There a very interesting article at http://www.forteantimes.com
The jist of it is that if you compare the results of giving patients a placebo with the results of not treating them at all, the difference is almost nothing. They’re saying that that drug studies should be re-designed to have three groups of people: One given the real drug, one given a placebo, and one given nothing.
This probably belongs in GQ (or even GD).
I heard the same story on NPR.
Here is the story from the New England Journal of Medicine
The link is to this article in the Times.. It may be that the article misrepresents the study, but it seems to me that this misses the point of trials with placebos, which loosely speaking is attempting to isolate the (reported or otherwise measured) effects of what is in the pill from the effects of taking a pill.