Coding mammograms using the classification "probably benign finding--short interval follow-up suggested".

Abstract

Many benign breast lesions revealed by mammography show features indicating that the lesions have a high, but not complete, likelihood of being benign. The Breast Imaging Reporting and Data System (BI-RADS) allows radiologists to classify these mammograms as "probably benign finding-short interval follow-up suggested" (category 3). We explored whether certain factors are associated with the use of category 3 in a national cancer detection program. We analyzed data from the National Breast and Cervical Cancer Early Detection Program, a comprehensive nationwide program that provides cancer screening for low-income and medically underserved women. The study population included all women at least 40 years old who had undergone mammography on or before September 30, 1996 (n = 372,760). Of the 372,760 mammograms, 7.7% were classified as category 3. The probability of receiving a category 3 classification decreased as patients' ages increased. Women who were symptomatic were nearly twice as likely as women who were asymptomatic to receive a category 3 classification, and women whose clinical breast examinations had abnormal findings were more than twice as likely as women with examinations having normal findings to receive a category 3 classification. The percentage of mammograms classified as category 3 by state or tribal organization ranged from 1.4% to 14.0%. Several patient variables, including patient symptomatology, were associated with the probability of having a mammogram classified as category 3. One of the most important determinants was where the patient underwent mammography, which suggests that variability exists among radiologists themselves in using this BI-RADS code for "probably benign" mammographic lesions.