Drs Pelosi and Birchwood (Reference Pelosi and Birchwood2003) have provided some stimulating thoughts about the implementation of early intervention for psychosis. Perhaps one of the underlying difficulties that may lead to the dichotomy of views expressed by the two authors is a confusion about what constitutes ‘early intervention’. Pelosi rightly identified both the lack of evidence and theoretical restriction in clinical usefulness based on the epidemiology of schizophrenia and the sensitivity and specificity of screening for the disease. It seems reasonable to question the widespread and costly implementation of a service based on such shaky evidence.

However, there is a sharp contrast between the concept of early intervention as a service aimed at secondary prevention, with treatment in prodromal phases of schizophrenia, and the way in which it is defined in the UK Government's Mental Health Policy Implementation Guide (Department of Health, 2001). Here, it is clear that the service should primarily be focused on interventions in people who have already developed psychotic symptoms, with various broad-ranging strategies to ensure early identification and referral and good links with employment and education institutions ensuring a high-quality and holistic service.

None of this is rocket science and the argument that it could be provided by existing community mental health teams might seem attractive were it not for the failure over many years of existing teams to truly address these issues. Experience from other areas of health care, such as cancer services, suggests that specialisation often leads to improvements in quality of services and the same might be expected within the context of early intervention for psychosis.

Early intervention provides an opportunity for significant improvements in the way in which young people with devastating illnesses are managed, and it is essential that psychiatrists lend the full weight of their experience and expertise to ensuring the success of these teams.