BACKGROUND: Simulation-based training is essential for preparing medical interns to manage high-stakes emergencies. Although virtual reality (VR)-based simulation has been rapidly integrated into medical education, there remains limited evidence directly assessing its effectiveness relative to established high-fidelity simulation (HFS) methodologies. OBJECTIVE: This study aimed to assess the perceived educational effectiveness of VR and HFS in enhancing novice physicians' confidence, satisfaction, and perceived preparedness for managing acute oxygen desaturation. METHODS: A randomized controlled trial was conducted with 168 medical interns from Seoul National University Hospital. Participants were randomly assigned to VR group (n = 81) or HFS group (n = 87). Four participants were excluded due to incomplete surveys, leaving 164 for analysis (VR:79; HFS: 85). Both groups were trained to manage simulated patients with low oxygen saturation. Confidence (10-point Likert scale) and satisfaction (7-point Likert scale) were measured using pre- and post-training surveys. Usability was assessed with the User Experience Questionnaire-Short. Between-group comparisons were conducted using t-tests and chi-square tests, while within-group confidence changes were analyzed using paired t-tests and repeated-measures analysis of variance. To account for correlated data and estimate effect sizes, generalized estimating equations were applied, with statistical significance set at P < .05. Focus group interviews (FGIs) at one- and five-months post-training explored real-world application and behavior transfer. Transcripts were independently reviewed by two researchers and thematically analyzed to identify recurring patterns and insights related to clinical behavior. RESULTS: Confidence in managing oxygen desaturation significantly improved from a mean SD 3.78 2.12 to 6.20 2.02 across VR and HFS groups (t(163) = -14.04, P < .001), with no significant difference between groups (F(1,162) = 3.28, P = .07). Satisfaction was high overall (mean SD 6.07 1.02), but significantly greater in the HFS group than in the VR group (6.23 0.92 vs. 5.89 1.10; t(162) = 2.29, P = .02). HFS participants rated tutor guidance (6.49 0.86 vs. 6.10 1.02; P = .008) and authenticity (6.24 1.05 vs. 5.77 1.15; P = .006) higher, whereas both groups scored usability above 5 on all items. Qualitative analyses revealed complementary strengths. Interns valued VR for its immersive environment, focused repetition, and reduced distractions that facilitated stepwise problem-solving. HFS was praised for palpable realism, hands-on practice with equipment, and immediate feedback that reinforced team communication and role clarity. Across follow-up interviews, interns reported improved recognition of desaturation, more structured initial responses (airway assessment, oxygen delivery adjustments, escalation), and greater willingness to act promptly under pressure-suggesting perceived transfer of learning to clinical practice beyond the simulation lab. CONCLUSIONS: VR may complement HFS in emergency response training. Both modalities were associated with improvements in interns' self-reported confidence and perceived preparedness. The scalability and accessibility of VR suggest its potential value in diverse training contexts. CLINICALTRIAL: ClinicalTrials.gov NCT06295887.
Huh et al. (Mon,) studied this question.