Prognostic factors for joint destruction in rheumatoid arthritis:: A prospective longitudinal study of 318 patients

Objective. To quantify articular damage and to investigate prognostic factors for joint damage progression in rheumatoid arthritis (RA). Methods. RA patients satisfying the 1987 American College of Rheumatology criteria and with disease duration under 5 years were sampled from the EURIDISS longitudinal cohort study in Norway, The Netherlands, and France. Hand radiographs were assessed at baseline and at 2 to 3 year followup using Sharp score modified by van der Heijde. Assessment of erosion and joint space narrowing, performed in sequential order by a single reader blinded to patients' characteristics, had high intraobserver reproducibility (intraclass correlation coefficient 0.98-0.99). Baseline prognostic factors were analyzed in a multivariate linear regression model. Results. A total of 318 patients with RA aged 52.4 years (70.4% were female) and with a mean 2 years' disease duration at baseline were followed over 30 months. Median (quartiles) baseline and followup modified Sharp scores were 3 (0-11) and 9 (1-27), respectively, with 35.8% and then 22.3% of patients with no radiological damage. Controlling for age, sex, and country, the final joint damage was predicted by baseline modified Sharp score, rheumatoid factor positivity, time from disease diagnosis, patient global health assessment, and erythrocyte sedimentation rate, and by followup duration, explaining 76.8% of the outcome variance. Conclusion. This multinational study confirmed the prognostic role in RA of a set of features previously identified in smaller cohorts. It indicates which disease characteristics should be focused on in the early years of RA to identify patients at higher risk of developing severe disease and who are candidates for aggressive therapy.