Access to practice opportunities for history-taking in undergraduate medical education can be resource-limited. Large language models are a potential avenue to address this. This study sought to characterize changes in learner self-reported confidence with history-taking before and after a simulation with an LLM-based patient and understand learner experience with and the acceptability of virtual LLM-based patients. This was a multi-method study conducted at McMaster University. Simulations were facilitated with the OSCEai tool. Data was collected through surveys with a Likert scale and open-ended questions and semi-structured interviews. A total of 24 participants generated 93 survey responses and 17 interviews. Overall, participants reported a 14.6% increase in comfort with history-taking. Strengths included its flexibility, accessibility, detailed feedback, and ability to provide a judgement-free space to practice. Limitations included its lower fidelity compared to standardized patients and at times repetitive and less clinically relevant feedback as compared to preceptors. It was overall viewed best as a supplement rather than a replacement for standardized patients. In conclusion, LLM-based virtual patients were feasible and valued as an adjunct tool. They can support scalable, personalized practice. Future work is needed to understand objective metrics of improvement and to design curricular strategies for integration.

Large language model-enabled virtual patients may expand access to clinical skills practice by supporting explicitly defined practice tasks. This program evaluation examined medical students’ awareness, use, and perceptions of the McMaster Virtual SP Tool, a custom generative artificial intelligence tool developed to supplement clinical skills practice. We conducted an anonymous, single-institution cross-sectional survey of students across three cohorts at McMaster University. Quantitative responses were summarized descriptively, and free-text responses were analyzed using qualitative content analysis informed by task-aligned fidelity, deliberate practice, learner-centred feedback, and simulation instructional design. Thirty-five students responded; 23 (65.7%) were aware of the tool and 16 (45.7%) had used it. Among the 16 users, 15/16 (93.8%) found the tool at least somewhat easy to navigate; 13 (81.3%) would use it again; and 13 (81.3%) would recommend it. Across all respondents, 14/35 (40.0%) reported using AI-enabled virtual patients for OSCE preparation. Users valued its accessibility, independent low-stakes rehearsal, and usefulness for focused history-taking, question wording, and clinical reasoning. Perceived limitations included reduced human connection, nonverbal and emotional realism, physical examination practice, and feedback specificity. AI-enabled virtual patients may therefore be considered as adjuncts for selected cognitive and structural rehearsal tasks. Objective educational outcomes were not assessed.