It goes without saying that the drugs you take for a headache, or high blood pressure, or even depression should work better than a Tic-Tac. That’s what drug trials are for: researchers give a group of subjects either the drug under investigation or a placebo to check that the medicine is significantly more effective than a sugar pill. Plus, the trials can reveal any potentially harmful side effects. In theory, this is a great way to weed out useless or actively harmful drugs. But it fails when drug manufacturers cherry-pick their data, publishing papers on the positive trials and sweeping the unsuccessful ones under the rug. And this behavior is completely legal.
Science writer and medical doctor Ben Goldacre wrote a book, with a long excerpt published at the Guardian, about how this process leads to approval for drugs that don’t actually work. And as he explains, when widely used drugs—such as the diabetes medication rosiglitazone—have harmful side effects, they sometimes remain in common use.
In 2003 the Uppsala drug monitoring group of the World Health Organisation contacted [pharmaceutical company GlaxoSmithKline] about an unusually large number of spontaneous reports associating rosiglitazone with heart problems. GSK conducted …
from Discover Magazine