Predicting the progression of difficult-to-treat rheumatoid arthritis by a machine learning scoring system, from the FIRST registry
Authors' conclusion
This study developed and validated a scoring system to predict D2T RA progression and identify patients at high risk and suggested the advantages of IL-6Ri.
Study context
- Population
- Patients with RA in the FIRST registry initiating their first b/tsDMARD
- Participants
- 1,221 patients
- Outcome
- Progression to difficult-to-treat RA (D2T RA)
Methodological detailsmulticentre registry, machine learning, internal validation, 25 prediction models generated using 43 baseline characteristics
- Multicentre registry
- Machine learning
- Internal validation
- 25 prediction models generated using 43 baseline characteristics
Performance
As reported in the abstract
Progression to D2T RA; model performance
Progressed to D2T RA
193
15.8%
Median time to D2T RA (months)
54.0
Best model: Lasso logistic regression, AUC
0.71
Further findings
| Outcome | RR |
|---|---|
| D2T RA progression with IL-6R inhibitors in patients at high risk by the scoring system (subgroup)Significantly reduced D2T RA progression | 0.48 |
Baseline features of patients progressing to D2T RA
- Higher tender joint count, global assessment, pain scale, HAQ score and CDAI score, and more frequent coexisting lung disease
Original publication
Predicting the progression of difficult-to-treat rheumatoid arthritis by a machine learning scoring system, from the FIRST registry. RMD Open. Published 16 January 2026. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →