Share this article on:

Mammography: Changing Understandings about Harms & Benefits

DiGiulio, Sarah

doi: 10.1097/01.COT.0000444893.06548.0c

SAN ANTONIO—Screening mammography should be considered a choice, rather than a public health imperative. That was the message from H. Gilbert Welch, MD, MPH, Professor of Medicine at the Dartmouth Institute for Health Policy and Clinical Research, speaking here at the first plenary lecture of the CTRC-AACR San Antonio Breast Cancer Symposium.

Figure. No caption a...

There are benefits to screening for breast cancer with mammography, but there are also human costs, he said—and until relatively recently there has been a tendency in the past to ignore those harms. Women should understand the benefits and the harms, and make the decision with their doctor whether or not they want to be screened.

Figure. No caption a...

Symposium Co-director Kent Osborne, MD, Director of the Dan L. Duncan Cancer Center and the Lester and Sue Smith Breast Center at Baylor College of Medicine agreed: Most of the American public do not think they have a choice whether or not to get screened because it has been ground into people's minds—both patients and doctors—that early diagnosis is important, he said in an interview after the talk. “It's not a black-and-white issue—you couldn't possibly argue that.”

Also at the meeting, Robert A. Smith, PhD, Senior Director of Cancer Screening at the American Cancer Society, presented an analysis of four major systematic reviews that estimated the benefits and harms of mammography. The results showed that the differences between the outcomes of each of the reviews—the major discrepancies that have been reported in evaluating the effectiveness of mammography—were more similar than previously thought when the data were reanalyzed using “common denominators.”

His takeaway: the balance of the benefits and harms of screening with mammography “substantially” favor regular screening.

Back to Top | Article Outline

Talking about the Harms

“There's a fair amount of uncertainty in breast cancer screening—and a lot of professional disagreement,” Welch said in an interview after the meeting. “But it's very important to talk about the harms for two reasons: one, that it's a choice and people should understand what they're getting into; and two, it's only when the harms—false-positive results and overdiagnosis—are acknowledged and we start talking about them that we can actually reduce them.”

One study, Welch noted, found that three years after the time a woman finds out she had a false-positive result (and does not have cancer), the anxiety she experiences is somewhere between the anxiety of a breast cancer patient three years after she was diagnosed via mammogram and a healthy patient who had not received a false-positive result three years after a mammogram (Ann Fam Med 2013;11:106–115). Those women also reported a negative impact on sexuality, behavior, and sleep; worries about breast cancer; and loss of inner calm.

“One of the things you need to worry about when you're dealing with a population-based strategy is how much ‘dis-ease’ you're putting into a population,” Welch said.

He compared overdiagnosis—when a mammogram finds a real cancer and that cancer is treated, but that cancer was never destined to cause the patient harm in her lifetime, neither symptoms nor death—to a black hole. Neither overdiagnosis nor black holes are directly observed, but their existence is inferred by what goes on around them.

Back to Top | Article Outline

Optimistic and Pessimistic Estimates

Although the data vary drastically on what the real extent of these harms actually are, women should know both the most optimistic and the most pessimistic estimates, Welch said. He presented the ranges of these benefits and harms—estimates of deaths avoided, false-positive results, and overdiagnoses—which are now available online ahead of print as a Special Communication in a JAMA Internal Medicine paper he coauthored with Honor J. Passow, PhD (doi:10.1001/jamainternmed.2013.13635).

For a group of 1,000 women over age 50 undergoing annual mammography for 10 years, somewhere between 0.3 and 3.2 women will avoid a breast cancer death (0.3 is a five percent reduction in mortality, and 3.2 is a 37 percent reduction in mortality, the highest ever observed). In that same group, 490 to 670 will have at least one false alarm, 70 to 100 of whom will undergo a biopsy, and three to 14 of whom will be overdiagnosed and treated.

In his SABCS talk, Welch also presented data from an earlier paper, the effects of three decades of screening mammography on stage-specific breast cancer incidence (NEJM 2012;367:1998–2005). After analyzing breast cancer incidence during three decades of mammography use, overall incidence should remain stable—screening should not cause more cancers, he said. “You expect over time to see more early-stage cancers, but you also expect less diagnosis of late-stage cancers.”

But the data showed a doubling in the amount of early-stage breast cancer in women in the U.S. over age 40 following the introduction of screening mammography (from 1975 to 2008), and little compensatory decrease in the number of women presenting with late-stage breast cancer. The imbalance suggests that screening mammography finds a lot of extra cancers that are not destined to appear, Welch said. “We estimate that over a million women in this period have been overdiagnosed.

“Our bottom line is these data raise valid concerns about the value of screening mammography. They clarify that the mortality benefit is likely smaller and the harm of overdiagnosis likely larger than has been previously recognized. Women should realize this work does not answer the question, should I be screened for breast cancer—but they can rest assured that the question has more than one right answer.”

Back to Top | Article Outline

Reexamining Data

Smith reached a different conclusion in his re-examination of the Nordic Cochrane review (2011), the U.K. Independent Breast Screening Review (2012), the U.S. Preventive Services Task Force review (2009), and the European Screening Network review (EUROSCREEN, 2012).

Those studies, he explained, varied by as much as 20-fold in the estimate of the absolute benefit of screening, because they differed in how the data were used to estimate the benefits and harms of screening, in spite of the fact that three of the reviews observed similar results from their analysis of randomized trials.

“Our question was: how is it that three systematic reviews—each working from the same relative risk—produce such different estimates of absolute risk?”

Two of the reviews measured the absolute benefit of mammography using the number of women needed to invite to screening for 10 years to prevent one breast cancer death: for the UPSTF review, 377 to 1,904 women (depending on the age and number of screening rounds in the different randomized trials); and for the Nordic Cochrane Review, 2,000 women.

Whereas the other two reviews measured the absolute benefit of mammography using the number of women needed to screen to prevent one breast cancer death: for the U.K. review, 180 women screened every three years over a 20-year period; and for the EUROSCREEN review, 111 screened every two years over a 20-year period.

Smith explained that in the studies that measure invitation to screen, approximately 30 to 40 percent of the women in those trials did not attend all screening rounds, so the effectiveness measures decrease based on non-attendance in the women who never actually had a mammogram, despite being invited to screen, and still ended up dying of breast cancer. “Measuring exposure to screening is preferable, since invitation to screen is not likely to help a woman unless she actually attends screening.”

The other variable that would could largely skew results was the duration of follow up in each of the studies, which ranged from 10 to 30 years in the four reviews. With longer follow-up data, the number of women needed to screen to prevent one breast cancer death steadily improves, he said—i.e., more deaths are prevented because the number of deaths among women not exposed to screening steadily increases over the follow-up period.

In explaining why longer follow-up is needed in these studies, Smith referenced the Swedish Two-County Trial, which compared the absolute benefit of screening in 133,065 women after 10 years of follow-up with 29 years of follow-up. The number of women needed to undergo screening for seven years to prevent one breast cancer death decreased from 922 at 10 years to 414 after 29 years of follow-up (Radiology 2011;260:658–653).

At least 20 years of follow-up is needed to observe at least most of the benefit of mammography, he said: “We have not actually observed even half of the deaths prevented in 10 years of follow-up.”

Back to Top | Article Outline

Adjusted Absolute Benefit

In their analysis, Smith and his colleagues adjusted the data in the Nordic Cochrane Review, the USPSTF Review, and the EUROSCREEN to follow the parameters used in the U.K. Review: including women ages 50 to 51, screened every three years, for 20 years, with 25 years of follow-up, removing the number of deaths observed from cancers diagnosed in the pre-screening period (and thus considering deaths only from the women who could have benefitted from screening), and considering the absolute benefit in terms of the number of women needed to screen (rather than the number needed to invite).

The absolute benefit of mammography screening for the four reviews ranged from 96 women needed to screen to prevent one breast cancer death to 257 after being adjusted—dropping from a 20-fold difference in the findings of the four reviews to a 2.5-fold difference.

“Our conclusion is that deriving an absolute benefit of mammography—which is controversial—from common data sources is contextual,” Smith said. “In short, the absolute benefit once you standardize [the findings]—to a common population, a common screening scenario, and a common duration of follow-up—these differences become not so significant or important at all.”

Back to Top | Article Outline

Still the Patient's Choice

Smith's results help explain why some of these major studies had such different results, Osborne noted. “The true benefit of screening will be underestimated if you're counting people who never got screened in the screened group. [Smith] has found a way to analyze the studies on the same playing field. And when you do it that way, they have more similar results. I think that was the benefit of his analysis—it makes intuitive sense.”

The fact that there is a benefit in most of the studies—albeit small—was a reason to screen, Osborne added. The harms need to be taken into account, but “balanced by the fact that it looks like there is a reduction in mortality—not huge—but there is a mortality reduction in getting a mammogram. You present those two things to the patient and let her have a choice. We need, to the very best of our ability, to educate patients that mammograms have their problems and they probably also help to some extent.”

Back to Top | Article Outline

25-Year Follow-Up of Canadian National Breast Screening Study

As this issue of OT was being finalized, the latest large-scale, long-term study was published—the 25-year follow-up of the Canadian National Breast Screening Study of 89,935 women (BMJ doi: 10.1136/bmj.g366)—concluding that in women age 40 to 59, screening with annual mammography did not reduce mortality from breast cancer compared with annual physical breast examination and standard care.

Back to Top | Article Outline

The Downside of the Debate

Carol Lee, MD, a diagnostic radiologist who specializes in breast imaging at Memorial Sloan Kettering Cancer Center, said in an interview for this article after the meeting that the downside of the debate about mammography is the confusion it causes patients.

Lee noted she did not attend SABCS this year but is familiar with both Welch's and Smith's presentations, as well as the standing debate. She is a Fellow of the Society of Breast Imaging and Chair of the Communications Committee of the Breast Imaging Commission of the American College of Radiology, and has served on the National Mammography Quality Assurance Advisory Committee.

Figure. No caption a...
Back to Top | Article Outline

Dr. Welch and Dr. Smith seem to agree and disagree—what would you say is the takeaway?

“Looking at these exact same data, different groups come to different conclusions. What is unfortunate is that it serves to confuse people. It serves to cast doubts on the utility of mammography when study after study has shown decreased mortality, decreased deaths from breast cancer among screened populations.”

Back to Top | Article Outline

Should there be more of an emphasis put on the patient having the choice whether or not to screen in the guidelines?

“I don't think you need to put that into a guideline—patients always have a choice. A guideline is just that—it's a guideline. Whether or not the patient chooses to adhere to the guidelines is completely up to him or her. We as a medical community need to come up with guidelines based on evidence—and the evidence is that screening with mammography decreases deaths from breast cancer. I don't like to give the message that screening mammography may not work, because that's definitely not true.”

Back to Top | Article Outline

Do you agree that for some patients the harms do outweigh the benefits?

“That's something that can be addressed—a discussion that needs to take place between a woman and her health care provider. There are some downsides. But, I think what gets lost in all of this discussion is the fact that screening with mammography decreases mortality from breast cancer.”

Back to Top | Article Outline

iPad Extra!

Listen on the iPad edition of this issue as Kent Osborne, MD, discusses in a podcast with OT Assistant Editor Sarah DiGiulio what makes the findings of mammography screening studies so controversial.

If you are not yet receiving our iPad issues, download the free Oncology Times app from the App Store today! Visit, search in the App Store, or follow the link on

Back to Top | Article Outline

iPad Extra!

Figure. No caption a...

Watch on the iPad edition of this issue as Robert A. Smith, PhD, explains in a video interview with OT reporter Dan Keller, why four major reviews of the benefits of mammography screening have reached widely varying conclusions.

If you are not yet receiving our iPad issues, download the free Oncology Times app from the App Store today! Visit, search in the App Store, or follow the link on

Figure. No caption a...
© 2014 by Lippincott Williams & Wilkins, Inc.
Home  Clinical Resource Center
Current Issue       Search OT
Archives Get OT Enews
Blogs Email us!