Besides, screening studies must be careful to screen out the so-called "screening effect." Women who are healthier, more educated and eat more fruits and vegetables, and thus less likely to get sick in the first place, are more likely to attend screenings when invited to them.

The Danish group got their results using breast cancer deaths from 1997-2006 pulled from a death registry hosted by the country's National Board of Health.


Acknowledged Criticisms

In the paper, Jorgensen and his co-authors admit to some built-in weaknesses of the study. It looked at a cohort of women, not individuals, so there was no way to track women who moved between screened and non-screened areas, which could distort the mortality rates. But the authors conclude this would be "unlikely." They also doubted that different regions have different risk factors, such as variations in the use of hormone replacement therapy, a known breast cancer risk, because Denmark is a homogeneous country.

However, other critics, including some contacted by DOTmed News, have observed what they think are fatal flaws in the paper. They note that about half of all breast cancer deaths over a 10-year period are caused by cancers diagnosed before the period began. And the current study includes no mechanism for reliably parceling out deaths from cancers diagnosed after screening started, which could be lower.

Critics also note that while the paper implies that it compares same-age women from screening and non-screening groups, in fact the paper only looks at people from areas where screenings were offered so the methodology does not allow researchers to truly match up women who actually had a screening mammogram with those who didn't.

U.S. Evidence

Across the Atlantic, the study is sure to fan the flames ignited last year by a controversial United States Preventive Services Task Force report. The American government-convened group concluded that breast cancer screening had only a modest impact on mortality rates, and recommended raising the age of screening in the United States to 50 and only having screenings once every two years, instead of once every year, bringing the practice in line with most countries in Europe, such as Denmark.

But a paper cited in the Danish article, using statistical models, found a median 15 percent decline in mortality rates in the U.S. due to breast cancer screening, a result Jorgensen feels agrees with other assessments. However, Jorgensen cautions that the statistical models could rely on false assumptions.

More Industry Headlines