Monday, June 09, 2014

The benefit-to-harm ratio of antidepressants

A Lancet Psychiatry article attacks Peter Gøtzsche (who I have mentioned in a previous post) for saying that antidepressants do more harm than good. Although his Guardian article has this headline, in fact what he wrote was "the way we currently use psychiatric drugs is causing more harm than good" [my emphasis]. He doesn't seem to be advocating not using antidepressants at all, but "much less, for shorter periods of time, and always with a plan for tapering off, to prevent people from being medicated for the rest of their lives".

As I commented in my previous post about Gøtzsche, the problem is that attacking him in this way deflects from the validity of what he is saying. I wish mainstream psychiatry would engage more with me. I do know it is important to avoid overstatement (see previous post). The Lancet Psychiatry article emphasises the effect size found in antidepressant trials. What it doesn't discuss is whether this finding could be explained by placebo amplification due to unblinding in clinical trials (eg. see previous post). Nor does it discuss the bias in the literature introduced through selective publication. These days this is because data submitted by the drug companies to the regulatory authorities is not always examined completely. Previously, with older trials, this was because of what was called the 'file drawer problem' (see further information) in that negative trials did not always get published. This may well have affected the amitriptyline data that the Lancet Psychiatry article cites.

As I said in a comment on a previous post, it's important not to get too hung up by the size of effect in clinical trials. Maybe I'm too sceptical, but the finding that non-psychiatric drugs have similar effect sizes in clinical trials does not imply that psychiatric drugs are effective, but that medical as well as psychiatric clinical trials are subject to the same biases of unblinding (see previous post).

The authors of the article speculate about the reasons for doctors questioning the effectiveness of antidepressants, suggesting it is the anti-psychological bias of doctors that makes them want to believe that there can't be a physical treatment that could possibly be effective for mental illness. So, they try to turn the tables on critics by suggesting it's the critics not them that are stigmatising mental illness. They feel insulted by the critique but it is important to be open to argument and not prejudiced.

5 comments:

Chrys Muirhead said...

Interesting post Duncan, thanks.

One question. Who is the man in the photo?

Adzcliff said...

Hi Chrys. That is Prof. David Nutt. He is perhaps most famous for being the sacked head of the Government's independent Advisory Council on the Misuse of Drugs, whose independent scientific findings were not consistent with government drugs policy. However, I recall he has also been involved in a spat with Prof. Richard Bentall about the effectiveness of psychological therapy versus psychiatric drug treatment (I think). I think he provides an excellent example against the argument from authority, as he is both a personal hero and irritant of mine depending on what issue he's addressing.

Ta.

Adzcliff said...

I notice Nutt et al. have opted for the straw man by going after – ‘resurrecting’ even – the nebulas and undeniably controversial 'anti-psychiatry' movement. I would much have preferred a proper scientific critique of the misleadings of Kirsch et al. (2008) and the atropine studies, and why anti-depressants with ostensibly opposite actions yield similar effect sizes (e.g. Tianeptine versus SSRIs). I am very much open to having my mind changed on these issues .

As it stands, the Lancet letter comes across as a desperate attempt to defend academic turf. Note also how rare and bizarre anti-depressant side-effects are, probably entirely sensibly, relegated to 'unexplained medical symptom', but the sexual interference of patients by psychotherapists is raised as their second and last indentified potential side-effect of psychotherapy.

I suppose when you're a psychopharmacologist, everything looks like a nail. Not sure...

John Hoggett said...

Critisising anti-depressnts gets up Nutt's nose. He has got hot under the collar before about David Healy but I have not heard him produce figures countering these people's arguments.

If he has done some serious analysis that counters these peoples analysis can someone post a link?

Anonymous said...

Gøtzsche has long been a nasty thorn in the side of mainstream medicine. A very common way to neutralize such a threat is by "attacking him" and by trying to "turn the tables on critics by suggesting it's the critics not them." Gøtzsche is most well-known for his many mammography-critical studies. In that field too, some pro-mammogram researchers had called him a pseudo-critic or twisted and distorted Gøtzsche's position on mammography (google/bing "A Mammogram Letter The British Medical Journal Censored"), and some have accused him of only wanting to sell his book.

Yet anyone who analyzes Gøtzsche's material will recognize that he is most accurate and the orthodoxy is mostly wrong. The medical industry's evidence on SSRIs is about as shallow as their ad hominem attacks.