Mammography in hospital patients: use and misuse.

Of 1433 consecutive patients who underwent mammography, 90 were subsequently found to have histologically confirmed carcinoma of the breast (85 invasive carcinoma, three lobular carcinoma in situ, two intraduct carcinoma). Sixty (67%) of the mammograms of the patients with cancer were reported correctly as carcinoma. There were 27 (30%) false negative reports and three (3%) equivocal reports. In 14 (16%) patients with false negative reports the biopsy was delayed by between 2 and 24 months. The false reassurance of the mammogram report was responsible for the delay in diagnosis in nine and may have contributed in the remainder. The mammogram was most likely to be misleading when used as an investigation for a palpable lump in the breast. The mean age of patients with a false negative mammogram (44 years) was significantly less than that of patients with a true positive mammogram (57 years). In 23 of the 60 patients with positive mammograms management was changed as a result of the investigation. The mammogram was most helpful when no breast lump was palpable and there was either some other clinical evidence of local malignancy (ten patients, 11%) or metastatic disease (three patients, 3%), or the mammogram had been performed to screen a high-risk patient (three patients, 3%). The positive mammogram was also useful in two patients with residual or recurrent disease following local excision of a breast cancer. There were 15 patients with bilateral breast cancer (17%). Three (3%) were diagnosed within 2 months of the first tumour but mammography did not contribute to the diagnosis of any of these.