The asymptotic theorems of adaptive designs for clinical trials with generations depending on the estimated success rates
Feifang Hu,Li-Xin Zhang
2004-01-01
Abstract:Abstract The randomized play-the winner (RPW) rule is very useful in clinical trials for patient allocation with two treatments. Wei (1979) introduces the generalized Friedman’s urn (GFU) model to clinical trials of K treatment (as an extension of the RPW). In this paper, we consider a general kind of adaptive designs, whose generations depend on an estimated parameter, for example the estimated success rate. Using this kind of adaptive designs, more informations of previous stages are used to update the model at each stage, and more patients may be assigned to a better treatment than the RPW rule and Wei’s rule. The strong consistency and the asymptotic normalities are established for both the urn composition and the number of patients assigned to each treatment. A adaptive design proposed by Bai, Hu and Shen (2001) recently is a simple special example. 1 Introduction and main results 1.1 Introduction and the Model In comparing of K treatments, suppose that patients are available for the study sequentially and must be assigned to a treatment group immediately upon arrival. Adaptive designs, which use accumulating information to assign patients to treatments, tend to assign more patients to better treatments in clinical trials. A typical probability model for adaptive clinical trials is the Generalized Friedman’s Urn (GFU) model (also named as Generalized P´olya Urn (GPU) in literature). The model is described as follows. Consider an urn containing particles of K types, respectively representing K ’treatments’ in a clinical trial. The patients are to be allocated sequentially to the treatments. At the beginning, the urn contains Y0 = (Y0;1;:::;Y0;K) particles, where Y0;k > 0 denotes the number of particles of type k, k = 1;:::;K. At stage i, i = 1;2;:::, a particle is drawn from the urn and replaced. If the particle is of type k, then the treatment k is assigned to the i-th patient, k = 1;:::;K, i = 1;2;:::. After observing a random variables »i;k which may be considered as the response of the i-th patient on the treatment k, an additional Di(k;q;»i;k) particles of type q,