The criticisms of the mammogram study are valid
Matthew Martin 2/26/2014 01:37:00 PM
Aaron Carroll reminds me of the people who were unconvinced by the latest mammogram study that showed, basically, that detecting breast cancer before a noticeable lump exists has no value-added. In post after post, Carroll responded to various criticisms of the mammogram study that argued, for example, that the technology in that study was two decades old, and thus the results are not generalizeable to today. As Carroll points out, that critique can always be applied to pretty much any large long-term study in medicine; the critique is technically valid, but not a good reason for ignoring the evidence.
It's very true that we are in the process of researching and improving treatment options for breast cancer, and thus what is true about early detection with mammograms today might not be true in the near future. There's no way to know. But there's a right way and a wrong way to handle that uncertainty.
The right way is to demand more mammogram studies, while recommending against universal mammogram screenings in the meantime. Mammogram screenings represent more than minimal risk to patients: even ignoring the substantial economic costs (resources that could have been used more productively), they lead to more biopsies and other proceedures in women who do not have cancer, which can cause serious complications in a small fraction of people[*] that should not be ignored in the cost-benefit calculus.
The wrong way to handle this uncertainty is to continue to advocate universal mammogram screening inspite of the fact that the evidence says they do more harm than good.
That's not to say that no one should get mammograms until further notice. People in particularly high-risk subgroups, such as women in families with high rates of breast cancer, should probably still get regular screenings. And generally the first thing you'll do after feeling a lump in the breast is get a mammogram, which may also be a good idea. But, a healthy 40 year old woman with no prior breast cancer risks should not be getting mammograms, until further notice.
[*]I use the gender neutral "people" instead of "women" because breast cancer also occurs in men. Sometimes, mammograms are also performed on men, though this is less common because it tends to be more painful and has less chance of detecting cancer than it does for women.