Screening Tests Part 2: Mammograms and Breast Cancer Screening.

Nothing is more controversial and yet so established as breast cancer screening through mammograms. As we begin to discuss the risks and benefits of screening, it is always best to start with mammography because it is illustrative of the complexity and uncertainty that beset most screening tests. A recent Danish study fueled the controversy by showing that one out of three women treated for breast cancer detected by mammography did not need the treatment that they received. In other words, the breast cancer would not have killed them if left alone. Researchers like H. Gilbert Welch at Dartmouth have cited similar numbers, showing that 20% of breast cancers may regress even without treatment, and in a New York Times article stating that 3-14 women out of 1000 who receive mammograms are treated unnecessarily due to false positives . Groups such as radiology societies argue that mammography is life saving and that the over-treatment ascribed to them is hyperbole. They often site a 20% reduction in breast cancer death among women who obtain annual mammograms.

As is often the case with screening, the truth overlaps both sides of the debate: there is risk and benefit, and each individual, with the help of a medical professional and accurate data, must decide his/her own comfort level with the testing. Of note, Medicare’s quality indicators insist that all women between the age of 50-75 get a mammogram at least every two years, and if they do not then their primary care doctor will be financially penalized through the new “quality and value” report care that will determine a substantial part of physician pay starting in 2017. Unfortunately, as with all screening tests, Medicare’s mandate fails to acknowledge the uncertainty of this and many other tests, and thus dissuades a woman from making a decision that is best for her, while discouraging doctors from discussing risks and benefits.

If you have not read it already, it is worth spending 3 minutes looking at Jay Hancock’s article in Kaiser Health News and the accompanying video that describes how BRCTs can simplify decisions about breast cancer screening. In fact, the 20% reduction in death cited by mammography advocates (a misleading relative number) translates to mean that out of 1000 women who have lifetime screening, one will avert a breast cancer death, as shown in the BRCT at the top of this article. That number is higher for people at increased risk of breast cancer (such as those with a strong family history), but drops lower as people age. With minor fluctuations, that number has been consistent among most studies. Thus, when deciding whether to get a mammogram, the benefit of 1/1000 should be considered and then individualized to each patient’s particular circumstances by her doctor. In most cases 999/1000 people (99.9%) will derive no benefit from annual mammography.

What is the risk? There is a very high rate of false positive mammograms that lead to unnecessary tests and treatments in people without cancer, as well as true positives (as the above quoted Danish study and others demonstrate) that find cancers that would have been harmless if left alone. As with all cancer tests, the false positive rate and the incidence of harmless cancers that are discovered need to be considered by any patient who is going to engage in the screening process. We will see with PSA testing how important those determinations are.

Currently, according to the most robust studies and independent health data sites such as Cochrane and the US Preventive Services Task Force, approximately 500/1000 people who undergo lifetime mammography screening will have at least one false positive test. That means that out of 1000 people who undergo lifetime screening mammograms, 500 of them will be told at least once that their mammogram is abnormal and could be indicative of cancer. In most cases, those women will be called back for additional testing and found to be cancer-free. 64 woman out of 1000 screened will have at least one biopsy during their lives (see BRCT to the left) due to a false positive read, and 10 out of 1000 screened women will have treatment for cancers that either they do not have or would not have harmed them. False positives and unnecessary biopsies can cause stress; unnecessary biopsies and treatments can cause harm. It is estimated that the cost to insurances like Medicare for false positive mammograms exceeds $4 billion a year.

Ultimately the risks and benefits of mammograms to individual patients, to the insurance industry, and to women’s health are not clear. Each patient along with her doctor should assess her willingness to accept the inherent uncertainty of screening. Similarly, insurance companies may want to reconsider compelling doctors to order these tests, when there is very little evidence that they confer high value and save the system money. Rather than grade doctors using clinical guidelines that assess performance on the very nebulous contention that all women must get mammograms for their wellbeing, such guidelines would be more helpful if they encouraged shared decision making using tools such as BRCTs. As with most screening tests, mammograms are best assessed using easily understood and accurate information, and there is no right answer as to whether the test’s merits or flaws are more compelling to each woman.