Perhaps unsurprisingly, something has annoyed me today. Today’s something has to do with diagnostic rigour, and the impact that has on the appropriateness of the treatments offered to patients.
We often see people referred following some stress or conflict or other. Sometimes there is psychiatric illness behind it; often there is not. Sometimes when there is not, a suggestion is still made (not by me, nor generally by any member of my team) that involves medication (usually antidepressants, sometimes – perhaps often – quetiapine), and even referral for ongoing follow-up by a psychiatric team.
My question: why? Why would you give a psychiatric medication to someone not psychiatrically ill? Why would you have a psychiatric team have ongoing contact with someone not psychiatrically ill?
The usual response? They have “some depressive features”.
Honestly, that is like being a little bit pregnant, or having some features of a heart attack. It’s stupid.
In other areas of medicine, including pregnancy and myocardial infarctions, we have investigations which can tell us if the condition really is present: we can do a serum βHCG, and ultrasound scans; we can check your troponin-I, and do ECGs.
In psychiatry we don’t. That’s why we have lists of symptoms -crucially, with thresholds: for both numbers and duration of symptoms. Without that, everyone in the world could be diagnosed with some psychiatric illness or other. The thresholds (eg at least 5 of the listed symptoms, for at least 2 weeks, for major depressive disorder) are there in an attempt to differentiate illness, or at least legitimate cause for clinical attention, from not.
Now, I’m one of the first to point out the problems with DSM (and probably would do so with ICD if I was as familiar with it). However, that doesn’t give me licence to ignore it. For people to prescribe antidepressants to someone who does not have a depressive disorder, only “some depressive features” goes against all the work that has gone into delineating major depressive disorder as a real entity. It is also not supported by evidence: not only are the trials that show antidepressants to be effective done in groups of people with actual major depressive disorder (not “some depressive features”), recent meta-analyses suggest (despite their shortcomings) that antidepressants might not even be much better than placebo for mild or moderate depression … So how likely is it that they would help “some depressive features”?
It’s too common though. A recent study (sorry, I can’t find the reference right now) looked at antidepressant prescribing – in the USA, but I don’t see that we have cause to assume we’d be different – and found that around a third of the prescriptions were written in the absence of major depression.
Any question as to why so many people find that their antidepressants don’t work?