The percentage of BI-RADS 3 lesions varies in literature from 2,34% to 11,2%. Malignancy is detected in less than 2% of the cases, usually in the first follow-up, resulting in early cancer detection, with no impact on the prognosis. 

In regards to mammography, BI-RADS 3 category can be used just after diagnostic workup has been made, with a complete imaging evaluation (complementary views) and correlation with other imaging methods. The BI-RADS 3 category is not used in screening mammograms. Recall is used to exclude malignant features, increase the identification of typically benign lesions (reduce follow-up interval and patient anxiety) and identify rapidly growing lesions.

The follow up protocol of BI-RADS 3 lesions includes a 6-month examination followed by 12-month, 24-month and optionally 36-month diagnostic examinations. If the lesion increases in size or presents any suspicious feature during the follow-up, it will be categorized as BI-RADS 4 or 5 and, in this way, a biopsy will be necessary. On the other hand, in case of regression or benign features, a follow-up is no longer necessary. It is essential to always use the most suspicious finding whether it is in mammography or ultrasound. A histopathological study may also be indicated in some specific cases such as: a) patient anxiety; b) uncertain follow-up; c) additional lesion in patients with cancer diagnosis.

The main objectives of the BI-RADS 3 category is to reduce the number of false positive biopsies, perform early malignant diagnosis and increase cost-effectiveness in screening while maintaining a high detection rate for early-stage cancer.

It is important to highlight that there are specific lesions which integrates this category. However, BI-RADS 3 category usually has high interobserver variability and inappropiate use in undetermined lesions. Therefore, it is the BI-RADS category with the highest error rate. Lehman et al. demonstrate that only 20% of the lesions categorized as BI-RADS 3 met the necessary criteria. The most common causes of misclassification in B-IRADS 3 category includes: technical parameters (different techniques make comparison difficult), lack of complementary views and interobserver variability.