The number of deformed joints as a surrogate measure of damage in rheumatoid arthritis.

OBJECTIVE: To evaluate the reliability and validity of the number of deformed joints (NDJ) as a surrogate measure of joint damage in rheumatoid arthritis (RA).
METHODS: We tested interrater reliability and validity in determining the NDJ as a surrogate for joint damage in consecutive patients with RA. We rated each of 48 joints as normal or abnormal in terms of alignment and range of motion, and expressed the results as the total number of deformed joints. We compared the NDJ with the severity of damage on a plain radiograph of the hands, scored using Sharp's technique, as the gold standard measure of joint damage. We also compared the correlation between the NDJ and radiographic joint damage, on the one hand, and disease duration, performance-based measures of physical function, and the self-reported level of disability.
RESULTS: The interrater reliability of the NDJ was excellent, with an intraclass correlation among four examiners of 0.94. To assess validity of the NDJ, we studied 273 RA patients from 5 clinical settings. Their average NDJ was 14 (range 0-43), and their average Sharp's score for joint space narrowing and erosions combined was 106 (range 4-309). The NDJ and the total Sharp's score were highly correlated (r = 0.83). Both measures were correlated to a similar degree with disease duration (r = 0.51 for each measure), grip strength (r = -0.49 for NDJ, and r = -0.51 for Sharp's score), walking velocity (r = -0.44 for NDJ, and r = -0.45 for Sharp's score), the timed button test (r = -0.62 for NDJ, and r = -0.57 for Sharp's score), and the Modified Health Assessment Questionnaire (r = 0.38 for NDJ, and r = 0.38 for Sharp's score). Both the Sharp's score and the NDJ worsened significantly in 38 patients for whom 1-2 year followup data were available.
CONCLUSION: The NDJ is reliable and is strongly associated with the standard measure of joint damage in RA. Because it is easily performed in a clinical setting, it could be used as an economical surrogate of joint damage in studies of the long-term outcome of RA.