Clinical joint examination is less time-consuming than ultrasound in rheumatoid arthritis. Knowledge of clinical and ultrasound concordance of joints groups could help in selecting joints for a best ultrasonographic assessment.

To evaluate concordance between clinical examination and ultrasound of joints in a heterogeneous group of patients with rheumatoid arthritis.

Forty patients were included in a prospective, transversal, single-center study, whatever disease activity, duration or treatment. In each patient, 40 joints were evaluated for a total of 1600 joints. Synovitis was scored using clinical examination, B-mode, power Doppler and both B-mode and power Doppler. Concordance between swelling joint by clinical examination, synovitis thickening by B-mode (grade 1 or higher) and inflammation by power Doppler (grade 1 or higher) was assessed by computing the kappa coefficient.

Clinical joint examination and ultrasound concordance was very low at the shoulders and metatarsophalangeal joints (κ<0.1) and was low at wrists (κ: 0.23 to 0.30). B-mode and power Doppler found 2.4 and 1.4 more synovitis than swollen joint count using clinical examination and up to 30 times more at metatarsophalangeal joints. Concordance was strong at tibio-talar joints (κ: 0.65 to 0.82) and moderate at others joints sites (κ: 0.4 to 0.6).

Assessment of a heterogeneous group showed that ultrasound adds information to clinical examination, most notably at the shoulders, wrists and metatarsophalangeal joints. Concordance was moderate to strong at other joint sites.