The XX Factor

The Risk-Benefit Calculation of Mammograms

Mammograms aren’t useless. Whether you should get one, though, depends on a variety of factors.

Photo by ROBYN BECK/AFP/GettyImages

Last week on Slate, physician Meri Kolbrener lamented that she didn’t have a good way to explain the latest evidence on mammography to her patients. When patients ask her if they should get a mammogram, Kolbrener’s answer is, “I don’t know.”

Kolbrener is correct that the science on mammography is difficult to frame as a yes/no decision, but there is a more helpful way to answer the question.

First, explain how our thinking on breast cancer has evolved. Mammography was a no-brainer when we thought that breast cancer was a relentlessly progressive disease, predestined to kill you unless you can find it and remove it in time. This model of cancer suggests that every breast cancer is curable, if only we can remove it before it metastasizes and thus becomes deadly.


But scientists now know that breast cancers do not behave in a uniform way. Some aren’t destined to metastasize. These cancers grow slowly and they will never harm you. Others do follow a relentless progression pattern—they start out relatively benign, but eventually become deadly. Still others are so invasive and fast-growing that mammograms and other screening tests simply can’t detect them until they’ve already spread.

Mammography is very good at finding benign cancers and all but incapable of finding the super-invasive cancers in time to make a difference. Mammography saves lives when it finds the middle group of cancers, the ones that eventually progress. But there’s no way to tell whether a breast cancer found on a mammogram is a benign one or a progressive one. The statistics show that if you have regular mammograms, you’re about ten times as likely to be diagnosed with (and treated for) a cancer that would have never harmed you as you are to have your life saved.


Many doctors overlook the overdiagnosis problem, because it’s invisible to them. Women who have an indolent cancer detected on a mammogram end up undergoing cancer treatments they never needed. They may lose a breast because of a cancer that would have never hurt them, but they’re not angry about this, they’re grateful. They believe that their lives were saved, even though the studies show that most of them were instead unnecessarily treated.


The decision about whether to get mammograms comes down to this: How much harm are you willing to risk in return for a very small chance that you might achieve a very large benefit—that benefit being your life, saved?

I asked H. Gilbert Welch, author of the recent New England Journal of Medicine study that Kolbrener cites in her piece, to help me dissect the data. Here’s how the numbers stack up:


If you’re 50 years old, and you have an annual mammogram for the next ten years:
—Your risk of dying from breast cancer is between 0.34 and 0.37 percent.
—Your risk of being treated for a cancer found by screening that won’t harm you is between 0.4 and 1 percent.
—Your risk of a false alarm from screening is between 25 and 48 percent.

If you skip mammograms between age 50 and 59:
—Your risk of dying from breast cancer is between 0.39 and 0.48 percent.
—Your risk of being treated for a cancer found by screening that won’t harm you is 0 percent.
—Your risk of a false alarm from screening is 0 percent.


So: Having an annual mammogram during your 50s reduces your risk of dying from breast cancer by less than two-tenths of a percent, increases your risk of being treated for a cancer that would not have harmed you by as much as one percent, and ups your risk of a false alarm by as much as 50 percent.

For women in their 40s, the benefits of mammography are even lower—that’s why the U.S. Preventative Services Task Force decided, for that age group, to leave the decision up to women and their doctors. Breast cancer risk increases with age, and the benefits of mammography are greatest for women in their 60s.

There are no right answers here. But women deserve a chance to make their decisions with the numbers in hand.