Screening Mammography for Women in Their Forties

Article type: Research Article

Authors: Berry, Donald A.

Affiliations: Institute of Statistics and Decision Sciences, Duke University, ISDS Box 90251, Durham, NC 27708-0251, USA

Abstract: Nonrandomized screening studies and some analyses of randomized data are subject to enormous biases. Many such analyses have the goal of showing that screening mammography is beneficial. They masquerade as science but are extreme examples of lying with statistics. Randomized trials have provided mixed results concerning whether screening reduces breast cancer mortality. Estimates from these trials are subject to a great deal more variability than traditional meta-analyses belie. A blanket recommendation for the screening of women in their forties is inappropriate, and it is a disservice to these women. Regular screening mammography of women in this age group is not an imperative health measure. If it is beneficial, its benefits are limited: taking the results of the randomized trials at face value, regular screening adds about five days in life expectancy per woman. In addition, screening has many significant risks.Those who are most concerned about their health will probably discount the risks and choose regular screening, and reasonably so. Others will eschew screening, also reasonably so.

DOI: 10.3233/BD-1998-103-405

Journal: Breast Disease, vol. 10, no. 3-4, pp. 23-32, 1998