AI is expanding across the full chain of cardiac arrest care. The next step involves testing algorithms in multicenter clinical settings, improving interpretability, and ensuring improved patient outcomes.
Summary
A review published in the World Journal of Emergency Medicine (March 15, 2026) provides an overview of AI applications across the full cardiac arrest care continuum, covering prediction, resuscitation support, prognosis, and other emerging areas.
Methods
The review followed PRISMA guidelines, searching PubMed, Embase, Cochrane Library, and Web of Science from inception to June 10, 2025. From 2,108 records, the team included 114 studies and assessed 92 distinct AI models.
Key Findings
- Pre-Arrest Prediction: A multilayer perceptron model for in-hospital cardiac arrest achieved the highest reported AUC of 0.998. For out-of-hospital cardiac arrest prediction, extreme gradient boosting and random forest models reached an AUC of 0.950.
- Resuscitation Support: Convolutional neural network models showed strong performance, with a best reported AUC of 0.990.
- Prognosis: A multilayer perceptron model for out-of-hospital cardiac arrest reached an AUC of 0.976.
- Emerging Directions: Studies covered large language models (e.g., GPT), emergency call recognition, wearable-based detection, and AI-assisted education.
Practical Implications
AI may help identify patients at risk of in-hospital cardiac arrest, support emergency dispatch and CPR feedback, and assist with prognosis and clinical trial design.
Limitations
Barriers include data imbalance, limited external validation, infrastructure gaps, privacy concerns, and algorithmic bias. Future work should focus on prospective trials, explainable models, and equitable deployment.
Source
Luo, X., et al. (2026). Beyond the chain of survival: a scoping review of artificial intelligence applications in cardiac arrest. World Journal of Emergency Medicine. DOI: 10.5847/wjem.j.1920-8642.2026.025.