The role of HRT exposure in breast cancer etiology: a causal inference approach

Cancer etiology is a topic of great interest in medical research. Assessing and quantifying causal links between an exposure (e.g. a treatment) and an outcome of interest (such as cancer prognosis or survival) is often not a trivial task. Methods have been proposed to deal with the problem ofbias in estimation of causal effects. A method is presented here, applied to the field of Hormone Replacement Therapy and Breast Cancer, which allows to obtain an (at least locally causal) estimate of a continuous exposure effect on an outcome of interest within the framework of Principal Stratification, under a case-control study design.