Psychomotor skills and cognitive load training on a virtual reality laparoscopic simulator for tubal surgery is effective.

OBJECTIVES

Validation of a virtual reality (VR) simulator for the training and assessment of laparoscopic tubal surgery and mapping of cognitive load.

STUDY

DESIGN

Prospective cohort study conducted at the Imperial College Virtual Reality Surgical Skills laboratory amongst 25 trainees and nine senior gynaecologists. Participants performed two sessions of salpingectomy and salpingotomy procedures on a VR simulator to assess construct validity. Nine novices performed ten such sessions to enable assessment of the learning curve. The relationship between cognitive load and the dexterity parameters was assessed. Simulator fidelity was reported by experienced and intermediate level gynaecologists. Statistical analyses utilised non-parametric tests, Kruskall-Wallis and Mann-Whitney U tests. Learning curves were assessed using the Friedman test and Wilcoxon Signed Ranks test. Relationship between dexterity metrics and cognitive load was performed using Spearman's rank order correlation.

RESULTS

Salpingectomy demonstrated construct validity for time taken by experienced, intermediate and novice gynaecologists (median 170 vs. 191 vs. 313s (P=0.003) respectively) and movements (median 200 vs. 267 vs. 376s, P=0.045). Salpingotomy demonstrated construct validity for time taken (median 183 vs. 191 vs. 306s, P=<0.001) and movements (median 210 vs. 233 vs. 328s, P=0.005). Learning curve analysis for salpingectomy displayed a plateau for time taken after the eighth session, and the fourth session for movements. Salpingotomy displayed a plateau after the eighth session for both time taken and movements. Cognitive load correlated significantly with dexterity parameters. The fidelity scores were not significantly different between the two procedures (P=0.619).

CONCLUSION

The LAP Mentor VR laparoscopic simulator is a valid and effective tool for training novice surgeons in ectopic pregnancy surgery. Reduction in cognitive load significantly correlates with the learning curves.