Revising DAS28 scores for remission in rheumatoid arthritis.

Abstract

Current initiatives to treat rheumatoid arthritis (RA) to target remission have resulted in the widespread use of composite outcome measures to quantify disease activity. Simplified Disease Activity Index (SDAI) ≤3.3 is the gold standard of remission. Previous work has suggested that the remission threshold of DAS28-ESR or DAS28-CRP ≤2.6 is known to be not strict enough and should be revised. There is no previous study that looks at the equivalent DAS28-CRP value that best reflects SDAI remission in a real-life cohort. Consecutive cases fulfilling ACR/EULAR classification criteria for RA from one centre were included if they had an optimum number of visits with recording of all raw data to calculate DAS28-ESR, DAS28-CRP and SDAI. Data from seven visits each of 76 patients, providing 532 data points was examined. Spearman's correlation between DAS28-ESR, DAS28-CRP and SDAI was 0.91-0.97 (p < 0.001). A Bland-Altman plot demonstrated a mean difference of 0.37 units between DAS28-ESR and DAS28-CRP (p < 0.001). ROC and kappa analysis provided values of 2.58 and 2.55 for DAS28-ESR4V and 2.09 and 2.05 for DAS28-CRP4V for SDAI value of 3.3, respectively. The two versions of DAS28 using ESR and CRP and SDAI correlate but are not equivalent or interchangeable for an individual patient. The DAS28-CRP overestimates remission and should be revised downwards to a proposed value of 2.1.