Maxime Verhoeven

227 Summary and general discussion In conclusion, based on this thesis and literature, we assume that a treat-to-target treatment strategy initiating a csDMARD in combination with GC is the most cost- effective strategy nowadays, which is widely available as well. However, with bDMARDs becoming less costly, in specific subgroups, e.g., MTX non-responders, initial treatment with a bDMARD may possibly be more effective as well as cost-effective when combined with an active tapering protocol. Whatever the initial strategy, many RA patients eventually will (concomitantly) need a bDMARD. What is the optimal instrument to monitor disease activity in RA patients? Measurement instruments used for monitoring disease activity such as DAS28 currently are not always easy to implement in daily practice, as they are time consuming, and rather subjective. 8,19,20 A DAS28 measurement will take approximately 5-10 minutes of rheumatologists’ or healthcare professionals’ time, which is quite a proportion of the total time generally available for an out-patient-clinic visit. Further, as most of these measurements (except CDAI) include an acute phase reactant (APR), rheumatologists have to be aware that (at least) one of the current medical therapies (i.e., TCZ) strongly directly affects this factor, which in theory may result in underestimation of disease activity score and/or overestimation of treatment effect. 21 The HandScan could provide a more objective and faster measurement, and might be an alternative method for assessing disease activity. However, the correlation between DAS28 and OST-scores at one time point is moderate only, 22 and the model’s ability to estimate DAS28 over time (i.e., monitor disease activity) by OST-scores was low. Besides, OST-scores seem to be influenced by gender ( chapter 10, 11 and 12 ): OST-scores were higher in males compared to in females. This could be in part a system/device error; future research is warranted to exclude this. The OST-score can be considered an imaging tool; other studies, mostly investigating a simple correlation between an imaging tool (MRI or ultrasound) and a clinical disease activity tool, show similar results, i.e., a moderate correlation. 23–25 As a consequence, some clinical trials applied a treatment target based on a combination of clinical remission AND remission defined on imaging. However, in these studies outcomes for RA patients didn’t improve. On the contrary, it rather seemed to result in additional DMARD therapy, which might be considered as overtreatment. 26,27 We argue that replacement of the time consuming joint count assessments, by the OST-score in a composite index like DAS28 could be favourable as OST quantifies inflammation faster. Replacement would make more sense than adding the OST-score to the composite index, given that the advantage of saving time would be lost, and that adding an OST-based target would probably not result in better patient outcomes, as described above for other imaging modalities. The concept, DAS-OST, wherein joint counts have been replaced by OST, is tested against DAS28 and the rheumatologist’s 13