Laparoscopic virtual reality training: Are 30 repetitions enough?

Background. Current literature suggests that novices reach a plateau after two to seven trials when training on the MIST VR laparoscopic virtual reality system. We hypothesize that significant benefit may be gained through additional training. Materials and methods. Second-year medical students (n = 12) voluntarily enrolled under an IRB-approved protocol for MIST VR training. All subjects completed pre- and posttraining questionnaires and performed 30 repetitions of 12 tasks. Performance data were automatically recorded for each trial. Learning curves for each task were generated by fitting spline curves to the mean overall scores for each repetition. Scores were assessed for plateaus by repeated measures, slope, and best score. Results. On average, subjects completed training in 7.1 h. (range, 5.9-9.2). Two to seven performance plateaus were identified for each of the 12 MIST VR tasks. Initial plateaus were found for all tasks by the 8th repetition; however, ultimate plateaus were not reached until 21-29 repetitions. Overall best score was reached between 20 and 30 repetitions and occurred beyond the ultimate plateau for 9 tasks. Conclusions. These data indicate that a lengthy learning curve exists for novices and may be seen throughout 30 repetitions and possibly beyond. Performance plateaus may not reliably determine training endpoints. We conclude that a significant and variable amount of training may be required to achieve maximal benefit. Neither a predetermined training duration nor an arbitrary number of repetitions may be adequate to ensure laparoscopic proficiency following simulator training. Standards which define performance-based endpoints should be established. (C) 2004 Elsevier Inc. All rights reserved.