Abstract:
|
In a stepped wedge study design, clusters start with the baseline treatment, and then crossover to the intervention at randomly determined times. Such study designs are useful when the intervention must be delivered at the cluster level and is expected to have a carryover effect. If the outcome is death or serious morbidity, researchers may have an ethical imperative to monitor and stop the study before maximum enrollment if the new therapy is proven to be beneficial. Additionally, sequential monitoring can make a research program more efficient by allowing for early termination of trials when a significant benefit for new therapy has been ruled out. However, use of the stepped wedge study design complicates the implementation of sequential monitoring methods. Both the amount of correlation within a cluster and the timing of the crossovers of clusters from one treatment to the other impact the rate of information growth, an important component at an interim analysis. We describe the information growth of these stepped wedge trials, potential problems if information growth is misspecified, and make recommendations for implementing group sequential methods in these trials.
|
ASA Meetings Department
732 North Washington Street, Alexandria, VA 22314
(703) 684-1221 • meetings@amstat.org
Copyright © American Statistical Association.