Enhancing OSCE Performance with AI-based Virtual Patients

Multi-Center Research

Problem Statement:

Traditional OSCE preparation using standardized/simulated patients (SPs) is effective but resource-intensive and difficult to scale with growing student cohorts. There is a need for more accessible, efficient training methods without compromising educational outcomes.

Research Hypothesis:

Training with AI-based virtual simulated patients (VSPs) can improve or match the effectiveness of traditional SP-based preparation for OSCE performance.

Aim:

To compare OSCE outcomes between students trained with traditional SP-based simulation and those using VSPs, focusing on clinical reasoning, diagnostic accuracy, and communication.

Methodology:

Three student groups will undergo distinct OSCE preparation modalities:

(1)  Control group receiving traditional SP-based training;
(2)  Research group A trained exclusively with AI-based VSPs;
(3)  Research group B receiving a blended approach: SP+VSP.

All participants will complete a baseline assessment, a post-intervention OSCE evaluated by blinded assessors, and a follow-up OSCE after three months to measure skill retention.

If you are interested in participation in this reseach please contact via "Contact Us" form below.