Abstract
Background: Mandibular reconstruction is challenging for experienced and resident surgeons. Three-dimensional (3D) biomodeling creates accurate physical models of patients' craniofacial skeletons, which can potentially assist reconstruction. However, this capacity has not been objectively examined.
Objective: The purpose of this study was to assess 3D biomodels in performing and learning mandibular reconstruction through surgical simulation.
Design: Prospective cohort study.
Setting: Tertiary care academic referral centre.
Methods: Ten experienced and 10 naive resident surgeons were asked to bend and fixate a titanium reconstruction plate, for a standardized anterior hemimandibular defect, on a 3D biomodel by freehand or 3D biomodel-assisted means. Participants were randomized to which technique was performed first. Twenty-four to 48 hours later, participants performed the opposite technique.
Main Outcome Measures: Accuracy was measured by anterior mental projection and intercondylar and interangular splay. The results per technique were compared to a complete (control) mandible. The time of reconstruction and usability of each technique, as per an International Standards Organization-based questionnaire, were also determined.
Results: Three-dimensional biomodel-assisted reconstruction led to plates with statistically indifferent projection and splay compared to the control (p < .05) for both groups. Conversely, freehand constructs significantly deviated in projection and splay for either group (p < .05). No difference in reconstruction time by technique was found (p < .05). Usability favoured 3D biomodel-assisted bending, with significantly higher ratings in either group (p < .05).
Conclusions: Three-dimensional biomodels provide a usable and accurate means of mandibular reconstruction for experienced surgeons. Moreover, when used in surgical simulation, they provide an effective tool for teaching residents.