A1667
Title: Sample size requirements to detect an intervention by time interaction in four-level longitudinal CRT
Authors: Priyanka Majumder - Indian Institute of Science Education and Research Thiruvananthapuram (India) [presenting]
Siuli Mukhopadhyay Siuli Mukhopadhyay - IIT Bombay (India)
Abstract: Cluster/group randomized controlled trials (CRTs) have a long history in the study of health sciences. The aim is to design and analyze four-level longitudinal cluster randomized trials. The main interest is to study the difference between treatment groups over time for such a four-level hierarchical data structure. This motivation is based on a real-life study of education-based HIV prevention. Such trials are popular not only for administrative convenience, ethical considerations, and subject compliance but also to help reduce contamination bias. A random intercept mixed model affects linear regression, including a time-by-intervention interaction used for modeling. Closed-form expression of the power function to detect the interaction effect is determined. Sample size equations depend on correlations among schools but not on correlations among classes or students, while the power function depends on the product of the number of units at different levels. The optimal allocation of units under a fixed cost by minimizing the expected standardized variance is also determined and is shown to be independent of correlations among units at any level. Results of detailed simulation studies find the theoretical power estimates based on the derived formulae close to the empirical estimates.
