A prospective study of the use and effects of screening mammography in women aged 70 and older.

Abstract

To prospectively examine how age and comorbidity affect screening mammography use, cancer detection, and overall survival. Prospective, with median follow-up of 10.2 years. A population-based cohort of Vermont women with data in the Vermont Breast Cancer Surveillance System. Women aged 70 and older with no history of breast cancer (N = 20,697). Rates of screening, diagnostic procedure use, and breast cancer diagnosis were examined according to age and comorbidity. The effect of breast cancer on overall survival was assessed in relation to detection mode, tumor characteristics, and treatment. Screening declined 9% per year after age 70 and 18% with each unit increase in comorbidity score, with corresponding increases in clinically detected breast cancer. Invasive cancer was associated with greater overall mortality (screen detected: hazard ratio (HR) = 1.22, 95% confidence interval (CI) = 1.07-1.40; clinically detected: HR = 1.68, 95% CI = 1.43-1.96). The latter HR reflects a much greater absolute risk of death for women in a population with high baseline mortality. Use of breast-conserving surgery as the only treatment for Stage I cancer increased markedly with age and was associated with shorter overall survival than radiation or mastectomy (relative risk = 2.23, 95% CI = 1.42-3.47). Lower screening mammography use by older women is associated with more clinically detected breast cancers, which are associated with poorer survival. Treatment received for early-stage cancer influences the effect of screening on survival.