Related Experiment Video
Updated: Jul 17, 2025
Real-time Pressure-volume Analysis of Acute Myocardial Infarction in Mice
Published on: July 2, 2018
Identification of distinct clinical phenotypes of cardiogenic shock using machine learning consensus clustering
Li Wang1, Yufeng Zhang2, Renqi Yao3,4,5
1Department of Nephrology, Changzheng Hospital, Naval Medical University, Shanghai, China.
Insights
Machine learning identified two distinct cardiogenic shock (CS) phenotypes. One phenotype shows better outcomes, aiding in personalized treatment strategies for CS patients.
Area of Science:
- Critical Care Medicine
- Data Science in Healthcare
- Cardiology
Background:
- Cardiogenic shock (CS) presents complex challenges due to diverse causes and outcomes.
- Differentiating between various CS phenotypes remains clinically difficult.
- Identifying distinct CS phenotypes is crucial for targeted interventions and improved patient prognoses.
Purpose of the Study:
- To investigate the utility of machine learning (ML) consensus clustering for identifying distinct cardiogenic shock (CS) phenotypes.
- To explore if these identified phenotypes possess unique clinical profiles and prognoses.
- To leverage ML for a data-driven approach to CS classification.
Main Methods:
- Utilized unsupervised ML consensus clustering on a large cohort of 21,925 patients with CS from the eICU Collaborative Research Database.
- Determined the optimal number of clusters using consensus matrix heatmap, CDF, cluster-consensus plots, and PAC analysis.
- Analyzed key features differentiating clusters using standardized mean difference and logistic regression for clinical endpoint associations.
Main Results:
- Identified two distinct CS clusters (Cluster 1: 9,848 patients; Cluster 2: 12,077 patients).
- Cluster 1 exhibited lower blood pressure, poorer renal function (lower eGFR, higher BUN/creatinine), and higher severity scores (SOFA, APS III, APACHE IV).
- Cluster 2 was significantly associated with lower in-hospital mortality (OR 0.374), ICU mortality (OR 0.349), and reduced incidence of acute kidney injury (AKI) (OR 0.478).
Conclusions:
- ML consensus clustering effectively synthesizes clinical and laboratory data to reveal distinct CS phenotypes.
- The identified phenotypes possess significantly different clinical profiles and prognoses.
- This approach offers a promising avenue for stratifying CS patients and guiding clinical decision-making.
Background:
Cardiogenic shock (CS) is a complex state with many underlying causes and associated outcomes. It is still difficult to differentiate between various CS phenotypes. We investigated if the CS phenotypes with distinctive clinical profiles and prognoses might be found using the machine learning (ML) consensus clustering approach.
Methods:
The current study included patients who were diagnosed with CS at the time of admission from the electronic ICU (eICU) Collaborative Research Database. Among 21,925 patients with CS, an unsupervised ML consensus clustering analysis was conducted. The optimal number of clusters was identified by means of the consensus matrix (CM) heat map, cumulative distribution function (CDF), cluster-consensus plots, and the proportion of ambiguously clustered pairs (PAC) analysis. We calculated the standardized mean difference (SMD) of each variable and used the cutoff of ± 0.3 to identify each cluster's key features. We examined the relationship between the phenotypes and several clinical endpoints utilizing logistic regression (LR) analysis.
Results:
The consensus cluster analysis identified two clusters (Cluster 1: n = 9,848; Cluster 2: n = 12,077). The key features of patients in Cluster 1, compared with Cluster 2, included: lower blood pressure, lower eGFR (estimated glomerular filtration rate), higher BUN (blood urea nitrogen), higher creatinine, lower albumin, higher potassium, lower bicarbonate, lower red blood cell (RBC), higher red blood cell distribution width (RDW), higher SOFA score, higher APS III score, and higher APACHE IV score on admission. The results of LR analysis showed that the Cluster 2 was associated with lower in-hospital mortality (odds ratio [OR]: 0.374; 95% confidence interval [CI]: 0.347-0.402; P < 0.001), ICU mortality (OR: 0.349; 95% CI: 0.318-0.382; P < 0.001), and the incidence of acute kidney injury (AKI) after admission (OR: 0.478; 95% CI: 0.452-0.505; P < 0.001).
Conclusions:
ML consensus clustering analysis synthesized the pattern of clinical and laboratory data to reveal distinct CS phenotypes with different clinical outcomes.
More Related Videos
06:10Utilizing Percutaneous Ventricular Assist Devices in Acute Myocardial Infarction Complicated by Cardiogenic Shock
Published on: June 12, 2021
07:39Use of a Percutaneous Ventricular Assist Device/Left Atrium to Femoral Artery Bypass System for Cardiogenic Shock
Published on: August 16, 2021
Related Concept Videos
01:30Heart Failure IV: Classification and Diagnostic Evaluation
01:25Cardiomyopathy I: Introduction and Classification
01:30Cardiomyopathy II: Dilated Cardiomyopathy
01:24Blood Pressure Imbalances and Circulatory Shock
Blood Pressure: Hypertension and Hypotension
Normal blood pressure is 120/80 mm Hg. Elevated blood pressure is 120-129/under 80 mm Hg. Hypertension, warranting treatment at 130/80 mm Hg, is often asymptomatic and can lead to severe cardiovascular events, aneurysms, peripheral arterial disease, chronic renal disease, or cardiac...
01:25Cardiopulmonary Resuscitation IV: Pharmacological Management
01:26Heart Failure III: Clinical Manifestations