Arrhythmias & electrophysiology
Phenotyping of atrial fibrillation with cluster analysis and external validation
Authors' conclusion
Machine learning-based cluster analysis identified five different phenotypes of AF with unique clinical characteristics and different clinical outcomes.
Study context
- Population
- Patients with atrial fibrillation (AF) in Japanese registries (SAKURA AF derivation; RAFFINE validation)derivation
- Participants
- 3,055 patients
- Comparison
- Five AF phenotype clusters
- Cluster 1n = 414
- Cluster 2n = 1,003
- Cluster 3n = 517
- Cluster 4n = 652
- Cluster 5n = 469
- Outcome
- All-cause mortality and composite CV events
Methodological detailsmulticentre, K-prototype cluster analysis (14 clinical variables)
- Multicentre
- K-prototype cluster analysis (14 clinical variables)
Performance
As reported in the abstract
Risk of all-cause mortality increased across clusters
log-rank p<0.001
Risk of composite CV events increased across clusters
log-rank p<0.001
Further findings
- Similar results in the external validation cohort
Original publication
Phenotyping of atrial fibrillation with cluster analysis and external validation. Heart. Published 1 June 2023. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →