Demystifying Medicine One Month at a Time


Mammograms: Find Your Sanity

Fairly typical week in health news: Mammograms.

The big story is that the American Cancer Society issued an updated guideline recommending that women undergo mammography less frequently than before.

mammogramThis announcement was denounced on both “sides” of the perennial debate. Those in the “mammograms save lives” camp are outraged that a scientific society dedicated to cancer prevention and treatment would issue a proclamation that seems to run counter to the notion that “early detection saves lives.”

Those in the “putting scientific evidence in the forefront” camp are actually somewhat pleased that the ACS is finally “moving in the right direction,” but displeased that the society didn’t get all the way to the vicinity of, for example, the US Preventive Services Task Force, which has the most heavily-weighted (and least stringent) screening mammography recommendations: for women at average risk (i.e. those that don’t have a mother or sister with breast cancer), start breast cancer screening at age 50 and get a mammogram every 2 years until age 74.

The new ACS guideline: start screening at age 45 (well, 40 if you want to) and have mammograms annually until age 55, at which point you can go to every other year.

If a woman at average risk for breast cancer follows the USPSTF guideline to the letter (and is lucky enough to avoid a ‘call-back,’ i.e. further looks for a possible abnormality), she’d have 13 mammograms over 25 years. If she follows the new ACS guidelines to the letter, she’d have 20 mammograms, possibly more. Of course, every mammogram not only increases the cumulative total of lifetime radiation exposure, it increases the odds that an abnormality will be found and a call-back will be issued.

The best analysis regarding the new ACS recommendation (and actually, one of the best pieces about the whole breast cancer screening issue in general) is from FiveThirtyEight’s lead science writer, Christie Aschwanden, whose piece is titled, “Science Won’t Settle the Mammogram Debate.” Aschwanden correctly points out the ‘right thing’ depends on you, the patient, and your values. There is no right answer.

For some, not getting mammograms annually (or even at all) is the right choice. For the rest, following the ‘rules’ such as they are provides the best piece of mind.

And that’s OK.

Here’s the thing: because choosing to have mammograms or not is a personal decision, we should refrain from blaming people who choose one way or the other. People have their reasons. As with many social and medical issues, the personal has become very political, because people’s beliefs are strongly held. Ultimately, a lot of economics is impacted by the politics here. Pro-screening partisans are always uneasy when edicts cutting back on screening are issued, because the fear is that the health care ‘establishment’ (i.e. insurance companies) will stop covering the tests.

That’s simply not going to happen with mammography.

If we strip the emotion out of the issue and just try to stick to facts, what, at heart, is undergoing a mammogram like?

The video below comes from the the UK’s Cancer Institute. It’s just more than a minute, and is very matter-of-fact.



Dear Glass:

I loved your recent piece on unnecessary screening tests in the Atlantic.

“Letters! We’ve got letters!”

You explain things so simply and clearly. Kudos!

One question, though. I understand when you say in general that CT scans for lung cancer screening can be dangerous. Excess radiation exposure, incidentalomas, cost, etc. I get it. But for those occasions when someone gets a lung tumor discovered early, don’t you think the screening is actually worth it? I mean, how can finding cancer earlier NOT prolong life?


Paula from Paducah

Good question. Two related answers.

The first is the concept of lead-time bias. [Look at figure 2 in the article linked here.]

Simply defined, lead-time bias gives a the false impression of prolonged life, because the time of diagnosis occurs earlier. But in diseases for which the treatment we have doesn’t alter the course of the illness (like lung cancer), we do not actually prolong life. So the “early” diagnosis is in fact an artifact.

Reason 2 is that lung cancer discovered on screening CTs is usually too far advanced to make early treatment meaningful in terms of survival benefit. Put another way, even if you know earlier that you have the illness, it doesn’t make a difference (in aggregate) for people with the disease.

This stuff is hard because if you’re the one (or you have a family member or friend) with the disease, it seems logical that it HAS to make a difference to find the information sooner rather than later.

Sadly, at this point, it’s just not true.

© 2021 GlassHospital

Theme by Anders NorenUp ↑