Physician and practice characteristics, frequency of performance, and the Resource-Based Relative Value Scale

The Resource-Based Relative Value Scale is based in part on the ratings of the work of services obtained from a random sample of physicians in a specialty. Ratings are used without regard to board-certification or other characteristics of the physician, or to the physician's experience with the service. Critics have suggested that all physicians may not be equally qualified to rate the work of services. Using data obtained from the Resource-Based Relative Value Scale surveys and analyzed using multiple regression methods, the authors found that physician and practice characteristics explain, at most, a small fraction of the variation in ratings of work. Any increase in the precision of the work scale obtained by adjusting physicians' work ratings according to physicians' characteristics could be achieved at lower cost by a slight increase in sample size. Associations between frequency of performing a service and ratings of work are about as likely to be in one direction as another. Most of the differences between estimates of work, excluding and including physicians who have not performed a service, are less than 2% in absolute value, and all are less than 10%. Estimates of work using ratings of physicians in the upper half in frequency of performance are usually within 10% of estimates using other ratings. Even if the observed associations are not due to chance, the potential improvement in accuracy of estimates of work appears too small to justify using data on frequency of performance.