Mammography screening study praised, criticized: Some hail its size and scope, but others say it has critical design flaws. (45% Decrease in Breast CA Mortality).

A recent large-scale Swedish study linking routine screening mammography to a 45% decrease in breast cancer mortality is being criticized by some experts as having major flaws that undermine its findings.

The retrospective observational study, the latest entry in the ongoing debate over the risks versus benefits of screening mammography, examined the effect of organized mammography service screening in seven Swedish counties comprising a third of the country's eligible female population.

In an editorial accompanying he study, Dr. Stephen Feig said the impressive size and scope of the study make its results irrefutable. But other experts interviewed for this article criticized the study for its design and for not taking into account improvements in breast cancer treatment efficacy over time.

The study compared mortality from breast cancers diagnosed in the periods before and after screening was available and included data from four decades. Before screening was offered there were 5,728 breast cancers and 1,169 deaths from the disease. After the screening program was established there were 8,364 cases but only 875 deaths, reported Stephen Duffy of the department of mathematics, statistics, and epidemiology at Cancer Research UK, London, and his colleagues (Cancer 95[3]:458-69, 2002).

Among women who took advantage of the screening, there was an overall 45% reduction in breast cancer mortality. Up to 25% of the eligible women did not take advantage of the screening when it was offered. When these women were included in the analysis, the overall breast cancer mortality rate reduction was about 30%.

In his editorial, Dr. Feig, director of breast imaging at Mount Sinai Medical Center in New York, wrote: "Already proven by randomized clinical trials, breast cancer mortality reduction through screening mammography has now been confirmed with organized service screening programs" (Cancer 95[3]:451-457, 2002).

But those who have questioned the findings of previous mammography studies aren't convinced this one ties up the debate. Dr. Peter C. Gotzsche, director of the Nordic Cochrane Center in Copenhagen, said in an interview that the Swedish study contains a bias that invalidates the results. "They're not looking at all breast cancers. They are selecting some here and some there," using only cancers that are diagnosed within the study period and also result in ...