Applications of Artificial Intelligence in Out-of-Hospital Cardiac Arrest: A Systematic Review

• Published in: Curēus (Palo Alto, CA) Vol. 17; no. 4; p. e82320
• Main Authors: Cheriachan, Doju, Desai, Heet N, Sangurima, Leslie, Malik, Maujid Masood, Ganatra, Nency, Siby, Rosemary, Kumar, Sanjay, Khan, Sara, Jayaprakasan, Srilakshmi K, Hamid, Pousette
• Format: Journal Article
• Language: English
• Published: United States Springer Nature B.V 15.04.2025; Cureus
• Subjects: Artificial intelligence; Cardiac arrest; Cardiology; Cardiopulmonary resuscitation; Clinical outcomes; CPR; Deep learning; English language; Healthcare Technology; Hospitals; Internal Medicine; Machine learning; Medical Subject Headings-MeSH; Neural networks; Patients; Respiratory failure; Systematic review; management of cardiac arrest; cardiac arrest outcome; emergency response; ai protocols for healthcare; cardiac resuscitation; artificial intelligence in health care; out-of-hospital cardiac arrest
• ISSN: 2168-8184; 2168-8184
• DOI: 10.7759/cureus.82320

Artificial intelligence (AI) refers to a computer system capable of performing complex tasks that require human skills. AI is increasingly being utilized in the healthcare sector; therefore, the aim of this review was to explore the role of AI in enhancing immediate response and successful management of cardiac arrest by improving the recognition of cardiac arrest from emergency calls in an out-of-hospital setting. The preferred design for this study was a literature review. To get the relevant articles, an in-depth literature search for primary and secondary studies was carried out over various databases, namely ProQuest, CINHAL, PUBMED, and Google Scholar. The results revealed that in five out of the 12 studies, there was a total of 98,922 participants. Three were reviews of past studies, and one involved the examination of a number of emergency calls, but the number of participants was not specified, while the other examined AI and self-care. A thematic analysis of the 10 articles was performed and four dominant themes were identified, namely AI improves self-care; AI improves clinical outcomes with a positive predictive value of (33.0%, p < 0.001); improved decision making with an accuracy of 0.908, which increases the survival rate with an accuracy of 0.896; and the prediction of cardiac arrest with an accuracy of 0.8 in predicting cardiovascular risks including cardiac arrest. Hence, it is concluded that the integration of AI facilitates the early prediction of potential cases of cardiac arrest in out-of-hospital settings. Early detection is associated with improved decision-making regarding the next action steps to take. AI enhances self-care, whereby through virtual doctors and online applications, patients can take proactive measures when they are susceptible to a heart attack. As a result, the integration of AI significantly improves patient outcomes.