Maxime Verhoeven

177 Utility of the HandScan in monitoring disease activity in RA 2. SJC and TJC changes of one SD unit were on average related to changes of 0.18 (95%CI 0.05–0.31) and 0.16 (95%CI 0.05–0.25) SD units of OST-score respectively. The explained variance increased with 4% and 3% respectively (to 32% and 43% respectively) when adding OST-score to the models. The association with SJC (but not TJC, p=0.52 for the interaction term) was found to be modified by RA stage (p=0.03 for the interaction term). Stratified analyses showed that one SD unit of OST-score change was on average related to 0.08 (95%CI -0.08–0.14) and 0.37 (95%CI 0.15–0.59) SD units in SJC, respectively, for early and established RA. No association of OST-score with HAQ nor EQ5D-5L was found (results not shown). Figure 2 Observed DAS28 vs. estimated DAS28 (using full model with OST-score). DAS28= disease activity score assessing 28 joints; Predicted DAS28= DAS28 as estimated by the model, with OST-score, visit month and DAS28 at the previous visit as variables. Baseline OST-score (odds ratio (OR) 0.93, 95%CI 0.83–1.04; standardized OR 0.67, 95%CI 0.37–1.22), and the short-termOST-score change (OR 1.04, 95%CI 0.90–1.19; standardized OR 1.18, 95%CI 0.65–2.13) were both not statistically significantly predictive for EULAR response at 3 months. Baseline DAS28 neither was a significant predictor (OR 0.76, 95%CI 0.41–1.41; standardized OR 0.74, 95%CI 0.38–1.44), whereas short-term DAS28 change (OR 4.47, 95%CI 1.73–11.58; standardized OR 3.96, 95%CI 1.65–9.52) was. Results for EULAR response at 6 months were in line with the above, see Table 3. For ACR50 response at 3 months, none of the variables were significant predictors. Short-term DAS28 change (OR 3.92, 95%CI 1.57–9.28; standardized OR 3.69, 95%CI 1.65–9.52) was 9