SAT0211 Interobserver agreement of ultrasound examination in assessment of the small joints of the hands and feet in rheumatoid arthritis

Background Ultrasound examination (US) of the small joints of the hands and feet in rheumatoid arthritis (RA) has the potential to become a rheumatological bedside procedure to assess early joint pathology. There is a growing number of studies and possible practical applications of US in this field. However, US has not been sufficiently validated with respect to common and early joint manifestations of RA. Objectives To evaluate the interobserver agreement of ultrasound examination in assessment of finger and toe joints in rheumatoid arthritis. Methods B-mode and power Doppler US was performed on 150 small joints of 30 patients with active RA (ACR 1987 criteria). Nineteen patients had a disease duration of 2 years or more and eleven patients had a disease duration of less than 2 years. Sixty-one of the joints were clinically active (tender and/or swollen), as evaluated by an experienced rheumatologist. US was performed with a LOGIQ GE 500 unit by means of a 7–13 MHz linear array transducer. In each patient, five preselected small joints were examined: 2nd and 3rd metacarpophalangeal joints, 2nd proximal interphalangeal joint, 1st and 2nd metatarsophalangeal joints. The number and location of bone erosions, synovial thickening, joint effusion and power Doppler signal in these joints were evaluated and classified on four-grade semi-quantitative scales. Both US examinations were performed independently on the same day by two observers – a radiologist with expertise in musculoskeletal US and a rheumatologist trained in US evaluation of small joints of the extremities. The observers reached consensus as to the evaluation scale prior to the beginning of the study. Results In the examined joints, exact agreement between the two observers was seen in 84% of the examinations with regard to bone erosions, in 78% as to synovitis, in 84% as to joint effusions and in 82% as to power Doppler signal assessments. A difference of more than one grade on the evaluation was found in 6, 11, 1 and 3%, respectively. Corresponding unweighted kappa values were 0.69, 0.63, 0.63 and 0.51. Conclusion The present study showed a good degree of interobserver agreement for the examined parameters. Our results suggest that ultrasound examination is a reproducible method for assessment of RA joint pathology of the finger and toe joints.