Fewer Recalls With Consensus Double Read Mammograms

ATLANTA—Compared with independent double reading of mammograms, consensus double reading detects slightly more cancers while significantly decreasing recall rates, thereby minimizing the anxiety that women might experience from undergoing a second mammogram, Susan Harvey, MD, said at the 102nd Annual Meeting of the American Roentgen Ray Society (abstract 63).

Dr. Harvey is assistant professor of radiology, University of Vermont College of Medicine, and lead author of the study.

With independent double reading of mammograms, two radiologists read each mammogram separately, and if either radiologist wants a recall of a woman for additional imaging studies, the recall takes place.

"Therefore, by the nature of it, it has to increase the recall rate," Dr. Harvey told ONI in an interview.

With consensus double reading, two radiologists read the mammogram and they both must agree on the findings and whether a recall is needed. If they disagree, a third radiologist is asked to review the case.

Over a 10-month period, Dr. Harvey and her colleagues performed consensus double reading of 15,985 screening mammograms, which led to 391 biopsies (2.4%). A total of 103 cancers were detected, and the second reader was responsible for detection of 10 of the cancers, a 9.7% increase in cancer detection. The recall rate was 13.2%.

In comparison, in their previous study of independent double reading, 25,369 mammograms read over 18 months led to 676 biopsies (2.7%). A total of 143 cancers were detected, and the second reader detected 11 of these cancers, for a 7.7% increase in detection. The recall rate was 14.2%.

Increased Cancer Detection Rate

The researchers conclude that double consensus reading increased the cancer detection rate by 9.7%, similar to the 7.7% increase seen with independent double reading, and, compared with independent double reading, decreased the recall rate by 7%.

Dr. Harvey noted that consensus double reading takes more radiologists’ time at the time of the double read, "but if it decreases the recall rate, then the overall time the radiologists spent on that group of women is decreased," she said. "A limitation is that you have to have enough staffing to provide this service."