Custom-built web application integrating an AI chatbot into a three-part diagnostic reasoning case¶
Design · All designs
Evidence · no claims cited
Description¶
A custom-built web application integrated a GPT-4o-powered AI chatbot into a three-part acute kidney injury case requiring iterative Bayesian reasoning. Within each part of the case, students first ranked their differential diagnoses without AI access, and the integrated AI chatbot then became available for students to refine their reasoning and adjust their rankings on a voluntary basis. Students could submit up to 10 prompts across the whole case.
Design Implications¶
Context¶
Requirements¶
- Students must select and include relevant clinical information in their prompts, as the chatbot did not have automatic access to the clinical case data
Constraints¶
- The chatbot was powered by GPT-4o, the most capable model available at implementation in August 2025
Target Learners¶
- Second-year pre-clerkship medical students
Learning Goals¶
- Diagnostic reasoning
- Differential diagnosis ranking
Claims¶
Related Designs¶
- Aspen: a prompt-engineered AI chatbot peer mentor for inquiry revision
- Jordan Chatbot: a unit-specific GPT-4o pedagogical support agent grounded in constructivist and socio-cultural learning theory
- Custom-built online waves module with simulations integrated with a chatbot
- Custom LLM practice platform with validated offloading classifier for extent-of-support labeling
Examples¶
-
Key Sources¶
- Hayden Robert, Blanco Maria, Ramesh Saradha, Epstein Scott, Trowbridge Robert. (2026). Diagnostic reasoning with and without AI: automation bias in pre-clerkship medical students. BMC Medical Education. https://doi.org/10.1186/s12909-026-09979-x
Page metadata
| Field | Value |
|---|---|
type |
design |
id |
ai-integrated-diagnostic-reasoning-web-application |
title |
Custom-built web application integrating an AI chatbot into a three-part diagnostic reasoning case |
description |
A custom-built web application integrated a GPT-4o-powered AI chatbot into a three-part acute kidney injury case requiring iterative Bayesian reasoning. |
status |
review |
generated |
by process:settle-candidates on 2026-11-04 |
trust |
unverified — no one has confirmed this page's content |
sources |
1 — hayden-robert-2026 |
Frontmatter is the page's machine-readable record — see the Schema & Guide.