RSNA: Rads read breast density differently, even in same practice
“We knew there would be some variations because we’re aware of our partners’ tendencies, but we didn’t realize that the differences would be so great,” study author Stephanie L. van Colen, DO, from Caritas Good Samaritan Medical Center in Brockton, Mass., told Health Imaging News. In fact, prior to the study, van Colen and her colleagues at Cambridge Health Alliance in Cambridge, Mass., hypothesized that radiologists’ density description would range only 1 to 2 categories for most patients, thinking that there would be agreement among the majority of the radiologists on most exams.
|Slideshow | Variation in Reported Breast Density Among Radiologists|
|Stephanie van Colen DO, Carol Hulka MD, Janet Baum MD |
All radiologists viewed the studies on a GE Healthcare mammography reading station flat-panel monitors in the same reading room to eliminate external variation. Using a pre-printed answer sheet, the radiologists selected one ACR-recommended tissue density option for each exam: dense, heterogeneously dense, scattered fibroglandular tissues and fatty.
The researchers reported “great variability” in breast density reporting among the radiologists. In approximately 26 percent of exams, there was total agreement on density. In approximately 48 percent of exams, the majority (7/8 and 6/8) agreed on density. In approximately 17 percent, five of eight radiologists agreed. The remaining 9 percent of mammograms had 50 percent or less concordance of density reporting.
In approximately 32 percent of the cases, there was some radiologist disagreement between the categories more than 50 percent density (dense and heterogeneously dense) and less than 50 percent density (fatty and scattered fibroglandular tissue). In some cases, more than one radiologist assigned a density greater than 50 percent on the same exam that others assigned a density less than 50 percent, van Colen pointed out.
“The problem is,” she explained, “when a patient’s breast is designated as more than 50 percent dense, they may need to undergo more screening tests, such as MRI or ultrasound, especially if they have an additional risk factor. Therefore, underestimating breast density could potentially limit a patient's ability to have additional diagnostic imaging and could affect screening recommendations.”
Based on their findings, the researchers recommended that standardization of breast density reporting is “necessary and may be achieved through additional training or computer breast density estimation software.”
van Colen hypothesized that these types of disparities are probably “quite common” in real-life clinical practice. She noted that readings might be more uniform via practice-wide training sessions for mammographers and suggested that if computer-aided detection software continues to improve, it could provide insight into breast density. She acknowledged that this technology is still being developed and tested in clinical trials. However, it may prove useful down the line.