When breast cancer is detected by physical examination or imaging and the p
atient has previously undergone a mammographic study that did not identify
malignancy, the study may be considered under the category "false-negative
mammogram". Some lesions cannot be seen in retrospect, some should be consi
dered as subthreshold, and some are appropriately classified as missed. Alt
hough little can be done to reduce delay in diagnosis secondary to the firs
t category, errors in detection or diagnosis contribute to the latter two c
ategories that may lend themselves to analysis and efforts at quality impro
vement. Understanding conditions that influence detection and diagnostic er
rors may reduce the number of false-negative mammograms.