SAT0072 Performance of Patient Reported Outcomes in the Assessment of Rheumatoid Arthritis Disease Activity: The Experience of the Espoir Cohort

Background Activity of rheumatoid arthritis (RA) can be assessed by several outcome measures: joint counts, inflammatory syndrom but also auto-questionnaires such as global patient assessment. The importance of Patient Reported Outcomes (PROs) has been recently put forward. Objectives To determine whether patient self-assessment can reflect RA disease activity. Methods Data from patients included in the early arthritis ESPOIR cohort and fulfilling ACR/EULAR 2010 RA criteria at month 12 were used. The following PROs (Visual Analog Scales for fatigue, pain, patient assessment of disease activity; HAQ; SF36; EMIR-court and RAPID3) and their association with disease activity assessed by DAS28-3 variables, were measured. DAS28-3 variables (including tender joint count, swollen joint count and erythrocyte sedimentation rate) was preferred to DAS28-4 variables in order to limit the input of subjective and patient-derived appreciation when defining the level of disease activity. We compared the influence of disease characteristics on each PRO by Student t tests. The association between PROs and disease activity was assessed in several ways: part of explained variance, correlation and performance of each PRO in differentiating low versus higher disease activity states. Results 677 patients (83.2%) of the 813 patients of the ESPOIR cohort, were analyzed as they respondend to the ACR/EULAR 2010 criteria of RA at month 12.Their mean age was 48.6±12.3 ans with 77.4% of female, with a mean duration of disease of 3.41±1.74 months, 60.5% were rheumatoid factor or anti-citrullinated protein positives, 16.4% had anxiodepressive disorder and 15.1% had radiological lesions according to ACR/EULAR criteria. Disease activity assessed by DAS28-3 avriables was, at inclusion, 5.0±1.2, with a decrease during the follow-up. Whatever the status of disease activity, patients with the lowest disease activity always had significantly less impaired PROs. All analyzed PROs showed moderate correlations with RA disease activity. The PRO showing the best association with DAS28-3v in determining RA disease activity states, was RAPID3 (Pearson correlation coefficient between 0.45-0.55, explained variance between 30-45%, sensitivity between 69-100% and specificity between 55-78%). PROs with the highest association with disease activity were global PROs (RAPID3, EMIR-court) followed by those assessing physical function. Conclusions Association between PROs and RA disease activity as measured by DAS28-3 variables remains moderate. RAPID3, a global PRO, showed the best association with disease activity compared to the other analyzed PROs. Evaluation of disease activity with only an auto-questionnaire seems to be difficult. Disclosure of Interest None declared