This course will focus on biomarkers that are used to guide treatment decisions—called predictive, prescriptive, or treatment selection biomarkers.

There is a heightened interest in treatment selection biomarkers, especially in clinical settings where the treatment is toxic or costly, or where the condition being treated is thought to be heterogeneous. These might be something as simple as subject demographics or clinical characteristics, or the results of gene expression technology or medical imaging.

This course will describe preferred statistical approaches to evaluating treatment selection biomarkers, with a focus on the ideal randomized trial setting. Recent methodology for screening candidate biomarkers and for deriving biomarker combinations and treatment rules will be reviewed. Study designs for discovering and evaluating treatment selection biomarkers will be compared and contrasted.

Students will receive training on using R packages for biomarker evaluation and for combining biomarkers and developing treatment rules. Several datasets will be provided and will be used to illustrate methods throughout the course.