Surgeon estimation of retropubic trocar position in blind 3D space

Retropubic midurethral sling surgery involves the blind passage of trocars near vital organs. We quantified the proximity of surgeons' mental representation of trocar position relative to actual position using a pelvis simulation platform. We hypothesized that novice surgeons, compared with experts, would estimate the trocar's location to be further from the actual location.

Novice and expert surgeons performed bilateral retropubic trocar passes of a Gynecare TVT trocar (#810041B-#810,051) on the simulation platform. We measured the trocar tip's position using a motion capture system, and recorded vocalizations when they perceived contacting the bone and crossing three landmark-oriented planes. We calculated differences (∆_Bone, ∆_Turn, ∆_Top, ∆_Pop) between vocalization times and when the trocar crossed the corresponding plane. We performed Mann-Whitney and Chi-squared tests to investigate differences between novices and experts and Levene's test to assess equality of variances for subject-level variation.

A total of 34 trials, including 22 expert and 12 novice trials, were performed by six participants. ∆_Bone was significantly smaller among novice surgeons (1.27 vs 2.81 s, p=0.013). There were no significant differences in the remaining three deltas or in vocalizing early versus late. Levene's test revealed no significant differences in within-subject variability for any of the four deltas. Novices passed the trocar anterior to the pubic bone on three passes.

Novices were similar to expert surgeons in their estimation of the trocar's location and may have relied more heavily on anticipatory mechanisms to compensate for lack of experience. Teaching surgeons should make sure the novice surgeon trocar pass starts posterior to the bone.