Proper inference from Simon's two-stage designs.

摘要: Simon's two-stage designs are very popular for phase II clinical trials. A literature review revealed that the inference procedures used with almost always ignore actual sampling plan used. Reported P-values, point estimates and confidence intervals response rate not usually adjusted design's adaptiveness. In addition, we found sample size second stage is often different from planned. We present here a method inferences using both planned sizes. The conventional preferred yield similar P-values rate. inference, however, may contradict result of corresponding hypothesis testing.