Inferences for the Lead Time in Breast Cancer Screening Trials under a Stable Disease Model

We estimated the effects on the lead time for women participating in a long-term breast cancer screening program when the screening sensitivities and disease progression are independent of age. The lead time, or time by which a diagnosis is advanced by screening, is one of the major concerns of any cancer screening program, which we consider to include both a mammogram and physical examination. Using estimates of test sensitivities and mean sojourn times previously calculated by other authors from observed data, we estimated properties of the lead time. We utilized the model for the lead time derived by other authors and ran simulations for different screening program designs, concentrating on screening interval lengths of 0.5 years, 1 year, 1.5 years, and 2 years. These estimates were based on a long-term screening program from age 50 to 80. For each six-month decrease in screening interval length, we estimated the percent increase in mean lead time, as well as the percent increase in the proportion of clinical patients who will have their cancer detected at a screening exam.