"It Talks Like a Patient, But Feels Different": Co-Designing AI Standardized Patients with Medical Learners

TL;DR

The paper investigates how medical learners experience standardized patient (SP) training and how to design AI-based SPs (AI-SPs) that balance standardization, clinical realism, and relational authenticity. Through 12 clinical-year student interviews and three co-design workshops, it identifies six learner-centered needs and translates them into six AI-SP design requirements, culminating in a conceptual workflow for AI-SPs that supports deliberate practice with multimodal evidence gathering and dual-loop feedback. The findings argue that instructional usability, including mode-specific fidelity and transparent rules, drives trust and educational value more than raw conversational realism. Practically, AI-SPs are positioned as complementary educational infrastructure that enables scalable, on-demand practice, while maintaining alignment with real-world clinical goals and training ecologies.

Abstract

Standardized patients (SPs) play a central role in clinical communication training but are costly, difficult to scale, and inconsistent. Large language model (LLM) based AI standardized patients (AI-SPs) promise flexible, on-demand practice, yet learners often report that they talk like a patient but feel different. We interviewed 12 clinical-year medical students and conducted three co-design workshops to examine how learners experience constraints of SP encounters and what they expect from AI-SPs. We identified six learner-centered needs, translated them into AI-SP design requirements, and synthesized a conceptual workflow. Our findings position AI-SPs as tools for deliberate practice and show that instructional usability, rather than conversational realism alone, drives learner trust, engagement, and educational value.