Clinical trials are often carried out in a sequential fashion, for the reason of ethics, efficiency and costs. At the end of a sequential trial, results such as response rate of a treatment, side effects and important clinical endpoints are often reported. But because of the sequential nature of a sequential trial, conventional methods of presenting estimates are invalid. Currently, the methods that are available to answer to this need are either difficult to apply or possess questionable qualities. This has led to the slow acceptance of these methods in the reporting of results in sequential studies. The primary goal of this proposal is to define, in a quantitative and succinct way, practical methods of presenting parameter estimates from a sequential clinical study. The fundamental theoretical basis of this proposal is that the characteristics of the observed data in a study can be studied through random samples drawn from the observed data. Therefore, any inadequacies of using conventional estimation methods will be reflected and can be corrected.
The specific aim of the proposal are to study the behavior of this method in the context of giving point and interval estimates in sequential trials. The small sample properties of this method will be studied through simulations and the large sample properties of this method will be studied through theoretical analysis. The application of this method to group sequential trials will be explored, with special emphasis on tissues such as non-uniform accrual and times of testing. The use of this method in multiple reporting, as in the case of interim analysis will be investigated. Finally, the method will be applied to studies at Memorial Sloan-Kettering Cancer Center.
| Leung, D H; Wang, Y (2001) Isotonic designs for phase I trials. Control Clin Trials 22:126-38 |
| Leung, D H; Wang, Y G (2001) A Bayesian decision approach for sample size determination in phase II trials. Biometrics 57:309-12 |
| Leung, D H; Wang, Y G (2001) Optimal designs for evaluating a series of treatments. Biometrics 57:168-71 |
| Wang, Y G; Leung, D H (1998) An optimal design for screening trials. Biometrics 54:243-50 |
