Maxime Verhoeven

179 Utility of the HandScan in monitoring disease activity in RA association of the HandScan with DAS28, which would be a prerequisite for using such an instrument for monitoring disease activity over time, the added value (explained variance) was low. This limits the use of the HandScan as a comprehensive instrument for monitoring individual patients’ disease activity. A plausible explanation for the low ability to estimate DAS28 with OST-scores, may be that in addition to the number of tender and swollen joints (from 28), an acute phase reactant (i.e., ESR) and a VAS expressing the patients’ assessment of disease activity are part of DAS28, 13 whereas the OST-score measures only RA inflammation of hand and wrist joints (with a maximum of 22 joints). Therefore we evaluated also components of DAS28 separately. Indeed the association with TJC and, especially, SJC was (somewhat) stronger than with DAS28 as apparent from the higher standardized regression coefficients and increase in explained variance by adding OST-score to the longitudinal model. We could not establish a predictive association of baseline OST-score or short-term changes in OST-score with later response to treatment. This lack of predictive ability may again, partly, be due the fact that OST-scores only reflects joint inflammation in a limited set of joints and response criteria are based on composite scores. 14 As shown in Supplementary Table S1, GC therapy was used in early as well as in established RA. GC therapy diminishes disease activity, but this will likely have equally been the case for DAS28 and the OST-score. Therefore we think this has had no or only limited influence on the results of our main analysis, i.e., the longitudinal association between OST-score and DAS28. Since inflammation of osteoarthritic joints is generally considerably less than in RA joints, and the DIP joints (mostly affected in osteoarthritis) are not assessed in the HandScan, we expect that the influence of concomitant osteoarthritis on OST-score results have been limited. A limitation of this study is that the sample size is modest. The intention of the current study was to explore whether a longitudinal association of OST-score with DAS28 existed, which is a prerequisite for using OST-scores as a disease activity measurement in patients over time. Therefore, we aimed to include at least 30 early as well as 30 established RA. In the analysis we combined early (n=32) and established (n=32) RA patients in the analyses, correcting for RA stage. It turned out that RA stage did not influence the longitudinal association between OST-score and other outcomes, except for SJC. A possible explanation might be the fact that in early RA patients the SJC was often 0 at follow-up due to the strict treat-to-target treatment approach, possibly obscuring small changes over time, whereas SJC was generally higher in established RA patients. 15 In addition, the predictive association between short-term changes in OST-score and longer-term response was also not influenced by RA stage. Furthermore, one may expect that type of treatment may influence the ability of OST-scores to detect 9