Evidence on screening for breast cancer from a randomized trial


  • Sam Shapiro

    Director, Corresponding author
    1. Health Services Research and Development Center, and Professor of Health Services Administration, The Johns Hopkins University, Baltimore, MD
    • Health Services Research and Development Center, The Johns Hopkins University, Baltimore, MD 21205
    Search for more papers by this author


Results in the breast cancer screening project of the Health Insurance Plan of Greater New York that started the end of 1963 have been updated through December 31, 1975. The HIP study is a randomized trial designed to test whether periodic screening with clinical examination and mammography results in reduced breast cancer mortality among women aged 40–64 years at the start. Study women were offered screening examinations; 65% appeared for initial examinations and a large majority of these women had at least one of the three additional screenings at annual intervals. The control group of women continued to receive their usual medical care. New data support earlier reported results on benefits. During the nine years following date of entry there were 128 breast cancer deaths in the control group as compared with 91 in the study group (screenees plus refusers). The impact of the screening program continues to be confined to women 50 years of age and over with no benefit at ages 40–49 years. Seven-year case fatality rates show similar relationships. Several issues related to screening benefit are considered. Of major importance is the observation of no reduction among women under 50 in breast cancer mortality. The possibility that under different screening conditions (e.g., with current mammography equipment) a benefit would be found needs to be investigated. There is a clear need for rapidly determining whether a new randomized trial is the only way to answer the question and whether experience in the 27 NCI-ACS demonstration projects can provide useful data. Another critical issue concerns the incremental value of mammography in a screening program. Over an eight-year period after diagnosis, breast cancer cases that were positive only on mammography when screened had a case fatality rate of 14%; this compares with 32% for cases positive only in the clinical examination and 41% for cases positive on both modalities. Excluding mammography would have reduced the benefit of screening by an estimated one-third. With regard to risk associated with screening, it is concluded that the increment in risk resulting from radiation exposure in mammography does not offset the benefits of screening above 50 years of age. Below that age, although the risk increment is small, the risk-benefit balance is negative because of the absence of a demonstrated benefit. Another source of risk is related to the possible increase in biopsies. Assessment of the HIP experience suggests that only timing of biopsies was affected, but the potential for considerable variation if screening is adopted widely exists. Based on current findings in the HIP study, there appears to be strong support for periodic screening at ages 50 years and over with clinical examination and mammography; to justify screening under 50, new information from other studies is required.