[Skip to Content]
[Skip to Content Landing]
April 19, 1985

Physicians' Abilities to Detect Lumps in Silicone Breast Models

Author Affiliations

From the Departments of Medicine (Dr Fletcher and Mr O'Malley) and Epidemiology (Dr Fletcher), University of North Carolina at Chapel Hill (Ms Bunce).

JAMA. 1985;253(15):2224-2228. doi:10.1001/jama.1985.03350390066027

Little is known about how well physicians detect breast lumps in clinical breast examinations. We studied 80 general medicine, family medicine, general surgery, and obstetrics/gynecology physicians to determine their abilities to detect lumps in manufactured breast models. The mean number of lumps detected was 8.0 (44%), with a range of three (17%) to 15 (83%). Detection varied significantly by size (87% of 1.0-cm and 14% of 0.3-cm lumps) and hardness (56% of hard and 40% of soft lumps), but not depth; by specialty (from 50% for general internists to 40% for obstetricians), but not by level of training or experience; and by search duration (r=.59). On multiple regression analysis, only search duration was consistently associated with increased detection. Modest detection rates and wide variation suggest breast lump detection can be improved among physicians. Adequate search duration may be important for high detection rates.

(JAMA 1985;253:2224-2228)