Introduction:
Mandibular reconstruction with microvascular free tissue transfer is a competency of head and neck microvascular surgery. Given the critical nature of the procedure, learning opportunities can be limited in the operating room. Simulation has been a tool used increasingly for surgical training. We thus explored whether a virtual surgical planning simulation could be beneficial to teach residents mandibular reconstruction with a fibular graft. The utility of discovery learning was examined and whether the order of didactic instruction and hands on experience affected learning outcomes.

Methods: Surgical residents and fellows at a tertiary care centre were enrolled in a mandibular reconstruction simulation course in 2019 and 2021. An in-house virtual surgical planning model was used for the simulation. Construct and face validity were validated. Both cohorts received didactic instruction with demonstration and hands on practice. One cohort received hands-on practice first (Do then See) while the other received didactic instruction first (See then Do). Participants then performed guided mandibular reconstruction with a simulated mandible and fibula model. Outcomes included pre-simulation and post-simulation surveys assessing difficulty and understanding of reconstruction, and learner confidence on a Likert scale from 1-7. Accuracy of reconstruction was measured by assessing differences in length, width and projection of reconstructed mandibles compared to an ideal reconstructed model.

Results: 10 learners performed Do then See while 8 learners performed See then Do. Learners rated that models had dimensions and a layout similar to those found in a real patient (5.94) confirming face validity. There was no significant difference between accuracy in reconstruction between junior and senior residents when comparing differences in width (0.86 vs 0.88 cm, p=0.94), length (0.65 vs. 0.55 cm, p=0.71) or projection (0.98 vs. 0.65 cm, p=0.20). Learners had significantly higher post-simulation vs. pre-simulation ratings on the likelihood of performing a mandibular reconstruction adequately in Do then See (4.98 vs 3.80, p=0.01) and See then Do (5.22 vs 3.70, p<0.01). There was no significant difference between post-simulation ratings on likelihood of performing adequate reconstruction when comparing Do then See to See then Do (4.98 vs. 5.21, p=0.24). Accuracy was not significantly different between Do then See and See then Do when assessing width (0.59 vs 1.03cm, p=0.07), length (0.66 vs. 0.61cm, p=0.87) and projection (0.77 vs. 0.87cm p=0.72)

Discussion: The face validity of this low cost simulation was confirmed. Construct validity was not confirmed as junior and senior learners performed similarly on the reconstruction. However, this may be as a result of the sample size and limited experience of senior learners in performing mandibular reconstructions.Learners rated a perceived increase in understanding of mandibular reconstruction, anatomy of reconstruction and technical skills. There was no difference in outcomes between the two learning groups. Simulation is an effective tool to teach learners basic fundamental skills and allows opportunities to gain experience in high stakes or rare procedures. Overall, this demonstrates the potential learning benefits of a simulation model for mandibular reconstruction. Further studies with a larger sample size would strengthen the support for this model.