PURPOSE: To measure U.S. radiologists' workload and the variation and trends in workload. MATERIALS AND METHODS: A stratified, random sample of 794 radiology practice groups were surveyed about hiring and workload in 1996; 78% responded. Responses were weighted to represent all U.S. radiologists. Types of procedures performed in Medicare patients in 1991 and 1995 were analyzed. RESULTS: In 1995-1996, the average workload per full-time equivalent diagnostic radiologist was 11,600 procedures +/- 200 (standard error) annually, an apparent (but not statistically significant) 5% increase from 4 years earlier. Average relative value units per Medicare procedure increased by 8%, due to the increasing role of magnetic resonance imaging, computed tomography, and interventional radiology. The annual average number of procedures varied substantially by group type (e.g., 13,200 procedures +/- 500 in groups of two to four radiologists vs 10,300 +/- 300 in groups of 11 or more radiologists). Within any group type, radiologists at the 75th percentile in terms of workload typically performed at least 50% more procedures than did radiologists at the 25th percentile. CONCLUSION: Given the large variability in the annual number of procedures within and across group types, averages should not be taken as norms. Contrary to concerns about a possible surplus of radiologists, the workload per radiologist has increased substantially in the past few years.
PURPOSE: To measure U.S. radiologists' workload and the variation and trends in workload. MATERIALS AND METHODS: A stratified, random sample of 794 radiology practice groups were surveyed about hiring and workload in 1996; 78% responded. Responses were weighted to represent all U.S. radiologists. Types of procedures performed in Medicare patients in 1991 and 1995 were analyzed. RESULTS: In 1995-1996, the average workload per full-time equivalent diagnostic radiologist was 11,600 procedures +/- 200 (standard error) annually, an apparent (but not statistically significant) 5% increase from 4 years earlier. Average relative value units per Medicare procedure increased by 8%, due to the increasing role of magnetic resonance imaging, computed tomography, and interventional radiology. The annual average number of procedures varied substantially by group type (e.g., 13,200 procedures +/- 500 in groups of two to four radiologists vs 10,300 +/- 300 in groups of 11 or more radiologists). Within any group type, radiologists at the 75th percentile in terms of workload typically performed at least 50% more procedures than did radiologists at the 25th percentile. CONCLUSION: Given the large variability in the annual number of procedures within and across group types, averages should not be taken as norms. Contrary to concerns about a possible surplus of radiologists, the workload per radiologist has increased substantially in the past few years.