More Mammogram Madness

A new study shows women who won't benefit still get the tests.

Not to keep beating up on mammograms, as I've done here and here, but the latest study of who is receiving them doesn't inspire confidence that doctors are following evidence-based practices and putting their patients' welfare first.

In a nutshell: a significant percentage of elderly women with severe dementia are getting screened. Such women have an average life expectancy of only 3.3 years. Yet science-based guidelines from the American Cancer Society and other experts say that women with a life expectancy of less than five years should not be screened (because any cancer that's found will not grow fast enough to cut into her remaining years). Even more disturbing, if an elderly woman with severe dementia is also married and with a net worth of $100,000 or more, she is more than twice as likely to get these inappropriate mammograms as her poorer peers.

For the study, which was just posted in the online edition of the American Journal of Public Health, scientists combed through the health and Medicare records of 2,131 women over 70. Seventy-two percent had normal cognitive status, 29 percent had mild-to-moderate cognitive impairment, and 9 percent had severe cognitive impairment. The good news is that "only" 18 percent of the women with severe dementia received screening mammography, compared with 45 percent of women with normal cognitive status. (The rate of 45 percent among older women with normal cognitive status seems about right: those in good health who are expected to live for at least another five years can benefit from screening and should get it, while those in poorer health with less than five years to live are unlikely to benefit and are right to forgo screening.) I call the 18 percent good news because at least it's lower than the rate for mentally healthy elderly women: screening mammography is inappropriate for elderly women with severe dementia, and not only because their short life expectancy makes them unlikely to benefit. In addition, mammograms often lead to follow-up tests such as biopsies that are stressful enough for healthy women, but women with dementia may not even understand why they are getting the tests. The tests can also lead to surgery for growths that would never have caused problems in the woman's lifetime. And mammograms plus follow-up divert time and focus away from the medical care that makes much more difference to someone with severe dementia.

That's why "even a screening rate of 18 percent may be considered too high,'' says Kala Mehta, a geriatrics researcher at the University of California, San Francisco, who led the study. To benefit from screening mammography, "a woman must have a life expectancy of at least four to five years,'' whereas the severely cognitively impaired women had an average life expectancy of 3.3 years. "The potential harms are likely to outweigh the benefits.''

More troubling, among elderly women with severe cognitive impairment who were married and had a net worth of more than $100,000, 47 percent got screening mammograms. Given that the rate should be about zero, it's clear that some physicians are ignoring the science and, arguably, the humanity of their elderly, confused patients. "I don't think most people would say that putting a very demented older woman through screening mammography is a good thing," says Louise C. Walter, an associate professor of medicine at UCSF who oversaw the study. "Screening an active, healthy, independent 75-year-old woman who is likely to live for another 15 or 20 years is one thing, but severely cognitively impaired older women with limited life expectancies have immediate health challenges that must take priority."

Why, then, are so many elderly, cognitively impaired women getting inappropriate mammograms—and why are they more likely to get them if they're married and wealthy? The marriage connection makes sense: the loving husband, knowing his wife dutifully got her mammogram every year, keeps taking her for her screening, not knowing (and perhaps not being told) that it is unlikely to help her and may hurt her. The money connection is trickier, since Medicare reimburses for mammograms, and doctors who accept Medicare patients agree to accept that reimbursement as full payment. But some doctors do not accept Medicare, and so may be more likely to recommend screening if the woman can pay out of pocket or has private, supplemental insurance—both more likely if she is wealthy. Indeed, a 2008 study also found that "wealthier older women with limited prognoses are often screened when they are unlikely to benefit."

If the women in the UCSF study are typical, then more than 120,000 screening mammograms are performed on elderly, severely cognitively impaired women in the U.S. every year. Medicare reimburses providers $81 for a plain-film mammogram and $130 for a digital mammogram, so the cost of these unnecessary screenings tops $12 million a year.

But—and this keeps getting lost in the health-care debate about medical treatments that don't benefit—it is not primarily about the money. It is about care that provides no benefit and may prove harmful. Since elderly patients with dementia generally live less than five years, the scientists write, they "are unlikely to benefit from screening mammography," and are more likely to "experience harm." Giving mammograms to elderly, cognitively impaired women with only a few years to live may not be as outrageous as giving annual Pap tests to women without a cervix and no risk factors for cervical cancer, but it's pretty darn close.