Attitudes to the treatment of schizophrenia-spectrum disorders, often now simply called ‘psychoses’, are changing. Whereas until recently Department of Health policy documents on mental health were focused on the treatment of people ‘with severe and enduring mental illness’, now we read exhortations to provide early intervention. For example, the recently published National Service Framework for Mental Health states:

In a supplement of the British Journal of Psychiatry, ‘Verging on Reality’, edited by McGorry, from which is taken our first quote, recent thinking on early intervention is well represented (Reference McGorryMcGorry, 1998). The key issues that emerge are: early detection; the concept of a critical period; social and emotional problems; cognitive therapy approaches; and multi-component early intervention.

In this editorial we will discuss some of this new thinking about psychosis, describe our experience of developing cognitive therapy for early psychosis and consider the implications for models of service provision.

Early detection

Studies of first-episode psychosis show that the average time between onset of symptoms and first effective treatment is often one year or more (Reference McGlashanMcGlashan, 1998). For at least two reasons, this long duration of untreated psychosis is undesirable. First, the person developing psychosis is at risk of serious consequences occurring from changes in mental state and behaviour, such as loss of job or disruption to important relationships and self-harm or suicide. Second, there is evidence that delayed treatment leads to poorer long-term outcome, even when attempts have been made to control for those factors which may confound delay and long-term morbidity (Reference Wyatt, Damiani and HenterWyatt et al, 1998). Larsen et al (Reference Larsen, Johannessen and Opjordsmoen1998) investigated the factors associated with a long duration of untreated psychosis in a group of consecutively admitted people with first episodes of psychosis in Norway. They found that those with a long delay to treatment (mean over four years) more frequently had a premorbid deteriorating course, were socially withdrawn with poorer social networks and were less likely to be in work. Frequently, attempts to initiate treatment had been made early, but failed and were not followed up, and key people (such as relatives and teachers) were not informed about the nature of the illness. Larsen et al are, therefore, currently engaged on a major programme incorporating both educating professionals about early detection of psychosis and a public information campaign to seek to change the pathways to care for people with an emerging psychosis. There can be no doubt that this work is relevant in a British context. We recently conducted a small survey of our local service in inner-city London. In over 70% of cases of first-episode psychosis, the formal recognition and initiation of treatment by the mental health service came only after a crisis had occurred, resulting in emergency involuntary admission to hospital, frequently following a long duration of untreated psychosis (Reference Garety and RiggGarety & Rigg, 1998).

Critical period

A view that has been gaining ground is that the early phase of psychosis has a major influence on the long-term outcome of the disorder, and should be seen as a ‘critical period’ with implications for the secondary prevention of the impairments and disabilities which accompany psychosis (Reference Birchwood, Todd and JacksonBirchwood et al, 1998). Researchers have found that the greatest deterioration in cognitive and social functioning occurs early in the course of the illness, reaching a plateau within two years. Birchwood et al argue that the data suggest that deterioration, although variable, does occur in the period before the psychosis occurs and early in the course of psychosis (treated and untreated), but that this often stabilises after two to five years. They conclude that intervention targeted in the early years after onset is likely to have a disproportionate impact relative to interventions later in the course.

Social and emotional problems

The social and emotional problems of early psychosis are increasingly being recognised. The loss of social roles may precede, accompany and be a consequence of an episode of psychosis, and poor social functioning at first episode predicts poorer social functioning later (Reference Birchwood, Todd and JacksonBirchwood et al, 1998). Depression and anxiety are common in people with psychosis and have been well documented (Reference Addington, Addington and PattenAddington et al, 1998; Reference Freeman and GaretyFreeman & Garety, 1999). Despite these facts, very little is typically offered by services as part of early treatment to reduce the social and emotional impact of psychosis. Relevant interventions, such as employment support programmes, are often targeted at those with ‘severe and enduring mental illness’. Even if programmes focused on social functioning are made available to people with early psychosis, they may prove unattractive to them, being perceived as meeting the needs of an older, more disabled group. Treatment for emotional problems is similarly weak and has been almost entirely restricted to the prescription of antidepressants in addition to neuroleptic medication, although some services may also offer community psychiatric nurse support.

Cognitive therapy for early psychosis

Given the problems and needs of those presenting early in the course of a psychotic illness and the importance of early intervention, the potential benefit of the development of cognitive interventions for this group is clear. The efficacy of cognitive therapy delivered adjunctively to standard care in reducing distress and symptomatology for those with medication-resistant psychotic symptoms is now relatively well established (Kuipers et al, Reference Kuipers, Garety and Fowler1997, Reference Kuipers, Fowler and Garety1998; Reference Tarrier, Yusupoff and KinneyTarrier et al, 1998; Reference Garety, Fowler and KuipersGarety et al, 2000). Cognitive therapy for early psychosis has broadly adopted these protocols, usually comprising: an assessment and engagement period; reformulating and developing an individualised model of psychosis; addressing residual psychotic symptoms from a cognitive perspective; work on dysfunctional assumptions about self, others and the world; and relapse prevention. However, the specific problems of those facing psychotic illness for the first time suggests a shift in focus is necessary.

In our recent pilot work we have placed relatively greater emphasis on the person's understanding and appraisal of the experience of psychosis, particularly in terms of the view of the self, and the potential impact on future plans and relationships. Consideration of the implications of having had a psychotic episode in terms of engagement with other appropriate services and treatments is also a key component. Jackson & Birchwood (Reference Jackson and Birchwood1996) highlighted the prevalence of post-traumatic type symptomatology following a first episode of psychosis, and of secondary depression, and we have therefore integrated emotional processing work into our approach, and maintained the emphasis on affective disturbance. We have offered some brief family contact, although surprisingly few of those we have seen have been living with or in close contact with family members.

There is little published at present in the way of evaluation of cognitive therapy for early psychosis. Therapeutic emphasis and goals have varied greatly across studies, as has the duration, intensity and timing of therapy. More detailed evaluation of what works for whom is needed. This should particularly address the work of McGlashan (Reference McGlashan1987) and Birchwood and colleagues (Reference Drayton, Birchwood and TrowerDrayton et al, 1998). They postulate a continuum of recovery style following a psychotic episode, ranging from ‘integration’ to ‘sealing over’. They also suggest that a more integrative style may respond better to a therapeutic emphasis on discussing and processing psychotic and emotional experience, while evidence of sealing over may indicate a more behavioural, coping-oriented intervention. Nevertheless, existing evidence suggests cognitive therapy is an appropriate and beneficial treatment for this group. Drury et al (Reference Drury, Birchwood and Cochrane1996) showed that intensive group and individual cognitive therapy with brief family intervention reduced positive symptoms and recovery time when targeted at those presenting in the acute phase of psychotic illness, about a third of whom were experiencing their first episode, with gains maintained at follow-up.

A similar intensive, in-patient focus on the acute phase of the illness has been adopted by Tarrier et al in a recently completed large multi-centre trial (SOCRATES). The therapy targets symptom management, and early results appear to suggest more rapid reductions in symptoms for the cognitive therapy group, compared to supportive psychotherapy and standard care, but the full analysis is awaited. Jackson et al (Reference Jackson, McGorry and Edwards1998) have recently completed a pilot evaluation of cognitively oriented psychotherapy for early psychosis (COPE) in the context of their specialist early psychosis service. They have targeted those early in the illness, but worked over a longer time period on adjustment to illness and the impact of the experience of psychotic illness on the self, showing promising improvements in clinical functioning.

We have generally engaged with people during their in-patient stay but the bulk of therapy has occurred on an out-patient basis, lasting up to six months (18 sessions), although indications from our ongoing study are that for those who engage and benefit from the work, sessions lasting up to a year may be indicated. This is work in progress, but early indications of acceptability of the therapy to service users once engaged, and of positive outcome, are encouraging. Our thinking is that this longer term approach, bridging the transition from hospital to community, and facilitating adjustment to the experience of psychosis and re-engagement in social networks, addresses more of the specific problems of those with early psychosis, and is likely to promote improved long-term outcome.

Service provision: multi-component early intervention

The implications of the issues highlighted here are evident. There are strong grounds for attempting to detect psychosis early and to intervene before deterioration in clinical and social functioning occurs. Relapse prevention is an important goal. The emotional impact of psychosis should also be addressed. A service response is needed which combines best practice in pharmacological interventions, family work, vocational and social programmes and cognitive approaches. Models of good practice can be seen in Birmingham, where an early intervention service under the leadership of Max Birchwood has been developed and in Australia, where Patrick McGorry has developed the pioneering Early Psychosis Prevention and Intervention Centre (EPPIC) service with his colleagues (Reference McGorry, Edwards and MihalopoulosMcGorry et al, 1996). Both provide specialised early psychosis services, with a strong youth focus, multi-component interventions, and follow-up over at least two years, in a single service. There are encouraging preliminary data from these services, but as yet, no randomised controlled trials for such an approach.

In Lambeth, a multicultural, inner-city London borough, the South London and Maudsley Trust is developing a new early intervention service, which is being evaluated in a randomised controlled trial. The project, partially funded by the King's Fund, will combine dedicated in-patient beds in a ward for first service contact patients with a newly established assertive outreach team based in the community. Staff have specialist skills in vocational training and placement, substance misuse, cognitive therapy, family work, multicultural work and pharmacological treatment. Service users are involved in planning and delivering the service. It aims to provide rapid access to assessments in community settings, good links with a wide range of community agencies, assertive follow-up and a full range of evidence-based interventions. We also hope to undertake community information and education programmes to facilitate early detection. The evaluation will, we hope, provide important pointers to whether our enthusiasm for early intervention is justified.