Background/Purpose: Clinical measures of disease activity, such as the DAS-score and ultrasound (US) scores of disease activity can be sometimes yield discordant results. Little research has attempted to understand the reasons for these discordances. Moreover, it is not well known what the implications of such discordances are or how clinical and US assessments evolve over time in this situation

Objectives :First: to determine the percentages of patients presenting discordances between DAS and US assessments in a real-life cohort. Second: to describe associated factors. Third: to evaluate the evolution of both measures of disease activity over time.

Methods: All patients with at least one concomitant US- and DAS-score assessment, performed since the introduction of validated US (SONAR) score in the SCQM between 2009 and January 2017 were included. Disease activity was categorized as remission, low, moderate and high activity based on previously established cut-offs for the DAS and the US-score. Potential predictors of discordance were extracted from the SCQM database, including age, gender, seropositivity, duration of illness, number of swollen and tender joints, global physician assessment of disease activity, HAQ, presence of fibromyalgia, type of treatment, patients followed in private practice and other. A longitudinal analysis was performed in all the patients with at least two subsequent visits with US and DAS assessment performed simultaneously

Results: 2367 assessments could be analyzed, of which 1072 (45%) were considered concordant based on identical disease activity states with the DAS- and the US score. The proportion concordant assessments significantly differed by clinical disease status (p<0.001); more frequent in clinical remission (78.4% of agreement) compared to active disease status (21% in low or moderate DAS, 45% in high DAS). Among the discordant assessments, disease activity state tended to be more frequently over-estimated by the DAS compared to US-score (38%), than the other way round (17%), (p<0.05) .

Factors associated with the presence of discordant results were the swollen joint count (p <0.001), the overall estimation of the disease activity (p<0.001) by the clinician, the duration of the disease (p<0.02)

For 1181 patients, several DAS and US assessments were available during follow-up. The proportion of discordances during the follows up was similar to the initial evaluation. Initial discordance/concordances could however change status without obvious reason, especially in the moderate and low disease activity subgroup (75 % new discordances).

Conclusion: Discordances between DAS and US assessments appear to be higher than expected in real life. Both outcome measures can lead to over- or under-estimations of the true disease activity, Discordant assessments seem to be linked essentially to inaccuracies in the clinical evaluation in particular (inadequate swollen joint counts) and /or to limitations of the US procedure (especially poor distinction between moderate and low activity disease

« Back to ACR/ARHP Annual Meeting 2017

ACR Meeting Abstracts - https://acrabstracts.org/abstract/comparison-of-clinical-and-ultrasound-measures-of-disease-activity-in-a-large-national-real-life-cohort-of-ra-patients/