Applications of artificial intelligence in advanced clinical simulation for anesthesia education: a scoping review

Anesthesiology is characterized by its development in complex, dynamic, and high-risk clinical environments, where decision-making under pressure, effective team coordination, and the precise execution of technical and non-technical skills are crucial for patient safety. In this context, advanced clinical simulation has become a fundamental educational strategy for the training and evaluation of anesthesiologists, enabling the practice of highly complex, low-frequency scenarios that are difficult to address safely and systematically in real-world clinical practice (1,2). Therefore, simulation has evolved from a complementary tool to a structural component of modern anesthesia curricula. In line with the above, numerous high-impact studies and reviews have demonstrated that advanced clinical simulation improves technical performance, strengthens non-technical skills, and contributes to the development of clinical reasoning and crisis management in anesthesia (3,4). Furthermore, when simulation is integrated with sound pedagogical principles such as deliberate practice, structured feedback, and alignment with curricular objectives, it is associated with sustained improvements in learning and greater transfer to the real clinical setting (5,6). However, despite its widespread adoption, significant challenges remain related to the standardization of scenarios, the objective evaluation of performance, and the personalization of training according to the individual needs of learners. In relation to these challenges, anesthesiology has been one of the pioneering specialties in the use of simulation for training in crisis management and teamwork, particularly through approaches such as crisis resource management and structured debriefing (7,8). However, even in well-established simulation programs, performance assessment continues to rely heavily on human observation, which introduces inter-rater variability, limits program scalability, and hinders personalized feedback based on objective data (4,9). Consequently, the need to incorporate new tools that optimize the measurement, analysis, and feedback of performance in advanced clinical simulation has been highlighted. In this context, artificial intelligence emerges as a technology with the potential to substantially transform medical education and, in particular, clinical simulation. Artificial intelligence has demonstrated increasing capabilities to analyze large volumes of data, identify complex patterns, and generate adaptive systems that can support educational, evaluative, and decision-making processes (10,11). In the field of medical education, recent high-impact reviews have documented a sustained increase in the use of artificial intelligence for teaching, performance assessment, and supporting clinical reasoning, although they have also highlighted methodological fragmentation and the need for clear conceptual frameworks for its responsible implementation (12). Specifically, the convergence of artificial intelligence and advanced clinical simulation offers unique opportunities for anesthesia education. These include the dynamic generation of simulated scenarios, real-time adaptation of case complexity based on learner performance, automated analysis of technical and non-technical variables, and support for debriefing through advanced performance analytics (13,14). Consequently, AI-based simulation could overcome some of the current limitations of traditional models, facilitating more personalized, objective, and scalable training. However, despite the growing interest in these applications, the available evidence remains scattered and heterogeneous. Studies differ widely in the types of algorithms used, simulation environments, fidelity levels, target populations, and educational outcomes assessed. Furthermore, ethical, pedagogical, and validity considerations have not yet been systematically addressed, particularly regarding algorithmic transparency, bias, and curriculum integration (11,15). Therefore, it is necessary to map and characterize the existing body of evidence in a structured way. Therefore, a scoping review is warranted to identify, describe, and synthesize current applications of artificial intelligence in advanced clinical simulation for anesthesia education. This approach will allow us to explore the types of interventions developed, the educational contexts in which they have been implemented, the outcomes evaluated, and the main knowledge gaps, providing a solid foundation to guide future research and the design of evidence-based educational programs.