Patterns and Determinants of Time to First DMARD Prescription in US Adults with Rheumatoid Arthritis: A Retrospective Cohort Study Using a Large Electronic Health Record Database.

introductionTimely initiation of disease-modifying antirheumatic drugs (DMARDs) is critical for achieving disease control, preventing irreversible joint damage, and improving the long-term outcomes of rheumatoid arthritis (RA). Despite guideline recommendations, real-world delays remain common and may be influenced by demographics, clinical, and geographical factors. Therefore, this study aimed to evaluate initiation patterns and identify demographic, clinical, and regional predictors associated with time to first DMARD initiation and assess differences in timing by drug class and age group among adults with RA in a large US electronic health record (EHR) cohort.

methodsA retrospective cohort study was conducted using EHR data of adults with incident RA diagnosed between November 2012 and August 2024. The primary outcome was the time from confirmed RA diagnosis to the first DMARD prescription. Covariates included age, sex, race and ethnicity, marital status, geographic location, and comorbidity. Descriptive analyses of initiation patterns were performed using Kaplan-Meier curves. Furthermore, multivariable Cox proportional hazards ratio (HR) regression models along with 95% confidence intervals (CI) were used to evaluate factors associated with time to DMARD initiation.

resultsAmong 197,107 patients with incident RA, 57.53% initiated DMARD within the first 3 months of diagnosis, with a mean delay of 18.18 months (SD 41.8); 28.6% initiated treatment after 12 months of diagnosis. Compared with the age group 18-29 years, initiation rates increased across older age groups, reaching an HR of 1.390 (95% CI 1.340-1.442) among patients aged ≥ 80 years. Furthermore, higher comorbidity burden was associated with a faster initiation (HR 1.045, 95% CI 1.042-1.047). Female sex was associated with slower initiation than male patients (HR 0.945, 95% CI 0.934-0.955). American Indian/Alaska Native patients showed 22% faster initiation compared with white patients (HR 1.292, 95% CI 1.219-1.369; p < 0.001). Moreover, longer delays were observed among patients in the Northeast and South than in the West (19% and 4% lower, respectively). Patients who were initiated on biologic DMARDs (bDMARDs) or targeted synthetic DMARDs (tsDMARDs) experienced a delay compared to conventional synthetic DMARDs (csDMARDs) initiation (bDMARD: HR 0.807, 95% CI 0.798-0.817; tsDMARD: HR 0.732, 95% CI 0.711-0.754).

conclusionIn this real-world study, patients with RA experienced a substantial delay in DMARD initiation, despite the availability of effective therapies and compelling evidence supporting early treatment. Furthermore, demographic, clinical, geographic, and treatment class characteristics were independently associated with the timing of DMARD initiation. These findings emphasize the need for targeted interventions to address treatment delays and ensure equitable access to early and effective RA therapy.