Drop in breast ca mortality linked to mammography: study criticized as 'seriously flawed'. (40-Year Retrospective Study).

A new Swedish study linking routine screening mammography with a 44% decrease in breast cancer mortality hasn't convinced some experts, who claim it suffers from the same flaws that plagued a 2002 study by the same authors.

The new investigation, a 40-year retrospective study by Dr. Stephen Duffy and his colleagues, compared breast cancer deaths in two Swedish counties over two periods: the 20 years before routine screening mammography was offered (1958-1977) and the 20 years after it was offered (1978-1997). The study clearly indicates a 44% decrease in breast cancer mortality in the screening epoch among women aged 40-69, reported Dr. Duffy of Cancer Research United Kingdom, London (Lancet 361[9367]:1405-10, 2003).

Dr. Duffy published a similar study in 2002, which examined breast cancer mortality in seven Swedish counties before and after screening mammography was offered to residents. It found a 45% decrease in breast cancer mortality among those who took part in screening (Cancer 95[3]:451-57, 2002). Some have criticized those findings, however, saying that selection bias invalidated the results and that the mortality decrease could have been attributed to better breast cancer therapy in the study's later years.

Donald Berry, Ph.D., a member of the Physician Data Query Screening and Prevention Editorial Board, an independent panel that advises the National Cancer Institute, said the observed decrease in mortality may be correct, but there's not enough evidence to link it directly to screening mammography.

"Even though I believe their overall conclusion that screening has a benefit, I think that the study is seriously flawed," Dr. Berry said in an interview. "The principal flaw is that they ignore therapeutic improvements, which occurred at or near the times screening was being introduced. Moreover, the effectiveness of both chemotherapy and hormonal therapy has improved substantially over time as new agents have replaced older agents and we have learned better how to use the available agents."

Such therapeutic improvements could well account for the reductions in breast cancer mortality, he said.

But Robert Smith, Ph.D., director of cancer screening at the American Cancer Society; Atlanta, and coauthor of the new study, said it addresses this question by comparing the breast cancer mortality of women aged 20-69 before screening was offered with mortality rates in the post-screening period in the following groups: women aged 40-69 who accepted screening, women aged 40-69 who were offered screening and refused, and women aged 20-39, who were never offered screening.