Sunday, May 13, 2018

Valid psychiatric diagnosis is unreliable

It's touching to see Paul Salkovskis & Irene Sutcliffe apparently hankering after DSM-III in their The Mental Elf blog post. For some reason they associate DSM-III with ICD-9, which was actually very similar to DSM-II not III. Surely ICD-9 still had some of the "notoriously unreliable" diagnostic definitions Salkovskis & Sutcliffe seem to so dislike. True, they do recognise that biological research has failed to substantiate reliable diagnostic categories, as in DSM-5 (see previous post).

What I'm objecting to is the way Paul Salkovskis & Irene Sutcliffe seem to follow Robert Spitzer in his view that "assuredly an unreliable system must be invalid" (Spitzer & Fleiss, 1974). As I said in my talk (see powerpoint slides), psychiatry should not be panicked by the unreliability of psychiatric diagnosis. If psychiatric diagnosis is going to be meaningful, there will be inevitable inconsistencies. In fact, paradoxically, increasingly internal consistency may well create an overly narrow measure that does not measure the construct optimally, which is called the attenuation paradox.

I'm not against Salkovskis & Sutcliffe encouraging debate about whether psychiatric diagnosis is finished (see previous post). But they speak against the Division of Clinical Psychology (DCP) position statement encouraging people to give up the disease model of mental disorder, which is actually what I think should happen (see previous post).

Most of The Mental Elf post is about the Power Threat Meaning (PTM) Framework, on which I have also blogged (eg. see previous post). Personally I'm not worried that DCP spent money on the launch of the Framework, as it needs to do something with its money. I agree with Salkovskis & Sutcliffe about the Framework being hard to follow, but I've tended to assume it will get easier once I've got used to it. Maybe though it does need a more user-friendly version.

I actually think the summary by Salkovskis & Sutcliffe of the document is quite good. What worries me though is that they seem to still believe in scientific positivism in mental health. They need to accept more the limitations of psychiatric diagnosis (see previous post), and take the PTM Framework more seriously. Lucy Johnstone, one of the main authors of the Framework, is merely arguing for service users to have more choice as to whether they take on a psychiatric label as part of their understanding of their problems (see my book review). Some service users may well find the PTM Framework helpful.

No comments: