Mammography Screening Is a Failed Experiment

A federal task force recommends that women start getting mammograms at age 40 in spite of evidence that mammograms harm more women than they help. Image from Wikipedia.

May 12, 2023. After writing three columns in a row on cancer (here, here and here), I need a break. But I have to write about mammograms, which made headlines this week.

The U.S. Preventive Services Task Force, a panel of experts that issues influential health-care guidelines, is recommending that women get mammograms every two years starting at age 40. This draft guidance, which could be modified, represents a dramatic shift from the previous recommended age of 50. “The change is bound to be controversial,” STAT reports, quoting an expert who sees no “compelling reason” for the shift.

The task force justifies its decision by citing recent increases in breast cancer among women in their 40s and higher-than-average mortality rates among black women. This justification makes no sense, because mammograms do not help women live longer—according to the task force itself! In fact, there is evidence that mammography cuts short more lives than it saves.

Before I delve into that evidence, some background: When researchers evaluate a medical intervention for a disease, they traditionally look at whether the intervention reduces the risk of dying from that disease. This outcome is called disease-specific mortality. The problem is that attributing death to a specific cause can be subjective, and it might under-estimate harm resulting from the intervention.

Moreover, mammograms can lead to false positives and an even more insidious problem, overdiagnosis. A false positive can be overturned by a biopsy, but not overdiagnosis, which occurs when screening detects tumors and other anomalies that never would have compromised health if left alone. Overdiagnosis often leads to unnecessary treatment, including surgery, chemotherapy and radiation.

The Preventive Services Task Force asserts that the benefits of mammography, which ideally detects cancer early when it is more treatable, outweigh harms stemming from false positives and overdiagnosis. But the alleged benefits of screening appear only in studies that measure breast-cancer mortality, which does not capture harm stemming from overdiagnosis.

For example, radiation therapy can damage the heart, but if someone treated for breast cancer dies of heart failure, that death will probably not be attributed to the breast cancer. Same with suicide resulting from the physical, psychological and financial trauma of being a cancer patient. Omitting these deaths discounts the harm from overdiagnosis.

For these reasons, researchers increasingly favor “all-cause mortality,” death from any cause, as a measure of screening’s effectiveness. Death, period, is a hard endpoint, with no subjective wriggle room. Various studies have found that mammography does not extend life when all-cause mortality is measured. That is why some experts urge that mammography screening be discontinued.

I expected the Preventive Services Task Force to justify its new guidelines with new data showing that mammograms reduce all-cause mortality. But no, the task force bases its recommendations only on trials that measure breast-cancer mortality. The task force acknowledges: “None of the trials nor the combined meta-analysis demonstrated a difference in all-cause mortality with screening mammography.”

I’ve italicized this statement, because it contradicts a headline on the task force website: “Breast Cancer Screening Saves Lives.” That brings me back to the evidence I mentioned above, that screening cuts short more lives than it saves. Amanda Kowalski, an economist specializing in health care, presents this evidence in “Mammograms and Mortality: How Has the Evidence Evolved?”, published in the Journal of Economic Perspectives in 2021.

Kowalski focuses on the Canadian National Breast Screening Study. Begun in 1980, the study divided almost 90,000 women ages 40-59 into two groups: one was offered annual mammograms for 5 years; the control group received “usual care.” The subjects were tracked for 20 years, an unusually long time, and 85 percent of the women offered mammograms got them, an unusually high compliance rate.

Over two decades, women offered screening died at a significantly faster rate than women in the control group. Kowalski notes that screened women had an elevated risk of dying of lung and esophageal cancer; she cites evidence that radiation treatment for breast cancer increases patients’ odds of getting lethal lung and esophageal cancer.

Ideally, trials should show the life-saving benefits of screening growing over time. But over the course of the Canadian trial, mammography’s lethal consequences became more pronounced. “With each additional year that passes after enrollment,” Kowalski writes, “an additional 7 excess deaths per 100,000 women become apparent among intervention arm participants relative to control arm participants. To put this trend in perspective, annual road traffic deaths in the United States are 12.4 per 100,000.”

Kowalski’s analysis is cautious, couched with caveats. Recent advances in screening and treatment, she notes, might boost the benefits of screening. Here’s a caveat of my own: mammograms might benefit women with a higher-than-average risk of breast cancer, such as those whose relatives have succumbed to the disease. But Kowalski’s findings have a devastating implication: screening healthy, asymptomatic women ends up killing more women than it saves.

This conclusion corroborates one reached a decade ago by cancer surgeon Michael Baum. After helping establish mammography screening in the United Kingdom, Baum became convinced that screening does more harm than good. Writing in a British medical journal in 2013, he estimates that for every woman whose life is saved by mammography, as many as three women die. In a followup essay in 2015, he calls mammography screening a “failed experiment” that should be abandoned.

Writing about mammography in the past, I’ve cited an influential 2013 review of mammography carried out by Cochrane, a group of independent medical experts. The review concludes that “for every 2,000 women invited for screening throughout 10 years, one will avoid dying of breast cancer and 10 healthy women, who would not have been diagnosed if there had not been screening, will be treated unnecessarily.”

Female friends who get mammograms have told me that they accept these odds. That is, they are willing to run the risk of being overdiagnosed and treated unnecessarily if there is even a small chance that their lives will be saved. But the chance that they will be saved by a mammogram is outweighed by the chance that they will die prematurely because of it. Or so the analyses of Kowalski and Baum suggest.

Given all these problems, how can experts, including those on the Preventive Services Task Force, still insist that mammography benefits women? Greed cannot be discounted. Breast-cancer care is a vast, profitable enterprise fueled by women’s fear of breast cancer. Two out of three American women aged 40 and over have gotten a mammogram within the last two years, according to the CDC. Caring for the 3.8 million breast-cancer survivors in the U.S. is also lucrative.

Another factor, I suspect, is that advocates of mammography screening can’t face the fact that they are perpetuating a colossal failed experiment. They can’t accept that they are hurting those they purport to help, including women in their 40s and women of color. And so the experts keep insisting that mammography “saves lives,” women keep getting screened, the failed experiment continues.

The National Cancer Institute estimates that 297,790 women will be diagnosed with breast cancer in 2023, and 43,170 women will die of the disease. These alarming numbers motivate women to get mammograms. The Cancer Institute should estimate how many women will suffer unnecessarily and die prematurely this year because they got mammograms. And if the Preventive Services Task Force really cares about women, it should call for an end to mammography screening rather than an expansion of it.

Further Reading:

See my previous columns on cancer: The Cancer Industry: Hype Versus Reality, We’re Too Scared of Skin Cancer and Do Colonoscopies Really Saves Lives?

If you want a break from cancer, check out my other “Cross-Check columns, like Conservation of Ignorance: A New Law of Nature.