A Maximum Likelihood Approach to Power Calculations for the Risk Difference in a Stepped Wedge Design

• Activity Number: 610
• Type: Contributed
• Date/Time: Wednesday, August 12, 2015 : 2:00 PM to 3:50 PM
• Sponsor: Biometrics Section
• Author(s): Lauren Kunz* and Sharon-Lise Normand and Donna Spiegelman
• Companies: NIH/NHLBI and Harvard Medical School and Harvard School of Public Health
• Keywords: stepped wedge design ; power calculation ; study design ; cluster randomization

Abstract

Stepped wedge designs (SWDs) randomize clusters to the period at which they receive the intervention in a sequential rollout over time periods used to assess an intervention. We developed numerical methods to determine asymptotic power of a range of SWDs using Romberg integration over the unobservable cluster random effects in the common setting of a binary outcome comparing two interventions. Using two-sided Wald tests, we compared the power for detecting risk differences ranging from 0.0125 to 0.1 in designs based upon a new likelihood framework to designs based upon a closed-form approximation given by Hussey and Hughes (HH) (2007). Over a range of parameters, our exact method provided designs 1.09 to 2.4 times more efficient than those given by HH. Utilizing a theoretical asymptotic approach to power calculations will provide more efficient study designs for detecting risk differences of pre-specified magnitude than the previously available method, suggesting that the SWD may be more feasible than previously appreciated. The new method was applied to the design of a study of mechanical device implantation for therapy in patients with severe heart failure.