Background Discrepancy between patient (PTGA) and physician (MDGA) global assessments in RA can adversely affect therapeutic decisions in many cases. Understanding the factors influencing this discrepancy may improve clinical outcomes.

Objectives The purpose of this study is to evaluate clinical factors associated with PTGA-MDGA discrepancies.

Methods Patients with RA were recruited from the Canadian Early Arthritis Cohort (CATCH) a prospective cohort where data is collected according to a standardized protocol. CATCH patients were considered for this analysis if they initiated MTX at baseline, were biologic naïve and had ≥6 months follow up. PTGA and MDGA were scored out of 100. PTGA-MDGA discrepancy was calculated by subtracting MDGA from PTGA at baseline. A clinically meaningful discrepancy was considered a difference of ≥30. Spearman correlations were used between MDGA, PTGA and other clinical factors. Linear regression analysis was used to evaluate factors associated with the PTGA-MDGA discrepancy when adjusted for potential confounders. To address the role of clinicians we included CATCH recruiting “site” as one of the predictors. Sites with more than 25 patients were considered for analysis.

Results Baseline characteristics of the 953 RA patients who met inclusion criteria for this study included:71% female, mean age 53, disease duration 0.5 years, TJC 9.6, SJC 8.5 (of 28), DAS28 5.2, ESR 28.6, CRP 15.0 mg/L, PTGA 58.7 and MDGA 51.6.Spearman rank correlations revealed that the MDGA was moderately to highly correlated with the TJC, SJC and patient pain (r=0.43-0.56). PTGA was weakly correlated with the TJC and SJC (r=0.29 and 0.37); PTGA was highly correlated with pain (r=0.84). Using Spearman correlations there was a lower correlation between PTGA and MDGA when patient pain score and SJC were higher. Further analysis was performed for 820 patients who were recruited from sites with more than 25 patients and revealed a clinically relevant discrepancy in 324 (36%) patients. Of these 111 patients (34%) exhibited a negative discrepancy (MDGA>PTGA) while 213 (66%) exhibited a positive discrepancy (PTGA>MDGA). Differences in baseline characteristics between patients with and without a discrepancy were accounted by the TJC, SJC and pain scores (p<0.05). Multivariate linear regression analyses to identify factors associated with discordance between PTGA and MDGA were adjusted for demographics, clinical factors and treating “site”s, stress (yes/no) and depression (yes/no). The results showed that the PTGA-MDGA discrepancy was significantly influenced by SJC, pain, ESR and gender (male) (p<0.05). The patient pain and SJC were two strong predictors with higher SJC and pain associated with higher discrepancy.

Conclusions In patients with early RA the pain score and SJC have a significant influence on PTGA-MDGA discrepancy. Previous studies have emphasized the pain score as a major factor affecting the PTGA-MDGA discrepancy, we have also demonstrated a significant influence of the SJC. Further studies of early RA cohorts are necessary to validate our findings.

Disclosure of Interest P. Akhavan: None Declared, V. Bykerk Grant/Research support from: CATCH is supported by unrestricted research grants from Amgen, Pfizer, Abbott, BMS, Janssen, Y. Sun: None Declared, M. Choi: None Declared, J. Pope: None Declared, C. Hitchon: None Declared, G. Boire: None Declared, B. Haraoui: None Declared, C. Thorne: None Declared, D. Tin: None Declared, E. Keystone: None Declared