A Novel Approach for Developing and Interpreting Treatment Moderator Profiles in Randomized Clinical Trials

Abstract

Our results suggest that the strongest determinations for personalized treatment selection will likely require simultaneous consideration of multiple moderators, emphasizing the value of the methods presented here. After validation in a randomized clinical trial, a mental health practitioner could input a patient's relevant baseline values into a handheld computer programmed with the weights needed to calculate M*. The device could then output the patient's M* value and suggested treatment, thereby allowing the mental health practitioner to select the treatment that would offer the greatest likelihood of success for each patient.