Commentary on “Considerations Before Selecting a Stepped-Wedge Cluster Randomized Trial Design for a Practice Improvement Study”

In many implementation studies, the primary units of analysis at the level of the setting and/or health practitioner. Recruiting sufficient numbers settings to power analyses can be challenging. To address this challenge, a growing number of researchers are choosing to use Stepped-Wedge Cluster Randomized Trials (SW-CRT) designs for their implementation studies.  While we know SW-CRT increases statistical power, we know less about the practical challenges and benefits to using this design. Nguyen and colleagues address this question by building on the experience of investigators who participated in the AHRQ-funded EvidenceNOW initiative, which funded seven 36-months grants to study the implementation of cardiovascular disease prevention guidelines in primary care. Seven regional cooperatives, each of which included at least 200 clinics, participated in EvidenceNOW. This paper reports the findings from interviews with 17 key informants, including all EvidenceNOW principal investigators as well as other members of their teams. Interview findings detail investigators’ reflections on the benefits of SW-CRT. In addition to increasing statistical power, the design allows research teams to economize their investments in implementation strategies. Because implementation occurs across cohorts in a stepped sequence, researchers are able to shift resources (e.g., facilitation or coaching staff) from one cohort to the next. The design also ensures that all participating settings receive support for implementation.

In reflecting on the challenges of SW-CRT, investigators reported on how difficult it was to recruit all participating settings prior to the start of the trial, even settings that may not receive implementation strategies for months or even years.  They also reported difficulty retaining settings over time.  Anyone who has done research in primary care or other clinical or community settings knows how challenging it is to retain settings, which may experience turnover in leadership, close, or reorganize. Furthermore, much of the initial enthusiasm for the initiative may be lost overtime as settings wait for their turn in the stepped wedge design. One of several other challenges outlined in the article is the challenge of temporal trends. Since this article was written, the COVID-19 pandemic has highlighted the challenges created when temporal trends change the implementation context and create barriers to analyzing findings across settings that were stepped into a trial before and after March 2019.  Reflections on the real world challenges to a study design, such as this paper offers, provide investigators with the foundation for in-depth discussions with members of the study team, practice partners, and other key stakeholders. Through these discussions, research and practice stakeholders can anticipate challenges and develop strategies to overcome them.

References:

1. Nguyen, A. M., Cleland, C. M., Dickinson, L. M., Barry, M. P., Cykert, S., Duffy, F. D., . . . Walunas, T. L. (2022). Considerations Before Selecting a Stepped-Wedge Cluster Randomized Trial Design for a Practice Improvement Study. Ann Fam Med, 20(3), 255-261. doi:10.1370/afm.2810