Doctor-Patient Relationship Predicts Medication Effectiveness
Whether we like it or not, medications only have a modest effect on improving major depression. This effect could be constrained by large responses to non-drug-related features of clinical situations. A study  was done to examine such an influence, which looked at the patient-doctor relationship, pill administration, and the expectation of a response.
About ninety people participated in an eight-week treatment. They were randomly assigned to either supportive care alone or combined with a pill, in the latter group neither the participant nor the treater knew if the pill was placebo or medication. In addition to symptomatic improvements, participant's perceptions of general treatment effectiveness and therapeutic alliance were measured.
It appears "pill-taking" adds significantly to the benefits of supportive care for depression. These benefits cannot be explained by supportive interaction with the provider and are influenced by a patient's relationship with their prescriber.