The prevalence and types of discordance between physician perception and objective data from standardized measures of rheumatoid arthritis disease activity in real-world clinical practice in the US

Abstract

Background: Heterogeneity in assessments of rheumatoid arthritis (RA) disease remission, based on physician judgment and patient self-reports versus standardized measures, have previously been reported. This study explored the prevalence and types of discordance between physician perception versus objective data of RA disease activity in real-world clinical practice in the US. Methods: Data were from the Adelphi RA Disease Specific Programme (DSP; January to March 2014), a cross-sectional survey of US rheumatologists and their patients. RA remission based on physician judgment versus Disease Activity Score in 28 joints (3)-erythrocyte sedimentation rate (DAS28(3)-ESR) and Clinical Disease Activity Index (CDAI) scores were compared using descriptive analyses; patient and physician factors associated with discordance were identified using bivariate and multivariate analyses. Results: Of 101 rheumatologists participating (completing patient-record forms for 843 patients), 56.4% based assessment of remission on clinical judgment alone. Of 531 patients eligible for the discordance analysis, 49.7% were in remission based on rheumatologists' evaluation, and 30.7% were eligible based on DAS28(3)-ESR. Compared with DAS28(3)-ESR criteria, 25.8% of patients' disease remission was negatively discordant (overestimated remission) based on clinical perception. These patients were mostly administered biologic disease-modifying antirheumatic drugs and were without a treat-to-target strategy followed by their rheumatologist (P < 0.05). These patients were also more likely to have experienced a higher level of pain as well as increased joint inflammation and damage (e.g. destruction of cartilage, thinning of bone, and/or synovium inflammation) compared with concordant patients (P < 0.005). Conversely, 6.8% of rheumatologists were positively discordant (under estimated remission) versus the DAS28(3)-ESR. Sensitivity analysis indicated different levels of discordance using CDAI, with 35.6% negative discordance and 1.3% positive discordance of rheumatologist-assessed disease remission compared with objective data. Conclusion: There is discordance between RA remission as assessed by rheumatologist perception versus standardized measures among those in the US DSP sample. Our study identified the factors associated with the discordance which may inform strategies to enhance assessments of RA disease remission.