Why do general practitioners request rheumatoid factor? A study of symptoms, requesting patterns and patient outcome

Abstract

Background: To investigate the reasons why general practitioners (GPs) request rheumatoid factor (RF) assays, we studied 200 consecutive requests for RF from general practice in 1995. Method: By means of an audit questionnaire, we studied 100 negative, 50 positive and 50 borderline RF results and compared these with the presenting symptoms that prompted the request, the GPs' understanding of the significance of the result, the referral intention and behaviour of the GP, and finally, the patient outcome after 5 years. Results: There was an 80% response rate. The presenting symptoms closely matched the American Rheumatism Association revised criteria for the classification of rheumatoid arthritis, indicating that the requests were made on valid clinical grounds, with polyarthralgia, morning stiffness and joint pain being the most common. Most GPs considered a negative or positive result to be meaningful, in that a positive RF meant that a referral was more likely than with a negative or borderline result, even in the presence of appropriate symptoms in all three groups. Seventeen to thirty per cent felt that the test excluded or confirmed RA. The result appeared to influence this decision to a greater extent than it should. A 5-year follow-up on these patients showed that 26/40 patients with positive RF were referred, and that 25 of them developed a rheumatic disease of some kind, with 17 patients eventually being diagnosed with RA. Only 17/80 patients with negative RF were referred, the remainder having no autoimmune problem evident after 5 years, 11 of them developing a rheumatic disease, and only three being diagnosed with RA. Conclusions: Although this is a locally based study, we believe the conclusions would be applicable to all laboratories and GPs undertaking these tests. RF requests are made on valid clinical grounds by GPs, but there may be an over-reliance on the results as regards referral behaviour. If patients were referred on clinical grounds, this would significantly lengthen consultants' waiting lists.