Few neurologists doubt that a significant proportion of the patients who consult them have a psychiatric disorder of at least moderate severity. It is well established that conditions such as epilepsy, Parkinson's disease, multiple sclerosis, and cerebrovascular disease are associated with an increased vulnerability to anxiety disorders, affective disorders, and psychoses.1 The predisposition probably results both from the functional disability associated with the neurological condition and also from disturbance of intracerebral neurotransmitter pathways. In addition to those with established organic disease neurologists are consulted by a considerable number of somatising patients, those with little or no organic pathology but who have various neurological symptoms masking an underlying psychiatric disorder.2

The paper by Carson et al 3 on pages 202–206 of this volume indicates the extent of this phenomenon in general neurological outpatient practice and confirms previous studies in this area. Only a minority of all patients with an emotional disorder were considered by their general practitioner or neurologist to require some form of psychological or psychiatric treatment. The exception was the group with unexplained symptoms, in over half of whom the neurologist thought that such treatment was required.

In a previous survey of neurological inpatients the prevalence of psychiatric morbidity, estimated by a standardised two stage assessment, was 39%, of which 72% was not recognised by the neurologists.4 Patients were divided on whether they wished to discuss their mood with the clinician. Fifty five per cent thought that such an inquiry would have been unhelpful. Reasons given included the lack of privacy of hospital wards and the inability to see the consultant without the presence of an accompanying entourage of junior doctors and other professional staff. However, few patients with an emotional disorder expressed a need for psychiatric or psychological treatment, raising the question of what is the most appropriate management for their problems.

Neurologists, in keeping with other clinicians, must by now be bored by critical comments from psychiatrists that they fail to recognise psychiatric disorders in their patients. The study of Carsonet al indicates that even when the disorders are recognised, patients are reluctant to embark on any form of psychological treatment. They resist psychiatric referral for various reasons, including the perceived stigma of psychiatric illness and also because of their conviction that their symptoms are due to a physical rather than a psychological process. It is likely that for most patients, such as those identified in this study, management of their psychiatric problems will have to be undertaken jointly by the neurologist and general practitioner. This particularly applies to those patients whose emotional disorders coexist with an underlying neurological disorder. The fledgling specialty of liaison psychiatry does not have the resources to deal with the numbers of patients Carsonet al have identified. For practical reasons only those patients with complex and treatment resistant conditions should be referred. The psychiatrist needs to have a close working relationship with neurological colleagues and particular experience of assessing patients with neurological symptoms.

Patients who somatise are much more likely to accept psychiatric treatment if an attempt is made to modify the way they perceive their symptoms. The simple techniques of cognitive behaviour therapy can help the patient reattribute the symptoms and recognise them as being linked with an underlying emotional problem which requires appropriate intervention. It has been demonstrated that the techniques of reattribution can be taught to general practitioners5 and when applied in practice they can reduce the levels of psychopathology. The acquisition of cognitive behaviour therapy skills will facilitate the management of such patients in the general practice surgery and neurology clinic.