Residual minimal disease activity in rheumatoid arthritis: a simple definition through an in-depth statistical analysis of the major outcome

Objective. To obtain the simplest definition of minimal disease activity (MDA) and to compare it with published proposed definitions of MDA in patients with RA.Methods. Two hundred and fourteen patients with long-standing RA (LSRA) were evaluated for clinical and laboratory parameters. Factor analysis was performed to remove redundant variables included in the core set measure for MDA definition stated by the OMERACT. Receiver operating characteristic (ROC) curves analysis allowed to obtain optimal cut-off predictors of a 28-joint disease activity score (DAS28) <= 2.85. These were tested in 112 LSRA and 95 early-onset RA (ERA) patients.Results. Factor and ROC curve analysis showed that the best predictors of a DAS28 <= 2.85 in LSRA cohort were: (i) ESR < 20 mm/h (sensitivity: 80%, specificity: 54%); (ii) swollen joint count (out of 28) <= 2 (sensitivity: 95%, specificity: 74%); (iii) patient global assessment (0-100) <= 15 (sensitivity: 78%, specificity: 78%); and (iv) HAQ (0-3) <= 0.5 (sensitivity: 91%, specificity: 61%). To each of these four criteria we assigned a value of 1 when it was satisfied (score ranging: 0-4). The cut-off with the highest overall accuracy for identifying RA patients with DAS28 <= 2.85 was a score >= 3. We adopted these four parameters in order to define the residual MDA (RMDA). Comparing RMDA criteria, in distinct 112 LSRA and 95 ERA patients, with OMERACT, Simplified Disease Activity Index and Clinical Disease Activity Index definitions of MDA, we found a good agreement in the LSRA cohort and moderate agreement in the ERA cohort.Conclusions. HAQ, PaGA, SJC28 and ESR allow identification of RA patients with an RMDA. The RMDA criteria behaves similarly to OMERACT definitions, but appears more simple and feasible.