Lupus & antiphospholipid syndrome
Systemic lupus erythematosus damage risk index (SLE-DRI): a simple machine learning-based tool for identifying patients at risk for early organ damage
Authors' conclusion
We developed and validated a clinician-friendly model for early organ damage prediction in SLE, facilitating risk stratification.
Study context
- Population
- Patients with systemic lupus erythematosus (SLE); derivation cohort of 914 and external cohortexternal cohort n=50 · external cohort 50
- Participants
- 914 patients
- Comparison
- Least absolute shrinkage and selection operator-logistic regression (LASSO-LR) model vs ACR-1997, SLICC-2012 and European League Against Rheumatism (EULAR)/ACR-2019 classification systems for predicting early organ damage
- Outcome
- Damage (SLICC/ACR Damage Index (SDI)) within 5 years since diagnosis
- Follow-up
- 5 years since diagnosis
Methodological detailsmachine learning, LASSO-LR, 10-fold cross-validation, external validation
- Machine learning
- LASSO-LR
- 10-fold cross-validation
- External validation
Performance
As reported in the abstract
Prediction of early organ damage, AUC
| Test or model | AUC derivation cohort | AUC external validation cohort | AUC | Specificity (binary classifier) |
|---|---|---|---|---|
| LASSO-LR model | 0.80 (95% CI 0.74 to 0.87) | 0.86 (95% CI 0.76 to 0.97) | — not stated in the abstract | 0.90, 95% CI 0.78 to 0.95 |
| ACR-1997 | — not stated in the abstract | — not stated in the abstract | 0.70; 95% CI 0.62 to 0.78 | — not stated in the abstract |
| SLICC-2012 | — not stated in the abstract | — not stated in the abstract | 0.74; 95% CI 0.66 to 0.81 | — not stated in the abstract |
| EULAR/ACR-2019 | — not stated in the abstract | — not stated in the abstract | 0.70; 95% CI 0.63 to 0.78 | — not stated in the abstract |
AUC derivation cohort
LASSO-LR model0.80 (95% CI 0.74 to 0.87)
ACR-1997— not stated in the abstract
SLICC-2012— not stated in the abstract
EULAR/ACR-2019— not stated in the abstract
AUC external validation cohort
LASSO-LR model0.86 (95% CI 0.76 to 0.97)
ACR-1997— not stated in the abstract
SLICC-2012— not stated in the abstract
EULAR/ACR-2019— not stated in the abstract
AUC
LASSO-LR model— not stated in the abstract
ACR-19970.70; 95% CI 0.62 to 0.78
SLICC-20120.74; 95% CI 0.66 to 0.81
EULAR/ACR-20190.70; 95% CI 0.63 to 0.78
Specificity (binary classifier)
LASSO-LR model0.90, 95% CI 0.78 to 0.95
ACR-1997— not stated in the abstract
SLICC-2012— not stated in the abstract
EULAR/ACR-2019— not stated in the abstract
— not stated in the abstract
Further findings
Simplified scoring system, threshold ≥3
- AUC
- 0.86 (95% CI 0.75 to 0.96)
Features most strongly associated with damage
- SLICC-2012 neurological disorder, EULAR/ACR-2019 class III/IV LN, and among non-criteria manifestations, myocarditis and ILD
Original publication
Systemic lupus erythematosus damage risk index (SLE-DRI): a simple machine learning-based tool for identifying patients at risk for early organ damage. RMD Open. Published 30 September 2025. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →