Heart failure & cardiomyopathiesCohort study
Outcome trajectories in hospitalized heart failure with preserved ejection fraction: a machine learning cluster analysis
- Prospective cohort study with machine learning cluster analysis
- consecutive patients
Authors' conclusion
Cluster analysis identified low-, intermediate-, and high-risk acute HFpEF phenotypes.
Study context
- Population
- Patients hospitalised for HF with LVEF >40%, evaluated at discharge1,052 screened
- Participants
- 471 patients
- Exposure
- Cluster 2 and Cluster 3 vs Cluster 1
- Outcome
- Composite of:
- All-cause death
- Urgent heart transplant
- HF hospitalisation
- ED visit for decompensated HF
- Follow-up
- 12 months
Main finding
As reported in the abstract
Each group vs cluster 1Hazard ratio (95% CI)P value
Cluster 2
2.9(1.7–5.2)
P < 0.001Cluster 3
4.8(2.7–8.4)
P < 0.001Observational data: associations, not effects.
Further findings
Composite endpoint in the vulnerable period · 3 months
- Outcome results consistent
Predictors of adverse outcomes
- NYHA class and neutrophil-lymphocyte ratio (NLR) emerged as key prognostic nodes
Original publication
Outcome trajectories in hospitalized heart failure with preserved ejection fraction: a machine learning cluster analysis. Eur J Heart Fail. Published 10 February 2026. doi:10.1093/ejhf/xuag037
Summarised from the published abstract; every number is checked against it