Background/Purpose: In recent years there is increasing evidence of the relevance of including the assessment of the different aspects of rheumatoid arthritis (RA) from the patient’s perspective through patient reported outcomes (PRO). Previous studies showed discrepancies in the disease global disease assessment reported by patients (PtGA) or physicians (PhGA).

Our aim was to evaluate differences between PtGA and PhGA and analyze the variables related to these differences in an early arthritis cohort.

Methods: Cross-sectional study analyzing data from the baseline visit of patients included in the PEARL study (Princess Early Arthritis Register Longitudinal Study) in which demographic, laboratory and clinical characteristics including PtGA and PhGA (0-100 mm) are systematically collected.

The main variable was the difference between PtGA and PhGA (ΔGA). The descriptive analysis was performed using the Kruskal-Wallis, Mann-Whitney or Pearson correlation tests as appropriate. Stadistical analysis was perfomed with STATA 14.

A multivariate linear regression model was developed with ΔGA as a dependent variable. All those predictors available at the baseline visit reaching a p < 0.15 in the univariate analysis were included in the initial model. The final one was obtained through the progressive elimination of those variables not showing an improvement in the model as assessed by the adjusted R2 parameter. To categorize the dependent variable, we considered relevant differences between PtGA and PhGA greater than 5 mm.

Results: 530 patients were included, 422 (79.6%) were women with a mean age of 55.3 +/- 16 years, 21.2% current smokers, 54% and 50.4% rheumatoid factor and anti-CCP positive respectively. A 43.3% had moderate activity and 33.6% high, measured by DAS28-VSG.

The median of ΔGA was 4, (interquartile range; -10 to 18; see figure1). In 22% of the cases patients and physicians scored the same, in 46.5% the patients’ scores were higher and physicians’ were so in 31.5% of the cases.

The variables that explained DGA after adjusting the multivariate model were pain [β=0.36 (95%CI 0.28 to 0.44)], number of swollen joints [β= -3.19 (95%CI -3.7 to -2.7)] and ESR [β=-0.11 (95%CI -0.2 to -0.03)]. Pain had a greater influence on patients’ opinion while the number of swollen joints and ESR were more relevant for physicians. Other variables such as race, marital status, profession, sex, smoking, seropositivity or disease activity were not relevant in the prediction of ΔGA.

Conclusion: In our sample, disagreements between PtGA and PhGA were observed. Patients scored higher based on painful perception and physicians based on objective evidence of inflammation.

Figure 1. Representation of DGA distribution. DGA: difference between PtGA and PhGA; IQR: interquartile range.

« Back to ACR Convergence 2020

ACR Meeting Abstracts - https://acrabstracts.org/abstract/early-arthritis-global-assesment-what-differences-exist-between-patients-and-physicians/