A Comparison of Randomized Trials and the Multi-phase Optimization Strategy for Behavioral Interventions

The multi-phase optimization strategy (MOST) is an experimental framework for achieving the goals of developing, optimizing and evaluating multi-component behavioral interventions. MOST was proposed by Collins et al. [8] as a precursor to a randomized confirmatory trial. MOST consists of three ordered phases: screening, refining, and confirming ; and it utilizes fractional factorial designs when there are a large number of possible components. The goal of this paper is to compare this new procedure (MOST) with the traditional approach of formulating a multi-component intervention and immediately proceeding to a confirmatory randomized trial, through a simulation study.