Evaluating the Effectiveness of Prognostic Models in Patients with Cardiogenic Shock Undergoing VA-ECMO Support - Report - MDSpire

Evaluating the Effectiveness of Prognostic Models in Patients with Cardiogenic Shock Undergoing VA-ECMO Support

  • By

  • Maria Calvo-Barceló

  • Finn Boyhan Irvine

  • Niamh Tierney

  • Shivani Ayyar

  • Taylor Devine

  • Vasileios Panoulas

  • Maria Montegudo-Vela

  • Fernando Riesgo Gil

  • Clara Hernandez Caballero

  • I. Dolores Poveda Pinedo

  • Javier Bautista

  • Jason Van Schoor

  • Donna Hall

  • Sofia Pinto

  • Eftychia Galiatsou

  • Alex Rosenberg

  • April 28, 2026

  • 0 min

Share

Clinical Report: Evaluating the Effectiveness of Prognostic Models in Patients with Cardiogenic Shock Undergoing VA-ECMO Support

Overview

Expand on the implications of poor model discrimination for clinical decision-making.

Background

Accurate risk stratification for patients undergoing veno-arterial extracorporeal membrane oxygenation (VA-ECMO) for cardiogenic shock is crucial for clinical decision-making. Existing prognostic scores have shown limited real-world applicability, raising concerns about their utility in guiding treatment. Understanding the factors influencing survival can help improve patient outcomes and inform ECMO initiation strategies.

Data Highlights

OutcomePercentage
Survival to decannulation52%
Survival to ICU discharge41%
Survival to six months37%

Key Findings

  • Discrimination of prognostic models was poor, with AUCs ranging from 0.52 to 0.68.
  • The Alfred score had the best performance among the models evaluated (AUC 0.68).
  • Cardiogenic shock aetiology was the only independent predictor of survival.
  • Survival rates varied significantly by aetiology, with myocarditis showing the highest survival at 76%.
  • General ICU scores performed poorly, likely due to their reliance on variables less informative in ECMO patients.
  • Models should incorporate readily available data for effective pre-cannulation assessment.

Clinical Implications

Clinicians should be aware of the limited utility of existing prognostic models in predicting outcomes for patients undergoing VA-ECMO. Emphasis should be placed on aetiology and other readily available clinical data to guide decision-making regarding ECMO initiation.

Conclusion

The study highlights the need for improved prognostic tools that can reliably assess risk in patients with cardiogenic shock undergoing VA-ECMO. Future models should focus on integrating aetiological factors with easily obtainable clinical variables.

References

  1. Intensive Care Medicine, 2022 -- Mortality Prediction Models for Patients Undergoing ECMO: A Systematic Review of Their Characteristics and Performance
  2. Critical Care (Springer), 2025 -- Machine learning models for predicting limb ischemia during VA-ECMO: an analysis of the Chinese extracorporeal life support registry
  3. Intensive Care Medicine, 2023 -- ECMO Survival Prediction: Implementing Deep Learning Models in Venoarterial Extracorporeal Membrane Oxygenation
  4. Clinical Research in Cardiology, 2020 -- Evaluation of Mechanical Circulatory Support: A Propensity-Matched Study of Venoarterial Extracorporeal Membrane Oxygenation versus Impella in Patients Experiencing Cardiogenic Shock
  5. Mechanical Circulatory Support in Acute Myocardial Infarction-Cardiogenic Shock: 2025 Acute Coronary Syndrome Guideline in Context - PMC
  6. ExtraCorporeal Life Support for acute myocardial infarction complicated by cardiogenic shock - American College of Cardiology
  7. Performance of prognostic scores for patients with cardiogenic shock supported With VA-ECMO | Critical Care | Springer Nature Link
  8. Current State of Prognostication for VA-ECMO
  9. ExtraCorporeal Life Support for acute myocardial infarction complicated by cardiogenic shock - American College of Cardiology
  10. Performance of prognostic scores for patients with cardiogenic shock supported With VA-ECMO | Critical Care | Springer Nature Link

Original Source(s)

Evaluating the Effectiveness of Prognostic Models in Patients with Cardiogenic Shock Undergoing VA-ECMO Support

  • by Maria Calvo-Barceló, Finn Boyhan Irvine, Niamh Tierney, Shivani Ayyar, Taylor Devine, Vasileios Panoulas, Maria Montegudo-Vela, Fernando Riesgo Gil, Clara Hernandez Caballero, I. Dolores Poveda Pinedo, Javier Bautista, Jason Van Schoor, Donna Hall, Sofia Pinto, Eftychia Galiatsou, Alex Rosenberg

  • April 28, 2026

Related Content

GLP-1 Drugs Tied to Lower Addiction Risk

In a target-trial emulation of more than 600,000 veterans, GLP-1 RA initiators saw fewer new substance use disorders—and patients with existing SUDs had fewer overdoses, hospitalizations, and deaths.

Case Report: When inotropy backfires—dobutamine-associated eosinophilic myocarditis

Quality of Life in Cardiac Arrest Recovery

Post–cardiac arrest care is increasingly structured around survivorship, with standardized patient-reported assessments supporting multidisciplinary follow-up.