Overdiagnosis in breast cancer screening: women have minimal prior awareness of the issue, and their screening intentions are influenced by the size of the risk
- Correspondence to
: Dr Mette Kalager
Harvard School of Public Health, 677 Huntington Avenue, Boston, MA 02115, USA;
Implications for practice and research
Women have little knowledge about overdiagnosis of breast cancer.
If the overdiagnosis rate is 50%, some women will consider declining breast cancer screening.
Doctors need to be aware of the issue of overdiagnosis and women need access to balanced information.
There has been increasing awareness recently of the problems of overdiagnosis in breast cancer screening. The amount of overdiagnosis varies across different studies. The aim of this study by Hersch and colleagues was to generate insight into women's understanding and interpretation of overdiagnosis of breast cancer screening. Further, to understand how different estimates of overdiagnosis influence screening attitudes and how women view information given on overdiagnosis. Overdiagnosis was defined as the detection of ‘a cancer that would have neither caused death nor presented clinically during the woman's lifetime’.
Women aged 40–79 years with different socioeconomic backgrounds, varying levels of education and living in the Sydney suburbs were eligible for the study. There were 50 women included in eight focus groups where an audiovisual presentation of the benefits of mammography screening, and three different estimates of overdiagnosis (1–10%, 30% and 50% overdiagnosis among women participating in mammography screening) were presented and discussed. The women were encouraged to seek clarification as often as they needed. In addition to the focus groups, a written questionnaire including demographics, preferences on mammography screening and participants' understanding of overdiagnosis was given at the start and the end of each focus group.
Participants had diverse cultural and educational backgrounds. Among women older than 50 years, 77% had previously been screened, compared to 37% below 50 years. Many women were not aware of overdiagnosis, but most women came to understand the issue during the study. The estimate of 50% overdiagnosis made some women perceive the need for more careful personal decision-making about screening, whereas the lower estimates had no impact on intended screening behaviour. Information preferences varied—many considered it to be important to take overdiagnosis into account and make informed choices, but many women wanted to be encouraged to be screened.
This is an interesting and well-conducted study. Not surprisingly, many women were not aware of overdiagnosis and raised concerns that there was a hidden governmental plan to shut down mammography screening. Some were struggling to understand that there are no means of distinguishing between cancers that are overdiagnosed versus those that are fatal. There was disagreement about the notion that overdiagnosis could be ‘a necessarily negative occurrence’. Additionally, some women would rather consider watchful waiting if they were diagnosed with breast cancer, than consider to decline screening. Women seemed to accept screening if the amount of overdiagnosis was 30% or less. This means that they accepted that for every five women treated unnecessarily we save one from dying of breast cancer. This is an interesting outcome. We know that both doctors and women are overestimating the benefits and underestimating the harms of mammography screening1 ,2; thus, it appears that both the public and the medical community are misinformed.3
The authors conclude that it is vital to determine optimal methods to support informed choice in mammography screening. However, as it appears, the default is that mammography screening is beneficial. This unbalanced message of mammography screening creates a potential larger threat to informed choice than different estimates of overdiagnosis. The finding that, only at a rate of 50% overdiagnosis women might reconsider their view on mammography screening, is worrisome. The worst possible consequence of cancer treatment is death, thus, the potential benefit of reduced breast cancer mortality might be dwarfed by higher mortality among women overdiagnosed (randomised trials of mammography screening showed no impact on overall mortality).4 For women of all ages, the risk of breast cancer exceeds the risk of dying from breast cancer, so a relative increase of 30% for both breast cancer death and overdiagnosis means that a woman's absolute risk of overdiagnosis is 5–6 times higher than her risk of dying from breast cancer.