Scholarlink

Introduction <p>Generative artificial intelligence (GenAI) has transformative potential in healthcare simulation (HCS), where developing clinical reasoning skills is crucial for patient safety and quality of care. Despite a growing interest on GenAI in HCS, no systematic review has examined how GenAI-powered HCS supports clinical reasoning skill development. This review addresses this gap by synthesizing current evidence on GenAI applications, outcomes and challenges in this context.</p> Methods <p>Following the 2020 PRISMA guidelines, a systematic search was conducted across PubMed (MEDLINE), ScienceDirect, Scopus, Web of Science, and Google Scholar for studies published between January 2020 and May 2026. Eligible studies were assessed for risk of bias using the ROBINS-I and RoB2 Cochrane tools. The review protocol was registered with PROSPERO (ID: CRD420251034581).</p> Results <p>22 studies met inclusion criteria. ChatGPT was the most widely used GenAI tool to empower HCS for clinical reasoning skills development. Identified applications included history-taking interviews, diagnostic accuracy, problem-solving, critical thinking, clinical judgment, and decision-making. Reported pedagogical outcomes encompassed learners’ satisfaction, and motivation for self-directed learning, improved medical diagnostic and interview performance, enhanced dialogue authenticity. However, challenges included excessive verbosity, artificial intelligence hallucinations, the lack of non-verbal and empathetic elements essential to clinical reasoning development. Furthermore, ethical, linguistic, and technical barriers also limited GenAI-powered HCS implementation.</p> Conclusion <p>GenAI shows strong promise for enhancing healthcare students’ clinical reasoning through personalized feedback and adaptive learning experiences. Healthcare educators are invited to integrate GenAI tools to complement traditional simulation modalities with a necessity of human-in-the-loop supervision. Nevertheless, continued rigorous research is required to overcome current ethical, pedagogical, and technological limitations.</p> Summary Statement <p>Generative artificial intelligence (GenAI) technology holds a promising potential to improve healthcare students clinical reasoning outcomes allowing for scalable, immersive, and personalized learning experiences. While learner engagement, knowledge gains, history-taking, diagnostic accuracy, clinical-decision making, and cost-efficiency are clear strengths, challenges in realism, feedback, AI verbosity and hallucinations, validity, technical and linguistic barriers, and ethical oversight must be addressed. With iterative refinement and pedagogical alignment, GenAI can serve as a valuable complement, not a replacement, to human-guided clinical simulation education.</p> <p>GenAI in healthcare simulation training serves as a reliable adjunct, particularly in low-resource settings or where standardized patients are impractical. GenAI, when thoughtfully integrated, it can amplify human clinical reasoning, expose cognitive blind spots, and democratize access to high-fidelity simulation training across different disciplines and settings. However, pedagogical integration must be intentional, with clear learning objectives, robust debriefing, and human oversight to mitigate AI limitations.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

A Systematic Review of Generative Artificial Intelligence-powered Healthcare Simulation for Clinical Reasoning Skills Development: Applications, Outcomes and Challenges

  • 期刊论文

  • 发表日期：

    2026年7月3日

查看全文
  • Kamal Takhdat,
  • Abdelmonim El fadely,
  • Eddabbah Mohamed,
  • Ahmed Ouaamr,
  • Ahmed Rhassane El Adib

摘要

Introduction

Generative artificial intelligence (GenAI) has transformative potential in healthcare simulation (HCS), where developing clinical reasoning skills is crucial for patient safety and quality of care. Despite a growing interest on GenAI in HCS, no systematic review has examined how GenAI-powered HCS supports clinical reasoning skill development. This review addresses this gap by synthesizing current evidence on GenAI applications, outcomes and challenges in this context.

Methods

Following the 2020 PRISMA guidelines, a systematic search was conducted across PubMed (MEDLINE), ScienceDirect, Scopus, Web of Science, and Google Scholar for studies published between January 2020 and May 2026. Eligible studies were assessed for risk of bias using the ROBINS-I and RoB2 Cochrane tools. The review protocol was registered with PROSPERO (ID: CRD420251034581).

Results

22 studies met inclusion criteria. ChatGPT was the most widely used GenAI tool to empower HCS for clinical reasoning skills development. Identified applications included history-taking interviews, diagnostic accuracy, problem-solving, critical thinking, clinical judgment, and decision-making. Reported pedagogical outcomes encompassed learners’ satisfaction, and motivation for self-directed learning, improved medical diagnostic and interview performance, enhanced dialogue authenticity. However, challenges included excessive verbosity, artificial intelligence hallucinations, the lack of non-verbal and empathetic elements essential to clinical reasoning development. Furthermore, ethical, linguistic, and technical barriers also limited GenAI-powered HCS implementation.

Conclusion

GenAI shows strong promise for enhancing healthcare students’ clinical reasoning through personalized feedback and adaptive learning experiences. Healthcare educators are invited to integrate GenAI tools to complement traditional simulation modalities with a necessity of human-in-the-loop supervision. Nevertheless, continued rigorous research is required to overcome current ethical, pedagogical, and technological limitations.

Summary Statement

Generative artificial intelligence (GenAI) technology holds a promising potential to improve healthcare students clinical reasoning outcomes allowing for scalable, immersive, and personalized learning experiences. While learner engagement, knowledge gains, history-taking, diagnostic accuracy, clinical-decision making, and cost-efficiency are clear strengths, challenges in realism, feedback, AI verbosity and hallucinations, validity, technical and linguistic barriers, and ethical oversight must be addressed. With iterative refinement and pedagogical alignment, GenAI can serve as a valuable complement, not a replacement, to human-guided clinical simulation education.

GenAI in healthcare simulation training serves as a reliable adjunct, particularly in low-resource settings or where standardized patients are impractical. GenAI, when thoughtfully integrated, it can amplify human clinical reasoning, expose cognitive blind spots, and democratize access to high-fidelity simulation training across different disciplines and settings. However, pedagogical integration must be intentional, with clear learning objectives, robust debriefing, and human oversight to mitigate AI limitations.