Mammography: Applying Sound Mind to an Emotional Controversy


Synopsis: Three papers recently published in the Annals of Internal Medicine bring to light the recent debate about the use of mammography for women in their 40s. It appears that the procedure is only marginally effective in that age group, and, when examined in the context of cost effectiveness, the results are staggering. If women in this age group are routinely screened, the estimated cost per year of life saved is approximately $150,000. Some have argued quite cogently that cost-effectiveness data are irrelevant and dehumanizing and should not be included in a debate about saving lives. The American Cancer Society and the National Cancer Institute have responded by recommending mammography for women in their 40s, but it is quite clear the discussions will continue. Taken together, these three reports urge a level headed re-examination of all of the issues involved in the recommendations regarding mammography in this age group and present a prescription for a more even-tempered discussion on the matter.

Sources: Salzmann P, et al. Ann Intern Med 1997; 127:955-965; Ransohoff DF, Harris RP. Ann Intern Med 1997;127:1029-1034. Eddy DM. Ann Intern Med 1997;127:1035-1036.

The issue of mammogram recommendation for women in the 40-49 year age group has received an inordinate amount of attention in the last few months, and much of it focused around an NIH Consensus Development Conference in January 19971 that concluded "the available data do not warrant a single recommendation for mammography for all women in their 40s." The American Cancer Society, examining the same data, which include at least eight randomized and three case-controlled trials, published a contrasting recommendation, yearly mammography for women in this age group.2 The issue went back to the NIH. The NCI, upon review by the National Cancer Advisory Board, suggested mammography every 1-2 years for women 40-49 years of age.3 However, it seems the controversy hasn’t ended. In three articles recently published in the Annals of Internal Medicine (an original research paper, a "perspective," and an editorial), the issue is re-examined. In the research presented by Salzmann and colleagues, a cost-effectiveness analysis using Markov and Monte Carlo models revealed that screening women from 50-69 years of age improved life expectancy by 12 days at a cost of $704 per woman, resulting in a cost of $21,400 per year of life saved. However, extending screening to include women 40-49 years of age improved life expectancy by only 2.5 days at a cost of $676 per woman. By adding the younger age group, there was an increase in cost of screening by $105,000 per year of life saved.

Eddy, in his editorial, points out that cost effectiveness is only one component of the debate. Another issue pertains to effectiveness rather than cost effectiveness. He points out that the data regarding effectiveness are incomplete, but what data are available indicate a very small and delayed benefit from routine screening in this group (about 1 life saved per 1000 women screened after 10 years), and, because breast cancer is more rare in this age group, the likelihood that a woman would have a false positive mammogram (usually necessitating a biopsy) approached 25% over 10 years. He criticizes the prior debates about screening in younger women as irrational and emotional without a careful evaluation of the evidence. He also predicts that such would occur again, unless the process of review and analysis is agreed upon before entering into a debate. Ransohoff and Harris draw the same conclusion. They point out that the whole process has become highly charged by "single minded advocacy groups and a kind of gender rivalry." They suggest that the contentious nature of the debate would be defused if both sides agreed on several things: 1) the evidence from the clinical trials is widely agreed upon and, thus, a main task now is to factor in the values of individual women who are making the decisions; 2) that the values of women may differ substantially and those differences should be respected; 3) both individuals and the public should be fully and fairly informed about the pros and cons of screening; and 4) that cost effectiveness should at least be considered during the decision-making process.


Investigators generally agree that screening interventions that cost less than $50,000 per year of life saved are considered cost effective. Thus, screening for breast cancer in the 50-69 year old age group is considered cost effective at $21,400 per year of life saved, but the cost is nearly seven times higher if women in their 40s are screened. The reasons for this are complex but certainly include the low incidence of disease in this age group, the small effectiveness of the screening tool (mammography), and the long delay before benefit appears. Some have argued that relying on cost effectiveness assessment is dehumanizing and that even though the costs are large, if only a handful of women are saved, it is worth the expense. It’s hard to argue against that point of view, and it is at just this point where the discussion leaves the realm of science and enters that of emotion.

We are not living in a society of unrestricted assets, and to be fully engaged in this debate we need to accept the responsibility of accounting for the distribution of limited resources. Eddy points out in his editorial that cost-effectiveness data are only one component of the debate and are best considered when put into a more global perspective. Thus, when resources are limited, wouldn’t it be a more prudent use of resources to make a stronger effort to screen the women in the 50-69 age group who go unscreened (estimated to be 40%)—in whom the effectiveness is established and the costs per year saved are relatively low? This strategy would save many more lives. Certainly, other interventions may save more lives from other diseases at a lower cost, but their availability may be restricted because of the resources drained by the high cost of screening young women for breast cancer with very low benefit.

The debate, to date, has been in large part dysfunctional with strong input from pro-mammography advocacy groups. The importance of these three articles is that, taken together, they clarify just how powerful public sentiment and political factors are in the establishment of policy. Of course, there is always the hope that new, improved mammography techniques will be developed that are more sensitive, more specific, and less expensive. Short of that, the debate should not be closed. Instead, women in the 40-49 year age group should be fully informed of the very high rate of false positive mammograms, the very small but measureable benefit, and each woman should decide for herself whether the value of the test is worth the risk.


1. NIH Consensus Statement: Breast cancer screening for women ages 40-49. J Natl Cancer Inst 1997;89: 1015-1120.

2. American Cancer Society. Workshop on Guidelines for Breast Cancer Detection. Chicago, March 7-9, 1997.

3. National Cancer Institute. National Cancer Advisory Board issues mammography screening recommendations. National Institutes of Health. 27 March 1997.

Regarding recommendations for regular screening with mammography to detect early breast cancer, which of the following statements is true?

a. Screening women between the ages of 40-49 is as effective as in older age groups, but it presents an economic challenge because of American population demography and the large number of individuals in that age group now as well as those projected for the next thirty years.

b. Screening in the 50-65-year-old age group is both currently recommended and highly performed, with approximately 80% of individuals in this age group being screened at regular intervals.

c. Screening in the 40-49 year old age group is not only less cost effective, it is less effective when compared to screening in the 50-65 year old age group.

d. Screening in all women over the age of 40 is recommended because it is has been shown to be both effective and cost effective.

e. Health services researchers examining cost benefit ratios of screening interventions consider a test to be cost effective if the calculated cost-effective ratio (estimated cost for one added year of life) is less than $150,000.