In the United States, a whopping 20 percent of breast cancers go undetected in the yearly mammograms that are recommended by the US government to women 40 and over. Cancer research proved years ago that mammogram detection rates are 10 percent higher when two, rather than just one, radiologists examine the x-rays, and since then it’s been standard practice in the UK and much of Europe to do so. In the US, however, standard practice still calls for only one radiologist, ostensibly because it is not convenient, i.e. there are not enough radiologists, for hospitals to require two. (The reality is that American insurance companies don’t cover a second radiologist, so it’s not financially convenient for hospitals to mandate it.) Over the last ten years, US policy has resulted in about 10 percent more missed tumors among American women than among British women—an “inconvenience” which the US health care system has apparently been willing to accept.
The good news for American women and lonely radiologists alike is that computer-aided detection (CAD), which was approved a decade ago and is currently used in about a third of US mammograms, has just been shown to be about as good at reading mammogram x-rays as a second radiologist. A study recently published in the UK found that one expert working together with a CAD spotted almost exactly the same number of breast cancers (198 out of 227) as two human experts (199 out of 227).
The study was funded by the British government along with Cancer Research UK, a charity organization.
CAD was originally developed to assist radiologists in spotting more cancers, but only a third of American mammography centers have adopted it due to concerns about the accuracy of the technology. Hopefully, the new research will encourage more US hospitals to purchase and use the software—particularly since Medicaid is willing to reimburse them an extra $15 for using it.