For this patient, and many other women, mammography screening has become a double-edged sword. Responding to strong recommendations from the medical community, some 80% of women over 40 with no known risk factors for breast cancer dutifully undergo the annual screening. The conventional wisdom among advocates of regular screening is that with mammography, cancers are detected earlier and can be treated before they have a chance to spread. Often, we are told that middle-aged women who have annual screenings reduce their chance of dying from breast cancer “by about a third.” Early diagnosis is also cited as contributing to the fact that the five-year survival rate for women diagnosed with breast cancer is close to 90%.
Behind the Numbers
But scratch the surface of those figures and the benefits are not so clear. Fiver year survival rates are high in the U.S.. because so many woman are screened, and their cancers are detected very early. But that doesn’t mean they live longer than woman in countries where mammograms are not as common. Early detection does not guarantee a cure. Imagine that I live in the U.S. and at age 40, a mammogram shows that I have breast cancer; my twin sister, who lives in France doesn’t go for mammograms, and doesn’t discover she has cancer until she is 43. We both die at age 46. I survived for six years following detection, so I add to this country’s “five-year-survival” rate. My sister lived for only three years after her cancer was diagnosed. But we died at the same age. My only “advantage” is that I knew I had cancer for six years, while she carried this burden for only three years.
As for the idea that a middle-aged woman’s chances of dying from breast cancer is reduced by a third, first, keep in mind that the chances that breast cancer will kill you are not that big to begin with. Let’s say that in your age group, 3 out of 1,000 women will die of breast cancer over a 10-year period if they don’t go for screening. If they do have mammograms, only 2 out of 1,000 would die. The risk is reduced by a third—from 3 to 2—but we’re talking about only 1 woman in 1,000 being saved after all of them undergo mammograms, annually, for ten years.
In addition, in 2001, researchers from the Nordic Cochrane Center in Sweden published a review of relevant studies on mammography and found that screening was likely to reduce the relative mortality risk of breast cancer by 15%, not the 30% that most groups quote. Their conclusion: “For every 2,000 women [age 50-69] invited for screening throughout 10 years, one will have her life prolonged. In addition, 10 healthy women who would not have been diagnosed if there had not been screening, will be diagnosed as breast cancer patients and will be treated unnecessarily. It is thus not clear whether screening does more good than harm.”
Research also has begun to calculate the psychological downside to mammography. Many of us harbor a deep fear of being struck with no warning, of finding a lump, of receiving a diagnosis of “cancer” that instantaneously moves us from what Susan Sontag called the “kingdom of the well” to become citizens of that other place, “the kingdom of the sick.” We have all known someone who has been diagnosed with breast cancer; many of us have friends or relatives who have died from the disease. We have been conditioned to see regular mammography screening as a defense against breast cancer, something we owe our families and loved ones as much as ourselves.
More than one in ten women experience psychological distress after mammography, primarily because of the high rate of false positives and the attendant biopsies that result. The problem is that radiologists often have a difficult time distinguishing between cancerous and benign lesions on the film. In fact, some 90% of what looks like cancer in routine mammograms turns out not to be. And after ten years of routine screening, a woman stands a 50% chance of receiving a false positive and a call-back for more testing. The benefit of yearly mammography screening for low-risk, younger women age 40-50, is at least as equivocal. A panel convened by the National Cancer Institute in 1997 found that in this age group, less than one life would be saved for every 1,000 women screened for an entire decade. One problem in this younger group is that they are more likely to be diagnosed with fast-growing, aggressive tumors that show up in between mammograms, eluding early diagnosis. Another issue is that younger women tend to have denser breast tissue, making mammography less accurate and false positives and negatives more likely.
An Epidemic of Diagnoses
The problems of over-diagnosis and over-treatment have increased as advances such as digital mammography, computer-aided diagnosis and MRI scanning make screening more sensitive to tiny changes in breast tissue. Besides the shadows and microcalcifications that turn out to be innocuous artifacts of the imaging process, there has been a more than eight-fold increase in the diagnosis of ductal carcinoma in situ (DCIS), small “precancerous” lesions that are formed in the ducts from the milk glands. Often a mammogram will uncover these DCIS.
“We are witnessing an epidemic of DCIS,” says Nortin M. Hadler, a physician and professor at the University of North Carolina at Chapel Hill, in his book, Worried Sick: A Prescription for Health in an Overtreated America. That’s not because women are getting more cancers, he says, but because “U.S. women are getting more breast biopsies thanks to mammography. DCIS is the incidental finding of this exercise.”
How often DCIS, left untreated, would lead to invasive breast cancer is not known. But what is known is that at least 30% of these tiny lesions would not have caused a problem, and in fact, some proportion would have gone away on their own. Hadler bemoans the fact that in most cases, a diagnosis of DCIS leads to a lumpectomy, adjuvant radiation treatment and chemotherapy. “Finding DCIS and resecting DCIS and irradiating DCIS are enormous industries in the United States . . . I suspect that little of importance, perhaps nothing, would be lost if all these women with tiny lesions, only detected on mammography, never knew they had DCIS.”
Ambiguity and Vested Interests
Women looking for a clear, ultimate truth about mammography will be disappointed. The predominant view comes from a medical establishment that includes the enormous industry built up around screening and treatment. Radiology, pathology and surgery departments are big revenue centers for hospitals; some doctors have financial interests in the mammography centers where they refer their patients.. The world market for breast cancer therapies was estimated to be $11.3 billion in 2007, growing at 22% each year. The aggressive treatment of early-stage lesions like DCIS is a new industry in itself—with women at all points in the age spectrum, including elderly women, increasingly being diagnosed with the condition.. Cancer screening is also strongly supported by physicians who worry about the legal and personal ramifications of “missing something” in a patient who goes on to be diagnosed with late-stage disease.
“There will always be an asymmetry among testing proponents and detractors,” says H. Gilbert Welch, a physician and professor at Dartmouth Medical School, in his book Should I Be Tested For Cancer? “Testing proponents have a very strong interest (often financial) in promoting tests—much more so than the researchers trying to critically evaluate those tests. Proponents have powerful anecdotes, about individuals whose lives may have been “saved” because their cancer was caught early (if perhaps unnecessarily).”
In the end, the benefits of mammography are not as clear cut as women have been led to believe. The aggressive outreach for yearly screening does not reflect a realistic projection of how useful this technology is to women, especially those at the younger and older ends of the recommended spectrum. There is certainly little mention of the high rate of false positives, or increasingly, the chance that the cancer that is found might not be harmful and might even disappear on its own.
Most women do not know about the increased burden of unnecessary radiation and surgery—including biopsies, lumpectomies and mastectomies that results from this country’s zeal for cancer screening. This information might not—and should not—change the fact that many women still will want to be screened at all costs. But in the interest of making an “informed choice” women should know the real facts—or at least to be told that much of what they know about mammography is hyperbole. (For more about “informed choice” see this HealthBeat post) They should not be made to feel guilty if, with no family history of breast cancer or other significant risks, they decide to choose yearly clinical breast exams with less frequent mammogram screening.
Individuals Need a Chance to Make an “Informed Choice”
For groups that question the effectiveness of mammography, the goal is not to set limits on those women over 40 who feel strongly about being screened yearly. The real goal is to provide comprehensive, realistic information about mammography to help women and their doctors engage in the process of informed decision-making. This is a key component of patient advocacy; making sure that the choice to be screened is right for the individual based on her age, risk factors, personal beliefs and ability to deal with the psychological stress of false diagnosis.
Two years ago, the American College of Physicians issued new breast cancer screening guidelines suggesting that a routine annual mammogram for low-risk women in their 40s may cause more harm than good, and that the decision to be screened should be individual, based on a doctor-patient discussion of all the risks and benefits. These recommendations were published in the April 3, 2007 issue of the Annals of Internal Medicine, and deviate from the blanket guidelines issued by the American Cancer Society and many other groups that continue to recommend that all normal-risk women 40 or older get annual or biennial mammograms as well as physical exams to screen for breast cancer.
When it comes to women over 69, the ACP and other groups again stress the importance of individual decision-making. By contrast, the American Cancer Society continues to recommend regular mammography screening for women up to age 85 who are in good health, even though the new cancers that do appear in this group are usually slow-growing and the women are much more likely to die of something else before the breast cancer catches up with them.
A better recommendation comes from the authors of a 1999 article in the Journal of the American Medical Association, “Continuing Screening Mammography in Women Aged 70 to 79 Years,”
“Elderly women who are bothered by medical tests, visits to physicians, and the discomfort of undergoing mammography, or who experience significant anxiety waiting for test results, and who are willing to accept a very small incremental risk of death from breast cancer, may rationally decline screening.”
Some breast cancer advocacy groups also are taking a more individualized view of mammography. The National Breast Cancer Coalition Fund (NBCCF), whose membership includes cancer support, information and service groups, as well as women’s health and provider organizations, takes this thoughtful position on a complicated issue:
:
“The scientific evidence from randomized trials on the impact of screening mammography in saving lives is conflicted, and the quality of the individual trials limited. The National Breast Cancer Coalition Fund (NBCCF) believes, on the basis of recently published reviews, that the benefits of screening mammography in reducing mortality are modest and there are harms associated with screening. No individual woman can be assured that screening mammography will be effective for her, and from a public health perspective, the harms and public health costs of screening mammography may outweigh the modest benefits of the intervention. Mammography does not prevent or cure breast cancer, and has many limitations. Therefore, a woman’s decision to undergo a screening mammogram must be made on an individual level, based on quality information about her specific risk factors, and her personal preferences.”
Where should we go from here? It’s time to stop reevaluating data from the same seven large-scale studies on mammography that have been fodder for numerous review articles on the subject. By kicking out the results of some studies considered skewed or of poor quality—as the Danish researchers did in the Cochrane Study—a strong argument can be made that the risks of yearly screening for women under 50 exceed the benefits.
By going back to these same old studies—as the American Cancer Society and many other groups do–the opposite point can be made, over and over again.
Rather than recycling the old conventional wisdom, perhaps we should focus on the shortcomings that we now recognize as inherent in mammography and let them spur a concerted effort to improve the way we diagnose and treat breast cancer. This would mean spending money on research efforts to help differentiate between early-stage tumors and precancerous lesions that are likely to grow and become invasive and those that are likely to never progress—or even disappear. In the meantime, researchers should be also be exploring strategies such as “watchful waiting”—continuing regular screening on DCIS, but perhaps delaying surgical biopsies and further interventions until there are signs that a lesion is actually progressing.
In the end, “Mammographyy remain a pretty crude tool,” acknowledges the surgeon/scientist/ who writes Respectful Insolence under the nom de blog,“Orac.” . “The reason it persists is because it is inexpensive, at least compared to newer modalities. But unfortunately . . .newer, more sensitive modalities like MRI suffer from the same problem in spades, the MRI, it is even less able to distinguish between tumors. That’s why I tend to believe that ever more sensitive detection modalities are not the answer. Rather, the development of better molecular diagnostic tests that more accurately distinguish between aggressive tumors and tumors that are unlikely ever to trouble the patient will be far more likely to improve the “signal-to-noise” ratio and decrease the unwanted phenomenon of overtreatment.
Orac is right. Mammography is not the optimal solution for lowering mortalities from breast cancer. The sooner we admit that and start looking for better options, the sooner we will see real progress in reducing the number of women who are hurt—either by the dreaded disease itself, or by the fears that leads to “iatrogenic suffering;” misery caused, inadvertently, by medical treatment.
This is really a well laid out website. I like how you have presented the information in full detail. Keep up the great work and please stop by my site sometime. The url is http://healthy-nutrition-facts.blogspot.com
Mammography has certainly been presented to the public as more useful than it really is.
Many European countries screen at a much higher rate than we do. Even in the US, rates vary geographically, with the test not utilized to the same degree in Florida, for example.
In the US, biopsy is pursued at the expense of observation and followup mammography. This is a manifestation of the same mentality that gets us CT scans in the ER rather than observation and clinical re-evaluation, gets us surgery for infantile pyloric stenosis, which in England, for the most part the little boy with this condition is allowed to follow its normal course and outgrow it without surgery…etc etc..
Another limitation of all cancer detection in young women is that below 35 or 40 years of age early detection does not confer a better outcome. These cancers are generally much more malignant.
The “lead time falicy” you allude to is well known to epidemiologists, and corrected for. The nordic countries are the world leaders in breast cancer detection. Mammography is not perfect.
No one can make you have a mammogram, if you don’t want to. If you do you need to find a center that stresses a low false positive rate, rather than a low cancer miss rate. In that setting, you will have a much lower call back and biopsy rate, but still harvest most of the cancer detection benefits at much lower (psychic as well as monetary ) cost.
There is no medical idea which is good as intended but somewhat toxic when carried to its logical extreme. Back surgery, coronary stents, end of life care come to mind in this context.
Great post.
Here is where The Standard of Care, conflicts with Peace of Mind.
The main problem is that although mammography suffers from all the defects that Naomi Freundliche describes, it is still the best tool we have right now. Based on past studies, we are pretty sure that if we don’t do mammography, many more women will present with lethal cancers detected only by physical exam. More research to find more definitive information would be useful, but is unlikely to be performed because of the ethics of the potential risks to women in the control group, plus the difficulty of getting enough women to be willing to be in the control group.
The great Achilles heel of mammography is the issue of false positives – abnormalities found on mammograms that turn out not to be cancers. The false positive rate is anywhere from 3 to 10 times as high as the true positive rate. I assume that it is these false positives that the article refers to when it says that for every woman who benefits ten will be diagnosed as breast cancer patients. They will not actually be diagnosed as cancer patients, but they will be forced to undergo uncomfortable and psychologically stressful biopsies to prove they don’t have cancer. Fortunately, thanks to techniques developed in Europe and popularized in the US by Colorado radiologist Steve Parker and others, most of those biopsies no longer involve surgery and anesthesia, but are done by minimally invasive techniques. That does not make the stress of the period of waiting for the diagnosis less, and may in fact lead radiologists to be more aggressive in calling for biopsies, since they are now less invasive. At today’s state of the art, there are many more women who have to suffer through the anxiety and discomfort of biopsy for benign lesions than for cancers.
As the article states, there is good evidence that as many as one third of cancers detected as ductal carcinoma in situ – cancers that have not invaded outside the milk ducts – will not lead to potentially killing cancers. Even a certain number of invasive cancers, especially small ones of certain types, may not end in killing the patient, especially if the patient is older and dies of something else first.
However, all of these problems put us in a situation of trying to guess who will fall among those who do not need more aggressive treatment and those who do. As a result, unless the patient wants to roll the dice and risk dying of a potentially treatable cancer, everyone with the diagnosis ends up being treated.
One other problem with mammography not discussed in the article is that in studies done at major mammography centers both in the US and abroad, as many as 65% of cancers are visible in retrospect on earlier mammograms. The gray zone between normal and abnormal is very large, even for the most skillful interpreters.
Unfortunately, it turns out that efforts to bridge that gray zone are less than totally
A great article on Mammography from The New Yorker is available online at this link:
http://www.newyorker.com/archive/2004/12/13/041213fa_fact
As someone who has interpreted mammography for over 20 years, I strongly agree that mammography is not a very good test. It has a high false positive rate and a significant false negative rate. In fact the two are linked, in order to lower the false negative rate, we have accepted a fairly high false positive rate.
Part of the reason that the false positive rate is so high is the degree of hysteria that surrounds breast cancer. Breast cancer is treated differently by the patients and by the government than other more serious (based on mortality) tumors. There are separate inspection and reporting requirements that do not apply to more serious tumors (lung, colon, ovarian).
And because expectations are so high for this relatively inaccurate test, the opportunities for lawsuits abound. When a women develops a breast cancer on mammography, it is often possible to go back to the prior mammogram and “see” (or at least claim to see) similar findings a year earlier. You can imagine how much the lawyers like this.
There have been a number of attempts to replace mammography over the years including; thermography, screening ultrasound and others. However, none of these alternative methods have been shown to be even as good as mammography.
Now along comes MRI, with which (thankfully) I am not involved. I suspect that by the time the “magnetic fields have quenched” – “the dust has settled”, we will find out that it is a useful adjunct to mammography, but is not a replacement due to cost and a high false positive rate.
Reading a mammogram is the most dangerous thing a radiologist can do, more dangerous than performing invasive procedures or trying to diagnose more lethal tumors. It is also not well paid – despite what has been stated. Fortunately, for the past several years I have not read any mammograms, a privilege that many of my colleagues are jealous of.
Reform of the whole mammography/breast cancer “industrial complex” is long overdue but unlikely to happen. Too many vested interests, too much hysteria, too many politicians, lawyers and women’s health advocates anxious to show how much they “care”.
welcome naomi. very nice piece.
a recurring tension here involves challenges to the conventional wisdom, which is central to science, but extremely difficult for public policy — and the public — to deal with.
you challenge it at your peril. if you start telling women under 50 that they were misinformed about mammograms, the danger is that your message will be construed as undermining conventional wisdom generally. maybe exercise doesn’t make much sense after all. maybe obesity isn’t an issue.
OH-OH! My post got cut off part way through (undoubtedly pilot error.)
Here is the rest:
Unfortunately, it turns out that efforts to bridge that gray zone are less than totally successful. Double and even triple reading (having each mammogram read by more than one radiologist) and computer based mammogram interpretation aids cause increases in false positives, often without much effect on diagnosis of cancers. MRI has a high sensitivity for cancers, but also introduces a whole new set of false positive results.
In the end, our best hope is discovery of chemotherapy treatments that can treat advanced breast cancers successfully, making early detection unnecessary. That day has not arrived yet.
Meanwhile, let me give some advice. First, make sure the center performing your mammogram and the radiologist reading it are experienced, certified, and perform a large number of mammograms. In many countries, radiologists need to interpret from 2000 to 4000 mammograms a year to qualify to read screening studies. In the US, they can read as few as 475 a year. Second, you may wish to seek out a center that provides prompt interpretation of mammograms and quick access to any necessary biopsies. There are a few centers providing interpretation before the patient leaves the department and same day service for biopsies, and many that will have results to your doctor in less than 24 hours and biopsies within less than a week. Third, in order to avoid the ordeal that the woman mentioned at the beginning of the article, if you change mammogram centers, try to either bring your old mammograms with you or arrange in advance to have them sent to the new center.
One other service I know of that might be worth looking for – I worked at a center where every woman who had an abnormal mammogram was assigned a nurse practitioner who oversaw the rest of her experience, discussing the results, accompanying her on visits to other specialists, and being available on a 24 hour basis to answer questions. The NP stayed with the patient until her problem was resolved, whether by a negative biopsy or by treatment and management of a breast cancer. We found that patients reported this to be very helpful in dealing with the stress of the situation.
Christopher,
Mammography rates are much higher in the U.S. than in the U.K., due to a national policy in the U.K. that recommends screening every three years only for women in the 50-64yr. age group. In 2004, only 2% of women in the U.K. who were under 50 years old received mammograms. And after the age of 65 years, the frequency of screening declines dramatically, whereas one-third of all mammograms were performed on American women in this age group. The rate of DCIS diagnosis is higher in the U.S. than in the U.K. as is the rate with which women are called back for biopsy. But the overall rates of cancer detection and mortality from breast cancer are similar.
Your point about women basing their decision about where to have a mammogram on “low false positive rates” is a good one, but unfortunately this is not a measure that readily accessible to most consumers.
Jim, Legacy Flyer, Pat;
Conventional wisdom is especially hard to challenge when it concerns such an emotionally-charged issue as breast cancer. Mammography has long been seen as a powerful weapon in the “war” on cancer. Evidence of its role in reducing mortality rates is modest, at best. We are loathe to give up this weapon–or at least acknowledge its weakness–because it forces us to admit how far we still have to go in conquering breast cancer and avoiding the collateral damage of overtreatment.
The role of DCIS in medical practice is similar to the role of the HMO in regulatory “practice”. Dispite hopes to the contrary, finding boatloads of DCIS has had no effect on survival, but we still look for it. Similarly, many regulatory schemes “should” have worked==but didn’t, and are still in place, adding to overhead without the promised savings.
The problem is that even bad ideas develop a constituency that makes modification difficult. Interestingly enough, medicine has a much better record of eliminating bad medical procedures than reformers have removing failed regulatory efforts.
Two additional comments:
1. The best criticism of mammography is that the cancers are actually found as a result of regular doctor visits and clinical exam, and that Mammography offers little ADDITIONAL benefit to regular primary care. (Maggie might like that one. For me, it is a strong one.)
2. My principle objection to endless patient call backs and stupid biopsies it that it causes women to lose faith in Mammography in general, and then they drop out of screening altogether. If you believe the data (and why would anyone who watches “The Wire” do that?) skipping mammography, on balance, would be a tragedy.
I have no problem, as I think I made clear, with acknowledging the weaknesses of mammography. It is, like Winston Churchill said about democracy, a horrible system, whose only strength is that it is better than any other system we have now. We need better ideas, but they are not here yet.
Looking for mammography centers with low false positive rates is great, as long as you know what you are doing when you look for them. In doing quality control for mammography programs for 20 years, I can tell you that there are two ways to attain lower false positive rates. One is to be really good, the other is to be really bad. People who have low false positive rates because they never call anything but very obvious cancers are not doing anyone any favors. I once worked with a radiologist who had a false positive rate of only 30%, which would have been great accept that he had never been able to find a cancer less than 3 cm. in size. What you really want to look at is an even more obscure statistic – the rate of stage 1 cancers in the radiologist’s biopsy total – the number of biopsies she needs to call for in order to find an invasive cancer that can be cured with a high rate of success.
Finally, the notion that physical exam has any ability to discover cancers better than mammography was put to bed 30 years ago, by the American Cancer Society’s Breast Cancer Detection Demonstration Project. It proved that only 60% of cancers detected by mammography were detected by physical exam, and that the majority of the cancers detected by physical exam were stage 2 and stage 3 (advanced) cancers with much higher death rates. Physical exam and breast self exam remain important parts of breast cancer detection, but need to be used with, not instead of, mammograms in women of appropriate age and risk patterns.
I’ll end with a quote from the excellent 2004 New Yorker article by Malcolm Gladwell, already cited by Christopher George, which acknowledges the state of affairs. After seven pages discussing everything negative about mammography that has already been said in this discussion, he says:
“The answer is that mammograms do not have to be infallible to save lives. A modest estimate of mammography’s benefit is that it reduces the risk of dying from breast cancer by about ten per cent—which works out, for the average woman in her fifties, to be about three extra days of life, or, to put it another way, a health benefit on a par with wearing a helmet on a ten-hour bicycle trip. That is not a trivial benefit. Multiplied across the millions of adult women in the United States, it amounts to thousands of lives saved every year, and, in combination with a medical regimen that includes radiation, surgery, and new and promising drugs, it has helped brighten the prognosis for women with breast cancer. Mammography isn’t as a good as we’d like it to be. But we are still better off than we would be without it.”
Dr. Pat,
I don’t really disagree with what you are saying, if I understand you correctly.
Just to clarify: A screening examination should minimize false positives. This is completely unrelated to poor reader performance. One is designed to set the test’s discriminate value to minimize unnecessary workup and anxiety. Sloppy work is associated with a poor reciever operator curve. A poor reader performs poorly, regardless of where you set the cutoff value.
One can provide mammographic screening with extremely low false positive rates, at the expense of a slightly higher false negative rate. This is how it is done in Northern Europe, with low call back rates.
It is essential to understand this somewhat subtle concept. Screening can be effective even if it finds some, but not necessarily all, occult cancers.
Just because you are being tortured with call back imaging and biopsies, you may still be getting poor interpretations. Sometimes the radiologist inflicts his indecision on the patient through never-ending additional imaging and inconclusive reports, with which I am familiar.
But I couldn’t resist a smile when you stated that anything in mammography has been “put to bed”. Open a journal, sometime. Dozens of studies have failed to find a benefit to mammography. Dozens more show benefit. This is the problem of evidence based. What evidence are we to believe?
Mammography, almost alone in medical practice, is the one thing in which experience does not confer an advantage in results. No study has ever convincingly shown that people who read a large volume do a better job than those who read less, as we expect a busy heart surgeon to perform better than an occasional heart surgeon. It speaks to a procedure which is just at the ragged edge of medical value. And it is the most effective cancer screening tool that we have!
I agree that attaining a low level of call backs, false positives, and additional exams should be a goal of any good mammography program. I was just pointing out that just asking about false positive rates was an unsatisfactory way of looking at mammogram excellence. Unfortunately, there are a surprising number of radiologists who attain low levels of false positives by ignoring anything but the most obvious tumors. I will restate my point, which is that the number you want to look at is the number of biopsies a radiologist requires to find a cureable cancer. This addresses both the issue of false positives, since it will increase the number of biopsies done, and of ability to find what we are looking for.
As to the contention that experience does not improve performance on mammography, that could not be further from the truth. Study after study in the US, Canada, the British Isles, Europe, and Australia have shown that the best mammographers are those who read large numbers of studies. This is quite different from just the number of years reading mammograms. The other important factor is ongoing careful evaluation of results — reviewing every biopsy mammogram and missed cancer after the fact in the light of pathologic results, and learning from your own performance. Like everything else in medicine, experience and careful review of the lessons of that experience are the best teachers.
That is why almost all foreign countries require much higher levels of experience to read screening studies.
The Scandanavians, as you suggest, lead the world in terms of the best results. They not only have lower false positive rates, but lower false negative rates. One important tool they use to attain this result is limit mammogram interpretation to a small number of people in central locations (the mammograms themselves are often taken by mobile units that travel from village to village, so it is the films, not the patients, that travel.) These designated mammographers read huge numbers of studies, often as many as 300 per day. They are also required to undergo extensive training under the supervision of established mammographers before being allowed to even read their first screening study on their own.
Unfortunately, the politics of mammography in the US makes this type of program impossible, since competition among independent hospitals, clinics, and radiology groups makes this unacceptable. US women are paying the price.
Pat S: Your post says, “I assume that it is these false positives that the article refers to when it says that for every woman who benefits ten will be diagnosed as breast cancer patients.”
However, it is not “false positives” that the Nordic Cochrane Center, which published this analysis, is referring to. They mean that a woman is ten times more likely to be treated for breast cancer – surgery for removal of all or part of a breast, radiation, chemo – even though she does not have cancer. Not just a biopsy (or some other procedure). Their exact words:
“If 2,000 women are screened regularly for ten years, one will benefit as she will avoid dying from breast cancer.
At the same time ten healthy women will be treated unnecessarily, having part or the whole of a breast removed and receiving radiotherapy and sometimes chemotherapy. A further 200 healthy women will have a false alarm.”
The debate over this issue is far more advanced in Europe. Why the powers that be in the US don’t think that we deserve this information so that we can make an informed choice is beyond me. It is terribly cruel. Perhaps the recent findings on the PSA test will help bring the overtreatment issue to the fore. The equivalent figure for men getting the PSA test is that they are 48 times more likely to be harmed by it than to benefit. The harms are unnecessary surgery, impotence, incontinence etc. Such unnecessary misery – it’s terribly sad.
I see…
People are not treating women for breast cancer who do not have breast cancer. They are treating women for breast cancers that may not be lethal. What the Nordic Cochran Center is referring to is women who have breast cancer based on pathology but would not die from breast cancer — women with either slow growing invasive cancers or ductal carcinoma in situ who would not develop lethal cancers during their normal lifespan.
The women are healthy in the sense that they don’t have a lethal disease, since some breast cancers are not lethal or will not be lethal during the lifetime of the woman.
This is pretty much the same issue as the prostate screening, except for the fact that prostate cancer is even more a disease of old age than breast cancer making the likelihood of death from prostate cancers even lower.
The point is well taken however. The problem, as has been discussed earlier in this thread, is making the distinction between potentially lethal cancers and non-lethal cancers. Although epidemiologically we (and the Cochran epidemiologists) know that many cases of DCIS and some cases of invasive cancers will not be lethal, we don’t have a good way of telling which ones are deadly and which are not as of now.
The question for women then is “is it worth reducing my risk of dying of breast cancer in my lifetime by 10% (from 9% to 8%) if it means a risk that I may be treated for a cancer that will not kill me, or should I ignore the whole issue and not get mammography, avoiding the risk of overtreatment but having a 1% greater risk of death from breast cancer?”
If the criticism of breast cancer screening becomes will enough understood, maybe the political climate surrounding breast cancer will change enough so that we can finally get a well designed large scale study using adaquate technical quality mammography and definitively answer the question.
Pat, I don’t want to argue, you seem like a really nice person, but I believe that a woman’s lifetime risk of dying of breast cancer is something like 1 in 28 or 3.6%. Also, the Nordic Cochrane analysis refers to healthy women who get a cancer diagnosis as having a “pseudo-disease” – yes it looks like cancer, but would never affect their health, and the only reason it was even discovered was because they were screened. It is, in essence, a “disease” created by screening.
The entire Nordic Cochrane plain-language breast cancer screening pamphlet is available at http://www.screening.dk – the harms of overtreatment are discussed far better than I could do it. Cheers.
This stuff is well known to us docs, but somehow doesn’t get through to our patients. Wonder why. My wife is a 44year-old family physician who absolutely refuses to get a mammogram. I support her decision fully. When I am 50, I will refuse a PSA (but will get a colonoscopy).
Mammography has been an enormous benefit – for Trial Lawyers. Plaintiff’s attorneys have reaped windfall profits from “Failure to Diagnose Breast Cancer Cases.” It’s gotten so bad that many Radiology centers in the NYC area are having difficulty finding physicians willing to read the studies.
Let’s say we decide that mammography is only marginally valuable. There is plenty of literature to support that position.
Now what?
Does that mean that a woman can’t get coverage for a mammogram from Medicare? From private insurance? Right now, if you don’t want to get a mammogram, you don’t have to. What we are really talking about is telling a woman or her doctor, one or both of whom wants her to get the exam, that she can’t get a mammogram.
Does that mean no PCP can be sued for not suggesting a mammogram when a woman develops breast cancer? Does it mean that the PCP’s quality indicator goes down if her patient gets a mammogram. Currently, it goes up, the more pap smears, mammograms, cholesterol screen the doctor does.
The mammography is an important question because there are a thousand similar questions in medicine where a marginal improvement or no improvement is purchased at the expense of a lot of money for a potentially unnecessary or even dangerous procedure. This relates to colonoscopy, back surgery, stents, physical therapy after a stroke, CT in the ER for headache and on and on.
Looking at the big picture this amounts to HMO redux. Instead of having the HMO steer doctors to cheap alternatives and sending the unspent money to shareholders, we are now proposing that overlords somewhere else (chosen for their flawless credentials) effectively outlaw (restrict, raise co-pays, or administratively impede with enless pre-certification ) expensive treatment and use the savings to expand coverage or reduce global warming or build another Federal office building in West Virginia named for Robert Bird. The government has one advantage that the HMOs lacked. HMOs were always afraid that the doctor wanting the test was correct, and that their refusal would have financial cocequences for the HMO; so, they never exactly refused. They just made it very very difficult. The government can simply refuse.
HMO round one was defeated because the public didn’t want to be told to ice their collective knees and see what happened. They wanted the MRI immediately and a consultation with an orthopedic surgeon who actually was an expert in knee injuries.
Let’s see what happens this time.
Ginny C. —
You are right — the number I cited was the risk of getting breast cancer. The death rate is about 3.5%. Screening, based on the most conservative estimates, would only save about 10% of those, or about one in every 300 women.
I have read the Nordic Cochrane study. It says exactly what we are both saying about non-lethal cancers. Strictly speaking, the women are not “healthy,” but rather have non-fatal illnesses that would not have been detected without screening.
The problem is that we have no good way of distinguishing in advance which women with cancer will live without treatment — who have pseudo-disease if you will –and which will die. That information is only available as population data, not individual information.
If we stopped screening, we would solve that problem, but at the expense of a few thousand exta women dying of breast cancer every year. I would not be opposed to that happening if that is what women want, since I believe it is certainly the right of people to choose or not choose what they want their health care to be like. As has been said by others, most doctors would love to be rid of screening mammography.
Finally, as the Cochrane group themselves say, all of the studies of breast cancer and screening — they are all from the 70’s and 80’s — are poor quality due to either poor statistical design, poor technical quality of mammograms, or both. The two most important studies — the Swedish Two County Study, which gives the strongest support for mammography, and the Canadian study, which raises the strongest questions about mammography — were particularly flawed, by poor statistical management in the Swedish study and by poor technical quality and by violation of the statistical design by some centers in the Canadian study.
The problem with meta-studies like the Cochrane group is the old problem of garbage in, garbage out. Piling 9 bad studies together does not make one big good study, but rather one big bad study.
In reality there is no good evidence for screening mammography or good evidence against screeing mammography. Obviously, in that situation the default position is not to screen and it is not rational to advocate screening. However the politics of breast cancer — a disease even more beset by politics than AIDS — are such that it is very hard to fight against. This political situation is worsened by the fact that even in the most conservative estimates several thousand women a year are saved by screening, albeit at considerable expense in both dollars and suffering by women who do not benefit but are in fact made worse.
The questions about mammography suggest that what really should be done is to do a good study, one featuring good technique, a large enough sample to get good statistics, and a long enough run time to deal with the problem of variable survivor time in breast cancer.
Until we do that kind of a study, we are all just chattering about our own ideas, not talking about science. Advocating either screening or not screening is not supported by any good evidence. That is the point of the Cochrane group report.
Very good posting and comments.
I don’t know where this came from (comment by Christopher George) and would be interested in a reference.
“The best criticism of mammography is that the cancers are actually found as a result of regular doctor visits and clinical exam, and that Mammography offers little ADDITIONAL benefit to regular primary care.”
In our community this is not the case.
In mammography’s defense with it there is no doubt that tumors are diagnosed earlier and that means a greater chance of removal before spread–cure. Look at Gina Kolata’s article in today’s NY Times for the outlook for patients with disease that has spread.
Mammography is not by any means perfect; it may be overused; there are too many false positives; and I share the concern about over treatment of DCIS; but it does offer the best chance of early diagnosis for people who actually have the problem.
http://www.medicynic.com
Very good comments indeed – the NYTimes article was about the lack of progress in bringing down cancer death rates -thus the title “Advances Elusive in Drive to Cure Cancer”.
The breast cancer victim featured in the article had an “insignificant” mammogram, and 6 months later had to get a walnut-sized tumor removed. Her cancer has returned despite aggressive post-surgical treatment.
She was shocked when she was diagnosed. I believe that the prevailing “wisdom” out there is that this will not happen to people who get mammograms.
This woman was not helped by mammography – her cancer was not found early. To my mind, this is no endorsement of screening mammography, which just does not do a good job in catching aggressive cancers. As a “weapon” in the war on breast cancer, it’s a dud. And if this was the only problem with mammography it would not be that terrible. Unfortunately, screening mammography is very good at catching insignificant changes, resulting in an unnecessary cancer diagnosis. Cycledoc, you are sincere, I’m sure, but the Times article is not a good endorsement of mammography. Nothing makes more intuitive sense than “catch it before it spreads”, but the article reminds us that it’s not that simple. The woman in the article was not helped or harmed, but plenty of women receive harm from mammography, and no benefit whatsoever.
Here is a very illustrative quote, the first sentence from the Nordic Cochrane’s breast cancer pamphlet that sums it up quite nicely:
“It may be reasonable to attend for breast cancer screening with mammography, but it may also be reasonable not to attend, as screening has both benefits and harms.”
Not quite the definitive guidance that most patients would like to hear.
The NY Times story illustrates something I said in an earlier post, that one of the other problems with mammography is that in good studies as many as 65% of breast tumors are visible in retrospect on earlier mammograms. Efforts to try to avoid the false positives and diagnosis of insignificant cancers are part of the reason, but the main reason is, as the article in the New Yorker says, the huge overlap between normal and abnormal.
In the long run, the best hope for cancers is to develop therapy that can cure advanced cancers — more Lance Armstrong style treatments like those for some testicular cancers and for choriocarcinoma in women that can actually cure cancers that have spread to the brain, lungs, liver and so on. Those advances could make screening and early detection unnecessary, since we could wait till tumors were obvious, then treat them.
Unfortunately, those cures are slow in coming.
Meanwhile, screening remains the best hope for detection before tumors spread and become untreatable. However, all screening tests are associated with missed cancers, false positives, and detection of cancers that probably would not kill the patients under ordinary circumstances. We have already discussed breast cancer and prostate cancer, but unfortunately efforts to detect lung cancer have worked out even worse. Pap smears for cervical cancer have lots of problems, and there is emerging evidence that colonoscopy is less than perfect as well, with reports of failures from the Society of Gastoenterology itself and in the NY Times in the last year.
The pattern for almost all screening tests is that some patients will benefit, some patients will fail to get diagnoses of cancers that are in fact present, some patients will suffer from overdiagnosis and undergo unnecessary surgery and treatments, and for some studies like colonoscopy some patients will suffer severe and sometimes fatal complications from the procedure itself.
We need to decide whether the benefits are worth the problems and the cost by doing good quality large controlled studies of the efficacy of the tests, particularly in reference to their ability of lack of ability to save lives. We need to make patients more aware of the potential problems, although that will probably lead to more rejection of screening procedures and consequently more deaths from potentially cureable tumors.
There is no easy road here, and anyone on either side of the issue who suggests there is is not telling the truth.
I find it disturbing that such high tech procedures and instruments that we use for making medical diagnoses can be so inaccurate and unreliable. On the other hand, the use of these tests, such as mammograms, have saved the lives of many women who have been diagnosed and treated within the proper time frames. Doctors and radiologists are not infalliable and when the results and images of man-made machines have to be relied on to make such critical diagnoses, it is expected that false positives and misinterpretations will occur. I understand that such a situation can be highly stressful and cause much worry for someone who falls victim of this but personally, I am just grateful for technology such as this that with the right diagnosis could potentially save my life one day. As with anything, the pros and cons must be weighed and in situations such as the diagnosis of breast cancer that can mean life or death, I would prefer to have the choice of mammography, although imperfect, for my screenings and preventative care.
Cycledoc, Pat S. and Christopher;
I agree that the studies used to determine the true benefits of mammography are less than stellar–some are just plain bad. This is clearly an example of the difficulty inherent in evidence-based medicine–which evidence to use?
Different groups–including the Cochrane folks–choose to kick out studies they believe are flawed and come to very different conclusions about the benefits of screening mammography.Perhaps another, better designed study does need to be undertaken to come to a less equivocal conclusion. Maybe those studies should focus on younger women or those over 69 where the benefits of screening are more marginal. But I believe that we should concentrate our limited resources on finding new insights that can allow technology to provide more meaningful information to women and their doctors. This includes learning how to differentiate between “pseudo-disease” and lesions that will progress.
Ginny C. and Christopher,
The approach to mammography in Europe is quite different than in the U.S. as you, Ginny, point out.In the U.K., for example, women age 50-70 are “invited” (meaning that’s what the National Health Service pays for)to have a mammogram every three years. I’m not advocating for that policy here and I wouldn’t want to think that Medicare or private insurers would stop paying for women to have mammograms if they want them.
But, I do feel that asking questions about the benefits of widespread, annual screening for women age 40-85 can only help the cause of patient advocacy. The ACS and other groups–including insurers and agencies that use regular mammography screening as a measure of quality care–are making a case for benefits that are not so clear. In the interest of informed decision-making, we need to keep asking questions, if only to drive improvements in breast cancer detection that really will save lives.
The Cochrane folks did not kick out the studies they considered poorly designed or conducted. In the preface to their discussion of the 9 studies they included in their study they point out that only 2 of them are technically good quality.
The problem is that you cannot talk about mammography by talking about good quality studies, since there are no real good quality studies out there.
That is probably the main reason that Cochrane says that they cannot draw a clear conclusion about mammography, stating that there are rational arguments for mammography and rational arguments against.
Good studies of women over 70 or under 50 would be a useful start, but would leave out most of the women discussed in the lead article. We need good studies, period. Good data is the key to intelligent health care policy.
Also, Britain is not a surrogate for Europe in mammography. For example, the Scandanavian countries, including Denmark, the home of the Cochrane group, offer screening on a yearly basis. The only major difference from the US is that all of the European countries have an upper age limit, ranging from 70 to 80, for screening mammography.
Britain and the US are outliers on either end of this issue.
Dear Naomi:
I missed all these comments at the other posting of this article. These are all excellent questions, probably ones I would not hear at a radiologist convention unfortunately. Here are some simple facts: The death risk without screening from breast cancer is only 1% over 15 years for a women age 53. The lifetime death risk is about 2.24% without screening. For a 40 year old woman, the survival percentage goes from 99.5% with no screening, to 99.6% with screening. A recent trial in England for women age 40 found a statistically non-significant risk reduction. Why hasn’t the ACS publicized this data to better inform women? Please see our two recent papers at open access biomedcentral.com: Medical Informatics and Decision Making. Great job.
I had my first and last mammogram last year.
I spent 48 hours worrying about my future – I was sure I had cancer.
I spent all day having repeat mammograms, ultrasounds, physical exams and biopsies. I was told the next day there was no cancer – it was a false positive.
I’ve since read up about mammograms and find this is not unusual – the test produces lots of false positives and unnecessary follow-up. The anxirty caused by a false positive is enormous…just awful.
There is also the worry of ductal carcinoma in situ which is found in 40% of older women – a cancer that rarely bothers you, it’s so slow growing, but once biopsied, it can become aggressive and once diagnosed, the breast usually comes off…and the new research that mammograms may cause breast cancer due to trauma to delicate breast tissue.
No thanks, I’ve now rejected all cancer screening – I’ll take my chances and deal with symptoms.
One final point – why is there such dishonesty surrounding cancer screening? Why did I have to research the topic? Why are the cancer screening brochures totally inadequate – only showing the upside of screening.
Why didn’t my Dr mention the risks of screening?
It seems we have to protect ourselves these days and can no longer rely on our doctors – they might be chasing a screening target and the financial incentive that will follow or following recommendations from their associations whether they agree with them or not…profits, votes…it seems our health comes last.
Deborah, I am also very concerned about the one-sided information provided by the promoters of cancer screening. I would like to formally request that a group of credible, eminent persons follow the lead of such a group in England and write a letter to the editor of a major newspaper (the NY Times would be good) and present some of the actual facts about screening. Please see this link to read a great example of such a letter to the London Times. We deserve a solid basis for informed consent in the US. Can someone PLEASE help us out? http://www.timesonline.co.uk/tol/comment/letters/article5761650.ece
Please consider helping to develope a mail from home breath screen for Breast and Lung cancer. More info @ dogs4cancer.com/newsletter
John,
Thanks so much for the references, those are papers I had not known about but add to the literature questioning the benefits of widespread mammography screening.
Ginny, I’ll send a response.
I believe US women are over-examined, over-treated and over-tested to their detriment.
I have issues with cervical screening as well.
It’s an unreliable test for an uncommon cancer which means lots of false positives.
In the States, they push annual screening which means 95% of women will have a colposcopy and biopsies…with only a very small number having any cancer. (RM DeMay, “Should we abandon pap smear testing” Jnl of Cl. Pathology 2000)
Dr Angela Raffles (UK cancer screening expert) says that 1000 women need to be screened regularly for 35 years to save ONE woman from cervical cancer!
Why aren’t women given this information?
Also, the practice of tying birth control with cancer screening still continues in the States – cancer screening has nothing to do with birth control – this is a tactic to increase coverage and ignores the need for informed consent.
Also, US women are pressured to have annual pelvic, rectal and breast exams from their teens.(and annual smears – and all this includes virgins!) The medical thinking in other countries is quite clear – these exams are unnecessary in asymptomatic women and can be harmful.
The exceptions are – women over 40 or 45 may choose to have a CBE every year or two and cervical screening is offered every 2, 3 or 5 years depending on the country. Virgins are not included in these programs.
I’m not surprised my US colleagues choose to get their medical care in Hong Kong and avoid all this unnecessary testing and exams.
Some do not have screening at all, others follow the Dutch or Finnish cervical screening program which produces fewer false positives (55% of women will have biopsies) – these programs do not screen women under 30 and only test every 5 years stopping at 55 or 60. Young women have a high risk of producing a false positive & testing too frequently increases the number of false positive with no real benefit.
Biopsies are not to be taken lightly- they’re unpleasant, painful and it can take weeks or months to heal…damage to the cervix can cause problems with fertility, during pregnancy and some women have psycho-sexual problems.
Mammograms – read anything by Prof Michael Baum – the UK breast cancer surgeon who helped set up the first breast screening centre. He now has serious concerns about mammograms (and cervical screening)
Women (and now men with prostate screening) must appreciate that doctors have a conflict of interest – you may not get the information you need from your Dr. It’s easy with the Internet to do your research.
You’ll be surprised when you get to the truth.
The Nordic Cochrane Institute have also been very critical of the breast screening brochure and have asked for it to be redrafted…the current brochure IMO overstates the benefits and doesn’t mention the many risk factors.
When you have the facts, the upside and the downside of screening,you’re better able to make an informed decision and it’s easier to make clear your wish not to be included in these programs or you have the test knowing the risks…either way, you’ve made an informed decision.
Malignant tumors are difficult to classify, and a single, universally accepted classification is not developed. However, the most general level, these tumors can be divided into actual cancers (carcinoma), sarcoma and lymphoproliferative disease. Carcinoma – a malignant tumor composed of cells of epithelial or endothelial origin (eg, breast cancer, skin or lungs), usually invading the surrounding tissue and metastasizing through the lymphatic vessels. Sarcoma composed of cells of mesodermal derivatives such as connective tissue, muscle, cartilage and bone, and metastasis mainly on blood vessels. By lymphoproliferative diseases include leukemia, lymphoma (both Hodgkin’s disease or lymphogranulomatosis and non-Hodgkin’s lymphoma type) and polycythemia vera. These tumors arise in lymphoid tissue of lymph nodes and spleen cells or are formed in the bone marrow.
Impressed with this info on mammogram ..need to be proactive. thanks for sharing this with me!!
I consider one practice in the US very unethical. I come from Chile, where the law is that every film (x-ray, MRI, etc) that is taken of the patient belongs to the PATIENT. After all, it’s his or her body. Thus, when one gets a mammogram, in a day or two, one picks up and signs for a copy of the films. This is crucial to one’s ability to get a second opinion.
In the US, even in the most reputable of teaching hospitals, I cannot get my films unless I pay extra. The doctor cannot show me ON THE FILMS what the problem is. I cannot go to someone else for a second opinion. It is a closed system and thus, very subject to the doctor’s or hospital’s self-interest. “Oh, here’s a bit of a shadow, let’s milk this patient for all she’s worth. Let’s get a second mammogram, an MRI, a biopsy, etc. $$$$$” If ever I get such a ‘diagnosis’ my first step will be to get an airplane ticket and go home! Even if my country doesn’t have the ‘latest’ treatment gadget.