Objective: To assess the criteria for remission based on Disease Activity Score 28 (DAS28) and DAS28-3 (excluding patients' evaluation of disease activity) compared to American College of Rheumatology (ACR) preliminary criteria in established rheumatoid arthritis (RA), and to examine the value of each ACR criterion individually.
Methods: The EMECAR study was designed to assess the burden of comorbidity and inflammatory activity for RA in Spain. A random sample of 788 patients with RA from 34 Spanish centers was selected. Remission was defined by preliminary ACR criteria applied specifically and the clinical activity assessed by the DAS28 and the DAS28-3. A receiver operating characteristics curve analysis was performed to identify cutoff values with the highest usefulness in defining remission on both DAS indices.
Results: Thirty-two patients (4.1%) were in ACR-defined remission, 62 (7.9%) if fatigue was excluded from the criteria. The frequency of any single criterion that patients in remission fulfilled: no fatigue and joint pain by anamnesis in 31 patients (96.9%); morning stiffness < 15 min in 26 (81.3%); no swelling in joints in 21 (65.6%); normal erythrocyte sedimentation rate (ESR) in 29 (90.6%); and no joint tenderness in 21 (65.6%) patients. The positive predictive value for remission of each criterion: normal ESR 6.5%; morning stiffness < 15 min 8.4%; no fatigue 8.7%; no joint tenderness 13%; no swelling in joints 15.8%; and no joint pain by anamnesis 27.7%. The DAS28 cutoff values with higher discriminatory power for remission were 3.14 (sensitivity 87%; specificity 67%) when all the ACR criteria were used, and 2.81 (sensitivity 84%; specificity 81%) when fatigue was omitted. The equivalent cutoffs for the DAS28-3 were 3.52 (sensitivity 84%; specificity 66%) and 2.95 (sensitivity 82%; specificity 83%), respectively.
Conclusion: DAS28 and DAS28-3 are good tools to define remission in established RA. No joint pain by anamnesis is the criterion with the highest value in defining remission, while normal ESR, an absence of morning stiffness, and fatigue are the least effective.