You raise a good point. When you know the mechanism of the disease and you can tell who is susceptible and who is not (perhaps, say, by a genetic test), then the interpretation of a clinical study is flawed if you don't separate out different types of patients. However, if you don't know the mechanism, and you cannot tell who is susceptible (and you are not even sure it is a real disease), there is little difference between saying "all people are the same and everyone has a 10% chance of being sick" versus "10% of the people are susceptible and will get sick, and 90% are immune and will not get sick." The patient in one case wonders "will I be unlucky" and in the other case wonders "am I susceptible". There is little difference in what the patient thinks or how the doctor treats the patient.

The whole point of a placebo is to eliminate the impact of a "positive or negative viewpoint". A good placebo will give no indication if it is real or fake, so on average there will be no affect on the viewpoint. True, people with chronic negative attitudes may be sicker than optimistic people, but that has nothing to do with the disease being tested. In a large random study, both the treated and placebo groups should have similar numbers of each group. The goal is to separate out a treatment of a disease from any effect of viewpoint, so you do a randomized, double-blind clinical study with a large number of participants to remove viewpoint artifacts.

Thanks for you comments. I appreciate your taking the time to read my work. I hope this clarification helps.