AN APPROACH TO ESTIMATING EXPOSURE-SPECIFIC RATES OF BREAST-CANCER FROM A 2-STAGE CASE-CONTROL STUDY WITHIN A COHORT

The Breast Cancer Detection and Demonstration Project (BCDDP) included a large cohort of women followed for incidence of breast cancer and f rom whom an initial case-control sample was drawn and standard risk fa ctors obtained. In order to study the effect of mammographic features on breast cancer risk, a nested subsample of cases and controls was dr awn. Therefore, these data can be viewed as two-stage case-control dat a within a cohort, or as cohort data with two nested levels of missing ness, since basic characteristics like age were measured on all member s of the cohort, standard risk factors were elicited only in the initi al case-control sample, and mammographic features were assessed only i n the nested subsample of cases and controls. We present a Poisson pse udo-likelihood approach to estimating age- and exposure-specific breas t cancer incidence rates based on the three types of variables, This a pproach takes into account the nested missingness as well as two other type of missingness, namely, that for basic variables and standard ri sk factors, some levels (i) were omitted by design in the nested subsa mple of case and controls or (ii) were empty because of the sparsity o f the data in that subsample. Estimates of standard errors are obtaine d from a parametric bootstrap. The approach seems to be efficient when applied to the BCDDP data and is flexible for modelling breast cancer rates and taking the special missingness features of these data into account.