Reported problems and responses during the conduct of stepped-wedge cluster randomized trials in healthcare settings: a qualitative systematic review

Background: The stepped-wedge cluster randomized trial (SW-CRT) is a pragmatic complex design that can be difficult to implement. We aimed to summarize the reported problems and responses to problems in studies recently published after the publication of the reporting guidelines for SW-CRTs. Methods: We searched the literature for SW-CRTs published between 9 November 2018 and 23 February 2021 to identify reported SW-CRT-related problems (defined as relating to the components of the design, i.e. involving clusters and the staggered intervention implementation) and responses to problems. We carried out a thematic analysis to derive descriptive themes and overarching analytical themes. Results: Among 84 included SW-CRTs, 62 (74%) reported 107 problems related to the SW-CRT design and 38 responses to 36 problems were reported by 24 trials. The “problems” formed six descriptive problem themes: “participant recruitment,” “cluster issues” (e.g. cluster merger or dropout), “internal factors” (e.g. logistic or administrative issues), “external factors” (e.g. weather or religious events), “outcome measurement” (e.g. practicalities around measurement of repeated outcomes), and “intervention implementation” (e.g. delays or contamination). The “responses” formed six descriptive themes: “adding new clusters,” “modifying the randomization,” “reducing contamination,” “changing outcomes,” “intention-to-treat,” and “modifying the analysis.” Conclusion: SW-CRTs commonly run into problems. Two overarching and conflicting analytical problem themes emerged: the “struggle to adhere to the protocol,” given the defining features of the SW-CRT design, when faced against “real-life pressures” created by internal and external factors. Further research is needed to explore whether responses to these problems have resource or integrity ramifications.