Background/Purpose: Concordant 0-10 estimates of global status by physicians (DOCGL) and patients (PATGL) are associated with greater expectations for improvement and better outcomes¹, and appear desirable in shared decisions concerning therapy. However, previous studies have shown that physicians and their patients with rheumatoid arthritis (RA) and other rheumatic conditions have discordant global estimates in about 50% of instances^2,3. Although recent evidence suggests that disease burden RA is similar or lesser than in osteoarthritis OA, OA generally remains regarded as being less severe than RA. We analyzed discordance between PATGL and DOCGL in OA vs RA patients seen in routine care at an academic rheumatology setting.

Methods: All patients in this setting complete a multidimensional health assessment questionnaire / routine assessment of patient index data (MDHAQ/RAPID3) at each visit, which includes 0-10 scores for physical function, pain, and PATGL compiled into a 0-30 RAPID3, as well as fatigue VAS, 60 symptom checklist, RADAI self-report joint count, and demographic data. Rheumatologists complete a RheuMetric checklist, which includes a physician global (DOCGL). Mean differences between PATGL & DOCGL were calculated. Patients with primary OA or RA were classified into 3 categories: PATGL ≥ DOCGL by 2/10 units, PATGL = DOCGL, and DOCGL ≥ PATGL by 2 units, and compared with other measures using chi-square tests, ANOVA, and logistic regression.

Results: The study included 216 patients with RA and 243 with OA. Patients with OA were older and showed higher scores for PATGL, pain, fatigue, and number of symptoms (data not shown). A higher percentage of patients with OA showed discordance with PATGL>DOCGL in comparison with RA (34% vs 18%) while concordance and discordance with DOCGL > PATGL were higher in RA patients (67% vs 56% and 15% vs 10%) (p<0.001). In general, patients with PATGL>DOCGL had lower formal education levels (p<0.05 in RA but not OA), and had higher scores for pain, function, fatigue, number of symptoms (p<0.05 in OA but not RA), and self-report RADAI (Table). In logistic regressions, pain was the only significant predictor of discordance in both RA [odds ration (OR) 1.47, 95% confidence interval (CI) 1.12-1.93] and OA (OR 1.40, 95% CI 1.04-1.89), but not age, gender, race, education, function, symptom checklist or self-report RADAI.

Conclusion: Patients with OA are more likely to have their clinical status underestimated by rheumatologists than patients with RA, although the burden of disease was greater in OA than in RA. The data suggest that rheumatologists and the medical and general public might revise generally held views that OA is less severe than RA. References: 1. Am J Public Health 1981, 71(2):127-131. 2.Arthritis Care Res (Hoboken) 2014;66:934-42. 2. 3.Arthritis Care Res (Hoboken) 2012;64:206-14.

« Back to 2016 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/estimates-of-global-status-by-physicians-and-patients-are-more-likely-to-be-discordant-in-osteoarthritis-than-in-rheumatoid-arthritis/