They call it the Abandoned Illness, in the Schizophrenia commission's report – but not, they emphasise, because it is an illness society can afford to abandon. In fact, schizophrenia costs the health service more than cancer or heart disease. It's the most common cause of hospitalisation, and – since it won't go away on its own – will last a lifetime with the level of care patients often receive.
There is a high level of coercion; every year, more people are admitted to hospital against their will as surrounding services are cut. The conditions in mental health units are so demoralized, overcrowded, grotty and often dangerous, that every time you're admitted against your will, that experience in itself will make you progressively less likely to go in of your own accord. Coercive care is the most expensive form of treatment you could ever devise; last year it cost £1.2bn, about 19% of the mental health budget.
Patients are often put on drugs which then aren't monitored. They're given no access to talking therapies, and after a decade or two the side-effects of the drugs may have become more problematic and more defining than the illness itself. Partly as a consequence of this, partly because the time isn't taken to involve them in the treatment of their physical health, people with severe mental illnesses die 15 to 20 years earlier than the rest of the population.
If engaging with the NHS is difficult, then engaging with the surrounding network, the benefits system, is rendered more so by an institutional dimwittedness that often sounds deliberate.
My stepmother, whose son has a diagnosis of schizophrenia, read me the questions on the disability living allowance form: "I am not motivated to wash: how often? How long each time? I am not aware of common dangers: how often? How long each time?; I might wander: how often? How long each time?" The traits ascribed to serious mental illness are often wildly off, as if the person devising the form couldn't be bothered to look up the illness on Wikipedia and didn't even aim for an internal logic to their own questions. And all that is pre-Atos, whose assessments on mental illness are so ignorant that in July the Public Law Project won the right to take them to judicial review.
Slipping through benefits assessments; being left on drug regimes that are accompanied by many other problems (weight gain sounds trivial, but as a cause of premature death, it isn't); never getting the cognitive behavioural therapy Nice recommends – all these things heap on pressure, and the result is often crisis hospitalisation.
In a way, this situation is totally predictable. You take a diagnosis that is at once very fixed (a life sentence, incurable) but at the same time, very fluid (taking in so many symptoms, covering so much ground) and it is unsurprising to find its treatment marked by low morale and inertia.
But that's nothing like the full story of this report. Nearly a decade ago an early intervention programme (EIP) was started whose defining features weren't, as its name suggests, just arriving fast on the scene of a recent diagnosis. Instead, as consultant clinical psychologist Dr Alison Brabben explains: "It was quite a break away from traditional mental health services. Previously, schizophrenia was seen as a purely biological condition, the diagnosis was made and then people were given a pill to try to make the symptoms better. No one ever asked about people's lives."
EIP staff had small caseloads and were highly trained, and they could refer people to cognitive behavioural and family therapies – that's the bit that cost the money. The part patients valued was that they were asked real questions about their concerns. Plenty of people can live with delusions and voices: it's some other factor that makes their lives unliveable. It's probably related to money or relationships, like everybody else's problems are. It sounds so obvious, doesn't it? When you listen to people they engage more, they can make use of the support you're offering and they're less likely to end up in hospital.
Nevertheless, it remains contentious, because part of the treatment involves allowing for the possibility that the disease was caused by trauma.
Medicine, and indeed society, frames its questions in a binary way: it's either chemical or it's psycho-social. It's either incurable or it's curable. In fact, it has been clear for a long time that the chemical explanation for psychosis was incomplete. If you look at the constituency of people with the diagnosis, black groups are far more likely to be represented, and yet these high rates aren't found in Africa or the Caribbean. Being black isn't the problem. It's being black in Britain. Being poor, being discriminated against, being bullied. "If you could remove early adversity, you would probably remove a third of cases of psychosis," Brabben says. And yet the trauma explanation isn't complete either – what about the other two-thirds?
Ultimately, treatment will lie in the grey areas – the causes that remain unknown, the interventions that can't be measured in milligrams, whose success is defined in terms of "personal recovery" and doesn't look the same in any two people. More pressingly, local commissioning bodies must resist the urge to cut costs by driving up caseloads and driving down training; the process of sucking the time and warmth and energy out of the relationship which, in many places, has already begun. It might sound cheaper, but it won't be. Local commissioners, it will get you where it hurts, right in your fiendishly expensive, locum-staffed secure units, which should be a last resort – and too often have been the only resort.
Twitter: @zoesqwilliams