A woman's chance of getting a false-positive result on a screening mammogram depends a lot on the radiologist interpreting the image, a new study shows.

"We found there is a lot of variability among community radiologists in their false- positive rates," said Joanne G. Elmore, MD, MPH, lead author of the report in the Sept. 18 Journal of the National Cancer Institute (JNCI, Vol. 94, No. 18: 1373-1380).

In this study, "false-positive" means any additional study or examination a radiologist requested after reviewing a screening mammogram, and the woman did not have breast cancer diagnosed within one year.

The study looked at false-positive rates among 2,169 women referred by their HMO to local radiologists for routine screening mammography.

Among the 24 radiologists interpreting the X-ray films, some produced false positives as seldom as about once for every 29 mammograms (3.5%) while others reported a false positive for every 13 mammograms (7.9%).

Radiologists more recently out of medical school, with less mammography-reading experience, had two-to four-times the false-positive rates of older, more experienced radiologists.

Although these doctors may also have found more cancers, this information wasn't available, the researchers said.

The study also showed that false-positive rates almost doubled over the six years from 1985 to 1987 and 1991 to 1993.

Doctors' increasing concern they might be sued if a breast cancer is missed could be one reason for some radiologists' ordering more X-rays and examinations, and resulting in high false-positive rates, said Elmore.

Ways need to be found to reduce false-positive rates, but it's not easy to do that without missing more cancers, Elmore said.

Women returning to the same mammography facility every year may reduce their chances of a false positive, since radiologists will have their earlier mammogram films for comparison, Elmore noted.

And premenopausal women may wish to have mammograms when they are not menstruating, since there is some evidence to indicate that the increase in breast density during menstruation slightly reduces accuracy.

With an overall 10% chance of a false positive mammogram in the US, women should not be too concerned if they get a call suggesting they come back in for further tests, said Elmore.

Americans need to be thinking as a national community about what radiologists should be doing to reduce the number of false positive mammograms, said M. Robyn Anderson, PhD, MPH, assistant member of the Fred Hutchinson Cancer Research Center in Seattle and one of three co-authors of an editorial in the same issue of the JNCI.

"Are we willing to accept having huge proportions of women getting false positives, in order to ensure that every cancer that ever could be seen on a mammogram is caught, or do we want mammography to be more selective even if it isn't able to find every cancer?" she asked.

False positives could be reduced by requiring two radiologists to interpret each mammogram, said the editorialists.

Better training and more research to improve mammography could help also, they noted.

Call-backs for a second look are very alarming to women, but most clarify that cancer is absent, not present, noted Anderson.

In recent years, insurers' lower payments for mammography and the increasing risks of lawsuits have closed many mammography centers, yet accuracy depends on experience, said LaMar McGinnis, MD, medical consultant to the American Cancer Society (ACS).

Double reading is worth exploring, but insurers may hesitate to pay for it, he said.

Greater use of digital mammography may help in the future, since this technology can be easily forwarded to high-volume mammogram reading centers, possibly lowering costs while increasing accuracy. Digital mammography also has been shown to have a lower false positive rate.

Robert Smith, PhD, director of cancer screening for the ACS, pointed out a false-positive screening, as described in this study, usually results only in additional X-ray views of the breast. "And, after that," said Smith, "the woman's exam is usually defined as normal."

Smith also said that other research has shown that "women are aware of the likelihood of false positive results on mammography." He noted that women were willing to accept false positive screening results in order to save lives.

"Women place a higher value on saving lives from breast cancer," said Smith. "But that should not detract from our attempts to insure that radiologists have had adequate training in mammography screening. They should be able to rapidly gain experience so that the avoidable rate of false positives is as low as possible."