Sunday, October 16, 2011

Psychiatrists should know what they're doing with medication

An editorial this month in the British Journal of Psychiatry is entitled 'No psychiatry without psychopharmacology'. It suggests there is insufficient prioritisation of psychopharmacology in psychiatric training. It asks for psychopharmacology to be affirmed as an integral and significant component of psychiatry, and a consequent expectation of a commensurately high level of knowledge and practice.

As I mentioned in my previous post, I've just come back from a conference in Newcastle. Paul Harrison, the main author of the editorial, gave a presentation at the meeting with the same title as the editorial. He thought that the low priority apparently given to psychopharmacology may be related to what he called the 'anti-pharmacological lobby'.  I suppose I would be seen as part of this lobby, but, despite what Harrison implies, I agree that psychiatrists should be well trained in psychopharmacology.

What I didn't hear in his talk, but is mentioned twice in the editorial, is unquestioned belief in the efficacy of psychiatric medication. To quote from the editorial, research shows that "contemporary psychotropic medications are effective". Again, it says that "the evidence that psychotropic drugs are beneficial when used in the right way and for the correct indications is unequivocal".

Good psychopharmacological training will highlight the weaknesses and bias of randomised controlled trials (eg. see my Bias in controlled trials webpage). The trouble with much prescribing is that it is beyond the evidence (and not even conforming to guidelines such as those produced by NICE). Psychiatrists are insufficiently critical of the evidence base for medication.

Are antidepressants really placebos?

I've just come back from Newcastle, where the Faculty of General and Community Psychiatry of the Royal College of Psychiatrists has been holding its Annual General Meeting (see programme). One of the talks was by Professor Ian Anderson entitled 'Are antidepressants really placebos?' As he himself said, as might be expected, his answer was no.

I'm not sure if the talk really had much more in it than a letter he had published in the British Journal of Psychiatry several years ago. Anderson makes a lot of the fact that continuation studies show high relapse rates. Following the results of the meta-analysis by Geddes et al (2003), the average rate of relapse on placebo is 41% compared with 18% on active treatment. In other words, continuing treatment with antidepressants reduces the odds of relapse by 70%. Anderson doesn't think this could possibly be a placebo effect.

Doctors have always underestimated the difficulties of discontinuing antidepressants (see my Antidepressant discontinuation reactions website). If patients are significantly unblinded in discontinuation studies, the negative placebo (nocebo) response could explain these results because of how reliant people have become on their medication. Any change threatens an equilibrium related to a complex set of meanings that their medication has acquired.