Introduction: To understand factors leading to biologic switches and to develop a readily usable model with data collected in clinical care at preceding visits, with the overall aim to predict the probability of switching biologic at a subsequent clinic visit in patients with rheumatoid arthritis (RA).

Methods: Participants were adults with RA participating in the CorEvitas RA registry. The study matched patients who switched biologics or targeted synthetic disease-modifying anti-rheumatic drugs (tsDMARDs) with control patients who had not switched biologics/tsDMARDs; the cohort was divided into a training and test set for prediction model development and validation. Using the training set, the best subset regression, lasso, and elastic net methods were used to determine the best potential models. Area under the ROC curve (AUC) was used for the final selection of the best model, and estimated coefficients of this model were applied to the test dataset to predict switching.

Results: A total of 5050 patients were included, of whom 3016 were in the training set and 2034 were in the test dataset. The average age was 59.6 years, the majority were female (3998, 79.2%), and the average duration of RA at the time of switch or control visit was 12.8 years. The final model included prior Clinical Disease Activity Index (CDAI) by category, prior patient pain measurement, change in CDAI from baseline, age group, and number of prior biologics, all of which were significantly associated with switching biologics. The AUC was 0.690 for this model with the training dataset. The model was then applied to the test data with similar performance; the AUC was 0.687.

Conclusion: We have developed a simple model to determine the probability of switching biologics for RA at the following clinic visit. This model could be used in practice to provide clinicians with more information about their patient's trajectory and likelihood of switching to a new biologic.