Economic Savings and Costs of Periodic Mammographic Screening in the Workplace

We easily believe that which we wish.-Pierre Corneille (1606-1684)

This politically correct paper suffers from two basic problems common to the medical literature focusing on costs, effectiveness, cost-effectiveness, and benefits. First, society can go medically bankrupt providing care that is "cost-effective." Second, wishing won't make it so.

Bankrupting Society With Cost-Effective Care

This paper suggests that the health benefit of mammographic screening in the workplace is substantial and falls within the range of cost-effective medical care. No one will argue that mammography for women over the age of 50 reduces the risk of death from breast cancer by approximately 30%. Few other interventions are as clearly effective in screening.

The problem is in paying for this benefit. The benefit of mammographic screening in women over the age of 50 is substantial enough and the cost of the test low enough that the cost-effectiveness ratio is within accepted bounds. Eddy and others have estimated this ratio to be $20,000 to $50,000 per year of life saved, less than the currently accepted "benchmark" of about $50,000 [1,2].

However, society can go bankrupt even when providing "cost-effective" care. Remember that this is additional money that must be expended. It is not cost-savings by a long stretch. Only pneumococcal and flu vaccines yield true cost savings in adults; all other medical interventions add cost even it they provide some benefit. If we keep adding procedures that are "cost-effective," we could clearly be spending 20% of our gross national product on health care within a few years. There are several ways to divide the "medical commons," but it must be recognized that once all the "grass" is gone, there is no more green stuff.

The cost-effectiveness ratio of mammographic screening in women under the age of 50 is even more problematic. Assuming that it works (an assumption I believe but cannot prove), and using David Eddy's estimate of the benefit (which is reasonably generous), the cost is still $232,000 per year of life saved [3]. So, here is an intervention that is effective but probably too effective for society to afford, and it was not included in a recent evidence-based benefit package [4].

Wishing Won't Make It So

I disagree with some of the wished-for conclusions of the authors. I do this with reluctance because I hope that they are correct, but I simply cannot prove it.

First, there is no evidence to date that a group of women at high risk for breast cancer can be effectively screened any more than can overall populations. These women at high risk can be identified [5]. It certainly makes sense that women with the highest risk for breast cancer would be those who might benefit from intensive screening, because the detection rate would be higher. It may also be true that the denser mammographic pattern in high-risk women makes the test less efficient in detecting curable disease, and leads to a much higher rate of nonproductive biopsies. However, there are no reliable data to support these contentions.

Second, the authors hope that screening programs in the workplace will be of high quality. Note that there is substantial variation in mammographic screening technique as well as interpretation. It is my own bias that screening mammography should be done only in programs certified by the American College of Radiology (ACR), and that the last thing the United States needs is more small-volume, non-centrally-located, non-ACR-certified units, with mammograms read by people for whom it is a hobby rather than a career. Mobile mammography units may assist in solving this conundrum of getting high-quality units to the people who need it most.

The authors suggest that the cost per screening mammogram can be reduced. If fewer views are done or images are of lower quality, effectiveness will diminish.

The authors mention the costs incurred by a company due to lost productivity during screening. I do not know a single company that hires a "temp" for yearly mammography; the losses should be trivial, equivalent to those for any routine medical care. Furthermore, all other cost-effectiveness ratios take into account only direct medical care costs. We should not define a new standard here to make on-site mammography look better than mammography for women who do not work outside the home, or the treatment of hypertension in men.

The authors note the recent National Cancer Institute study showing that costs increase as cancer is diagnosed earlier [6]. I cannot tell if they are using this as an argument for or against mammographic screening. It certainly suggests that cost-savings may not accrue to mammographic screening, but the public good is so high for women over age 50 that costs are almost irrelevant.

Finally, the authors suggest that the employer assume responsibility for coordination of care and counseling. Few employers have the means or desire to do this, and data on efficacy and costs are lacking.