← All research

Enhanced Early Risk Stratification in Cardiogenic Shock Using Machine Learning: A Hierarchical Clustering Analysis.

JACC. Advances · 2026-10-13

Shuojohn Li, Abduljabar Adi, Alejandro Quintero Villegas et al.

This multicenter retrospective registry included 2,906 adults with cardiogenic shock admitted to ICUs between 2016 and 2022. Hierarchical clustering was applied independently to clinical variables from ICU days 1, 2, and 3. Day-1 Cardiac, Cardio-Renal, and Cardio-Metabolic phenotypes had reported mortality of 15%, 27%, and 55%, respectively. Similar phenotypes persisted through days 2 and 3, and phenotype classification enhanced mortality stratification within each SCAI stage across the first three ICU days. The authors concluded that prospective validation is required before clinical implementation.

Clinical bottom line

Early clinical phenotypes may provide additional mortality stratification within SCAI stages in cardiogenic shock. These retrospective findings support further validation, not established clinical use.

Study limitation: The retrospective design limits inference; the abstract does not specify the mortality timepoint or quantify incremental predictive performance.

Read the original on PubMed

AI-generated from the published abstract with automated source checks. For research discovery; consult the original paper before applying findings.