When I was in medical school, a professor told us how he used to write prescriptions for Obecalp. He'd tell his patients it was a new drug that they could only get at the hospital pharmacy. (ObecalpHe'd tell his patients spells placebo backwards.) In today's medical climate, that would be unethical. But it seems that even if the patient knows she is getting a placebo it is still effective. In the U.S. opioids are only around 9% more effective than placebos. And opioids cause 44 deaths a day nationwide.
All of this, and more is discussed in an excellent NY Times Op-Ed piece A Placebo Treatment for Pain. (January 10, 2016)
The author asks, With placebo responses in pain so high — and the risks of drugs so severe — why not prescribe a course of “honest” placebos for those who wish to try it, before proceeding, if necessary, to an active drug?
For a more in-dept discussion see Michael Spector's "The Power of Nothing" in the December 12, 2011 New Yorker.