April 16, 1998
According to new research, nearly a third of all women screened for breast cancer are wrongly told they need further tests. Two researchers discuss the findings with Margaret Warner.
MARGARET WARNER: More than half of all American women over the age of 40 receive regular mammograms, a breast x-ray to detect the early signs of cancer. A new study in The New England Journal of Medicine today suggests that mammograms and clinical breast exams, while often saving lives, also produce a higher rate of false alarms than previously thought. The study analyzed ten years worth of breast exams performed on 2,400 women ages 40 through 69 at a Boston HMO. Here to tell us about it is Dr. Joann Elmore, the study's lead author. She's an internist at the University of Washington School of Medicine. Also joining us is Dr. Stephen Feig, a radiologist. He's director of breast imaging at Thomas Jefferson University Hospital in Philadelphia.
A RealAudio version of this segment is available.
April 6, 1998
A report on tamoxifen's impact on breast cancer rates.
March 12, 1998
Cancer incidents and deaths in America are down, a report finds.
January 12, 1998
A series of designer estrogens are being developed that could help menopausal women.
June 19, 1997
A debate over estrogen replacements.
January 24, 1997
Experts debate the need for yearly mammograms.
November 3, 1995
Is there a breast cancer gene?
Browse the NewsHour's coverage of health issues.
The New England Journal of Medicine's collection of breast cancer articles.
Information about breastcancerinfo.com, a Web site sponsored by the Susan G. Komen Breast Cancer Foundation.
The National Cancer Institute home page.
And, Dr. Elmore, how frequently do these false alarms occur?
DR. JOANN ELMORE, University of Washington School of Medicine: Well, in our study we found that over the ten-year period one out of three women had a false alarm. Now, these women were not having screening every year. On average, these women had four screening mammograms and five physical breast examinations. We used the data to come up with estimates so that we can tell our patients what is their risk after having 10 screens. And with this data we now can tell our patients that after having ten screening mammograms, they have a 50 percent chance of having a false alarm. We can also tell them that after having ten physical breast examinations, that they have a 25 percent chance of having a false alarm.
MARGARET WARNER: And when we talk about a false alarm or a false positive, we're talking about essentially a woman who does not have cancer, nonetheless, the result is either ambiguous enough or suspicious enough that she--it's suggested that she take additional tests of some nature, is that right?
DR. JOANN ELMORE: That's correct. That's correct.
Why are there so many false alarms?
MARGARET WARNER: Now, why do you think there are so many false positives or false alarms?
DR. JOANN ELMORE: Well, it's a fine balance. We don't want to miss any cancers, and I think that's very important.
MARGARET WARNER: Explain a little more.
DR. JOANN ELMORE: Well, I think that we are afraid of missing cancers, and if you lower the false positive rate, some individuals are afraid that they may miss cancers. I am concerned, however, that the false positive rate is so high. My hope is that now we have shown what the numbers actually are that we can look at perhaps why this is happening and try to investigate ways of lowering it.
MARGARET WARNER: Dr. Feig, do you consider that this study identified a significant problem?
Dr. Stephen Feig: "It really is analogous to say a smoke detector in your kitchen...You want it to go off occasionally when you're cooking something on the oven, and there's smoke in the air."
DR. STEPHEN FEIG, Thomas Jefferson University Hospital: We always want to do better. The initial purpose of the radiologist in interpreting a screening mammogram is to sort the women into two groups, one group that compromises about 95 percent of women having a single mammogram have completely normal studies. The other five percent having mammograms each year need to have additional mammographic views or perhaps ultrasound to evaluate a questionable area. In nearly all of these cases of the five percent it turns out to be normal. And biopsy is necessary in much less than 1 percent of patients. But getting these extra views does reduce the number of unnecessary biopsies. If we reduce them too far, we're going to miss cancers. It really is analogous to say a smoke detector in your kitchen. If you have a smoke detector that only goes off when half your house is burning down, it's not sensitive enough. You want it to go off occasionally when you're cooking something on the oven, and there's smoke in the air. So the reality of the situation is that it's a balancing act, and it's a trade-off. But if women are properly apprized that they may need to have extra views after their initial screening study and if they're told that in most cases this is not going to indicate any substantial abnormality, and then when they're told that the results may need additional views if they're approached properly, there really shouldn't be any anxiety. And that's what our patients tell us.
MARGARET WARNER: What, that they are being approached properly and, therefore, they--
DR. STEPHEN FEIG: They are, that they're being reassured of the situation. And then most of them take this quite well. It's an inconvenience to some patients, but it really should not provoke an undue amount of anxiety if the patients are properly informed about their results.
MARGARET WARNER: Dr. Elmore, from your experience, do you think women are being advised sort of not to panic and that--and told that there are this many false positives?
The emotional and monetary cost of false alarms.
DR. JOANN ELMORE: Well, I'm a primary care clinician and an internist, and over the years I've received a lot of frantic phone calls from patients. They are anxious; they want more information; and they need to be reassured that the great majority of these women with abnormalities noted on screening don't have cancer. Maybe because I'm a primary care clinician my patients call me instead of the radiologist, but I do think it's a big problem. And that's what stimulated me to do this study.
MARGARET WARNER: Now, when you said "big problem," what do you mean? You mentioned the costs involved. Describe that a little more.
DR. JOANN ELMORE: Well, by big problem I was mostly referring to the anxiety, but there is a cost associated with false positive results as well.
MARGARET WARNER: You mean a financial cost?
DR. JOANN ELMORE: Yes. In our study we found that for every $100 that was spent on the screening itself an extra $33 was spent to pay for all of the diagnostic tests, all of these ultrasounds and extra mammograms and even the biopsies that were done.
MARGARET WARNER: Dr. Feig, do you think that either the technique of screening, that is, I guess I'm talking about the technology and/or the interpretation, which is what your profession does, of those results could be improved?
DR. STEPHEN FEIG: It could always be improved, but actually we're doing quite well right now. I think that Dr. Elwell's [Elmore's] estimates of cost are too high, but even if--even if we accept those, mammography is still much more cost-effective or comparable cost-effectiveness to other commonly accepted procedures like screening for carcinoma of the cervix, screening for colorectal cancer. Women who go actually for screening for pap smears, about 10 percent of the examinations may require a repeat pap smear. And mammography is actually much more cost-effective than intervention such as automobile seat belts and air bags.
MARGARET WARNER: Dr. Elmore, would you agree with that--if you're weighing these various kinds of studies?
DR. JOANN ELMORE: Well, I actually think that false positives are more common that Dr. Feig is stating. Perhaps we should be sending our patients to him if his false positive rate is five percent. There was a study that showed, on average in the U.S., about 10 percent of women that go in to get a mammogram have some sort of false alarm.
MARGARET WARNER: And you're talking about in any given year, rather than 10 years' worth?
DR. JOANN ELMORE: That's correct.
Are clinical or physical breast exams more effective than mammograms?
MARGARET WARNER: All right. Let me ask you for women watching this, I notice that--and you said that the rate of false positives is much higher really for mammograms than for clinical breast exams or physical breast exams. Why is that?
DR. JOANN ELMORE: Why is it higher for mammograms?
MARGARET WARNER: Yes.
DR. JOANN ELMORE: I don't know. It's important to point out a few things, though. First of all, mammograms are actually picking up a lot more breast cancers than we're picking up with our physical examinations.
MARGARET WARNER: So you're saying, in a way, they're more effective?
DR. JOANN ELMORE: In a way they are, although the have a higher cost of this higher false positive rate.
MARGARET WARNER: But then could Dr. Feig be right that it--I mean, it's better to err on the side of being hypersensitive because a false positive is a lot better than a false negative.
DR. JOANN ELMORE: That's correct. But I think that now that we have pointed out how common these are, I really hope we can try to see if we can't reduce the false positive rate.
MARGARET WARNER: Dr. Feig, do you think there's a lesson in this about whether a woman should have a mammogram versus a clinical or physical exam, or should she have both?
DR. STEPHEN FEIG: She should have both, although there are more false positive on the basis of mammography. Mammography is going to pick up many more cancers, and, more importantly, it's going to pick up earlier cancers. I also actually think that there's a basic fallacy in the analysis, and that is that if you have a certain number of call backs say in patients who had four mammograms over 10 years, as was the case in the study. You can't simply multiply it by ten-fourths and project the number of extra views that would be taken in ten years because that's not as frequent. When patients come for their initial mammogram, about 10 percent of them may need extra views. On subsequent mammograms it should be about five percent. And that's what we're seeing around the country. But when women come sporadically, rather than every year for mammography, then the recall rate is going to be higher; it's going to tend towards 10 percent. And many of these women in the study that were referred to actually did come as sporadically for mammography rather than regularly. It certainly was the case here.MARGARET WARNER: And is that because--are you saying because if a woman goes every year, then the radiologist has a base line, can keep comparing it year after year, whereas, if it's sporadic, it's almost as if she's coming for the first time?
DR. STEPHEN FEIG: Yes, that's exactly the case. On the first mammogram perhaps a radiologist detects a mass, such as a cyst or Phibro anonyme. And you can't tell initially that that's not a cancer in many cases. But then when she comes for a second and a third mammogram and you see the mass is still there and it's unchanged, that indicates that it's benign.
Are younger women more likely to receive a false positive?
MARGARET WARNER: All right. One final question to you, Dr. Elmore, one final finding that I'd like to ask you about. You also found that the number of false positives or false alarms was considerably higher for younger women in their 40's than it was for women over 50. How do you explain that, and what does that tell you?
DR. JOANN ELMORE: We didn't set out to study any differences between younger women versus older women. All that we found is that younger women--in other words those women 40 to 49 years of age--do have a slightly higher chance of a false positive. But this has been shown by other investigators, and it wasn't a significant finding.
MARGARET WARNER: And, Dr. Feig, your view on that, why that is the case.
DR. STEPHEN FEIG: Well, you know, the false positive rate in the United States was compared in this paper to the comparable rate for Sweden, and I think it's very important to point out that although we have a higher false positive rate here than those that were mentioned in the Swedish studies, we're also detecting cancers here at a much earlier stage than were found in the Swedish studies conducted in the 1980s where these numbers were taken from. For example, in those studies, about 50 percent of the cancers were stage one, a relatively low stage, whereas, in North America, about 85 percent of the cancers are stage one. In those Swedish studies, 10 to 15 percent of the cancers were stage zero, that is, still confined to the ducts of the breast, ductile carcinoma in situ. Whereas, in the United States, it's about 30 percent of the cancers detected on screening that are carcinomas in situ that are a very curable stage of carcinoma.
MARGARET WARNER: All right. Well, thank you, Dr. Feig, and, Dr. Elmore, thanks to you both very much.
|Support the kind of journalism done by the NewsHour...Become a member of your local PBS station.|