ABSTRACT
Abstract Background Early rheumatoid arthritis (RA) offers an opportunity for better treatment outcomes. In real-life settings, grasping this opportunity might depend on access to specialized care. We evaluated the effects of early versus late assessment by the rheumatologist on the diagnosis, treatment initiation and long-term outcomes of RA under real-life conditions. Methods Adults meeting the ACR/EULAR (2010) or ARA (1987) criteria for RA were included. Structured interviews were conducted. The specialized assessment was deemed "early" when the rheumatologist was the first or second physician consulted after symptoms onset, and "late" when performed afterwards. Delays in RA diagnosis and treatment were inquired. Disease activity (DAS28-CRP) and physical function (HAQ-DI) were evaluated. Student's t, Mann-Whitney U, chi-squared and correlation tests, and multiple linear regression were performed. For sensitivity analysis, a propensity score-matched subsample of early- vs. late-assessed participants was derived based on logistic regression. The study received ethical approval; all participants signed informed consent. Results We included 1057 participants (89.4% female, 56.5% white); mean (SD) age: 56.9 (11.5) years; disease duration: 173.1 (114.5) months. Median (IQR) delays from symptoms onset to both RA diagnosis and initial treatment coincided: 12 (6-36) months, with no significant delay between diagnosis and treatment. Most participants (64.6%) first sought a general practitioner. Notwithstanding, 80.7% had the diagnosis established only by the rheumatologist. Only a minority (28.7%) attained early RA treatment (≤ 6 months of symptoms). Diagnostic and treatment delays were strongly correlated (rho 0.816; p < 0.001). The chances of missing early treatment more than doubled when the assessment by the rheumatologist was belated (OR 2.77; 95% CI: 1.93, 3.97). After long disease duration, late-assessed participants still presented lower chances of remission/low disease activity (OR 0.74; 95% CI: 0.55, 0.99), while the early-assessed ones showed better DAS28-CRP and HAQ-DI scores (difference in means [95% CI]: −0.25 [−0.46, −0.04] and − 0.196 [−0.306, −0.087] respectively). The results in the propensity-score matched subsample confirmed those observed in the original (whole) sample. Conclusions Early diagnosis and treatment initiation in patients with RA was critically dependent on early access to the rheumatologist; late specialized assessment was associated with worse long-term clinical outcomes.