The Case of The Woman Who Refused Her Mammogram

Recently a colleague of mine expressed frustration about a 52-year-old woman who refused mammography. Given my interest in preventive health, he came to me to ask how he should go about convincing the patient to get her mammogram. He was surprised by my response: as long as the patient understands the risks and benefits, then it's reasonable for her to refuse.

On one level, this response is downright shocking. Mammograms are proven to save lives; we must convince the patient to do what is best. On other hand, my response is straightforward. Patients have the right to refuse medical services, in the most cynical view, my response was a medical-legal one (document understanding of risk and benefits in case she later develops breast cancer and sues). This becomes the classic paternalistic versus patient autonomy debate. But, in truth, my response was a reflection of neither viewpoint. It was based on a basic fact: it is as reasonable to undergo mammography as it is to forgo it. Mammograms are no panacea and have significant limitations that for certain women make the test not worth the risks.

Let's begin by reviewing the the benefits and risks of mammography:

The Benefits

Reduced death from breast cancer. Mammograms sometimes lead to the detection of cancer earlier than it would be found otherwise; sometimes earlier detection leads to earlier treatment that reduces morbidity and mortality. (Just because mammography finds breast cancer earlier does not mean it necessarily leads to improved outcomes; hence sometimes.) For women ages 50 to 69 mammography has been shown to reduce deaths from breast cancer by 33 percent. This percentage though is a relative risk reduction; absolute risk reduction is much smaller. A less cited statistic is that to prevent one death from breast cancer, 1000 women ages 50 to 69 need to be screened (that is, the number needed to screen is 1000).

The Risks

• False negatives: Sometimes mammograms fail to detect breast cancers and thus give a false negative result. It is estimated that one in five breast cancers present at the time of screening is missed. False negatives cause harm by delaying diagnosis and treatment and creating a false sense of security.
• False positives: Sometimes mammograms suggest breast cancer is present when it is not there. This leads to unnecessary testing including biopsies, which are anxiety-provoking, expensive, and potentially disfiguring.
• Overdiagnosis: Much less talked about is the risk of overdiagnosis. Overdiagnosis refers to cancers that will never cause harm if untreated. As I discussed in a recent post, cancer often solicits a knee-jerk, "get this thing out of me!" response. But the truth is that many cancers, if left untreated, would never cause harm to us during our lifetimes. Sometimes this is because the cancer is dormant or regresses; other times it is because we die of other causes before the cancer becomes clinically relevant. The trouble is that with the current state of technology, doctors cannot always tell which cancers will cause harm and which ones will not, and as a result treat all cancers aggressively. The net result is that we risk the chance that women will undergo treatment for breast cancer including surgery and radiation without any health benefits. A recent study suggested that one in three breast cancers detected in population-based screening programs are overdiagnoses (see reference below).

An editorial in the British Medical Journal published just last week presented an excellent framework for translating these risks and benefits in a way that patients can use to make informed decisions. The following table is taken directly from this editorial. It shows the credits (benefits) and debits (harms) of screening 1000 50-year-old women with mammography every year for 10 years. For example, it shows that for every 1000 women screened, 10 to 15 women will be diagnosed with breast cancer earlier than they would be otherwise but without any improvement in prognosis (or outcome).

In the editorial, the authors suggest that more data is needed to calculate more precise estimates of the credits and debits. They note that for many women the critical threshold for being in favor of or against mammography may be within the wide ranges of these estimates.

Given the risks and benefits of mammography, it is reasonable that my colleague's patient, or any woman for that matter, refuse mammography. The question for health care providers is not, "How can I get my patient to get her mammogram?" But rather, "How can I make sure my patient understands the benefits and risks of mammography and makes the right decision for herself?" Instead of focusing on getting 100 percent of our patients to meet the guidelines (as quality improvement agencies and pay-per-performance would have us do) what we should be focusing on is making sure that 100 percent of our patients are making informed healthcare decisions based on a solid understanding of the best available evidence. Getting a mammogram is not an obvious choice for all women, so we need to step up our efforts at full consent and make sure we are not over-selling the benefits and under-appreciating the risks of breast cancer screening.

Copyright Shantanu Nundy, M.D.