Electroencephalogram after first unprovoked seizure in children: Routine, unnecessary or case specific.

A child is brought into a paediatric emergency unit with an unprovoked, afebrile first seizure. We conduct a clinical assessment of the child and rule out any acute metabolic, traumatic or infectious causes and consequently, make the diagnosis of an epileptic seizure. The International League against Epilepsy (ILAE) suggests that following such a diagnosis, the next step should be the appropriate classification of the seizure type, after which an appropriate syndrome diagnosis should be made. (1) Should an EEG be arranged for this child and if so, should it be arranged within 24 hours or within the next week? If we decide not to arrange an EEG this time and to do so if any further seizures occur, are we practicing evidence based medicine? A recent guideline published by the Royal College of paediatrics and child health (RCPCH) asserted: "There is no need for an EEG following a first simple afebrile seizure". (2) This is a very bold and clear statement but what evidence and what quality of evidence is this statement based upon? This review analyses and discusses prominent literature regarding this widely-discussed topic.

Introduction

A child is brought into a pediatric emergency unit with an unprovoked, afebrile first seizure. We conduct a clinical assessment of the child and rule out any acute metabolic, traumatic or infectious causes and consequently, make the diagnosis of an epileptic seizure. The International League against Epilepsy suggests that following such a diagnosis, the next step should be the appropriate classification of the seizure type, after which an appropriate syndrome diagnosis should be made.[1]

Should an electroencephalogram (EEG) be arranged for this child and if so, should it be arranged within 24 h or within the next week? If we decide not to arrange an EEG this time and to do so if any further seizures occur, are we practicing evidence based medicine? A recent guideline published by the Royal College of pediatrics and child health (RCPCH) asserted: “There is no need for an EEG following a first simple afebrile seizure.”[2] This is a very bold and clear statement however, what evidence and what quality of evidence is this statement based upon?

This review analyses and discusses prominent literature regarding this widely-discussed topic.

EEG: The Evidence for Its Use After a First, Unprovoked Seizure

Whilst literature on this controversial subject matter is very diverse, a large proportion of the debate surrounds the utility of the EEG for the counselling of parents/family following a first seizure; particularly as to whether it is a viable tool to predict seizure recurrence.

One study conducted by Stroink et al. aimed to assess the accuracy of the diagnosis of a first unprovoked seizure in childhood, the recurrence rate within 2 years, the risk factors for recurrence and the long-term outcome 2 years after recurrence.[3] A panel of three neurologists confirmed the diagnosis of seizure in children based on pre-described diagnostic criteria. All the children deemed as having a seizure had a standard EEG, if it did not disclose an epileptiform abnormality, a second EEG was performed after partial sleep deprivation, or during daytime sleep. One of the most interesting results was the recurrence rate of 71% in children who had an epileptiform EEGs in standard EEG.[3] They concluded by stating that an EEG with epileptiform abnormalities proved to be the main risk factor for recurrence. However, we must consider whether having this information after the first seizure would change our practice; it must be noted that a significant proportion (29%) of children with epileptiform EEGs did not have a recurrence.[3]

Shinnar et al. also conducted a study that assessed EEG abnormalities in this patient cohort.[4] The authors, in their prospective cohort study on 347 children with a first, unprovoked afebrile seizure showed that children with an idiopathic first seizure who had an abnormal EEG were at an increased risk of seizure recurrence. 54% of those with epileptiform EEGs experienced a second seizure, in contrast to 25% recurrences in those who exhibited a normal EEG.[4] They concluded that an abnormal EEG, particularly an epileptiform abnormality, is associated with a higher risk of seizure recurrence in children with an idiopathic first seizure.[4] Whilst Shinnar et al., have clearly demonstrated the higher risk of seizure recurrence in children with interictal EEG abnormalities, it is essential that we remember that a significant proportion of children with an abnormal EEG had no seizure recurrence. Although, an abnormal EEG would increase the risk of recurrence by nearly two folds, whether this has to be included in counseling patients and families is debatable, as an abnormal EEG is not an absolute predictor of seizure recurrence, and a normal EEG may be falsely reassuring.

A review of literature conducted by Pohlmann-Eden et al. on the use of the EEG and neuro-imaging following an epileptic seizure also focused on the potential interpretations and limitations of the EEG following a first seizure.[5] They reviewed literature (children and adults) on seizure recurrence, timing of EEG, yield of abnormalities on EEG, yield and timing of sleep deprived EEG and focal epileptiform activity in relation to recurrence. They concluded that an early abnormal EEG especially, when showing focal epileptiform activity seems to be an excellent predictor for seizure recurrence.[5]

Would a physician be able to effectively counsel children and carers based on the percentages identified in the studies appraised above and more importantly, is having a recurrence rate for any type of seizure contributory to the management or counselling of patients? This remains unclear.

Another argument put forward by advocates of the use of EEG-after first seizure is its application as a prognostic marker. Dr Panayiotopolous, from the department of clinical neurophysiology and epilepsy at St. Thomas′s Hospital, recommended that a revision of the current practice in pediatrics of not requesting an EEG after the first afebrile seizure may be needed. He argued that the current practice of treating after two seizures is not in keeping with the best-available evidence.[6] He commented that an abnormal EEG, particularly generalized spike wave discharges have been reported as a consistent predictor of recurrence in nearly all studies. He discussed the benefits of having an EEG after first seizure; these include diagnosis of a specific epileptic syndrome, making a diagnosis of benign childhood partial seizures (as these children may not have more than a single episode) and also assessing the diagnostic and management applications of minor seizures (such as absences and myoclonic jerks).[6] He clearly defined the advantages of an EEG after the first seizure, stating that the child and the family are entitled to a diagnosis, prognosis and management that is specific and precise, even though this may only be possible in a select-proportion of patients after the first seizure.[6]

EEG: The Evidence Against Its Use After a First, Unprovoked Seizure

In his recommendations, Panayiotopoulos also commented that the EEG was a harmless investigation; this point brought disagreement from pediatric neurologist, Richard Appleton. He emphasized that although the EEG was harmless with regards to its invasiveness, it could be potentially harmful in terms of its interpretation.[7] Due to the variability in the quality of EEG reports and interpretation, there may be a significant number of inaccurate diagnoses, which could consequently lead to serious medical, psychological and social consequences.[7] He also added that irrespective of the EEG changes, many children would not be treated after a first seizure owing to the high non-recurrence rates after first seizures.[7] He has also pointed towards the fact that most children are managed outside tertiary epilepsy centers and may not have access to medical personnel who should be undertaking and reporting EEGs.[7] The author has clearly highlighted very important issues that would be of immense practical importance and would also initiate thinking in these under-emphasized areas of this contentious topic.

Some of the pitfalls of pediatric EEG interpretations were also discussed by Tan et al.; their article was written in response to the clinical guidelines published by the National Institute of Clinical Excellence (NICE) in 2004.[8] They stated that the EEG has rarely been able to function as a “black or white” or “yes or no” diagnostic test for epilepsy, whilst also commenting that the normal EEG evolves and matures with time, reaching the normal adult pattern by approximately 12 years of age.[8] They proceeded to describe EEG reading as an “inexact science,” which is learnt through experience and stated that the accuracy of EEG interpretation relies heavily on the clinical history provided by the physician, which on many occasions is insufficient. They discuss the fact that in the United Kingdom, EEGs are reported by different medical practitioners including neurophysiologists, neurologists (both pediatric and adult), psychiatrists and even general practitioners with a special interest in the EEG.[8] Many of these practitioners will have had very little exposure to children and pediatric epilepsy, thus, it is difficult to monitor the quality of EEG reporting. The authors have clearly highlighted the varied quality of EEG reports in various settings and therefore, this has to be considered when using the EEG to guide management of an individual case, especially, if the EEG has been acquired after a first unprovoked seizure.

Another argument that has been put forward against the routine use of the EEG following the first seizure is whether or not it will contribute to diagnosis or management. Gilbert et al. conducted a MedLine search from 1980 to 1998 to quantify and analyze the value of expected information from an EEG after first unprovoked seizure in childhood.[9] They employed an evidence-based analytical approach to published literature to select studies providing reasonable estimates of the value of EEG in the general population of children with first unprovoked seizure. Next, the information from these studies was quantified using linear information theory and their results showed that the likelihood of making a clinically useful diagnosis by performing an EEG in every child after first seizure was low.[9] In addition to this, they commented that as a result of the low sensitivity and specificity of the EEG, many false predictions could occur, thus, resulting in unnecessary patient distress or reassurance.[9] The quality of expected information from the EEG is too low to affect treatment recommendations in most patients, thus, they concluded by stating, EEG should be ordered selectively, not routinely, after first unprovoked seizure in childhood.[9]

Their conclusion was countered by Berg et al. who stated that Gilbert et al. focused on the impact of EEG on “treatment decisions” and neglected much of the additional value of an EEG.[10] He argued that an EEG following the first seizure may provide valuable diagnostic and prognostic information despite epilepsy being a clinical diagnosis.[10] Furthermore, an abnormal EEG in an appropriate clinical context may help support the diagnosis of a seizure disorder, whilst also helping to distinguish between partial, and generalized seizures.[10] In their study, Gilbert and Buncher also made reference to a study conducted by King et al. on the utility of EEG for diagnosis following first seizure, which stated that differentiating partial from generalized seizures was non-specific; Berg et al. argued that this “non-specific” distinction has fundamental treatment implications if recurrent seizure occur, as the first line anti-epileptic drugs are different for partial and generalized epilepsy.[9-11] However, it is important to note that although differentiating partial and generalized epilepsy may have treatment implications (it is not common practice to treat after first seizures) these are not specific syndromic diagnoses and thus, having this information may not contribute to the first counseling the child or carers receive.

Conclusion

The EEG has a very important role in the investigations and classifications of epilepsies. Even though EEG is pivotal, epilepsy remains primarily a clinical diagnosis. Both NICE and Scottish Intercollegiate Guidelines Network in their guidelines acknowledged that epilepsy is primarily a clinical diagnosis and that, in isolation, the EEG should not be used to make a diagnosis of epilepsy.

At present, the majority of pediatricians do not request an EEG after a first seizure and this practice has been emphasized at a recent RCPCH guideline appraisal. EEG carried out at this point has a low sensitivity and specificity for recurrence rates, which may lead to false predictions. Therefore, the information on recurrence may not be helpful in counseling; however, an abnormal EEG would increase the risk of recurrence by nearly two folds. There are also a significant proportion of children with first unprovoked seizure and abnormal EEG who may not have any further seizures. Therefore, including the information in counseling may lead to unnecessary psychological stress for both children and carers as the finding of an abnormal EEG does not imply that the child will invariably have a further seizure. Most parents would opt against giving their children anti-epileptic medications after a first seizure, provided they are given the right information that it is possible that there may be no recurrence. The other aspect to consider is the quality of EEG interpretation that a physician receives to base his judgment on and as mentioned above, quality control of the EEG is very difficult to regulate.

On the other hand, there are several clinical situations in which conducting an EEG after a single seizure can lead to the diagnosis of specific epileptic syndrome. Secondly, minor seizures such as absences and myoclonic jerks could be picked up, both for which prognostic and treatment implications could be deduced. Children and their carers are entitled to a specific diagnosis, a prognosis, and a precise management plan. Therefore, we need to consider these factors when we counsel them.

Recommendations

Diagnosis of a seizure, epilepsy type and an epileptic syndrome should be primarily based on clinical grounds. An EEG should not be routinely requested without a purpose. If it is requested, then it should be to answer a specific question, which in turn would help to aid the diagnosis of epilepsy. The quality of the EEG report and its interpretation has to be considered, and if in doubt, further specialized advice should be sought.

Recommendations for further research

As there is limited literature on EEGs after first seizure in varying age groups, conducting a prospective study to evaluate the EEG after first seizure is needed. In particular, to assess the influence of the EEG results on treatment decisions, social and emotional implications and seizure recurrence.