Screening mammography and breast cancer treatment patterns in older women

Objective. To examine the impact of mammography screening on treatment opti ons received by a cohort of older breast cancer patients. Setting and population. We studied 718 newly diagnosed breast cancer patien ts, 67 years and over, diagnosed with TNM Stage I and II disease between 19 95 and 1997 at 29 hospitals in five regions. Methods. Data were collected from patients, surgeons, and medical records. A breast cancer diagnosis was considered to have been by screening mammogra phy if so reported by both patient and medical records. Bivariate and logis tic regression were used to identify predictors of a women having her cance r detected by screening mammography and the relationships between mode of d etection, stage of disease at diagnosis, and local treatment. Results. Women with high school or greater education were 1.75 times (95%, CI 1.11-2.75) more likely to have their cancers diagnosed by screening mamm ography than women who had not completed high school, controlling for other factors. Screening found earlier stage disease: 96% of women with mammogra phically diagnosed cancer had T1 lesions, compared to 81% of women diagnose d by other means (p = 0.001). Women with mammography detected lesions were more likely to have ductal cancer, and to be referred to radiation oncologi sts more than women diagnosed by other means. Controlling for stage and his tology, screening remained associated with a higher likelihood of receiving breast conserving surgery (BCS) with radiation (RT) (OR 1.56, 95%, CI 1.10 -2.22) than other local therapies. Conclusions. Beyond the impact on stage, ductal cancers were more likely to be diagnosed by screening. Mammographically detected lesions were associat ed with referrals to radiation oncologists and higher rates of BCS and RT. Research is needed to explain the residual independent effects of mammograp hy screening on breast cancer treatment.