• epilepsy;
  • intellectual disability guidelines consensus


  1. Top of page
  2. Abstract
  3. Introduction
  4. Method
  5. Key points
  6. Conclusion
  7. References

Background  Epilepsy has a pervasive impact on the lives of people with intellectual disability and their carers. The delivery of high-quality care is impacted on by the complexity and diversity of epilepsy in this population. This article presents the results of a consensus clinical guideline process.

Results  A Delphi process identified a list of priority areas for the development of evidence-based guidelines. All guidelines were graded and consensus on scoring was achieved across the guideline group.

Conclusion  There is a dearth of high-quality evidence from well-constructed studies on which to base guidance. However, the development of internationally derived consensus guidelines may further support the management of epilepsy in adults with an intellectual disability.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Method
  5. Key points
  6. Conclusion
  7. References

Adults with an intellectual disability (ID) comprise a significant proportion of the prevalent epilepsy population (Lhatoo & Sander 2001). Moreover, the association between ID and refractory epilepsy means that they compose a major portion of the ‘hard to treat’ epilepsy population.

The population with ID offers certain unique challenges to the epilepsy specialist as a result of the complex aetiology and commonly associated behaviour and communication disorders (Bowley & Kerr 2000).

In recognition of these special clinical issues, guidelines were produced in 2001 highlighting key management issues (IASSID Guidelines Group 2001). Since their publication there has been an increase in evidence-based guidance for people with epilepsy. In the UK, the Scottish Intercollegiate Guidelines Network (S.I.G.N. 2003) and the National Institute for Health and Clinical Excellence (NICE) (Clinical Guideline 20 2004) guidelines have produced detailed guidance, as have the American Neurology Association in North America and the ILAE (Glauser et al. 2006). These have greatly clarified treatment approaches to people with epilepsy in general; however, there remains a dearth of guidance on the complex clinical areas particularly pertinent to adults with ID.

In order to address this need, we set out to establish pragmatic evidence-based guidance for key issues in the management of epilepsy in people with ID. This paper describes the process and the clinical practice points.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Method
  5. Key points
  6. Conclusion
  7. References

The guidelines were established through evidence-based modified Delphi process, a process successfully employed in the production of other epilepsy guidelines (Lux & Osborne 2004). Professionals with a special interest and experience in the management of epilepsy in people with ID were identified and invited to participate through the International Society for the Scientific Study of Intellectual Disability (IASSID). The group were individuals who had contributed to the previous guidelines and others who had shown an interest in contributing through the special interest health group of IASSID. Fifteen individuals from six countries participated throughout the Delphi Process. The individuals were comprised from adult neurology, nursing, paediatric neurology, psychiatry and physicians with an interest in the care of adults with an ID.

All experts were sent a detailed evidence base derived from a Cochrane review of interventions for epilepsy in people with ID. The search method for this can be seen in the two Cochrane reviews (Beavis et al. 2007a,b). These reviews identified, with a full electronic and hand search methodology, evidence for interventions both pharmacological and nonpharmacological. The participating experts were sent a CD-Rom and associated literature containing the abstracts from the review; including all interventional trials identified. Each expert was then invited to submit up to 10 areas where it was felt new clinical guidance statements were necessary. An initial sample of the 30 key clinical areas was identified through compilation of the submissions from the experts (Table 1). A further mailing of the group was undertaken on which the experts were asked to rate, ranking 1 to 10, the 10 areas of most importance from the 30. These priority data were analysed and a large separation in ranking score was seen after the 11th area, thus 11 priority areas were put forward for the development of guidelines (Table 2).

Table 1.  Clinical areas of relevance
  1. EEG, electroencephalogram; MRI, Magnetic Resonance Image; NICE, the National Institute for Health and Clinical Excellence; AED, anti-epileptic drug.

Down's syndrome and epilepsy
Psychogenic non-epileptic attack disorder
Patient and carer education & communication
Sleep disturbance and seizures
Access, referral requirement and preparation for EEG and MRI
Links and similarities with other published guidelines e.g. ILAE/NICE or other country guidelines
Differential diagnosis of epilepsy
Treatment approach to refractory seizures
Managing drug side effects and potential switch from ‘sedating’ AEDs
Drug interactions
Referral procedures for epilepsy surgery
Assessment of treatment emergent behavioural difficulties
Syndrome-dependant treatment in adults e.g. Dravets
Bone health issues
Use of rescue medication
Impact of comorbidity
Impact of associated feeding difficulties
Autism and epilepsy
Assessing the impact of seizures
Impact of AED on behaviour/cognition
Risk assessment e.g. drowning/sudep?
Impact of psychotropic and other medication on seizure threshold
When to discontinue AEDs in seizure free patients
Monotherapy choice
Indication for genetic investigation
Which specialist should treat people with ID and epilepsy
Investigation of first seizure
Drug-induced psychosis
Assessment of quality of life
Seizure disorder in the elderly
Table 2.  Prioritised guidelines areas
  1. AED, anti-epileptic; EEG, electroencephalogram; MRI, Magnetic Resonance Image.

 1. Treatment approach to refractory seizures
 2. Impact of AED on behaviour/cognition
 3. Managing drug side effects and potential switch from ‘sedating’ AEDs
 4. Assessment of treatment emergent behavioural difficulties
 5. Differential diagnosis of epilepsy
 6. Drug interactions
 7. Autism and epilepsy
 8. Impact of comorbidity
 9. Referral procedures for epilepsy surgery
10. Access, referral requirement and preparation for EEG and MRI
11. Down's syndrome and epilepsy

Individuals within the group were asked to champion the development of draft statements on the basis of their particular expertise. These were then sent to group members for comment and were scored and graded as per the evidence-based tables (Tables 3,4). Grading was by mode score. The final statements were resent to the group to confirm agreement on the grading by evidence level before the process was closed.

Table 3.  Grading of evidence
Level of evidenceType of evidence
  1. RCT, Randomised Controlled Trial.

IaEvidence from systematic reviews or meta-analysis of RCTs
IIbEvidence from at least one RCT
IIaEvidence from at least one controlled study without randomisation
IIbEvidence from at least one other type of quasi experimental study
IIIEvidence from nonexperimental descriptive studies, such as comparative studies, correlation studies and case–control studies
IVEvidence from expert committee reports or opinions and/or clinical experience of respected authorities
Table 4.  Grading of recommendations
Level of recommendationType of recommendation
ABased on hierarchy I evidence
BBased on hierarchy II evidence or extrapolated from hierarchy I evidence
CBased on hierarchy III evidence or extrapolated from hierarchy I or II evidence
DDirectly based on hierarchy IV evidence or extrapolated from hierarchy I, II or III evidence

Key points

  1. Top of page
  2. Abstract
  3. Introduction
  4. Method
  5. Key points
  6. Conclusion
  7. References

Key points from the guidelines include the following.

Differential diagnosis of epilepsy in people with intellectual disability

In the group of people who have epilepsy and ID inappropriate or misdiagnosis is prevented by following an appropriate diagnostic pathway. (Grade C)

An eyewitness account of the seizure, with an accurate description of the suspected seizure, is needed in every case. This description should be recorded soon after the event; however, an exact description may often be too challenging for even an experienced observer when simultaneously having to care for the patient. A simple video recording of seizures (increasingly possible with modern mobile phones) can be a valuable tool and is particularly useful where frequent seizures occur. (Grade D)

An assessment of the patient's capacity to participate in treatment decisions is always necessary. It is important to try to evaluate the patient's experience of the seizure and to ascertain whether there are warning signs and precipitating or perpetuating factors in the development of seizures. Where such a history is not available the clinician often has to depend on relatives, carers or other professionals involved in trying to infer this relevant information. It is recognised that proxy informants such as care providers can alter greatly from family in judging, for example, the severity of the epilepsy or the impact on an individual (Grade D)

A careful assessment of potential seizure provoking factors should be made: fever, infection, stress, excessive waking, alcohol withdrawal, hyperventilation, some medications, sudden discontinuation of sedative drugs and specific activities (in the case of activity induced seizures) should be taken into account. (Grade C)

Non-convulsive epileptic phenomena and even partial seizures may be difficult to diagnose in people with ID. This may lead to an under-diagnosis of these particular seizure types. (Grade C)

Attributing a diagnosis of epilepsy to non-epileptic events in people with ID is a significant problem. Sudden aggression, motor stereotypies and vacant staring are the most important reasons for the over-diagnosis of epilepsy, and consequently also for inappropriate use of anti-epileptic (AED) medication. (Grade C)

Sleep disorders are often present in people with ID who have difficult-to-treat epilepsy and behavioural problems (Dorris et al. 2008). It is important to understand the sleep disorder, categorise the problem, investigate appropriately and then come to a diagnosis. It is important that treatment of sleep disorder should be available for this group as it is a significant quality of life issue. (Grade C)

Successful electroencephalogram (EEG) recording needs the patient to co-operate and possibly sleep during the investigation. It is helpful, if possible, for prior acquaintance with the facilities to be made, in order for the patient to be less anxious and to maximise the opportunity of each appointment. Some patients may need melatonin or chloral hydrate (though National Licensing may restrict their use), and these do not affect the EEG findings. (Grade C)

During the recording the clinician should be sensitive to the patient's reactions and proceed gently. Prolonged video-EEG monitoring is of use in order to distinguish between epileptic and non-epileptic seizures and in selecting candidates for epilepsy surgery. If this investigation is not available, portable cassette recording of the EEG may also be of considerable value. (Grade C)

Magnetic Resonance Image (MRI) scanning of the brain is important in the diagnosis of conditions such as: malformations due to abnormal neuronal migration (lissencephaly, heterotopia), malformations due to abnormal cortical organisation (polymicrogyria, schizencephaly, cortical dysplasia), neoplastic lesions (e.g. ganglioglioma, dysembryoplastic neuroepithelial tumour, hamartoma of the hypothalamus), malformations of posterior fossa structures, ischaemic and haemorrhagic lesions (of the newborn), and neurometabolic disorders/leucoencephalopathies. If MRI is not available, Computed Tomography (CT) is recommended. Criteria for investigation should be agreed locally for people with epilepsy. Some people with ID may need anaesthesia for MRI scanning; this bears another set of risks that should be noted but should not mean exclusion from investigation. Brain CT, with contrast, is particularly useful in emergency situations. (Grade C)


Refractory epilepsy

There is no unifying definition of refractory epilepsy. Often, seizures are referred to as being refractory or treatment resistant when three or more AEDs have failed (Kwan & Brodie 2000). Selection of the appropriate drug for a given individual must be based on an understanding of the individual, each drug's pharmacology, side effect profile, interaction and risks. For the individual with ID, the impact of AEDs on their cognition, behaviour and mobility should be particularly considered. (Grade C)

There is no Level I evidence comparing new AEDs with the old, or the new AEDs with each other in patients with refractory epilepsy. Studies of AEDs are often short-term placebo controlled trials that are designed to meet the requirements of drug regulatory authorities rather than inform clinical decisions. (Grade C)

On the basis of a recent Cochrane review (Beavis et al. 2007b), there is no evidence from randomised controlled trials that informs drug choices for refractory epilepsy in people with ID. Drugs should be chosen on the basis of the patients' seizure type, seizure syndrome, and aetiology and being cogniscent of the side effect profile of the drug. In general this evidence will come from trials in non-ID populations. (Grade C)

Drug side effects

In the general population with epilepsy, individual AEDs show both positive and negative effects on cognition. There is insufficient evidence of these effects in people with ID. The evidence from the general population may not simply be extrapolated to people with ID because of the confounding impact of pre-existing cognitive disability, the difficulty of recognising and identifying causes of drowsiness, and communication difficulties in assessing drug impact on their quality of life. It is therefore recommended that the impact of the drug in the general population be carefully considered and reflected in clinician choice (Grade C).

New prescription of phenobarbitone (Phenobarbital) (de Silva et al. 1996) is discouraged because of the high incidence of behavioural side effects in paediatric studies. However, it is recognised that the data are based on paediatric studies and that in some countries fiscal constraints mean it is a drug of choice. Any use should be closely monitored for behavioural effects. In individual cases, the drug may be used as a third-line AED if other, more suitable options have been used without success. (Grade B)

In a placebo-controlled study of topiramate as add-on therapy in patients with ID no significant behaviour changes between groups were noted using a validated measure of behaviour change. (Kerr et al. 2005) (Grade B)

No recommendation can be given for a specific drug of choice in patients with epilepsy and ID. Similarly, there is no drug that would have to be avoided in those patients for its cognitive or behavioural side effects. (Grade B)

Baseline cognitive and behavioural assessments should be made and then remeasured after drug changes, preferably by using validated measures (Espie et al. 1997;Kerr & Espie 1997), before introducing a new drug and on follow-up. Such vigilance enables the identification of changes in cognition and/or behaviour. (Grade C)

It is important to recognise that the majority of patients in this population are unable to self-report side effects. So, subtle effects (e.g. double-vision) may remain undetected and only overt ones (e.g. vomiting) may come to the attention of doctors, care-givers and relatives. Non-specific symptoms including unwillingness to take medication or behavioural disturbances may be due to – otherwise undetected – side effects such as nausea. (Grade D)

It may be difficult to distinguish long-term side effects from the consequences of underlying or accompanying diseases versus an AED influence. For example, the development of osteoporosis is not necessarily attributable solely or partially as a side effect of drug therapy but may be at least partly due to diet (low D-vitamin serum levels) or impaired mobility as a consequence of physical disability. (Grade D)

Drug interactions

Because of the possibility of drug interactions, complicated by the patients' difficulties in communicating their symptoms, AED treatment should be re-evaluated whenever individuals have a change in their non-AED medication (Perucca 2006). (Grade C)

Patients with the highest risk of drug interactions are patients with comorbid disorders such as psychiatric illness, HIV/AIDS, cardiovascular disease, or cancer, patients with difficult-to-control or refractory epilepsy and geriatric patients. (Grade C)

Physicians should be aware of what drugs the patient is taking, including non-AEDs, over-the-counter medications, and supplemental nutrients and understand the underlying metabolic pathways of these agents. Just because a medication is ‘over-the-counter’, does not mean that it is necessarily devoid of interactions. (Grade C)

Consideration of the measurement of concentrations of an AED, if available, should be made at the time a patient presents with a clinically significant possible dose-related adverse event. The clinical assessment of the patient is critical and often more important than measuring the drug level. Several individual clinical situations, such as pregnancy, may merit specific drug level monitoring (Patsalos et al. 2008). (Grade C)

Patients on phenytoin are an exception and need regular, at least yearly, serum drug concentration measurement. Drug monitoring must, however, be combined with clinical examination for clinical signs of side effects (e.g. Nystagmus). (Grade B)

Predicting AED drug interactions can assist in quickly and efficiently achieving and maintaining therapeutic target levels. The potential for drug interactions is not a contraindication. The key is recognising that an interaction is possible, then taking steps to prevent or minimise complications from occurring. (Grade C)

When patients are on phenytoin or carbamazepine in combination with valproate an assay of free, unbound, phenytoin or carbamazepine may be a better reflection of potential toxicity as these AEDs have high protein binding. (Grade C)


An initial clear description of any behavioural presentation should be made. In particular, the type of behaviour, precipitants and past history of similar behaviour should be recorded. (Grade D)

Intercurrent physical illness (including exclusion of a metabolic cause of ID) should be excluded as a cause of behavioural disturbance irrespective of the apparent association with the timing of drug change. In particular, direct questioning as to whether the symptoms could be explained by gastro-oesophageal reflux disease which can cause sleep disturbance and appetite change, should be made. (Grade D)

The majority of problem behaviours seen in people with ID represent longstanding disability-related factors such as level of ability, communication difficulties, environmental issues or the presence of autistic traits. Clinicians should clearly communicate this to carers and ensure that the patient is receiving the maximum social support. (Grade C)

Where the behaviour has occurred in association with significant seizure reduction a consideration that the patient is experiencing so-called ‘forced normalisation’ (Mula & Sander 2007) should be made. ‘Forced normalisation’ is the concept that behaviour deterioration occurs as the EEG normalises but in clinical practice often refers to seizures stopping, or significantly reducing, without corroborative EEG evidence. There is no trial data to guide response to this situation when it occurs. Identification of the phenomenon, behavioural support to the patient and family, treatment of the emergent psychiatric condition and a risk assessment of a return of seizures are all potential elements of the clinical response. (Grade D)

Clinical psychiatric assessment should be considered, as the behaviour may be the presentation of a treatable condition such as anxiety or depression. (Grade D)

Clinical psychological assessment should be considered both to identify the causation of the behaviour and to offer therapeutic strategies. (Grade D)

Impact of comorbidity

The co-occurrence of epilepsy with other common physical and psychological conditions is common. All initial epilepsy assessments should identify these comorbid conditions as they may impact on treatment decision-making and treatment outcome assessment. (Grade C)

Psychiatric comorbidity, depression, anxiety, behaviour problems, autism and attention deficit and hyperactivity disorder are all relatively common in this population (Smiley et al. 2007). The diagnoses of these conditions should be sound. Such confirmed comorbidities should be thoroughly treated optimising both nonpharmacological and pharmacological therapies. In the case where psychopharmacological interventions are deemed necessary, the risk for seizure worsening is in most cases small and needs to be balanced against the impact of chronic untreated psychological disorder. (Grade C)

Referral procedures for epilepsy surgery

The presence of ID is not a contraindication to neurosurgery for epilepsy. All patient fulfilling referral criteria should, on the basis of individualised assessment, be referred to appropriate specialist services. (Grade C)

A key determinant in an individuals' ability to progress through to neurosurgery is to meet the criteria for neurosurgical procedure:

  • 1
    Diagnosis of epilepsy must be ascertained;
  • 2
    An adequate trial of drug therapy; that is, the correct drugs used in the correct dosage, carefully monitored for an appropriate time; and
  • 3
    The electroclinical syndrome must be defined. (Grade C)

If a patient is believed to be suitable for surgery an MRI scan is essential. MRI facilities able to provide anaesthetic cover should be available. (Grade B)

A syndromal approach to management

Lennox–Gastaut syndrome

No comparative evidence exists for the treatment of adults with seizures in Lennox–Gastaut syndrome. Level I evidence does exist for the impact of lamotrigine (Motte et al. 1997) and topiramate (Sachdeo et al. 1999) on drop attacks. (Grade A)

Rufinamide (Glauser et al. 2008) has also been shown in a Level I study to have an impact on drop attacks in Lennox–Gastaut syndrome; however, it has only recently been released and clinical experience, to date, is limited. (Grade A)

Level I evidence exists for felbamate (The Felbamate Study Group in Lennox-Gastaut Syndrome 1993); however, it is rarely used because of serious fatal side effects. Its use is only recommended in specialist centres. (Grade A)

Other medications, which can be effective in generalised seizure disorders, and which can be used in Lennox–Gastaut syndrome, include sodium valproate, levetiracetam and zonisamide. (Grade C)

It is not uncommon to see patients with Lennox–Gastaut syndrome on a range of therapy for both partial and generalised seizures. Cautious drug reduction may be tried for any drug that appears not to have worked in such patients in order to establish whether mono or duo therapy may prove as successful as multiple drug therapy, which is often associated with significant side effects in people with ID. (Grade C)

Down's syndrome and epilepsy

The treatment of epilepsy in people with DS follows general guidelines. The drug treatment is planned according the type of epilepsy or seizures. In dementia-linked epilepsy, broad-spectrum drugs (valproate, levetiracetam) should be used as first-line drugs. If using valproate, keep the potential risk for chronic valproate encephalopathy, which may resemble a dementia-like clinical picture, in mind. (Grade D)

All adults with DS who present with epilepsy over the age of 30 should be screened for the presence of a coexisting dementia, usually of an Alzheimer's type. (Grade C)

Autism and epilepsy

Extreme diagnostic caution is needed in assessing whether behaviour associated with autism, such as an outburst of behaviour disturbance or repeated stereotypies might be due to epilepsy. A video of the events and ictal EEG are sometimes needed. In the absence of these tests decisions should only be made by a clinician expert in assessing the epilepsy diagnosis in patients with ID and autism with behavioural concerns. (Grade D)

Assessing the impact of treatment change in individuals with autism is particularly difficult. A baseline assessment of behaviour and physical status is regarded as essential. (Grade D)

There is no quality evidence to suggest that treatment with AEDs has an impact on autistic traits. (Grade D)


  1. Top of page
  2. Abstract
  3. Introduction
  4. Method
  5. Key points
  6. Conclusion
  7. References

The management of epilepsy in people with an ID demands high professional standards across a range of clinical domains. This guideline has shown that consensus can be achieved in these areas by a range of professionals from differing countries. Unfortunately the quality of the available evidence is often very limited. Several key areas remain poorly served. These include, surprisingly, pharmacotherapy despite the huge reliance on such medication in this population. Additionally there is little information to support clinical decision-making in areas of risk; this is most clear in issues relating to behavioural change such as forced normalisation where clinical decisions are made in an essentially evidence-free setting. There is a mandate for those who fund as well as those who deliver health care to improve these gaps in our knowledge.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Method
  5. Key points
  6. Conclusion
  7. References