Assessing the efficacy of ChatGPT as a virtual patient in nursing simulation training: A study on nursing students' experience

Integrating healthcare simulators in nursing education allows for innovative training approaches. ChatGPT as a virtual patient is one such innovation. This study assessed ChatGPT's effectiveness in nursing simulation. Twelve nursing students from Morocco's Higher Institute of Health Sciences engaged with ChatGPT in a dyspnea scenario. Data collection included acceptability, accessibility, engagement ratings, and assessing students' virtual patient interaction skills. Students embraced ChatGPT (accessibility: 4.3 ± 0.5, engagement: 4.3 ± 0.5). They recognized its value in training (average: 4.2 ± 0.5). Student interaction skills positively correlated with overall performance. Notably, students' interaction skills demonstrated a positive correlation with their overall performance. Further analysis revealed significant correlations between specific skills: clarity and comprehensibility of responses (r = 0.701), relevance of responses (r = 0.444), and the ability to provide useful information (r = 0.597). Incorporating communication-based simulations and patient engagement in nursing education is crucial. ChatGPT as a virtual patient holds promise in enriching clinical learning. Integrating such technology effectively prepares students for real patient interactions.