ASPO Abstracts
Accuracy of Women's Self-reported Breast Density
Category: Early Detection & Risk Prediction
Conference Year: 2020
Abstract Body:
Purpose: To assess the accuracy of women's self-reported breast density against Breast Imaging Reporting
and Data System (BI-RADS) breast density assessed on most recent mammogram.
Methods: Women participating in an online or paper survey conducted by the Breast Cancer Surveillance
Consortium (BCSC). Eligible women were ages 40-74 years, had no history of breast cancer, received a digital
mammogram or digital breast tomosynthesis within 12 months of survey response, or breast MRI within 24
months before response. Women who reported they were told by a health care provider that they have dense
or non-dense breasts were asked to identify their BI-RADS breast density category. We compared women's self-reported binary and four-category density to their density from most recent screening exam, and assessed
the accuracy of self-reported density by percentage correct per density category.
Results: Among 2,177 survey participants, 29% were ages 40-49, 43% 50-64, and 29% 65-74 years; 73%
were white; 82% reported undergoing annual breast cancer screening; and 67% had a 4-year college or higher
degree. The distribution of BCSC 5-year breast cancer risk was 19% low, 34% average, 26% intermediate, and
22% high. 29% (639/2177) of women reported never being told or not knowing whether they have dense or
non-dense breasts, and 1% (18/2177) did not respond. Among the 70% (1,520/2,177) of women who reported
ever being told they have dense or non-dense breasts, 93% (128/137) accurately reported having non-dense
breasts and 73% (1016/1383) accurately reported having dense breasts. When asked which of the four BI-
RADS categories their breast density was, 56% (844/1520) reported not knowing their categorical breast
density, and 8% (128/1520) did not respond. Among the 548 who gave a density category, accuracy by category
was 74% (37/50) fatty, 60% (78/130) scattered fibroglandular density, 46% (122/264) heterogeneously dense,
and 82% (85/104) extremely dense.
Conclusions: Among women who reported ever being told if they have dense or non-dense breasts, the
majority were accurate in their knowledge of binary density, but did not know their BI-RADS breast density.
Additionally, accuracy was low for some BI-RADS density categories.
Keywords: Breast Density, Screening, Cancer, BI-RADS