SAT0624 VALIDATION PROCESS OF CASES OF RHEUMATOID ARTHRITIS IN A LARGE PROSPECTIVE COHORT OF FRENCH ADULT WOMEN: THE E3N COHORT

Background Rheumatoid arthritis (RA) is a complex multifactorial autoimmune disease in which genetic and environmental factors interact in the pathogenesis of the disease to trigger auto-immunity. Except for tobacco smoking, the role of environmental factors has been suggested yet poorly investigated, and results were rarely reproducible. More observational studies are requested to address the question. Cohort studies offer the advantage over case-control studies of having a prospective collection of environmental factors before disease onset, thus avoiding recall bias. However, collected information about disease phenotypes is usually limited, and a rigorous process of case validation is needed. Objectives To detect RA cases in a large prospective cohort of healthy French adult women and to assess the performance of the validation methods. Methods The French E3N cohort included 98,995 healthy women prospectively followed since 1990. Self-administered questionnaires were sent every 2–3 years to collect medical events, and general, lifestyle, and environmental characteristics. Potential cases of inflammatory rheumatic diseases (IRD), including RA cases, were identified through self-reports in three consecutive questionnaires. Self-reported RA cases were validated with two methods including sending of a specific validation questionnaire and the use of the reimbursement database. The sensitivities and specificities of each method were calculated using as a reference the analysis of available medical records reviewed by a panel of expert rheumatologists. Results Among the 3,192 identified potential IRD cases, 964 RA cases were validated, including 698 incident cases and 266 prevalent cases. Of them, 314 (32.6%) were seropositive, 23 (2.4%) seronegative and 627 (65.0%) had unknown antibody status. Mean age at diagnosis was 57.4 ± 13.9 years (40.9 ± 10.4 years for prevalent cases, and 63.8 ± 9.0 for incident cases). Sensitivities and specificities of the validation methods were 92.2 and 83.7% for the specific validation questionnaire and 69.8 and 97.0% for the reimbursement database. Conclusion This study enabled us to detect a large number of RA cases in a large general population prospective cohort of women with acceptable sensitivity and specificity. This will allow investigating a large number of potential endogenous and exogenous risk factors of RA in women. Acknowledgement We would like to thank the Foundation for Research in Rheumatology (FOREUM) and the French Society for Rheumatology (SFR) for their financial support. Disclosure of Interests Yann Nguyen: None declared, Carine Salliot: None declared, Gaelle Gusto: None declared, Elise Descamps: None declared, Xavier Mariette Grant/research support from: Servier, Consultant for: AstraZeneca, Bristol-Myers Squibb, GlaxoSmithKline, Janssen, Pfizer, UCB Pharma, Marie-Christine Boutron-Ruault: None declared, Raphaele Seror: None declared