"Our result that reader volume affects cancer-detection accuracy is not very surprising," concludes Laura Esserman, M.D., M.B.A., University of California, San Francisco, and her coworkers. Past studies have shown that the number of patients who died from medical procedures while under the care of a physician decreased dramatically if that physician has performed a certain number of procedures. For mammography, a minimum of 2,500 interpretations per year has been associated with improved cancer detection rates.
Radiologists in the United Kingdom are required to interpret a minimum of 5,000 mammograms per year to meet the national standard, while U.S. radiologists need only interpret 480 mammograms per year to meet the national standard. However, U.K. mammographers perform fewer biopsies than U.S. mammographers.
To test their hypothesis that reader experience, not number of interventions performed, determines the quality of mammographic interpretations, Esserman and her coworkers asked 60 U.S. radiologists and 194 U.K. radiologists to interpret 60 mammography films that contained 13 undisclosed cancers.
The U.S. radiologists were divided into three groups: low-volume radiologists who read 100 or fewer mammograms per month, medium-volume radiologists who read 101 to 300 mammograms per month, and high-volume radiologists who read more than 300 mammograms per month. All the U.K. radiologists were high-volume readers. Among the U.S. radiologists, those who interpreted the most mammograms per month detected the most cancers, but it was still not as many as the U.K. radiologists detected.
"Our finding that higher volume improves diagnostic performance suggests that there may be an opportunity to improve quality and efficiency by re-engineering the organization of U.S. mammography screening programs," the authors conclude. "Higher quality does not need to come at the price of more interventions."
In an editorial, Joann G. Elmore, M.D., of the University of Washington School of Medicine, and Patricia A. Carney, Ph.D., of Dartmouth Medical School, say the results of this study deserve follow-up, but point out that perhaps U.S. radiologists who "enjoy" interpreting mammograms feel more confident about their skills and, as a result, read more films. Moreover, they say, artificial test environments may not reflect real-world consequences.
They conclude, "Increasing the volume of interpretations for each mammographer in the United States may not result in improved accuracy if influences, such as the fear of medical malpractice, financial rewards, or differing levels of comfort with ambiguity in clinical decision making, remain unchanged."
Contact: Eve Harris, University of California, San Francisco, (415) 885-7277; fax (415) 502-1804, eharris@pubaff.ucsf.edu
Editorial: Susan Gregg-Hanson, University of Washington Harborview Medical Center, (206) 731-4097; fax: (206) 731-8605, sghanson@u.washington.edu
Esserman L, Cowley H, Eberle C, Kirkpatrick A, Chang S, Berbaum K, Gale A. Improving the accuracy of mammography: volume and outcome relationships. J Natl Cancer Inst 2002;94:369-75.
Editorial: Elmore JG, Carney JG. Does practice make perfect when interpreting mammography? J Natl Cancer Inst 2002;94:321-3.
Attribution to the Journal of the National Cancer Institute is requested in all news coverage.
Journal
JNCI Journal of the National Cancer Institute