February 22, 2013


Breast practices: The mammogram dilemma : Your annual screening may cause you more harm than good. (H. Gilbert Welch, February 21, 2013, LA Times)

For decades, researchers have documented the problem of false positive mammograms. These are the mammograms that are judged to be possibly indicative of cancer but are subsequently proved not to be. In the interim, many healthy women have the scare of their life.

There will necessarily always be some false positive mammograms. But their frequency in the U.S. is extreme: Somewhere between 25% and 45% of women will have one in a 10-year course of mammography.

More recently, researchers have focused on the harm of overdiagnosis: the detection of abnormalities that meet the pathologic definition of cancer but are not destined to cause problems. The problem here is that anything called "cancer" gets treated with surgery, radiation and/or chemotherapy.

There will also always be some overdiagnosis -- it's a side effect of trying to catch cancer early. But it appears that somewhere between a quarter and one-half of all cancers detected during routine screenings fall in this category, and that is totally unacceptable.

False positives and overdiagnosis have the same root cause. They are the product of the conventional paradigm of cancer screening: Look harder and harder to find smaller and smaller abnormalities.

Call it the "find more" approach. Digital mammograms find more cancer than plain films, so they must be an improvement. Because breast MRIs find more cancer than digital mammograms, they must be better yet. It's why newly touted 3D mammograms will undoubtedly be said to be better than anything else.

It's a cycle of increasing intervention, a cycle that aggravates both the false positive and overdiagnosis problem. And it's not clear it adds anything (but cost).

There is a fundamental asymmetry to screening: Only a very few can possibly benefit (those women who would die if their breast cancer wasn't detected and treated), but any participant can be harmed. It requires a more elegant approach, one that finds the cancers that matter while minimizing the collateral damage.

Posted by at February 22, 2013 5:18 AM

blog comments powered by Disqus