Effect of breast-tissue characteristics on the outcome of clinical breast examination training.

Abstract

To compare the effects of clinical breast examination (CBE) training on lump-detection rates in simulated premenopausal and postmenopausal breast tissue. Two sets of six silicone models were made with background breast tissue simulating premenopausal tissue (most nodular, least soft) and postmenopausal breast tissue (least nodular, most soft) respectively. Eighteen lumps were located in each set of models. In September 1994, 82 housestaff and attending physicians with outpatient practices in the Division of General Internal Medicine at Northwestern University Medical School were randomized to a CBE-teaching intervention or a control group. Lump-detection rates for the two sets of models were measured before and after the teaching intervention. Analysis of covariance was used to analyze the effect of CBE training on examination sensitivity and specificity, controlling for baseline rates. CBE training increased lump-detection rates similarly and significantly in models simulating premenopausal and postmenopausal tissue, respectively. Specificity declined after training in models simulating postmenopausal tissue (p = 0.02) but was unchanged in models simulating premenopausal tissue (p = 0.54). CBE training had greater influence on sensitivity among house-staff than among attending physicians (p = 0.02). CBE training similarly affects lump detection in simulated premenopausal and postmenopausal breast tissue, but adversely affects specificity in simulated postmenopausal tissue only.