THE NATIONAL BREAST AND CERVICAL-CANCER EARLY DETECTION PROGRAM - REPORT ON THE FIRST 4 YEARS OF MAMMOGRAPHY PROVIDED TO MEDICALLY UNDERSERVED WOMEN

OBJECTIVE. We describe results from 284,503 mammographic examinations and associated diagnostic workup provided to medically underserved wom en in an ongoing nationwide breast cancer early detection program. MAT ERIALS AND METHODS. We report the results of mammographic examinations and diagnostic workups on 230,143 medically underserved women 40 year s old or older who underwent at least one mammographic examination fro m July 1991 through June 1995. Mammograms were obtained in hundreds of mammography and clinical facilities throughout the United States, inc luding community health centers, health department clinics, private pr actitioners' offices, university-based facilities, and mobile mammogra phy units. Our analysis included rates of mammograms with abnormal fin dings (reported according to the categories of the American College of Radiology Breast Imaging Reporting and Data System), breast cancer de tection rates, numbers of diagnostic procedures performed, stage and s ize distribution of breast cancers, and positive predictive value of m ammograms and biopsies with abnormal findings-all presented according to screening round and 10-year age intervals. RESULTS. Mammograms with abnormal findings constituted 5% of mammograms in the first round and 4% in subsequent rounds, both proportions declining by approximately one third from the youngest (40-49 years) to the oldest (70 years and older) age group. Breast cancer detection rates per 1000 mammographic examinations were 5.1 for the first round and 2.0 for subsequent round s; from the youngest to the oldest age group, the first-round rates do ubled and the subsequent-round rates tripled, Early-stage cancers acco unted for 54% of first-round cancers and 81% of subsequent-round cance rs. Percentage of invasive cancers at least 2 cm in size declined from 51% in the first round to 33% in subsequent rounds; however, we found little change in the proportion of lesions smaller than 1 cm. Positiv e predictive values declined from 9.5 cancers per 100 mammograms with abnormal findings in the first round to 5.6 cancers per 100 mammograms with abnormal findings in the subsequent rounds. CONCLUSION. A large nationwide breast cancer early detection program conducted through hun dreds of diverse facilities has provided results that, although not a statistically representative sample of mammography services, are proba bly the best available characterization of the current state of breast cancer screening practices as they actually occur in the 1990s in the United States, These results should be useful to clinicians, research ers, and public health personnel in counseling patients, planning new studies, and improving efforts to control breast cancer.