It is now standard practice to incorporate formal sequential monitoring procedures into the design and conduct of long term randomized clinical trials. Interim monitoring provides an advance determination of efficacy and safety so that a data and safety monitoring board can decide whether to terminate the study early. However the integrity of a study can be seriously compromised if the accruing data are monitored inappropriately. One then runs the risk of attributing chance fluctuations in the data to a real therapeutic effect. Hence one needs specialized statistical methods and accompanying software that can balance the ethical and economic advantages of stopping a study early against the risk of an incorrect conclusion. Research has progressed rapidly in this area over the past two decades and methods have been developed to meet the key practical needs. These methods are not yet accessible to clinical trials practitioners because there exists no friendly, professionally developed and validated software implementing them on a popular platform such as Microsoft Windows. This Phase II SBIR proposal is intended to fill this void. The goal is to complete the development of WinEaSt, a Windows package for the design and interim monitoring of group sequential clinical trials. When completed, WinEaSt will: (1) Perform group-sequential inference for the parameters of interest from practically any type of mathematical model including generalized linear models, parametric and semi-parametric survival models and mixed models for longitudinal studies. (2) Permit adaptive re-estimation of sample size, accrual period, total study duration and other design parameters, at each interim monitoring time-point using information-based monitoring.
NOT AVAILABLE
