RESUMO
OBJECTIVES: To assess the inter-observer agreement of standard joint count between experienced Rheumatology professor (Prof) and young Rheumatology fellow (candidate), and to compare disease global assessment between professor, young candidate and patients. METHODS: This study included one hundred rheumatoid arthritis patients. For all patients independent clinical evaluation was done by two rheumatologists (professor and candidate) for detection of tenderness in 28 joints and swelling in 26 joints. The study also involved global assessment of disease activity by the provider (Prof and candidate) (EGA) as well as by the patient (PGA). The EGA was determined without previous knowledge of the patient's laboratory test results. RESULTS: A highly significant accordance (correlation) between professor and candidate was found in both the number of tender joints (p<0.001) (r=0.946), and the number of swollen joints (p<0.001) (r=0.797). Regarding swollen joints, the highest agreement was in right knee (0.929), while poor agreement was found in the right 5th MCP (0.049). Regarding tender joints, the highest analogy was in the right elbow (0.899), in contrast to the left 3rd PIP (0.462) which showed the least congruence. Agreement study using kappa measurement for disease global assessment showed: moderate agreement (between professor and candidate) (0.405), fair agreement between (professor and patient) (0.213), fair agreement between (candidate and patient) (0.367). CONCLUSION: Inter-observer reliability was better for TJCs than SJCs. Regarding SJCs agreement was better in large joints such as the knees compared to the small joints such as the MCPs. Disease global assessment may show discrepancy between patients and physicians.