Physician Workload in Mammography

Abstract

United States Food and Drug Administration (FDA) guidelines for certification require that radiologists interpret > or = 960 mammography examinations within each 2-year period (approximately 480 annually). The purpose of our study was to estimate per-physician annual volumes of mammography interpretation. Our study includes 4.2 million mammography examinations performed at 196 facilities between 1998 and 2004. We calculated the annual interpretive volumes per physician, the proportion of mammography examinations interpreted by radiologists in specified volume categories, and the impact on mammography capacity if annual interpretive volume requirements increased. The mean annual mammographic interpretive volume was 1,777. Approximately 31% of radiologists interpreted < 1,000 mammography examinations annually, yet these low-volume radiologists interpreted only 10% of all mammograms. The 10% of radiologists who interpreted > or = 3,000 mammography examinations annually interpreted 32% of all examinations. Rural radiologists interpreted fewer examinations annually compared with urban radiologists. If the minimum annual volume requirement were increased to 1,000 mammograms per year, only 10% of the overall U.S. mammography capacity would be affected. If the requirement were increased to 2,000 mammograms annually, 47% of capacity would be eliminated, and a major rearrangement of workload would be required because most radiologists would no longer interpret enough examinations to meet the revised standards. Doubling physician annual volume requirements would result in a small impact on overall mammography capacity. Increasing volume requirements to 2,000 mammography examinations annually would require a dramatic increase in the number of mammography examinations interpreted by the higher volume radiologists. Unless previously low-volume radiologists increased their volumes, raising requirements to 2,000 examinations could curtail access to mammography, particularly in rural areas.