onset partial epilepsy. Exclusion: Children with neurological or psychiatric comorbidities, including developmental delay Non-epilepsy population: control group not suspected of epilepsy	Routine interictal EEG recording, using international 10-20 system. Functional network approach: Periods of resting-state EEG, free of abnormal slowing or epileptiform activity, were selected to construct functional networks of correlated activity.	The clinical diagnosis of epilepsy was defined by at least two unprovoked seizures within one year, judged by two neurologists to be of epileptic origin.
Varma, 1996²⁰³	N= 20; UK; mean age 35.3years; 50% female Inclusion: Patients referred to neurosurgery unit and diagnoses with NES or epilepsy; diagnosis based on video EEG findings Exclusion: People with dual epilepsy/PNES; brain lesions on CT/MRI Non-epilepsy population: PNES	Hexamethyl propylene amine oxime single photon emission tomography (HMPAO SPECT) brain imaging	VIDEO EEG
Verhoeven, 2018²⁰⁵	N=75; Switzerland, Belgium and Austria; mean age 31.7 years; 52.5% female Inclusion: drug resistant TLE, or ‘healthy’ Exclusion: not reported Non-epilepsy population: any suspected of epilepsy	Resting-state high-density EEG recording data was used. Epochs without interictal spikes were selected. The cortical source activity was obtained for 82 regions of interest and whole brain directed functional connectivity was estimated in the theta, alpha and beta frequency bands. These connectivity values were then used to build a classification system based on two two-class Random Forests classifiers: TLE vs healthy controls and left vs right TLE.	Drug resistant TLE was definitively diagnosed as follows: unilateral anteromedial localization of the epileptogenic zone confirmed by good surgical outcome (Engel’s class I or II, after at least 12 months post-operative follow-up), intracranial EEG or concordant presurgical evaluation methods and the existence of at least a 10–15 min resting state eyes-closed high-density EEG recording (96–256 channels).
Vukmir, 2004²⁰⁹	N=200; USA; age and gender not reported Inclusion: Patients who presented to the emergency department with a clinical symptom complex consistent with seizure, manifested as near or total loss of consciousness, accompanied by abnormal motor activity and/or a post-ictal phase. Exclusion: <18 years Non-epilepsy population: any suspected of epilepsy	Serum prolactin level	A hospital discharge diagnosis of seizure either initially or at the end of the stay. The diagnosis was recorded from ED records if discharged or inpatient discharge record if admitted. The presence of an abnormal electroencephalogram indicated by abrupt onset and termination of repetitive rhythmic activity usually consisting of a sharp or spike wave pattern, during the hospital stay if performed was included as well.
Watson, 2012²¹³	N= 630; UK; mean age 49.5 years; gender not reported Inclusion: People with EEGs done in the department between July 2006 to December 2009 Exclusion: not reported Non-epilepsy population: any suspected of epilepsy	Routine EEG	Final diagnosis of epilepsy/ no epilepsy, based on all information, including laboratory results, MRI/CT/X ray imaging.
Wilkus, 1984²¹⁵	N=20; USA mean age 28.2 years; gender unknown Inclusion: Patients referred for inpatient EEG/CCTV monitoring Exclusion: not reported Non-epilepsy population: PNES	See Wilkus classification guideline (Slater, 1995)	VIDEO EEG
Willert, 2004²¹⁶	N=52; Germany; mean age 34.7years; 41.6% female Inclusion: Single seizures with an interval of at least 24 hours before and after the seizure; normal levels of NSE, PRL and CK at baseline Exclusion: Acute disorders of the CNS or endocrinological diseases; pregnancy; medication other than anticonvulsants Non-epilepsy population: PNES	Serum neuron-specific enolase (NSE) Serum prolactin (PRL) Serum creatine kinase (CK)	VIDEO EEG

Table 3Clinical evidence summary: diagnostic test accuracy of different symptoms/signs/semiology for detection of epilepsy

Where detection is of a specific type of epilepsy, rather than epilepsy overall, this is stated clearly in the first column. Each index test is positive if the described symptom is present.

Index Test	Number of studies	n	Interpreter of index test	Gold standard used in study	Sensitivity (95% CI)	Specificity (95% CI)	Risk of bias	Indirectness	Inconsistency	Imprecision	GRADE
Tongue biting / oral lacerations during seizure	2²⁵ ¹⁴⁵	194	NR/ medical scientist	Video EEG Non-epilepsy group: syncope / population suspected of epilepsy	0.22 [0.10, 0.39] 0.26 [0.16, 0.38]	0.99 [0.93, 1.00] 1.00 [0.81, 1.00]	Sensitivity
							Very serious^a	serious^b	NA	None^c	VERY LOW
							Specificity
							Very serious^a	serious^b	NA	None^c	VERY LOW
Incontinence during seizure	1¹⁴⁵	84	Medical scientist	Video EEG Population suspected of epilepsy but no definite differential diagnoses	0.23 [0.13, 0.35]	0.94 [0.73, 1.00]	Sensitivity
							serious^a	none	NA	None^c	MOD
							Specificity
							serious^a	none	NA	serious^c	LOW
Urine loss during seizure DETECTING ABSENCE SEIZURES IN INFANTS	1¹⁶³	40	Physician	Video EEG Non-epilepsy group: population suspected of epilepsy	0.12 [0.01, 0.36]	1.00 [0.85, 1.00]	Sensitivity
							serious^a	none	NA	None^c	MOD
							Specificity
							serious^a	none	NA	serious^c	LOW
Oral lacerations AND incontinence during seizure	1¹⁴⁵	84	Medical scientist	Video EEG Population suspected of epilepsy but no definite differential diagnoses	0.08(0.03-0.18)	1.0(0.78-1.0)	Sensitivity
							serious^a	none	NA	None^c	MOD
							Specificity
							serious^a	none	NA	serious^c	LOW
Sign observed by epileptologist on video during seizure - eye opening or widening at onset	1¹⁹¹	36	epileptologist	Video EEG Non-epilepsy group: PNES	1.00 [0.79, 1.00]	0.85 [0.62, 0.97]	Sensitivity
							serious^a	serious^b	NA	serious^c	VERY LOW
							Specificity
							serious^a	serious^b	NA	serious^c	VERY LOW
Sign observed by epileptologist on video during seizure - abrupt onset	2³⁹^, ¹⁹¹	79	epileptologist	Video EEG Non-epilepsy group: PNES	0.94 [0.70, 1.00] 1.00 [0.87, 1.00]	0.55 [0.32, 0.77] 0.13 [0.02, 0.38]	Sensitivity
							serious^a	serious^b	none	serious^c	VERY LOW
							Specificity
							serious^a	serious^b	none	serious^c	VERY LOW
Sign observed by epileptologist on video during seizure – postictal confusion/sleep	1¹⁹¹	36	epileptologist	Video EEG Non-epilepsy group: PNES	0.81 [0.54, 0.96]	0.70 [0.46, 0.88]	Sensitivity
							serious^a	serious^b	NA	serious^c	VERY LOW
							Specificity
							serious^a	serious^b	NA	serious^c	VERY LOW
Sign observed by epileptologist on video during seizure – eyes fixed	1¹⁹¹	36	epileptologist	Video EEG Non-epilepsy group: PNES	0.57 [0.34, 0.77]	0.92 [0.62, 1.00]	Sensitivity
							serious^a	serious^b	NA	serious^c	VERY LOW
							Specificity
							serious^a	serious^b	NA	serious^c	VERY LOW
Sign observed by epileptologist on video during seizure – unilateral head turning	1¹⁹¹	36	epileptologist	Video EEG Non-epilepsy group: PNES	0.30 [0.13, 0.53]	1.00 [0.74, 1.00]	Sensitivity
							serious^a	serious^b	NA	None^c	LOW
							Specificity
							serious^a	serious^b	NA	serious^c	VERY LOW
Sign observed by epileptologist on video during seizure – non-sensical speech	1¹⁹¹	36	epileptologist	Video EEG Non-epilepsy group: PNES	0.00 [0.00, 0.15]	0.92 [0.62, 1.00]	Sensitivity
							serious^a	serious^b	NA	None^c	LOW
							Specificity
							serious^a	serious^b	NA	serious^c	VERY LOW
Sign observed by epileptologist on video during seizure – clenched mouth	1¹⁹¹	36	epileptologist	Video EEG Non-epilepsy group: PNES	0.09 [0.01, 0.28]	0.25 [0.05, 0.57]	Sensitivity
							serious^a	serious^b	NA	None^c	LOW
							Specificity
							serious^a	serious^b	NA	None^c	LOW
Sign observed by epileptologist on video during seizure – hand automatisms	2³⁹^, ¹⁹¹	79	epileptologist	Video EEG / surgical or long term follow up Non-epilepsy group: PNES	0.26 [0.10, 0.48] 0.52 [0.32, 0.71]	1.00 [0.74, 1.00] 0.94 [0.70, 1.00]	Sensitivity
							serious^a	serious^b	NA	serious^c	VERY LOW
							Specificity
							serious^a	serious^b	NA	serious^c	VERY LOW
Sign observed by epileptologist on video during seizure – ictal scream	1¹⁹¹	36	epileptologist	Video EEG Non-epilepsy group: PNES	0.22 [0.07, 0.44]	1.00 [0.74, 1.00]	Sensitivity
							serious^a	serious^b	NA	None^c	LOW
							Specificity
							serious^a	serious^b	NA	serious^c	VERY LOW
Sign observed by epileptologist on video during seizure - grasping	1¹⁹¹	36	epileptologist	Video EEG Non-epilepsy group: PNES	0.09 [0.01, 0.28]	1.00 [0.74, 1.00]	Sensitivity
							serious^a	serious^b	NA	None^c	LOW
							Specificity
							serious^a	serious^b	NA	serious^c	VERY LOW
Sign observed by epileptologist on video during seizure – post-ictal nosewiping	1¹⁹¹	36	epileptologist	Video EEG Non-epilepsy group: PNES	0.09 [0.01, 0.28]	1.00 [0.74, 1.00]	Sensitivity
							serious^a	serious^b	NA	None^c	LOW
							Specificity
							serious^a	serious^b	NA	serious^c	VERY LOW
Sign observed by epileptologist on video during seizure - epostical aphasia	1¹⁹¹	36	epileptologist	Video EEG Non-epilepsy group: PNES	0.09 [0.01, 0.28]	1.00 [0.74, 1.00]	Sensitivity
							serious^a	serious^b	NA	None^c	LOW
							Specificity
							serious^a	serious^b	NA	serious^c	VERY LOW
Sign observed by epileptologist on video during seizure – postictal snoring	2¹⁶^, ¹⁹¹	104	epileptologist	Video EEG Non-epilepsy group: PNES	0.35 [0.16, 0.57] 0.34 [0.20, 0.50]	1.00 [0.74, 1.00] 1.0 [00.86, 1.00]	Sensitivity
							Very serious^a	serious^b	NA	None^c	VERY LOW
							Specificity
							Very serious^a	serious^b	NA	serious^c	VERY LOW
Sign observed by epileptologist on video during seizure – abrupt offset	1³⁹^, ¹⁹¹	79	epileptologist	Video EEG Non-epilepsy group: PNES	0.75^e 0.74 [0.54, 0.89]	0.7^e 0.31 [0.11, 0.59]	Sensitivity
							serious^a	serious^b	NA	Serious^c	VERY LOW
							Specificity
							serious^a	serious^b	NA	none^c	LOW
Sign observed by epileptologist on video during seizure – continuous movements	1¹⁹¹	36	epileptologist	Video EEG Non-epilepsy group: PNES	0.57 [0.34, 0.77]	0.67 [0.35, 0.90]	Sensitivity
							serious^a	serious^b	NA	serious^c	VERY LOW
							Specificity
							serious^a	serious^b	NA	serious^c	VERY LOW
Sign observed by epileptologist on video during seizure – eyes rolled back into head	1¹⁹¹	36	epileptologist	Video EEG Non-epilepsy group: PNES	0.52 [0.31, 0.73]	0.67 [0.35, 0.90]	Sensitivity
							serious^a	serious^b	NA	serious^c	VERY LOW
							Specificity
							serious^a	serious^b	NA	serious^c	VERY LOW
Upward eye movements DETECTING ABSENCE SEIZURES IN INFANTS	1¹⁶³	40	Physician	Video EEG Non-epilepsy group: population suspected of epilepsy	0.35 [0.14, 0.62]	0.91 [0.72, 0.99]	Sensitivity
							serious^a	none	NA	serious^c	LOW
							Specificity
							serious^a	none	NA	serious^c	LOW
Sign observed by epileptologist on video during seizure – postictal exhaustion	1¹⁹¹	36	epileptologist	Video EEG Non-epilepsy group: PNES	0.52 [0.31, 0.73]	0.42 [0.15, 0.72]	Sensitivity
							serious^a	serious^b	NA	serious^c	VERY LOW
							Specificity
							serious^a	serious^b	NA	serious^c	VERY LOW
Sign observed by epileptologist on video during seizure – postictal stertorous/loud/deep breathing	4¹⁶^, ³⁹^, ¹⁷⁶^, ¹⁹¹	183	epileptologist	Video EEG, or overall clinical findings over prolonged follow up Non-epilepsy group: PNES	0.43 [0.23, 0.66] 0.22 [0.09, 0.42] 0.52 [0.37, 0.68] 0.96[0.80, 1.0] Pooled (95% CrIs): 0.57(0.14 – 0.93)	0.50 [0.21, 0.79] 1.00 [0.79, 1.00] 0.79[0.58, 0.93] 1.0 [0.90,1.0] Pooled (95% CrIs): 0.89 (0.46 – 0.99)	Sensitivity
							Very serious^a	serious^b	none	Very serious^c	VERY LOW
							Specificity
							Very serious^a	serious^b	none	Very serious^c	VERY LOW
Sign observed by epileptologist on video during seizure – looking around	1¹⁹¹	36	epileptologist	Video EEG Non-epilepsy group: PNES	0.48 [0.27, 0.69]	0.25 [0.05, 0.57]	Sensitivity
							serious^a	serious^b	NA	serious^c	VERY LOW
							Specificity
							serious^a	serious^b	NA	None^c	LOW
Sign observed by epileptologist on video during seizure – epileptic aura	1¹⁹¹	36	epileptologist	Video EEG Non-epilepsy group: PNES	0.5^e	0.17^e	Sensitivity
							serious^a	serious^b	NA	NA^c	LOW
							Specificity
							serious^a	serious^b	NA	NA^c	LOW
Sign observed by epileptologist on video during seizure - gradual behavioural build-up to peak intensity, but within 70 seconds	1³⁹	43	epileptologist	Surgical or long term follow up Non-epilepsy group: PNES	0.81 [0.62, 0.94]	0.94 [0.70, 1.00]	Sensitivity
							serious^a	serious^b	NA	serious^c	VERY LOW
							Specificity
							serious^a	serious^b	NA	serious^c	VERY LOW
Sign observed by epileptologist on video during seizure – eyes closed at peak	1³⁹	43	epileptologist	Surgical or long term follow up Non-epilepsy group: PNES	0.00 [0.00, 0.14]	0.20 [0.04, 0.48]	Sensitivity
							serious^a	serious^b	NA	none	LOW
							Specificity
							serious^a	serious^b	NA	none	LOW
Sign observed by epileptologist on video during seizure – waxing / waning event tempo	1³⁹	43	epileptologist	Surgical or long term follow up Non-epilepsy group: PNES	0.04 [0.00, 0.19]	0.31 [0.11, 0.59]	Sensitivity
							serious^a	serious^b	NA	none	LOW
							Specificity
							serious^a	serious^b	NA	none	LOW
Sign observed by epileptologist on video during seizure – non-synchronous movements	1³⁹	43	epileptologist	Surgical or long term follow up Non-epilepsy group: PNES	0.07 [0.01, 0.24]	0.56 [0.30, 0.80]	Sensitivity
							serious^a	serious^b	NA	none	LOW
							Specificity
							serious^a	serious^b	NA	serious^c	VERY LOW
Sign observed by epileptologist on video during seizure – side to side head movements	1³⁹	43	epileptologist	Surgical or long term follow up Non-epilepsy group: PNES	0.00 [0.00, 0.13]	0.75 [0.48, 0.93]	Sensitivity
							serious^a	serious^b	NA	none	LOW
							Specificity
							serious^a	serious^b	NA	Very serious^c	VERY LOW
Sign observed by epileptologist on video during seizure – pelvic thrusting	4¹⁶^, ³⁹^, ⁸²	594	Epileptologist/neurologist	Surgical or long term follow up / Video EEG Non-epilepsy group: PNES and other Epi types	0.04 [0.00, 0.19] 0.11 [0.07, 0.17] 0.02 [0.00, 0.12] Pooled (95%CrIs): 0.055(0.0066-0.227)	0.69 [0.41, 0.89] 0.83 [0.74, 0.90] 0.92 [0.73, 0.99] Pooled (95%CrIs): 0.834(0.52-00.961)	Sensitivity
							Very serious^a	serious^b	none	none	VERY LOW
							Specificity
							Very serious^a	serious^b	none	serious^c	VERY LOW
Pelvic thrusting during seizure DETECTING RIGHT TLE (not included in above meta-analysis because the data already included in the overall epilepsy data)	1⁸²	261	neurologists	Critical care continuous EEG Non-epilepsy group: PNES / other epilepsy types	0.08 [0.02, 0.19]	0.85 [0.80, 0.90]	Sensitivity
							Very serious^a	serious^b	NA	none^c	VERY LOW
							Specificity
							Very serious^a	serious^b	NA	none^c	VERY LOW
Pelvic thrusting during seizure DETECTING LEFTTLE (not included in above meta-analysis because the data already included in the overall epilepsy data)	1⁸²	261	neurologists	Critical care continuous EEG Non-epilepsy group: PNES / other epilepsy types	0.04 [0.00, 0.14]	0.84 [0.79, 0.89]	Sensitivity
							Very serious^a	serious^b	NA	none^c	VERY LOW
							Specificity
							Very serious^a	serious^b	NA	none^c	VERY LOW
Pelvic thrusting DETECTING FLE (not included in above meta-analysis because the data already included in the overall epilepsy data)	1⁸²	261	neurologists	Critical care continuous EEG Non-epilepsy group: PNES / other epilepsy types	0.24 [0.13, 0.38]	0.89 [0.84, 0.93]	Sensitivity
							Very serious^a	serious^b	NA	none^c	VERY LOW
							Specificity
							Very serious^a	serious^b	NA	serious^c	VERY LOW
Sign observed by epileptologist on video during seizure – expression of pain	1³⁹	43	epileptologist	Surgical or long term follow up Non-epilepsy group: PNES	0.00 [0.00, 0.13]	0.75 [0.48, 0.93]	Sensitivity
							serious^a	serious^b	NA	none	LOW
							Specificity
							serious^a	serious^b	NA	Very serious^c	VERY LOW
Sign observed by epileptologist on video during seizure – motor behavioural onset	1³⁹	43	epileptologist	Surgical or long term follow up Non-epilepsy group: PNES	0.22 [0.09, 0.42]	0.81 [0.54, 0.96]	Sensitivity
							serious^a	serious^b	NA	none	LOW
							Specificity
							serious^a	serious^b	NA	Very serious^c	VERY LOW
Sign observed by epileptologist on video during seizure – head version	1³⁹	43	epileptologist	Surgical or long term follow up Non-epilepsy group: PNES	0.22 [0.09, 0.42]	0.94 [0.70, 1.00]	Sensitivity
							serious^a	serious^b	NA	none	LOW
							Specificity
							serious^a	serious^b	NA	Serious^c	VERY LOW
Sign observed by epileptologist on video during seizure – eye deviation	1³⁹	43	epileptologist	Surgical or long term follow up Non-epilepsy group: PNES	0.20 [0.07, 0.41]	1.00 [0.78, 1.00]	Sensitivity
							serious^a	serious^b	NA	none	LOW
							Specificity
							serious^a	serious^b	NA	Serious^c	VERY LOW
Sign observed by epileptologist on video during seizure – repetitive eye blinks	1³⁹	43	epileptologist	Surgical or long term follow up Non-epilepsy group: PNES	0.04 [0.00, 0.20]	0.80 [0.52, 0.96]	Sensitivity
							serious^a	serious^b	NA	none	LOW
							Specificity
							serious^a	serious^b	NA	Very serious^c	VERY LOW
Sign observed by epileptologist on video during seizure – facial grimacing	1³⁹	43	epileptologist	Surgical or long term follow up Non-epilepsy group: PNES	0.11 [0.02, 0.29]	0.88 [0.62, 0.98]	Sensitivity
							serious^a	serious^b	NA	none	LOW
							Specificity
							serious^a	serious^b	NA	Serious^c	VERY LOW
Sign observed by epileptologist on video during seizure – abnormal posturing	1³⁹	43	epileptologist	Surgical or long term follow up Non-epilepsy group: PNES	0.37 [0.19, 0.58]	0.63 [0.35, 0.85]	Sensitivity
							serious^a	serious^b	NA	none	LOW
							Specificity
							serious^a	serious^b	NA	Serious^c	VERY LOW
Sign observed by epileptologist on video during seizure – clonic activities	1³⁹	43	epileptologist	Surgical or long term follow up Non-epilepsy group: PNES	0.30 [0.14, 0.50]	0.81 [0.54, 0.96]	Sensitivity
							serious^a	serious^b	NA	none	LOW
							Specificity
							serious^a	serious^b	NA	Very serious^c	VERY LOW
Sign observed by epileptologist on video during seizure – vocalisation/speech	1³⁹	43	epileptologist	Surgical or long term follow up Non-epilepsy group: PNES	0.37 [0.19, 0.58]	0.69 [0.41, 0.89]	Sensitivity
							serious^a	serious^b	NA	none	LOW
							Specificity
							serious^a	serious^b	NA	Serious^c	VERY LOW
Sign observed by epileptologist on video during seizure – thrashing/writhing	1³⁹	43	epileptologist	Surgical or long term follow up Non-epilepsy group: PNES	0.15 [0.04, 0.34]	0.69 [0.41, 0.89]	Sensitivity
							serious^a	serious^b	NA	none	LOW
							Specificity
							serious^a	serious^b	NA	Serious^c	VERY LOW
Neurologist observation of video: eyes open during seizure	1¹⁶	68	neurologist	Video EEG Non-epilepsy group: PNES	1.00 [0.92, 1.00]	0.88 [0.68, 0.97]	Sensitivity
							Very serious^a	serious^b	NA	none	VERY LOW
							Specificity
							Very serious^a	serious^b	NA	serious^c	VERY LOW
Neurologist observation of video: Ictal vocalisation	1¹⁶	68	neurologist	Video EEG Non-epilepsy group: PNES	0.64 [0.48, 0.78]	0.88 [0.68, 0.97]	Sensitivity
							Very serious^a	serious^b	NA	serious^c	VERY LOW
							Specificity
							Very serious^a	serious^b	NA	serious^c	VERY LOW
Neurologist observation of video: Ictal side to side head and body turning	1¹⁶	68	neurologist	Video EEG Non-epilepsy group: PNES	0.39 [0.24, 0.55]	0.38 [0.19, 0.59]	Sensitivity
							Very serious^a	serious^b	NA	none	VERY LOW
							Specificity
							Very serious^a	serious^b	NA	none	VERY LOW
Neurologist observation of video: Ictal asynchronous extremity motion	1¹⁶	68	neurologist	Video EEG Non-epilepsy group: PNES	0.48 [0.32, 0.63]	0.04 [0.00, 0.21]	Sensitivity
							Very serious^a	serious^b	NA	serious^c	VERY LOW
							Specificity
							Very serious^a	serious^b	NA	none	VERY LOW
Neurologist observation of video: Post ictal breathing regularity	1¹⁶	68	neurologist	Video EEG Non-epilepsy group: PNES	0.50 [0.35, 0.65]	0.79 [0.58, 0.93]	Sensitivity
							Very serious^a	serious^b	NA	serious^c	VERY LOW
							Specificity
							Very serious^a	serious^b	NA	serious^c	VERY LOW
Neurologist observation of video: Post ictal agitation	1¹⁶	68	neurologist	Video EEG Non-epilepsy group: PNES	0.34 [0.20, 0.50]	0.88 [0.68, 0.97]	Sensitivity
							Very serious^a	serious^b	NA	none	VERY LOW
							Specificity
							Very serious^a	serious^b	NA	serious^c	VERY LOW
Neurologist observation of video: Post ictal confusion	1¹⁶	68	neurologist	Video EEG Non-epilepsy group: PNES	0.76 [0.56, 0.90]	0.88 [0.68, 0.97]	Sensitivity
							Very serious^a	serious^b	NA	serious^c	VERY LOW
							Specificity
							Very serious^a	serious^b	NA	serious^c	VERY LOW
Twitching arms or legs during seizure DETECTING ABSENCE SEIZURES IN INFANTS	1¹⁶³	40	Physician	Video EEG Non-epilepsy group: population suspected of epilepsy	0.24 [0.07, 0.50]	1.00 [0.85, 1.00]	Sensitivity
							serious^a	none	NA	None^c	MOD
							Specificity
							serious^a	none	NA	serious^c	LOW
Occurrence of seizure when tired DETECTING ABSENCE SEIZURES IN INFANTS	1¹⁶³	40	Physician	Video EEG Non-epilepsy group: population suspected of epilepsy	0.59 [0.33, 0.82]	0.74 [0.52, 0.90]	Sensitivity
							serious^a	none	NA	serious^c	LOW
							Specificity
							serious^a	none	NA	serious^c	LOW
Twitching arms or legs OR urine loss during seizure DETECTING ABSENCE SEIZURES IN INFANTS	1¹⁶³	40	Physician	Video EEG Non-epilepsy group: population suspected of epilepsy	0.35 [0.14, 0.62]	1.00 [0.85, 1.00]	Sensitivity
							serious^a	none	NA	serious^c	LOW
							Specificity
							serious^a	none	NA	serious^c	LOW
Upward eye movement during seizures and occurrence of seizures when tired DETECTING ABSENCE SEIZURES IN INFANTS	1¹⁶³	40	Physician	Video EEG Non-epilepsy group: population suspected of epilepsy	0.29 [0.10, 0.56]	0.96 [0.78, 1.00]	Sensitivity
							serious^a	none	NA	None^c	MOD
							Specificity
							serious^a	none	NA	serious^c	LOW
Eye witness (family/relative) account of eye opening or widening at onset during seizure	1¹⁹¹	36	epileptologist	Video EEG Non-epilepsy group: PNES	0.83 [0.61, 0.95]	0.25 [0.05, 0.57]	Sensitivity
							serious^a	serious^b	NA	serious^c	VERY LOW
							Specificity
							serious^a	serious^b	NA	None^c	LOW
Eye witness (family/relative) account of abrupt onset during seizure	1¹⁹¹	36	epileptologist	Video EEG Non-epilepsy group: PNES	0.48 [0.27, 0.69]	0.25 [0.05, 0.57]	Sensitivity
							serious^a	serious^b	NA	serious^c	VERY LOW
							Specificity
							serious^a	serious^b	NA	None^c	LOW
Eye witness (family/relative) account of post-ictal confusion/sleep	1¹⁹¹	36	epileptologist	Video EEG Non-epilepsy group: PNES	0.78 [0.56, 0.93]	0.00 [0.00, 0.26]	Sensitivity
							serious^a	serious^b	NA	serious^c	VERY LOW
							Specificity
							serious^a	serious^b	NA	None^c	LOW
Symptom questionnaire for patients – existence of headache after seizure?	1⁷⁴	39	NR	Video EEG Non-epilepsy group: PNES	0.38 [0.15, 0.65]	0.96 [0.78, 1.00]	Sensitivity
							Very serious^a	serious^b	NA	serious^c	VERY LOW
							Specificity
							Very serious^a	serious^b	NA	serious^c	VERY LOW
Symptom questionnaire for patients – existence of fatigue or lethargy?	1⁷⁴	39	NR	Video EEG Non-epilepsy group: PNES	0.56 [0.30, 0.80]	0.87 [0.66, 0.97]	Sensitivity
							Very serious^a	serious^b	NA	serious^c	VERY LOW
							Specificity
							Very serious^a	serious^b	NA	serious^c	VERY LOW
Symptom questionnaire for patients – existence of confusion alone?	1⁷⁴	39	NR	Video EEG Non-epilepsy group: PNES	0.13 [0.02, 0.38]	0.88 [0.69, 0.97]	Sensitivity
							Very serious^a	serious^b	NA	none	VERY LOW
							Specificity
							Very serious^a	serious^b	NA	serious^c	VERY LOW
Symptom questionnaire for patients – existence of no symptoms?	1⁷⁴	39	NR	Video EEG Non-epilepsy group: PNES	0.00 [0.00, 0.21]	0.52 [0.31, 0.72]	Sensitivity
							Very serious^a	serious^b	NA	none	VERY LOW
							Specificity
							Very serious^a	serious^b	NA	serious^c	VERY LOW
Reports of physical symptoms: light-headedness	1⁵⁶	69	NR	Video EEG Non-epilepsy group: PNES	0.10 [0.02, 0.27]	0.21 [0.09, 0.36]	Sensitivity
							serious^a	Serious^b	NA	none	LOW
							Specificity
							serious^a	Serious^b	NA	none	LOW
Reports of physical symptoms: sensory disturbances/dysaesthesias	1⁵⁶	69	NR	Video EEG Non-epilepsy group: PNES	0.17 [0.06, 0.35]	0.38 [0.23, 0.55]	Sensitivity
							serious^a	Serious^b	NA	none	LOW
							Specificity
							serious^a	Serious^b	NA	none	LOW
Reports of physical symptoms: hot flushes	1⁵⁶	69	NR	Video EEG Non-epilepsy group: PNES	0.00 [0.00, 0.12]	0.74 [0.58, 0.87]	Sensitivity
							serious^a	Serious^b	NA	none	LOW
							Specificity
							serious^a	Serious^b	NA	Serious^c	VERY LOW
Reports of physical symptoms: palpitations	1⁵⁶	69	NR	Video EEG Non-epilepsy group: PNES	0.03 [0.00, 0.17]	0.79 [0.64, 0.91]	Sensitivity
							serious^a	Serious^b	NA	none	LOW
							Specificity
							serious^a	Serious^b	NA	Serious^c	VERY LOW
Clinical signs of non-convulsive seizures (unexplained deterioration of consciousness, subtle motor activity, pupillary and ocular movement abnormalities) DETECTING NCSE	1¹¹⁴	NC	neurologists	Critical care continuous EEG Non-epilepsy group: population suspected of epilepsy	0.929^e	0.631^e	Sensitivity
							serious^a	none	NA	NA^c	MOD
							Specificity
							serious^a	none	NA	NA^c	LOW
Clinical signs of non-convulsive seizures (unexplained deterioration of consciousness, subtle motor activity, pupillary and ocular movement abnormalities) AND early sporadic epileptiform discharges OR Early rhythmic and periodic EEG patterns of ‘ictal-interictal uncertainty’ DETECTING NCSE	1¹¹⁴	NC	neurologists	Critical care continuous EEG Non-epilepsy group: population suspected of epilepsy	0.786^e	0.892^e	Sensitivity
							serious^a	none	NA	NA^c	MOD
							Specificity
							serious^a	none	NA	NA^c	LOW
Clinical signs of non-convulsive seizures (unexplained deterioration of consciousness, subtle motor activity, pupillary and ocular movement abnormalities) OR early sporadic epileptiform discharges OR Early rhythmic and periodic EEG patterns of ‘ictal-interictal uncertainty’ DETECTING NCSE	1¹¹⁴	NC	neurologists	Critical care continuous EEG Non-epilepsy group: population suspected of epilepsy	1.0^e	0.492^e	Sensitivity
							serious^a	none	NA	NA^c	MOD
							Specificity
							serious^a	none	NA	NA^c	LOW
Ictal duration >60s (measured by epileptologist using video)	1¹⁷⁷	782	epiletologist	Video EEG Non-epilepsy group: PNES	0.35 [0.30, 0.40]	0.29 [0.24, 0.34]	Sensitivity
							serious^a	serious^b	NA	none	LOW
							Specificity
							serious^a	serious^b	NA	none	LOW
Ictal duration >120s (measured by epileptologist using video)	1¹⁷⁷	782	epiletologist	Video EEG Non-epilepsy group: PNES	0.07 [0.05, 0.10]	0.48 [0.43, 0.54]	Sensitivity
							serious^a	serious^b	NA	none	LOW
							Specificity
							serious^a	serious^b	NA	none	LOW
Ictal duration >180s (measured by epileptologist using video)	1¹⁷⁷	782	epiletologist	Video EEG Non-epilepsy group: PNES	0.02 [0.01, 0.04]	0.63 [0.58, 0.68]	Sensitivity
							serious^a	serious^b	NA	none	LOW
							Specificity
							serious^a	serious^b	NA	serious^c	VERY LOW
Ictal duration >240s (measured by epileptologist using video)	1¹⁷⁷	782	epiletologist	Video EEG Non-epilepsy group: PNES	0.01 [0.01, 0.03]	0.71 [0.66, 0.75]	Sensitivity
							serious^a	serious^b	NA	none	LOW
							Specificity
							serious^a	serious^b	NA	none	LOW
Ictal duration >300s (measured by epileptologist using video)	1¹⁷⁷	782	epiletologist	Video EEG Non-epilepsy group: PNES	0.01 [0.00, 0.03]	0.79 [0.74, 0.83]	Sensitivity
							serious^a	serious^b	NA	none	LOW
							Specificity
							serious^a	serious^b	NA	none	LOW
Paroxysmal Event Profile Questionnaire – ‘factor scores’ (PNES as non-epilepsy group). No details of scoring or thresholds used.	1¹⁶⁰	200	NR	Video EEG Non-epilepsy group: PNES	0.72 [0.62, 0.81]	0.78 [0.69, 0.86]	Sensitivity
							Very serious^a	serious^b	NA	none	VERY LOW
							Specificity
							Very serious^a	serious^b	NA	none	VERY LOW
Paroxysmal Event Profile questionnaire – ‘patient information’ (PNES as non-epilepsy group). No details of scoring or thresholds used.	1¹⁶⁰	200	NR	Video EEG Non-epilepsy group: PNES	0.46 [0.36, 0.56]	0.74 [0.64, 0.82]	Sensitivity
							Very serious^a	serious^b	NA	none	VERY LOW
							Specificity
							Very serious^a	serious^b	NA	none	VERY LOW
Paroxysmal Event Profile questionnaire – ‘combined’(PNES as non-epilepsy group). No details of scoring or thresholds used.	1¹⁶⁰	200	NR	Video EEG Non-epilepsy group: PNES	0.74 [0.64, 0.82]	0.80 [0.71, 0.87]	Sensitivity
							Very serious^a	serious^b	NA	none	VERY LOW
							Specificity
							Very serious^a	serious^b	NA	none	VERY LOW
Paroxysmal Event Profile questionnaire – ‘factor scores’ (syncope as non-epilepsy group). No details of scoring or thresholds used.	1¹⁶⁰	200	NR	Video EEG Non-epilepsy group: syncope	0.83 [0.74, 0.90]	0.87 [0.79, 0.93]	Sensitivity
							Very serious^a	serious^b	NA	none	VERY LOW
							Specificity
							Very serious^a	serious^b	NA	serious^c	VERY LOW
Paroxysmal Event Profile questionnaire- ‘patient info’ (syncope as non-epilepsy group). No details of scoring or thresholds used.	1¹⁶⁰	200	NR	Video EEG Non-epilepsy group: syncope	0.68 [0.58, 0.77]	0.88 [0.80, 0.94]	Sensitivity
							Very serious^a	serious^b	NA	serious^c	VERY LOW
							Specificity
							Very serious^a	serious^b	NA	serious^c	VERY LOW
Paroxysmal Event Profile – ‘combined’ (syncope as non-epilepsy group). No details of scoring or thresholds used.	1¹⁶⁰	200	NR	Video EEG Non-epilepsy group: syncope	0.91 [0.84, 0.96]	0.92 [0.85, 0.96]	Sensitivity
							Very serious^a	serious^b	NA	serious^c	VERY LOW
							Specificity
							Very serious^a	serious^b	NA	serious^c	VERY LOW
>1 comorbidity on medical records	1⁶⁰	280	NR	Video EEG Non-epilepsy group: PNES	0.27 [0.19, 0.36]	0.34 [0.27, 0.42]	Sensitivity
							Very serious^a	serious^b	NA	none	VERY LOW
							Specificity
							Very serious^a	serious^b	NA	none	VERY LOW
Use of video information alone during seizure (from Video EEG) without other data to form ‘diagnosis’.	3³⁹^, ⁷³^, ⁹⁰	170	Epileptologist/neurologist	Surgery or long term observation / Video EEG Non-epilepsy group: PNES / suspected of epilepsy but no differential diagnoses	0.93 [0.76, 0.99] 0.75 [0.59, 0.87] 1.00 [0.48, 1.00] Pooled (95% CrIs): 0.892(0.534-0.996)	0.94 [0.70, 1.00] 0.95 [0.87, 0.99] 0.71 [0.29, 0.96] Pooled (95% CrIs): 0.917(0.603-0.987)	Sensitivity
							serious^a	none	none	serious^c	LOW
							Specificity
							serious^a	none	none	serious^c	LOW
Use of Clinical history / interview to form ‘diagnosis’	2¹⁴⁶ ⁹⁰	354	NR/neurologist	Medical record review / Video EEG Non-epilepsy group: healthy controls / suspected of epilepsy but no differential diagnoses	0.96 [0.92, 0.98] 0.80 [0.28, 0.99]	0.93 [0.88, 0.96] 0.86 [0.42, 1.00]	Sensitivity
							Very serious^a	serious^b	NA	serious^c	VERY LOW
							Specificity
							Very serious^a	serious^b	NA	serious^c	LOW
Use of history and physical examination only to form ‘diagnosis’	1¹⁹³	530	expert	Medical record review / Video EEG Non-epilepsy group: suspected of epilepsy but no differential diagnoses	1.00 [0.97, 1.00]	0.89 [0.85, 0.92]	Sensitivity
							serious^a	none	NA	none	MOD
							Specificity
							serious^a	none	NA	serious^c	LOW
Use of medical record only to form diagnosis INFANTS	1⁹⁶	NC	expert	Medical record review / Video EEG Non-epilepsy group: suspected of epilepsy but no differential diagnoses	0.849^e	0.399^e	Sensitivity
							serious^a	none	NA	NA	MOD
							Specificity
							serious^a	none	NA	NA	MOD
Use of medical record and 1 minute video of event to form ‘diagnosis’ INFANTS	1⁹⁶	NC	expert	Medical record review / Video EEG Non-epilepsy group: suspected of epilepsy but no differential diagnoses	0.888^e	0.514^e	Sensitivity
							serious^a	none	NA	NA	MOD
							Specificity
							serious^a	none	NA	NA	MOD
Use of smartphone video taken by witness to form ‘diagnosis’ (by experts and residents)	1¹⁹³	530	Experts and residents (ALL)	Medical record review / Video EEG Non-epilepsy group: suspected of epilepsy but no differential diagnoses	0.60 [0.51, 0.68]	0.91 [0.88, 0.94]	Sensitivity
							serious^a	none	NA	serious^c	LOW
							Specificity
							serious^a	none	NA	serious^c	LOW
Use of smartphone video taken by witness to form ‘diagnosis’ (by experts only)	1¹⁹³	530	Experts only	Medical record review / Video EEG Non-epilepsy group: suspected of epilepsy but no differential diagnoses	0.77 [0.69, 0.83]	0.93 [0.90, 0.96]	Sensitivity
							serious^a	none	NA	none	MOD
							Specificity
							serious^a	none	NA	none	MOD
Use of smartphone video taken by witness to form ‘diagnosis’ (by residents only)	1¹⁹³	NC	Residents only	Medical record review / Video EEG Non-epilepsy group: suspected of epilepsy but no differential diagnoses	0.42 [0.33, 0.50]	0.88 [0.85, 0.91]	Sensitivity
							serious^a	none	NA	none	MOD
							Specificity
							serious^a	none	NA	serious^c	LOW

(a): Risk of bias was assessed using the QUADAS-2 checklist. The evidence was downgraded by 1 increment if the majority of studies were rated at high risk of bias, and downgraded by 2 increments if the majority of studies were rated at very high risk of bias.
(b): Indirectness was assessed using the QUADAS-2 checklist items referring to applicability. The evidence was downgraded by 1 increment if the majority of studies were seriously indirect, and downgraded by 2 increments if the majority of studies are very seriously indirect
(c): Imprecision was assessed based on inspection of the confidence region in the diagnostic meta-analysis or, where diagnostic meta-analysis has not been conducted, assessed according to the range of confidence intervals in the individual studies. The evidence was downgraded by 1 increment when the confidence interval around the point estimate crossed one of the clinical thresholds (0.90 or 0.60), and downgraded by 2 increments when the confidence interval around the point estimate crossed both of the clinical thresholds. The upper clinical threshold marked the point above which recommendations would be possible, and the lower clinical threshold marked the point below which the tool would be regarded as of little clinical use.
(d): Inconsistency was assessed by visual inspection of the sensitivity/specificity plots, or data (if 2 studies). The evidence was downgraded by 1 increment if there was no overlap of 95% confidence intervals. For single studies no evaluation was made and ‘not applicable’ was recorded.
(e): No confidence intervals were presented because there was insufficient information available or there was a mismatch between the raw data and the accuracy results.

Table 4Clinical evidence summary: diagnostic test accuracy of different serum measurements for detection of epilepsy

Where detection is of a specific type of epilepsy, rather than epilepsy overall, this is stated clearly in the first column.

Index Test	Number of studies	n	Interpreter of index test	Gold standard used in study	Sensitivity (95% CI)	Specificity (95% CI)	Risk of bias	Indirectness	Inconsistency	Imprecision	GRADE
serum prolactin level at threshold >29.9 mg/dl (indicating epilepsy). This was measured in the ED for patients presenting with recent seizure	1²⁰⁹	200	NR	Discharge diagnosis. Non-epilepsy group: range of people without epilepsy initially suspected of epilepsy (not restricted to one differential diagnosis)	0.42 [0.33, 0.52]	0.82 [0.73, 0.90]	Sensitivity
							Serious^a	None^b	NA	None^c	MOD
							Specificity
							Serious^a	None^b	NA	Serious^c	LOW
Paired serum prolactin >1025 microU/ml (indicating epilepsy) in immediate post-seizure period	1⁷	58	NR	Video EEG Non-epilepsy group: PNES	0.34 [0.20, 0.51]	1.00 [0.83, 1.00]	Sensitivity
							Serious^a	Serious^b	NA	None^c	LOW
							Specificity
							Serious^a	Serious^b	NA	Serious^c	VERY LOW
Paired serum prolactin RI > 5.5 in post seizure period (5.5 × increase in serum prolactin between 15 mins post-seizure and 2 hours after baseline sample)	1⁷	58	NR	Video EEG Non-epilepsy group: PNES	0.21 [0.10, 0.37]	1.00 [0.83, 1.00]	Sensitivity
							Serious^a	Serious^b	NA	None^c	LOW
							Specificity
							Serious^a	Serious^b	NA	Serious^c	VERY LOW
Paired serum prolactin RI > 2 in post seizure period (2 × increase in serum prolactin between 15 mins post-seizure and 2 hours after baseline sample)	1⁷	58	NR	Video EEG Non-epilepsy group: PNES	0.68 [0.51, 0.82]	0.75 [0.51, 0.91]	Sensitivity
							Serious^a	Serious^b	NA	Serious^c	VERY LOW
							Specificity
							Serious^a	Serious^b	NA	Very serious^c	VERY LOW
Paired serum prolactin >1025 microU/ml (indicating epilepsy) in immediate post-seizure period DETECTING COMPLEX PARTIAL SEIZURES	1⁷	40	NR	Video EEG Non-PC epilepsy group: PNES	0.35 [0.15, 0.59]	1.00 [0.83, 1.00]	Sensitivity
							Serious^a	Serious^b	NA	None^c	LOW
							Specificity
							Serious^a	Serious^b	NA	Serious^c	VERY LOW
Paired serum prolactin RI > 5.5 in post seizure period (5.5 × increase in serum prolactin between 15 mins post-seizure and 2 hours after baseline sample) DETECTING COMPLEX PARTIAL SEIZURES	1⁷	40	NR	Video EEG Non-PC epilepsy group: PNES	0.28^e	1^e	Sensitivity
							Serious^a	Serious^b	NA	NA^c	LOW
							Specificity
							Serious^a	Serious^b	NA	NA^c	LOW
Paired serum prolactin RI > 2 in post seizure period (2 × increase in serum prolactin between 15 mins post-seizure and 2 hours after baseline sample) DETECTING PARTIAL COMPLEX SEIZURES	1⁷	40	NR	Video EEG Non-PC epilepsy group: PNES	0.61^e	0.74^e	Sensitivity
							Serious^a	Serious^b	NA	NA^c	LOW
							Specificity
							Serious^a	Serious^b	NA	NA	LOW
Paired serum prolactin >1025 microU/ml (indicating epilepsy) in immediate post-seizure period DETECTING GENERALISED CLOINIC TONIC SEIZURES	1⁷	36	NR	Video EEG Non-GCS epilepsy group: PNES	0.38 [0.15, 0.65]	1.00 [0.83, 1.00]	Sensitivity
							Serious^a	Serious^b	NA	Serious^c	VERY LOW
							Specificity
							Serious^a	Serious^b	NA	Serious^c	VERY LOW
Paired serum prolactin RI > 5.5 in post seizure period (5.5 × increase in serum prolactin between 15 mins post-seizure and 2 hours after baseline sample) DETECTING GENERALISED CLOINIC TONIC SEIZURES	1⁷	36	NR	Video EEG Non-GCS epilepsy group: PNES	0.2	1	Sensitivity
							Serious^a	Serious^b	NA	NA^c	LOW
							Specificity
							Serious^a	Serious^b	NA	NA^c	LOW
Paired serum prolactin RI > 2 in post seizure period (2 × increase in serum prolactin between 15 mins post-seizure and 2 hours after baseline sample) DETECTING GENERALISED CLONIC TONIC SEIZURES	1⁷	36	NR	Video EEG Non-GCS epilepsy group: PNES	0.94 [0.70, 1.00]	0.75 [0.51, 0.91]	Sensitivity
							Serious^a	Serious^b	NA	Serious^c	VERY LOW
							Specificity
							Serious^a	Serious^b	NA	Very serious^c	VERY LOW
Capillary prolactin level above 6.7 ng/ml at 15 minutes post-seizure	1⁶⁹	50	Nursing staff	Video EEG Non-epilepsy group: PNES	0.69 [0.52, 0.84]	0.93 [0.66, 1.00]	Sensitivity
							Serious^a	None^b	NA	Serious^c	LOW
							Specificity
							Serious^a	None^b	NA	Serious^c	LOW
2 fold decrease in capillary prolactin level, between 15 min sample and sample obtained 1 hr later	1⁶⁹	50	Nursing staff	Video EEG Non-epilepsy group: PNES	0.69 [0.52, 0.84]	0.86 [0.57, 0.98]	Sensitivity
							Serious^a	None^b	NA	Serious^c	LOW
							Specificity
							Serious^a	None^b	NA	Very serious^c	VERY LOW
15 min cap prolactin level above 6.7 ng/ml AND a 2 fold decrease between 15 mins and 1 hour post-seizure	1⁶⁹	50	Nursing staff	Video EEG Non-epilepsy group: PNES	0.56 [0.38, 0.72]	1.00 [0.77, 1.00]	Sensitivity
							Serious^a	None^b	NA	Serious^c	LOW
							Specificity
							Serious^a	None^b	NA	Serious^c	LOW
Serum prolactin >23 microg [women]/>16.5 [men] at 10mins post seizure	1²¹⁶	44	NR	Video EEG Non-epilepsy group: PNES	0.88 [0.71, 0.96]	0.58 [0.28, 0.85]	Sensitivity
							Serious^a	Serious^b	NA	Serious^c	VERY LOW
							Specificity
							Serious^a	Serious^b	NA	Serious^c	VERY LOW
serum prolactin >23 microg [women]/>16.5 [men] at 20mins post seizure	1²¹⁶	44	NR	Video EEG Non-epilepsy group: PNES	0.88 [0.71, 0.96]	0.67 [0.35, 0.90]	Sensitivity
							Serious^a	Serious^b	NA	Serious^c	VERY LOW
							Specificity
							Serious^a	Serious^b	NA	Serious^c	VERY LOW
Serum prolactin >23 microg [women]/>16.5 [men] at 30mins post seizure	1²¹⁶	44	NR	Video EEG Non-epilepsy group: PNES	0.84 [0.67, 0.95]	0.75 [0.43, 0.95]	Sensitivity
							Serious^a	Serious^b	NA	Serious^c	VERY LOW
							Specificity
							Serious^a	Serious^b	NA	Very serious^c	VERY LOW
Serum prolactin >23 microg [women]/>16.5 [men] at 60mins post seizure	1²¹⁶	44	NR	Video EEG Non-epilepsy group: PNES	0.63 [0.44, 0.79]	0.92 [0.62, 1.00]	Sensitivity
							Serious^a	Serious^b	NA	Serious^c	VERY LOW
							Specificity
							Serious^a	Serious^b	NA	Serious^c	VERY LOW
Serum prolactin >23 microg [women]/>16.5 [men] at 6 hours post seizure	1²¹⁶	44	NR	Video EEG Non-epilepsy group: PNES	0.22 [0.09, 0.40]	0.83 [0.52, 0.98]	Sensitivity
							Serious^a	Serious^b	NA	none^c	LOW
							Specificity
							Serious^a	Serious^b	NA	Very serious^c	VERY LOW
Serum prolactin >23 microg [women]/>16.5 [men] at 12 hours post seizure	1²¹⁶	44	NR	Video EEG Non-epilepsy group: PNES	0.19 [0.07, 0.36]	0.83 [0.52, 0.98]	Sensitivity
							Serious^a	Serious^b	NA	none^c	LOW
							Specificity
							Serious^a	Serious^b	NA	Very serious^c	VERY LOW
Serum prolactin >23 microg [women]/>16.5 [men] at 24 hours post seizure	1²¹⁶	44	NR	Video EEG Non-epilepsy group: PNES	0.13 [0.04, 0.29]	0.92 [0.62, 1.00]	Sensitivity
							Serious^a	Serious^b	NA	None^c	LOW
							Specificity
							Serious^a	Serious^b	NA	Serious^c	VERY LOW
Serum neuron-specific enolase >12 microg/L at 10 minutes post seizure	1²¹⁶	44	NR	Video EEG Non-epilepsy group: PNES	0.06 [0.01, 0.21]	1.00 [0.74, 1.00]	Sensitivity
							Serious^a	Serious^b	NA	None^c	LOW
							Specificity
							Serious^a	Serious^b	NA	Serious^c	VERY LOW
Serum neuron-specific enolase >12 microg/L at 20 minutes post seizure	1²¹⁶	44	NR	Video EEG Non-epilepsy group: PNES	0.06 [0.01, 0.21]	1.00 [0.74, 1.00]	Sensitivity
							Serious^a	None^b	NA	None^c	MOD
							Specificity
							Serious^a	None^b	NA	Serious^c	LOW
Serum neuron-specific enolase >12 microg/L at 30 minutes post seizure	1²¹⁶	44	NR	Video EEG Non-epilepsy group: PNES	0.06 [0.01, 0.21]	1.00 [0.74, 1.00]	Sensitivity
							Serious^a	None^b	NA	None^c	MOD
							Specificity
							Serious^a	None^b	NA	Serious^c	LOW
Serum neuron-specific enolase >12 microg/L at 60 minutes post seizure	1²¹⁶	44	NR	Video EEG Non-epilepsy group: PNES	0.03 [0.00, 0.16]	1.00 [0.74, 1.00]	Sensitivity
							Serious^a	None^b	NA	None^c	MOD
							Specificity
							Serious^a	None^b	NA	Serious^c	LOW
Serum neuron-specific enolase >12 microg/L at 6 hours post seizure	1²¹⁶	44	NR	Video EEG Non-epilepsy group: PNES	0.13 [0.04, 0.29]	1.00 [0.74, 1.00]	Sensitivity
							Serious^a	None^b	NA	None^c	MOD
							Specificity
							Serious^a	None^b	NA	Serious^c	LOW
Serum neuron-specific enolase >12 microg/L at 12 hours post seizure	1²¹⁶	44	NR	Video EEG Non-epilepsy group: PNES	0.09 [0.02, 0.25]	1.00 [0.74, 1.00]	Sensitivity
							Serious^a	None^b	NA	None^c	MOD
							Specificity
							Serious^a	None^b	NA	Serious^c	LOW
Serum neuron-specific enolase >12 microg/L at 24 hours post seizure	1²¹⁶	44	NR	Video EEG Non-epilepsy group: PNES	0.00 [0.00, 0.11]	1.00 [0.74, 1.00]	Sensitivity
							Serious^a	None^b	NA	None^c	MOD
							Specificity
							Serious^a	None^b	NA	Serious^c	LOW
Serum creatine kinase >2.8 [women]/>3.25 [men] at 10 minutes post seizure	1²¹⁶	44	NR	Video EEG Non-epilepsy group: PNES	0.00 [0.00, 0.11]	1.00 [0.74, 1.00]	Sensitivity
							Serious^a	None^b	NA	None^c	MOD
							Specificity
							Serious^a	None^b	NA	Serious^c	LOW
Serum creatine kinase >2.8 [women]/>3.25 [men] at 20 minutes post seizure	1²¹⁶	44	NR	Video EEG Non-epilepsy group: PNES	0.00 [0.00, 0.11]	1.00 [0.74, 1.00]	Sensitivity
							Serious^a	None^b	NA	None^c	MOD
							Specificity
							Serious^a	None^b	NA	Serious^c	LOW
Serum creatine kinase >2.8 [women]/>3.25 [men] at 30 minutes post seizure	1²¹⁶	44	NR	Video EEG Non-epilepsy group: PNES	0.00 [0.00, 0.11]	1.00 [0.74, 1.00]	Sensitivity
							Serious^a	None^b	NA	None^c	MOD
							Specificity
							Serious^a	None^b	NA	Serious^c	LOW
Serum creatine kinase >2.8 [women]/>3.25 [men] at 60 minutes post seizure	1²¹⁶	44	NR	Video EEG Non-epilepsy group: PNES	0.00 [0.00, 0.11]	1.00 [0.74, 1.00]	Sensitivity
							Serious^a	None^b	NA	None^c	MOD
							Specificity
							Serious^a	None^b	NA	Serious^c	LOW
Serum creatine kinase >2.8 [women]/>3.25 [men] at 6 hours post seizure	1²¹⁶	44	NR	Video EEG Non-epilepsy group: PNES	0.09 [0.02, 0.25]	1.00 [0.74, 1.00]	Sensitivity
							Serious^a	None^b	NA	None^c	MOD
							Specificity
							Serious^a	None^b	NA	Serious^c	LOW
Serum creatine kinase >2.8 [women]/>3.25 [men] at 12 hours post seizure	1²¹⁶	44	NR	Video EEG Non-epilepsy group: PNES	0.16 [0.05, 0.33]	1.00 [0.74, 1.00]	Sensitivity
							Serious^a	None^b	NA	None^c	MOD
							Specificity
							Serious^a	None^b	NA	Serious^c	LOW
Serum creatine kinase >2.8 [women]/>3.25 [men] at 24 hours post seizure	1²¹⁶	44	NR	Video EEG Non-epilepsy group: PNES	0.19 [0.07, 0.36]	1.00 [0.74, 1.00]	Sensitivity
							Serious^a	None^b	NA	None^c	MOD
							Specificity
							serious^a	Serious^b	NA	Serious^c	LOW
Anion gap in first 2 hrs after seizure event (threshold at >10 mEq/L)	1¹²⁵	54	NR	Video EEG Non-epilepsy group: PNES	0.81 [0.62, 0.94]	1.00 [0.87, 1.00]	Sensitivity
							Serious^a	None^b	NA	Serious^c	LOW
							Specificity
							Serious^a	None^b	NA	Serious^c	LOW
serum lactate 2 hrs post ictal (threshold >=2.2 mmol/L)	1⁶¹	270	NR	Final definitive diagnosis with CT/MRI, EEG and ECG data with observable clinical signs and symptoms Non-epilepsy group: PNES and syncope	0.85 [0.78, 0.90]	0.82 [0.74, 0.89]	Sensitivity
							Very serious^a	None^b	NA	None^c	MOD
							Specificity
							Very serious^a	None^b	NA	None^c	LOW
Post-seizure (within 6 hours) serum glial fibrillary astrocytic protein levels at threshold of >=2.71 ng/ml	1¹⁸⁰	63	NR	Video EEG Non-epilepsy group: PNES	0.72 [0.56, 0.85]	0.60 [0.36, 0.81]	Sensitivity
							Serious^a	None^b	NA	serious^c	LOW
							Specificity
							Serious^a	None^b	NA	serious^c	LOW
baseline serum ammonia at cut-off of >=80 micromol/L DETECTING GENERALISED CLONIC TONIC SEIZURES	1⁶	26	NR	Video EEG Non-GCS epilepsy group: people initially suspected of epilepsy but with no definite differential diagnoses	0.53 [0.28, 0.77]	1.00 [0.66, 1.00]	Sensitivity
							Very serious^a	none^b	NA	Serious^c	VERY LOW
							Specificity
							Very serious^a	none^b	NA	Serious^c	VERY LOW

(a): Risk of bias was assessed using the QUADAS-2 checklist. The evidence was downgraded by 1 increment if the majority of studies were rated at high risk of bias, and downgraded by 2 increments if the majority of studies were rated at very high risk of bias.
(b): Indirectness was assessed using the QUADAS-2 checklist items referring to applicability. The evidence was downgraded by 1 increment if the majority of studies were seriously indirect, and downgraded by 2 increments if the majority of studies are very seriously indirect
(c): Imprecision was assessed based on inspection of the confidence region in the diagnostic meta-analysis or, where diagnostic meta-analysis has not been conducted, assessed according to the range of confidence intervals in the individual studies. The evidence was downgraded by 1 increment when the confidence interval around the point estimate crossed one of the clinical thresholds (0.90 or 0.60), and downgraded by 2 increments when the confidence interval around the point estimate crossed both of the clinical thresholds. The upper clinical threshold marked the point above which recommendations would be possible, and the lower clinical threshold marked the point below which the tool would be regarded as of little clinical use.
(d): Inconsistency was assessed by visual inspection of the sensitivity/specificity plots, or data (if 2 studies). The evidence was downgraded by 1 increment if there was no overlap of 95% confidence intervals. For single studies no evaluation was made and ‘not applicable’ was recorded.
(e): No confidence intervals were presented because there was insufficient information available or there was a mismatch between the raw data and the accuracy results.

Table 5Clinical evidence summary: diagnostic test accuracy of ECG tests for detection of epilepsy

Where detection is of a specific type of epilepsy, rather than epilepsy overall, this is stated clearly in the first column.

Index Test	Number of studies	n	Interpreter of index test	Gold standard used in study	Sensitivity (95% CI)	Specificity (95% CI)	Risk of bias	Indirectness	Inconsistency	Imprecision	GRADE
ECG. Interictal. No details of measures or thresholds used.	1⁴³	142	NR	EEG plus clinical findings, over prolonged follow up. Non-epilepsy group: range of people without epilepsy initially suspected of epilepsy (not restricted to one differential diagnosis)	0.14 [0.02, 0.43]	0.73 [0.65, 0.81]	Sensitivity
							serious^a	none^b	NA	none	MOD
							Specificity
							serious^a	none^b	NA	none	MOD

(a): Risk of bias was assessed using the QUADAS-2 checklist. The evidence was downgraded by 1 increment if the majority of studies were rated at high risk of bias, and downgraded by 2 increments if the majority of studies were rated at very high risk of bias.
(b): Indirectness was assessed using the QUADAS-2 checklist items referring to applicability. The evidence was downgraded by 1 increment if the majority of studies were seriously indirect, and downgraded by 2 increments if the majority of studies are very seriously indirect
(c): Imprecision was assessed based on inspection of the confidence region in the diagnostic meta-analysis or, where diagnostic meta-analysis has not been conducted, assessed according to the range of confidence intervals in the individual studies. The evidence was downgraded by 1 increment when the confidence interval around the point estimate crossed one of the clinical thresholds (0.90 or 0.60), and downgraded by 2 increments when the confidence interval around the point estimate crossed both of the clinical thresholds. The upper clinical threshold marked the point above which recommendations would be possible, and the lower clinical threshold marked the point below which the tool would be regarded as of little clinical use.
(d): Inconsistency was assessed by visual inspection of the sensitivity/specificity plots, or data (if 2 studies). The evidence was downgraded by 1 increment if there was no overlap of 95% confidence intervals. For single studies no evaluation was made and ‘not applicable’ was recorded.
(e): No confidence intervals were presented because there was insufficient information available or there was a mismatch between the raw data and the accuracy results.

Table 6Clinical evidence summary: diagnostic test accuracy of different imaging tests for detection of epilepsy

Where detection is of a specific type of epilepsy, rather than epilepsy overall, this is stated clearly in the first column.

Index Test	Number of studies	n	Interpreter of index test	Gold standard used in study	Sensitivity (95% CI)	Specificity (95% CI)	Risk of bias	Indirectness	Inconsistency	Imprecision	GRADE
Echocardiogram. Interictal. No details of measures or threshold available.	1⁴³	63	NR	EEG plus clinical findings, over prolonged follow up Non-epilepsy group: those initially suspected of epilepsy but with no differential diagnoses	0.00 [0.00, 0.46]	0.96 [0.88, 1.00]	Sensitivity
							Serious^a	None	NA	none	MOD
							Specificity
							Serious^a	None	NA	serious^c	LOW
Brain CT. Interictal. No details of measures or threshold available.	1⁴³	33	NR	EEG plus clinical findings, over prolonged follow up Non-epilepsy group: those initially suspected of epilepsy but with no differential diagnoses	0.20 [0.01, 0.72]	0.79 [0.59, 0.92]	Sensitivity
							Serious^a	None	NA	serious^c	LOW
							Specificity
							Serious^a	None	NA	Very serious^c	VERY LOW
Single photon emission computed tomography (SPECT) - post-ictal abnormal measure	1⁷⁵	22	nuclear medicine specialists	Video-EEG Non-epilepsy group: PNES	0.64 [0.31, 0.89]	0.73 [0.39, 0.94]	Sensitivity
							None	Serious^b	NA	serious^c	LOW
							Specificity
							None	Serious^b	NA	Very serious^c	VERY LOW
Single photon emission computed tomography (SPECT) - inter-ictal abnormal measure	1⁷⁵	22	nuclear medicine specialists	Video-EEG Non-epilepsy group: PNES	0.36 [0.11, 0.69]	0.73 [0.39, 0.94]	Sensitivity
							None	Serious^b	NA	serious^c	LOW
							Specificity
							None	Serious^b	NA	Very serious^c	VERY LOW
Hexamethyl propylene amine oxime single photon emission tomography (HMPAO SPECT) brain imaging. Interictal. (positive=hypoperfusion not including equivocal hypoperfusion)	1²⁰³	20	nuclear medicine specialists	Video-EEG Non-epilepsy group: PNES	0.80 [0.44, 0.97]	0.80 [0.44, 0.97]	Sensitivity
							Very serious^a	Serious^b	NA	Very serious^c	VERY LOW
							Specificity
							Very serious^a	Serious^b	NA	Very serious^c	VERY LOW
Hexamethyl propylene amine oxime single photon emission tomography (HMPAO SPECT) brain imaging. Interictal. (positive=hypoperfusion including equivocal hypoperfusion)	1²⁰³	20	nuclear medicine specialists	Video-EEG Non-epilepsy group: PNES	1.00 [0.69, 1.00]	0.70 [0.35, 0.93]	Sensitivity
							Very serious^a	Serious^b	NA	serious^c	VERY LOW
							Specificity
							Very serious^a	Serious^b	NA	Very serious^c	VERY LOW
HMPAO-SPECT using visual analysis: SPECTS considered positive for status Epilepticus when there was at least one area of Focal Uptake compared to the adjacent or contralateral areas of the brain. ICTAL DETECTING NCSE	1¹⁰⁰	55	3 experts in nuclear medicine	consensus based on all data, inc EEG Non-epilepsy group: those initially suspected of epilepsy but with no differential diagnoses	0.81 [0.64, 0.92]	0.89 [0.67, 0.99]	Sensitivity
							none	none	NA	serious^c	MOD
							Specificity
							none	none	NA	serious^c	MOD
HMPAO-SPECT - QtSPECTCOM using quantitative analysis: Results were compared to a normal database and the difference in terms of the Z score was quantified. ICTAL DETECTING NCSE	1¹⁰⁰	55	3 experts in nuclear medicine	consensus based on all data, inc EEG Non-epilepsy group: those initially suspected of epilepsy but with no differential diagnoses	0.83 [0.67, 0.94]	0.79 [0.54, 0.94]	Sensitivity
							none	none	NA	serious^c	MOD
							Specificity
							none	none	NA	Very serious^c	LOW
Perfusion computed tomography using hyperperfusion detection. ICTAL. DETECTING STATUS EPILEPTICUS	1⁸⁶	29	Experienced neuroradiologist	Ictal EEG and clinical semiology Non-epilepsy group: those initially suspected of epilepsy but with no differential diagnoses	0.79 [0.54, 0.94]	0.90 [0.55, 1.00]	Sensitivity
							serious^a	None	NA	Very serious^c	VERY LOW
							Specificity
							serious^a	None	NA	Very serious^c	VERY LOW
Brain MRI. Interictal. No details of measures or threshold available.	1⁴³	13	NR	EEG plus clinical findings, over prolonged follow up Non-epilepsy group: those initially suspected of epilepsy but with no differential diagnoses	0.20 [0.01, 0.72]	0.88 [0.47, 1.00]	Sensitivity
							Serious^a	None	NA	serious^c	LOW
							Specificity
							Serious^a	None	NA	Very serious^c	VERY LOW
MRI (IN CHILDREN). No details of measures or threshold available.	1¹⁹⁴	NC	NR	49 month follow up Non-epilepsy group: those initially suspected of epilepsy but with no differential diagnoses	0.36^e	0.74^e	Sensitivity
							Serious^a	None	NA	serious^c	LOW
							Specificity
							Serious^a	None	NA	Very serious^c	VERY LOW
4T MRI: the presence/absence of MTS in TLE was based on hippocampal subfield volumetry. Appears to be interictal. DETECTING TLE with MTS	1¹³⁶	80	NR	Video EEG Non-epilepsy group: healthy controls and other types of epilepsy	0.84 [0.60, 0.97]	0.87 [0.76, 0.94]	Sensitivity
							Very serious^a	serious^b	NA	serious^c	VERY LOW
							Specificity
							Very serious^a	serious^b	NA	serious^c	VERY LOW
4T MRI: the presence/absence of MTS in TLE was based on hippocampal subfield volumetry. Appears to be interictal. DETECTING TLE without MTS	1¹³⁶	80	NR	Video EEG Non-epilepsy group: healthy controls and other types of epilepsy	0.73 [0.50, 0.89]	0.86 [0.75, 0.94]	Sensitivity
							Very serious^a	serious^b	NA	serious^c	VERY LOW
							Specificity
							Very serious^a	serious^b	NA	serious^c	VERY LOW
4T MRI. Appears to be interictal. DETECTING FLE	1¹³⁶	80	NR	Video EEG Non-epilepsy group: healthy controls and other types of epilepsy	0.64 [0.35, 0.87]	0.86 [0.76, 0.94]	Sensitivity
							Very serious^a	serious^b	NA	serious^c	VERY LOW
							Specificity
							Very serious^a	serious^b	NA	serious^c	VERY LOW
Positron Emission Tomography with 2-deoxy-2[18F] fluro-D-glucose (FDG-PET). Interictal. DETECTING TLE	1¹⁸⁶	NC	board certified neuroradiologists	Video EEG Non-epilepsy group: those initially suspected of epilepsy but with no differential diagnoses	0.7^e	0.56^e	Sensitivity
							Very serious^a	None	NA	NA	LOW
							Specificity
							Very serious^a	None	NA	NA	LOW
Positron Emission Tomography with 2-deoxy-2[18F] fluro-D-glucose (FDG-PET). Interictal. DETECTING FLE	1¹⁸⁶	NC	board certified neuroradiologists	Video EEG Non-epilepsy group: those initially suspected of epilepsy but with no differential diagnoses	0.57^e	0.45^e	Sensitivity
							Very serious^a	None	NA	NA	LOW
							Specificity
							Very serious^a	None	NA	NA	LOW
Positron Emission Tomography with 2-deoxy-2[18F] fluro-D-glucose (FDG-PET). Interictal. DETECTING parietal – occipital lobe epilepsy	1¹⁸⁶	NC	board certified neuroradiologists	Video EEG Non-epilepsy group: those initially suspected of epilepsy but with no differential diagnoses	0.59^e	0.6^e	Sensitivity
							Very serious^a	None	NA	NA	LOW
							Specificity
							Very serious^a	None	NA	NA	LOW

(a): Risk of bias was assessed using the QUADAS-2 checklist. The evidence was downgraded by 1 increment if the majority of studies were rated at high risk of bias, and downgraded by 2 increments if the majority of studies were rated at very high risk of bias.
(b): Indirectness was assessed using the QUADAS-2 checklist items referring to applicability. The evidence was downgraded by 1 increment if the majority of studies were seriously indirect, and downgraded by 2 increments if the majority of studies are very seriously indirect
(c): Imprecision was assessed based on inspection of the confidence region in the diagnostic meta-analysis or, where diagnostic meta-analysis has not been conducted, assessed according to the range of confidence intervals in the individual studies. The evidence was downgraded by 1 increment when the confidence interval around the point estimate crossed one of the clinical thresholds (0.90 or 0.60), and downgraded by 2 increments when the confidence interval around the point estimate crossed both of the clinical thresholds. The upper clinical threshold marked the point above which recommendations would be possible, and the lower clinical threshold marked the point below which the tool would be regarded as of little clinical use.
(d): Inconsistency was assessed by visual inspection of the sensitivity/specificity plots, or data (if 2 studies). The evidence was downgraded by 1 increment if there was no overlap of 95% confidence intervals. For single studies no evaluation was made and ‘not applicable’ was recorded.
(e): No confidence intervals were presented because there was insufficient information available or there was a mismatch between the raw data and the accuracy results.

Table 7Clinical evidence summary: diagnostic test accuracy of EEG methods for detection of epilepsy

Where detection is of a specific type of epilepsy, rather than epilepsy overall, this is stated clearly in the first column.

Index Test	Number of studies	n	Interpreter of index test	Gold standard used in study	Sensitivity (95% CI)	Specificity (95% CI)	Risk of bias	Indirectness	Inconsistency	Imprecision	GRADE
Routine Interictal EEG – abnormal (i.e. epileptiform waveforms) [Most studies detecting epilepsy overall, but van diessen²⁰⁰ detecting partial epilepsy specifically, and Kimiskidis¹⁰⁹ detecting genetic generalised epilepsy]	9⁴³^, ⁹⁴^, ¹⁰⁹^, ¹¹¹^, ¹⁷⁹^, ¹⁸⁴^, ¹⁹⁴^, ²⁰⁰^, ²¹³ Stroink¹⁸⁴ has 2 cohorts (single and multiple seizures) and Watson, 2012²¹³ has 3 cohorts (ages 16-39, 40-64 and 65 or over). Thus, there are 12 datapoints from 9 studies	2348	Neurophysiologist, epileptologists, clinical physiologists and pediatric neurologists	Detailed clinical findings over prolonged follow up period Non-epilepsy group: Population suspected of epilepsy but with no known differential diagnoses [however, for Kimiskidis (2017) non-epilepsy group were healthy controls]	0.40 [0.26, 0.56] 0.80 [0.52, 0.96] 0.24 [0.09, 0.45] 0.33 [0.24, 0.44] 0.40 [0.31, 0.50] 0.40 [0.30, 0.50] 0.60 [0.47, 0.72] 0.40 [0.28, 0.52] 0.55 [0.43, 0.66] 0.70 [0.66, 0.75] 0.56 [0.48, 0.63] 0.77 [0.60, 0.90] Pooled (95% CrI): 0.508(0.393-0.625)	0.95 [0.87, 0.98] 0.80 [0.59, 0.93] 1.00 [0.72, 1.00] 0.87 [0.82, 0.91] 0.95 [0.90, 0.99] 0.99 [0.96, 1.00] 0.88 [0.74, 0.96] 0.99 [0.96, 1.00] 0.77 [0.70, 0.83] 0.77 [0.69, 0.84] 0.78 [0.64, 0.88] 0.91 [0.77, 0.98] Pooled (95% CrI): 0.920(0.846-0.966)	Sensitivity
							serious^a	serious^b	serious^d	serious^c	VERY LOW
							Specificity
							serious^a	serious^b	serious^d	serious^c	VERY LOW
Sleep-deprived interictal EEG – abnormal (i.e. epileptiform waveforms)	3⁸¹^, ⁸⁴^, ¹⁵⁹	499	Resident/consultant in neurology	Collegial discussion of detailed clinical findings over prolonged follow up period Non-epilepsy group: Population suspected of epilepsy but with no known differential diagnoses [for Kimiskidis (2017): healthy controls]	0.25 [0.15, 0.36] 0.45 [0.27, 0.64] 0.41 [0.33, 0.50] Pooled (95% CrI): 0.362(0123-0.699)	0.99 [0.97, 1.00] 0.90 [0.70, 0.99] 0.91 [0.83, 0.96] Pooled (95% CrI): 0.962(0.697-0.997)	Sensitivity
							serious^a	none	none	none	MOD
							Specificity
							serious^a	none	none	serious^c	LOW
24 hour sleep deprivation interictal EEG–abnormal (i.e. epileptiform waveforms) DETECTING FOCAL EPILEPSY Not included in meta-analysis above as same participants already included in Renzel (2015) ‘overall epilepsy’ cohort	1¹⁵⁹	226	Interpreted by resident and consultant in neurology and clinical neurophysiology	Collegial discussion following ILAE guidelines, and EEG evidence Non-epilepsy group: Population suspected of epilepsy but with no known differential diagnoses	0.17 [0.09, 0.29]	0.99 [0.97, 1.00]	Sensitivity
							serious^a	none	NA	none^c	MOD
							Specificity
							serious^a	none	NA	none^c	MOD
24 hour sleep deprivation interictal EEG–abnormal (i.e. epileptiform waveforms) DETECTING GENERALISED EPILEPSY Not included in meta-analysis above as same participants already included in Renzel (2015) ‘overall epilepsy’ cohort	1¹⁵⁹	179	Interpreted by resident and consultant in neurology and clinical neurophysiology	Collegial discussion following ILAE guidelines, and EEG evidence Non-epilepsy group: Population suspected of epilepsy but with no known differential diagnoses	0.64 [0.31, 0.89]	0.99 [0.97, 1.00]	Sensitivity
							serious^a	none	NA	serious^c	LOW
							Specificity
							serious^a	none	NA	none^c	MOD
Ambulatory interictal EEG (16-24 hrs, including sleep) – abnormal (i.e. epileptiform waveforms)	1⁸¹	52	Resident/consultant in neurology	Clinical record surveyed for clinical, imaging and diagnosis at 1 year data (ILAE) Non-epilepsy group: Population suspected of epilepsy but with no known differential diagnoses [for Kimiskidis (2017): healthy controls]	0.63 [0.44, 0.79]	0.95 [0.75, 1.00]	Sensitivity
							serious^a	none	NA	serious^c	LOW
							Specificity
							serious^a	none	NA	serious^c	LOW
Prolonged ambulatory interictal EEG using epileptiform discharges only as definition of a positive test	1¹⁰²	72	Electroenceph alographers	Summation of retrospective medical records and expert opinion Non-epilepsy group: Population suspected of epilepsy but with no known differential diagnoses	0.58 [0.43, 0.72]	0.95 [0.77, 1.00]	Sensitivity
							none	none	NA	serious^c	MOD
							Specificity
							none	none	NA	serious^c	MOD
Prolonged ambulatory interictal EEG using either epileptiform discharges or non-epileptiform abnormalities as definitions of a positive test	1¹⁰²	72	Electroenceph alographers	Summation of retrospective medical records and expert opinion Non-epilepsy group: Population suspected of epilepsy but with no known differential diagnoses	0.78 [0.64, 0.88]	0.59 [0.36, 0.79]	Sensitivity
							none	none	NA	none	HIGH
							Specificity
							none	none	NA	serious^c	MOD
Routine interictal EEG with provocation with hyperventilation, intermittent phototic stimulation and eye opening/closing, using epileptiform discharges as definition of positive test	1¹⁰²	72	Electroenceph alographers	Summation of retrospective medical records and expert opinion Non-epilepsy group: Population suspected of epilepsy but with no known differential diagnoses	0.26 [0.15, 0.40]	1.00 [0.85, 1.00]	Sensitivity
							none	none	NA	none	HIGH
							Specificity
							none	none	NA	serious^c	MOD
Routine interictal EEG with provocation with hyperventilation, intermittent phototic stimulation and eye opening/closing, using either epileptiform or non-epileptiform abnormalities as definitions of a positive test	1¹⁰²	72	Electroenceph alographers	Summation of retrospective medical records and expert opinion Non-epilepsy group: Population suspected of epilepsy but with no known differential diagnoses	0.62 [0.47, 0.75]	0.55 [0.32, 0.76]	Sensitivity
							none	none	NA	serious^c	MOD
							Specificity
							none	none	NA	serious^c	MOD
Early sporadic epileptiform discharges (first 30 minutes of the EEG recordings) DETECTING NCSE	1¹¹⁴	NC	neurophysiology experts	Critical care continuous EEG Non-epilepsy group: Population suspected of epilepsy but with no known differential diagnoses	0.214^e	0.908^e	Sensitivity
							serious^a	none	NA	NA^c	MOD
							Specificity
							serious^a	none	NA	NA^c	MOD
Computational biomarker looking at the synchrony between EEG channels and the normalised power spectrum from a short resting state interictal EEG (does not require epileptiform discharges). Details of the threshold of synchrony not given.	1¹⁷¹	68	Trained clinical EEG technician	EEG monitoring Non-epilepsy group: Healthy controls	0.57 [0.37, 0.75] The above data is based on the fact that at 100% specificity we have 56.7% sensitivity The paper also reports (based on the ROC curves) that at 100% sensitivity, 65.8% specificity is attainable	1.00 [0.91, 1.00]	Sensitivity
							Very serious^a	serious^b	NA	serious^c	VERY LOW
							Specificity
							Very serious^a	serious^b	NA	none	VERY LOW
Synchronisation likelihood (SL) based on standard EEG after a first seizure. The Theta band SL values were tested for accuracy, but details or specific threshold not given	1⁶²	161	NR	Medical chart review with a 1 year follow up (ILAE) Non-epilepsy group: unclear	0.61 [0.48, 0.74]	0.76 [0.67, 0.84]	Sensitivity
							Very serious^a	serious^b	NA	serious^c	VERY LOW
							Specificity
							Very serious^a	serious^b	NA	none	VERY LOW
Interictal fast ripple (250-500Hz) events, based on scalp EEG. Single 10-minute epoch per patient. Existence of fast ripples = positive test. (INFANTS WITH TUBEROUS SCLEROSIS COMPLEX-ASSOCIATED EPILEPSY)	1²⁸	11	Trained clinicians	Video EEG Non-epilepsy group: healthy controls	1.00 [0.59, 1.00]	1.00 [0.40, 1.00]	Sensitivity
							serious^a	Serious^b	NA	Very serious^c	VERY LOW
							Specificity
							serious^a	Very serious^b	NA	Very serious^c	VERY LOW
Functional network approach. Periods of resting-state EEG, free of abnormal slowing or epileptiform activity, were selected to construct functional networks of correlated activity. The statistical interdependencies for each pair of EEG electrode time series are considered as functional connectivity and used to construct a functional network per subject for each of the four epochs and were averaged per subject. Details of thresholds not provided DETECTING PARTIAL EPILEPSY	1²⁰⁰	70	Clinical epileptologist	EEG/clinical and 1 year follow up Non-epilepsy group: healthy controls]	0.96 [0.78–1.00]	0.95 [0.76–1.00]	Sensitivity
							Very serious^a	serious^b	NA	serious^c	VERY LOW
							Specificity
							Very serious^a	serious^b	NA	serious^c	VERY LOW
Early rhythmic and periodic EEG patterns of ictal-interictal uncertainty (RPPIIU) DETECTING NCSE	1¹¹⁴	NC	neurophysiology experts	Critical care continuous EEG Non-epilepsy group: Population suspected of epilepsy but with no known differential diagnoses	0.643^e	0.846^e	Sensitivity
							serious^a	none	NA	NA^c	MOD
							Specificity
							serious^a	none	NA	NA^c	MOD
Early sporadic epileptiform discharges OR Early rhythmic and periodic EEG patterns of ‘ictal-interictal uncertainty’ DETECTING NCSE	1¹¹⁴	NC	neurophysiology experts	Critical care continuous EEG Non-epilepsy group: Population suspected of epilepsy but with no known differential diagnoses	0.857^e	0.754^e	Sensitivity
							serious^a	none	NA	NA^c	MOD
							Specificity
							serious^a	none	NA	NA^c	MOD
Resting state 10-15 min high density EEG. The cortical source activity was obtained and whole-brain directed functional connectivity was estimated in the theta, alpha and beta frequency bands. No threshold information available DETECTING TEMPORAL LOBE EPILEPSY	1²⁰⁵	75	NR	EEG/clinical Non-epilepsy group: healthy controls]	0.95 [0.83, 0.99]	0.86 [0.70, 0.95]	Sensitivity
							Very serious^a	serious^b	NA	serious^c	VERY LOW
							Specificity
							Very serious^a	serious^b	NA	serious^c	VERY LOW
Routine EEG using Salzburg criteria. ICTAL for Jaraba¹⁰⁰ but unclear for other two studies DETECTING NCSE	3⁸⁷^, ¹⁰⁰^, ¹²⁴ Note there are 2 cohorts from Goselink, 2019⁸⁷ – patients suspected of NCSE and patients not suspected of NCSE	366	Nuclear medicine or neurophysiology experts	All data including clinical, EEG, imaging, lab tests etc Non-epilepsy group: Population suspected of epilepsy but with no known differential diagnoses	0.98 [0.88, 1.00] 0.61 [0.43, 0.77] 0.67 [0.35, 0.90] 1.00 [0.03, 1.00] Pooled (95%CrIs): 0.838(0.430-0.986)	0.90 [0.81, 0.95] 0.89 [0.67, 0.99] 0.89 [0.81, 0.95] 0.89 [0.81, 0.95] Pooled (95%CrIs): 0.899(0.782-0.959)	Sensitivity
							serious^a	none	none	serious^c	LOW
							Specificity
							serious^a	none	none	serious^c	LOW
Ictal EEG (without access to video or observation) – abnormal (i.e. epileptiform waveforms)	1³⁹	43	fellowship trained epileptologist	Surgical or by long term follow up Non-epilepsy group: PNES	0.89 [0.71, 0.98]	0.94 [0.70, 1.00]	Sensitivity
							serious^a	serious^b	NA	serious^c	VERY LOW
							Specificity
							serious^a	serious^b	NA	serious^c	VERY LOW
Quantitative ICTAL EEG interpreted by PICU clinicians in real time – abnormal waveforms (INFANTS)	1¹⁶⁶	101	PICU clinicians	Clinical neurophysiologist retrospective review qEEG Non-epilepsy group: Population suspected of epilepsy but with no known differential diagnoses	1.00 [0.74, 1.00]	0.88 [0.79, 0.94]	Sensitivity
							serious^a	none	NA	serious^c	LOW
							Specificity
							serious^a	none	NA	serious^c	LOW
Headset-type continuous video EEG monitoring – detection of abnormal patterns, such as periodic discharges, rhythmic delta activity, spikes and wave and continuous slow discharges DETECTING NCSE	1⁶⁸	50	1 neurointensivist and one board certified neurophysiologist	Video EEG Non-epilepsy group: Population suspected of epilepsy but with no known differential diagnoses	0.71 [0.44, 0.90]	0.97 [0.84, 1.00]	Sensitivity
							serious^a	none	NA	Very serious^c	VERY LOW
							Specificity
							serious^a	none	NA	serious^c	LOW
No event video EEG (at least 16 hours)	1¹¹¹	340	NR	Full definitive diagnosis based on full medical records and a minimum of 1 clinic visit in 1 year of follow up Non-epilepsy group: Population suspected of epilepsy but with no known differential diagnoses	0.54 [0.44, 0.64]	0.88 [0.83, 0.92]	Sensitivity
							Very serious^a	serious^b	NA	serious^c	VERY LOW
							Specificity
							Very serious^a	serious^b	NA	serious^c	VERY LOW

(a): Risk of bias was assessed using the QUADAS-2 checklist. The evidence was downgraded by 1 increment if the majority of studies were rated at high risk of bias, and downgraded by 2 increments if the majority of studies were rated at very high risk of bias.
(b): Indirectness was assessed using the QUADAS-2 checklist items referring to applicability. The evidence was downgraded by 1 increment if the majority of studies were seriously indirect, and downgraded by 2 increments if the majority of studies are very seriously indirect
(c): Imprecision was assessed based on inspection of the confidence region in the diagnostic meta-analysis or, where diagnostic meta-analysis has not been conducted, assessed according to the range of confidence intervals in the individual studies. The evidence was downgraded by 1 increment when the confidence interval around the point estimate crossed one of the clinical thresholds (0.90 or 0.60), and downgraded by 2 increments when the confidence interval around the point estimate crossed both of the clinical thresholds. The upper clinical threshold marked the point above which recommendations would be possible, and the lower clinical threshold marked the point below which the tool would be regarded as of little clinical use.
(d): Inconsistency was assessed by visual inspection of the sensitivity/specificity plots, or data (if 2 studies). The evidence was downgraded by 1 increment if there was no overlap of 95% confidence intervals. If a meta-analysis had been carried out, sub-grouping was carried out when I² was >50%, according to the strategies listed in the protocol. However, in no circumstance did sub-grouping explain the heterogeneity observed, and so sub-grouping was not carried out. For single studies no evaluation was made and ‘not applicable’ was recorded.
(e): No confidence intervals were presented because there was insufficient information available or there was a mismatch between the raw data and the accuracy results.

Table 8Clinical evidence summary: diagnostic test accuracy of different Magnetoencephalography / Transcranial Magnetic Stimulation tests for detection of epilepsy

Where detection is of a specific type of epilepsy, rather than epilepsy overall, this is stated clearly in the first column.

Index Test	Number of studies	n	Interpreter of index test	Gold standard used in study	Sensitivity (95% CI)	Specificity (95% CI)	Risk of bias	Indirectness	Inconsistency	Imprecision	GRADE
Magnetoencephalography with simultaneous EEG (MEG-EEG). Interictal. No details of threshold available.	1⁶⁵	52	Trained physicians	1 year follow up, including all data Non-epilepsy group: those initially suspected of epilepsy but with no differential diagnoses	0.41 [0.21, 0.64]	0.93 [0.78, 0.99]	Sensitivity
							Serious^a	Serious^b	NA	serious^c	VERY LOW
							Specificity
							Serious^a	Serious^b	NA	serious^c	VERY LOW
Paired pulse Transcranial Magnetic Stimulation with EEG (TMS-EEG) immediately after hyperventilation. Interictal. No details of threshold available.	1¹⁰⁹	36	NR	consensus by 2 experienced epileptologists who reached consensus based on clinical and lab data Non-epilepsy group: healthy controls	1.00 [0.86, 1.00]	0.73 [0.39, 0.94]	Sensitivity
							Serious^a	Serious^b	NA	serious^c	VERY LOW
							Specificity
							Serious^a	Serious^b	NA	Very serious^c	VERY LOW
Paired pulse TMS-EEG during hyperventilation. Interictal. No details of threshold available.	1¹⁰⁹	36	NR	consensus by 2 experienced epileptologists who reached consensus based on clinical and lab data Non-epilepsy group: healthy controls	0.78^e	0.89^e	Sensitivity
							Serious^a	Serious^b	NA	NA	LOW
							Specificity
							Serious^a	Serious^b	NA	NA	LOW
Paired pulse TMS-EEG at rest. Interictal. No details of threshold available.	1¹⁰⁹	36	NR	consensus by 2 experienced epileptologists who reached consensus based on clinical and lab data Non-epilepsy group: healthy controls	0.85^e	0.89^e	Sensitivity
							Serious^a	Serious^b	NA	NA	LOW
							Specificity
							Serious^a	Serious^b	NA	NA	LOW
Single pulse TMS-EEG at rest. Interictal. No details of threshold available.	1¹⁰⁹	36	NR	consensus by 2 experienced epileptologists who reached consensus based on clinical and lab data 7Non-epilepsy group: healthy controls	0.6^e	0.82^e	Sensitivity
							Serious^a	Serious^b	NA	NA	LOW
							Specificity
							Serious^a	Serious^b	NA	NA	LOW

(a): Risk of bias was assessed using the QUADAS-2 checklist. The evidence was downgraded by 1 increment if the majority of studies were rated at high risk of bias, and downgraded by 2 increments if the majority of studies were rated at very high risk of bias.
(b): Indirectness was assessed using the QUADAS-2 checklist items referring to applicability. The evidence was downgraded by 1 increment if the majority of studies were seriously indirect, and downgraded by 2 increments if the majority of studies are very seriously indirect
(c): Imprecision was assessed based on inspection of the confidence region in the diagnostic meta-analysis or, where diagnostic meta-analysis has not been conducted, assessed according to the range of confidence intervals in the individual studies. The evidence was downgraded by 1 increment when the confidence interval around the point estimate crossed one of the clinical thresholds (0.90 or 0.60), and downgraded by 2 increments when the confidence interval around the point estimate crossed both of the clinical thresholds. The upper clinical threshold marked the point above which recommendations would be possible, and the lower clinical threshold marked the point below which the tool would be regarded as of little clinical use.
(d): Inconsistency was assessed by visual inspection of the sensitivity/specificity plots, or data (if 2 studies). The evidence was downgraded by 1 increment if there was no overlap of 95% confidence intervals. For single studies no evaluation was made and ‘not applicable’ was recorded.
(e): No confidence intervals were presented because there was insufficient information available or there was a mismatch between the raw data and the accuracy results.

Table 9Clinical evidence summary: diagnostic test accuracy of different psychological measurements for detection of epilepsy

Where detection is of a specific type of epilepsy, rather than epilepsy overall, this is stated clearly in the first column.

Index Test	Number of studies	n	Interpreter of index test	Gold standard used in study	Sensitivity (95% CI)	Specificity (95% CI)	Risk of bias	Indirectness	Inconsistency	Imprecision	GRADE
Personality Assessment scale: Psychogenic nonepileptic seizures (PNES) scale; threshold <1	1¹⁹⁶	184	NR	Video EEG Non-epilepsy group: PNES	0.85 [0.77, 0.91]	0.59 [0.47, 0.70]	Sensitivity
							Very serious^a	Serious^b	NA	Serious^c	VERY LOW
							Specificity
							Very serious^a	Serious^b	NA	Serious^c	VERY LOW
Personality Assessment scale: SOM-C (conversion) scale; threshold <70	1¹⁹⁶	184	NR	Video EEG Non-epilepsy group: PNES	0.83 [0.75, 0.90]	0.59 [0.47, 0.70]	Sensitivity
							Very serious^a	Serious^b	NA	None^c	VERY LOW
							Specificity
							Very serious^a	Serious^b	NA	Serious^c	VERY LOW
Personality Assessment scale: SOM (somatic complaints); threshold <70	1¹⁹⁶	184	NR	Video EEG Non-epilepsy group: PNES	0.73 [0.64, 0.81]	0.56 [0.44, 0.67]	Sensitivity
							Very serious^a	Serious^b	NA	None^c	VERY LOW
							Specificity
							Very serious^a	Serious^b	NA	Serious^c	VERY LOW
Personality Assessment scale: SOM-S (somatisation); threshold <70	1¹⁹⁶	184	NR	Video EEG Non-epilepsy group: PNES	0.82 [0.73, 0.88]	0.45 [0.34, 0.57]	Sensitivity
							Very serious^a	Serious^b	NA	none^c	VERY LOW
							Specificity
							Very serious^a	Serious^b	NA	none	VERY LOW
Personality Assessment scale: DEP-P (Depression-physiological); threshold <70	1¹⁹⁶	184	NR	Video EEG Non-epilepsy group: PNES	0.86 [0.78, 0.92]	0.49 [0.38, 0.61]	Sensitivity
							Very serious^a	Serious^b	NA	Serious^c	VERY LOW
							Specificity
							Very serious^a	Serious^b	NA	Serious^c	VERY LOW
Personality Assessment scale: DEP-P (Depression); threshold <60	1¹⁹⁶	184	NR	Video EEG Non-epilepsy group: PNES	0.61 [0.52, 0.71]	0.63 [0.51, 0.74]	Sensitivity
							Very serious^a	Serious^b	NA	Serious^c	VERY LOW
							Specificity
							Very serious^a	Serious^b	NA	Serious^c	VERY LOW
Personality Assessment scale: ANX-P (Anxiety-Physiological); threshold <60	1¹⁹⁶	184	NR	Video EEG Non-epilepsy group: PNES	0.68 [0.58, 0.77]	0.57 [0.45, 0.69]	Sensitivity
							Very serious^a	Serious^b	NA	Serious^c	VERY LOW
							Specificity
							Very serious^a	Serious^b	NA	serious^c	VERY LOW
Wilkus measure of hysteria and hypochondriasis: A patients has pseudoseizures if any of the following are true: a) hysteria or hypochondriasis score >=70 and one of the two highest points in the profile (disregarding the masculinity-femininity and social introversion scales, b) hysteria or hypochondriasis score >=80 and not necessarily among the two highest points, c) hysteria and hypochondriasis both >59 and both 10 points higher than the depression scale. In a sample where ONLY epilepsy and PNES patients are known to exist then this test could be used to show that epilepsy exists if NONE of these conditions exists.	2¹⁸¹^, ²¹⁵	69	Trained psychometrists	Video EEG Non-epilepsy group: PNES	0.74 [0.54, 0.89] 0.80 [0.44, 0.97]	0.59 [0.36, 0.79] 0.90 [0.55, 1.00]	Sensitivity
							Very serious^a	Serious^b	NA	Serious^c	VERY LOW
							Specificity
							Very serious^a	Serious^b	NA	Very serious^c	VERY LOW
Structured Interview of malingered Symptomatology questionnaire; threshold <14	1²⁶	120	NR	Video EEG Non-epilepsy group: PNES	0.55 [0.36, 0.74]	0.76 [0.66, 0.84]	Sensitivity
							Serious^a	Serious^b	NA	Serious^c	VERY LOW
							Specificity
							Serious^a	Serious^b	NA	none^c	VERY LOW
Structured Interview of malingered Symptomatology questionnaire; threshold <16	1²⁶	120	NR	Video EEG Non-epilepsy group: PNES	0.69 [0.49, 0.85]	0.71 [0.61, 0.80]	Sensitivity
							Serious^a	Serious^b	NA	Serious^c	VERY LOW
							Specificity
							Serious^a	Serious^b	NA	none^c	VERY LOW
multivariate model of psychometric testing using 4 measures of cognitive ability – vocabulary, information, Boston naming test and letter fluency (unclear description in article)	1¹⁹⁹	105	Masters level psychometrist, predocintern or postdoc fellow	Video EEG Non-epilepsy group: PNES	0.92 [0.83, 0.97]	0.45 [0.28, 0.64]	Sensitivity
							Serious^a	Serious^b	NA	Serious^c	VERY LOW
							Specificity
							Serious^a	Serious^b	NA	serious^c	VERY LOW
Number of panic attack symptoms <5	1⁹²	354	NR	Video EEG Non-epilepsy group: PNES	0.65 [0.57, 0.74]	0.70 [0.64, 0.76]	Sensitivity
							Very serious^a	Serious^b	NA	Serious^c	VERY LOW
							Specificity
							Serious^a	Serious^b	NA	Serious^c	VERY LOW
lifetime axis 1 (no details or score threshold available)	1¹⁰	41	Trained psychiatrist	Video EEG Non-epilepsy group: PNES	0.52 [0.32, 0.71]	0.29 [0.08, 0.58]	Sensitivity
							Serious^a	Serious^b	NA	serious^c	VERY LOW
							Specificity
							Serious^a	Serious^b	NA	none	LOW
Current axis 1 (no details or score threshold available)	1¹⁰	41	Trained psychiatrist	Video EEG Non-epilepsy group: PNES	0.30 [0.14, 0.50]	0.57 [0.29, 0.82]	Sensitivity
							Serious^a	Serious^b	NA	none^c	LOW
							Specificity
							Serious^a	Serious^b	NA	serious^c	VERY LOW
Current axis II (no details or score threshold available)	1¹⁰	41	Trained psychiatrist	Video EEG Non-epilepsy group: PNES	0.19 [0.06, 0.38]	0.64 [0.35, 0.87]	Sensitivity
							Serious^a	Serious^b	NA	None^c	LOW
							Specificity
							Serious^a	Serious^b	NA	Serious^c	VERY LOW
Any psychological trauma (yes/No). Criteria not given.	1¹⁰	41	Trained psychiatrist	Video EEG Non-epilepsy group: PNES	0.33 [0.17, 0.54]	0.14 [0.02, 0.43]	Sensitivity
							Serious^a	Serious^b	NA	None^c	LOW
							Specificity
							Serious^a	Serious^b	NA	none^c	VERY LOW

(a): Risk of bias was assessed using the QUADAS-2 checklist. The evidence was downgraded by 1 increment if the majority of studies were rated at high risk of bias, and downgraded by 2 increments if the majority of studies were rated at very high risk of bias.
(b): Indirectness was assessed using the QUADAS-2 checklist items referring to applicability. The evidence was downgraded by 1 increment if the majority of studies were seriously indirect, and downgraded by 2 increments if the majority of studies are very seriously indirect
(c): Imprecision was assessed based on inspection of the confidence region in the diagnostic meta-analysis or, where diagnostic meta-analysis has not been conducted, assessed according to the range of confidence intervals in the individual studies. The evidence was downgraded by 1 increment when the confidence interval around the point estimate crossed one of the clinical thresholds (0.90 or 0.60), and downgraded by 2 increments when the confidence interval around the point estimate crossed both of the clinical thresholds. The upper clinical threshold marked the point above which recommendations would be possible, and the lower clinical threshold marked the point below which the tool would be regarded as of little clinical use.
(d): Inconsistency was assessed by visual inspection of the sensitivity/specificity plots, or data (if 2 studies). The evidence was downgraded by 1 increment if there was no overlap of 95% confidence intervals. For single studies no evaluation was made and ‘not applicable’ was recorded.
(e): No confidence intervals were presented because there was insufficient information available or there was a mismatch between the raw data and the accuracy results.

Table 10Clinical evidence summary: diagnostic test accuracy of different linguistic tests for detection of epilepsy

Where detection is of a specific type of epilepsy, rather than epilepsy overall, this is stated clearly in the first column.

Index Test	Number of studies	n	Interpreter of index test	Gold standard used in study	Sensitivity (95% CI)	Specificity (95% CI)	Risk of bias	Indirectness	Inconsistency	Imprecision	GRADE
Linguistic analysis following guidelines from the German EpiLing project (rater 1) – threshold of >4.5 Unclear if the accuracy data refer to detection of epilepsy or PNES	1¹⁶¹	20	Neurologist 1	Video EEG. Non-epilepsy group: PNES	0.86 [0.42, 1.00]	0.85 [0.55, 0.98]	Sensitivity
							Serious^a	Serious^b	NA	Very serious^c	VERY LOW
							Specificity
							Serious^a	Serious^b	NA	Very serious^c	VERY LOW
Linguistic analysis following guidelines from the German EpiLing project (rater 2) with threshold of >7.5 Unclear if the accuracy data refer to detection of epilepsy or PNES	1¹⁶¹	20	Neurologist 2	Video EEG. Non-epilepsy group: PNES	0.71 [0.29, 0.96]	0.92 [0.64, 1.00]	Sensitivity
							Serious^a	Serious^b	NA	Very serious^c	VERY LOW
							Specificity
							Serious^a	Serious^b	NA	Serious^c	VERY LOW

(a): Risk of bias was assessed using the QUADAS-2 checklist. The evidence was downgraded by 1 increment if the majority of studies were rated at high risk of bias, and downgraded by 2 increments if the majority of studies were rated at very high risk of bias.
(b): Indirectness was assessed using the QUADAS-2 checklist items referring to applicability. The evidence was downgraded by 1 increment if the majority of studies were seriously indirect, and downgraded by 2 increments if the majority of studies are very seriously indirect
(c): Imprecision was assessed based on inspection of the confidence region in the diagnostic meta-analysis or, where diagnostic meta-analysis has not been conducted, assessed according to the range of confidence intervals in the individual studies. The evidence was downgraded by 1 increment when the confidence interval around the point estimate crossed one of the clinical thresholds (0.90 or 0.60), and downgraded by 2 increments when the confidence interval around the point estimate crossed both of the clinical thresholds. The upper clinical threshold marked the point above which recommendations would be possible, and the lower clinical threshold marked the point below which the tool would be regarded as of little clinical use.
(d): Inconsistency was assessed by visual inspection of the sensitivity/specificity plots, or data (if 2 studies). The evidence was downgraded by 1 increment if there was no overlap of 95% confidence intervals. For single studies no evaluation was made and ‘not applicable’ was recorded.
(e): No confidence intervals were presented because there was insufficient information available or there was a mismatch between the raw data and the accuracy results.

Table 11Clinical evidence summary: diagnostic test accuracy of EMG tests for detection of epilepsy

Where detection is of a specific type of epilepsy, rather than epilepsy overall, this is stated clearly in the first column.

Index Test	Number of studies	n	Interpreter of index test	Gold standard used in study	Sensitivity (95% CI)	Specificity (95% CI)	Risk of bias	Indirectness	Inconsistency	Imprecision	GRADE
Single channel surface EMG (on biceps muscle belly). ICTAL. Decision based on expert review, but criteria unclear.	1⁹⁷	34	Board certified neurologists	Video EEG. Non-epilepsy group: PNES	0.77(0.64-0.86)^e	0.96(0.89-0.99)^e	Sensitivity
							Serious^a	Serious^b	NA	none^c	LOW
							Specificity
							Serious^a	Serious^b	NA	serious^c	VERY LOW
Single channel surface EMG (on biecps muscle belly). ICTAL. Decision based on automated criteria (score between 0-25 with a score of 8 or above = epilepsy).	1⁹⁷	20	Automated	Video EEG. Non-epilepsy group: PNES	0.87 [0.60, 0.98]	0.79 [0.54, 0.94]	Sensitivity
							Serious^a	Serious^b	NA	serious^c	VERY LOW
							Specificity
							Serious^a	Serious^b	NA	Very serious^c	VERY LOW

(a): Risk of bias was assessed using the QUADAS-2 checklist. The evidence was downgraded by 1 increment if the majority of studies were rated at high risk of bias, and downgraded by 2 increments if the majority of studies were rated at very high risk of bias.
(b): Indirectness was assessed using the QUADAS-2 checklist items referring to applicability. The evidence was downgraded by 1 increment if the majority of studies were seriously indirect, and downgraded by 2 increments if the majority of studies are very seriously indirect
(c): Imprecision was assessed based on inspection of the confidence region in the diagnostic meta-analysis or, where diagnostic meta-analysis has not been conducted, assessed according to the range of confidence intervals in the individual studies. The evidence was downgraded by 1 increment when the confidence interval around the point estimate crossed one of the clinical thresholds (0.90 or 0.60), and downgraded by 2 increments when the confidence interval around the point estimate crossed both of the clinical thresholds. The upper clinical threshold marked the point above which recommendations would be possible, and the lower clinical threshold marked the point below which the tool would be regarded as of little clinical use.
(d): Inconsistency was assessed by visual inspection of the sensitivity/specificity plots, or data (if 2 studies). The evidence was downgraded by 1 increment if there was no overlap of 95% confidence intervals. For single studies no evaluation was made and ‘not applicable’ was recorded.
(e): No confidence intervals were presented because there was insufficient information available or there was a mismatch between the raw data and the accuracy results.

Table 12Clinical evidence summary: diagnostic test accuracy of accelerometer tests for detection of epilepsy

Where detection is of a specific type of epilepsy, rather than epilepsy overall, this is stated clearly in the first column.

Index Test	Number of studies	n	Interpreter of index test	Gold standard used in study	Sensitivity (95% CI)	Specificity (95% CI)	Risk of bias	Indirectness	Inconsistency	Imprecision	GRADE
Wrist accelerometer. ICTAL. (Bayly, 2013 used visual review of time-frequency maps by epileptologist, but criteria unclear. Kusmakar, 2018 used review of the Poincare-derived temporal variations by epileptologists but again criteria unclear)	2²⁰^, ¹¹⁶	124	epileptologists	Clinical consensus /Video EEG. Non-epilepsy group: PNES	0.75 [0.35, 0.97] 0.87 [0.72, 0.96]	0.93 [0.80, 0.98] 0.70 [0.53, 0.84]	Sensitivity
							none^a	Serious^b	NA	Very serious^c	LOW
							Specificity
							none^a	Serious^b	NA	serious^c	LOW
Wrist accelerometer. ICTAL. (automated). Bayly, 2013 used the co-efficient of variation of the frequency of movements, using a threshold of 32% [<32% = PNES and >=32% = epilepsy]). Kusmakar, 2018 used an automated classifier built using TI and DDI of Poincare-derived temporal variations, but thresholds not provided. Naganur, 2018 used K-means clustering and support vector machines, but details not available.	3²⁰^,¹³⁷^,¹¹⁶	163	Automated	Clinical consensus/Video EEG. Non-epilepsy group: PNES	0.91 [0.59, 1.00] 0.73 [0.39, 0.94] 0.95 [0.83, 0.99] Pooled (95% CrIs): 0.895(0.558-0.986)	0.93 [0.82, 0.99] 1.00 [0.75, 1.00] 0.95 [0.85, 0.99] Pooled (95% CrIs): 0.955(0.805-0.994)	Sensitivity
							Serious^a	Serious^b	none	serious^c	VERY LOW
							Specificity
							Serious^a	Serious^b	none	serious^c	VERY LOW

(a): Risk of bias was assessed using the QUADAS-2 checklist. The evidence was downgraded by 1 increment if the majority of studies were rated at high risk of bias, and downgraded by 2 increments if the majority of studies were rated at very high risk of bias.
(b): Indirectness was assessed using the QUADAS-2 checklist items referring to applicability. The evidence was downgraded by 1 increment if the majority of studies were seriously indirect, and downgraded by 2 increments if the majority of studies are very seriously indirect
(c): Imprecision was assessed based on inspection of the confidence region in the diagnostic meta-analysis or, where diagnostic meta-analysis has not been conducted, assessed according to the range of confidence intervals in the individual studies. The evidence was downgraded by 1 increment when the confidence interval around the point estimate crossed one of the clinical thresholds (0.90 or 0.60), and downgraded by 2 increments when the confidence interval around the point estimate crossed both of the clinical thresholds. The upper clinical threshold marked the point above which recommendations would be possible, and the lower clinical threshold marked the point below which the tool would be regarded as of little clinical use.
(d): Inconsistency was assessed by visual inspection of the sensitivity/specificity plots, or data (if 2 studies). The evidence was downgraded by 1 increment if there was no overlap of 95% confidence intervals. For single studies no evaluation was made and ‘not applicable’ was recorded.
(e): No confidence intervals were presented because there was insufficient information available or there was a mismatch between the raw data and the accuracy results

Table 13Clinical evidence summary: diagnostic test accuracy of initial diagnosis at admission for detection of epilepsy

Where detection is of a specific type of epilepsy, rather than epilepsy overall, this is stated clearly in the first column.

Index Test	Number of studies	n	Interpreter of index test	Gold standard used in study	Sensitivity (95% CI)	Specificity (95% CI)	Risk of bias	Indirectness	Inconsistency	Imprecision	GRADE
ED assessment. Included full blood examination and tests for blood glucose levels, liver function, urea and electrolytes, as well as calcium and magnesium. Drug and ethanol levels were performed on a case-by-case basis. Computed tomography (CT) neuroimaging was usually performed for all patients presenting with first seizures, unless there is a contraindication. Cerebrospinal fluid (CSF) examination is performed when meningitis or encephalitis is suspected.	1⁹⁹	219	ED doctors	Final diagnosis using index test data plus imaging, EEG, longer follow up and consensus Non-epilepsy group: range of people without epilepsy initially suspected of epilepsy (not restricted to one differential diagnosis)	0.73 [0.66, 0.80]	0.32 [0.18, 0.49]	Sensitivity
							serious^a	none^b	NA	none	MOD
							Specificity
							serious^a	none^b	NA	none^c	MOD
Impression of admitting epileptologist, based on review of history, physical and available diagnostic testing as documented in the medical record prior to vEEG.	1¹⁴³	439	Admitting epileptologist	Clinical consensus/Video EEG. Non-epilepsy group: range of people without epilepsy initially suspected of epilepsy (not restricted to one differential diagnosis)	0.91 [0.82, 0.96]	0.86 [0.82, 0.90]	Sensitivity
							Serious^a	none^b	NA	serious^c	LOW
							Specificity
							Serious^a	none^b	NA	none^c	MOD
Initial Clinical diagnosis. Attending pediatric neurologist completed an extensive questionnaire on description of events, including postictal signs, possible provoking factors, medical history and family history. (CHILDREN)	1¹⁸⁴	536	Paediatric neurologist	Diagnosis based on 5 year follow up Non-epilepsy group: range of people without epilepsy initially suspected of epilepsy (not restricted to one differential diagnosis)	0.98 [0.96, 0.99]	0.86 [0.79, 0.91]	Sensitivity
							Serious^a	serious^b	NA	none^c	LOW
							Specificity
							Serious^a	serious^b	NA	serious^c	VERY LOW

(a): Risk of bias was assessed using the QUADAS-2 checklist. The evidence was downgraded by 1 increment if the majority of studies were rated at high risk of bias, and downgraded by 2 increments if the majority of studies were rated at very high risk of bias.
(b): Indirectness was assessed using the QUADAS-2 checklist items referring to applicability. The evidence was downgraded by 1 increment if the majority of studies were seriously indirect, and downgraded by 2 increments if the majority of studies are very seriously indirect
(c): Imprecision was assessed based on inspection of the confidence region in the diagnostic meta-analysis or, where diagnostic meta-analysis has not been conducted, assessed according to the range of confidence intervals in the individual studies. The evidence was downgraded by 1 increment when the confidence interval around the point estimate crossed one of the clinical thresholds (0.90 or 0.60), and downgraded by 2 increments when the confidence interval around the point estimate crossed both of the clinical thresholds. The upper clinical threshold marked the point above which recommendations would be possible, and the lower clinical threshold marked the point below which the tool would be regarded as of little clinical use.
(d): Inconsistency was assessed by visual inspection of the sensitivity/specificity plots, or data (if 2 studies). The evidence was downgraded by 1 increment if there was no overlap of 95% confidence intervals. For single studies no evaluation was made and ‘not applicable’ was recorded.
(e): No confidence intervals were presented because there was insufficient information available or there was a mismatch between the raw data and the accuracy results.

Table 14Clinical evidence summary: diagnostic test accuracy of other miscellaneous physiological scales for detection of epilepsy

Where detection is of a specific type of epilepsy, rather than epilepsy overall, this is stated clearly in the first column.

Index Test	Number of studies	n	Interpreter of index test	Gold standard used in study	Sensitivity (95% CI)	Specificity (95% CI)	Risk of bias	Indirectness	Inconsistency	Imprecision	GRADE
Hyperventilation and blood gas recovery. Interictal. If patient <65years, had an additional hyperventilation test (40 breaths per minute for 3 minutes. End tidal CO2 level had to be <2.5% after hyperventilation. Blood gases measured. Hyperventilation test considered negative if end tidal CO2 did not restore to >90% baseline value after 3 minutes recovery.	1⁹⁴	83	Neurophysiologist	Specific semiology Non-epilepsy group: Population suspected of epilepsy but with no known differential diagnoses	0.16 [0.06, 0.32]	0.43 [0.29, 0.59]	Sensitivity
							serious^a	serious^b	NA	none	LOW
							Specificity
							serious^a	serious^b	NA	none	LOW
Head up tilt test. Interictal. (No details available in paper)	1⁴³	49	NR	EEG plus clinical findings, over prolonged follow up Non-epilepsy group: Population suspected of epilepsy but with no known differential diagnoses	0.20 [0.01, 0.72]	0.09 [0.03, 0.22]	Sensitivity
							serious^a	none	NA	serious^c	LOW
							Specificity
							serious^a	none	NA	none	MOD
Epifinder application (a clinical decision support tool). Epifinder’s algorithm is a form of artificial intelligence that is based on pattern recognition. It utilises standardised terminology and heuristic algorithms that produce a list of differential diagnoses based on pattern recognition of a cluster of semiology against ILAE-defined epilepsy criteria	1¹⁴⁴	53	epilepsy trained neurologist	Video EEG Non-epilepsy group: Population suspected of epilepsy but with no known differential diagnoses	0.88 [0.70, 0.98]	0.85 [0.66, 0.96]	Sensitivity
							serious^a	none	NA	serious^c	LOW
							Specificity
							serious^a	none	NA	serious^c	LOW
Hypnosis Induction Profile (HIP) score (threshold of <=9). Interictal.	1¹⁰⁷	40	physician	Video EEG Non-epilepsy group: Population suspected of epilepsy but with no known differential diagnoses	0.69 [0.41, 0.89]	0.42 [0.22, 0.63]	Sensitivity
							serious^a	none	NA	serious^c	LOW
							Specificity
							serious^a	none	NA	serious^c	LOW
Not having an event during hypnosis	1¹⁰⁷	40	physician	Video EEG Non-epilepsy group: Population suspected of epilepsy but with no known differential diagnoses	0.88 [0.62, 0.98]	0.46 [0.26, 0.67]	Sensitivity
							serious^a	none	NA	serious^c	LOW
							Specificity
							serious^a	none	NA	serious^c	LOW
Review of systems questionnaire (threshold of <2.5)	1¹¹	60	physician	Video EEG Non-epilepsy group: PNES	0.90 [0.73, 0.98]	0.40 [0.23, 0.59]	Sensitivity
							Very serious^a	serious^b	NA	serious^c	VERY LOW
							Specificity
							Very serious^a	serious^b	NA	none	VERY LOW
Frontal Lobe Epilepsy and Parasomnias (FLEP) scale. Filled in on basis of reports from partners or relatives. Threshold not provided. DETECTING NOCTURNAL FRONTAL LOBE EPILEPSY	1⁵⁸	62	Research Assistant	Video EEG Non-PC epilepsy group: arousal parasomnia and sleep disorder	1.00 [0.89, 1.00]	0.90 [0.74, 0.98]	Sensitivity
							Serious^a	Serious^b	NA	serious^c	VERY LOW
							Specificity
							Serious^a	Serious^b	NA	Serious^c	VERY LOW
Frontal Lobe Epilepsy and Parasomnias (FLEP) scale. Filled in on basis of reports from partners or relatives. Threshold not provided. DETECTING NOCTURNAL FRONTAL LOBE EPILEPSY	1⁵⁸	62	Experienced physician	Video EEG Non-PC epilepsy group: arousal parasomnia and sleep disorder	1.0(0.86-1.00)	0.93 (0.79-0.98)	Sensitivity
							Serious^a	Serious^b	NA	serious^c	VERY LOW
							Specificity
							Serious^a	Serious^b	NA	Serious^c	VERY LOW
FLEP scale (excluding those with scores in uncertain range of 1-3). Filled in on basis of reports from partners or relatives. Threshold >3	1¹³¹	49	Medical doctor	Video EEG Non-epilepsy group: Parasomnias and idiopathic RBD	0.50 [0.16, 0.84]	1.00 [0.91, 1.00]	Sensitivity
							None^a	serious^b	NA	serious^c	LOW
							Specificity
							None^a	serious^b	NA	none	MOD
FLEP scale (including those with scores in uncertain range of 1-3 = NFLE). Filled in on basis of reports from partners or relatives. Threshold >0	1¹³¹	71	Medical doctor	Video EEG Non-epilepsy group: Parasomnias and idiopathic RBD	0.71 [0.42, 0.92]	0.72 [0.58, 0.83]	Sensitivity
							None^a	serious^b	NA	serious^c	LOW
							Specificity
							None^a	serious^b	NA	serious^c	LOW
Nocturnal frontal lobe epilepsy (including those with scores in uncertain range of 1-3 = NO NFLE). Filled in on basis of reports from partners or relatives. Threshold>3	1¹³¹	71	Medical doctor	Video EEG Non-epilepsy group: Parasomnias and idiopathic RBD	0.29 [0.08, 0.58]	1.00 [0.94, 1.00]	Sensitivity
							None^a	serious^b	NA	none	MOD
							Specificity
							None^a	serious^b	NA	none	MOD

(a): Risk of bias was assessed using the QUADAS-2 checklist. The evidence was downgraded by 1 increment if the majority of studies were rated at high risk of bias, and downgraded by 2 increments if the majority of studies were rated at very high risk of bias.
(b): Indirectness was assessed using the QUADAS-2 checklist items referring to applicability. The evidence was downgraded by 1 increment if the majority of studies were seriously indirect, and downgraded by 2 increments if the majority of studies are very seriously indirect
(c): Imprecision was assessed based on inspection of the confidence region in the diagnostic meta-analysis or, where diagnostic meta-analysis has not been conducted, assessed according to the range of confidence intervals in the individual studies. The evidence was downgraded by 1 increment when the confidence interval around the point estimate crossed one of the clinical thresholds (0.90 or 0.60), and downgraded by 2 increments when the confidence interval around the point estimate crossed both of the clinical thresholds. The upper clinical threshold marked the point above which recommendations would be possible, and the lower clinical threshold marked the point below which the tool would be regarded as of little clinical use.
(d): Inconsistency was assessed by visual inspection of the sensitivity/specificity plots, or data (if 2 studies). The evidence was downgraded by 1 increment if there was no overlap of 95% confidence intervals. For single studies no evaluation was made and ‘not applicable’ was recorded.
(e): No confidence intervals were presented because there was insufficient information available or there was a mismatch between the raw data and the accuracy results.

Table 15Clinical evidence summary: diagnostic test accuracy of different serum measurements for differentiation of people with autoimmune epilepsy from people with other epilepsy sub-types

Index Test	Number of studies	n	Interpreter of index test	Gold standard used in study	Sensitivity (95% CI)	Specificity (95% CI)	Risk of bias	Indirectness	Inconsistency	Imprecision	GRADE
Antibody prevalence in Epilepsy (APE) score; threshold >=4. Interictal. DETECTING AUTOIMMUNE EPILEPSY	1⁶⁴	387	NR	CNS-specific antibodies Non-autoimmune epilepsy group: other epilepsy groups	0.98 [0.88, 1.00]	0.78 [0.73, 0.82]	Sensitivity
							Serious^a	none^b	NA	Serious^c	LOW
							Specificity
							Serious^a	none^b	NA	None	MOD
Antibody prevalence in Epilepsy2 (APE2) score; threshold not reported. Interictal. DETECTING AUTOIMMUNE EPILEPSY	1¹³²	219	NR	Detection of NSAb Non-autoimmune epilepsy group: new onset focal epilepsy	0.435^e	0.791^e	Sensitivity
							Serious^a	none^b	NA	NA	MOD
							Specificity
							Serious^a	none^b	NA	NA	MOD

(a): Risk of bias was assessed using the QUADAS-2 checklist. The evidence was downgraded by 1 increment if the majority of studies were rated at high risk of bias, and downgraded by 2 increments if the majority of studies were rated at very high risk of bias.
(b): Indirectness was assessed using the QUADAS-2 checklist items referring to applicability. The evidence was downgraded by 1 increment if the majority of studies were seriously indirect, and downgraded by 2 increments if the majority of studies are very seriously indirect
(c): Imprecision was assessed based on inspection of the confidence region in the diagnostic meta-analysis or, where diagnostic meta-analysis has not been conducted, assessed according to the range of confidence intervals in the individual studies. The evidence was downgraded by 1 increment when the confidence interval around the point estimate crossed one of the clinical thresholds (0.90 or 0.60), and downgraded by 2 increments when the confidence interval around the point estimate crossed both of the clinical thresholds. The upper clinical threshold marked the point above which recommendations would be possible, and the lower clinical threshold marked the point below which the tool would be regarded as of little clinical use.
(d): Inconsistency was assessed by visual inspection of the sensitivity/specificity plots, or data (if 2 studies). The evidence was downgraded by 1 increment if there was no overlap of 95% confidence intervals. For single studies no evaluation was made and ‘not applicable’ was recorded.
(e): No confidence intervals were presented because there was insufficient information available or there was a mismatch between the raw data and the accuracy results.

Table 16Clinical evidence summary: diagnostic test accuracy of different psychological measurements for differentiation of people with autoimmune epilepsy from people with other epilepsy sub-types

Index Test	Number of studies	n	Interpreter of index test	Gold standard used in study	Sensitivity (95% CI)	Specificity (95% CI)	Risk of bias	Indirectness	Inconsistency	Imprecision	GRADE
Addenbrooke’s cognitive examination (ACE) attention domain (threshold >=0) Interictal. DETECTING AUTOIMMUNE EPILEPSY	1¹³²	219	NR	Detection of NSAb Non-autoimmune epilepsy group: new onset focal epilepsy	0.667^e	0.849^e	Sensitivity
							Serious^a	none^b	NA	NA	MOD
							Specificity
							Serious^a	none^b	NA	NA	MOD

(f): Risk of bias was assessed using the QUADAS-2 checklist. The evidence was downgraded by 1 increment if the majority of studies were rated at high risk of bias, and downgraded by 2 increments if the majority of studies were rated at very high risk of bias.
(g): Indirectness was assessed using the QUADAS-2 checklist items referring to applicability. The evidence was downgraded by 1 increment if the majority of studies were seriously indirect, and downgraded by 2 increments if the majority of studies are very seriously indirect
(h): Imprecision was assessed based on inspection of the confidence region in the diagnostic meta-analysis or, where diagnostic meta-analysis has not been conducted, assessed according to the range of confidence intervals in the individual studies. The evidence was downgraded by 1 increment when the confidence interval around the point estimate crossed one of the clinical thresholds (0.90 or 0.60), and downgraded by 2 increments when the confidence interval around the point estimate crossed both of the clinical thresholds. The upper clinical threshold marked the point above which recommendations would be possible, and the lower clinical threshold marked the point below which the tool would be regarded as of little clinical use.
(i): Inconsistency was assessed by visual inspection of the sensitivity/specificity plots, or data (if 2 studies). The evidence was downgraded by 1 increment if there was no overlap of 95% confidence intervals. For single studies no evaluation was made and ‘not applicable’ was recorded.
(j): No confidence intervals were presented because there was insufficient information available or there was a mismatch between the raw data and the accuracy results.

Table 17Electroencephalogram (EEG) unit costs

Conventional EEG, EMG or Nerve conduction Studies
Adults (19 years and over)
Currency code: AA33C	Activity	Unit Cost	Total Cost
Total	190,268	£199	£37,938,282
Elective	125	£1,952	£243,961
Non-elective long stay	157	£2,993	£469,837
Non-elective short stay	1,007	£827	£832,773
Day case	808	£807	£651,783
Regular day or night admissions	86	£993	£85,361
Outpatient procedures	141,294	£205	£28,914,172
Directly accessed diagnostic services	46,791	£144	£11,264,379
Children (18 years and under)
Currency code: AA33D	Activity	Unit Cost	Total Cost
Total	22,390	£340	£7,607,597
Elective	210	£1,186	£248,995
Non-elective long stay	77	£2,885	£222,125
Non-elective short stay	609	£1,422	£866,025
Day case	2,614	£651	£1,702,333
Regular day or night admissions	2	£1,092	£2,183
Outpatient procedures	18,591	£241	£4,471,167
Directly accessed diagnostic services	287	£330	£94,768
Complex Long-term EEG monitoring
Currency code: AA80Z	Activity	Unit Cost	Total Cost
Total	4,902	£2,067	£10,133,610
Elective	3,808	£2,126	£8,096,765
Non-elective long stay	476	£2,960	£1,409,167
Non-elective short stay	257	£1,182	£303,834
Day case	358	£901	£322,713
Regular day or night admissions	1	£674	£674
Outpatient procedures	-	-	-
Directly accessed diagnostic services	2	£228	£457
Standard Long-term EEG monitoring
Currency code: AA81Z	Activity	Unit Cost	Total Cost
Total	2,020	£491	£991,134
Elective	395	£994	£392,797
Non-elective long stay	118	£2,106	£248,475
Non-elective short stay	74	£860	£63,634
Day case	10	£1,217	£12,166
Regular day or night admissions	2	£1,809	£3,619
Outpatient procedures	1,308	£193	£252,104
Directly accessed diagnostic services	113	£162	£18,339

Table 18Electrocardiogram (ECG) unit costs

ECG monitoring or stress testing
Currency code: EY51Z	Activity	Unit Cost	Total Cost
Total	565,058	£102	£57,831,246
Elective	46	£643	£29,599
Non-elective long stay	4	£3,575	£14,300
Non-elective short stay	53	£783	£41,524
Day case	2,700	£464	£1,252,196
Regular day or night admissions	397	£457	£181,594
Outpatient procedures	330,956	£136	£45,047,653
Directly accessed diagnostic services	230,902	£49	£11,264,379

Table 19Magnetic Resonance Imaging (MRI) unit costs

Currency code	Currency description	Activity	Unit Cost	Total Cost
RD01A	MRI Scan of One Area, without Contrast, 19 years and over	1,440,377	£136	£196,146,270
RD01B	MRI Scan of One Area, without Contrast, between 6 and 18 years	62,170	£138	£8,592,099
RD01C	MRI Scan of One Area, without Contrast, 5 years and under	16,609	£135	£2,246,755
RD02A	MRI Scan of One Area, with Post-Contrast Only, 19 years and over	239,007	£151	£36,014,012
RD02B	MRI Scan of One Area, with Post-Contrast Only, between 6 and 18 years	7,569	£172	£1,301,693
RD02C	MRI Scan of One Area, with Post-Contrast Only, 5 years and under	1,374	£141	£193,099
RD03Z	MRI Scan of One Area, with Pre- and Post-Contrast	45,069	£215	£9,703,024
RD04Z	MRI Scan of Two or Three Areas, without Contrast	117,642	£142	£16,648,325
RD05Z	MRI Scan of Two or Three Areas, with Contrast	24,148	£204	£4,934,540
RD06Z	MRI Scan of more than Three Areas	45,209	£194	£8,771,400
RD07Z	MRI Scan Requiring Extensive Patient Repositioning	5,477	£263	£1,442,365

Table 20Computerised Tomography (CT) unit costs

Currency code	Currency description	Activity	Unit Cost	Total Cost
RD20A	CT Scan of One Area, without Contrast, 19 years and over	827,230	£83	£68,854,114
RD20B	CT Scan of One Area, without Contrast, between 6 and 18 years	13,504	£97	£1,308,085
RD20C	CT Scan of One Area, without Contrast, 5 years and under	13,579	£66	£894,029
RD21A	CT Scan of One Area, with Post-Contrast Only, 19 years and over	235,143	£107	£25,196,786
RD21B	CT Scan of One Area, with Post-Contrast Only, between 6 and 18 years	1,172	£133	£155,768
RD21C	CT Scan of One Area, with Post-Contrast Only, 5 years and under	695	£172	£119,719
RD22Z	CT Scan of One Area, with Pre- and Post-Contrast	24,731	£105	£2,586,066
RD23Z	CT Scan of Two Areas, without Contrast	55,248	£93	£5,123,143
RD24Z	CT Scan of Two Areas, with Contrast	230,506	£104	£23,883,214
RD25Z	CT Scan of Three Areas, without Contrast	24,080	£103	£2,475,934
RD26Z	CT Scan of Three Areas, with Contrast	358,745	£115	£41,322,696
RD27Z	CT Scan of more than Three Areas	83,205	£111	£9,201,145

Table 21Positron Emission Tomography (PET) and Single Photon Emission Computed Tomography (SPECT) unit costs

Currency code	Currency description	Activity	Unit Cost	Total Cost
RN07A	PET, 19 years and over	18,314	£830	£15,193,497
RN07B	PET, between 6 and 18 years	51	£215	£10,964
RN07C	PET, 5 years and under	5	£119	£595
RN08A	SPECT, 19 years and over	16,068	£319	£5,125,070
RN08B	SPECT, between 6 and 18 years	199	£332	£66,144
RN08C	SPECT, 5 years and under	26	£236	£6,145

Table 22Neurology appointment costs

Neurology appointments
Consultant led – adults
Non-Admitted Face-to-Face Attendance, Follow-up	£169
Non-Admitted Face-to-Face Attendance, First	£220
Multiprofessional Non-Admitted Face-to-Face Attendance, Follow-up	£237
Multiprofessional Non-Admitted Face-to-Face Attendance, First	£245
Non-consultant led – adults
Non-Admitted Face-to-Face Attendance, Follow-up	£115
Non-Admitted Face-to-Face Attendance, First	£113
Multiprofessional Non-Admitted Face-to-Face Attendance, Follow-up	£1,019
Multiprofessional Non-Admitted Face-to-Face Attendance, First	£127
Consultant led – children
Non-Admitted Face-to-Face Attendance, Follow-up	£305
Non-Admitted Face-to-Face Attendance, First	£435
Multiprofessional Non-Admitted Face-to-Face Attendance, Follow-up	£284
Multiprofessional Non-Admitted Face-to-Face Attendance, First	£412
Non-consultant led – children
Non-Admitted Face-to-Face Attendance, Follow-up	£240
Non-Admitted Face-to-Face Attendance, First	£851
Multiprofessional Non-Admitted Face-to-Face Attendance, Follow-up	£311
Multiprofessional Non-Admitted Face-to-Face Attendance, First	£445

Table 23PICO characteristics of review question

Population	Inclusion: Strata: Children and adults with suspected epilepsy. Children and adults with epilepsy, where uncertainty remains as to the type of epilepsy Exclusion: New-born babies with acute symptomatic seizures
Intervention	Any comparison of diagnostic strategies used in studies (these do not have to contain EEG or ECG but are likely to do so).
Comparison	Each other
Outcomes	mortality seizures (we will collect both binary data and time to event data) seizure frequency time to withdrawal of treatment quality of life (any validated scores) any adverse events Follow up: any available but stratify to <1 yr, 1-5 yrs, >5 yrs
Study design	RCTs only

Table 24Summary of studies included in the evidence review

Study

Intervention and comparison

Population

Outcomes

Rossetti, 2015¹⁶⁵

Continuous EEG (30-48 hours) versus routine EEG (2 × 30 mins over 48 hours)

inpatients from Switzerland in intensive care units with impaired consciousness; mean age 63.75 years. Inclusion: Inpatients >18 years in intensive or intermediate care units having impaired consciousness of any aetiology, defined as GCS of 11 or less or a FOUR score of 12 or less; referred from the treating team for EEG Exclusion: Weekend patients; patients in palliative care; those risking invasive procedures within 48 hours; those with recent (<36 hours) seizures or SE (96 hours)

Mortality at 6 months

Seizures at 6 months

Adverse events at 6 months

Zehtabchi, 2014²¹⁸

Micro EEG + routine care versus routine care

149 patients from USA; mean age 65. Inclusion All adult (18 year and older) ED patients with AMS, defined as any alteration in level of responsiveness or alertness or arousability, presenting as lethargy, delirium, confusion, agitation, coma, disinhibition, labile/blunted affects, or unexpected psychosis. Exclusion criteria included patients with immediately correctable causes of AMS (including finger stick or serum glucose less than 60 mg/dL); hypothermia (body temperature below 35.0°C); hyperthermia, heat exhaustion, or heat stroke; opioid overdose responding to naloxone; patients who were unable to undergo EEG recordings (e.g., severe scalp injury); hemodynamically unstable patients (systolic blood pressure < 90 mm Hg); uncooperative or combative patients; and patients who were discharged, admitted, or transferred before enrolment. Patients who had overt seizures in the ED were only included if they experienced prolonged postictal periods (at the discretion of the ED attending physician).

Mortality during inpatient period

Table 25Clinical evidence summary: continuous EEG vs Routine EEG

Outcomes	No of Participants (studies) Follow up	Quality of the evidence (GRADE)	Relative effect (95% CI)	Anticipated absolute effects
Outcomes	No of Participants (studies) Follow up	Quality of the evidence (GRADE)	Relative effect (95% CI)	Risk with control	Risk difference with intervention (95% CI)
Mortality	364 (1 study) 6 months	⊕⊕⊕⊝ MODERATE^a due to risk of bias	RR 1.01 (0.82 to 1.25)	Moderate
Mortality	364 (1 study) 6 months	⊕⊕⊕⊝ MODERATE^a due to risk of bias	RR 1.01 (0.82 to 1.25)	484 per 1000	5 more per 1000 (from 87 fewer to 121 more)
Health Related Quality of life	No evidence found
seizures	368 (1 study) 6 months	⊕⊕⊕⊝ MODERATE^a due to risk of bias	RR 3.59 (1.68 to 7.63)	Moderate
seizures	368 (1 study) 6 months	⊕⊕⊕⊝ MODERATE^a due to risk of bias	RR 3.59 (1.68 to 7.63)	44 per 1000	113 more per 1000 (from 30 more to 290 more)
Adverse events	368 (1 study) 6 months	⊕⊕⊝⊝ LOW^a^,^b due to risk of bias, imprecision	RR 0.83 (0.60 to 1.15)	Moderate
Adverse events	368 (1 study) 6 months	⊕⊕⊝⊝ LOW^a^,^b due to risk of bias, imprecision	RR 0.83 (0.60 to 1.15)	36 per 1000	52 fewer per 1000 (from 122 fewer to 46 more)
Seizure frequency	No evidence found
Time to withdrawal of treatment	No evidence found

a: The study had serious risk of bias due to possible selection bias
b: The confidence intervals crossed the lower MID of 0.8

Table 26Clinical evidence summary: micro EEG + routine care versus routine care

Outcomes	No of Participants (studies) Follow up	Quality of the evidence (GRADE)	Relative effect (95% CI)	Anticipated absolute effects
Outcomes	No of Participants (studies) Follow up	Quality of the evidence (GRADE)	Relative effect (95% CI)	Risk with control	Risk difference with intervention (95% CI)
Mortality	149 (1 study) unclear follow up	⊝⊝⊝⊝ VERY LOW^a^,^b due to risk of bias, imprecision	RR 1.04 (0.27 to 4.01)	Moderate
Mortality	149 (1 study) unclear follow up	⊝⊝⊝⊝ VERY LOW^a^,^b due to risk of bias, imprecision	RR 1.04 (0.27 to 4.01)	53 per 1000	2 more per 1000 (from 38 fewer to 158 more)
Health Related Quality of life	No evidence found
seizures	No evidence found
Adverse events	No evidence found
Seizure frequency	No evidence found
Time to withdrawal of treatment	No evidence found

a: The study had serious risk of bias due to possible selection bias
b: The confidence intervals crossed the upper and lower MIDS of 0.8 and 1.25

Table 27Electroencephalogram (EEG) unit costs

Conventional EEG, EMG or Nerve conduction Studies
Adults (19 years and over)
Currency code: AA33C	Activity	Unit Cost	Total Cost
Total	190,268	£199	£37,938,282
Elective	125	£1,952	£243,961
Non-elective long stay	157	£2,993	£469,837
Non-elective short stay	1,007	£827	£832,773
Day case	808	£807	£651,783
Regular day or night admissions	86	£993	£85,361
Outpatient procedures	141,294	£205	£28,914,172
Directly accessed diagnostic services	46,791	£144	£11,264,379
Children (18 years and under)
Currency code: AA33D	Activity	Unit Cost	Total Cost
Total	22,390	£340	£7,607,597
Elective	210	£1,186	£248,995
Non-elective long stay	77	£2,885	£222,125
Non-elective short stay	609	£1,422	£866,025
Day case	2,614	£651	£1,702,333
Regular day or night admissions	2	£1,092	£2,183
Outpatient procedures	18,591	£241	£4,471,167
Directly accessed diagnostic services	287	£330	£94,768
Complex Long-term EEG monitoring
Currency code: AA80Z	Activity	Unit Cost	Total Cost
Total	4,902	£2,067	£10,133,610
Elective	3,808	£2,126	£8,096,765
Non-elective long stay	476	£2,960	£1,409,167
Non-elective short stay	257	£1,182	£303,834
Day case	358	£901	£322,713
Regular day or night admissions	1	£674	£674
Outpatient procedures	-	-	-
Directly accessed diagnostic services	2	£228	£457
Standard Long-term EEG monitoring
Currency code: AA81Z	Activity	Unit Cost	Total Cost
Total	2,020	£491	£991,134
Elective	395	£994	£392,797
Non-elective long stay	118	£2,106	£248,475
Non-elective short stay	74	£860	£63,634
Day case	10	£1,217	£12,166
Regular day or night admissions	2	£1,809	£3,619
Outpatient procedures	1,308	£193	£252,104
Directly accessed diagnostic services	113	£162	£18,339

Table 28Electrocardiogram (ECG) unit costs

ECG monitoring or stress testing
Currency code: EY51Z	Activity	Unit Cost	Total Cost
Total	565,058	£102	£57,831,246
Elective	46	£643	£29,599
Non-elective long stay	4	£3,575	£14,300
Non-elective short stay	53	£783	£41,524
Day case	2,700	£464	£1,252,196
Regular day or night admissions	397	£457	£181,594
Outpatient procedures	330,956	£136	£45,047,653
Directly accessed diagnostic services	230,902	£49	£11,264,379

Table 29Magnetic Resonance Imaging (MRI) unit costs

Currency code	Currency description	Activity	Unit Cost	Total Cost
RD01A	MRI Scan of One Area, without Contrast, 19 years and over	1,440,377	£136	£196,146,270
RD01B	MRI Scan of One Area, without Contrast, between 6 and 18 years	62,170	£138	£8,592,099
RD01C	MRI Scan of One Area, without Contrast, 5 years and under	16,609	£135	£2,246,755
RD02A	MRI Scan of One Area, with Post-Contrast Only, 19 years and over	239,007	£151	£36,014,012
RD02B	MRI Scan of One Area, with Post-Contrast Only, between 6 and 18 years	7,569	£172	£1,301,693
RD02C	MRI Scan of One Area, with Post-Contrast Only, 5 years and under	1,374	£141	£193,099
RD03Z	MRI Scan of One Area, with Pre- and Post-Contrast	45,069	£215	£9,703,024
RD04Z	MRI Scan of Two or Three Areas, without Contrast	117,642	£142	£16,648,325
RD05Z	MRI Scan of Two or Three Areas, with Contrast	24,148	£204	£4,934,540
RD06Z	MRI Scan of more than Three Areas	45,209	£194	£8,771,400
RD07Z	MRI Scan Requiring Extensive Patient Repositioning	5,477	£263	£1,442,365

Table 30Computerised Tomography (CT) unit costs

Currency code	Currency description	Activity	Unit Cost	Total Cost
RD20A	CT Scan of One Area, without Contrast, 19 years and over	827,230	£83	£68,854,114
RD20B	CT Scan of One Area, without Contrast, between 6 and 18 years	13,504	£97	£1,308,085
RD20C	CT Scan of One Area, without Contrast, 5 years and under	13,579	£66	£894,029
RD21A	CT Scan of One Area, with Post-Contrast Only, 19 years and over	235,143	£107	£25,196,786
RD21B	CT Scan of One Area, with Post-Contrast Only, between 6 and 18 years	1,172	£133	£155,768
RD21C	CT Scan of One Area, with Post-Contrast Only, 5 years and under	695	£172	£119,719
RD22Z	CT Scan of One Area, with Pre- and Post-Contrast	24,731	£105	£2,586,066
RD23Z	CT Scan of Two Areas, without Contrast	55,248	£93	£5,123,143
RD24Z	CT Scan of Two Areas, with Contrast	230,506	£104	£23,883,214
RD25Z	CT Scan of Three Areas, without Contrast	24,080	£103	£2,475,934
RD26Z	CT Scan of Three Areas, with Contrast	358,745	£115	£41,322,696
RD27Z	CT Scan of more than Three Areas	83,205	£111	£9,201,145

Table 31Positron Emission Tomography (PET) and Single Photon Emission Computed Tomography (SPECT) unit costs

Currency code	Currency description	Activity	Unit Cost	Total Cost
RN07A	PET, 19 years and over	18,314	£830	£15,193,497
RN07B	PET, between 6 and 18 years	51	£215	£10,964
RN07C	PET, 5 years and under	5	£119	£595
RN08A	SPECT, 19 years and over	16,068	£319	£5,125,070
RN08B	SPECT, between 6 and 18 years	199	£332	£66,144
RN08C	SPECT, 5 years and under	26	£236	£6,145

Table 32Neurology appointment costs

Neurology appointments
Consultant led – adults
Non-Admitted Face-to-Face Attendance, Follow-up	£169
Non-Admitted Face-to-Face Attendance, First	£220
Multiprofessional Non-Admitted Face-to-Face Attendance, Follow-up	£237
Multiprofessional Non-Admitted Face-to-Face Attendance, First	£245
Non-consultant led – adults
Non-Admitted Face-to-Face Attendance, Follow-up	£115
Non-Admitted Face-to-Face Attendance, First	£113
Multiprofessional Non-Admitted Face-to-Face Attendance, Follow-up	£1,019
Multiprofessional Non-Admitted Face-to-Face Attendance, First	£127
Consultant led – children
Non-Admitted Face-to-Face Attendance, Follow-up	£305
Non-Admitted Face-to-Face Attendance, First	£435
Multiprofessional Non-Admitted Face-to-Face Attendance, Follow-up	£284
Multiprofessional Non-Admitted Face-to-Face Attendance, First	£412
Non-consultant led – children
Non-Admitted Face-to-Face Attendance, Follow-up	£240
Non-Admitted Face-to-Face Attendance, First	£851
Multiprofessional Non-Admitted Face-to-Face Attendance, Follow-up	£311
Multiprofessional Non-Admitted Face-to-Face Attendance, First	£445

FINAL

Evidence review underpinning recommendations 1.2.1 – 1.2.10 in the NICE guideline

Developed by the National Guideline Centre

Disclaimer: The recommendations in this guideline represent the view of NICE, arrived at after careful consideration of the evidence available. When exercising their judgement, professionals are expected to take this guideline fully into account, alongside the individual needs, preferences and values of their patients or service users. The recommendations in this guideline are not mandatory and the guideline does not override the responsibility of healthcare professionals to make decisions appropriate to the circumstances of the individual patient, in consultation with the patient and/or their carer or guardian.

Local commissioners and/or providers have a responsibility to enable the guideline to be applied when individual health professionals and their patients or service users wish to use it. They should do so in the context of local and national priorities for funding and developing services, and in light of their duties to have due regard to the need to eliminate unlawful discrimination, to advance equality of opportunity and to reduce health inequalities. Nothing in this guideline should be interpreted in a way that would be inconsistent with compliance with those duties.

NICE guidelines cover health and care in England. Decisions on how they apply in other UK countries are made by ministers in the Welsh Government, Scottish Government, and Northern Ireland Executive. All NICE guidance is subject to regular review and may be updated or withdrawn.

Bookshelf ID: NBK586312PMID: 36395299

Evidence review: Diagnosis of epilepsies

Authors

1. Diagnosis of epilepsy

1.1. Introduction

1.2. Review question: What is the most accurate approach for 1) diagnosis of epilepsy and 2) differentiation between types of epilepsy?

1.2.1. Summary of the protocol

1.2.2. Methods and process

1.2.3. Effectiveness evidence

1.2.3.1. Included studies

1.2.3.2. Excluded studies

1.2.4. Summary of clinical studies included in the evidence review

1.2.5. Quality assessment of clinical studies included in the evidence review

STRATUM 1: Detection of any epilepsy (differentiation from no epilepsy)

STRATUM 2: Differentiation between specific types of epilepsy

1.2.6. Economic evidence

1.2.6.1. Included studies

1.2.6.2. Excluded studies

1.2.7. Economic model

1.2.8. Unit costs

1.3. Review question: What is the most clinically and cost-effective approach for diagnosis of epilepsies?

1.3.1. Summary of the protocol

1.3.2. Methods and process

1.3.3. Effectiveness evidence

1.3.3.1. Included studies

1.3.3.2. Excluded studies

1.3.4. Summary of studies included in the effectiveness evidence

1.3.5. Summary of the effectiveness evidence

1.3.6. Economic evidence

1.3.6.1. Included studies

1.3.6.2. Excluded studies

1.3.7. Economic model

1.3.8. Unit costs

1.4. Evidence statements

1.4.1. Effectiveness/Qualitative

1.4.2. Economic

1.5. The committee’s discussion of the evidence

1.5.1. The outcomes that matter most

1.5.1.1. Diagnostic accuracy review

1.5.1.2. RCT review

1.5.2. The quality of the evidence

1.5.2.1. Diagnostic accuracy review

1.5.2.2. RCT review

1.5.3. Benefits and harms

Stratum 1: Differentiating between epilepsy and non-epilepsy

Semiology, signs and symptoms

Serum measures

ECG

Imaging tests

EEG tests

Magnetoencephalography / Transcranial magnetic stimulation tests

Psychological tests

Linguistic tests

Electromyography (EMG) and accelerometers

Initial diagnosis at admission

Miscellaneous tests

Stratum 2: Differentiating between epilepsy sub-types

1.5.4. Cost effectiveness and resource use

1.5.5. Recommendations supported by this evidence review

References

Appendices

Appendix A. Review protocols

A.1. Review protocol: Diagnostic accuracy of point of care devices

A.2. Review protocol for diagnostic strategies

A.3. Health economic review protocol

Appendix B. Literature search strategies

B.1. Clinical search literature search strategy

B.2. Health Economics literature search strategy

Appendix C. Clinical evidence selection

C.1. Flow chart of clinical study selection for the review of diagnostic accuracy

C.2. Flow chart of clinical study selection for the review of clinical efficacy of diagnostic strategies

Appendix D. Clinical evidence tables

D.1. Clinical evidence Diagnostic accuracy

Appendix E. Coupled sensitivity and specificity forest plots and sROC curves

E.1. Diagnostic accuracy

E.2. Diagnostic strategies

Appendix F. GRADE tables

Appendix G. Health economic evidence selection

Appendix H. Health economic evidence tables

Appendix I. Health economic model

Appendix J. QUADAS2 risk of bias assessment