Variability in the interpretation of screening mammograms by US radiologists. Findings from a national sample

To evaluate the effectiveness of screening mammography by estimating the variability in radiologists' ability to detect breast cancer within the US population of radiologists at mammography centers accredited by the American College of Radiology.

A two-way sample survey design was used as follows. Fifty mammography centers having an American College of Radiology-accredited unit were randomly sampled from across the United States. One hundred eight radiologists from these centers gave blinded interpretation to the same set of 79 randomly selected screening mammograms. The mammograms were from women who had been screened at a large screening center. Before their sampling, these women had been stratified by their breast disease status, established either by biopsy or by 2-year follow-up. Rates of biopsy recommendations were summarized by the mean, median, minimum, maximum, and range of sensitivity and specificity. Overall cancer detection ability was summarized by similar statistics for receiver operating characteristic curve areas. Ninety-five percent lower confidence bounds on the ranges in accuracy measures were established by boo-strapping.

There is a range of at least 40% among US radiologists in their screening sensitivity. There is a range of at least 45% in the rates at which women without breast cancer are recommended for biopsy. As indicated by receiver operating characteristic curve areas, the ability of radiologists to detect cancer mammograms varies by as much as 11%.

Our findings indicate that there is wide variability in the accuracy of mammogram interpretation in the population of US radiologists. Current accreditation programs that certify the technical quality of radiographic equipment and images but not the accuracy of the interpretation given to mammograms may not be sufficient to help mammography fully realize its potential to reduce breast cancer mortality.