Placebos work even when people know it's a placebo. We might want to establish a baseline, like the pretender has to remind the patient he's just a pretender once a year. But if a pretend doctor can help, then go for it.

It is unethical to say that you are administering a medicine when you are not doing it, except in trials to which the subject has agreed.

I 100% agree, but that doesn't mean we can't still administer placebos:

I've read elsewhere that you don't have to lie to the person to still have a successful placebo effect- they can still receive a benefit even knowing it's not a real drug.