Abstract 4143354: Novel Methods Improve Assessment of Virtual Reality’s Impact on Physician Mental Models During Pre-Surgical Planning

Background: Virtual Reality (VR) has found numerous applications in healthcare, including 3D presentation of patient anatomy modeled from traditional medical imaging. While individual VR systems have been evaluated for their utility in pre-surgical planning, there remains a gap in understanding of the impact VR has on physicians’ mental models. VR is often assessed on performance criteria, omitting relevant perspectives of the physician’s experience. Research Questions: We sought to explore the following: 1) How does VR affect physicians’ perceptions of complex congenital heart cases, and 2) What strategies do physicians employ while exploring patient anatomy in VR. Goals and Aims: The aim of this study was to analyze how mental models change after reviewing anatomy in VR, and to describe common patterns of behavior in VR which provide deeper insight into the complex task of pre-surgical planning. Methods: We deployed a novel methodology of undirected think-aloud sessions in VR which we anchored to pre and post-session questionnaires based on the NASA Task Load Index. VR sessions were recorded in first-person view and were qualitatively analyzed by five independent coders using a grounded theory approach where codes emerged from the data. Iterative refinement of codes led to identification of multiple patterns of behavior which were summarized thematically. Pre and post-session surveys were analyzed statistically. Results: 10 VR sessions were captured from 3 physicians totaling 2.5 hours of recordings. Physician confidence in their understanding of the anatomy increased from a mean of 4/5 before VR to 5/5 after (p=0.012). Themes emerging from VR sessions included: 1) Exploration of the 3D environment led to confirmation of understanding and complex decision making, 2) Emotions expressed in VR reflected strengths and weaknesses of the VR tool, and 3) Confidence may mask complex decision making crucial to the surgical planning process. Reported mental, physical, and time demands of VR all remained low (<3/10 median) with high accomplishment of goals (10/10 median). Conclusion: Eliciting rich descriptions of physician perspectives of patient anatomy in VR provides a more detailed glimpse at the decision making and exploration which helps to refine mental models during pre-surgical planning. The 3D VR environment helps enable exploration of the anatomy which in turn leads to complex decision making and refinement of understanding of the case before surgery.