Design and Analysis Considerations for Stepped Wedge Cluster Randomized Trials with Decayed Correlation
A stepped wedge cluster randomized trial is a type of longitudinal cluster design that sequentially switches clusters to intervention over time until all clusters are treated. While the traditional posttest-only parallel design requires adjustment for a single intraclass correlation coefficient, the stepped wedge design allows multiple outcome measurements from the same cluster and so additional correlation parameters are necessary to characterize the within-cluster dependency structure. Although a number of studies have differentiated between the concepts of within-period and inter-period correlations, few studies have allowed the inter-period correlation to decay over time. In this article, we consider the proportional decay correlation structure for a cohort stepped wedge design, and provide a matrix-adjusted quasi-least squares (MAQLS) approach to accurately estimate the correlation parameters along with the intervention effect. We further develop a corresponding sample size procedure accounting for the correlation decay, and numerically validate it for continuous outcomes in a simulation study. We show that the empirical power agrees well with the prediction even with a small number of clusters, when data are analyzed with MAQLS concurrently with a suitable bias-corrected sandwich variance. Two trial examples are provided to illustrate the new sample size procedure.
READ FULL TEXT