First, I would like to thank the authors for putting together this important survey paper. They have covered a wide range of methods that are applied in dose ranging studies, with the emphasis on model-based techniques, from Bayesian modeling to multiple comparison-modeling approach to model-based optimal designs to nonparametric estimation techniques. These methods are of personal interest to me and I expect that the paper will generate a lot of interest from the readers of the journal. Still there is one specific area which, in my opinion, should have been covered in more detail - namely, what the authors call "D-optimal response-adaptive", or "Dopt approach". I understand that the authors address a variety of methods, and the goals of this paper are different from providing a comprehensive review of each method. However a complete lack of references on optimal model-based designs seems rather strange.
展开▼
