The Landscape of Virtual Reality and Generative AI in Oncology Education: A Scoping Review of Clinical Reasoning Assessment Tools

Summary

This scoping review finds VR and GenAI offer new ways to assess clinical reasoning in oncology training. However, both require more robust validation and ethical frameworks before widespread educational use, as current evidence is preliminary.

Abstract

Oncology is a setting in which clinical reasoning is a fundamental competency necessitating the synthesis of complex clinical information in a high-stakes setting. Conventional means of assessment, such as written tests, Objective Structured Clinical Examinations (OSCEs) and workplace-based assessments, offer valuable knowledge, but are still not effective in reflecting the dynamic and iterative quality of clinical reasoning. New opportunities to overcome these limitations are presented by emerging digital technologies, especially Virtual Reality (VR) and Generative Artificial Intelligence (GenAI). VR facilitates simulation-based environments, which can be immersive, and which allow the evaluation of procedural and situational factors of reasoning based on observable clinical action. GenAI, in turn, provides a more dialogue-based interaction, which enables assessing cognitive and analytical processes, including key cognitive processes. Literature shows that, these technologies have complementary advantages but are conceptually disjointed. VR mainly records behavioral performance but GenAI emphasizes cognitive processes, and the two are not as integrated. Moreover, the present evidence base is typified by methodological heterogeneity, dependence on preliminary research, and lack of validity evidence, particularly regarding reliability and real-world impact. Other factors to take into account are effect of cognitive load, user variability, and ethical issues, e.g., algorithmic bias and transparency. The feasibility of implementation is further influenced by organizational factors such as infrastructure and institutional readiness. In general, VR and GenAI are encouraging, yet developing methods of assessment of clinical reasoning in oncology. In the future, the evolution of the field will require coherent assessment frameworks, rigorous validation, and clear ethical structures.

Introduction

Clinical reasoning, the cognitive and metacognitive processes underlying diagnosis and management, constitutes the cornerstone of safe and effective oncology practice¹. It is a distinctively complicated process in the context of oncology, as clinicians must navigate probabilistic decision-making in the face of uncertainty, synthesize quickly changing evidence of therapies, and participate in multidisciplinary discussions in tumor boards. Moreover, oncologists have to find a balance between biomedical and patient-centered thought, including telling patients of their poor prognoses, dealing with ethical issues, and adjusting treatment plans to th....