Recent proposed changes in US breast cancer screening recommendations raise interesting points about how we discuss risks, trends and numbers. A preventive medicine task force reviewed available evidence and concluded that low-risk women age 40 to 49 should forgo mammograms, and older women should have them less frequently than previously recommended. I won't weigh in on whether these recommendations make sense. Cancer screening is a personal decision, and everyone has the right to oppose proposed changes based on scientific evidence or informed clinical judgments. But the conversation has been fascinating.

Viewed from a population level, cancer risks are high. About 192,000 women will be diagnosed with breast cancer in 2009, some 40,000 will die, for a death rate of about 25 per 100,000. But for the average women, the risk is quite small. The average 40-49 year old has a 1 in 69, or 1.44%, chance of being diagnosed with breast cancer. That's an important age group for the current screening debate.

Opposition to reduced screening intensity implies that more screening is better. To a certain extent that's true, but eventually you reach a point where you are treating abnormalities that would never have evolved into cancer. We know that about 25% of treatment may be unnecessary because 1 in 4 breast cancers never evolve into malignancies that threaten health. Since we don't know which 25% are harmless, we treat all that we find. What this tells us is that we aren't doing as well as we think in terms of treatment outcomes. We may only be doing 75% as good a job as current statistics would imply, and 25% of women who receive mastectomies, radiation and/or chemotherapy may well be doing so for something that wouldn't have harmed them if left alone.

Our discomfort with capitulating in our cultural "war" on cancer has fueled a game of ideological twister between the sometimes contradictory views of public health (looking at populations) and clinical care (treating individuals.) It goes something like this: Public health weighs harms and benefits to the population as a whole, concluding that some benefits (a few lives saved) are outweighed by the overall cost (in dollars or health metrics). Clinicians counter that "some benefit" is actually patient lives being saved. Public health talks about impersonal statistics and risk ratios, clinicians counter with emotional anecdotes about people with names and stories.

The current debate twists this logic, enhancing the magnitude of risk posed by breast cancer by citing population statistics that apply to women as a group, since the threat to individual women is small (with some exceptions where these recommendations don't apply). This has been combined with the clinical perspective which paints colorful anecdotes about individual patients who appear to have been saved by early mammography. And some continue to imply that less mammography amounts to the devaluing of women's lives, but the same could've been said about men last year when prostate cancer screening guidelines called for a reduction in intensity.