The choice of treatment for cancer patients based on covariate information.

摘要: This paper discusses the analysis of data from clinical trials in an effort to determine whether comparisons treatment various subsets patients yield sufficiently different results justify idea that there may be optimal for each patient based on his individual characteristics. approach belongs more field exploratory than classical hypothesis testing. The treatment-covariate interactions is discussed and methods detecting them are presented using parametric survival models incorporating covariate information. A detailed example a trial estrogen prostatic cancer presented. In this study significant were detected. Subsidiary analyses indicated young with high grade tumors should have been treated estrogens, but older low harmed by treatment.