Background:Treatment of Rheumatoid Arthritis (RA) has improved significantly based on early treat-to-target (T2T) strategies. Still, decreased health related quality of life (QoL), restricted ability to work and other unmet needs are reported by RA patients even in the absence of disease activity. We previously identified 3 factors representing the broader impact of RA using exploratory factor analysis: a patient-reported factor (patients global health, pain, fatigue and HAQ), a clinical factor (physician’s global health, tender and swollen joint count), and a laboratory factor (ESR and CRP)¹.Objectives:To test whether the discordance between patient-reported (PRF) and clinical(CF)/laboratory(LF) measures can predict QoL, or has a mediating effect in predicting future disease burden based on disease activity (DAS28CRP).Methods:This is a post-hoc analysis of the 2-year CareRA trial. PRF, CF and LF scores were calculated as weighted (by factor loading) sum of their components at week 16, 52 and 104 after treatment initiation. A discordance score (DS) between PRF and the mean of the other two scores was also computed.Mediation analyses were fitted to test the hypothesis (Figure 1) that DS could be a mediator for predicting PRF, CF and LF at a future time point (week 16, 52 and 104) using DAS28CRP at a previous time point (baseline and week 16). Confidence intervals were estimated via 10 000 bootstraps. Finally, a linear regression was fitted for DS to predict future QoL (RAQoL questionnaire; range 0-30; higher values indicating worse QoL).Results:Patients with early RA (n=379) were included with a mean (SD) age of 53.9 (13.0), 77% seropositive and 69% women.The DS was shown to be mediating the effect of DAS28CRP on any future PRF (Table1). On the other hand, there was no mediation effect of the DS in the prediction of the CF and an inconsistent mediation effect when predicting the LF.Moreover, the DS at week 16 significantly predicted (p<0.0001) RAQoL scores at year 1 with an effect of β 19.05 (SE 1.58) and an R² of 0.30 (CI 0.22-0.38). Similarly, it predicted RAQoL (p<0.0001) at year 2 with a β 19.74 (SE 1.56) and R² of 0.32 (CI 0.24-0.40).Table 1.Results of mediation analyses for prediction of future burden based on previous DAS28CRP and mediated by discordance.TimepointPredictor variablesDirect Effect95% CIsR²Mediation effect Patient-reported factorW16DAS28CRP at BL-0.0091-0.0240, 0.00580.1450PresentDS at BL0.0246*0.0169, 0.03310.1784W52DAS28CRP at W160.0215*0.0010, 0.04190.3394PartialDS at W160.0580*0.0442, 0.07390.2749W104DAS28CRP at W160.0101-0.0102, 0.03050.2798PresentDS at W160.0528*0.0396, 0.06860.2749Clinical factorW16DAS28CRP at BL0.0153*0.0074, 0.02320.0599AbsentDS at BL0.0019-0.0010, 0.00480.1784W52DAS28CRP at W160.0365*0.0267, 0.04630.1944AbsentDS at W160.0034-0.0031, 0.00950.2749W104DAS28CRP at W160.0115*0.0024, 0.02070.0409AbsentDS at W160.0033-0.0019, 0.00890.2749Laboratory factorW16DAS28CRP at BL0.0063*0.0015, 0.01110.0634PartialDS at BL0.0030*0.0012, 0.00500.1784W52DAS28CRP at W160.0003-0.0063, 0.00680.0305PresentDS at W160.0051*0.0012, 0.00960.2749W104DAS28CRP at W16-0.0007-0.0079, 0.00640.0014AbsentDS at W160.0013-0.0019, 0.00460.2749W: week BL: baseline DS: discordance scoreDAS28CRP: disease activity score in 28 joints with C-reactive protein*p<0.01Conclusion:Early discordance between patient-reported and biological/clinical factors mediates the effect of disease activity on future patient-reported outcomes, but also predicts QoL. Paying attention to this early discordance might provide opportunities to prevent patient’s unmet needs by additional non-pharmacological interventions, hence broadening the scope of T2T.References:[1]Pazmino S, et al. Does Including Pain, Fatigue, and Physical Function When Assessing Patients with Early Rheumatoid Arthritis Provide a Comprehensive Picture of Disease Burden? J Rheumatol. 2020 Nov 15:jrheum.200758. doi: 10.3899/jrheum.200758. Ahead of print.Disclosure of Interests:None declared