Alt Text
Office of Neuroscience Research > Neuroscience Calendar > Neurology/Division of Biostatistics Seminar: Yan Li (University of Alabama, Birmingham)

Neurology/Division of Biostatistics Seminar: Yan Li (University of Alabama, Birmingham)

" Sample Size Re-Estimation for Flexible Multi-Arm Trials with Normal Outcomes"


Add to your
calendar
When: Thursday, September 14, 2017 at 4:00 PM to 5:00 PM
Where: Becker Library 502 (Medical Campus, 5th floor)

This is a special faculty candidate seminar sponsored by the Depatment of Neurology and the Division of Biostatistics.

Abstract:
The need for efficient clinical trial designs has led to the development of flexible multi-arm designs that allow interim treatment selection and adaptation, primarily sample size re-estimation (SSR). However, SSR methods for two-arm trials are not directly applicable to multi-arm trials due to the application of the closed test procedure. In this study, we derived procedures for SSR based on the conditional power approach in the context of confirmatory two-stage multi-arm trials with normal outcomes for three designs: inverse normal combination test (INC), Fisher’s combination test (FiC) and flexible group sequential design (FGS). An R package - MASSR was developed for easy implementation. We conducted extensive simulation studies to evaluate the performance of the three designs with and without SSR. Results show that, without SSR, INC and FGS are slightly more powerful than FiC. With SSR, FGS is the most powerful design, though it requires slightly higher sample sizes than INC. Meanwhile, with SSR, INC always outperforms FiC. In practice, the choice of the most appropriate method will depend on the actual treatment effect profile, the timing of the interim analysis, the futility analysis scheme, the treatment selection rule applied and the maximum allowed sample size.

 

Light refreshments will be provided.



Medical Campus Map  |  Danforth Campus Map