Background/Purpose: Systemic Lupus Erythematosus Disease Activity Index-2000 Responder Index-50 (S2K RI-50) is a reliable and valid index able to measure ≥ 50% improvement in disease activity. We aimed to determine the Minimal Clinically Important Difference (MCID) for improvement for S2K RI-50.

Methods: Analysis was conducted on patients seen during 2010-2012 at a single lupus centre. The S2K RI-50 data retrieval form was completed at each visit. Physician global assessment was determined at baseline visit on a visual analogue scale (0-10) and at follow up visits on a 7-point Likert scale (LS) (1-3 reflects worsening, 4 unchanged, 5 slightly improved, 6 moderate improved (≥ 50%) and 7 much improved). This analysis is focused on the first follow up visit. LS was collapsed into 2groups; LS 6-7 and LS 1-5. The change of SLEDAI-2K scores was calculated between baseline and follow up. The change of S2K RI-50 was calculated between baseline SLEDAI-2K and follow up S2K RI-50 scores. MCID was determined with the anchor-based and the distribution-based approaches. Anchor-based approach using the whole cohort: we modeled the 1st follow up LS ≥ 6 as the outcome variable, and continuous S2K RI-50 change as the predictor in Logistic Regression model. Area Under the Curve (AUC) was obtained to determine the predictive power of S2K RI-50. We also derived the best S2K RI-50 cutoff based on optimal sensitivity/specificity in receiver operating characteristic (ROC) determined by the Youden index. Distribution-based approach: Standardized Response Mean (SRM) was used to confirm the effect size. The Standard Error of Measurement (SEM) was derived based on the following equation:  SEM =

[S2K RI-50 test-retest reliability = 0.98 based on a previous publication from the same centre]  

Results: 509 patients were studied (age and lupus duration at baseline visit were 44.3 ± 14.7 and 15.2 ± 11.0 years respectively). 48 patients had an improvement (LS 6 or 7). The characteristics of the patients with LS 6-7 and LS ≤5 are represented in table 1. Anchor-based approach: MCID of improvement is equal to 1 based on the ROC analysis [AUC 0.82 (95% CI: 75-89), sensitivity 81% and specificity 72%] (Figure 1). Distribution-based approach: SRM of S2K RI-50 for LS 6-7 was -0.97 (95% CI: -1.3, -0.6) which is considered a large effect size. SEM was equal to 0.96 from the previous equation.

Conclusion: The estimated MCID of S2K RI-50 derived from both ROC analysis and SEM confirmed that it is close to 1. Thus a reduction of S2K RI-50 of 1 represents a relevant change in disease activity.