Randomized Controlled Trials: The Stepped Wedge Design

The stepped wedge design is a special randomized controlled trial to determine the efficacy of a new treatment with respect to a control treatment. All participants first start at the control treatment and then (groups of) participants switch to the new intervention at predefined switch moments and they stay on this new treatment until the end of the trial. Allocation of participants to the switch moments is randomized. Due to this one-directional structure, the statistical analysis is considered less straightforward than more classical designs. The talk starts with a short history of the SWD followed by its definition and the advantages and disadvantages. The main part of the talk is about the statistical contributions that have been established so far: statistical analysis, sample size calculation, and optimal designs. A specific case study is presented to illustrate the complexities. The final part discusses a few issues that are still open.