I agree with the posts above that placebo trials are needed. However, the experiment design must be controlled in a way such that ethical concerns are mitigated. Ethical concerns arise when a proven effective treatment exists for fear that the placebo exposes patients to the potential harms of non-treatment.
There are general guidelines for when it is ethical to use placebo controls.
1)When there is no proven effective intervention for condition under study or when placebo is compared against an investigational treatment added on to established treatment.
2)When withholding an established, effective intervention would expose subjects to, at most, temporary discomfort or delay in relief of symptoms. In this case, the placebo is ok if the patient goes untreated and there are relatively no negative consequences or no treatment is an acceptable alternative.
3)If trial outcomes show a high expected placebo response or mixed data on effectiveness of standard treatment. For example, depression has fluctuating symptoms and a high placebo response rate.
4) When participants are not deprived of interventions they would otherwise receive and when research intended to develop interventions that will benefit the host population.

Millum, Joseph, and Christine Grady. “The Ethics of Placebo-Controlled Trials: Methodological Justifications.” Contemporary clinical trials 36.2 (2013): 10.1016/j.cct.2013.09.003. PMC. Web. 27 Oct. 2017.