This ecological study (comparing groups, rather than individuals) covered the period of January 1, 2000 to December 31, 2010 and analyzed data from the Surveillance, Epidemiology, and End Results (SEER) cancer registries. The study population included more than 16 million women, aged 40 years or older, and residing in 547 U.S. counties in the year 2000. Of these women, 53, 207 were diagnosed with breast cancer that year, and were then followed for the next 10 years.

Although the extent of screening correlated with an increased incidence of breast cancer across counties, it did not correspond with a decline in deaths from breast cancer. Instead, the increased screening contributed to the detection of small cancers. There was no decrease in the incidence of larger cancers or any significant differences in mortality. The number of deaths remained the same among women who were routinely screened with mammograms compared to those who were not.

Interpretations and Implications

This ecological study of breast cancer is important because overdiagnosis (the detection of cancers that will not cause death or symptoms) is not directly observable in individuals. Thus, it is crucial to analyze patterns and frequencies. In addition, it is the first study to compare such a large number of counties with diverse screening rates. Despite the variation in routine mammography screens, the association between screening and breast cancer did not vary substantially by county population.

It is unclear why screening is not associated with the detection of larger breast cancers, and the researchers point out that longer follow-up (13 years) could reveal a reduction in mortality that is not apparent at ten years. There is also no way to know whether the women who died of breast cancer during the 10 years of study were the ones diagnosed with breast cancer during the study period.

A Cochrane review of the most reliable randomized controlled trials (RCTs) revealed that the most realistic estimate is a 10-15 percent relative reduction in breast cancer mortality in the trials, and overdiagnosis at 30 percent. One of the largest and longest-running studies of screening mammography, the Canadian National Breast Screening Study, is an RCT that compared breast cancer incidence and mortality for women ages 40 to 59 who did, or did not, have an annual screening mammogram. This study followed nearly 90,000 women for up to 25 years. It found that, compared with clinical exam or routine care, annual mammograms did not reduce breast cancer deaths at all. Taken together, these results corroborate the county-by-county findings of no difference between the screened and unscreened women.

In light of the new findings that the “most prominent effect of screening mammography is overdiagnosis,” Harding and colleagues argue that the “balance of benefits and harms is likely to be most favorable when screening is directed to those at high risk, provided neither too frequently nor too rarely, and sometimes followed by watchful waiting instead of immediate active treatment.” This statement does not address the crucial gap in the evidence about who is at high risk for what type of breast cancer and therefore, when and what type of screening would be appropriate.

To close the gap, a research team at UC San Francisco led by breast cancer researcher Laura Esserman, MD, MBA, professor of surgery and director of the Carol Franc Buck Breast Care Center at the UCSF Helen Diller Family Comprehensive Cancer Center will conduct a 5-year study (called the WISDOM study) of 100,000 California women 40 to 80 years old to assess their individual risk of breast cancer, using that information to suggest when to start screening, with what frequency, and when to stop.

Source: Harding C, Pompei F, Burmistrov D, Welch H, Abebe R, Wilson R. Breast Cancer Screening, Incidence, and Mortality Across US Counties. JAMA Intern Med. Published online July 06, 2015.