Simulation study maps students’ cautious use of GenAI in cancer care

Students working in interprofessional cancer-care teams perceived generative artificial intelligence as a useful partner only when they treated its outputs as provisional contributions rather than authoritative clinical advice. In a qualitative study of a simulation module developed by universities in Hong Kong and Mainland China, teams said GenAI helped broaden care planning, reveal concerns they had overlooked and organise contributions from different professions. Those perceived benefits emerged alongside clear reservations about accuracy, limited localisation, insufficient clinical detail and the technology’s inability to capture the emotional and relational dimensions of care.

The study examined 316 health-professions students from six disciplines who were organised into 30 interprofessional teams. The module was co-designed by the two partner universities and delivered at the Hong Kong institution. Each team first developed a cancer-care plan without generative AI, then used a GenAI tool of its choice to review and refine that plan. The activity required students to appraise the generated information critically, adapt it to their context and collectively decide whether to include it. This sequence was designed to make AI use part of shared clinical deliberation instead of an independent or automated decision-making step.

All 30 teams consented to the research use of their joint reflections. Researchers analysed 27 records containing substantive reflective accounts; three were marked “NIL.” The reflections described what the researchers called a form of critically engaged partnership between teams and GenAI. Students reported that the technology could widen the scope of their planning and prompt them to identify previously unnoticed issues. It also appeared to help organise ideas from different professions, although those benefits depended on active human involvement rather than passive acceptance of generated text.

Teams achieved this engagement through practices that included formulating and refining prompts, comparing outputs with clinical guidelines and disciplinary knowledge, and negotiating whether particular suggestions belonged in the care plan. Their willingness to accept GenAI recommendations was therefore conditional. Incomplete or insufficiently localised information could prompt scepticism, as could answers that lacked clinically relevant detail or failed to address the personal aspects of cancer care. Across the reflections, the students retained responsibility for final decisions: the generated material entered the care plan only after scrutiny, contextual judgement and collective approval by the professionals involved. The findings support educational designs built around unaided planning, shared verification and explicit professional accountability, but they do not show that GenAI improves reasoning, collaboration or the quality of care plans. Because the evidence came from brief team reflections recorded after one delivery of the module at one site, it captures students’ negotiated perceptions rather than measured performance or clinical outcomes. The result is best understood as an initial evaluation of how health-professions learners might use GenAI within interprofessional education, not as evidence that AI-generated recommendations are safe, accurate or ready for autonomous clinical use.

Subject of Research: Students’ critical use of generative AI in interprofessional cancer-care education.

Article Title: Exploring GenAI‑enabled interprofessional education: development and initial evaluation

Article References: Ganotice, F. A., Jr, Dizon, J. I. W. T., Shen, X., Xiang, J., & He, Q. (2026). Exploring GenAI‑enabled interprofessional education: development and initial evaluation. BMC Medical Education. https://doi.org/10.1186/s12909-026-10251-5

Image Credits: AI Generated

DOI: 10.1186/s12909-026-10251-5

Keywords: generative AI, interprofessional education, cancer care, health professions, simulation, clinical reasoning, AI literacy, student teams, Exploring, GenAI, enabled, interprofessional