| 简介： | Abstract: In biomedical studies, it is often of interest to estimate how the outcome variables (either the efficacy outcome such as survival or risk profile of an adverse event) are related to an intervention and other related biomarker variable.For example, in randomized controlled clinical trials of bivalent human papillomavirus (HPV) vaccine,investigators are interested to know how miscarriage rate relates to the timing of HPV vaccination. We developed hierarchical Bayesian Biomarker Threshold Models to make simultaneous inference on both the cut-points of the biomarker variable and the magnitude of the biomarker-treatment interaction. Hierarchical priors are proposed and used in Markov Chain Monte Carlo for statistical inference. We further implement the proposed method in an R package for Biomarker Threshold Models (‘bhm’). Several clinical trials examples will be demonstrated how to use the ‘bhm’ package to analyze outcome from linear models, generalized linear models and survival models. |