Background/Purpose: Rheumatoid arthritis (RA) is a chronic autoimmune disease associated with increased mortality, primarily due to systemic inflammation and comorbidities. While previous studies have identified some prognostic factors in RA, there remains a lack of robust, clinically applicable tools adapted to the Chinese population. This study aims to develop a predictive model for all-cause mortality in Chinese RA patients using routinely available clinical data.

Methods: The clinical data utilized in this study were obtained from the database of the Chinese Registry of Rheumatoid Arthritis (CREDIT), while the survival outcome data were from the Chinese Center for Disease Control and Prevention (CDC) from October 2003 to December 2021. Variable selection was performed using the least absolute shrinkage and selection operator (LASSO) and univariate Cox regression.A multivariate Cox proportional hazards model was constructed and calibrated. Model performance was assessed using Harrell’s C-index and Brier score. External validation was conducted using an independent data as validation set from central China provinces (Henan, Hubei, and Hunan) in China, and data from other provinces was randomly split into training and internal test set. A nomogram was developed based on the final multivariate model to facilitate clinical application.

Results: A total of 62,321 patients with complete clinical data (Mean age: 53.13 years, SD: 12.85 years; female: 49,790, 79.9%) from 450 medical centers across the country were included in this study. Among them, 1,901 deaths (3.0%) were recorded during follow-up. 45,126, 12,058, and 5,137 patients were included in training, validation, and test set, respectively. Assisted with LASSO method, a multivariate Cox model with sex, age, coronary artery disease (CAD), stroke, fragility fracture, interstitial lung disease (ILD), DAS28CRP was built and calibration curve showed great agreement between predicted and observed 3-year survival probabilities. The model demonstrated consistent discrimination across datasets, with Harrell’s C-indices of 0.841 (95% CI: 0.816–0.863) in the training set, 0.857 (0.779–0.911) in the test set, and 0.834 (0.784–0.874) in the validation set. Decision curve analysis indicated that the derived nomogram offers meaningful clinical benefit.

Conclusion: This study developed a robust prediction model for all-cause mortality in patients with RA using routinely collected baseline data. The model offers a practical tool for stratifying patients by prognostic risk, particularly through assessment of comorbidity burden.

« Back to ACR Convergence 2025

ACR Meeting Abstracts - https://acrabstracts.org/abstract/predicting-mortality-of-rheumatoid-arthritis-in-china-a-nation-wide-cohort-study/