Breast cancer is the second leading cause of cancer death in the United States, and each year more than 205,000 new cases are diagnosed – many of them first detected through screening mammograms. Women often feel a sense of relief when their tests come back negative, but if the test comes back positive, does that mean they have breast cancer? Not necessarily. Past studies have found that as many as 1 of every 10 screening mammograms is mistakenly interpreted as positive when the woman does not really have breast cancer. For over a decade, researchers have been looking at factors that may influence these false positive results and their financial, medical and psychological cost.

Research published in the September 18, 2002 Journal of the National Cancer Institute found that younger and less experienced radiologists have 2 to 4 times more false-positive mammogram interpretations than older radiologists. Unlike most previous studies, researches evaluated the performance of radiologists in a real world clinical environment (versus an artificial testing situation) to determine if the patient, the radiologist, and/or the testing characteristics influenced the number of false positives.

About the Study

Researchers at the University of Washington School of Medicine evaluated 8,734 mammogram interpretations from 24 community radiologists. These mammograms were obtained over an eight-and-a-half year period and came from 2169 women (average age 48) enrolled in a large health maintenance organization (HMO) in New England. The study included women between the ages of 40 and 69 who had no history of breast cancer, prophylactic mastectomy, or breast implants during the study period.

The women’s medical records were used to determine age at time of mammogram, menopausal status, hormone replacement therapy (HRT) use, body mass index at time of mammogram, race, family history of breast cancer, and history of breast biopsy or other diagnostic procedures. These records also provided information on the results of screening mammograms and the radiologist’s recommendations for additional testing. The mammogram interpretation was classified as positive if the results were indeterminate or suspicious for cancer, or if there was a recommendation for non-routine follow-up. A true-positive was recorded if breast cancer was diagnosed within 1 year of the test. All others were false-positives. The researchers obtained the radiologists' sex, age, and number of years since medical school graduation from the Massachusetts State Medical Registry and HMO administrative files.

The Findings

Younger radiologists and those who graduated from medical school within the past 15 years were 2 to 4 times more likely to the have a false-positive interpretation of a mammogram than those who were older and had been out of medical school longer. This means that, according to this study, if a woman went to two randomly chosen radiologists, her odds of having a false-positive would be about 2 to 4 times greater for the younger, less experienced radiologist compared with the older more experienced one.

Researchers also found a 40% reduction in the false positive rate (5.4% versus 9.0%) when a previous mammogram was available for comparison. In addition, women who had mammograms within 18 months of the previous mammogram were less likely to have a false-positive result compared to those women who waited longer or did not have a previous screening. Women who were younger, premenopausal, taking HRT, had a positive family history of breast cancer, or had a previous breast biopsy were also more likely to have a false-positive screening test result.

Adjustments were made for the number of mammograms each radiologist read, as well as each woman’s overall “risk” for having a false positive including: younger age, HRT use, family history of breast cancer, and history of breast biopsy. After making these adjustments the rates of variability remained basically the same.

Though this research sheds light on factors that may influence false-positive mammograms, there are some limitations to the study. First, the radiologists were not interpreting the same mammogram films, so the researchers could not make any direct comparisons between the performances of individual radiologists. Second, because of the eligibility criteria, a number of mammograms read by these radiologists were not included in the study. Third, only a small number of women were actually diagnosed with breast cancer (45 total), so researchers were not able to analyze the predictive value of mammography in this group of women--the probability of actually having cancer if the test is positive. Finally, these results may not apply to the general population of US women, because the false positive rates for this group were lower than the national average and the majority of women surveyed were white.

How Does This Affect You?

This study confirms the high rate of false positives associated with mammography, especially in younger women. While it is natural to assume the worse when faced with the bad news of a positive result, statistically speaking a positive mammogram is more likely not to be cancer.

New technologies for reducing the number of false-positives in mammogram screening are on the horizon. For instance, there are already computers programs designed to help radiologists target certain abnormalities that might be shown on a mammogram. While these technologies are promising, we are still years away from their routine use in a clinical setting.

So, what can be done in the meantime? Women in their twenties and thirties may want to consider a clinical breast exam (which is performed by a physician) every three years. When to start having mammograms and how often to have them is still hotly debated, especially in light of the fact that younger women tend to have more false-positive results. At this time, the National Cancer Institute recommends mammograms every one to two years for women over age 40. NCI also recommends that women with a family history of breast cancer seek expert medical advice about starting mammograms before age 40 and how often to be screened. Other authorities recommend routine mammograms beginning at age 50.

One of the most important issues that this study raises is the appropriate place of second opinions for medical procedures. Although there is no guarantee that another radiologist will read the mammogram more accurately, having two radiologists agree is certainly more reassuring. The findings of this study do suggest that, at least in the case of mammography, a second opinion may be worthwhile before proceeding with further tests.