Yet another major study has challenged the efficacy of mammograms. The study examined the claim that mammography has helped reduce breast-cancer mortality in North America and Europe by 15 percent or more over the past two decades.
Carried out by Archie Bleyer of the Knight Cancer Institute in Oregon and two colleagues and published in the International Journal of Cancer, the study found little correlation between screening and the decline in breast-cancer deaths.
“In the United States,” the authors state, “the magnitude of the mortality decline is greater in the unscreened, younger women than in the screened population and regional variation in the rate of breast cancer mortality reduction is not correlated with screening penetrance, either as self-reported or by the magnitude of screening-induced increase in early-stage disease.”
An examination of data from other countries points to the same conclusion. “Countries without a nationally-established screening program at the time of onset of their breast cancer mortality reduction (Norway, Belgium, Canada, Republic of Ireland, Northern Ireland) had rate reductions (mean 2.25% per year) equivalent to or greater than countries with implemented national programs (United States, Sweden, Netherlands) (mean 1.79% per year). The country with the highest participation rate in screening mammography had the slowest reduction (Sweden).”
The authors attribute the decline in breast-cancer mortality to advances in treatment, with screening playing a “minor role.” But they cite other studies that raise doubts about even a “minor” benefit; when all causes of mortality are considered rather than just breast-cancer mortality, women who are screened do not live longer than unscreened women.
At the end of their paper, Bleyer and his colleagues imply that not all analyses of mammography are carried out in good faith. Quoting another health-care analyst, they note that some researchers “torture the data to make it confess to what one knows to be the real truth!”
This quote came to mind when I read “Early Detection of Breast Cancer Backed by U.K. Study” by Melinda Beck of The Wall Street Journal. Beck reports on a study, published in Lancet, on screened British women diagnosed with abnormal milk duct cells, also called ductal carcinoma in situ, or DCIS.
“Some experts argue that DCIS shouldn’t be labeled cancer at all, since it’s unclear whether it will spread outside the milk ducts,” Beck states. But the British study found that “in 90% of the local screening areas, for every three cases of DCIS detected and treated, one fewer case of invasive cancer was diagnosed in the following three years than would have been without early intervention.”
The study provided no evidence that early diagnosis of DCIS extended women’s lives, and it implied that some women diagnosed with DCIS might have been treated unnecessarily. Stephen Duffy, the study’s lead author, nonetheless tells Beck that the findings “suggest that a substantial proportion of DCIS will become invasive if untreated, and it is therefore worth detecting and treating early.”
That sounds like torturing the data, especially in light of Bleyer’s study; or this 2013 analysis by the Cochrane Group, which is renowned for its impartial assessments of health-care practices; or other studies on which I have reported (see Further Reading).
See also this recent article in the online journal FiveThirtyEight by science journalist Christie Aschwanden, who has reported brilliantly and courageously on the mammography debate. She writes:
“Underneath the debate about at which age and at what frequency we should urge women to get mammograms, another important question looms: Is it reasonable to recommend a test that will produce false positives for something like half of the people who take it? Is it OK to risk harming hundreds of women in hopes of helping a handful avert a breast cancer death? The [American Cancer Society and U.S. Preventive Services Task Force] have concluded yes, but that’s a value judgment, not a scientific one.”
Aschwanden has decided to “skip mammograms altogether.”