This article was originally posted in the
The United States Preventive Services Task Force issued a last week for their breast-cancer screening guidelines. The proposal would involve changing the recommendation to start screening women at age 40 rather than age 50. On the surface, more cancer screening should be better. But counterintuitively, the issue is a bit more nuanced than that because more health care isn鈥檛 always better health care.
When it comes to breast-cancer screening, the problem isn鈥檛 a lack of data. There is a wealth of data to analyze and far more than we have for most other diseases. Nevertheless, some caveats are necessary. First, technology has evolved. The original mammogram studies were done with film radiography, whereas most labs now use digital mammography. We also have a growing use of MRI, although that is still largely reserved for high-risk younger women. Newer, more sensitive imaging modalities are inevitably going to change the estimate of how and when to screen for breast cancer.
As it stands, the collected data on mammography suggests it is very good at diagnosing breast cancer. What鈥檚 more, it does the very thing you want a cancer screening test to do: It reduces mortality. The totality of evidence suggests a breast-cancer mortality reduction of about when you synthesize all the available trial data. However, the benefit is not uniform across all age strata.
The shows a 32-per-cent reduction in women over 60, with a more modest 15-per-cent reduction in younger groups. There are multiple issues, but the main one is that mammograms do not perform as well in younger women because denser breast tissue makes them harder to interpret. MRIs are often used in higher risk younger women who needed screening for exactly this reason. And yet, the mortality benefit to screening mammography does extend to women in their 40s, even if it is less pronounced than in older age groups.
Why, then, the controversy? The to cancer screening are false positives and overdiagnosis. A false positive result means that a mammogram is read as abnormal when the patient doesn鈥檛 have cancer.
Tests are never perfect. You have to make a trade-off between sensitivity and specificity. As a rule, you want screening tests to be sensitive and not miss any cancer. But that means they suffer in terms of specificity and pick up clinically inconsequential abnormalities that are not malignant. Most people accept this tradeoff and are willing to endure subsequent testing after an abnormal mammogram (that ultimately comes back normal) if it means not missing a cancer diagnosis.
Overdiagnosis is a slightly different problem. Not all detected tumours will necessarily spread and cause clinical disease. Some could, in theory, be left alone, but there is no reliable way to be certain which new cancer diagnosis is dangerous and which is more benign. Given the uncertainty, when it comes to breast cancer, consensus is to treat all new diagnoses and accept that a certain amount of overdiagnosis happens.
There is no inherently right or wrong age to start screening for breast cancer. The higher your baseline risk, the sooner screening should occur. Debates for starting at 40 instead of 50, or for yearly versus biennial mammograms reflect debates about tradeoffs between benefits and drawbacks. It is a complex balance driven not just by science but also by health-care economics and societal values about a certain degree of false positives. Also, improvements in diagnosis and treatment keep changing the calculus. Balancing the harms and benefits of screening is complicated and nuanced and hard to encapsulate in a single paragraph. But barring a few nihilists who claim that mammograms don鈥檛 work, there is broad consensus that mammography has a real mortality benefit. We just need to fine tune how best to use it for maximum benefit.
