Rheumatoid arthritis (RA) is an autoinflammatory disease, its core treatment principle is to achieve remission as soon as possible. There is no good prediction model that can accurately predict the remission rate of patients to choose a good treatment scheme. Here, we aimed to verify the prognostic value of some inflammatory indicators in RA and establish a prediction model to predict the remission rate after treatment. A total of 223 patients were enrolled at Qilu Hospital from June 2014 to June 2020. Baseline clinical data were collected and plasma was obtained to detect the inflammatory indicators. All patients were treated with conventional synthetic disease-modifying antirheumatic drugs (csDMARDs). All patients were followed up and were recorded the time to reach the disease activity score-28 with erythrocyte sedimentation rate (DAS28-ESR) of <2.6. A total of 156 patients were randomly assigned to the development cohort, and 67 patients were assigned to the validation cohort. Inflammatory indicators in plasma were detected by enzyme-linked immunosorbent assay (ELISA). The predictive factors were screeded by using least absolute shrinkage and selection operator (LASSO) and Cox regression. The model was created and verified by using the standard method. A total of 6 independent risk factors were analyzed to construct a nomogram to predict the remission rate in 3, 6 and 12 months. The remission rates after treatment in 3, 6 and 12 months were 38.76%, 58.91%, and 81.40%, respectively. Patient age, C-reactive protein (CRP), interleukin (IL)-6, galectin-9 (Gal-9), health assessment questionnaire (HAQ), and DAS28-ESR were included in the prognostic model to predict the remission rate. The resulting model had good discrimination ability in both the development cohort (C-index, 0.729) and the validation cohort (C-index, 0.710). Time-dependent receiver operating characteristic (ROC) curve, calibration analysis, and decision curve analysis (DCA) showed that the model has significant discriminant power and clinical practicability in predicting the remission rate. We established a new predictive model and validated it. The model can predict the remission rate in 3, 6 and 12 months after receiving csDMARDs treatment. By using this model, we can facilitate the identification of high-risk patients early and intervene with them as soon as possible.
Read full abstract