My high school Latin is a little rusty. All I have left is a few proverbs bouncing around in my head. But one of those turns out to be relevant to today’s topic: the much-debated value of breast-cancer screening.
The phrase that comes to mind is: Post hoc, ergo propter hoc. Which, if I’m remembering correctly, means “after this, therefore because of this.” It’s a logical fallacy: one event comes after another, therefore it must have been caused by the other.
Or, to apply it to breast-cancer screening, a woman is successfully treated for breast cancer after having a mammogram–therefore, the success of the treatment must be due to the mammogram.
It seems hard to argue otherwise, and however careful we try to be, it’s difficult to really know whether, in any case, a mammogram saved somebody’s life (or extended it, or “cured” the cancer, whatever that might mean).
The trick, in sorting this out, is to ask the right question. Not: Do mammograms save lives? But this question, posed by Tara Parker-Pope in a smart post on her Well blog for The New York Times: How is it possible that finding cancer early isn’t always better?
That gets us to the heart of the issue. Here’s the misconception that we find so beguiling: Cancer starts small, and kills when it grows and spreads. Find it when it’s small, treat it, and prevent the fatal consequences. How, as Parker-Pope asks, can that be wrong?
Parker notes that there are four kinds of breast cancer found by mammograms:
First, there are slow-growing cancers that would be found and successfully treated with or without screening. Then there are aggressive cancers, so-called bad cancers, that are deadly whether they are found early by screening, or late because of a lump or other symptoms. Women with cancers in either of these groups are not helped by screening.
Then, she notes, there are cancers that will never amount to anything, but are treated when they are found on a mammogram. These women are not helped by mammograms, and they are hurt by the unnecessary treatment they undergo as the result of screening. So far, the value of mammograms for these three kinds of cancer is zero, zero, and negative–it either doesn’t help, or it hurts.
Finally, Parker-Pope comes to the fourth set of circumstances, in which a cancer is found by a mammogram at a time when treatment can change the course of the cancer. These are the people we are thinking of when it seems so clear that finding cancer early must be better.
But how many women who get mammograms fall into this category? Half? A quarter? One in ten? Parker-Pope:
Clinical trial data suggests that 1 woman per 1,000 healthy women screened over 10 years falls into this category, although experts say that number is probably even smaller today because of advances in breast cancer treatments.
So, to return to Parker-Pope’s question, How is it possible that finding cancer early isn’t always better? It is better for 1 in 1,000 women. For the others, it’s worthless or harmful.
I suspect this is not the last word on the question, but it’s as clear an analysis of the subject as I can recall. Parker-Pope’s post is based on a new study by two Dartmouth researchers, Dr. H. Gilbert Welch and Brittney A. Frankel. Welch made a similar argument in the Times himself in an opinion piece on Oct. 10. “Even with screening,” he wrote, “most people destined to develop deadly, untreatable cancers will still do so.”
He compares breast-cancer screening to screening for prostate cancer, and finds that similar questions arise with both. Screening is a gamble. That’s not what we want, and it’s hard to accept. But until researchers can do a better job of screening for the cancers that can be treated–it’s a fact.
– Paul Raeburn
(Thanks to Stephen Hart for the heads-up.)
Leave a Reply