Assessing clusters of comorbidities in rheumatoid arthritis: a machine learning approach
Authors' conclusion
While ML clustering is useful for modeling multi-omics, using clusters to predict clinical outcomes produces models with a similar fit as those with individual comorbidities.
Study context
- Population
- Patients with rheumatoid arthritis (RA) with longitudinal data over 6 years, US-based CorEvitas registry (mean age 59 years; 77% women)
- Participants
- 11,883 patients
- Exposure
- Models with machine learning (ML)-derived comorbidity clusters vs models with each of 24 comorbidities entered individually
- Outcome
- Model fit (adjusted R2, root mean square error) for predicting time-averaged clinical disease activity index (CDAI) and HAQ-DI
- Follow-up
- Latter 3 years of 6-year study period
Methodological detailsregistry-based (CorEvitas), machine learning clustering (K-mode, K-mean, regression-based, hierarchical)
- Registry-based (CorEvitas)
- Machine learning clustering (K-mode, K-mean, regression-based, hierarchical)
Main finding
As reported in the abstract
Model fit for time-averaged CDAI
- ML-derived comorbidity clusters produced similarly strong models as models with each of the 24 separate comorbidities entered individually
Observational data: associations, not effects.
Further findings
Model fit for time-averaged HAQ-DI
- Same patterns observed for HAQ-DI
Original publication
Assessing clusters of comorbidities in rheumatoid arthritis: a machine learning approach. Arthritis Res Ther. Published 22 November 2023. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →