A Risk Prediction Model for the Development of Rheumatoid Arthritis, Sjögren's Syndrome, Systemic Sclerosis in Patients With Systemic Lupus Erythematosus: Multicenter Approaches

Abstract

Aim

The aim of this study was to construct a predictive model to evaluate the risk of overlapping syndrome (OS), including the combination of systemic sclerosis, rheumatoid arthritis, and Sjögren's syndrome in patients with systemic lupus erythematosus (SLE) from multicenters.

Methods

This study included SLE patients in the development cohort (n = 4714) and the external validation cohort (n = 2271). SLE patients in the development cohort were randomly divided into the training cohort and the internal validation cohort at a 7:3 ratio. Laboratory variables were collected and compared by univariate logistic regression modeling. Multivariate logistic regression was further used to develop predictive models and plot a risk nomogram for OS. Receiver operating characteristic (ROC) and calibration curve analysis were used to assess model validity and accuracy. Decision curve analysis was used to assess net clinical benefit.

Results

Indexes including anti-SSA antibody, anti-SSB antibody, proteinuria, occult blood in urine, age, eosinophil ratio, hematocrit, platelet, direct bilirubin, indirect bilirubin, rheumatoid factor, immunoglobulin A, prothrombin time, and ferritin were included in the predictive model. The nomogram showed a wide range of predictive ability. The area under the curve (AUC) of the ROC curve for the training cohort was 0.874 (0.851–0.896), the AUC for the internal validation cohort was 0.877 (0.843–0.911), and the AUC for the external validation cohort was 0.760 (0.730–0.790).

Conclusion

The model has a good predictive performance and will be clinically valuable for the assessment of the risk of OS in SLE.