False negative mammography: causes and consequences

Abstract

A retrospective review was undertaken of 174 women presenting consecutively to a symptomatic clinic and subsequently found to have breast cancer. Their mammograms, clinical and pathological findings were reviewed. Any delay occurring between presentation and treatment was recorded. Mammography showed a sensitivity of 83.3% overall (95% confidence interval 77.0–88.5). Negative mammogram reports were associated with age under 50 ( P = 0.01−0.005), premenopausal status ( P = 0.01−0.005), dense breast parenchyma ( P = 0.005−0.001), in situ disease ( P = 0.0078), small tumour size < 20 mm ( P = 0.0173) and negative cytology ( P = 0.0173). A negative mammogram resulted in a delay to definitive treatment. Median time to treatment for negative versus positive mammography was 7 weeks (range 1–90) versus 3 weeks (range 0–18) P = < 0.001, mean times being 13.4 weeks and 3.5 weeks. There was clinical progression in over half of the women whose treatment was delayed ( P = 0.006). When clinical or cytological suspicion remains following triple assessment, negative mammography should not delay definitive histological diagnosis.