Wednesday, June 06, 2007

Bias in drug trials.

According to this, drug trials financed by pharmaceutical companies tend to produce the results the sponsors want. In particular, in trials of one statin drug vs. another, any given study produced results favourable to the manufacturer that sponsored the study. Gasp, etc. I can think of a good half-dozen explanations for this that aren't particularly conspiracy theory; but, just for kicks, I went to find the paper that tried to establish this: here.

According to the "methods and findings" section, almost all the trials examined contained methodological weaknesses and errors -- which would lead one to expect some non-experimental factor contaminating the data. In trials that were better -- with adequate blinding, for example -- it was less likely that the conclusion would be favourable to the sponsor's drug.

So, overall, this isn't a big lesson about keeping pharmaceutical companies from funding research or what have you. Instead, it's a lesson about doing the trials properly. (What's more worrying to me, honestly, is that bad studies ever got published. So much for peer review.) The paper itself suggests one such explanation of their results:
Some people have suggested that drug companies may deliberately choose lower dosages for the comparison drug when they carry out “head-to-head” trials; this tactic is likely to result in the company's product doing better in the trial.
In other words, if the trial design is deliberately botched, favourable results can be obtained.

No comments: