The Role Of Artificial Intelligence In Emergency Medicine Decision-Making

Abstract

Introduction Emergency departments (EDs) are high-acuity settings where rapid decision-making is critical. With rising patient volumes and complexity, AI and machine learning (ML) offer tools to support triage, diagnosis, risk stratification, and operational planning.

Aim: To systematically review recent studies assessing AI applications that support decision-making in ED and related emergency/critical care pathways, focusing on performance, implementation, and implications for critical care.

Materials and Methods: Databases searched according to PRISMA Guidelines include PubMed, Embase, Scopus, Web of Science, Cochrane CENTRAL, and Google Scholar. Date range: Jan 1, 2021 – Oct 2, 2025. Search terms combined AI/ML terms with emergency/ED terms. Peer-reviewed studies evaluating AI applied to ED decision-making (triage, diagnosis, prediction) were included, while non-English, and conference abstracts without data were excluded. Data extraction template included study design, sample, AI method, features, outcomes, validation, and risk-of-bias assessment.

Results: Of 1100 records identified, 13 studies met inclusion criteria. AI models demonstrated superior performance in diagnostic accuracy (pooled area under the receiver operating characteristic curve (AUC) 0.90, 95% confidence interval (CI): 0.87-0.95), and outcome prediction (pooled sensitivity for hospital admission: 0.92, 95% CI: 0.87-0.95) compared to traditional methods.

Conclusion: AI shows promise in improving ED decision-making processes. However, challenges remain in real-world implementation, ethical considerations, and long-term impact on patient outcomes. Future research should focus on large-scale validation studies and addressing ethical and safety concerns.