Who carries the load? Role-specific cognitive load and team performance in a multi-role virtual reality emergency simulation: an exploratory pilot study of collaborative clinical reasoning in undergraduate medical education

Abstract Background Clinical reasoning is increasingly understood as a collaborative, distributed team activity, yet virtual reality (VR) simulation in undergraduate medical education remains predominantly single-user based. Guided by Collaborative Cognitive Load Theory, we developed a structured peer-assisted learning and team-based VR emergency course for one VR-operating student and up to five peers to occupy complementary functional roles (Team Lead, two Guideline Research roles, Documentation, Handover). The aim of this exploratory study was to examine the distribution of cognitive load across the six functional roles, the relation between a role’s cognitive load and overall team performance, and the relation between cognitive load and individually perceived competence gain. Methods In a single-arm exploratory pilot study, 44 undergraduate medical students participated in nine peer-led, VR sessions with rotating roles. Perceived workload was measured with the NASA Task Load Index (NASA-TLX) in 218 single-role observations and analysed with linear mixed-effects models. Team performance was rated with the Team Emergency Assessment Measure (TEAM), anchored to each participant's first Team Lead rotation, and related to workload using Spearman correlations. Self-assessed competence was rated on five items before and after the sessions and analysed using Wilcoxon signed-rank tests (Holm-corrected). Results All five self-assessment items improved significantly (all adjusted p ≤ .001; rank-biserial r rb = 0.72–1.00). Workload differed strongly by role (χ²(5) = 71.98, p < .001). After removing VR-specific physical demand, the Team Lead carried the highest cognitive load, and the two guideline-research roles diverged significantly, with procedural sequencing more demanding than pharmacological research. Higher Team Lead workload was associated with lower observer-rated team performance (ρ = −0.37, p = .018), whereas workload in the other roles showed only weak, non-significant correlations with performance. Individual workload showed no clear association with perceived competence gain. Conclusions This study examined a resource-conscious multi-role VR emergency simulation format in undergraduate medical education, associated with competence gains and functionally interpretable, role-specific workload profiles. The Team Lead role showed the highest cognitive demands in distributed reasoning, suggesting targeted instructional support for this role and offering a measurement model for studying collaborative clinical reasoning in team training.