Surgical resident performance on a virtual reality simulator correlates with operating room performance.

Background: To define the ability of a virtual reality (VR) simulator to reflect clinical skill in surgical residents, we compared clinical laparoscopic performance and contemporary lab performance during curricular VR skills training.

Methods: Nine postgraduate year (PGY) 1 and 2 surgical residents were assessed during laparoscopic cholecystectomies and appendectomies using a web-based interactive database (OpRate)over a 6-mo period. Operative performance data were collected at the conclusion of procedures (mean responses of attending surgeons in nine areas pertaining to resident preparedness and technical skill). During this period, all residents undertook iterative laparoscopic training using a new VR trainer (SEP: SimSurgery AS, Oslo, Norway; METI, Sarasota FL). OpRate performance over 4-wk blocks and closest VR performance data (mean time, path length, and errors for three iterations of six basic skills tasks) were tested for correlation by linear (Pearson) correlation method.

Results: Residents performed 1 to 6 operative cases each (median = 3) during time blocks used for comparisons (median separation operative and SEP performance data 18 d). Significant correlation of operative and VR scores was found for time to task completion in 5 of 6 VR tasks. Results were most significant for a gallbladder dissection task (P = 0.0066, correlation coefficient = -0.6671). No significant correlation of path length or error data and operative performance was observed for any VR task.

Conclusions: These data indicate that time to task completion on a VR training device correlates with resident performance in the clinical operating room. Serial evaluations will determine if concurrent performance improvement over time can be demonstrated.