Advertisement

Electroencephalography after a single unprovoked seizure

Open ArchivePublished:April 25, 2017DOI:https://doi.org/10.1016/j.seizure.2017.03.001

      Highlights

      • Clinical guidelines recommend that an EEG be obtained after a single unprovoked seizure in children and adults.
      • Epileptiform abnormalities on electroencephalography (EEG) are related to seizure disorders and are a factor in determining risk of recurrence after a single unprovoked seizure.
      • The yield for recording an epileptiform abnormality is increased if the EEG recording occurs shortly after the initial seizure, if initially normal EEG studies are repeated, if EEG studies are prolonged and if EEG studies include sleep deprivation.

      Abstract

      Electroencephalography (EEG) is an essential diagnostic tool in the evaluation of seizure disorders. In particular, EEG is used as an additional investigation for a single unprovoked seizure. Epileptiform abnormalities are related to seizure disorders and have been shown to predict recurrent unprovoked seizures (i.e., a clinical definition of epilepsy). Thus, the identification of epileptiform abnormalities after a single unprovoked seizure can inform treatment options. The current review addresses the relationship between EEG abnormalities and seizure recurrence. This review also addresses factors that are found to improve the yield of recording epileptiform abnormalities including timing of EEG relative to the new-onset seizure, use of repeat studies, use of sleep deprivation and prolonged recordings.

      Keywords

      1. Introduction

      Electroencephalography (EEG) remains an essential diagnostic tool in the evaluation of seizure disorders. The most recent clinical guidelines from the American Academy of Neurology for both children [
      • Hirtz D.
      • Ashwal S.
      • Berg A.
      • Bettis D.
      • Camfield C.
      • Camfield P.
      • et al.
      Practice parameter: evaluating a first nonfebrile seizure in children: report of the quality standards subcommittee of the American Academy of Neurology, The Child Neurology Society, and The American Epilepsy Society.
      ] and adults [
      • Krumholz A.
      • Wiebe S.
      • Gronseth G.
      • Shinnar S.
      • Levisohn P.
      • Ting T.
      • et al.
      Practice parameter: evaluating an apparent unprovoked first seizure in adults (an evidence-based review): report of the Quality Standards Subcommittee of the American Academy of Neurology and the American Epilepsy Society.
      ] recommend that an EEG be obtained after a single unprovoked seizure. Indeed, EEG abnormalities may be useful in identifying epileptogenic foci, structural abnormalities, and/or electrographic patterns associated with specific epilepsy syndromes. However, it is the potential predictive value of EEG that can have the most prominent role in the evaluation of a single unprovoked seizure.
      The challenge in evaluating a patient after a single unprovoked seizure is identifying those patients that will go on to develop recurrent seizures from those who will have only a single seizure in their lifetime. The risk of seizure recurrence is greatest within the first 1–2 years (21–45%) after a single unprovoked seizure in adults [
      • Krumholz A.
      • Wiebe S.
      • Gronseth G.S.
      • Gloss D.S.
      • Sanchez A.M.
      • Kabir A.a.
      • et al.
      Evidence-based guideline: management of an unprovoked first seizure in adults: report of the Guideline Development Subcommittee of the American Academy of Neurology and the American Epilepsy Society.
      ]. Similarly, in children, the risk is also greatest within the first 1–2 years (14–65%) after a single seizure [
      • Hirtz D.
      • Berg a.
      • Bettis D.
      • Camfield C.
      • Camfield P.
      • Crumrine P.
      • et al.
      Practice parameter: treatment of the child with a first unprovoked seizure: report of the Quality Standards Subcommittee of the American Academy of Neurology and the Practice Committee of the Child Neurology Society.
      ]. The risk increases to 60–90% after a second unprovoked seizure [
      • Hauser W.A.
      • Rich S.S.
      • Lee J.R.
      • Annegers J.F.
      • Anderson V.E.
      Risk of recurrent seizures after two unprovoked seizures.
      ,
      • Fisher R.S.
      • Acevedo C.
      • Arzimanoglou A.
      • Bogacz A.
      • Cross J.H.
      • Elger C.E.
      • et al.
      ILAE official report: a practical clinical definition of epilepsy.
      ], at which point it is felt that the traditional definition of epilepsy (an enduring predisposition to generating recurrent seizures [
      • Fisher R.S.
      • Boas W.E.
      • Blume W.
      • Elger C.
      • Genton P.
      • Lee P.
      • et al.
      Epileptic seizures and epilepsy: definitions proposed by the International League Against Epilepsy(ILAE) and the International Bureau for Epilepsy(IBE).
      ]) has been met. As a clinical definition, this requires waiting until a second unprovoked clinical seizure has occurred before providing a diagnosis of epilepsy.
      More recently, an operational definition of epilepsy has been published, which proposes that the diagnosis of epilepsy can be applied after a single unprovoked seizure if there is evidence to suggest that the risk of recurrence is sufficiently increased (greater than 60%) such that it approaches the risk of recurrence associated with a second unprovoked clinical seizure [
      • Fisher R.S.
      • Acevedo C.
      • Arzimanoglou A.
      • Bogacz A.
      • Cross J.H.
      • Elger C.E.
      • et al.
      ILAE official report: a practical clinical definition of epilepsy.
      ]. Specifically, it proposes that epilepsy can be diagnosed if associated investigations are consistent with an identifiable epilepsy syndrome, or if investigations after a single unprovoked seizure lead to the determination that the patient is at high risk of seizure recurrence (similar to that of the recurrence risk associated with two unprovoked seizures) over the next 10 years.
      This operational definition of epilepsy places emphasis on the use of supplementary investigations after a single unprovoked seizure to help predict recurrence risk. In the case of EEG, the finding of an abnormality could impact the management of a patient with a newly diagnosed seizure based on risk of recurrence. The following is a review of the literature as it pertains to the use of EEG after a single unprovoked seizure. The objective of this review was to determine the relationship between EEG abnormalities (focal or generalized, epileptiform discharges or non-epileptiform slowing of background rhythms) and seizure recurrence. An additional objective of this review was to determine aspects of the EEG – timing relative to seizure occurrence, the utility of repeat EEG studies, the inclusion of sleep deprivation with recording of the sleep state and the use of prolonged EEG recordings – that are thought to improve the yield of identifying EEG abnormalities in the context of assessment after a single unprovoked seizure. EEGs included in the reviewed studies were typically standard outpatient studies of approximately 30 min duration using the international standard 10–20 system of electrode placement unless specified otherwise.

      2. Relationship between EEG abnormalities and seizure recurrence

      From the perspective of EEG, it is well documented that seizures are related to EEG abnormalities. Upwards of 12–50% of adults and 18–56% of children with single unprovoked seizures had epileptiform abnormalities on EEG [
      • Wirrell E.C.
      Prognostic significance of interictal epileptiform discharges in newly diagnosed seizure disorders.
      ]. The authors suggested that the EEG not only helped distinguish epileptic seizures from non-epileptic events, but EEG also helped identify specific seizure types. It was determined that seizure types influenced the likelihood of EEG abnormalities; absence seizures were associated with EEG abnormalities in 92%, atonic or myoclonic seizures in 85%, complex partial seizures in 59% and generalized tonic-clonic seizures in 44% [
      • Wirrell E.C.
      Prognostic significance of interictal epileptiform discharges in newly diagnosed seizure disorders.
      ,
      • Carpay J.A.
      • de Weerd A.W.
      • Schimsheimer R.J.
      • Stroink H.
      • Brouwer O.F.
      • Peters A.C.
      • et al.
      The diagnostic yield of a second EEG after partial sleep deprivation: a prospective study in children with newly diagnosed seizures.
      ].
      The presence of any EEG abnormality (an epileptiform abnormality and/or focal slowing of background rhythms) predicted seizure recurrence (e.g., [
      • Berg A.T.
      • Shinnar S.
      The risk of seizure recurrence following a first unprovoked seizure: a quantitative review.
      ,
      • Pohlmann-Eden B.
      • Newton M.
      First seizure: EEG and neuroimaging following an epileptic seizure.
      ]), although the presence of epileptiform discharges was thought to offer more predictive ability. A systematic review and meta-analysis of 15 studies examining the diagnostic accuracy of routine EEG after a single unprovoked seizure demonstrated, in adults, a sensitivity and specificity of interictal epileptiform discharge for seizure recurrence of 17.3% and 94.7% respectively [
      • Bouma H.K.
      • Labos C.
      • Gore G.C.
      • Wolfson C.
      • Keezer M.R.
      The diagnostic accuracy of routine electroencephalography after a first unprovoked seizure.
      ]. This corresponded to a 77% post-test probability of recurrent seizures when epileptiform discharges were seen and a 47% post-test probability of recurrent seizures when not. A previous meta-analysis showed a post-test probability of epilepsy in 49.5% of those with epileptiform abnormalities on EEG and 27.4% in those without [
      • Krumholz A.
      • Wiebe S.
      • Gronseth G.
      • Shinnar S.
      • Levisohn P.
      • Ting T.
      • et al.
      Practice parameter: evaluating an apparent unprovoked first seizure in adults (an evidence-based review): report of the Quality Standards Subcommittee of the American Academy of Neurology and the American Epilepsy Society.
      ]. It was felt that the difference might have been related to the addition of more recent studies and the strict eligibility criteria for study inclusion [
      • Bouma H.K.
      • Labos C.
      • Gore G.C.
      • Wolfson C.
      • Keezer M.R.
      The diagnostic accuracy of routine electroencephalography after a first unprovoked seizure.
      ]. Nevertheless, in both systematic reviews, the presence of epileptiform discharges on EEG in the context of a single unprovoked seizure was predictive of the development of recurrent seizures.
      In the same systematic review, when considering studies involving children, routine EEG after a single unprovoked seizure was found to have a sensitivity and specificity for seizure recurrence of 57.8% and 69.6% respectively. This corresponded to a post-test probability of developing recurrent seizures of 66% when epileptiform abnormalities were observed and 38% when absent [
      • Bouma H.K.
      • Labos C.
      • Gore G.C.
      • Wolfson C.
      • Keezer M.R.
      The diagnostic accuracy of routine electroencephalography after a first unprovoked seizure.
      ]. This is in keeping with previous findings in general pediatric populations (e.g., [
      • Hirtz D.
      • Berg a.
      • Bettis D.
      • Camfield C.
      • Camfield P.
      • Crumrine P.
      • et al.
      Practice parameter: treatment of the child with a first unprovoked seizure: report of the Quality Standards Subcommittee of the American Academy of Neurology and the Practice Committee of the Child Neurology Society.
      ]).
      Taken together, these studies indicate that EEG can improve the determination of recurrence risk after a single unprovoked seizure. Specifically, an epileptiform abnormality is associated with a higher risk of seizure recurrence. Therefore, it is important to identify and optimize factors that can improve the likelihood of recording an epileptiform abnormality on EEG after a new-onset seizure.

      3. Timing of EEG after a single unprovoked seizure

      One factor that is thought to influence the ability to record an epileptiform abnormality on EEG after a single unprovoked seizure is the timing of the EEG recording relative to the clinical event. Interestingly, although it has been identified as a potentially important factor in assessing the ability of EEG to predict recurrent seizures, relatively few studies have addressed this issue directly. From a neurophysiological perspective, Pohlmann-Eden and Newton [
      • Pohlmann-Eden B.
      • Newton M.
      First seizure: EEG and neuroimaging following an epileptic seizure.
      ] postulated that immediately after a seizure, the brain is in a state of hyperexcitability, which reduces over time. According to this hypothesis, an early EEG (shortly after seizure occurrence) should offer a higher yield with respect to recording epileptiform abnormalities, if present. Concerns surrounding the use of early EEG include logistical issues of organizing and obtaining an early EEG, the potential inability of obtaining a sleep-deprived EEG at that early time-point, and the difficulty in interpreting non-epileptiform abnormalities in a time-period where transient abnormalities (e.g., post-ictal slowing) might be present [
      • Wirrell E.C.
      Prognostic significance of interictal epileptiform discharges in newly diagnosed seizure disorders.
      ].
      For studies included in the systematic review by Bouma et al. [
      • Bouma H.K.
      • Labos C.
      • Gore G.C.
      • Wolfson C.
      • Keezer M.R.
      The diagnostic accuracy of routine electroencephalography after a first unprovoked seizure.
      ], the timing of EEG after first seizure was either not reported or variable (ranging from less than 48 h to several months). There is a limitation in the literature regarding studies directly comparing early and late EEG recordings. King et al. [
      • King M.A.
      • Newton M.R.
      • Jackson G.D.
      • Fitt G.J.
      • Mitchell L.A.
      • Silvapulle M.J.
      • et al.
      Epileptology of the first-seizure presentation: a clinical, electroencephalographic, and magnetic resonance imaging study of 300 consecutive patients.
      ] identified epileptiform abnormalities in 51% of patients (aged 5 years or older) who had an EEG within 24 h of the seizure, as compared to 34% in patients who had their initial EEG at a later point in time. In a pediatric population (1–17 years old), there was no statistical difference in identifying epileptiform abnormalities between EEGs obtained less than 48 h (37%) and those obtained greater than 48 h (40%) from the first seizure [
      • Hamiwka L.D.
      • Singh N.
      • Niosi J.
      • Wirrell E.C.
      Diagnostic inaccuracy in children referred with “first seizure”: role for a first seizure clinic.
      ] (although noting that only 20% of patients obtained an EEG prior to 48 h).
      Other studies have reported the non-comparative findings of EEGs completed shortly after a first seizure. Neufeld et al. [
      • Neufeld M.Y.
      • Chistik V.
      • Vishne T.H.
      • Korczyn A.D.
      The diagnostic aid of routine EEG findings in patients presenting with a presumed first-ever unprovoked seizure.
      ] reported epileptiform activity in 21% of EEGs completed within 48 h of a first seizure in an adult population (15–87 years old). Schreiner and Pohlmann-Eden [
      • Schreiner A.
      • Pohlmann-Eden B.
      Value of the early electroencephalogram after a first unprovoked seizure.
      ] identified epileptiform abnormalities in 26.8% of patients (17–84 years old) in whom an EEG was obtained within 48 h of their first seizure. A study demonstrating the logistic feasibility of an ‘acute’ EEG in an emergency room setting involving adult patients (16–74 years old) identified epileptiform abnormalities in approximately 50% of those with a clinical diagnosis of seizure confirmed by an epileptologist who saw the patient in an outpatient clinic setting within 2 weeks of presentation to the emergency department [
      • Paliwal P.
      • Wakerley B.R.
      • Yeo L.L.L.
      • Ali K.M.
      • Ibrahim I.
      • Wilder-Smith E.
      • et al.
      Early electroencephalography in patients with Emergency Room diagnoses of suspected new-onset seizures: diagnostic yield and impact on clinical decision-making.
      ]. In this study, the EEG was performed within 24 h of presentation to the emergency department, prior to discharge. Approximately 26% of patients had seizure recurrence within the follow-up time period. In another recent study investigating the ability of early comprehensive assessments to improve patient care after a single unprovoked seizure, showed that overall, approximately 14.5% of patients who had a standard EEG within 72 h (the majority of which were done within 24 h) had an epileptogenic focus identified on EEG [
      • Fisch L.
      • Lascano A.M.
      • Vernaz Hegi N.
      • Girardin F.
      • Kapina V.
      • Heydrich L.
      • et al.
      Early specialized care after a first unprovoked epileptic seizure.
      ].
      These studies suggest that obtaining an EEG shortly after a new-onset seizure can improve the yield of identifying epileptiform abnormalities. Moreover, these studies show that it is logistically feasible in the clinical setting. However, the current literature is limited with few studies addressing a direct comparison of ‘early’ (within 72 h) and ‘late’ EEGs for detecting epileptiform abnormalities. One group of studies which have indirectly addressed this issue concerns the use of serial EEG recordings to improve the yield of identifying epileptiform abnormalities if initial EEG studies were found to be normal. The concept is that, if the patient truly has a seizure tendency, repeating the EEG will increase the probability of identifying an abnormality.

      4. Use of serial EEG studies

      If the initial EEG after a single unprovoked seizure is found to be normal, regardless of timing, subsequent recordings can increase the yield of detecting epileptiform abnormalities. In a group of patients with epilepsy that demonstrate interictal epileptiform abnormalities on EEG, approximately 50 % of patients demonstrated epileptiform activity on the first EEG study, which increased to 92% or patients after 4 consecutive studies with much lower incremental yield after that [
      • Salinsky M.
      • Kanter R.
      • Dasheiff R.M.
      Effectiveness of multiple EEGs in supporting the diagnosis of epilepsy: an operational curve.
      ]. Similarly, a population based study involving patients with a single unprovoked seizure demonstrated an incremental cumulative increase in yield for detecting EEG abnormalities with multiple recordings [
      • Baldin E.
      • Hauser W.A.
      • Buchhalter J.R.
      • Hesdorffer D.C.
      • Ottman R.
      Yield of epileptiform electroencephalogram abnormalities in incident unprovoked seizures: a population-based study.
      ]. Across subjects, the incremental yield was 39.1% after the first EEG recording, 49.1% after the second, and 67.6% after the third. In this study, the median time to the first EEG recording was 3 days.
      These findings suggest that the highest yield for identifying epileptiform activity on EEG is greatest on the initial study yet, yield can seemingly be improved with repeated studies. This highlights a conceptual paradox such that a sample effect exists but is limited by the decreasing incremental yield over time.

      5. Inclusion of sleep deprivation and sleep in EEG studies

      Repeat EEG studies after an initial normal routine EEG often include sleep deprivation and a recording of the sleep state. Indeed, a survey of Dutch neurologists indicated that 48% of physicians requested a sleep-deprived EEG if the initial EEG and MRI were normal [
      • Askamp J.
      • Van Putten M.J.
      Diagnostic decision-making after a first and recurrent seizure in adults.
      ]. Carpay et al. [
      • Carpay J.A.
      • de Weerd A.W.
      • Schimsheimer R.J.
      • Stroink H.
      • Brouwer O.F.
      • Peters A.C.
      • et al.
      The diagnostic yield of a second EEG after partial sleep deprivation: a prospective study in children with newly diagnosed seizures.
      ] demonstrated that in a group of children and adolescents (1 month – 16 years of age) with a newly diagnosed seizure, 34% who initially had a normal standard EEG were found to have epileptiform abnormalities on a repeat EEG with sleep deprivation. This increased the yield of identifying epileptiform discharges in this population from 56% to 67%. Similarly, in a study of patients with new-onset seizures, sleep-deprived EEG demonstrated an epileptiform abnormality in 35% of those who had a normal initial awake EEG, increasing the yield of EEG abnormality from 43% to 61% [
      • King M.A.
      • Newton M.R.
      • Jackson G.D.
      • Fitt G.J.
      • Mitchell L.A.
      • Silvapulle M.J.
      • et al.
      Epileptology of the first-seizure presentation: a clinical, electroencephalographic, and magnetic resonance imaging study of 300 consecutive patients.
      ]. Interestingly, in the study by Schreiner and Pohlmann-Eden [
      • Schreiner A.
      • Pohlmann-Eden B.
      Value of the early electroencephalogram after a first unprovoked seizure.
      ] that assessed the use of early EEG, only 13.3% of patients with an initially normal awake EEG showed epileptiform discharges with repeat sleep EEG performed 3–7 days later. The authors argued that this relatively low yield was related to the delay in obtaining the repeat sleep-deprived study.
      Sleep deprivation can be considered an activating procedure for EEG because some stages of sleep can be associated with an increased probability of inducing epileptiform abnormalities in those with an underlying seizure tendency. Drowsiness and early stages of non-rapid eye movement (NREM) sleep states tend to induce ictal and inter-ictal epileptiform discharges more readily than rapid eye movement (REM) sleep states for most forms of epilepsy (e.g., [
      • Shouse M.N.
      • da Silva A.M.
      • Sammaritano M.
      Circadian rhythm, sleep, and epilepsy.
      ,
      • Herman S.T.
      • Walczak T.S.
      • Bazil C.W.
      Distribution of partial seizures during the sleep–wake cycle: differences by seizure onset site.
      ]).
      The use of a sleep deprived EEG as a follow-up examination to increase the yield of recording epileptiform discharges incites many interesting and unresolved issues. It has been suggested that sleep deprivation results in sleep that is more activating than natural sleep alone. It is unclear whether sleep deprivation itself could be related to an activated cortical state even in the absence of sleep (for a review regarding the controversies surrounding sleep deprivation and sleep, see [
      • Giorgi F.S.
      • Guida M.
      • Caciagli L.
      • Maestri M.
      • Carnicelli L.
      • Bonanni E.
      • et al.
      What is the role for EEG after sleep deprivation in the diagnosis of epilepsy? Issues, controversies, and future directions.
      ]). The use of medication to induce sleep remains controversial because of the potential effects that sedating medications can have on the EEG recording that could lead to confounders. None of the reviewed studies included the use of sedating medications to induce sleep for the EEG. It was either not specified or it was specifically stated that sleep was not induced [
      • Carpay J.A.
      • de Weerd A.W.
      • Schimsheimer R.J.
      • Stroink H.
      • Brouwer O.F.
      • Peters A.C.
      • et al.
      The diagnostic yield of a second EEG after partial sleep deprivation: a prospective study in children with newly diagnosed seizures.
      ,
      • Delil S.
      • Senel G.B.
      • Demiray D.Y.
      • Yeni N.
      The role of sleep electroencephalography in patients with new onset epilepsy.
      ].
      It remains unclear whether the increased yield of sleep deprived EEG studies could, in part, simply be related to sample effect, wherein a repeat sleep recording increases the yield of recording epileptiform discharges due to the serial study effect described above. Moreover, the studies involving sleep deprivation [
      • Carpay J.A.
      • de Weerd A.W.
      • Schimsheimer R.J.
      • Stroink H.
      • Brouwer O.F.
      • Peters A.C.
      • et al.
      The diagnostic yield of a second EEG after partial sleep deprivation: a prospective study in children with newly diagnosed seizures.
      ,
      • King M.A.
      • Newton M.R.
      • Jackson G.D.
      • Fitt G.J.
      • Mitchell L.A.
      • Silvapulle M.J.
      • et al.
      Epileptology of the first-seizure presentation: a clinical, electroencephalographic, and magnetic resonance imaging study of 300 consecutive patients.
      ]) allowed for prolonged recordings if necessary to capture the sleep state. The prolonged nature of these recordings could in theory improve the yield of recording epileptiform abnormalities by increasing the length of the recording. Nevertheless, there is a complex relationship between sleep deprivation, sleep and the generation of epileptiform discharges in those patients with an underlying seizure tendency that remains to be fully elucidated.

      6. Prolonged EEG studies

      In addition to prolonged sleep-deprived EEG, other forms of prolonged EEG studies include prolonged video-EEG and ambulatory-EEG [
      • Michel V.
      • Mazzola L.
      • Lemesle M.
      • Vercueil L.
      Long-term EEG in adults: sleep-deprived EEG (SDE), ambulatory EEG (Amb-EEG) and long-term video-EEG recording (LTVER).
      ]. The identification of epileptiform discharges on 24-h video-EEG after a first unprovoked seizure has been shown to be a risk factor for seizure recurrence [
      • 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.
      ]. In this study, epileptiform abnormalities were identified in 41.8% of patients (3–77 years old) who underwent 24-h video-EEG recording within 7 days of their first unprovoked seizure. Fisch et al. [
      • Fisch L.
      • Lascano A.M.
      • Vernaz Hegi N.
      • Girardin F.
      • Kapina V.
      • Heydrich L.
      • et al.
      Early specialized care after a first unprovoked epileptic seizure.
      ] determined that a prolonged video-EEG (approximately 18 h in duration) provided an additional yield of 30% after a normal standard EEG in patients (>15 years old). An ambulatory-EEG has been demonstrated to be as effective in providing an improvement in the yield of identifying an epileptiform abnormality in comparison to standard or sleep-deprived EEG, with the additional benefit of detecting seizures over the course of the prolonged (24 h) recording that were not seen on the sleep deprived (30–60 min) study [
      • Liporace J.
      • Tatum IV, W.
      • Lee Morris G.
      • French J.
      Clinical utility of sleep-deprived versus computer-assisted ambulatory 16-channel EEG in epilepsy patients: a multi-center study.
      ,
      • Michel V.
      • Mazzola L.
      • Lemesle M.
      • Vercueil L.
      Long-term EEG in adults: sleep-deprived EEG (SDE), ambulatory EEG (Amb-EEG) and long-term video-EEG recording (LTVER).
      ].

      7. Summary

      Whereas it is clear that two unprovoked seizures (i.e., a clinical definition of epilepsy) requires treatment with anti-epileptic medications in most circumstances, a clinical challenge remains in single unprovoked seizures, where a decision must be made between treatment with an anti-epileptic medication or conservative management with observation. Recent guidelines suggest that an abnormality in associated testing, such as MRI or EEG, may warrant treatment after a single unprovoked seizure because the risk of recurrent seizures in the context of an abnormality is sufficiently high [
      • Fisher R.S.
      • Acevedo C.
      • Arzimanoglou A.
      • Bogacz A.
      • Cross J.H.
      • Elger C.E.
      • et al.
      ILAE official report: a practical clinical definition of epilepsy.
      ,
      • Krumholz A.
      • Wiebe S.
      • Gronseth G.S.
      • Gloss D.S.
      • Sanchez A.M.
      • Kabir A.a.
      • et al.
      Evidence-based guideline: management of an unprovoked first seizure in adults: report of the Guideline Development Subcommittee of the American Academy of Neurology and the American Epilepsy Society.
      ]. Indeed, the literature suggests that the identification of EEG abnormalities (in particular epileptiform abnormalities) can predict seizure recurrence after a first unprovoked seizure. A mathematical model of risk for seizure recurrence found EEG abnormalities to be an important variable in model development [
      • Kim L.G.
      • Johnson T.L.
      • Marson A.G.
      • Chadwick D.W.
      Prediction of risk of seizure recurrence after a single seizure and early epilepsy: further results from the MESS trial.
      ].
      As described in this review, a number of factors have been identified that can improve the yield of recording epileptiform abnormalities. However, at present, the literature is limited with respect to comprehensive investigations assessing combinations of factors to optimize the yield of recording epileptiform abnormalities. The current literature suggests a conceptual paradox, in that yield is increased by increasing samples of EEG recordings, but this is limited by an overall incremental reduction in yield for each individual study as a function of time. Indeed, the literature supports the finding that the highest yield for capturing epileptiform abnormalities on EEG is shortly after the seizure itself, which could be related to the idea that the brain is in a hyper-excitable state in this immediate timeframe. If the initial standard EEG is found to be normal, repeating the EEG can increase the yield for identifying epileptiform abnormalities even though the overall yield for a delayed EEG decreases over time. This temporal decrease in the probability of recording an EEG abnormality can be potentially offset by the inclusion of sleep deprivation with (preferably) or without recording sleep or by increasing the sample period with formal or ambulatory prolonged EEG recordings. It remains to be determined if combining these factors (e.g., immediate admission to an epilepsy monitoring unit for early prolonged EEG monitoring including sleep deprivation and sleep), as speculated by Pohlmann-Eden and Newton [
      • Pohlmann-Eden B.
      • Newton M.
      First seizure: EEG and neuroimaging following an epileptic seizure.
      ], could be a mechanism by which these factors can be combined and optimized.
      Although this review has highlighted ways in which to improve the yield for recording epileptiform discharges on EEG, it must be acknowledged that the interpretation of the EEG remains subjective and is dependent on the reporting electroencephalographer. Whether a finding on an EEG is interpreted as technical artifact, a normal physiological variant or a true abnormality can vary from one electroencephalographer to another. Indeed, a meta-analysis assessing variation in EEG sensitivity and specificity for predicting recurrent seizures identified that the observed wide variation in sensitivity and specificity across studies is largely due to the threshold with which the reporting epileptologist reported a finding as normal or abnormal [
      • Gilbert D.L.
      • Sethuraman G.
      • Kotagal U.
      • Buncher C.R.
      Meta-analysis of EEG test performance shows wide variation among studies.
      ]. This was found for both epileptiform and non-epileptiform abnormalities. In other words, those readers who demonstrated more specificity were better able to predict future seizures than those who read with greater sensitivity. In this way, when the definition of an abnormality was more strictly applied, an EEG abnormality was more accurate in predicting seizure recurrence than when the definition of abnormality was less strictly applied. Thus, a strategy of being more selective with respect to reporting an abnormality, particularly an epileptiform abnormality, could risk a false negative diagnosis but will provide confidence in identifying patients with a truly increased risk of recurrent seizure by minimizing a false positive diagnosis. In this way, this strategy could minimize patients started on anti-epileptic medications to those who are at greatest risk of seizure recurrence.
      Recent guidelines suggest that identifying an abnormality on EEG may warrant treatment after a single unprovoked seizure because the risk of recurrent seizures in the context of an abnormality is sufficiently high [
      • Fisher R.S.
      • Acevedo C.
      • Arzimanoglou A.
      • Bogacz A.
      • Cross J.H.
      • Elger C.E.
      • et al.
      ILAE official report: a practical clinical definition of epilepsy.
      ,
      • Krumholz A.
      • Wiebe S.
      • Gronseth G.S.
      • Gloss D.S.
      • Sanchez A.M.
      • Kabir A.a.
      • et al.
      Evidence-based guideline: management of an unprovoked first seizure in adults: report of the Guideline Development Subcommittee of the American Academy of Neurology and the American Epilepsy Society.
      ]. The assumption is that this is regardless of the time that the abnormality is identified relative to the initial seizure. However, Lawn et al. [
      • Lawn N.
      • Chan J.
      • Lee J.
      • Dunne J.
      Is the first seizure epilepsy – and when?.
      ] argued that the predictive ability of EEG is not static, but rather decreases with time in relationship to clinical seizure freedom. In their retrospective study, the 10-year risk of seizure recurrence after a single unprovoked seizure with epileptiform abnormalities on EEG was initially 76%. In contrast, the 10-year risk of seizure recurrence decreased to less than 60% after 12 months of seizure freedom. In other words, the longer the patient remained seizure free after a first unprovoked seizure, the presence of EEG epileptiform abnormalities became less helpful in predicting the development of recurrent seizures. The impact of this finding on the use of EEG to inform treatment decisions remains to be determined.

      8. Conclusion

      A routine EEG (with or without sleep deprivation) remains a recommended test in the investigation of a single unprovoked seizure for both children [
      • Hirtz D.
      • Ashwal S.
      • Berg A.
      • Bettis D.
      • Camfield C.
      • Camfield P.
      • et al.
      Practice parameter: evaluating a first nonfebrile seizure in children: report of the quality standards subcommittee of the American Academy of Neurology, The Child Neurology Society, and The American Epilepsy Society.
      ] and adults [
      • Krumholz A.
      • Wiebe S.
      • Gronseth G.
      • Shinnar S.
      • Levisohn P.
      • Ting T.
      • et al.
      Practice parameter: evaluating an apparent unprovoked first seizure in adults (an evidence-based review): report of the Quality Standards Subcommittee of the American Academy of Neurology and the American Epilepsy Society.
      ]. The literature suggests that EEG abnormalities, and in particular epileptiform abnormalities, after a new-onset seizure predict seizure recurrence. There is a suggestion that the yield for recording an epileptiform abnormality is highest in the time period shortly after the seizure or with sleep deprivation. It is thought that it is during this time that the brain is in a hyperexcitable state. Although the incremental yield is thought to decrease over time, repeated EEGs and/or prolonged EEGs are thought to improve the yield of recording an epileptiform abnormality due to sampling effect. Further studies will be needed to optimize these factors to continue to improve the yield of recording and identifying epileptiform abnormalities if present, and to improve the sensitivity of EEG after a single unprovoked seizure. This, in conjunction with clinical factors and other ancillary tests (e.g., MRI), will further help inform patients and physicians with respect to risk of seizure recurrence, and hence, ultimately help inform treatment decisions.

      Conflicts of interest

      None declared.

      References

        • Hirtz D.
        • Ashwal S.
        • Berg A.
        • Bettis D.
        • Camfield C.
        • Camfield P.
        • et al.
        Practice parameter: evaluating a first nonfebrile seizure in children: report of the quality standards subcommittee of the American Academy of Neurology, The Child Neurology Society, and The American Epilepsy Society.
        Neurology. 2000; 55: 616-623https://doi.org/10.1212/WNL.55.5.616
        • Krumholz A.
        • Wiebe S.
        • Gronseth G.
        • Shinnar S.
        • Levisohn P.
        • Ting T.
        • et al.
        Practice parameter: evaluating an apparent unprovoked first seizure in adults (an evidence-based review): report of the Quality Standards Subcommittee of the American Academy of Neurology and the American Epilepsy Society.
        Neurology. 2007; 69: 1996-2007https://doi.org/10.1212/01.wnl.0000285084.93652.43
        • Krumholz A.
        • Wiebe S.
        • Gronseth G.S.
        • Gloss D.S.
        • Sanchez A.M.
        • Kabir A.a.
        • et al.
        Evidence-based guideline: management of an unprovoked first seizure in adults: report of the Guideline Development Subcommittee of the American Academy of Neurology and the American Epilepsy Society.
        Neurology. 2015; 84: 1705-1713https://doi.org/10.1212/WNL.0000000000001487
        • Hirtz D.
        • Berg a.
        • Bettis D.
        • Camfield C.
        • Camfield P.
        • Crumrine P.
        • et al.
        Practice parameter: treatment of the child with a first unprovoked seizure: report of the Quality Standards Subcommittee of the American Academy of Neurology and the Practice Committee of the Child Neurology Society.
        Neurology. 2003; 60: 166-175https://doi.org/10.1212/01.WNL.0000033622.27961.B6
        • Hauser W.A.
        • Rich S.S.
        • Lee J.R.
        • Annegers J.F.
        • Anderson V.E.
        Risk of recurrent seizures after two unprovoked seizures.
        N Engl J Med. 1998; 338: 429-434https://doi.org/10.1056/NEJM199802123380704
        • Fisher R.S.
        • Acevedo C.
        • Arzimanoglou A.
        • Bogacz A.
        • Cross J.H.
        • Elger C.E.
        • et al.
        ILAE official report: a practical clinical definition of epilepsy.
        Epilepsia. 2014; 55: 475-482https://doi.org/10.1111/epi.12550
        • Fisher R.S.
        • Boas W.E.
        • Blume W.
        • Elger C.
        • Genton P.
        • Lee P.
        • et al.
        Epileptic seizures and epilepsy: definitions proposed by the International League Against Epilepsy(ILAE) and the International Bureau for Epilepsy(IBE).
        Epilepsia. 2005; 46: 470-472https://doi.org/10.1111/j.1528-1167.2005.00273_1.x
        • Wirrell E.C.
        Prognostic significance of interictal epileptiform discharges in newly diagnosed seizure disorders.
        J Clin Neurophysiol. 2010; 27: 239-248https://doi.org/10.1097/WNP.0b013e3181ea4288
        • Berg A.T.
        • Shinnar S.
        The risk of seizure recurrence following a first unprovoked seizure: a quantitative review.
        Neurology. 1991; 41 (Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/2067659): 965-972
        • Pohlmann-Eden B.
        • Newton M.
        First seizure: EEG and neuroimaging following an epileptic seizure.
        Epilepsia. 2008; 49: 19-25https://doi.org/10.1111/j.1528-1167.2008.01445.x
        • Bouma H.K.
        • Labos C.
        • Gore G.C.
        • Wolfson C.
        • Keezer M.R.
        The diagnostic accuracy of routine electroencephalography after a first unprovoked seizure.
        Eur J Neurol. 2016; 23: 455-463https://doi.org/10.1111/ene.12739
        • King M.A.
        • Newton M.R.
        • Jackson G.D.
        • Fitt G.J.
        • Mitchell L.A.
        • Silvapulle M.J.
        • et al.
        Epileptology of the first-seizure presentation: a clinical, electroencephalographic, and magnetic resonance imaging study of 300 consecutive patients.
        Lancet. 1998; 352: 1007-1011https://doi.org/10.1016/S0140-6736(98)03543-0
        • Hamiwka L.D.
        • Singh N.
        • Niosi J.
        • Wirrell E.C.
        Diagnostic inaccuracy in children referred with “first seizure”: role for a first seizure clinic.
        Epilepsia. 2007; 48: 1062-1066https://doi.org/10.1111/j.1528-1167.2007.01018.x
        • Neufeld M.Y.
        • Chistik V.
        • Vishne T.H.
        • Korczyn A.D.
        The diagnostic aid of routine EEG findings in patients presenting with a presumed first-ever unprovoked seizure.
        Epilepsy Res. 2000; 42: 197-202https://doi.org/10.1016/S0920-1211(00)00183-2
        • Schreiner A.
        • Pohlmann-Eden B.
        Value of the early electroencephalogram after a first unprovoked seizure.
        Clin EEG (Electroencephalogr). 2003; 34 (Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/14521275): 140-144
        • Paliwal P.
        • Wakerley B.R.
        • Yeo L.L.L.
        • Ali K.M.
        • Ibrahim I.
        • Wilder-Smith E.
        • et al.
        Early electroencephalography in patients with Emergency Room diagnoses of suspected new-onset seizures: diagnostic yield and impact on clinical decision-making.
        Seizure. 2015; 31: 22-26https://doi.org/10.1016/j.seizure.2015.06.013
        • Fisch L.
        • Lascano A.M.
        • Vernaz Hegi N.
        • Girardin F.
        • Kapina V.
        • Heydrich L.
        • et al.
        Early specialized care after a first unprovoked epileptic seizure.
        J Neurol. 2016; https://doi.org/10.1007/s00415-016-8272-3
        • Salinsky M.
        • Kanter R.
        • Dasheiff R.M.
        Effectiveness of multiple EEGs in supporting the diagnosis of epilepsy: an operational curve.
        Epilepsia. 1987; 28: 331-334https://doi.org/10.1111/j.1528-1157.1987.tb03652.x
        • Baldin E.
        • Hauser W.A.
        • Buchhalter J.R.
        • Hesdorffer D.C.
        • Ottman R.
        Yield of epileptiform electroencephalogram abnormalities in incident unprovoked seizures: a population-based study.
        Epilepsia. 2014; 55: 1389-1398https://doi.org/10.1111/epi.12720
        • Askamp J.
        • Van Putten M.J.
        Diagnostic decision-making after a first and recurrent seizure in adults.
        Seizure. 2013; 22: 507-511https://doi.org/10.1016/j.seizure.2013.03.012
        • Carpay J.A.
        • de Weerd A.W.
        • Schimsheimer R.J.
        • Stroink H.
        • Brouwer O.F.
        • Peters A.C.
        • et al.
        The diagnostic yield of a second EEG after partial sleep deprivation: a prospective study in children with newly diagnosed seizures.
        Epilepsia. 1997; 38 (Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/9184606): 595-599
        • Shouse M.N.
        • da Silva A.M.
        • Sammaritano M.
        Circadian rhythm, sleep, and epilepsy.
        J Clin Neurophysiol. 1996; 13 (Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/8988284): 32-50
        • Herman S.T.
        • Walczak T.S.
        • Bazil C.W.
        Distribution of partial seizures during the sleep–wake cycle: differences by seizure onset site.
        Neurology. 2001; 56: 1453-1459https://doi.org/10.1212/WNL.56.11.1453
        • Giorgi F.S.
        • Guida M.
        • Caciagli L.
        • Maestri M.
        • Carnicelli L.
        • Bonanni E.
        • et al.
        What is the role for EEG after sleep deprivation in the diagnosis of epilepsy? Issues, controversies, and future directions.
        Neurosci Biobehav Rev. 2014; 47: 533-548https://doi.org/10.1016/j.neubiorev.2014.10.005
        • Delil S.
        • Senel G.B.
        • Demiray D.Y.
        • Yeni N.
        The role of sleep electroencephalography in patients with new onset epilepsy.
        Seizure. 2015; 31: 80-83https://doi.org/10.1016/j.seizure.2015.07.011
        • Michel V.
        • Mazzola L.
        • Lemesle M.
        • Vercueil L.
        Long-term EEG in adults: sleep-deprived EEG (SDE), ambulatory EEG (Amb-EEG) and long-term video-EEG recording (LTVER).
        Neurophysiol Clin. 2015; 45: 47-64https://doi.org/10.1016/j.neucli.2014.11.004
        • 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: 46-51https://doi.org/10.1016/j.seizure.2016.06.005
        • Liporace J.
        • Tatum IV, W.
        • Lee Morris G.
        • French J.
        Clinical utility of sleep-deprived versus computer-assisted ambulatory 16-channel EEG in epilepsy patients: a multi-center study.
        Epilepsy Res. 1998; 32: 357-362https://doi.org/10.1016/S0920-1211(98)00069-2
        • Kim L.G.
        • Johnson T.L.
        • Marson A.G.
        • Chadwick D.W.
        Prediction of risk of seizure recurrence after a single seizure and early epilepsy: further results from the MESS trial.
        Lancet Neurol. 2006; 5: 317-322https://doi.org/10.1016/j.neuroscience.2005.08.052
        • Gilbert D.L.
        • Sethuraman G.
        • Kotagal U.
        • Buncher C.R.
        Meta-analysis of EEG test performance shows wide variation among studies.
        Neurology. 2002; 60: 564-570https://doi.org/10.1212/01.WNL.0000044152.79316.27
        • Lawn N.
        • Chan J.
        • Lee J.
        • Dunne J.
        Is the first seizure epilepsy – and when?.
        Epilepsia. 2015; 56: 1425-1431https://doi.org/10.1111/epi.13093