My Take on the USPSTF Recommendations for Screening Mammography

For the couple of weeks, the United States Preventative Services Task Force new recommendations regarding screening mammography have sown confusion in the minds of many. I must say that, after reading the abstract (summary) of the article, I, too, am confused. Sometime in the next week or two, I will go to the medical library at our hospital and read the entire article. Until that time, I do have some preliminary concerns and opinions.

Those who know me or read my writing realize that I am very cynical of anything that comes out of Washington. In this case, I guess I am less cynical than most who think that this is a precursor to governmental rationing of health care under a government run health system. Don’t get me wrong, I think this will be used to ration mammograms, but I don’t think that was the purpose of this report.

My Facebook friends also know that I do a lot of mammography and, of every modality I do, it is the one I dislike the most. That is true for most radiologists. In fact, when a radiologist claims to like mammography, we generally make sure they never have access to sharp instruments in the workplace. In fact, mammography is drudgery. Reading mammograms is a 9 hr. eye test. It requires an incredible amount of concentration and the implications of a missed significant finding are great. Mammography has the poorest reimbursement rate of any studies we do, but is the number one cause of malpractice suits in radiology. My group barely breaks even on mammography and most groups actually lose money. Given these issues, I should be one of the happiest people in the world to have to do less mammography.

Breast cancer, though, can be a devastating disease. It is the leading cause of cancer deaths in women. Approximately 1 out of 10 women will develop breast cancer during the course of their lives. Even if not fatal, treatment can be disfiguring and disruptive. The earlier we can catch breast cancer, the less invasive treatment may be. For these reasons, I have great concerns about these new recommendations that I will attempt to address.

Purpose of Screening Studies

First, we need to understand the purpose of a screening study. A screening study is a study designed to detect pathology in an at-risk population in which there is no cause to suspect that the pathology exists, except for their membership in that population. In other words, screening studies are conducted on patients without symptoms. Therefore, if a woman or her doctor feels a lump in her breast, a mammogram to evaluate that lump is not a screening mammogram, it is a diagnostic mammogram and is not covered by this report.

By their very nature, screening studies are inefficient. They are designed for the highest degree of sensitivity, thereby necessarily sacrificing specificity. Allow me to illustrate the reason for this with the following, simplified, example drawn from mammography. One of the signs of cancer that I use when I look at a mammogram is calcifications. Now there are a lot of different types of calcifications in the breasts from various causes, most of which are not cancer. There are some calcifications that I see that I can almost 100% guarantee will be malignant, but there are also calcifications that I can 100% guarantee will be benign. If these were the only calcifications we saw in the breast, we would be golden, but unfortunately, there are a great number of calcifications between these two extremes of the spectrum, ranging from almost certainly cancerous to almost certainly benign and everything in between. If I were to refer only those calcifications that I knew 100% to be cancer for biopsy, I would have 100% specificity (every biopsy is positive for cancer), but very poor sensitivity (I would miss a lot of cancers among those equivocal calcifications). If I referred every calcification for biopsy, I would have 100% sensitivity (I would detect every cancer), but I would have poor specificity (we would have a lot of benign biopsies.) Because we consider having cancer worse than undergoing a biopsy for a benign condition, we intentionally try to sample all calcifications that have a chance of being malignant, only omitting those calcifications that we are certain are benign.

Every department that does mammography is required by law to conduct quality assurance of their screening examinations. In our practice, once or twice each week, a group of radiologists who read mammography conduct an audit of every biopsy that is performed at our institution and even those performed at another institution, if we read the initial mammogram. During this audit, each mammogram is reviewed and compared to the pathology report. If we recommended a biopsy that turned out to be malignant, we also look at previous mammograms to determine whether the abnormality was present on the previous studies. If we have ancillary studies (e.g. - breast ultrasound or MRI), those are also reviewed. Not only are these audits good on-going training for radiologists, but important data is collected, namely, statistics are generated on everyone who reads mammograms and are compared with national published averages, as well as compared to other radiologists in our practice. This allows us to see how we are doing as a group and individually and allows us to identify any outliers. Of the statistics that are generated, the most important is our false negative rate (i.e. - we read a mammogram as negative, but sometime within a year of that reading, a biopsy proven cancer was detected.) Another, almost as important, statistic we collect is our false positive rate (i.e.- we recommend a biopsy, but the biopsy reveals no cancer.) The national average for false positives, and for our group, is 80%. In other words, of every 10 biopsies we recommend, we expect 2 cancers. If we get more false positives, we are recommending too many biopsies and if we get less false positives, we worry that we are not biopsying enough and that we are missing cancers we should be detecting.

Why Don’t We Screen Everyone?

Now, the American Cancer Society recommends that women start receiving screening mammograms at age 40, yet we know of women in their 30’s and even in their 20’s who get breast carcinoma. So why don’t we conduct screening on women in these age groups? In fact, men can also get breast cancer, so why don’t we routinely screen men? If we truly wanted to catch all the breast cancers that are “catchable”, we would start screening both males and females at puberty and not stop until death. The reason we don’t is two-fold: 1) Mammograms have consequences, and 2) Money.

First, let’s look at the consequences of screening mammography. The consequence of false positives is the major reason given by the task force to stop doing mammography in women in their 40’s. Women who have a false positive report receive unnecessary procedures on their breasts, up to and including potentially disfiguring surgery. Biopsies are uncomfortable and the stress of a woman being told she needs a biopsy is not inconsiderable. There is always the possibility of complications, including infection or even anesthesia related death. Now, admittedly, most of these complications (excluding stress – which is near universal) are very rare, but they are possible.

There are also consequences to even true positives. There is the possibility that we may remove a very slow growing cancer in a woman that would never have become problematic. This is especially a problem with prostate screening in men. We know that many men develop prostate cancer that remains confined to the prostate and would remain subclinical all their lives, until they died of some unrelated illness. We don’t know how many breast carcinomas may remain subclinical, but it may be a not insignificant sum, especially among the cancer that we are most concerned about removing – ductal carcinoma in situ. In this case, a woman might undergo mastectomy, chemotherapy and/or radiation therapy for a cancer that would never have been a problem to her.

There are even consequences to normal mammograms. First, mammograms are uncomfortable for most women and can be very painful for some. Second, mammograms involve radiation. The fact of the matter is that we don’t know how much radiation is required to cause breast cancer, but we are almost certain that increasing radiation to the breasts increases the rate of breast cancer. This is especially true in those breasts that have not finished developing, especially in those women who have never been pregnant, and even among those women who are still subject to the hormonal influences of menstruation. The breasts of women who are post-menopausal are probably less susceptible to radiation (though hormonal replacement therapy may negate some of that protective influence.)

However crass it may sound, the second reason we don’t perform screening mammograms on everyone is that screening is very cost-inefficient. This is the reason, incidentally, that the idea that preventative medicine will save money in the long run is specious. In fact, for every breast cancer that is discovered by screening mammography in the United States, we, as a society, spend over $17,000. Now, just because we detect the cancer doesn’t mean the cancer can be cured. We know that screening mammography has reduced mortality (death rate) among women between 40-50 by about 4%. Thus, for every Quality Adjusted Life Year (QALY) saved, we, as a society, spend approximately $26,500 - $85,500, depending on the study.

Now, if your life or your loved-one’s life is the one saved, this is a small price to pay, but in these times of limited health care dollars, we need to consider cost to society of screening examinations. How much money is too much to save one life? One of the reasons that insurance costs have increased is that most states now require insurance to cover screening studies such as mammograms, PSA screens and colonoscopies.

Several years ago, I went to a major breast care conference during which a pioneer in the field of mammography was honored. This man was a giant in the field, contributing to many of the practices we use today and one of the researchers that proved that mammography saves lives (it was not always assumed so.) This was an annual award and the recipient of the reward traditionally would give a keynote address at the conclusion of the conference. In his keynote address, this mammographer stunned everyone in the audience when he stated that he concluded that it was not moral to continue to provide screening mammograms. His reasoning was that there are only a limited amount of health care dollars (he was a big proponent of a single-payer health care system) and that screening mammography was so expensive per life saved, that we would get more value for our buck if we would put that money into inner-city health clinics, vaccination programs, etc.

The fact is, younger patients have less breast cancer, so we get diminishing returns when we screen younger patients until, eventually, it becomes cost-prohibitive to do so. The problem is deciding where to draw that cost-effectiveness line.

Problems with the Task Force Recommendations

The Panel

One major problem I have with the task force recommendations is that the members of the task force did not include anyone that actually diagnoses or treats breast cancer. There was no radiologist, surgeon, or oncologist on the panel. In fact, I’m not even sure that there was anyone on the panel who still sees patients, at all. Most of the panel are epidemiologists, public health physicians and nurses, and mathematicians and economists. These people, therefore, are number crunchers and do not deal with the various concerns of patients dealing with breast cancer. This manifested itself most notably in the inflated value the panel assigned to complications of false positive studies. While there is no doubt that there is stress and discomfort involved with false positive results, these pale in comparison with the stress and discomfort, not to mention potential death that may occur in women who develop breast cancer that was potentially treatable.

This also caused the panel to lose sight of the real purpose of screening mammography, which is to detect those cancers that are potentially treatable and will extend the patient’s life, in favor of maximizing cost-effectiveness of the screening mammograms.

The References

Like I stated earlier, I have not read the complete paper, so I don’t know if there are reference studies that were used that were not cited in the abstract, though I don’t understand why there would be. When I reviewed the abstract, however, I was struck by the fact that most of the references cited were other organizations statistics concerning screening mammograms and cost-analysis studies. I didn’t recognize hardly any of the titles of the articles. What I did recognize, however, were notable absences of multiple landmark studies, including those that laid the groundwork for performing screening mammograms in women under 50 years of age.

Unusual Statements by the Panel

I was also struck by several statements by members of the panel that demonstrated their total lack of understanding of mammography. The task force abstract states, “Digital mammography detects some cases of cancer not identified by film mammography; film mammography detects some cases of cancer not identified by digital mammography.” This is false. While digital mammography can identify some cases of cancer that film mammography cannot, there is no evidence that film mammography may detect cancers not identified by digital mammography.

On a news show, one of the members of the panel stated that screening ultrasound is just as good as screening mammography and doesn’t have the risk of radiation. While he was correct about not using radiation, the statement that ultrasound is just as good as screening mammography is not only laughable, but it is a reckless statement for an “expert” to make in front of the public. In actuality, ultrasound is a complimentary study to mammography and is not adequate, by itself, to screen for breast cancer.

Recommendations Based on No Evidence

A very unusual recommendation of the Task Force is the recommendation that women over 74 not receive screening mammography. The reason given is that there is not enough research to prove that there is an increase in life span in those women that receive screening mammography. Now, there may be very good reasons not to screen women above a certain age, but the reason given that not enough research is available is just bizarre.

Recommendations Based on Less Than State-of-the-Art Equipment

In another move that is baffling to us that perform mammography is the reliance on studies using film mammography. In a time when most departments are replacing their film based mammogram systems with digital systems, that is kind of like basing a review of the fuel efficiency of cars on studies of the Chevy Nova. Sure, there may be a few still on the road, but they are antiquated and don’t provide a fair representation of the state-of-the-art today. Radiologists will tell you that digital mammography is superior to film mammography, but you don’t even have to take our word for it. In fact, multiple studies, contrary to the statements by the Task Force, also prove this contention. It is puzzling that, with a pronouncement of this magnitude, the latest research was not included.

Lack of Understanding of the Physiology of Breast Cancer

All of the above raise concern about the conclusions of the Task Force, but any, in and of itself, is not a disqualifying factor. What is most disturbing about the recommendations is that they lack an understanding of the physiology of breast cancer. Namely, while breast cancer is more rare in younger patients, the cancer that occurs in younger patients tends to be more aggressive than in older patients. Because of this, we can perform mammograms less often in older women, because the cancer grows more slowly. In a faster growing cancer, however, more frequent screenings are required since the longer interval between screenings allows the cancer to become larger and less amenable to cure. In addition, when talking about Quality Adjusted Life Years, a 40 year old has a longer expected life span than a 65 year old, so the screening in this case is actually more efficient per year of life saved.

My Recommendations

Until I see some new, more convincing data, my recommendations are to follow the American Cancer Society Guidelines. Women with no additional risk factors should have screening mammograms yearly, beginning at age 40. Women with increased risk factors, such as strong first-degree family history or the HrBCA gene should consult their doctor about when to begin screening mammograms.

Mammograms in Elderly Patients

Elderly or infirm women should consult with their physician regarding screening mammograms. It really does not make sense to perform a screening mammogram in a patient with only a couple of years of life expectancy remaining. To put an absolute age on when to discontinue screening mammograms makes very little sense. People are living healthy lives much longer now. There are many 75 year olds that may have a life expectancy of 20 years or more. To deny these women screening mammograms is a crime. On the other hand, if there is no intention to undergo treatment, should cancer be found, a screening mammogram is a waste of money.

Ramifications for Health Care Reform

We have been assured by the administration that this task force recommendation will not affect policy as regards PelosiCare. In fact, nothing could be further from the truth. Specifically, Section 2713 of the Senate Health Bill would give the recommendations of the U.S. Preventive Services Task Force the force of law by requiring all health insurance plans to provide coverage (with no patient co-pays) for “items or services that have in effect a rating of “A” or “B” [recommended] in the current recommendations of the United States Preventive Services Task Force.” White House budget director, Peter Orszag, states that with health care reform, "An independent Medicare commission … will ensure that reforming the health-care system is not a one-time event but an ongoing process that implements the most recent progress in medical science with the goal of improving care and lowering costs." (Washington Post; 11/20/09; Bills Offer Clear Path to Better Health Care; Orszag, Peter)

In addition, this is just the type of recommendation that insurance companies look for in order to deny services. So for the Obama administration to claim that these recommendations will not be used to "ration" medical services are mistaken, at best, and dishonest, at worst. All of us should keep this in mind when we consider increasing government involvement in our health care system.

Adopting these guidelines will cost lives. The media and Democrats went into a frenzy when Sarah Palin mentioned death panels. This may not be what she had in mind, but with recommendations like this, it is a death panel, nevertheless.