Background Discordance between the patient global assessment (PGA) and physician global assessment (PhGA) is well reported in rheumatoid arthritis (RA) and can lead to suboptimal care^[1]. There are no published studies on discordance in South African (SA) patients.

Objectives To measure and explore patient-physician discordance amongst RA patients attending a tertiary SA hospital.

Methods Consenting adults (≥18 years) with RA completed demographic, clinical and patient-reported outcome measures including Health Assessment Questionnaire-Disability Index (HAQ-DI),FACIT fatigue scale, Hospital Anxiety and Depression Scale (HADs), Brief Pain Questionnaire-short form (BPI-SF), and EuroQOL five dimensions questionnaire (EQ5D). Poor socio-economic status (SES) was defined using a pooled index. The Clinical Disease Activity Index (CDAI) was calculated based on the swollen and tender joint counts (SJC and TJC), PGA and PhGA visual analogue scale. Patients were categorized into high, moderate or low disease activity (HDA, MDA and LDA) states. Discordance was calculated by subtracting the PhGA score from the PGA and positive discordance was defined as a difference of > 2.5 points. Determinants of PGA and PhGA were also analyzed.

Results Of 550 patients (467 females), the mean (SD) age and disease duration were 55.8 (12.9) and 10.5 (9.7) years, most were in LDA (47.8%), and 371 (67.4%) had poor SES. Positive discordance was seen in 136 patients (24.7%). Fifteen patients with negative discordance (PGA- PhGA < - 2.5) were excluded from this analysis. Comparing concordant and positive discordant patients, there were no significant differences in age, sex, disease duration, SES, RA therapy or number of comorbidities, nor in fatigue, HAD-DI, or HADs. The discordant group had significantly lower SJC, TJC and PhGA. More patients in the discordant group were in LDA (58.1% vs. 45.4%), with fewer in MDA (32.4 % vs. 29.1 %) and HDA (9.6% vs. 25.6%), p = 0.001. In addition, the discordant group included more fibromyalgia (FMS) patients, reported higher pain severity score and a higher EQ5D level sum score, with problems with usual activities being the only statistically significant domain (p<0.001). In logistic regression analysis, pain severity score and problems with usual activities on EQ5D were predictors of positive discordance. Pain was the main determinant of PGA (R² 0.24, p<0.001) and SJC was the main determinant of PhGA (R² 0.44, p<0.001).

Conclusion In this cohort of indigent SA RA patients, patient-physician discordance was found in a quarter of participants, and was noted in the LDA states, FMS and in patients with high pain scores and those who have problems with usual activities. We have shown, similar to studies elsewhere, that patients and clinicians focus on different aspects of disease when making their global assessments. Studies to understand the impact of discordance and the role of interventions, including pain management and occupational therapy are planned.

Reference [1]Desthieux C et al. Arthritis Care Res. 2016

Acknowledgements: NIL.

Disclosure of Interests None Declared.