Rheumatoid arthritis is a chronic autoimmune condition in which the immune system erroneously attacks the synovial lining of joints, producing inflammation, swelling, pain, and progressive structural damage. Modern management follows a strategy known as Treat to Target, which aims to drive the disease into remission or low disease activity using composite scoring instruments such as the Clinical Disease Activity Index and the Simplified Disease Activity Index. These indices combine objective clinical measures, including the 28-joint tender and swollen joint counts, with two subjective global assessments scored from zero to ten, where zero represents very good and ten represents very poor. One of these is the patient global assessment, a patient-reported outcome capturing how the individual perceives their arthritis overall. The other is the physician global assessment, reflecting the clinician’s expert judgment of the same disease.

The patient global assessment is far more than a formality. It is a widely used patient-reported outcome measure that offers valuable insight into the experiential burden of the disease as the patient integrates pain, fatigue, sleep disturbance, mental health status, and functional limitation into a single number. Yet this same subjectivity means the score can be shaped by factors that have little to do with active joint inflammation. Socio-demographic characteristics, cultural context, mood disorders, and pain sensitization all contribute to how patients rate their condition. When the patient’s number diverges substantially from the physician’s, the resulting discordance reflects a multidimensional misalignment in disease perception rather than a simple arithmetic error.

Researchers from the Division of Rheumatology at Groote Schuur Hospital and the University of Cape Town set out to quantify this misalignment in a South African cohort, a population in which such research had been notably absent. Language barriers, cultural beliefs, socioeconomic disparities, and uneven healthcare access can all influence communication and understanding between patients and physicians. Because South Africa is extraordinarily diverse, studying patient-physician discordance in this setting may illuminate challenges that are specific to the local context and guide the development of culturally sensitive interventions. The study enrolled 550 consenting adults meeting the EULAR/ACR 2010 rheumatoid arthritis classification criteria, collecting demographic details, disease-specific information, joint counts, and a battery of patient-reported outcome measures including the EuroQol EQ-5D, the Health Assessment Questionnaire-Disability Index, the Hospital Anxiety and Depression Scale, the FACIT-Fatigue Scale, and the Brief Pain Inventory.

The cohort was predominantly female, at 84.9 percent, with a mean age of 55.8 years and a mean disease duration of 11.5 years. A significant proportion, 67.4 percent, were classified as living in poor socioeconomic settings using a pooled index that considered education level, household size, and access to domestic washing machines. Most patients had at least one comorbidity, and 91.6 percent were seropositive for rheumatoid factor or anti-cyclic citrullinated protein antibodies. The mean patient global assessment was 4.9, notably higher than the mean physician global assessment of 3.3, already hinting at the perceptual gap that the study would go on to dissect in detail.

To measure discordance, the researchers subtracted the physician global assessment from the patient global assessment, defining discordance as a difference of 2.5 or more on the zero-to-ten scale, a threshold consistent with published ranges of 2.0 to 3.0. Fifteen patients, or 2.7 percent, rated their disease lower than their physician and were excluded from the primary analysis due to insufficient numbers. Among the remaining patients, 27.5 percent were discordant, with the overwhelming majority rating their arthritis worse than their clinician did. This rate is somewhat lower than figures reported in other regions, where up to half of patients may score their disease differently from their physicians, but it still represents a considerable proportion of the cohort.

The statistical architecture of the study revealed a striking asymmetry in what drives each score. Using Spearman correlation and multivariable linear stepwise regression, the researchers found that pain severity, measured by the Brief Pain Inventory pain severity score, was the strongest determinant of the patient global assessment, with a correlation coefficient of 0.50 and a regression contribution explaining 25.7 percent of the variance on its own. Additional contributors included the tender joint count, the swollen joint count, problems with usual activities on the EQ-5D, prednisone use, and fibromyalgia. Together these variables accounted for 38.3 percent of the variance in the patient global assessment, leaving a remarkable 61.7 percent unexplained. In contrast, the physician global assessment was predominantly determined by the swollen joint count, which alone explained 44.9 percent of its variance, with additional contributions from the tender joint count, fatigue, prednisone use, and functional disability. Even so, 48.5 percent of the physician’s score remained unexplained by the measured variables.

This asymmetry is the crux of the discordance phenomenon. Physicians, trained to emphasize objective markers of inflammation such as the swollen joint count and serum acute phase reactants, anchor their global assessment in what they can palpate and measure. Patients, by contrast, integrate the full experiential weight of their disease, including pain, fatigue, mood, and functional limitation, into their rating. The regression findings align with studies conducted in various regions, confirming that patients perceive their disease through the lens of its impact on daily life, while physicians focus on objective signs of joint inflammation. The large unexplained variance in both scores suggests that individual perception, expectation, social support, and unmeasured disease manifestations contribute substantially to how each party rates the condition.

When the researchers compared discordant patients with concordant ones, a paradoxical picture emerged. Discordant patients had significantly lower tender and swollen joint counts, at 1.9 and 1.3 respectively, compared with 3.5 and 3.5 in concordant patients, and were more likely to be in remission or low disease activity. Yet they reported higher pain severity scores, worse EQ-5D level sum scores, more problems with usual activities, and a higher prevalence of fibromyalgia, at 12.5 percent versus 5.5 percent. Binomial logistic regression confirmed that lower swollen joint count, higher pain severity, and problems with usual activities independently predicted discordance. In other words, patients were rating their disease worse precisely when objective inflammation was lowest, suggesting that pain and functional limitation, rather than active synovitis, were driving their elevated scores.

This observation carries direct relevance for international remission criteria. The 2022 revision of the ACR/EULAR Boolean remission criteria raised the patient global assessment threshold to two or below, explicitly acknowledging that non-inflammatory factors such as pain and fatigue can raise the score and hinder remission classification despite controlled inflammation. The South African findings provide compelling support for this revision, demonstrating that discordant patients with persistently elevated global assessments had objectively lower joint counts and were more frequently in remission or low disease activity by CDAI. The authors caution that CDAI inherently includes both global assessments, creating a degree of circularity, but the fact that discordant patients still more frequently fell into remission or low disease activity categories, despite contributing higher patient scores, strikingly emphasizes the absence of objective inflammation in this group.

Notably, mental health scores were not associated with discordance in this cohort, even though a third of patients had significant anxiety or depression. This contrasts with findings from other studies, such as work by Barton and colleagues, where higher depressive symptoms were linked to discordance. The divergence may reflect cohort-specific factors, cultural expression of distress, or the cross-sectional design’s inability to capture mood fluctuations over time. Fibromyalgia, recorded as a pre-existing comorbidity rather than actively screened with formal diagnostic criteria, may have been underestimated in prevalence, and its association with discordance suggests that centralized pain processing plays a role in how patients rate their disease independent of joint inflammation.

The clinical implications are substantial. Discordance has been shown to negatively affect medical care, with poor treatment adherence, higher dissatisfaction, increased risk of joint damage, impaired health-related quality of life, and decreased work productivity. Addressing pain and activity limitations, particularly in patients with low joint counts, may reduce discordance and foster shared decision-making. The authors suggest that clinical review with musculoskeletal ultrasound, pain control strategies, occupational and physiotherapy assessment, and exercise may help narrow the gap. Notably, 89.1 percent of patients were receiving prednisone, most at low doses of five milligrams per day or less, an area the authors flag for improvement given that most patients were in remission or low disease activity and that corticosteroid toxicity requires education, particularly among discordant patients with low disease activity who need better pain management rather than more immunosuppression.

The study’s limitations are acknowledged by the authors. The cross-sectional design restricts the ability to establish causal relationships or capture disease activity fluctuations over time. Acute phase reactants, including C-reactive protein and erythrocyte sedimentation rate, were not measured as a cost-saving measure consistent with practice in resource-constrained settings, preventing the researchers from distinguishing pain-driven discordance from discordance attributable to residual subclinical inflammation not captured by joint counts. Musculoskeletal ultrasound synovitis scoring and radiographic erosion assessment were also not performed, both of which may contribute to the unexplained variance in the patient global assessment and are recommended for future prospective studies. The large proportion of unexplained variance, at 61.7 percent for the patient global assessment, highlights the subjective nature of the measure and the need for further research incorporating structural damage, health literacy, patient expectations, and social support.

Negative discordance, observed in fifteen patients who rated their disease lower than their physician, represents a distinct phenomenon warranting dedicated investigation in future studies. The authors call for longitudinal follow-up and qualitative interviews to better understand the variability in patient global assessments and to capture the full range of factors that shape how patients perceive their disease. Such research may allow a more comprehensive understanding of the patient’s perspective and inform the development of targeted interventions that address pain, fatigue, and functional limitation alongside inflammation.

In conclusion, this South African study demonstrates that over a quarter of patients with longstanding rheumatoid arthritis rate their disease activity higher than their clinicians, particularly when objective disease activity is low. Pain severity and functional limitations play a crucial role in shaping patient perceptions, and addressing these factors may reduce discordance, optimize disease management, and foster genuine shared decision-making. The findings underscore that the patient global assessment is not a flawed mirror of inflammation but a valid reflection of the experiential burden of disease, one that clinicians must learn to interpret alongside joint counts and laboratory markers if the Treat to Target strategy is to serve patients fully in diverse and resource-constrained settings.

Subject of Research: Patient-physician discordance in rheumatoid arthritis assessments among South Africans

Article Title: Discordance between patient and physician global assessments amongst South Africans with rheumatoid arthritis

Article References: Didi, S. A., & Hodkinson, B. (2026). Discordance between patient and physician global assessments amongst South Africans with rheumatoid arthritis. Frontiers in Musculoskeletal Disorders, 4, Article 1882633. https://doi.org/10.3389/fmscd.2026.1882633

Image Credits: AI Generated

DOI: 10.3389/fmscd.2026.1882633

Keywords: rheumatoid arthritis, patient-physician discordance, pain, disease activity, South Africa, patient-reported outcomes, rheumatology, Discordance, between, patient, physician, global