Unsupervised Machine Learning to Identify Subphenotypes Among Cardiac Intensive Care Unit Patients with Heart Failure
Unsupervised machine learning clustering can identify multiple distinct clinical HF phenotypes within the CICU population that display differing mortality profiles both in-hospital and at 1 year. Mortality was lowest for the uncomplicated HF phenotype and highest for the hypoperfused phenotype. The inflamed phenotype had comparatively higher in-hospital mortality yet lower post-discharge mortality, suggesting divergent short-term and long-term prognosis.
Authors
- Jacob Colin Jentzer (ORCID: https://orcid.org/0000-0002-6366-2859)
- Sabri Soussi (ORCID: https://orcid.org/0000-0002-1489-4768)
- Yogesh N.V. Reddy (ORCID: https://orcid.org/0000-0001-5322-0902)
- Patrick R. Lawler (ORCID: https://orcid.org/0000-0001-5155-5071)
- Ruben Crespo‐Diaz
- Parag C. Patel (ORCID: https://orcid.org/0000-0002-2761-5895)
- Alexandre Mebazaa (ORCID: https://orcid.org/0000-0001-8715-7753)
- Shannon M. Dunlay (ORCID: https://orcid.org/0000-0001-5145-7420)
Institutions
- University Health Network (CA)
- Mayo Clinic (US)
- Inserm (FR)
- University of Toronto (CA)
- Université Paris Cité (FR)
- Toronto Western Hospital (CA)
- Jacksonville College (US)
- McGill University Health Centre (CA)
- Sorbonne Paris Cité (FR)
- Assistance Publique – Hôpitaux de Paris (FR)
- Mayo Clinic in Arizona (US)
- Paris Cardiovascular Research Center (FR)
- Mayo Clinic in Florida (US)
Publication Details
- Journal
- ESC Heart Failure
- Published
- 2024-08-19
- DOI
- https://doi.org/10.1002/ehf2.15027
- Citations
- 7
- Primary Topic
- Heart Failure Treatment and Management
- Type
- article
- Field-Weighted Citation Impact
- 2.81