Background/Purpose: A total of 98.8% of French citizens are registered in the National Health Data System (SNDS), which includes claims data of all reimbursed health-related expenditures such as medication, professional healthcare services, and eligibility for full reimbursement of care related to chronic diseases (named Long-Term Disease (LTD) status).The identification of rheumatoid arthritis (RA) patients in the SNDS is generally based on the presence of at least one RA-related ICD-10 code (International Classification of Diseases, 10th Revision) recorded during hospitalization and/or as part of an LTD status. Based on these criteria, annual mapping by the French health insurance system estimated an RA prevalence of 0.47% in the adult population in 2022. (1) However, inclusion based on a single RA code may lack specificity, potentially leading to an overestimation of RA prevalence. Conversely, RA patients who are neither hospitalized nor covered by LTD status may not be identifiable using these criteria.The PREST study aims to refine algorithms for improving the estimation of RA prevalence in France. Additionally, it seeks to analyze care pathways of identified patients and explore factors associated with disease management.

Methods: Data were extracted from the SNDS covering the period 2010–2022, using the following criteria: hospitalization records and/or LTD status with ICD-10 codes M05 or M06, and/or the dispensing of csDMARDs (conventional synthetic Disease-Modifying Antirheumatic Drugs) or b/tsDMARDs (biologic/targeted synthetic DMARDs). Different RA identification algorithms were applied including additional criteria based on disease-management pathways. Various combinations of these criteria were applied to define multiple subgroups. Corresponding RA prevalences were estimated for 2019.

Results: The data extraction identified 914,278 adult patients alive on January 1, 2019, who met at least one of the following criteria: RA LTD status (N&#3f281,441), hospitalization with RA ICD-10 code (N&#3f205,232), or treatment for RA without differential diagnosis (N&#3f390,564). Patients without RA coding but with a differential diagnosis were excluded. Among them, 355,350 met the standard definition corresponding to a prevalence of 0.70%. The new algorithms yielded highly variable prevalence estimates depending on the criteria. The most stringent combination (patients with M05 or M06 codes receiving RA treatment) identified 249,240 cases (0.49%). Broadening the criteria to include untreated patients with at least two RA ICD-10 codes identified 275,694 patients (0.54%). Including previously excluded patients who had diagnostic procedures associated with an RA code or treatment further increased the number of identified patients to 392,974 (0.77%).

Conclusion: Depending on the identification criteria, RA prevalence in the French adult population in 2019 ranges from 0.49% to 0.82%, with the true prevalence most likely falling within this range. A previous study, EPIRHUM, conducted a general population survey in 2000 and reported an underestimated prevalence of 0.31% (2). The significant variation in prevalence based on the identification criteria highlights the complexity of identifying RA in the SNDS.

« Back to ACR Convergence 2025

ACR Meeting Abstracts - https://acrabstracts.org/abstract/rheumatoid-arthritis-prevalence-estimation-in-france-using-care-pathways-in-the-national-health-data-system-opportunities-and-limitations-prest-study/