A medical approach naturally relies on diagnosis. I shall expand on this issue in Chapter 2, but there is precious little science in psychiatric diagnosis. I chose to work in the mental health system anticipating that I would be part of a vital and life-saving system of committed and caring professionals, eager to understand their clients' lives. Over more than twenty years as an active clinician and researcher, I am reluctantly coming to the conclusion, shared by many of my colleagues, that the dominant diagnostic framework in mental health is unreliable, invalid and inhumane and even bizarre. Committees of experts (with clear financial as well as professional advantage in coming to a consensus that supports a 'disease model' in the face of scientific evidence) fail to agree on diagnostic criteria. Statistical analyses fail to identify clusters of symptoms that correspond to the putative 'disorders'. In fact, the abject failure of diagnostic approaches has led the enormously wealthy and enormously influential US NIMH (National Institute of Mental Health) to declare that its strategy 'cannot succeed' if it uses the diagnostic categories enshrined in the influential 'Diagnostic and Statistical Manual' of the American Psychiatric Association (which I shall describe in more detail in Chapter 2), and henceforth will 'be re-orient its research away from [those] categories'.26 And, as I shall detail in Chapter 2, there is widespread opposition from a wide group of professional and service user-led groups.
Despite this, in the current (flawed) system, diagnostic labels have huge implications - they affect access to healthcare (but they shouldn't), they affect legal issues (but they shouldn't), they affect employment rights (but they shouldn't) and they lead to discrimination and stigma. They are hugely 'sticky'; like zombies, they may well be dead, but they don't realise it and carry on walking. Fortunately, again, however, alternatives are available. A combination of simple problem identification and psychological formulations would address people's problems. As I explain in Chapter 2, a simple list of problems would permit sensible and appropriate planning, and would result in much higher levels of reliability. Validity, too, would be significantly improved, as invalidity in psychiatric diagnosis stems not from the identification of specific problems - whether 'symptoms' (effectively, ways of thinking and acting) or environmental stresses - but from the complex rules for combining these to form diagnoses and the meaning of those diagnoses as representing 'real illnesses'.