A critical factor determining mammographic sensitivity is the radiologist's interpretation. Studies have shown substantial variability in interpretation and reading accuracy among radiologists.[14,15,16,17,18,19,20,21,22,23] Some evidence suggests that using physician interpretation of actual mammograms influences sensitivity, specificity, or both, and a learning curve has been noted during the first few months of experience interpreting mammography examinations.[17,18,24,25] Whether this results from different overall accuracy or a shift in the trade-off between sensitivity and specificity, however, is not certain. The clinical significance of variability in radiologists' interpretations is not clear. Identifying a radiologist who is more accurate than another is difficult.
High breast density is associated with low sensitivity. At all ages, regardless of hormone therapy (HT), high breast density is associated with 10% to 29% lower sensitivity. HT, which increases breast density, is associated with both lower sensitivity and an increased rate of interval cancers. High breast density is an inherent trait, which can be familial [28,29] but also may be affected by age, endogenous  and exogenous [31,32] hormones, selective estrogen receptor modulators such as tamoxifen, and diet. Strategies have been proposed to improve mammographic sensitivity by altering diet, by timing mammograms with menstrual cycles, by interrupting HT use before the examination, or by using digital mammography machines.
The specificity of mammography is the likelihood of the test being normal when cancer is absent, whereas the false-positive rate is the likelihood of the test being abnormal when cancer is absent. If specificity is low, many false-positive examinations result in unnecessary follow-up examinations and procedures. (Refer to the Harms of Screening section of this summary for more information.) An improvement in reporting mammography results has been the adoption of Breast Imaging Reporting and Data System (BI-RADS) categories, which standardize the terminology used in assessing the significance of the findings and recommending future action. A study correlating needle localization biopsies with BI-RADS categories showed that categories 0 and 2 yielded benign tissue in 87% and 100%, respectively, of 65 cases. Category 3 (probably benign) yielded benign tissue in 98% of 141 cases, category 4 (suspicious) yielded benign tissue in 70% of 936 cases, and category 5 (highly suspicious) yielded benign tissue in only 3% of 170 cases. Studies have shown relatively little impact of false-positive test results on the use of subsequent mammography screening behavior, but false-positive test results may have long-term consequences, such as anxiety about breast cancer.
International comparisons of screening mammography have found that specificity is greater in countries with more highly centralized screening systems and national quality assurance programs.[39,40] For example, one study reported that the recall rate is twice as high in the United States as it is in the United Kingdom, with no difference in the rate of cancers detected. Such comparisons may be confounded, however, by other social, cultural, or economic factors that can influence the performance of mammography screening. No improvement in cancer detection was noted in these studies despite the higher recall rate.