Impact of Artificial Intelligence–Based Educational Programs on Clinical Decision‑Making and Medical Errors among Nursing Students: A Systematic Review and Narrative Synthesis

Background: Preventable medical errors are a major concern in clinical care, highlighting the importance of effective nursing education for clinical decision-making. Recently, AI-based educational strategies have emerged, but their actual impact on nursing students has not been systematically assessed. Objectives: This systematic review aimed to synthesize the available evidence on the impact of AI‑based educational programs on clinical decision‑making and medical errors among nursing students. Methods: This systematic review was conducted according to PRISMA 2020. PubMed, Scopus, Web of Science, Cochrane CENTRAL, and Embase were searched for relevant studies (2019–2024). Eligible studies were randomized controlled trials (RCTs) and quasi-experimental designs comparing AI-based interventions with traditional education in undergraduate nursing students. Study selection, data extraction, and quality appraisal (RoB-2 and ROBINS-I) were performed independently. Due to substantial heterogeneity among the included studies, a narrative synthesis was performed. Results: The initial search identified 1,487 records, of which 16 studies involving approximately 1,590 nursing students met the inclusion criteria. AI‑based interventions, mainly virtual patient simulations and adaptive learning systems, significantly enhanced clinical decision‑making skills in all included studies. Moreover, most studies assessing medical errors (9 of 10) reported notable reductions in medication miscalculations and diagnostic inaccuracies. Overall, the methodological quality of the included studies was rated as moderate to good. Conclusion: AI-based education shows strong potential to improve clinical judgment and promote patient safety in nursing students. Careful, context-sensitive integration into nursing curricula is recommended, with future research needed to address methodological heterogeneity and long-term effectiveness.