The new breast cancer screening guidelines demonstrate yet again why savvy patients need to understand the numbers behind risk/benefit studies before making the very personal decision about whether and how often to get a cancer screening test.
The recommendations of the U.S. Preventive Services Task Force that women hold off on routine mammograms until age 50, and then get one every couple of years instead of every year, are based on sophisticated statistical modeling that aimed to get the most benefit at the least harm. The benefit is saving lives. The harm is overdiagnosing, overtreating, and needlessly frightening women who receive “false alarms” with mammograms.
The key fact that women — and all patients — need to understand is that your risk of a “false alarm” depends on your risk for the disease. The lower your natural risk of disease, the higher the risk of a false alarm, and vice versa. The reason the task force said that women between ages 40 and 49 don’t need screening is that with the low risk of disease in that age group, the chance of needless false alarm is much higher, and the benefit of discovering the occasional cancer is much lower.
For example, one statistical model, called the Stanford model, which the task force looked at, concluded that if women between ages 40 and 69 got mammograms every year, there would be on average 2,250 false alarms — “false positive” results — in every 1,000 women — an average of two per woman tested over ten years. On the other hand, there would be a total of 164 years of life gained among these 1,000 women. When the mammograms were reduced to every other year, and only given between ages 50 and 69, the years of life gained were 99 and the false alarm rate went down to 780 per 1,000. That is the tradeoff.
Those extra 65 years (164 versus 99) of life gained among the 1,000 women also come at an economic cost: if 1,000 women have to get 10 mammograms each over 10 years, that cost is around $10 million (assuming $1,000 per mammogram).
We don’t like to think about “rationing” medical care, but sensible decisions on both a personal and societal level have to take into account how much we’re spending for how much gain, and what else the same amount of money could be spent on.
In my book, “The Life You Save: Nine Steps to Finding the Best Medical Care — and Avoiding the Worst,” I have an extensive chapter on the risks and benefits of breast cancer screening at different ages. Here is an excerpt:
What about saving lives with screening tests intended to catch early cancers? Here is how that statistic of a 25 percent reduction in deaths translates in the real world. If you thought that means that of every 100 women screened, 25 of them would be saved by mammograms,
you would be making a natural assumption that many others have made, but you would be very far off the mark. The actual numbers come from a series of studies in Sweden involving some 280,000 women. Of those over age forty who did not undergo mammograms,
4 in 1,000 died of breast cancer over the ten years of the study. Of those over age forty who did have mammograms, 3 in 1,000 died of breast cancer over the same ten years. The reduction from 4 to 3 per 1,000 is where the 25 percent number comes from. Put another way, for every 1,000 women who participate in mammogram screening for ten years,
1 of them will be saved from dying of breast cancer.7 The odds of saving 1 life are a little improved if screening begins only at age fifty instead of forty. Of every 270 women who start screening mammograms at age fifty and undergo one every other year for the next twenty years,
1 life will be saved-or about 4 in 1,000, which is a lot more than the 1 in 1,000 lives saved for starting mammograms at age forty.
The new statistics from the Preventive Services Task Force concluded that the benefit from mammograms in women aged 40-49 was even lower than the Swedish estimates — more on the order of one life saved for every 1,900 women screened, instead of one in 1,000.
Here’s the bottom line, as I put it in my book:
Are those kinds of odds worth it to undergo the trouble of regular screening? That is a personal decision. My only point is that to make that decision rationally and realistically, you need to look at the numbers with eyes wide open.