Laparoscopic skills training: the effects of viewing mode (2D vs. 3D) on skill acquisition and transfer

Background Three-dimensional (3D) visual displays have been suggested to aid laparoscopic skills training by providing the depth cues not present in traditional two-dimensional (2D) displays. However, few studies have robustly investigated the impact of viewing mode (2D vs. 3D) on learning outcomes. Purpose To examine how viewing mode (2D vs. 3D) impacts the acquisition and transferability of basic laparoscopic skills by comparing performance between transfer and control groups on a complete profciency-based training program. Method A counterbalanced between-subjects design was employed. Each participant was randomly allocated to one of four groups, comprising two transfer groups (trained in one viewing mode and tested in the alternate mode: the 2D→3D and 3D→2D groups) and two control groups (trained and tested in one viewing mode: the 2D→2D and 3D→3D groups). Participants completed profciency-based training in six laparoscopic training tasks. Testing included two further repetitions of all tasks under test conditions. Objective performance measures included the total number of repetitions to reach profciency, and total performance scores (i.e. time+error penalties across all repetitions) in training and testing. Results The groups trained in 3D demonstrated superior training performance (i.e. less time+errors) and took fewer repetitions to reach profciency than the groups trained in 2D. The groups tested in 3D also demonstrated superior test performance compared to those tested in 2D. However, training mode did not yield signifcant test diferences between the groups tested in 2D (i.e. 2D→2D vs. 3D→2D), or between the groups tested in 3D (i.e. 3D→3D vs. 2D→3D). Conclusion Novices demonstrate superior performance in laparoscopic skills training using a 3D viewing mode compared to 2D. However, this does not necessarily translate to superior performance in subsequent testing or enhanced learning overall. Rather, test performance appears to be dictated by the viewing mode used during testing, not that of prior training