Search

RATIONALE AND OBJECTIVES: The total number of procedures and their relative value units (RVUs) were used to measure the productivity of radiologists. Besides variations in productivity due to differences in the percentage of clinical effort, baseline productivity also varies among clinical subspecialty sections. The authors' previous research used the full-time equivalent (FTE) as the unit to adjust for differences in the percentage of clinical effort and a set of adjustment factors (or calibration constants) to modify the default work RVUs according to types of procedures. These adjustments led to comparable average productivity measurements across subspecialty sections. Since 2003, radiology practice has continued to change, including the introduction of positron emission tomography/computed tomography into clinical practice, suggesting a need to update the understanding of clinical productivity and refine the authors' adjustment procedure. In this study, the authors analyzed the most recent survey of academic departments and derived updated adjustment factors for differences in workload among subspecialty sections. The results can be used to determine faculty staffing requirements and evaluate radiologists' performance. MATERIALS AND METHODS: A survey performed by the Society of Chairmen of Academic Radiology Departments collected data in 2006 for 1,134 radiologists in 24 departments, including 10 departments that also reported productivity in an earlier 2003 survey. These data included the numbers of procedures (represented by Current Procedural Terminology [CPT] codes) performed by radiologists, percentage clinical effort, subspecialty sections, and the number of clinical days. The numbers of CPT codes were converted into total work RVUs per FTE faculty member. By grouping the CPT codes into 6 prespecified examination categories, adjustment factors were created to adjust the RVUs for CPT categories to ensure that the median total adjusted work RVUs from different subspecialty sections were comparable. RESULTS: Overall, the mean clinical workload in 2006 was 9,671 examinations, a statistically significant 15% increase from 2003. The mean number of work RVUs per FTE was 7,136, a 22% increase from 2003. The adjustment factors have been modified from those presented in the authors' earlier paper, including reductions for interventional radiology, computed tomography, magnetic resonance imaging, nuclear medicine, and a new adjustment factor for 'special procedures.' These adjustments reduced differences in adjusted RVUs per FTE between subspecialty sections. CONCLUSIONS: Clinical workload, as measured by RVUs per FTE and adjusted RVUs per FTE, is very useful for determining optimal staffing in subspecialty sections and in radiology departments in general. Workload continues to increase, both in examination complexity and in numbers of overall procedures. Adjustment factors make workload comparisons between subspecialty sections more valid and meaningful. [ABSTRACT FROM AUTHOR]

Background: Medicare data may be a useful source for determining the utilization of mammography among elderly women, but the accuracy of these data has not been established. Objective: We determined whether Medicare physician billing claims are an accurate reflection of mammography utilization among women ages 65 and older and whether they can be used to assess the use of screening as compared with diagnostic mammography. Data Sources: Mammography use was assessed using Medicare billing claims and radiology reports from 2 mammography registries; the San Francisco Mammography Registry and the New Mexico Mammography Registry. Methods: Completeness of the Medicare data was assessed by comparing mammography use based on Medicare, with radiology reports from the mammography registries, which served as the referent standard. Capture rates for Medicare claims for individual mammograms were examined, and women were characterized as having undergone at least 1 mammogram within each 2-year period based on the Medicare data, and these rates were compared with the referent standard. To determine whether Medicare data can distinguish between screening and diagnostic mammography, we performed a classification analysis using the mammography registries screening/diagnostic designation as the referent standard (dependent variable) and Medicare claim information as the independent/predictor variable. On the basis of the mammogram level classification analysis, women were categorized as having been frequently screened (at least 2 screening mammograms spaced by 12 to 36 months), screened (at least 1 screening mammogram), or not screened. Subjects: Women ages 65 and older, diagnosed with breast cancer between 1992-1999, who had at least 1 mammogram between 1992-1999 were examined. Results: A total of 3340 mammograms were obtained in 1371 women between 1992 and 1999. Overall, 83% of mammograms obtained by these women had a corresponding billing claim in Medicare. This increased from 65% in 1992 to 90% in 1999. Of women who underwent at least 1 mammogram during each 2-year period per the referent standard, 94% of women were accurately classified by Medicare claims as also having undergone mammography during the same 2-year period. In multivariable analysis, a mammogram was more likely to be associated with a billing claim over time, for women 80 years or older, and for white and Asian as compared with Hispanic women. Neither socioeconomic status nor screening/diagnostic designation affected the likelihood that a mammogram would be associated with a billing claim. The Medicare data accurately categorized a given mammogram as screening or diagnostic for 87.5% of mammograms. Lastly, there was moderate to substantial agreement in the categorization of women as frequently screened, screened or not screened between the 2 data sets (weighted kappa 0.74, 95% confidence interval 0.70-0.78). Conclusion: Medicare administrative claims are reliable for assessment of mammography utilization and have become more accurate over time. Medicare claims data also provide a mechanism for designating mammography as screening or diagnostic, which subsequently may allow accurate description of a woman's screening history. [ABSTRACT FROM AUTHOR]