Posted by: notdeaddinosaur | November 29, 2012

Mammograms for the Masses

Surprise! It turns out that if every woman in the US got a mammogram every year, almost exactly the same number of them would still die of breast cancer. Why? Watch this video. Read this by TBTAM. Bottom line: mammograms don’t find aggressive cancers soon enough (by definition, you can’t), while too many of the cancers they do find were never going to be fatal anyway. This is what “over-diagnosis” means.

The same problem arises when discussing other kinds of cancer screenings. Prostate cancer, for example, also tends to age more slowly than the men it’s growing in, usually outliving them (ie, patients die of something else first). Thus, the official acknowledgment that the harms of screening outweigh the benefits has resulted in the changed official recommendation from “screen” to “don’t screen”. Then again, I did just pick up a case of completely asymptomatic metastatic prostate cancer from a screening PSA in a 50-something year old patient. Go figure.

Even the venerable Annual Physical isn’t worth it. No one can show any actual health benefit to any kind of annual medical review (in the absence of symptoms, of course). It has been shown that patients who have a regular relationship with a primary care physician tend to do better than folks who never see a doctor. But in the absence of acute illness, how does one go about establishing that relationship? Some kind of periodic evaluation is obviously in order, but what “period” should we recommend (given that “annual” seems to be too frequent)?

The United States Preventive Services Task Force is one body charged with generating these kinds of official recommendations. This doesn’t stop all kinds of other folks (medical specialty societies, disease oriented groups, and so on) from also promulgating their own guidelines. Is it any surprise when these conflict? (hint: no)

The thing about guidelines is that, again by definition, they represent population-based recommendations. They are derived from population-based data (ie, studies done on large numbers of people), which result in statements that are intended to apply to populations, ie, large numbers of people.

Here’s my problem: Although I may take care of “large numbers of people”, I do not take care of populations. I take care of patients one at a time. Turns out that applying those population-based guidelines to the individual patient sitting in front of me isn’t always straightforward.

What to do?

TBTAM does a wonderful job of summarizing the discussions she presumably has with her patients about these kinds of issues (risks of birth control pills, hormone replacement therapy, and yes, mammograms). Similar talking points apply when discussing prostate cancer screenings with male patients. Most of these kinds of discussions end up with some version of, “joint decision making with a fully informed patient.”

A lot of the time, this still doesn’t help.

Never fear; the Dinosaur is here. Here is my answer to the newest age-old question, “How often should I…?” (see the doctor/get a mammogram/pap/PSA/etc):

Once in a while.

Mammogram every five years? Probably good enough to catch most of the slow-growing, non-fatal (unless completely ignored) malignancies. Paps every seven years? Ditto. We already know colonoscopies more often than every ten years* are probably overkill.

Screening cardiovascular blood work (cholesterol)? Every five years. That’s from Medicare’s preventive screening guidelines (ie, what they’ll pay for.)

Physical exams? Beats me. Most “physicals” are more an exercise in paperwork than anything else. Every five to ten years? Once every couple of years? Like so much else in medicine, it depends; mainly on the patient.

Think about it: the people who get into trouble are the folks who NEVER do any of these things. Checking in with a doctor once in a while, whatever seems right to a given individual, is probably good enough. Certainly better that not going at all.

Bear in mind, of course, that ALL of the above applies to people who are well. No symptoms. No pains. No lumps, no bumps. No overpowering thirst, constant urination, and blurred vision (the most common things I hear from “asymptomatic” patients I diagnose with diabetes upon discovering their blood sugar is 300). Screening means looking for things you have no reason to expect to find; by definition.

Because if you have a mass in your breast, you definitely need a mammogram.

Otherwise, just be sure to come see me once in a while.


* Typo edited; sorry/thanks.



  1. Did you mean that colonoscopies less than 10 years apart are probably overkill for the average, asymptomatic patient?

  2. I wish I could convince my 80-year-old mother that she no longer needs mammograms or colonoscopies, but I’m doomed as long as her internist keeps telling her to get them.

  3. The difficulty is the definition of “once in a whel.”

  4. What a very sensible post! Too bad that most MD’s don’t think like you…

  5. Lucy –

    Thanks for the link love!

    Of course, the challenging part is determining what “once in awhile means’ – it will be different for every patient. For women at average risk for breast cancer, it is an entirely reasonable choice to do mammograms less often than annually. For those with risks factors, it will be obviously be more often than for those without risks

    We must also understand that many women are more than willing to accept harms (ie overdiagnosis and unnecessary biopsies) in return for the real 15% or so reduction in mortality that mammogram affords. I would not want to deny those women a more frequent mammogram if that is what they desire. Especially as we do not have anything else at this time that works better. But the choice must be an informed one, not one based on hype.

    We have developed a web-based decision aid around mammograms for women in their 40’s. I’ll let you know when it is up and open to the public. We have to pilot it first.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s


%d bloggers like this: