Comparison of learning curves on the FLS and the Virtual Basic Laparoscopic Skill Trainer (VBLaST): pattern cutting and ligating loop tasks

Ali M Linsk, MD1, Emily E Diller, BS2, Kimberley R Monden, PhD3, Ganesh Sankaranarayanan, PhD3, Woojin Ahn, PhD4, Christopher S Awtrey, MD5, Suvranu De, ScD4, Daniel B Jones, MD5, Steven D Schwaitzberg, MD6, Scott K Epstein, MD7, Jesse M Rideout, MD, MPH7, Caroline GL Cao, PhD2. 1Beth Israel Deaconess Medical Center, Cambridge Health Alliance, 2Wright State University, 3Baylor University Medical Center at Dallas, 4Rensselaer Polytechnic Institute, 5Beth Israel Deaconess Medical Center, 6University at Buffalo School of Medicine and Biomedical Sciences, 7Tufts University School of Medicine

INTRODUCTION: VBLaST is a virtual reality-based simulator developed to replicate the FLS tasks. This study compared the learning curves of the pattern-cutting (PC) and ligating-loop (LL) tasks on the FLS trainer box with those on the VBLaST.

METHODS AND PROCEDURES: Forty-eight medical students were randomly assigned to one of six conditions: control PC (no training), training FLS-PC, training VBLaST-PC, control LL (no training), training FLS-LL, and training VBLaST-LL. The training subjects performed 150 trials over a three-week training period. Pretests, post-tests (immediately following training trials), and retention tests (two weeks after post-tests) on both FLS and VBLaST for the assigned task were also done.

Non-linear regression analysis was used to fit an inverse curve to the learning curve data for both completion time and normalized score. The average learning rate was calculated for each task on both simulators, and the performance plateau on the learning curves was determined. The learning plateau is the time to task completion or score at the asymptote of the curve and the learning rate is the number of trials required to reach 90% of the plateau.

RESULTS: Pattern-cutting task: The learning curves on the FLS and the VBLaST simulators were similar. Subjects reached the performance plateau at similar rates based on time to task completion (74.72 sec, 21 trials on FLS; 90 sec, 22 trials on VBLaST), and task score (77.93 score, 10 trials on FLS; 74 score, 10 trials on VBLaST).

Ligating-loop task: The learning curves on the FLS and the VBLaST simulators were also similar. Subjects who trained on the FLS reached a lower performance plateau (35.51 sec) than the VBLaST-trained subjects (29.94 sec), but at a faster rate (28 trials vs. 37 trials). Based on their scores, the FLS-trained subjects were able to reach a slightly higher score at performance plateau (101.11 score, 7 trials on FLS; 88.71 score, 8 trials on VBLaST) at the same rate.

CONCLUSIONS: Virtual reality training using the VBLaST simulator is equally effective in fostering laparoscopic skill acquisition when compared to the gold standard of the FLS box trainer. VBLaST offers additional advantages such as real-time feedback, instant scoring, and provides the ease of repeated trials without needing expensive consumables and trained proctors.

« Return to SAGES 2016 abstract archive