Cytel statisticians are looking foward to attending the Conference of the International Society for Clinical Biostatistics, which will be held in Vienna during the week of August 24th. Members of Cytel will be contributing to four sessions at this conference, including an invited talk on innovation entitled 'Beyond Wild Horses: Developing Innovation at Cytel.' They will also be contributing to a session called Adaptive Designs II, in which they will discuss Backward Image Confidence Intervals, a solution to the problem of parameter estimation at the end of an adaptive trial.
Sheetal Solanki and Namrata Deshpande will explain the method that was developed by Cytel Founder Cyrus Mehta and Cytel Statisticiain Lingyun Liu, in conjunction with Dr. Ping Gao of The Medicines Company. They will then provide several examples of how to apply this novel technique. The abstract of their talk is available below, as is the published paper.
Abtract of ISCB Presentation:
Backward Image Confidence Intervals for Adaptive Group Sequential Designs
An adaptive trial is defined as any clinical trial which uses accumulating data, possibly combined with external information, to modify aspects of the design without undermining the validity and integrity of the trial. Müller and Schäfer provided a methodology for conducting an adaptive
trial which guaranteed control of type 1 error while providing maximum flexibility.
However, corresponding solutions for the equally important and related problem of parameter estimation at the end of the adaptive trial have not been completely satisfactory. In their paper “Exact inference for adaptive group sequential designs” (April 2013), Ping Gao, Lingyun Liu and Cyrus Mehta introduce a method called Backward Image Confidence Intervals (BWCI) which is based on mapping the final test statistic obtained in the modified trial into a corresponding backward image in the original trial. It computes a two-sided confidence interval having exact coverage, along with a point estimate that is median unbiased for the primary efficacy parameter in a two-arm adaptive group sequential design.
This method will be discussed here with the help of several examples. Along with it, we will discuss advantages of this procedure over previously available methods, which either produced conservative coverage or no point estimates or provided exact coverage for one-sided intervals only. We will use simulation results generated by Cytel’s software East® for this purpose.
Paper:
Exact inference for adaptive group sequential designs
Methods for controlling the type-1 error of an adaptive group sequential trial were developed in seminal papers by Cui, Hung, and Wang (Biometrics, 1999), Lehmacher and Wassmer (Biometrics, 1999), and Müller and Schäfer (Biometrics, 2001). However, corresponding solutions for the equally important and related problem of parameter estimation at the end of the adaptive trial have not been completely satisfactory. In this paper, a method is provided for computing a two-sided confidence interval having exact coverage, along with a point estimate that is median unbiased for the primary efficacy parameter in a two-arm adaptive group sequential design. The possible adaptations are not only confined to sample size alterations but also include data-dependent changes in the number and spacing of interim looks and changes in the error spending function. The procedure is based on mapping the final test statistic obtained in the modified trial into a corresponding backward image in the original trial. This is an advance on previously available methods, which either produced conservative coverage and no point estimates or provided exact coverage for one-sided intervals only. Copyright © 2013 John Wiley & Sons, Ltd.
