A Partly Conditional Survival Approach for Modeling the Time-varying Association Between Breast Density and Breast Cancer Risk

PI Name: Cook A
Title: A Partly Conditional Survival Approach for Modeling the Time-varying Association Between Breast Density and Breast Cancer Risk
Institute: BCSC Statistical Coordinating Center

The investigators propose to develop a series of models using the Zheng and Heagerty partly conditional survival approach for modeling breast cancer risk as a function of time-varying covariates observed in the BCSC data and using these models to analyze data for the two proposals currently in queue (AB68KK and AB70KK). This will require developing a scientifically plausible, but flexible, model for the association between breast density and breast cancer risk that changes over time, age, and menopausal status. The investigators will need to derive the estimating equations for this model and program the fitting algorithm. Once this model is developed, they will be able to easily apply it to other proposals that use a cohort design to study breast cancer risk. To accomplish these goals, the investigators are requesting funds to cover 0.20 FTE for one year of a PhD biostatistician's time, Andrea Cook. Andrea's recent training in survival analysis and stochastic processes make her an ideal person to take on this project. Diana Miglioretti and Karla Kerlikowske will be collaborators on this project; however, their time will be covered by the BCSC grants at their sites. Given the large amount of biostatistician time required for model development and programming, this supplemental funding will allow the investigators to move forward with these important projects in a more timely fashion.