Visualizing Multi-Step Decision-Making at a Glance: Pairing Choose Your Own Adventure Style Simulated Cases with a Novel Mapping Framework

TL;DR

This study combines interactive adventure-style medical cases with visual mapping to better assess and understand clinical decision-making in medical education.

Contribution

A novel framework that pairs CYOA-style cases with dynamic mapping to visualize and evaluate multi-step clinical reasoning.

Findings

25.9% of users achieved expert-level decision-making in the simulated cases.

51.9% of users made at least one incorrect decision leading to adverse outcomes.

Visualizations enabled rapid comparison of decision patterns across different cohorts.

Abstract

Clinical reasoning is a critical component of medical education, yet assessing decision-making in real-world scenarios remains challenging. Traditional assessment methods, such as multiple-choice questions, fail to replicate the dynamic nature of clinical problem-solving, where decisions build upon one another over time. Choose Your Own Adventure (CYOA) Case studies offer a promising alternative by simulating non-linear, multi-step decision-making. This study aimed to pair CYOA-style clinical cases with dynamic mapping tools to visualize user decision-making patterns and Evaluate their effectiveness as a pedagogical tool. We developed CYOA branched narrative case studies to simulate clinical decision-making in the management of high-dose methotrexate treatment for patients with pediatric leukemia. Following completion of the case study, user decision-making was mapped using R-based DiagrammeR software, which visualized the paths taken by each participant through the case studies. The maps effectively visualized decision-making at each stage of the CYOA cases, revealing a variety of user pathways and decision-making behaviors. Analysis showed that 25.9% of users achieved expert-level decision-making, whereas 51.9% made at least one incorrect decision that led to adverse outcomes. Furthermore, the maps enabled comparisons of behavior between different cohorts, highlighting how various interventions influenced decision patterns within the case study. Notably, these visualizations and data tables were generated in under ten seconds, providing rapid insights into user performance. This study demonstrates the potential of combining CYOA-style case studies with dynamic mapping tools to assess clinical reasoning in a more nuanced and interactive way.

The online version contains supplementary material available at 10.1007/s40670-025-02505-6.

Figures