Analysis of False-Negative Cancer Cases Identified with a Mammography Audit

The purpose of this study was to analyze the causes of false-negative breast imaging in symptomatic and asymptomatic patients, as defined by cancer diagnosed within 1 year of a nonsuspicious mammogram. A computerized audit of 27,305 mammograms performed between November 1992 and December 1993 identified 50 patients who developed malignancy within 1 year of a mammographic report indicating negative, benign, or probably benign results. The audit revealed 26,661 mammograms interpreted as negative, benign, or probably benign. Of these, 50 patients were diagnosed with carcinoma within 1 year. Thirteen of the cancers were not visible retrospectively. Five were seen only in retrospect. Thirty-two were seen prospectively, of which 5 were interpreted as benign and 27 were interpreted as probably benign. Seventeen of the 27 probably benign were recommended for 6-month follow-up. Ten of the 27 probably benign had immediate ultrasound-guided fine needle aspiration cytology yielding unsuspected malignancy. In most (32/50) false-negative cases the lesions were seen prospectively but were interpreted as benign or probably benign. Ultrasound-guided aspiration averted a delayed diagnosis of malignancy in 20% (10/50) of the false-negative imaging interpretations. Six-month follow-up studies were helpful for 12 of 17 cases, where the lesions progressed within 1 year.