The misdiagnosis of epilepsy

The case of Dr Andrew Holton, consultant paediatrician at Leicester Royal Infirmary, highlights once again some of the dangers and pitfalls in the diagnosis and management of epilepsy. He has been suspended and referred to the General Medical Council after a review of 214 children seen by him showed that 171 gave definite or possible “cause for concern.” Just over a third of the children were not thought to have had epilepsy, and just under a third were thought to have been overtreated.¹ Both are common pitfalls in the management of epilepsy.

The review also made clear that Dr Holton's training fell well short of what would be required for his post. Although a consultant in paediatrics, Dr Holton was not a paediatric neurologist, of whom there are just 62 in the United Kingdom. The report also points to professional isolation and under-resourcing as important mitigating factors in Dr Holton's practice. This episode graphically illustrates the potential consequences of the shortcomings identified in 2000 by the Clinical Standards Advisory Group in its report on epilepsy services in the United Kingdom.²

Though the diagnosis of seizures and epilepsy can be straightforward, it can also be one of the greatest clinical challenges. An adequate diagnosis requires differentiation between seizures and other causes of transient neurological disturbance and collapse; differentiation between acute symptomatic and unprovoked truly epileptic seizures; and, in people with epilepsy, classification of the disorder and identification of the cause so as to optimise treatment.

The symptoms of epileptic seizures are varied and there are many imitators, ranging from convulsive syncope through to psychogenic events. The phenomenon most commonly mistaken for a convulsive seizure is syncope. Classical teaching is that collapse in syncope is flaccid and that no motor activity occurs.³ Commonly occurring motor and ocular phenomena are not widely recognised. Lempert et al induced syncope in 42 healthy volunteers, 90% of whom experienced myoclonus, usually multifocal.⁴ Additional features such as head turning, oral movements, or attempts to sit up occurred in 80%. These motor phenomena are often taken to indicate that a seizure has occurred. Similar convulsive episodes seen immediately after concussive head injury may also be mistaken for epileptic phenomena.⁵

Neurogenic syncope is provoked and involves brief loss of consciousness and rapid recovery. A detailed history will usually be all that is required by an experienced clinician to differentiate this from seizures. Unfortunately, in inexpert hands inappropriate investigation often takes precedence. Fainting is probably the single commonest reason for requesting an electroencephalogram, which in 20% of the population will reveal non specific abnormalities open to misinterpretation.⁶ Given that most requests emanate from non-specialist settings and most electroencephalograms are reported by neurophysiologists without great experience of epilepsy and its management, there is considerable potential for misdiagnosing faints as seizures.

Cardiac syncope often causes immediate loss of consciousness, tonic stiffening of body and limbs, and often myoclonic jerking. Testing with a tilt table and electrocardiographic monitoring may be needed to identify a specific cause and avoid the erroneous diagnosis of epilepsy.⁷,⁸

A population based study in Rochester, Minnesota, indicates that 40% of incident seizures are symptomatic of an acute systemic or cerebral disturbance rather than epilepsy.⁹ Fever in children and alcohol withdrawal in adults are the commonest causes of such acute symptomatic seizures. Recognition prevents unnecessary antiepileptic drug treatment and avoidance of the provocative factor, or treatment of the underlying cause, and permits early resumption of driving.

Dr Holton's experience is probably no more than the tip of the iceberg. Smith et al reported on 184 consecutive patients referred to one consultant neurologist.¹⁰ Forty six were believed to have been misdiagnosed, of whom 12 were referred with drug resistant epilepsy. Nineteen had experienced side effects of drugs, 12 had unnecessary driving restrictions, and five had serious employment problems with three dependent on state benefits. Zaidi et al, using standard laboratory tests, found that 13 of 36 (31%) patients with continued attacks despite antiepileptic drugs had treatable cardiovascular conditions.⁸ The direct costs to the NHS due to fruitless outpatient visits and unnecessary investigations and drugs are difficult to quantify. However, if one applies the figures from the cost of illness study by Cockerell et al to only those misdiagnosed patients with apparently drug resistant epilepsy the total cost of misdiagnosis is estimated to be £125m a year.¹¹

While the rate of misdiagnosis by itself is a strong argument for specialist care of epilepsy, the advances and increased choice of pharmacological and other treatments further support the case. We can no longer accept the delivery of epilepsy care from an unsupported general physician or paediatrician. The report of the Clinical Standards Advisory Group on epilepsy services recommends structured multidisciplinary specialist care for people with epilepsy with an appropriate interface with primary care.² The poor quality of epilepsy services and the preventable mortality of epilepsy have recently been highlighted in the chief medical officer's annual report and commitments made to address these issues.¹² It must now be evident that the human and financial costs of failing to implement the recommendations are too high to be acceptable.