Description:

Center for Dissemination and Implementation At Stanford (C-DIAS)

Prevention Science & Methodology Group (PSMG) Virtual Grand Rounds

Tuesday, March 17, 2026

12:00 – 1:30 PM CT

C. Hendricks Brown, PhD
Northwestern University

Gregory Simon, MD, MPH
Kaiser Permanente Washington Health Research Institute

Rollout Designs in Implementation Research: Often Necessary and Sometimes Preferred 

Background: Rollout designs, which include stepped wedge designs, are defined by staggered implementation of new or alternative programs or services. Critiques of stepped wedge and other rollout designs have raised concerns regarding the confounding of true implementation or program effects with unrelated, global changes in service delivery, with some recommending they only be used when traditional parallel-group designs are not practicable. However, rollout designs may sometimes be more suitable than traditional parallel group designs for ethical, scientific, or practical reasons.

Results: We define and provide rationale for and examples of stepped wedge and the larger class of rollout designs, in which all participating units receive a new program or service implementation. Staged implementation in a rollout design may be necessary when denying, rather than delaying, implementation of a known effective service is ethically unacceptable. Scientifically, stepped wedge has increased statistical power relative to an equivalent parallel group design, and some rollout designs have the capability to compare different phases of implementation and sustainment. A rollout design may be practically necessary either because of limited resources and other logistical challenges or community requirements that no site serve as a control. Examples of completed and ongoing rollout trials illustrate how these ethical, scientific, and practical considerations influenced trial designs.

Conclusions: Stepped wedge and other rollout trial designs may be well suited to evaluation of implementation strategies or policy changes. In implementation trials, rollout designs may be necessary for practical reasons, may be required for ethical reasons, and may be preferred for scientific reasons. We summarize when such rollout designs have advantages and drawbacks.

Please join our PSMG listserv to attend live at https://cepim.northwestern.edu/psmg-membership or you may view any of our past presentations on our archive which are open to all https://cepim.northwestern.edu/psmg-archive