Benefit and Risk of Mammography Screening – Considerations from an Epidemiological Viewpoint: Discussing the Benefit-Harm Debate

by Prof. Dr. rer. nat. Nikolaus Becker, Prof. Dr. med. Hans Junkermann in volume 08/2008

Becker and Junkermann's review article sheds light on a very important matter of health policy. Yet the problems begin as early as the title, which sets up a contrast between benefit and risk rather than between benefit and harm. The impression is conveyed that the benefit of mammography screening stands against a (merely statistical) risk. This, of course, is not the case; both the "benefit" and the risk are statistical extrapolations of an expected benefit and an expected harm. We consider the authors' assessment of benefit and harm to be imbalanced for a number of other reasons as well:

The authors illustrate the benefit of screening with a fictitious sample of 100 000 women who undergo screening every 2 years a total of 10 times. The harm, however, is illustrated with a fictitious group of only 1 000 women. For a group of 100 000 women in which screening would lead to the prevention of 540 deaths from breast cancer, the following types of harm would be expected to occur:

- 22 300 to 36 300 women would have false positive mammograms,
- 500 women would be overdiagnosed with breast cancer,
- 6300 women would have a false positive indication for a breast biopsy,
- about 500 women would go on to have a breast operation with a benign histological finding,
- and there would be 10 to 240 cases of radiation-induced breast cancer.

Thus, every third woman undergoing screening every 2 years between the ages of 50 and 69 would be subjected to unnecessary worry, many women would undergo invasive diagnostic or therapeutic procedures on the basis of a false positive finding or an overdiagnosis, and a not inconsiderable number of breast cancers would actually be induced. There is no critical discussion of the quality of life of these unnecessarily disturbed women and their families; such a discussion is indispensable if the benefits and harms of screening are to be considered fairly.

Thus, overall, the benefit-harm debate seems to lack balance. DOI: 10.3238/arztebl.2008.0421a

Prof. Dr. rer. nat. Karl-Heinz Jöckel
Dr. med. Barbara Hoffmann, MPH
Institut für Medizinische Informatik, Biometrie
und Epidemiologie
Medizinische Fakultät
Universität Duisburg-Essen
Hufelandstr. 55, 45147 Essen, Germany
[e-mail]

Prof. Dr. med. Andreas Stang, MPH
Sektion Klinische Epidemiologie
Institut für Medizinische Epidemiologie, Biometrie und Informatik
Medizinische Fakultät
Martin-Luther-Universität Halle-Wittenberg
Magdeburger Str. 8, 06097 Halle (Saale), Germany