Comment by disgruntledphd2

Experientially I agree with you, but therapy is a really really bad fit for controlled trials, as the active ingredient appears to be the patient provider relationship.