Clinical Question: Does screening mammography result in the identification and treatment of breast cancers that would never be apparent clinically?

Setting: Population-based

Study Design: Other

Allocation: Concealed

Synopsis: There are two possible harmful outcomes of screening for breast cancer. The first is a false-positive result, telling a woman she may have breast cancer only to find out on biopsy that she does not. A more subtle negative outcome is the risk of overdiagnosis, in which breast cancer is detected by mammography in a woman who would otherwise have lived her life never knowing that she had breast cancer. The authors conducting this analysis used the results of the Malmo mammographic screening trial. Over 10 years, 42,283 women were enrolled in the trial and randomized to receive mammography or no mammography. Follow-up continued for another 15 years.

As would be expected, significantly more women receiving mammography were identified as having breast cancer than were non-screened women. However, there continued to be more diagnoses of breast cancer in the screened group than in the nonscreened group, even though during follow-up women in both groups had regular mammography. Because any woman diagnosed with breast cancer was no longer part of the screening group, the initially unscreened women should have caught up and had more breast cancers identified when they started regular mammography. That they did not indicates that overdiagnosis occurred in the women initially assigned to the screening group.

Overall, about 10 percent (95% confidence interval, 1 to 18 percent) of the women who had mammography and were diagnosed with breast cancer were overdiagnosed, meaning they would never have known they had breast cancer—or have received treatment—except for the screening program.

Bottom Line: In addition to the possibility of having a false-positive result on mammography sometime in her life, one in 10 women who has a real diagnosis of breast cancer undergoes treatment that, although seemingly curative, ultimately has no beneficial effect on her because she would never have developed clinically apparent breast cancer. (Level of Evidence: 2b)