BI-RADS Classification for Management of

Abnormal Mammograms
Margaret M. Eberl, MD, MPH, Chester H. Fox, MD, Stephen B. Edge, MD,
Cathleen A. Carter, PhD, and Martin C. Mahoney, MD, PhD, FAAFP

The Breast Imaging Reporting and Data System (BI-RADS), developed by the American College of Radi-
ology, provides a standardized classification for mammographic studies. This system demonstrates good
correlation with the likelihood of breast malignancy. The BI-RADS system can inform family physicians
about key findings, identify appropriate follow-up and management and encourage the provision of ed-
ucational and emotional support to patients. (J Am Board Fam Med 2006;19:161– 4.)

The Breast Imaging Reporting and Data System ommendation, this system can both inform family
(BI-RADS) was developed in 1993 by the American physicians about findings and direct appropriate
College of Radiology (ACR) to standardize mam- follow-up and management.4
mographic reporting, to improve communication, The BI-RADS lexicon offers a number of
to reduce confusion regarding mammographic strengths, including the application of a standard-
findings, to aid research, and to facilitate outcomes ized common language to facilitate communication
monitoring.1 According to the Mammography between radiologists, referring physicians, and pa-
Quality Standards Act (MQSA) of 1997 [Final Rule tients. The system also clarifies the reporting of
62(208):55988], all mammograms in the United mammography results and will support the com-
States must be reported using one of these assess- pletion of quality improvement activities and clin-
ment categories.1,2 Each mammographic study ical research.
should be assigned a single assessment based on the The vast majority of screening mammograms
most concerning findings.1 are classified as BI-RADS 1 and 2. Between 5% and
9% of screening mammograms will require addi-
tional follow-up or biopsy including up to 7% of
The BI-RADS System
Table 1 presents BI-RADS classifications and man- mammograms classified as BI-RADS category 3 as
agement recommendations as an evidence table. well as the 2% of BI-RADS 4 or 5 mammo-
Classifications are divided into an incomplete as- grams.5–7 The positive predictive value of a biopsy
sessment (category 0) and completed assessments positive for malignancy increases from ⬍2% for
(categories 1, 2, 3, 4, 5, 6).1,3 Although there are 7 BI-RADS category 3 mammograms to 23% to 30%
assessment categories, only 4 outcomes are possi- for category 4 mammograms and to 95% for cate-
ble: (1) additional imaging studies, (2) routine in- gory 5 mammograms.8,9 Specific mammographic
terval mammography, (3) short-term follow-up, features with the highest positive predictive value
and (4) biopsy.2 All categories reflect the radiolo- of malignancy include masses with spiculated mar-
gist’s level of suspicion for malignancy, and these gins and/or irregular shape, as well as calcifications
assessment categories have been shown to be cor- with linear morphology and/or segmental distribu-
related with the likelihood of malignancy.2 Because tion.10
each BI-RADS category has only one specific rec- Table 2 summarizes findings from a population-
based mammography registry in New Hampshire
showing the proportion of breast cancers observed
Submitted 21 July 2005; accepted 29 July 2005. by BI-RADS category. The rate of breast cancer
From Roswell Park Cancer Institute (MME, SBE, CAC, among women with BI-RADS category 1, 2, or 3
MCM), Buffalo, NY; and Department of Family Medicine
(MME, CHF), State University of New York, Buffalo, NY. mammograms after 1 year of follow-up was approx-
Conflict of interest: none declared. imately 1/1000 compared with 136/1000 among
Corresponding author: Margaret M. Eberl, MD, MPH,
Roswell Park Cancer Institute, Elm and Carlton Streets, those with category 4 and 605/1000 with category
Buffalo, NY 14263 (Margaret.Eberl@RoswellPark.org). 5 mammograms.5

http://www.jabfm.org Managing Mammograms Using BI-RADS 161

Table 1. Evidence Table for Clinical Management Recommendations for Mammograms by Breast Imaging
Reporting and Data System (BI-RADS) Category
BI-RADS Clinical Management Strength of Comments on
Category Assessment Recommendation(s) Recommendation References References

0 Assessment incomplete Need to review prior A 3 All or none study; consensus

studies and/or complete guidelines
additional imaging
1 Negative Continue routine screening A 3, 8 Consensus guidelines;
validated clinical decision
tool
2 Benign finding Continue routine screening A 3, 8 Consensus guidelines;
validated clinical decision
tool
3 Probably benign finding Short-term follow-up B 3, 6, 8, Consensus guidelines; cohort
mammogram at 6 10–15 studies; large case series;
months, then every 6 to validated decision tool; less
12 months for 1 to 2 patient stress; lowered
years costs with surveillance
4 Suspicious abnormality Perform biopsy, preferably A 3, 8–10 All or none study; validated
needle biopsy clinical decision tool
5 Highly suspicious of Biopsy and treatment, as A 3, 8–10 All or none study; validated
malignancy; appropriate necessary. clinical decision tool
action should be taken.
6 Known biopsy-proven Assure that treatment is
malignancy, treatment completed
pending

Conclusion as well as women with suspicious mammograms

Given that BI-RADS can impact on patient care by (BI-RADS codes 4 and 5) undergo appropriate fol-
minimizing both over-utilization and under-utili- low-up. This might include the creation of an office
zation of follow-up tests/procedures, it is critical registry to assure optimal management, as well as
that family physicians, and other clinicians provid- necessary educational and emotional support.
ing care to women, be familiar with the interpre-
tation of and management strategy for each cate- Clinical Vignettes to Illustrate Clinical
gory. Management Using BI-RADS Codes
Primary care physicians would benefit from de- Case 1
veloping mechanisms, in partnership with their col- A 48-year-old female had a screening mammogram
laborating radiologists, to assure that all women showing rounded densities with possible irregular
needing further imaging (BI-RADS codes 0 and 3) borders amid dense breast tissue bilaterally.

Table 2. Mammography Assessment and Breast Cancers Detected in the New Hampshire Mammography Registry,
11/96 to 10/97, by Breast Imaging Reporting and Data System (BI-RADS) category
Baseline Mammography*
BI-RADS Breast Cancers Detected Anticipated Rate of
Category Number Percentage Rate/1000 Malignancy Following Biopsy

1 37,995 80.65% 0.7 —†

2 4,930 10.46% 1.2 —†
3 3,345 7.10% 8.1 ⬍2%6–8
4 766 1.63% 135.8 23%–34%12–15
5 76 0.16% 605.3 ⱖ95%1,4,12–15
Total 47,112 100.00%

* Data derived from Poplack et. al.5

† Categories 1 and 2 are considered to be negative test results.

162 JABFM March–April 2006 Vol. 19 No. 2 http://www.jabfm.org

Radiologic Interpretation: BI-RADS 0 (additional imaging sociated calcifications lateral to the left nipple area.
needed) This lesion was not present on prior mammograms.
One week later, she had spot compression views
that showed the nodules to be regular and sharply
defined. Ultrasound examination revealed cysts. Fi-
Radiologic Interpretation: BI-RADS 5 (highly suspicious of
nal classification as BI-RADS 2 (benign finding).
malignancy)
Results were reviewed with the patient, and needle
Patient should continue with routine breast cancer
biopsy was recommended. Vacuum-assisted needle
screening.
biopsy (Mammotome) was performed using mam-
mographic stereotactic localization. Pathology
Case 2 showed infiltrating ductal carcinoma, grade II.
A 57-year-old female completed a screening mam-
mogram showing calcifications in the right breast.
These lesions were confined to the upper outer Patient Education
quadrant but were scattered and round on magni-
fication views. The only prior mammogram was
● What is BI-RADS? BI-RADS is a system that
from 4 years ago, was of poor quality, and only
was developed by radiologists for reporting
showed a few scattered calcifications.
mammogram results using a common language.
The radiologist assigns a single digit BI-RADS
Radiologic Interpretation: BI-RADS 3 (probably benign) score (ranging from 0 to 5) when the report of
Despite the lack of a recent prior comparison mam- your mammogram is created.
mogram, the current calcifications were felt to be of ● What does BI-RADS 0 mean? BI-RADS 0
low suspicion. During a discussion the patient was identifies a mammogram study that is not yet
informed that the calcifications were felt to be of complete. You need to make sure that further
low suspicion. A repeat mammography was recom- evaluation is completed, perhaps extra mammog-
mended in 6 months. Follow-up mammogram at 6 raphy views or an ultrasound. Further informa-
months and subsequently at 1 year showed no tion is needed to make a final assessment (codes 1
change in these calcifications. to 5).
● What does BI-RADS 1 mean? BI-RADS 1
Case 3 means that the mammogram was negative (ie, no
A 53-year-old female had a screening mammogram cancer) and that you should continue your rou-
that showed linear calcifications clustered tightly in tine screening.
● What does BI-RADS 2 mean? BI-RADS 2 also
the upper outer quadrant of the right breast. Mag-
nification views confirmed these were clustered and means that your mammogram was normal (ie, no
that there was no associated mass. The calcifica- cancer), but other findings (eg, cysts) are de-
tions were not present on a mammogram obtained scribed in the report. You should continue your
12 months earlier. routine screening.
● What does BI-RADS 3 mean? BI-RADS 3
means that your mammogram is probably normal
Radiologic Interpretation: BI-RADS 4 (suspicious but a repeat mammogram should be completed
abnormality) in 6 months. The chance of breast cancer is
Results were reviewed with the patient and biopsy
approximately 2% in this category. You should
was recommended. Vacuum-assisted needle biopsy
make sure that these follow-up mammograms are
(Mammotome) was performed using mammo-
completed as requested.
graphic stereotactic localization. Pathology showed ● What does BI-RADS 4 mean? BI-RADS 4
atypical ductal hyperplasia. Subsequent excisional
means that the findings on your mammogram are
biopsy confirmed the absence of malignancy.
suspicious and that there is approximately a 23%
to 34% chance that this is breast cancer. You will
Case 4 need a biopsy to get a small tissue sample to make
A 62-year-old female completed a screening mam- a diagnosis. Talk to your doctors about any ques-
mogram showing a 1-cm spiculated mass with as- tions.

http://www.jabfm.org Managing Mammograms Using BI-RADS 163

● What is a biopsy? Biopsy is done to obtain a mography. Curr Opin Obstet Gynecol 2001;13:15–
piece of the breast tissue to determine whether 23.
there is cancer. The biopsy may be done using a 5. Poplack SP, Tosteson AN, Grove MR, Wells WA,
Carney PA. Mammography in 53,803 women from
needle technique (a “needle biopsy”) or may re-
the New Hampshire mammography network. Radi-
quire a surgical operation (a “surgical biopsy”). ology 2000;217:832– 40.
When a needle biopsy is an option, it is usually 6. Varas X, Leborgne F, Leborgne JH. Nonpalpable,
preferred. probably benign lesions: role of follow-up mammog-
● What does BI-RADS 5 mean? BI-RADS 5 raphy. Radiology 1992;184:409 –14.
means that your mammogram results are highly 7. Monticciolo DL, Caplan LS. The American College
suspicious with a 95% chance of breast cancer. of Radiology’s BI-RADS 3 classification in a nation-
You will need to have a biopsy for diagnosis. Talk wide screening program: current assessment and
comparison with earlier use. Breast J 2004;10:106 –
to your doctors about what course of action to
10.
take.
8. Lacquement MA, Mitchell D, Hollingsworth AB.
● What does BI-RADS 6 mean? BI-RADS 6 means Positive predictive value of the Breast Imaging Re-
that you have already been diagnosed with breast porting and Data System. J Am Coll Surg 1999;189:
cancer. Discuss your treatment plan with your 34 – 40.
doctors. 9. Orel SG, Kay N, Reynolds C, Sullivan DC. BI-
● Why do I need to know my BI-RADS score? RADS categorization as a predictor of malignancy.
Knowing your BI-RADS number can help to Radiology 1999;211:845–50.
make sure that you get proper follow-up after 10. Liberman L, Abramson AF, Squires FB, Glassman
JR, Morris EA, Dershaw DD. The breast imaging
your mammogram. It is a good thing for you to
reporting and data system: positive predictive value
know that score so you can keep track, along with of mammographic features and final assessment cat-
your physician, of what you need to do and ac- egories. AJR Am J Roentgenol 1998;171:35– 40.
tively participate in your medical care. 11. Sickles EA. Management of probably benign breast
lesions. Radiol Clin North Am 1995;33:1123–30.
References 12. Sickles EA, Parker SH. Appropriate role of core
1. American College of Radiology. The ACR breast breast biopsy in the management of probably benign
imaging reporting and data system (BI-RADS) lesions. Radiology 1993;188:315.
[web source]. November 11, 2003. Available from: 13. Lindfors KK, O’Connor J, Acredolo CR, Liston SE.
http://www.acr.org/departments/stand_accred/birads/ Short-interval follow-up mammography versus im-
contents.html. Accessed February, 27, 2004. mediate core biopsy of benign breast lesions: assess-
2. Liberman L, Menell JH. Breast imaging reporting ment of patient stress. AJR Am J Roentgenol 1998;
and data system (BI-RADS). Radiol Clin North Am 171:55– 8.
May 2002;40:409 –30. 14. Brenner RJ, Sickles EA. Surveillance mammography
3. Singletary E, Anderson B, Bevers T, Borgen P, Buys and stereotactic core breast biopsy for probably be-
S, Daly M. National comprehensive cancer network: nign lesions: a cost comparison analysis. Acad Radiol
clinical practice guidelines in oncology, breast cancer 1997;4:419 –25.
screening and diagnosis guidelines, v. 1.2003. 08/20/ 15. Sickles EA. Probably benign breast lesions: when
02. Available from: http://www.nccn.org/physician_ should follow-up be recommended and what is the
gls/f_guidelines.html. Accessed March 1, 2004. optimal follow-up protocol? Radiology 1999;213:
4. Bomalaski JJ, Tabano M, Hooper L, Fiorica J. Mam- 11– 4.

164 JABFM March–April 2006 Vol. 19 No. 2 http://www.jabfm.org