Spondyloarthritis & psoriatic arthritis
An interpretable machine learning approach for detecting psoriatic arthritis in a UK primary care psoriasis cohort using electronic health records from the Clinical Practice Research Datalink
Authors' conclusion
We were able to identify psoriasis patients at higher risk, and important indicators, of PsA in UK primary care.
Study context
- Population
- Psoriatic arthritis (PsA)-free patients with incident psoriasis, 1998–2019, UK primary care (Clinical Practice Research Datalink (CPRD))2,460 developed PsA
- Participants
- 122,330 patients
- Comparison
- Bayesian network vs random forest for detecting newly diagnosed PsA
- Outcome
- Model discrimination (AUC and PRAUC) for detecting PsA
Methodological detailsmachine learning prediction model, prospective cohort follow-up, permutation variable importance
- Machine learning prediction model
- Prospective cohort follow-up
- Permutation variable importance
Performance
As reported in the abstract
Discrimination for PsA
| Test or model | AUC | PRAUC |
|---|---|---|
| BN | 0.823 | 0.221 |
| Random forest (RF) | 0.851 | 0.261 |
AUC
BN0.823
Random forest (RF)0.851
PRAUC
BN0.221
Random forest (RF)0.261
Further findings
- Psoriasis duration, NSAID prescriptions, nonspecific arthritis, nonspecific arthralgia and CRP blood tests were all important variables
Original publication
An interpretable machine learning approach for detecting psoriatic arthritis in a UK primary care psoriasis cohort using electronic health records from the Clinical Practice Research Datalink. Ann Rheum Dis. Published 1 March 2025. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →