Thursday, February 23, 2017

What does it mean to say that antidepressants don't work?

Like Carmine Pariante, I have treated thousands of patients since I started as a trainee psychiatrist in 1985 and became a consultant over 25 years ago (see post on my personal blog). He is convinced that antidepressants work (see article), whereas I am more sceptical. He's correct that I'm in a minority. The trouble with relying on personal experience is that we can delude ourselves.

Pariante does recognise that "about one in three patients with serious clinical depression who takes them doesn't get better". Actually what he means is that about a third of patients do not seem to be helped in clinical trials. These trials are only short-term and people on antidepressants are compared with those taking placebo. About a half of patients in the clinical trials taking placebo also seem to improve. It's, therefore, misleading to imply that two-thirds of patients are getting better because of antidepressants (see post summarising evidence of outcomes with antidepressants). The difference between active and placebo treatment in clinical trials is much smaller than most people realise. In fact it's so small it could be an artefact (see previous post). 

1 comment:

Adam said...

The difference between active and placebo treatment in clinical trials is much smaller than most people realise. In fact it's so small it could be an artefact.

Indeed, and in trials where the active treatment is compared against an 'active placebo' (i.e. a non-antidepressant with similar side-effect profile), the effect vanishes almost entirely. It seems drugs don't need to be called antidepressants to produce effects, they just need to be distinguishable from inert placebo by trial participants (and not even in pleasant ways).