In a recent article in the Journal of the American Medical Association, Welch and Geisel colleague Honor J. Passow calculated that for every 50-year-old woman who avoids death from breast cancer through annual mammograms, at least 153 (and likely far more) suffer false alarms and at least four are "treated needlessly with surgery, radiation and/or chemotherapy."
In fact, over a decade of annual screenings, half or more of patients will be the victims of false positives that at best induce anxiety and at worst require surgery or other treatments for cancers that would not harm them. "My value judgment is that a population-based screening program that alarms half the population is outrageous," says Welch.
The point is not that the mass of American women should avoid annual mammograms. It's that they shouldn't do them without understanding that the procedure carries a small prospect of a large benefit and a large prospect of a small harm.
Women are not the only people who face this sort of dilemma. A widespread test for prostate cancer works almost identically. It detects a lot of cancers that are either unlikely to be fatal without treatment or very likely to be fatal even with treatment, while exposing many men to needless fear as well as treatments with serious side effects. The U.S. Preventive Services Task Force now recommends against it for routine screening.
Annual mammograms for breast cancer are expensive. Putting every woman through it annually starting at age 40 would cost a total of $10 billion a year. Starting the screenings at age 50 and doing them every other year until age 69 would cost $8 billion a year.
That's about twice what the government's National Cancer Institute spends annually on cancer research. Money spent on mammograms could be used in ways that would save more lives.
How should we feel about a health care system that has long put so much faith in such a flawed instrument? At least a little foolish.