Objective: To determine the extent to which characteristics of rheumatologists and their practices explain the variation in their use of prednisone and 2nd line agents for the treatment of rheumatoid arthritis (RA).
Methods: We used multiple logistic regression to examine the relationship between the use of prednisone, hydroxychloroquine, intramuscular gold, and methotrexate, and the following categories of rheumatologist characteristics: professional experience, primary payment method, practice setting, location of rheumatology training, and demographic characteristics. Our explanatory variables also included 12 patient characteristics and a random effect term.
Results: Much of the variation among rheumatologists in the use of these agents is explained by the rheumatologist characteristics. Depending on the agent, professional experience explained 15 to 54%, payment method 3 to 27%, practice setting 6 to 39%, training location 12 to 53%, and demographic characteristics 3 to 23% of the rheumatologist associated variation in use of each agent.
Conclusion: There are identifiable characteristics of rheumatologists and their practices that strongly influence their treatment decisions for RA. These findings have important policy implications.