Do mammograms save lives?

There is no evidence that breast-cancer screening with mammography saves women's lives, claims a report in the Lancet. Danish researchers who published that controversial conclusion last year have re-analysed their data, and confirmed their original conclusions.

Last year's report was criticised by breast cancer experts who questioned the researchers' reasoning and conclusions. So researchers in Copenhagen re-analysed their data according to the protocol of the Cochrane Collaboration, an international organisation that has established rigorous standards for conducting and publishing research reviews.

Their conclusion about mammography remains the same: the studies that have indicated mammography saves women's lives by catching breast cancer early employed flawed methods.

They reviewed each of seven large mammography trials, involving half a million women, and found:

Five of them were either of poor quality or so flawed they had to be discounted altogether.

The two remaining studies, which they said were of "medium quality," found no reduction in breast cancer deaths.

In contrast, the three studies considered poor quality, reported on average a 32 percent reduction in breast cancer mortality.

In their original study, the researchers cited problems with the way many mammography trials have been conducted – including imbalances in terms of the women's ages and other factors that they charge skewed the study results in favour of mammography.

But public health experts say this re-analysis doesn't refute the criticisms that have been lobbed at the Danish researchers. The American College of Physicians still recommends that women get a mammogram every year starting at age 40.

Breast cancer death rates have declined in the U.S. and U.K. in the 1990s. While better cancer treatment certainly contributed to this decline, experts say early detection made a significant difference.

Source: Ole Olsen and Peter C Gøtzsche, Cochrane review on screening for breast cancer with mammography, Lancet, October 20, 2001; Study recharges debate on the value of mammograms, Reuters Health, October 19, 2001.

For Lancet text (free registration required)
http://www.thelancet.com/journal/vol358/iss9290/full/llan.358.9290.original_research.18001.1
For more on Health http://www.ncpa.org/iss/hea/

FMF\30 October 2001