Immersive virtual reality versus traditional instruction for long‐term retention of cardiopulmonary resuscitation skills in medical student education: A randomized controlled trial

Abstract Background The rapid decay of cardiopulmonary resuscitation (CPR) skills remains a pervasive challenge in medical education. Although immersive virtual reality (VR) is increasingly used for training, its efficacy in mitigating long‐term skill decay compared to traditional methods remains unclear. Furthermore, the cognitive mechanisms—specifically how VR‐induced cognitive load influences learning outcomes—require further elucidation. Methods A randomized controlled trial was conducted with 128 novice medical students, randomly assigned to either a VR training group ( n = 64) or a traditional demonstration‐based control group ( n = 64). Both groups received identical content based on American Heart Association guidelines. Outcomes included theoretical knowledge (multiple‐choice questionnaire) and procedural skills (checklist), assessed immediately post‐training and at a 1‐month follow‐up. Cognitive load was measured using the NASA task load index (NASA‐TLX) instrument immediately after training. Results Although immediate theoretical knowledge scores showed no significant difference between groups ( p = 0.664), the VR group demonstrated significantly superior procedural skill retention after 1 month. The Control group experienced a substantial skill decay (−12.05 points), whereas the VR group maintained high proficiency with minimal decay (−4.36 points; p < 0.001). Notably, NASA‐TLX data revealed a cognitive divergence: the VR group reported significantly higher mental demand (56.33 vs. 45.59; p < 0.001) but significantly lower frustration levels (35.38 vs. 61.02; p < 0.001) compared to the Control group. Conclusion Immersive VR training is superior to traditional instruction in promoting the long‐term retention of CPR skills, effectively mitigating the “forgetting curve.” The coexistence of high mental demand and low frustration suggests that VR successfully converts cognitive effort into effective germane load, facilitating deep learning through embodied cognition. These findings support the integration of VR as a scalable tool for high‐frequency, low‐dose resuscitation retraining.