Antisocial behaviour is the most common reason for referral to child mental health services. It is also a clinical problem of considerable importance, because there is a marked tendency for it to persist, and the long-term outcome includes antisocial personality disorder and criminality. Furthermore, effective treatments are now available, although not yet widely used in Britain.

Those who work with older adults will inevitably find themselves working with families. Indeed assessment of an older person is often difficult if members of their family and social network are not involved. Referrals to social and psychiatric services are often precipitated by family changes. Ratna & Davis (1984) described 142 consecutive referrals to a community old age psychiatry service and found retirement, family conflict, departure of or illness of a carer, or bereavement precipitated 60% of referrals. Thus, work with older adults will often necessitate work with families, although it may not be seen in those terms.

The number of beds available in the UK for the treatment of mental illness has fallen by over two-thirds since its peak in the mid-1950s. There is a worldwide trend of closure of large mental hospitals and integration of services for the mentally ill in their own local community. Community surveys have shown that an overwhelming majority of mentally ill people are treated in primary care or by specialists, but without being admitted to a hospital. This is now the preferred clinical practice (Shepherd et al, 1981).

This study is presented to encourage recognition of the contribution mental health practitioners can make to the care of head-injured people.

The most taxing behavioural complications of severe head injury for relatives include irritability and aggression, as illustrated by J. B. Longitudinal studies show an increase over time in the frequency and severity of temper outbursts (Brooks et al, 1986; Hall et al, 1994), which take the form of explosive or impulsive aggression. I will comment on the nature and treatment of impulsive aggression as illustrated in case J. B., emphasising the interaction of neurobiology with motivational, situational and personality factors.

It has been estimated that every year 8 per 100 000 of the population suffer severe head injury, 18 per 100 000 suffer moderate head injury and 250–300 per 100 000 suffer mild head injury (Medical Disability Society, 1988). It is likely that 100–150 per 100 000 of the population have significant disability as the result of head injury (Department of Health, 1994).

Mental health professionals regard the assessment of suicide risk as part of normal clinical practice, but most shy away from assessment of risk of interpersonal violence, either doubting that such predictions are possible, or, if possible, are part of their role. In part, this fastidiousness reflects the influence of a consensus which emerged in the 1970s to the effect that there was no substantial relationship between mental illness and violent behaviour. The only basis for risk assessments which remained were those factors in the general population known to be associated with criminality. Even here, however, the accuracy of predictions made by clinicians had been subject to considerable critical scrutiny, which undermined any claims the profession may have had to expertise (Monahan, 1981; Brizer, 1989). There seems to have been wide acceptance that psychiatrists were doubly disabled in the business of predicting violence by having no relevant specialist knowledge and by being poor at applying knowledge gleaned from other disciplines such as criminology.

The problems associated with alcohol and drug misuse are diverse and make a major impact on the health of the nation (Anderson, 1991; Royal College of Psychiatrists, 1992). A recent survey showed that the prevalence of alcohol dependence in the population (aged 16–64 years) was 47 in 1000 and of drug dependence 22 in 1000. (This compares with a prevalence for functional psychosis of 4 in 1000) (Office of Population Censuses and Surveys, 1994).

The importance of personality is recognised tacitly by the development of the multiaxial classifications of ICD–10 (World Health Organization, 1992) and DSM–IV (American Psychiatric Association, 1993). The separation of axis 1 or current mental state diagnosis from axis 2, on which personality is described, is a recognition that personality is separate from other aspects of the patient's diagnostic status. Nevertheless, many clinicians are sceptical that personality can be assessed reliably and some hold that the diagnosis should be abandoned as being merely judgemental and pejorative (Lewis & Appleby, 1988). There is some basis for scepticism in respect of reliability in view of the evidence that even with operational criteria the clinical assessment of personality is not transportable, and the level of agreement between practising clinicians is much less than that achieved during field trials (Mellsop et al, 1982).