Can women do anything to improve their chances of receiving an accurate interpretation of a screening mammogram? A new study found that there are several important characteristics to look for when selecting a mammography facility. The finding is based on a survey of 44 U.S. mammography facilities published last month in the Journal of the National Cancer Institute.
To learn how women can act on the results of this study, Maryann Napoli, Center for Medical Consumers, interviewed one of its co-authors, Joann G. Elmore, MD, Professor of Medicine, Adjunct Professor of Epidemiology, University of Washington School of Medicine, Seattle.

MN: Doesn’t the physician usually choose the mammography facility?

JGE: Most often women see their primary doctors and that doctor will refer them to a mammography facility that is linked with the primary care office.

MN: So women don’t actually do the choosing?

JGE: Mammography screening is a unique medical test in that women can refer themselves. There are mobile vans going around the country, stopping at work sites. I don’t know the actual number of women who self refer, but I suspect many women are actively involved in choosing mammography facilities.

MN: You have co-authored dozens of mammography-related studies. I recall being shocked by your first one, published in The New England Journal of Medicine in 1994. You found that experienced board-certified radiologists interpreting the same images often come to different conclusions about the presence or absence of breast cancer.

JGE: For many years I’ve been trying to figure out why there is so much variability in the interpretation of mammograms. Then we went on to identify characteristics of the patients that put them at risk of having an inaccurate exam. Some characteristics you cannot change like having dense breasts, and some you can change. For example, menstruating women can make sure not to schedule a mammogram around the time of their periods when the breasts swell and become tender. A few studies show accuracy is lower at this time.

MN: And now you’ve taken on the characteristics of the mammography facilities in your latest study.

JGE: Yes, I have been working in this area for 15-20 years, but I had never really asked the question: Does it matter which facility women choose? We found that facilities that have breast imaging specialists on their staff had a higher degree of accuracy.

MN: Don’t all facilities have breast imaging specialists interpreting the mammograms?

JGE: No, in some facilities mammograms are interpreted by general radiologists who do not have specialty training in breast imaging. And these radiologists interpret the minimum number [of mammograms] required for accreditation, but they may also interpret chest x-rays, MRI scans, ultrasounds, etc. Mammography may be a small part of their clinical workload.

MN: So women should ask this question of the facilities on the phone, prior to making an appointment: Do you have a breast imaging specialist on the staff? I know this is an awful question but do you really think women will get an honest answer from facilities that have only general radiologists reading their mammograms?

JGE: Well yes, if women ask the question this way: Do you have a radiologist that spends more than 50% of the time working in breast imaging? That’s pretty clear-cut. That’s how we define a breast imaging specialist in our study.

MN: What else did you find that women should ask about the facility?

JGE: They should ask whether the facility conducts audit reviews two or more times a year. Our study found higher accuracy among the facilities that review their audit data with the radiologist on a regular basis. This finding makes sense to me because the radiologists are learning from the prior history and probably improving the quality of their interpretations.

MN: Audit reviews are a form of quality control. The facility should compare the mammogram interpretations with the results of the breast biopsies in order to regularly test the accuracy of the radiologists’ decisions.

JGE: If you read the fine print of the Mammography Quality Standards Act, you’ll see that the facilities don’t have to show the audit results to the individual radiologists.

MN: That’s disturbing. Why collect audit information if the facility doesn’t show it to the radiologist?

JEG: Yes, as a physician, I believe that we have the opportunity to improve when we get feedback.

MN: That changes the second question women should ask when choosing a facility. It should be: Does your facility do audit reviews at least twice a year, and is the radiologist routinely given the results of these reviews?

JEG: Yes.

MN: You did not find volume to be important to accuracy. One would think that the more images a radiologist reads, the more likely he or she will make accurate interpretations.

JGE: Many studies have evaluated whether high volume is associated with performance, and there’s an assumption that really high volume by the radiologist is good because it means that the radiologist has a lot of experience. But like all areas of life, it’s more difficult than that. It’s possible that the facility could have high volume, but the facility might not be collecting and reviewing the audits and therefore not following up on those many thousands of patients whose mammograms have been interpreted.

MN: What’s the definition of high-volume?

JGE: There have been a few studies, but many of them have contradictory results. In the U.S., radiologists must do a minimum 480 mammograms per year, as required by law.

MN: That doesn’t sound like much.

JGE: In Europe, the minimum in some countries is 5,000 cases a year. But the question is: Are some countries doing better because they have higher volume? The published studies have contradictory results. There’s an ongoing study funded by the American Cancer Society and the National Cancer Institute. It will be helpful to see if they can clarify this at all.
MN: What are you working on now? What are the other major unanswered questions about mammography that should get research attention?

JGE: I’m continuing my research efforts aimed at improving the quality of breast cancer screening and detection. I am investigating the efficacy of new technology, such as digital mammography and computer-aided detection, and how we can improve our communication related to mammography.

MN: Mammography screening has been aggressively promoted to women ever since the early 1970s. It’s interesting to see how long it takes to get some basic questions answered. You are to be congratulated for being in the forefront of mammography screening research and for continuing to ask the research questions that will lead to more improvements. See Mammography Leaflet from the Nordic Cochrane Centre.

Maryann Napoli Center for Medical Consumers ©
July 2008