A very large study found that using computers to help read mammograms -- called computer-aided detection (CAD) -- didn't improve detection of either early-stage or advanced-stage breast cancers and increased the risk of a false-positive reading.
A mammogram can identify an abnormality that looks like a cancer, but turns out to be normal. This "false alarm" is called a false positive.
The research was published in the July 2011 issue of the Journal of the National Cancer Institute.
With CAD, a computer program highlights areas on the mammogram image that MAY be abnormal. A radiologist has to decide if the areas really are abnormal. CAD isn't the same as having a second radiologist look at the mammogram. With CAD, only one radiologist looks at the mammogram. Other research has shown that having a second radiologist do an additional reading of a mammogram improves detection and reduces false positives.
This study was done through the Breast Cancer Surveillance Consortium, a group of 90 hospitals and mammography centers in seven U.S. states. The researchers reviewed more than 1.6 million mammogram readings done between 1998 and 2006. About 28% of the hospitals and mammography centers adopted CAD during the study, so the researchers compared the CAD-read mammograms to mammograms read without CAD and then compared the results to the health records of the women who had the mammograms.
CAD-read mammograms resulted in:
- no improvement in diagnosing small, early-stage invasive breast cancers
- no greater likelihood of diagnosing early-stage breast cancer with lymph node involvement (versus no lymph node involvement)
- no decrease in diagnoses of advanced-stage (versus early-stage) breast cancer
Women who had CAD-read mammograms were more likely to be called back for more tests of what turned out to be a false-positive reading. Besides the fear of a breast cancer diagnosis, a false positive usually means more tests (including biopsy) and follow-up doctor visits. The process can be very stressful and upsetting.
Some of the early studies of CAD-read mammograms have shown that using CAD was beneficial. That's why insurance companies typically will pay more for CAD mammogram readings, and most mammogram screening programs use CAD today.
Still, some recent studies, including this one, have found that CAD doesn't seem to improve screening outcomes. It may be that many doctors aren't using CAD the way it should be used. New technology offers promise, but there is always a learning curve. In this case, doctors may be relying too much on CAD results and not critically comparing the CAD reading to their readings. Or doctors may be intimidated by the CAD results and reluctant to question them.
Based on these results, it's reasonable to question whether CAD mammogram screening is increasing costs without adding benefits and possibly doing more harm than good because of the increased likelihood of false positives. But instead of abandoning CAD, experts should use the results to investigate how doctors use CAD and give more and better guidance about the best ways to evaluate suspicious areas highlighted by CAD.