B. J. Goudreau1, Y. Hu1, H. Kim1, E. Mace1, N. S. Schenkman1, L. A. Cantrell1, P. W. Smith1, S. K. Rasmussen1, P. T. Hallowell1 1University Of Virginia,Charlottesville, VA, USA
Introduction: Robotic surgery is becoming a fundamental component of general surgery practice and training. Although a sophisticated virtual reality trainer exists for the da Vinci robot, adaptive methods of measuring proficiency are lacking. The objective of this study was to use cumulative sum (Cusum) to characterize the learning curves of surgical novices and experts using the dV-Trainer robotic virtual simulation system.
Methods: Ten medical students enrolled in a robotic surgery training course completed 10 practice attempts on three tasks using dV-Trainer: Peg-Transfer 2, Ring-Walk 2, and Needle-Driving 1. Simulation proficiency was tracked using Cusum. The students were evaluated via pre- and post-tests using the da Vinci Si robot and two inanimate models: peg-transfer and ring-and-rail. For comparison, Cusum proficiency curves were generated for 7 surgical faculty and 9 surgery residents as they performed repetitions on the same three dV-Trainer tasks.
Results:Students’ scores on peg-transfer (87.3 vs 66.1, p = 0.004) and ring-and-rail (93.4 vs 77.4, p = 0.002) significantly improved following virtual training. Participants who remained sub-proficient by Cusum criteria at the Peg-Transfer 2 and Ring-Walk 2 virtual tasks were ranked 8thand 10thon post-testing for the corresponding inanimate models. Among residents, only one was sub-proficient on Peg-Transfer 2, while half were sub-proficient on Ring-Walk 2 and Needle-Driving 1. Experienced surgeons were Cusum-proficient on all tasks except for one surgeon who was sub-proficient on Needle-Driving 1.
Conclusion:Inherent variability exists in robotic surgery aptitude. Using Cusum to track training progress on basic technical skills can identify learners who are in need of more rigorous training.