Like many of my friends, I’ve been struck by the sudden backlash against the new federal guidelines recommending that women begin getting regular mammograms at age 50 rather than age 40. Surely most of the people carping (I thought) lacked the knowledge base required to know at what point the benefit of detecting and treating real cancers early would outweigh all the costs associated with testing, including false positives and potentially unnecessary treatments. But as I headed home for Thanksgiving, it occurred to me that my father, a breast cancer specialist, does have the requisite knowledge base, and so I might has well get his take.
The short version is that he didn’t think much of the new recommendations either. He thought the conclusions were driven by bird’s-eye meta-analysis that glossed a lot of important details. For instance, he suggested that some of the data driving that conclusion came from Canadian studies showing few benefits because they disproportionately drew on cases from institutions whose antiquated equipment had a much higher error rate than that currently in widespread use in the United States. From an econometrician’s perspective, he told me, the studies are methodologically exemplary, but few clinicians would regard them as relevant to contemporary practice. He also thought the analysis was too quick to play down the value of follow-up biopsies that uncovered atypical but not-currently-malignant tissue.
Needless to say, that doesn’t mean “I have this friend who caught a tumor at 35…” is a good argument for setting a particular population-wide default, or that less testing won’t sometimes be the right answer. But it does sound like there’s some reason to be skeptical about this one.
Addendum: Since, as commenters note, my hazy recollection of our ten minute chat in the car is not exactly a fully rigorous argument, Dad passes along this more detailed statement from the Society of Breast Imaging.