A graphical method to assess treatment-covariate interactions using the Cox model on subsets of the data

摘要: We introduce the subpopulation treatment effect pattern plot (STEPP) method, designed to facilitate interpretation of estimates derived from different but potentially overlapping subsets clinical trial data. In particular, we consider sequences subpopulations defined with respect a covariate, and obtain confidence bands for collection effects (here obtained Cox proportional hazards model) associated sequences. The method is aimed at determining whether magnitude changes as function values covariate. apply STEPP breast cancer data set evaluate oestrogen receptor content primary tumour.