Abstract:
|
The stepped wedge trial (SWT) is a type of cluster randomized trial with unidirectional crossover from control to intervention with randomly assigned crossover times for each cluster. The existing literature recommends including a fixed effect for time and a random effect for cluster. Under this modeling strategy, each cluster is assumed to have the same background temporal trend in outcome. In this paper, we describe our simulation studies for the effects of heterogeneous temporal trends at the cluster level on bias, standard errors, and power also explore the interactions between the effect of intervention and time.
|