False-positive mammograms and overtreatment of screen-detected tumors that may lie dormant for years cost the United States an estimated $4 billion annually, a new analysis suggests.
Mei-Sing Ong, Ph.D., and Kenneth D. Mandl, M.D., M.P.H., both of Harvard University in Cambridge, Mass., led the study. Its findings come when several organizations, including the American Cancer Society and the U.S. Preventive Services Task Force, are updating recommendations for when and how often average-risk women should be screened. The revised recommendations are expected before next year.
Using claim data filed with a major U.S. insurance plan, the researchers found that costs ran higher than previously documented, with false-positive readings accounting for most of the expense at $2.8 billion. The study, published in the April 15 Health Affairs, involved 702,154 women diagnosed and treated for breast cancer between 2011 and 2013. Roughly half the women screened were aged 40–49 years.
“I don’t think we’ve solved any problems in terms of what the screening process should be,” Mandl said, referring to the cancer community’s decades-old disagreement over the best preventive strategy. “But [the study] gives us a sense of the magnitude of harm when the benefits are controversial, at best.” Mandl is director of informatics at Boston Children’s Hospital and a professor of pediatrics at Harvard Medical School. Ong is a postdoctoral research fellow at Boston Children’s Hospital.
Controversy remains high regarding mammography’s net benefit to women in their 40s who have no known family history or other risk factors for disease.
The cancer society currently recommends that annual screening begin at age 40 years and continue throughout a woman’s lifetime, if she’s healthy. However, the 16-member task force of independent experts, which advises the government, recommends that average-risk women aged 50–74 years begin screening every other year.
Screening at an earlier age, the group said, should be left to a woman’s choice with her health care provider, whereas women older than 74 years should consult their physicians, as well, with the lack of useful data for or against screening. Similar advice is part of the task force’s updated guidelines, now under review, after a public comment period.
“Women may want to begin screening in their 40s,” said Kirsten Bibbins-Domingo, M.D., Ph.D., vice chair of the task force and a professor of medicine, epidemiology, and biostatistics at the University of California, San Francisco. “It’s an important tool. But in this age group, the benefits are closer to the harms.”Among the harms: false-positive mammograms that cause at least temporary emotional anxiety and distress, according to Bibbins-Domingo and others. These false alarms occur often in women in their 40s, whose dense breast tissue can make tumors harder to see. A risk factor for breast cancer, breast density occurs in at least half of women, and doctors still don’t know which patterns carry the greatest risk, Bibbins-Domingo said.
Investigators in the Harvard study estimated that false positives occur during mammography screening in roughly 11% of women overall. The figure rose to 13% when the algorithm incorporated ultrasound and magnetic resonance imaging, with women aged 40–49 years more likely to have these diagnostic workups than older women. And, from earlier studies, researchers cited a 61% cumulative probability of a false-positive recall in a woman aged 40–50 years after a decade of screening.
Mandl said he and Ong measured false-positive rates directly, examining claim data for follow-up tests that lacked a cancer diagnosis. Not so, however, for overdiagnosis—generically defined as the diagnosis of breast lesions unlikely to threaten a women’s health during her lifetime, yet treated all the same. To arrive at the study’s $1 billion annual cost estimate, according to Mandl, they relied on published overdiagnosis rates gleaned from several randomized trials. One trial included a 25-year follow-up of the Canadian National Breast Screening Study, which appeared in the British Medical Journal (BMJ 2014;348:g366). Twenty-two percent of women in the study’s screened cohort arm were diagnosed with breast cancers that never progressed but resulted in treatment anyway.
“It’s a convincing piece of possible overdiagnosis,” Mandl said. He attributed overdiagnosis primarily to interpreting mammographic images “based on a conceptual basis of cancer that might not be accurate,” as well as ductal carcinoma in situ. Such diagnoses jump after the introduction of screening mammography, Mandl said. “Some of it is clearly a disease of mammography, although that’s not to say every lesion is nonthreatening. Some are.”
Real Value
As the American Cancer Society grapples with how, or whether, to change its own screening recommendations, Richard Wender, M.D., the society’s chief cancer control officer, said his main criticism of the Harvard study is the implication that mammography carries cost without benefit to women in their 40s. “[Screening has] real value, particularly in preventing the death of otherwise healthy women in this age group,” he said. Some 17.7%, or nearly one in five women who die of breast cancer, receive a diagnosis of breast cancer in their 40s, according to Wender.
Wender also questioned the analytical approach used. “If you do modeling, there’s an obligation to publish a range of values,” he said. “But they took the high side of overdiagnosis (22%) to estimate cost.”
Although many cancer researchers agree that overdiagnosis does occur during screening mammography, Wender said, “we don’t know how much there is” or how many women undergo unneeded treatment. According to his own estimates, overdiagnosis happens in 3% of invasive breast cancers, 20%–30% of ductal carcinoma in situ, and 10% overall.
Kenneth D. Mandl, M.D., M.P.H.
Moreover, not all false-positive mammograms are alike, Wender said. Most can be resolved through additional screening views or a 6-month follow-up, whereas perhaps 15% require biopsy, he said. “I don’t mean to diminish the impact of living through the emotional anxiety of a biopsy, but that’s just the nature of looking for cancer. Eventually, you need tissue to confirm it.”
Finding Common Ground
Will the cancer community come together with uniform guidelines for mammography screening? Mandl said the economic impact of mammography needs to be part of any future discussions involving appropriate populations for screening. And, Wender said, he hopes 2015 will be the year that a clearer message emerges. “We need to emphasize areas of commonality,” he said. “We want the health message to be the predominant one.”
But before that happens, few would dispute the need for further improvements in assessing individual risk—possibly through molecular profiling—and better communicating those risks to women.
“There’s still so much we don’t know,” said Nancy Keating, M.D. M.P.H., professor of health care policy at Harvard Medical School. “In all women who have deadly tumors, mammography screening helps only a small percentage.” Women in their 40s, for example, have a low risk for breast cancer numerically, but mammography screening reduces mortality from these more aggressive cancers by only 15%. That means 85% will die, even with mammography, she said.
Still, Keating added, “I don’t think in America we are ready to stop screening women in their 40s. It’s not that there’s no benefit; it’s just quite small.” She called the recommendations of the task force reasonable, adding, “there are definitely women who are happy having mammograms every 2 years; others still want it every year.”
About the study by her Harvard colleagues, Keating said: Though others may quibble with its findings, or the final cost estimates, “society needs to know the cost is real.”
Meanwhile, the task force determined that existing evidence for newer digital mammography cannot yet balance benefits against harms or risk. Keating agreed. “The technique is so much better” than film-based mammography, she said. “But so is the treatment for breast cancer. Even when tumors are detected a little later, women are doing much better against this disease.”