Early intervention: Mission cramp versus mission creep?

Early intervention: an evidence-creating pillar of reform that works and saves money

Early diagnosis is so important because the earlier a mental illness can be detected, diagnosed and treatment can begin, the better off that person can be for the rest of his or her life. (Rosalynn Carter)

EI, previously missing in action in mental health care, has become one of the major frontiers for progress over the past quarter of a century. While its seeds were sown in the unpromising soil of schizophrenia, soil contaminated by a century of unremitting pessimism, and its growth has ebbed and flowed within a climate of innovation, scepticism and inertia and been retarded by a background of grossly inadequate investment in mental health care as a whole, EI has nevertheless flourished. We now know that it is possible to ‘bend the curve’ of recovery temporarily through EI and significantly improve symptomatic and functional outcomes in the early stages of psychotic illness, both prior to and following the first episode of psychosis (Riecher-Rossler and McGorry, 2016). If we invested more strongly in securing and safeguarding these early gains beyond 2 years post diagnosis, then there is no a priori reason they could not be maintained, and several studies are looking at this. EI has attracted a level of controversy not seen in cancer or cardiovascular medicine, where the premium placed on early diagnosis in no way detracts from later stage and palliative care. The serious underfunding of the latter in psychiatry has understandably clouded the thinking of some critics and caused them to question EI as a priority. Yet the fact is that few critics really question the logic or value of intervening early, unless they also doubt the value of mental health care more fundamentally. Debates have centred rather on whether it was even possible with the present state of knowledge, and if so, the optimal way to go about it. Hundreds of real world demonstrations around the world and a substantial body of Level 1 Cochrane evidence have now accumulated to show it is increasingly possible to improve short-term symptomatic and functional outcome and also suicide rates, while saving a lot of money. Hence, governments all over the world are moving to implement EI, with strong consumer and family support. In the recent Federal Election, we witnessed the Federal Government restate its general commitment to EI and stepped care and to specialised programmes in early psychosis in particular. The latter had been at risk. However, how to deliver EI remains a legitimate empirical question, and there are many issues that need to be considered. Age is clearly one of them.

Ageism in early intervention?

Early psychosis services everywhere have placed an upper limit on the maximum age of entry, based on a raft of previous incidence studies and even dating back to the original Kraepelinian concept of Dementia Praecox, which was fundamentally based on its typical onset in young people. However, data from the ÆSOP study (Lappin et al., 2016) show a pattern of incidence, which seems to depart significantly from this consensus and indeed many other landmark studies. This can be partly explained by the fact that clinical phenotypes that contain frank psychosis and allow a diagnosis of schizophrenia emerge late in the course of disorder, and treated incidence for psychosis in traditional hospital-based systems, prior to EI reforms, are a very crude marker of onset. We know that a genuine need for care presents much earlier, so the question arises as to when the true onset is. Also, if we consider the weight of all published incidence and age of onset data (Kessler et al., 2007), it seems clear that the incidence of schizophrenia and psychosis peaks in the third decade of life, with the median onset in the early 20s, as confirmed by all the major birth cohort studies, so the extent to which later onsets occur may not be so marked as suggested in the AESOP study (median age of 29), which may be a relative outlier. Indeed, Peter Jones, one of the co-authors of Lappin et al. (2016) and a leading epidemiologist, stated elsewhere (Jones, 2013) that in schizophrenia,

The earliest signs of any mental disturbance occurred before puberty, but psychotic symptoms and index admissions for the full syndrome were confined to the decade and half after puberty … These findings have been repeatedly replicated, and represent some of the most robust findings in psychiatric epidemiology.

While there is an apparent partial disconnect between the AESOP findings and the remainder of the literature, the key point made by Lappin et al. (2016) that provision of EIfor a subgroup (the size of whichmay vary and can be debated) of patients older than 25 or even 35 must be addressed, remains a valid one, particularly since they also appear to suffer from delayed access and outcomes that are equally poor as in younger patients, and which could almost certainly be improved by more intensive and sustained EI-style care than traditional, unfocused services can offer. The question is how to do it.

When we began our focus on first episode psychosis in Australia in the 1984, our original age range was 16–45 years. A minority, yet finite number, of cases presented in the age range 30–45 years, but the bulk were emerging adults under 30. When Early Psychosis Prevention and Intervention Centre (EPPIC) was established in 1992, we reduced the age range to 15–30 years to allow more of the younger patients’ access across a larger catchment. A decade later when we moved to a broader EI focus on other emerging disorders, as Orygen and headspace came into being, the upper age limit fell further to 25 years, based on the best available transdiagnostic national epidemiological data and an analysis of developmental needs and youth culture. We felt that the developmental, family and cultural elements were as just as critical as stage of illness in determining where to draw boundaries for service provision; 12–25 years seemed to be the best fit in trying to capture the onset phase for the greatest number of cases within a culturally appropriate, emerging adult framework. It also provided the opportunity to provide care very much earlier for people whose onsets were in this period of transition to adulthood, but who otherwise would not access care until much later. Originally, and we have been proved at least partly wrong here, we did not anticipate that a focus on only one group of disorders, namely the psychoses, would be sufficient to drive a widespread structural reform. We believed that a broader EI paradigm blended with the developmental paradigm of transition to adulthood would ultimately be more sustainable (McGorry, 1996). This is the road down which Australia’s national reforms since 2004 have moved, and they have involved a stepped care approach, beginning with the enhanced primary care gateway of headspace and with some belated investment in a linked secondary care EI focus for psychosis. The latter now needs to be scaled up fully and complemented by specialised programmes for other major diagnostic streams, notably complex mood disorders, personality disorders, eating disorders and substance use disorders. Implementation remains the challenge.

Yet for the 25% of cases of mental disorders with a new onset over 25, many of whom will have a psychotic illness it seems, it is quite true as Lappin et al. (2016) argue, that we need to address their needs for EI in a different way. This means tackling delay in access, addressing a quite different set of needs to other patients of similar age but different stage of illness, though perhaps from a better base of developmental achievement than those with a younger onset. While there is no ideal time to develop a serious mental illness, one might expect that it could be less devastating if there are ‘more runs on the board’. Conversely, issues of threatened or real loss, rather than lost potential may need to be faced. This is a challenge for policy-makers and mainstream adult services to consider.

So we face some choices. One would be to limit the focus to EI to psychosis, where most of the current evidence resides, keep building early psychosis services as part of the specialist mental health system, but lift the upper age limit and allow people with an onset of psychosis at any age to enter. This would mean privileging the EI principle over developmental, family and cultural needs. So should we prioritise stage or age? we always believed mixing emerging adults with the middle aged, especially the severely disabled with long-term illnesses, was not appropriate, and this error would be compounded if the relatively rare prepubertal psychoses and the late onset middle and older age people were all treated within a ‘one size fits all’ EI system. I recall such mixing of age groups early in my career, and it was chaotic and harmful.

Second, notwithstanding the data from AESOP, 75% of all mental disorders in Australia do appear before age 25 even if they don’t access care until later. So a multi-tiered focus on services for children and emerging adults up to around 25 will deliver the bulk of EI. At least two subcultures are needed here – prepubertal up to around 12 years and pubertal to mature adulthood (perhaps closer to 30 rather than 25). Age-based boundaries should, however, be porous and fluid. The remaining EI goal then would be to create a space or stream of care for genuine new onsets over 25 alongside people of the same age with illnesses whose onset was earlier but which have persisted or recurred. This could be done both in older adult mental health services and old age psychiatry.

Mission creep and transdiagnostic early intervention

How can you govern a country which has 246 varieties of cheese? (Charles de Gaulle)

The Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5) has over 500 varieties and 151 defined categories of mental disorder, and simply for reasons of credibility and utility, we need to reduce this to many fewer dimensions or categories to govern ourselves in psychiatry. Conversely, the mission of EI must be explored beyond the safe haven of psychosis if it is to achieve its potential in mental health care. There is a major trend towards transdiagnostic paradigms that recognise the arbitrary nature of traditional diagnostic boundaries in psychiatry and the fluidity of syndromes during onset. This supports the youth mental health focus, and 75% of onsets will be captured this way. The Government, through the National Mental Health Commission, and crucially, during the election period, recommitted strongly to EI, stepped care and specifically to the building blocks for this; namely headspace, fully funded early psychosis services and new models, yet to be developed, to respond to other forms of potentially severe mental illness in young people. However, as the past two decades have taught us, only reform that is guided by scientific evidence and enables such evidence to be assembled as it proceeds, will really work, and only reform that receives widespread community and political support will survive and flourish.