Diagnostic error in breast disease.

Fifty-six breast biopsies, incorrectly assessed by preoperative clinical or mammographic examination, were reviewed to define the characteristics in the tumor or patient that caused the clinician and mammographer to be diagnostically inaccurate. The most important patient characteristic associated with error was the use of hormones. Failure to recognize that oral contraceptive use significantly reduces the incidience of benign breast disease contributed to the frequent misdiagnosis of lesions in those patients. Twelve of 16 masses in oral contraceptive users were malignant. In seven, their resemblance to cysts or fibroadenomas resulted in treatment delay of two weeks or 18 months. Because benign disease is uncommon in women who have used contraceptives two or more years, all new lesions in those women should be studied by biopsy promptly. Neither clinical nor mammographic evaluation of lesions in postmenopausal women who used estrogens was accurate. Twelve postmenopausal patients with carcinoma had used estrogens. Three of these lesions were considered benign clinically and four, by mammogram. In one, treatment was delayed four years. In women over 50 not using hormones, clinical diagnosis of malignancy was accurate. Ten carcinomas in those women were missed by mammogram. Eight had negative nodes; thus a negative mamogram when the clinical diagnosis is correct may be an effective guide in predicting the status of axillary nodes. Paget's disease was not recognized clinically in two of eight patients with that disease, and an additional two were not recognized on mammography. The initial examiner did not identify three of six inflammatory carcinomas. Ten percent of benign lesions were intraductal hyperplasia or papillomatosis with atypia and were the benign lesions most often misdiagnosed clinically and by mammogram. No microscopic lesions were noted on mammography without an associated palpable mass. Twenty-five per cent of the lesions in women aged 40-49 were incorrectly assessed by mammography or clinical examination. Four (15%) of the 27 carcinomas in this age group were not recognized by either modality. Mammography helped delineate the characteristics of masses in premenopausal women. With recognition that any mass that appears in a woman using oral contraceptives must be studied by biopsy, combined clinical and mammographic study in primenopausal women should minimize diagnostic error. Unfortunately, neither clinical nor mammographic evaluation of the women with irregular periods approaching menopause or within three years past menopause is accurate. It is in that age group that new diagnostic modalities are needed and when reported their efficacy in that age group should be stressed.