Validation of a Novel Virtual Reality Training Curriculum for Robotic Cardiac Surgery: A Randomized Trial

Objective: Robotic cardiac surgery training has relied entirely on classical methods of surgical teaching. We sought to evaluate the impact of a virtual reality (VR) simulation curriculum to improve skill acquisition in robotic cardiac surgery.

Methods: We randomly assigned 20 surgical trainees to undergo a 9-exercise VR curriculum on a robotic surgical simulator or a control group that received no additional training. The trainees were then evaluated in a blinded fashion by assessing their de-identified video recordings of the following: (1) standardized robotic internal thoracic artery harvest and (2) mitral valve annuloplasty performed in porcine models, using a validated time-based scoring system and an objective intraoperative scoring tool. Postintervention assessments were compared to baseline.

Results: Trainees randomized to the VR group were faster than the control group for both the internal thoracic artery harvest (957.3 ± 98.9 vs. 749.1 ± 171.9; P = 0.004) and mitral annuloplasty (580.4 ± 14.4 vs. 463.8 ± 86.4; P < 0.001) and scored significantly higher with the intraoperative scoring tool (22.8 ± 2.7 vs. 11.0 ± 4.5; P < 0.001). Additionally, the VR group achieved a proficiency level similar to our experts for both time-based scores (P = 0.624 and P = 0.967), and the intraoperative assessment (P = 0.110), whereas the control group was not able to meet this level of proficiency for any of the primary outcomes. The average duration of training to successfully complete all required tasks was 9.3 hours.

Conclusions: We have demonstrated that a VR simulation curriculum can significantly improve the efficiency and quality of learning in robotic cardiac surgery. Further evaluation of this curriculum is required for its widespread implementation in surgical training (ClinicalTrials.gov, NCT#02357056).