The truth,and nothing but the truth,about early intervention in psychosis

Introduction

Mental illness remains a major source of burden to individuals, families and, more broadly, society. The so-called ‘severe’ mental illnesses, schizophrenia and related disorders and bipolar disorder (henceforward ‘psychotic illnesses’), are particularly disabling. The recently completed Australian Study of High Impact Psychoses (SHIP) (Morgan et al., 2011) reinforces the enduring symptoms, social exclusion and physical health problems that so many people with such conditions endure. In the hope that these outcomes can be avoided or at least ameliorated, substantial investment has been made into establishing early intervention services for psychotic disorders, both within Australia and abroad. However, good evidence that such services really make a difference to patients in the long term is lacking. Furthermore, the expansion of such services has been mirrored by a reduction in other services for people with a mental illness, notably access to psychological services under the Better Access initiative. This article, an extension of earlier critiques (Bosanac et al., 2010; Castle, 2011), asks whether early intervention should be such a major focus in mental health reform. However, first, I pose a question about ‘caseness’.

What is a psychosis ‘case’ and what are the implications for outcomes?

The epidemiology of psychotic disorders has been well established, with the careful World Health Organization (WHO) multinational studies of psychosis showing incidence rates of around 7–14 per 100,000 per year (Jablensky et al., 1992). Yet, in the era of early intervention, rates of psychosis have increased dramatically, such that in the UK one centre reports over 50 incident cases per 100,000 per year (Cheng et al., 2011) and for the Early Psychosis Prevention and Intervention Centre (EPPIC) catchment in north-west Melbourne, I have estimated rates to be of the order of 100 per 100,000 per year (Castle, 2011). These increases in psychosis incidence raise the question of who is actually being treated, and whether they actually need treatment. The early intervention advocates state these are all help-seeking individuals, but help-seeking is a behaviour, not a diagnosis, and the treatment evidence base rests on diagnosis. This large increase in incident cases also raises questions about the potential for labelling as pathological what are probably normal developmental processes and/or expressions of a degree of ‘psychosis proneness’ that is arguably normally distributed in the population (van Os et al., 2009). The potential for unnecessary exposure of such individuals to antipsychotic medications is of major concern, particularly given the evidence that these agents are associated with adverse general health outcomes (particularly metabolic problems) and potential brain grey and white matter loss (Ho et al., 2000). Also of concern are the service implications, as scarce mental health dollars are diverted away from those with chronic illnesses.

Another issue is that with the inclusion of such milder forms of psychosis as ‘cases’ in early intervention services, the outcome data from such services would be expected to be rather good. That is to say, the early intervention advocates suggest that they have taken a cleaver to Kraepelin and proven him wrong about the inevitably poor outcomes associated with schizophrenia. This is disingenuous at best, as it is quite clear that Kraepelin’s dementia praecox defined a severe early-onset form of illness that is only a subgroup of what we today consider ‘schizophrenia’. What is also clear is that for as long as long-term outcome studies have been conducted, many patients with schizophrenia have a very good outcome (Shepherd et al., 1989).

Finally, and most worryingly, is the ultra-high risk (UHR) experiment. Whether this is to be included in current service proposals is unclear, and indeed the early psychoses field itself seems confused about the whole issue. In any event, the UHR criteria proposed by the EPPIC service in Melbourne are very heterogeneous, including various combinations of psychotic symptoms, psychosocial decline and family history. It is thus hardly surprising that the positive predictive value of these criteria is very poor: even EPPIC is reporting now that fewer than 10% of such individuals actually ‘transition’ to a psychotic illness, irrespective of whether treated or not (Yung et al., 2011). As Addington et al. (2011) have shown, those individuals who do not ‘transition’ show improvement in symptoms and functioning over time. The whole treatment issue is also fraught, with very real concerns regarding exposure of UHR individuals to antipsychotic medications in particular (see above).

Are early intervention services associated with enhanced longer-term outcomes?

There are serious questions regarding whether early intervention actually improves the course of illness. The promise was that schizophrenia could be prevented, or at least that poor outcomes could be ameliorated. No published study has supported this assertion. For example, Robinson et al. (1999) from the USA reported cumulative relapse rates for 104 early psychosis patients to be 82% for a first episode and 78% for a second episode, over 5 years. The Lambeth Early Onset (LEO) study in London found benefits from their specialist intervention in terms of hospitalization rates and vocational and social functioning while the service was being provided, but there was a rapid increase in readmissions after the end of the specialist intervention (Gafoor et al., 2010). Similarly, the Danish OPUS study found that 2-year benefits in terms of improved psychotic symptoms and reduced substance use largely dissipated by 5 years (Bertelsen et al., 2008). From the EPPIC service in Melbourne, only 14.9% of non-affective psychosis patients showed remission in terms of psychosocial and symptomatic domains at 7 years, despite the inclusion of schizophreniform psychoses (Henry et al., 2010) and from Sweden, Bodén et al. (2010) reported that modified assertive community treatment (mACT) for early psychosis did not deliver better 5-year outcomes across multiple domains.

In contrast, Norman et al. (2011) from Canada found ongoing benefits from a high-fidelity early intervention, when a lower-intensity specialized intervention was offered after the first 2 years. This reinforces that all that the early psychosis programmes have shown is that good care is good for people but only works while it is being delivered. Of course this has been very widely acknowledged for decades. In particular, the assertive community treatment (ACT) literature is much larger than that for early intervention, and shows, according to a comprehensive review, ‘positive effects … in reducing hospital use and increasing patient and family satisfaction’ (Burns and Santos, 1995) while the 1998 Cochrane review concludes: ‘ACT is a clinically effective approach to managing the care of severely mentally ill people in the community … Policy makers, clinicians, and consumers should support the setting up of ACT teams’ (Marshall and Lockwood, 1998). By contrast, the 2011 Cochrane review of early intervention for psychosis states that there is ‘some support for specialised early intervention services, but further trials are desirable, and there is a question of whether gains are maintained’ (Marshall and Lockwood, 2011).

Another approach has been specifically to target the so-called ‘duration of untreated psychosis’ (DUP), which is the period of active psychotic symptoms antedating initial treatment. Longer DUP is associated with worse outcomes, but this may largely be due to the inherent poorer prognosis of people with an indolent type of illness onset. Also, DUP is notoriously difficult to measure accurately, especially when it is insidious: in cases with a well-defined and more abrupt onset, the illness is arguably of a more benign type. In any event, attempts to shorten DUP have been expensive and have reaped few rewards. Certainly early psychosis services alone and general practitioner campaigns have largely failed to reduce DUP (Lloyd-Evans et al., 2011). Large-scale public awareness campaigns have been shown to have a modest effect on DUP, but whether this actually makes any clinically significant impact on outcomes is arguable (Castle, 2011).

Why just young people?

Most early psychosis programmes target young people and the youth focus has been reflected in recent policy announcements. While the early psychosis advocates continue to assert that the majority of mental illness onsets in youth, they do not highlight the fact that in at least a third of cases the onset is over 25 (Castle et al., 1998), a fact reinforced in the SHIP study (Morgan et al., 2011). These later-onset cases have, like younger people, particular developmental issues to face in the context of a psychotic illness, not least dealing with partners and children. Their exclusion from early intervention seems inequitable.

Why stand-alone services?

Many early psychosis practitioners claim that stand-alone services are critical to maintain the integrity of what they do, and that joining with or being embedded within mainstream services would perturb their fidelity. Indeed, a leading group of experts in the field (McGorry et al., 2010) have labelled generic services ‘pessimistic’ and implied that they are responsible for the fact that patients do not do as well once the intensive intervention ceases.

However, no comparison of stand-alone and integrated services has, to my knowledge, been performed. Furthermore, we have shown that a high-fidelity early psychosis service can be embedded in generic mental health services (Petrakis et al., 2011). As I have suggested elsewhere (Castle, 2011), there are significant problems associated with stand-alone services, including silo effects, the potential deskilling of the general mental health workforce in the area of early psychosis and the difficultly of transitions between services for patients, their families and clinicians. Friis (2010) has also raised the ‘loss’ experienced by the patient on transition from early psychosis intervention services: surely the response to that is to create services that look after people for as long as such care is required, which is what happens in generic services.

What are the alternatives?

I am not, as some of the early psychosis advocates would have it, against early intervention or for ‘late intervention’ in psychosis. What I do believe is that the current focus on youth-focussed early intervention silos is ill-considered and will have long-term damaging consequences for our mental health services. Despite what the early intervention advocates say, I am not remotely alone in this view: a recent poll showed 60% of Australian psychiatrists ‘disagree’ or ‘strongly disagree’ that the models proposed are ‘appropriate’, while only 12% ‘strongly agree’ with them (Slezak, 2011). I would like to see an integrated model of care that has no silos, no age cut-offs and no impediment to individuals with psychotic illnesses receiving the very best of care as early as and for as long as they need, in a continuum of care model. I would also like to see a much more considered, bipartisan, whole-of-government approach to mental health, rather than it being as driven by the political cycles and informed by a few influential individuals, as it seems to be. As Carr has stated on this matter: ‘mental health care is broader than that and much too important to sell it so far short of what it deserves’ (Carr, 2010).