Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

That's why it's normal these days, for clinical studies, for the control group to receive current best available treatment, rather than a placebo.

After all, what you're generally trying to find out is whether the new treatment is better than what you already have - just being better than a placebo isn't very informative.

If you're interested in this sort of thing, I strongly recommend reading Ben Goldacre's work[1].

[1] Start at badscience.net.



> That's why it's normal these days, for clinical studies, for the control group to receive current best available treatment, rather than a placebo.

That's one of the reasons such studies are so unreliable -- they depend on a comparison of two or more drugs that may operate in different ways, rather than comparing a drug to the absence of a drug.

The result is predictable -- unbiased meta-analyses show that antidepressants don't work:

http://www.plosmedicine.org/article/info%3Adoi%2F10.1371%2Fj...

This is not to say that wouldn't have been the outcome anyway, regardless of the methods used, only that testing two drugs at once introduces confounding factors.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: