Measuring Stress and Perceptions for a Virtual Reality-Based Pericardiocentesis Procedure Simulation for Medical Training: Usability Study

Abstract

Background: Virtual reality (VR) is increasingly used in medical education, providing immersive environments for training in high-risk procedures such as pericardiocentesis. This lifesaving procedure requires technical precision and induces cognitive and physiological stress. Evaluating both usability and stress responses in a VR-based pericardiocentesis simulation is essential. Heart rate variability (HRV) serves as an objective stress marker, while prior VR experience may influence usability and stress perception.

objectives: This study aimed to assess the usability of a VR-based pericardiocentesis simulation, examine the relationship between usability perceptions and physiological stress (HRV), and determine the impact of prior VR experience on usability scores and stress responses.

Methods: A total of 119 final-year medical students participated in a VR pericardiocentesis simulation. Usability was evaluated using the System Usability Scale (SUS), the Post-Study System Usability Questionnaire, the Presence Questionnaire, and the Simulator Sickness Questionnaire. Physiological stress was assessed through HRV parameters, including the root-mean-square of successive differences (rMSSDs), percentage of differences greater than 50 ms (PNN50), low-frequency to high-frequency ratio, and nonlinear HRV indices (SD1/SD2 ratio, Poincaré area). Statistical analyses included descriptive statistics, Spearman correlations, and Mann-Whitney U tests to explore relationships between usability, stress, and prior VR experience.

Results: The VR simulation received a mean SUS score of 75.00 (SD 6.41; 95% CI 73.42-76.58), exceeding the general usability threshold of 68 (P=.002). The mean Post-Study System Usability Questionnaire score of 2.92 (SD 1.83; 95% CI 2.55-3.29) indicated moderate satisfaction, while the mean Presence Questionnaire score of 109.46 (SD 9.12; 95% CI 107.88-111.04) reflected strong immersion. Simulator sickness symptoms were mild (mean Simulator Sickness Questionnaire score 12.43, SD 15.41; 95% CI 9.28-15.58), although novice users reported significantly higher nausea levels (P=.02). Physiological stress analysis revealed a mean rMSSD of 281.27 (SD 98.99; 95% CI 259.45-303.09) ms and PNN50 of 56.85% (SD 19.70%; 95% CI 52.23%-61.47%), indicating moderate autonomic balance. A significant negative correlation was observed between HRV parameters (rMSSD and PNN50) and simulator sickness (P=.04; Spearman ρ=-0.23), suggesting that higher physiological stress was associated with increased simulator sickness symptoms. Prior VR experience was linked to higher usability scores (SUS +5.2; 95% CI 3.12-7.28; P=.03) and lower simulator sickness symptoms (P=.02) but did not significantly affect HRV markers.

Conclusions: VR-based simulations for high-risk medical procedures are effective training tools with high usability (SUS=75) and strong immersion. Simulator sickness correlated with physiological stress, emphasizing the need for design refinements to improve user comfort. Prior VR experience improved usability and reduced simulator sickness but did not significantly impact HRV markers. Future research should refine VR interfaces to balance immersion with minimized cognitive and physical discomfort.

Keywords: Presence Questionnaire; System Usability Scale; VR; heart rate variability; pericardiocentesis simulation; physiological stress; simulator sickness; usability assessment; virtual reality.

Figure 1.. Participant interacting with the virtual reality simulation for pericardiocentesis using Oculus Rift S. The simulation replicates the procedural steps required in a real clinical setting. The session lasted approximately 10 minutes and included real-time physiological stress monitoring via heart rate variability measurements.

Figure 2.. Virtual hospital environment for pericardiocentesis training. The virtual reality simulation presents a high-fidelity virtual hospital room containing a patient bed, vital signs monitor, procedural instruments, and interactive medical equipment necessary for pericardiocentesis. The immersive design was developed using Unity 2021.3.24f1 and aimed to enhance skill acquisition and clinical decision-making among medical students. (A) General view of the simulated room environment. (B) Lateral view of the scenario showing the patient and surrounding procedural equipment. (C) Opposite-side perspective providing a panoramic overview of the full training environment.

Figure 3.. Virtual hand interface used for procedural interactions. The virtual reality system uses diegetic interaction controls, where users interact with virtual elements directly replicating real-world medical tools. The hands, visible within the simulation, dynamically update to include surgical gloves when appropriate. This design eliminates elements not present in the real world such as teleportation mechanics or external heads-up display, ensuring a realistic and immersive training experience.

Figure 4.. User engagement within the virtual reality (VR) simulation. Still image from a recording of a participant performing pericardiocentesis inside the VR environment. The participant is navigating the simulation using handheld Oculus Rift controllers, interacting with virtual instruments to complete procedural steps. The VR session was part of a controlled study evaluating usability, immersion, and physiological stress responses in final-year medical students.

Figure 5.. Scatterplot: HRV (PNN50) versus Simulator Sickness. This scatterplot illustrates the correlation between HRV (PNN50) and Simulator Sickness scores, showing a negative trend where higher parasympathetic activity is associated with lower simulator sickness symptoms. HRV: heart rate variability; rMSSD: root-mean-square of successive differences; SUS: System Usability Scale.

Figure 6.. Scatterplot: HRV (root-mean-square of successive differences [rMSSD]) versus usability (System Usability Scale score). This scatterplot shows the lack of significant correlation between HRV (rMSSD) and usability scores, supporting the finding that physiological stress does not strongly influence usability perceptions. HRV: heart rate variability; PNN50: proportion of successive heartbeat intervals differing by more than 50 ms.