Thursday, February 28, 2008

Publication bias and my brilliance at darts

I’m a brilliant darts player. But more on that later.

Thanks to Matt for pointing me towards Ben Goldacre’s blog, for people who like his Bad Science column in the Guardian but don’t want to wait a whole week for their next fix.

Why have SSRI antidepressants now been found to be largely ineffective? Or, to put it another way, why had they previously been thought to be more effective?

The answer is that a lot of the clinical trials are performed by the drugs companies on their own drugs – and, as Goldacre says, they “have repeatedly been shown to bury unflattering data”. The old studies weren’t wrong, they were just heavily cherry-picked. The new study uses the US Freedom of Information Act to include data from some of the unpublished trials, thus vastly diluting the previously published positive results and giving a more rounded picture.

Goldacre argues that this behaviour “breaks a key moral contract between patient and researcher” – patients participate in trials on the understanding that they’re helping to further understanding of which treatments do – and don’t – work.

He recommends: “Nobody should get ethical approval to perform a clinical trial unless there is a clear undertaking that the results will be published, in full, in a publicly available forum” and also that there should be a compulsory international trials register so that disappointing data can’t be hidden.

Good call.

Now, back to darts, at which I am brilliant. You doubt me? Fine: to prove my skill, I will post a video of me throwing a perfect 180.

Might take a while to film it, though. You needn’t worry about the out-takes.

1 comment:

Incitatus4Congress said...

It’s not really that straightforward, though. There’s a good article over on Slate, by Peter Kramer, that’s worth a read.

Being a company man himself, I also recommend checking out Derek Lowe’s blog, In The Pipeline ,for an insiders view.