Review of the use of pattern analysis to differentiate true drug and placebo responses

Placebo response in patients assigned active drug is a troubling source of variance in antidepressant studies. This paper summarizes a series of studies utilizing pattern analysis to distinguish between placebo and true drug responses. Analysis of the persistence, speed, and timing of onset of patients' improvement during antidepressant therapy reveals distinct patterns of response which are likely to be attributable to placebo and true drug effects. While true drug effects seem to be characterized by a 2-week delay in onset followed by persistent improvement, placebo effects seem to be characterized by abrupt, transient improvement. Gradual responses on placebo may be due to spontaneous remission. The heuristic and clinical implications of pattern analysis are discussed.