AI-Generated Patients in Suicide Assessment Skills Training: The Prometheus Educational Innovation Project

Abstract

Background: Advanced training in suicide prevention requires well-developed skills in risk detection and clinical interviewing. Teaching these abilities using realistic contexts helps improve the understanding and integration of content. However, this may entail risks in sensitive areas such as suicidal behavior. This project seeks the development of virtual patients to support training in interviewing skills in the context of suicide.

Methods: This educational innovation project comprises several phases. First, focus groups will be conducted with professionals and health sciences students to identify perceived training needs. Subsequently, with Metamedics and professors from UDIMA’s Master’s in Suicide Prevention, ten clinical cases will be developed, each one focused on training specific abilities in suicide prevention. These virtual patients will function as chatbots with which students can interact through voice or text. To assess competency acquisition and satisfaction, a pre–post design will be implemented across participating classrooms. Competence will be evaluated using standardized clinical case scenarios, while satisfaction will be measured using the Students’ Satisfaction and Self-Confidence Scale and the Expectations Fulfilment Scale.

Expected results: The project is expected to have a positive impact on students by fostering professional competencies and improving self-confidence and satisfaction. AI-based, deliberate practice with clinical cases is anticipated to yield better outcomes than traditional strategies.

Discussion: Deliberate practice with virtual patients may provide a safe and immersive, environment for developing communicative and clinical competencies in suicide risk interviewing. Additionally, it can achieve this without requiring direct exposure to crisis situations, increasing ethical safeguards for both patients and students.