Background Physician's global assessment of disease activity (PhGA) has been shown to be a determinant factor for therapeutic decisions in Rheumatoid Arthritis (RA). The adequacy of this assessment is, therefore crucial, to assure that the best options are taken in the perspective of the patient's interest.

Objectives To evaluate the determinants of PhGA, including clinical variables and patient reported outcomes.

Methods Consecutive RA patients followed in a Tertiary Rheumatology Department were included in this cross-sectional study. Patient demographics and clinical assessments were collected through a standardized protocol which included age, gender, disease duration, DAS(28)4v-CRP (and its individual measures), DAS(28)4v, pain (VAS 0–100mm), Patient Global Assessment (PGA) of disease activity (VAS 0–100mm) and function (Health Assessment Questionnaire (HAQ)). From physicians were collected PhGA (VAS 0–100 mm), summary score of the patient's mobility and deformity (VAS 0–100mm).

To identify independent predictors of PhGA, Pearson's Correlation Coefficient was used to select variables (p<0.05) to subsequent inclusion in multivariable linear regression (stepwise model).

Results 311 patients were included: 81.7% females, 60±12 years, 11±9 years of disease duration, 8±5 years of formal education and a mean DAS(28)4v-CRP=2.9±1.2. PhGA was statistically and strongly associated only with swollen joint count (SJC28) (r=0.71). Moderate correlations were observed with tender joint count (TJC28) (r=0.41), DAS(28)3v–CRP (r=0.655) and DAS(28)3v (r=0.550). Weak correlations (r<0.4) were observed with pain, PGA, C-reactive protein (CRP), erythrocyte sedimentation rate (ESR) and HAQ, as well as with mobility and deformity in the physician's perspective. In multivariable analysis, SJC28 (β=5.02, 95%CI: 4.29;5.73, p<0.001) and CRP (β=3.0, 95%CI: 1.56;4.46, p<0.001) were the most significant predictors of PhGA, with a small contribution of the physicians assessment of patient mobility.

Conclusions Physicians consider objective measures when assessing disease activity, apparently disregarding pain and the PGA. This may lead to suboptimal management approaches from the patient's perspective, especially regarding disease dimensions that are not directly affected by immunosuppressive therapy.

Disclosure of Interest None declared