Study Finds Detecting Breast Cancer Depends on Radiologist

There is a new study on the effectiveness of diagnostic mammograms in actually detecting cancerous masses, and the news is sadly grim. Previous studies have shown substantial differences of opinion in radiologists when viewing screening mammograms (up to 25% of cases), but no one knew if this inaccuracy extended to diagnostic mammograms, situations where women have lumps or other symptoms.

The results are not good - radiologists miss an average of 2 in every 10 cases, with some missing as much as 7 out of every 10. The numbers, for those inclined, from the Chicago Tribune article (registration required), are:

123 radiologists who interpreted nearly 36,000 diagnostic mammograms between 1996 and 2003 at 72 U.S. facilities. All the mammograms had been ordered for women who found a lump themselves or whose doctors discovered something of concern.

The researchers found that sensitivity—the ability to detect cancer when it's present—ranged from 27 percent to 100 percent, with a median of 79 percent. The false-positive rate—women who got a tentative diagnosis of cancer when they didn't have it—ranged from zero to 16 percent, with a median of 4.3 percent.

But even for radiologists at teaching hospitals, median sensitivity was only 88 percent, which means that on average 12 out of every 100 cases of breast cancer in symptomatic patients were not detected.

This is kind of scary. These are women who already have cause to consider breast cancer a risk, be it lump or pain or leaking fluids, and they are still receiving false negatives. And given that the rate of breast cancer is 10fold higher among women with physical symptoms who go in for diagnostic (as opposed to screening) mammograms, you have to wonder how many women who receive those false negatives actually go and get a second opinion? How many primary care doctors insist on a second opinion, period, regardless of the result (but especially if the test result appears to disagree with physical symptoms)?

The results do suggest one clear thing: it's better to have your images read by someone who specializes in diagnostic breast imaging. The radiologists with the highest and most consistent correct reads were those in academic medical centers, and then those who spend at least 20% of their time on breast imaging. Unfortunately, most women in America have their images read by general radiologists in their community; in this study, 1/4 of those radiologists see less than 1,000 mammograms a year.

I haven't been able to access the article itself to read (but Andrea provided the link to the abstract, thanks!), but one other question does come strongly to mind. In the cases where the researchers found radiologists missing as many as 7 out of 10 cases of breast cancer, did the researchers say anything to the radiologists? To the center the radiologists work at? To someone, anyone? Did they track down those women with missed cancer cases who haven't already found out they were mis-diagnosed?

I realize this is a common question when it comes to research ethics, and perhaps my point of view is just too badly skewed by watching my mother die from cancer, but I can't imagine how you could know that people were out there without receiving treatment, going to die because of a missed diagnosis that you know about. And how can you not say anything about the person(s) continuing to risk lives by not being good at their jobs?

-Kelly