The New Mammogram Guidelines – What You Need to Know

Unless you’ve been living on another planet, you know that in mid-November, the US Preventive Services Task Force released new recommendations on screening mammography, in which they recommended against routine mammogram screening in women under age 50, and recommended that mammograms now be every two years in women ages 50-74.

What you may not have heard is that the Task Force has acknowledged that the mammogram guidelines were poorly worded, and have revised their original statement to clarify their intentions, mostly by removing those two little words “recommends against”.

Here’s how the guidelines now read (changes in red)
  • The USPSTF recommends biennial screening mammography for women aged 50 to 74 years.
  • The decision to start regular, biennial screening mammography before the age of 50 years should be an individual one and take patient context into account, including the patient’s values regarding specific benefits and harms.
  • The USPSTF concludes that the current evidence is insufficient to assess the additional benefits and harms of screening mammography in women 75 years or older.
  • The USPSTF concludes that the current evidence is insufficient to assess the additional benefits and harms of either digital mammography or magnetic resonance imaging (MRI) instead of film mammography as screening modalities for breast cancer.

They’ve also included this statement right in with the guidelines –

“So, what does this mean if you are a woman in your 40s? You should talk to your doctor and make an informed decision about whether mammography is right for you based on your family history, general health, and personal values.”

What the USPSTF meant to say

What the Task Force is saying is simply this – On a population basis, the net gain from adding 10 years of mammography in all women is small in relation to the risks of over-diagnosis, over treatment, unnecessary biopsies and anxiety. But you, as a patient, in consult with your physician and assessing your own personal risks of breast cancer, may decide you want to get a mammogram anyway.

What they meant to do was to take mammography out of the realm of the knee-jerk, automatic and into the realm of informed decision making. They meant to inform women that mammography’s 15% or so reduction in mortality comes at a price – a price that is physical, emotional and financial, in the form of false positive results, unnecessary biopsies and the anxiety and dollar cost that accompanies them. They also meant to dispel popular overblown notions about what mammograms actually do by clarifying both their benefits and their risks, so that women are making the most informed decision they can about whether or not to have this potentially lifesaving test.

Unfortunately, they blew their 15 minutes. Which leaves it to the rest of us to clean up the mess. So, let’s see if I can add my two sense to the party.

What you need to know about the USPSTF

First off, let’s dispel the conspiracy theories. The US Preventive Services Task Force is an independent panel charged with making health care recommendations based on current scientific evidence. They do not make health care policy or decide insurance coverage.

The task force members should have anticipated that the timing of their recommendations coincident with health care reform would lead to misunderstanding about their role. Their cluelessness in this regard alone should be proof that they have no ties with the stakeholders in health care reform, who clearly would have managed the spin upfront.

Which is not to say that the task force’s recommendations won’t be used to guide policy decisions, which is why everyone is taking this all so seriously.

What you need to know about mammograms

The lay public has an almost magical thinking about what mammograms actually do. This is not surprising given the intensity with which we have been advising them to have mammograms over the years. So it is not unexpected that women have been taken aback by the hard reality about mammograms that they are now being asked to accept. That said, here’s what you need to know –

1. Mammograms don’t prevent cancer. They diagnose it. It’s a simple but important distinction that gets clouded by the magical thinking surrounding this screening test. The value of mammography lies in its potential to diagnose cancers at an earlier stage, allowing life-saving treatment to begin earlier.

2.Because they use radiation, mammograms can actually cause cancers. Though a single mammogram has a low risk in this regard, the radiation exposure from annual mammograms over many years adds up. The task force estimated that on a population basis, annual mammograms from age 40-50 would induce 8 breast cancers for every 100,000 women.

3. Mammograms are not a perfect test. In general, they miss about 10% of cancers, more if you have dense breasts, which are more common in women under age 50. In addition, mammograms have a high false positive rate, meaning that if you have an abnormal mammogram, the odds are high that your biopsy will be benign, and technically unnecessary.

The task force estimated that the cumulative risk for a false-positive mammogram with 10 years of annual screening was about 50%. The younger you are, the higher the chance your abnormal mammogram will be a false alarm. The higher your risk of breast cancer going into screening, the lower your risk of a false positive result.

4. Mammograms may be better at diagnosing slower-growing cancers than more aggressive tumors. Think about it. If a tumor is growing slowly, testing once a year will find it sooner rather than later. If it’s a fast growing, aggressive tumor that spreads out of the breast at a smaller size, a test that is done only once a year may not pick it up before it has spread beyond the breast. So we may be finding and over-treating tumors that may never cause much problem, while missing the bad players. (I myself have a harder time accepting this as an argument for cutting back on screening in women under age 50 than for women over age 70.)

In this regard, one of the most problematic diagnoses made by mammography is that of DCIS, or ductal carcinoma in situ, a non-invasive neoplastic growth that looks like breast cancer by has not invaded beyond the duct wall, and may never become invasive. Mammograms are really good at finding DCIS, since its hallmark is calcifications, which tend to show up pretty well even in dense breasts. So we end up treating and even performing a lot of mastectomies because of DCIS, without knowing if we are impacting mortality.

Finally, if mammograms were as good as everyone thinks they are, then we should expect over the years to find less and less advanced breast cancers, since we should be picking them up earlier and treating them. Unfortunately, this has not yet been proven.<

5. Mammograms are a better screening tool in older versus younger women. In women ages 40-49, 1900 mammograms must be performed to prevent a single death in this age group, compared with 1339 women age 50-60, and 377 women age 60-69. This is because breast cancer risk increases with age (meaning a positive result is more likely to be a true positive) and because older women have less dense breasts, so that there are less false negative mammograms.

Measuring mammogram success by years of life saved instead of mortality alone, mammograms starting at age 40 look better as a screening tool, but still perform better in women over age 50.

6. The benefit of annual vs. biennial mammograms is negligible. Meaning you can go every other year without sacrificing much in the way of benefit (about 1-2% absolute risk reduction benefit), and save additional radiation exposure.

7. Despite their imperfections, Mammograms save lives. To the tune of about a 15-20% reduction in women ages 40-49, the group most affected by the new recommendations. This is an important fact that, in my opinion, keeps getting lost in the discussion about the guidelines.

Which brings me to the elephant in the room.

The Elephant in the Room

Breast cancer causes about 4,500 deaths annually in women ages 40-49, and is one of the leading causes of death in women in this age group. (This data does not include cancer deaths occurring after age 49 in women diagnosed in these years.) In the 10 year interval between 40 and 49, then, about 45,000 lives are lost to breast cancer. That’s no small number, and it’s why breast cancer advocates are up in arms at the recommendations.

Which brings me to the real crux of the question – how many of these breast cancer deaths is mammogram preventing in women ages 40-49? Put another way, if you forgo mammograms in that age group, what are your odds of dying as a result of that choice?<

A age 40, what are your odds of dying in the next 10 years from breast cancer?]

This was not an easy number to find. SEER data on cancer mortality groups ages from 35-44, 45-55 and so on, so it’s taken me a long time to find the data. But I finally found it.

At age 40, your chance of developing breast cancer in the next 10 years is 1.44% or about 1 in 69. Your chance of dying from breast cancer in that interval is about 1 in 480. (This compares to a risk of about 1 in 280 for a woman at age 50, 1 in 146 for a woman at age 60, and 1 in 108 at age 70, and so on.) Here’s how that risk looks visually, in the thousand dot graph below, with the red dots representing breast cancer deaths among 1,000 women.

So if mammograms prevent 15% of breast cancer deaths, then if you are 40, and have mammograms for the next 10 years, your chance of dying from breast cancer is reduced from 1 in 480 t0 about 1 in 564.

USA today estimates that annual mammograms reduce the 10 year mortality risk for women ages 40-49 from 1 in 300 to 1 in 357, as compared to women age 50-59 whose risk is reduced from 1 in 112 to 1 in 144.

That’s not a big individual reduction as far as cancer screening goes, especially when one compares it to, say, colon cancer screening, which reduces deaths from colon cancer by as much as 60%.

Looking at the numbers from a population rather than individual standpoint, assuming a US population of about 21 million women age 40-49, routine mammograms in this age group prevents about 680 deaths per year. Is that really worth having 21 million women get an annual test that over 10 years will result in 50% of them having an unnecessary breast biopsy? It certainly does not stand up to the standards we’ve set for screening tests in the past.

But breast cancer advocates will argue that every one of those 680 lives represents someone’s friend, spouse, parent or relative. How can we say those lives aren’t worth saving? But with that kind of argument, we’d be mammogramming 20 year olds. If mammograms were free and perfect, that would be a good argument. But they are neither.

I think when a screening test has such a high potential for false positives and invasive biopsies over time, it makes sense to allow individuals to make their own decisions about that screening. I also believe that breast cancer, because it is a leading cause of death in women age 40-50, deserves to be addressed as a risk, even if it is to decide in an individual to forgo screening.

What if You are High Risk?

The data the task force used to make their recommendations encompassed all women having screening, including both low and high risk women. But what if you are at increased risk?

You can calculate your individual risk for breast cancer by using one of several risk assessment tools – the most commonly used one being the Gail Model. The Gail model can give you your individual risk of being diagnosed with breast cancer in the next 5 years. You can then us this number to discuss with your doctor whether or not you want to start mammograms before age 50. I don’t know that the model can be used to predict mortality reduction from mammography in high risk women, but would say that if your risk for breast cancer approaches that of a 50 year old woman, you should start routine screening mammograms.

An important high risk group not addressed by the guidelines are African American women, who in general are diagnosed at more advanced stages of breast cancer and have higher breast cancer mortality rates than Caucasian women. Given that much of the data being used to support the USPSTF guidelines come from Scandinavian countries, one must question their application to non-white populations, including Hispanic and Asian women. Fortunately, the Gail model does include ethnicity in its risk calculation.

Bottom Line

Mammograms in women under age 50 are less efficient than in women over age 50, and come at a higher cost in terms of over-diagnosis and potential over-treatment. The USPSTF made a decision that the cost differential was enough to recommend against knee-jerk, routine mammograms in all women under age 50, and instead recommend that women discuss the decision with their doctor before deciding to start screening.

The American Cancer Society, the American College of Obstetricians and Gynecologists and the American College of Radiology continue to recommend routine mammogram screening every 1-2 years starting at age 40.

What do I recommend?

I’ve addressed this issue before, and have not changed my practice, which at this point is to offer mammograms starting at age 40 in all my patients. However, I am now framing it as an option rather than an undebatable recommendation for my low risk patients, which means we spending more time discussing the issue before I place the order.

So far, when presented with the data, every one of my low-risk patients age 40-50 has decided to have their mammograms. However, more than a few made that decision only after confirming that their insurer would continue to pay for the test. I’ve queried a few as to how much they would pay to have that mammogram if their insurer declined to pay – about $200 seems to be the break point above which those few low risk patients would decline the test based on cost alone. Most women are either willing to pay or fight for payment whatever the cost. (This is by no means a scientific sample, but I think captures the gestalt in my practice, which happens to include a fair number of high risk women.) In the absence of any other screening, most women seemed willing to accept the high rate of biopsy in return for a mortality reduction, however small.

I also frequently order screening sonograms in high risk women with dense breasts, and MRI in women with a first degree relative with premenopausal breast cancer or other risk factors for whom this testing has been recommended.

I am comfortable spacing mammogram screening to every other year, especially since that’s about the frequency many of my patients end up getting them anyway. The task force recommendations have certainly made me more comfortable reassuring the patient who calls a few weeks before her annual mammo is due and can’t be fit into the radiologist schedule for several months.

I have to admit I have some concerns about my risks if and when a patient declines routine mammograms. Will I get sued if I don’t urge her to get a routine mammogram and she ends up with an advanced stage breast cancer at some point in the future? Should I have her sign something to protect myself? What is the minimum I need to document to cover my tail? I’m also wondering how long it will be before the first lawsuit against a doc who follows the taskforce guidelines is filed. Will they try to sue the taskforce members themselves? (I wouldn’t put it past some of the lawyers.)

A Call for a Decision Tool

The Australian Screening Mammogram Decision Trial has a wonderful web-based tool to assist women age 40 in making a decision about mammography. I’d like to see the USPSTF develop a similar tool for American women incorporating the latest data they used. It’s the least they can do to help American women and their physicians begin to incorporate their recommendations into practice.

Recommended reading
Photo credit Wikipedia
Note – I clarified morality statistics from a previous version of this post, and apologize if they appeared misleading. They were technically correct, but I think this is clearer. Looking at deaths in this age group overall, cancers as a group account for about 30% of deaths, and breast cancer a third of these, or 10% of deaths overall. Heart disease as a group accounts for about 20% of deaths, with heart attacks about 5% of deaths overall.

32 Responses to The New Mammogram Guidelines – What You Need to Know

  1. Really good summary. Thanks! I'd really like to see the USPTF's own systematic review numbers, however.

    The problem I see is that we also need to have the overall mortality reported. So what if we reduce breast cancer deaths if these women are dying of other types of diseases.

    In the one study that specifically looked at mammography for ages 40-49, there was no benefit with 50,000 patients (!!!). You'd think if there was a difference, they would have found it with that amount of power.

  2. This is a helpful explanation; I like the little dot chart, as always. Seeing the percentages visually helps me understand them better.

  3. Graham – Thanks for your comments. The numbers the task force used can be found at their website, which I link to at the end of the post.

    They used RR and have some nice graphs showing the benefit risk ratio of various approaches to screening, all of which support their recommendations.

    Overall, there are 46,000 deaths annually in women age 40-49, who comprise about 22 million women. Of these deaths, 10% are from breast cancer (cancers overall account for 30% of deaths).

    Here is the reference –, and circulatory diseases about 20%

  4. I Redid the mortality stats, and have made some changes in the post, sticking to numbers rather than ranking. Although breast cancer is the single leading cause of deaths in women ages 40-49, if you group heart attacks (5%) and heart disease (other causes). then heart disease as a group indeed surpasses breast cancer as a cause of death by a few percent. Cancer overall causes 30% of deaths in this group, and remains at the top of the list until age 70, and I've referenced it up there.

    I apologize if I mislead anyone. Not my intention, These data are a bit hard to come by in the groupings I was looking for.

  5. No I don't think you misled. I just think it's important that whenever we're talking about mortality we should also include overall mortality.

    Also, if my back of the envelope calculations are right, the NNT is about 500-600, which isn't out of the range of other primary prevention measures people try, but I think it's important to consider that to save 1 life, 500-600 women have to get a mammogram, and how many of those will be false positives? (Most of them.)

  6. You did a great job reviewing all the recent surrounding controversy and clarifying it in simple terms. thanks. hope you have a wonderful holiday season.

  7. Thanks for a great post! I love the dot graphs.

    What do you think about doing a sonogram for a low risk 40 to 50 year old with dense breasts and no other risk factors, and starting biennial mammograms at 50?

  8. Graham-you are very close. I think when I was looking at the rationale papers on the Task Force site, the number was close to 1000 for NNT.

    Thanks for the great post.

  9. "So far, when presented with the data, every one of my low-risk patients age 40-50 has decided to have their mammograms."

    Does that make you wonder whether you are really framing the argument in truly non-biased fashion. We all act as choice architects, framing "informed decisions" with our own biases. You might believe your patient population is biased toward desiring more intensive screening in the first place.

    However, given the data on how many patients need to be screened to detect one relevant cancer which will be positively impacted by earlier detection (a whole lot!), a distribution of varying preferences which should be found in the population should yield many more no's in your patients, assuming they have really been presented with an unbiased view of the trade offs.

  10. MC-
    The answers I am getting make me think I am presenting things objectively – answers like –

    "I hear what you are saying, but I don't mind the risk of false positives"
    "As long as it is getting paid for I'm going to get it."
    "That's a small risk, but I'd do anything to lower it even further"

    Remember theses are New Yorkers coming for care at an academic medical center, so their reaction is not unexpected. You'd probably get a whole 'nuther reaction at a midwifery practice in Maine. I'm just telling it like I see it.

    Thanks for your thoughtful response.

  11. MC-

    Then again, a decision tool as i said would be most welcome and would keep things assuredly more objective

  12. great post, thanks for the efforts in breaking this complicated issue down. Unfortunately there are still no formal legal protections to protect doctors against the consequences of "informed decisions" not to screen for breast or prostate cancer, and so when you and the patient "miss one" things may deteriorate.

  13. I know of no legal theory under which the authors of the USPSTF could be sued by a woman who develops BC after forgoing a mammogram. Certainly, the authors could not be sued for medical malpractice, as they would not be in a doctor-patient relationship with the patient. If you know otherwise, please discuss.

    Of course, it is always possible to be sued, even if the suit is frivolous. However, most lawyers are not in the business of filing frivolous suits. Such suits are expensive, time-consuming, and in the case of repeat offenders, can lead a court to impose sanctions. They can also hurt a lawyer's credibility and reputation.

  14. Ms Mann-

    That's reassuringly good news for the members of the USPSTF. Any legal theory protecting me? 🙂

    Thanks for reading.


  15. Peggy

    As you probably guessed, I'm a lawyer. However, I have never done any medical malpractice work (my area is securities law). If you had a discussion with the patient of the pros and cons of mammography and the patient declined, and that was documented in your notes, wouldn't that provide some protection?


  16. Thanks very much for this comprehensive overview. I printed it and took it with me to an appointment to discuss my plan of care. Your comment about the physical and emotional price of mammograms in the form of unnecessary biopsies and procedures struck a chord. This article helped open up a conversation with my doctor.

  17. As with other commentators, I really like the dot graphs and your high-risk discussion.

    But, and maybe it is just me, it seems that your dot graphs are missing something that would help women look at the cost/benefit with new eyes (though the data simply may not exist for what I want).

    In addition to dot graphs showing *all* women 40-49 with their risks of death from breast cancer (with and without mammography), how about the same dot graphs but showing just 40-49 women at low risk.

    After all, it seems the USPTF's new guidelines are really aimed at evaluating mammogram use for this population. Evaluating just how much, or how little, difference the extra screening makes in this population is really the crux of the issue.

  18. Jehrler-

    Thank you for your comment – Those dot graphs are meant to represent the kind of data that were used to make the decisions about screening a population. The data each woman needs to make her own decision would be based on the Gail model or other tool for assessing individual risk.

    I point you to the Australian decision making tool I mentioned in my post for more targeted dot graphs. Its the kind of tool I beleive we need to assist patients in making the decision about mammography.

    Thanks again for reading.

  19. One of the difficulties for those of us in our forties is that we do not actually know that we are high risk, and, therefore, may not be able to make informed decisions with regard to screening. For example, my premenopausal 41 year old sister, the youngest of four sisters, was diagnosed with breast cancer five years ago. (Unfortunately, she is now at Stage IV.) Five months after my sister's diagnosis, I underwent my annual mammogram (45 years old, second youngest of the sisters, and premenopausal) and was also diagnosed with breast cancer. As far as either of us knew in the early part of that year, we were "low risk." And suddenly, we both had breast cancer and our two older sisters, one 50 and one 48, became "high risk." This example simply presents the potential difficulty of the typical risk assessment for screening purposes, and its usefulness for women in their forties who have not yet discovered what their risk status actually is.
    This article is very interesting and helpful, and I agree that discussions between the patient and doctor are critical. Following my own diagnosis, my surgeon graciously spent hours discussing Tamoxifen with me, and encouraged me to consider declining it following lumpectomy and radiation. Although I got the sense the oncologists did not necessarily agree with the surgeon, I took the surgeon's advice and declined the Tamoxifen (after a belated, very short trial of it prompted by my sister's first recurrence). I have now remained cancer-free for five years, although my sister's cancer has continued to progress to her liver and bones.
    Of course, for many of us, it is very hard to remain scientifically rational with regard to breast cancer screening because so many of us know young women with young children who are ravaged by breast cancer. The statistics just do not get us past those faces.

  20. In Australia mammograms are offered 2 yearly from 50. Women in their 40's could previously request a mammogram, but that is being discouraged now.
    I have chosen not to have mammograms after a year of research and soul-searching.
    I found one article very helpful and would recommend it to any woman thinking of having mammograms. The Nordic Cochrane Institute was so concerned that women were having mammograms without being advised of the risks and limited benefits, that they produced, "The risks and benefits of mammograms". It's worth a read.
    I think a decision to screen, or not, or to screen according to your own schedule is influenced by lots of things…your perception of risk, your personal risk assessment, family history, whether you believe in screening…
    No decision is ever wrong, it's a matter of reaching a decision that sits best with you.
    Good luck everyone!
    ( "Risks and Benefits of mammograms". (Screening Wars – Prof Michael Baum)

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.