Key points are not available for this paper at this time.
INTRODUCTION: Artificial intelligence (AI) has advanced rapidly in healthcare; however, its application in emergency nursing remains underexplored. This study aimed to map and synthesise existing evidence to clarify current applications, gaps, and practical implications. METHODS: A scoping review was conducted using seven databases from inception to 22 November 2024. From 1,885 initial records, 27 studies were included in the final analysis. The review followed the PRISMA-ScR guidelines, with data extracted in standardised formats and analysed using Arksey and O'Malley's framework. RESULTS: The study was conducted across North America, Asia, Europe, the Middle East, and Oceania. A total of 16 studies focused on triage and decision support, while others addressed direct nursing practices or emergency department operations. Evaluation of AI performance was reported in 20 studies, revealing considerable variability across algorithms, models, and metrics, with machine learning being the predominant approach; however, operational ethical aspects were explicitly discussed in just 11 studies. CONCLUSION: AI demonstrates strong potential for triage, workflow efficiency, and patient safety in emergency nursing but remains poorly integrated into clinical practice. Sustainable progress requires high-quality data, rigorous validation, auditability, and ethical safeguards. Institutional and governmental support, multidisciplinary collaboration, and nurse capacity building are critical for safe, equitable, and scalable implementation.
Cha et al. (Wed,) studied this question.