We surveyed general and family practitioners to evaluate their patterns of referring musculoskeletal disease patients to rheumatologists and orthopedists. Patients who had rheumatoid arthritis, systemic lupus erythematosus, and ankylosing spondylitis were most often referred to rheumatologists, whereas patients with osteoarthritis, persistent low back pain, and post-traumatic knee pain were most often referred to orthopedists. As conditions worsened in severity, referrals were more frequent. Patients with conditions that were difficult to diagnose, such as possible shoulder tendinitis that was unresponsive to initial nonsteroidal therapy, undiagnosed polyarthritis, and intermittent knee swelling with pain, were most often treated without referral and, when referred, were most often sent to orthopedists. Belief in the effectiveness of rheumatologists or orthopedists correlated strongly with reported referral behavior, yet most respondents considered themselves capable of managing the majority of patients with musculoskeletal diseases. Neither practice arrangement, board certification, nor educational background affected referral behavior. However, younger physicians were Submitted for publication November 19, 1984; accepted in revised form March 11, 1985. more likely (P = 0.002) to refer patients to rheumatologists. Multivariate analysis showed that the significant predictors of global referral behavior were belief in the effectiveness of subspecialists and a small number of musculoskeletal problems seen by the generalist. The predictors of referral to rheumatologists were belief in rheumatologist efficacy and young physician age.The referral of patients with musculoskeletal disorders is of vital interest to rheumatologists. The prevalence of primary care visits for rheumatic conditions is extremely high, yet the referral rate for these conditions is quite low (1). A small change in the percentage of patients referred would have an enormous impact on the demand for rheumatologists and the spectrum of diseases treated by rheumatologists. Nonetheless, subspecialists in general, and rheumatologists in particular, have not systematically studied which patients are being referred and which are not. Furthermore, although social scientists have investigated social networking and physician demographic characteristics and attitudes as determinants of referral behavior (2,3), the reasons behind the decision to refer some patients with specific diseases, while not referring others, have not been studied.We surveyed a group of general and family practitioners to determine how often specific musculoskeletal conditions are referred and to investigate the reasons behind the referrals.
METHODSWe developed a questionnaire designed to assess the frequency of referrals of patients with musculoskeletal diseases and to evaluate the reasons for the referral. The questionnaire underwent pilot testing with 10 general inter-
