Summary

This prospective, controlled and multi centric study about the transfer validity between VR simulator and real condition demonstrate that the simulator is not sufficient to learn a complex and sequential procédure like double row

Abstract

Purpose

To investigate the transfer validity of an arthroscopic cuff repair from virtual reality simulation to real conditions.

Methods

Thirty orthopedic residents and first year orthopedic fellows were enrolled in our study from six French university hospitals. At baseline, they were asked to answer a survey about the number of arthroscopic procedures they performed as an operating assistant or main surgeon. The attendees were also asked about their confidence in performing an arthroscopic cuff repair. Following this survey, they were randomized in two different groups: virtual reality training (VR+) or no virtual reality training (VR-). At the beginning of our study, both groups received one session of training on the simulator ArthroS® (VirtaMed AG, Zurich, Switzerland) and a theoretical course on cuff repair. Thereafter the VR+ group underwent a monthly based training program with increasingly difficult procedures. The training sessions were standardized, lasted 1 hour, and were performed under the supervision of the same assessor. After 6 months, all the attendees were asked to perform an arthroscopic double row supraspinatus repair on a 3D printed shoulder model using real anchors. Two independent and blinded shoulder specialized surgeons rated the attendees using the Arthroscopic Surgical Skill Evaluation Tool (ASSET) score. Statistical analysis was performed with a non-parametric t-test Mann-Whitney was used. The results are expressed with median and standard deviation.

Results

Both groups were initially comparable. There is no significative difference between the two group with the total ASSET score (63+/-4,7 versus 58+/-8,3, p=0,11). The VR+ group presented a better ASSET Global rating scale than the VR- group (35+/-3,1 versus 32+/-5,7, p=0,046). The task specific checklist for the procedure did not differ between VR+ and VR- (26,8+/-2 versus 25,6+/-2,8, p=0,24). The VR+ group was faster than the VR- group (1705+/-215 seconds versus 1890+/-247 p= 0.0063).

Conclusion

A monthly based VR program of 6 months improves the arthroscopic performance of orthopedic surgeons in formation but is not sufficient to master a procedure