The case for early, medium and late intervention in psychosis

As McGorry et al point out, the model for early intervention in psychosis draws on physical illness (typically cancer), where the idea is that early detection leads to treatment that is less radical, more successful and averts a poor or fatal outcome. Unfortunately in psychosis there is neither an early nor a specific biological marker, so that early intervention is really not early at all, but closer to secondary prevention where symptoms are already present, even if they are not yet severe. This means that all prodromal services can do is offer treatments to help seekers, up to 80% of whom will never make the transition. Prodromal services, by definition, do not offer help to those who deny they have problems and who may be at the more severe end of the spectrum with longer duration of untreated psychosis (DUP), more negative symptoms and poorer out-comes after an episode. Similarly, early intervention services can only offer help to those who will stay engaged.

Thus, the early intervention medical model is not correct for psychosis; those treated, or those who will accept treatment, by definition are unlikely to be those who will need it most. This is the first difficulty that services face, and until more specific markers are discovered, it will remain a stumbling block to the hope of preventing episodes or of offering comprehensive services to everyone at risk of an episode: a true early intervention model.

Of course, there are humanitarian reasons for offering services early; these are mainly to reduce the DUP, associated with a poorer response to antipsychotic medication 1, and the sometimes brutal and shocking realities of sectioning and admission that individuals can face if problems are left until a crisis. Offering a service that people collaborate with and take up before crises develop is entirely laudable. However, we have no evidence yet, apart from the DUP evidence, that such early treatment changes longer term course. We are still not able to look at 10 to 20 year follow-ups of early intervention, including deaths from all causes.

Further into his article, McGorry et al promote their idea of a “staged” model. Again this is a transfer of ideas from physical medicine. While a useful research programme, we have no way of yet knowing what markers, biological or social, predict better or worse outcomes, or would respond to less treatment (perhaps not needing medication for instance). While of interest, we are not in a position to implement anything like this kind of detailed and specified service delivery for psychosis.

McGorry et al touch on, but do not elaborate, the point that most successful early intervention psychosis treatment includes considerable social and vocational input. Young people with psychosis typically wish to reduce their social exclusion – they wish to “get back to normal”, and have easy access to meaningful activity (jobs), study and relationships. Early intervention services typically include a large “dose” of vocational help. This suggests that it is not only psychosis that needs treating, but society's and the individual's attitude to the difficulties it can cause. Easing people back into “normal” environments, despite problems such as their sensitivity to stress and possible poor concentration, is made more difficult because of the poor public understanding, fear and stigma, that surrounds these diagnoses and prevents re-integration. The current anti-stigma campaigns in some countries are trying to improve this aspect.

However, it is not just society's reaction. Illness perception research shows that, as with physical illness, people with psychosis, and their carers, can have understandably negative views about the consequences of, and their ability to control problems, which can affect decisions about treat-ment. Because of this, as John Weinman has pointed out 2, “illness perceptions account for a significant and important amount of variance in outcome in physical illness”. People with psychosis share these attitudes 3,4. Certainly we know that rejecting medication, because of its side effects, as well as failing to engage with our services, remain concerns for this population.

Thirdly, McGorry et al only touch on the issue of family intervention for early psychosis. There is some evidence that it is helpful 5,6. However, we also know that there are more carers at early episodes, perhaps 60%, and that these carers experience similar difficulties and reactions as do later carers 7. We also know that the impact of care for relatives is related to long-term depression, and higher levels of stress and exhaustion while the caring role continues 8. There-fore, in terms of improving service user and carer outcomes, offering early intervention, including family intervention, from the beginning of episodes has to be sensible.

Finally, it is hard to argue against the idea that early detection, prodromal and early intervention services are a “good” thing. It must be good practice to offer the best service we can. As Max Birchwood has noted, early intervention aims particularly to reduce the chaos and high suicide rates of the first “critical” years of psychosis 9. However, we only have emerging evidence that it can reduce relapse and improve engagement 10and none showing that longer term course will improve. As I have suggested before 11, offering high quality, comprehensive, needs led services at all stages of presentation, early, medium or later, including offering optimism and hope of recovery 12, would seem to be a more reasonable strategy.