You don't need to give people a placebo, you can test against a known good alternative. This is frequently used in just such cases, where an effective treatment

You don't need to give people a placebo, you can test against a known good alternative. This is frequently used in just such cases, where an effective treatment is already known, and denying patients in a trial that opportunity would be (as you say) unethical. https://en.wikipedia.org/wiki/Scientific_control#Positive https://en.wikipedia.org/wiki/Scientific_control#Positive