Header image for Vitalis 2026
Profile image for AI-based tool for healthcare professional training in person-centred conversations [PCC121]

AI-based tool for healthcare professional training in person-centred conversations [PCC121]

Tuesday May 5, 2026 12:00 - 11:15 Poster Arena

Presenter: Malin Bengtsson

Track: Poster

Background  Conversations between patients and healthcare professionals is fundamental to person-centred care. This approach requires healthcare professionals to be skilled in active listening and able to adapt to both the clinical setting and the individual needs of the conversation partner. However, there is a lack of specific training opportunities in person-centred conversations. Artificial intelligence (AI) offers new opportunities for realistic scenario-based training with individualized performance feedback. An AI-based tool for training person-centred conversation targeting healthcare professionals is currently under development within an innovation project. An interdisciplinary team conducts research on the development as well as explores the relevance and usability of the tool. The focus of this presentation is the co-design and iterative evaluation of this AI-based training tool.   Method  The design and evaluation follows the ADDIE model for instructional design and The UK Medical Research Council’s structured framework for developing and evaluating complex interventions, comprising three phases: development of realistic patient cases and AI-generated feedback; usability testing; and evaluation of effect. Tool development is user-centred and conducted in collaboration between tech developers, researchers, healthcare professionals and patient representatives from three care settings. This presentation will focus on the first two phases.  Results  Fifteen patient cases have been developed, aimed at conversations in primary health care, psychiatric care, and palliative care settings. Initial testing with healthcare professionals and patient representatives indicates that the patient cases are perceived as authentic and realistic. However, refinements are needed, including adjustments in complexity and sequencing. This presentation will highlight the development process and preliminary findings from the usability testing phase.  Conclusion  AI-based communication training shows promise in meeting the need for realistic training opportunities in healthcare. The tool has the potential to enable safe practice of difficult conversations. 
Language

English

Conference

GCPCC

GCPCC Code

PCC121

Lecturers

Profile image for Malin Bengtsson

Malin Bengtsson Presenter

PhD student
Sahlgrenska University Hospital

Malin Bengtsson, Jesper Fransson, Stina Nyblom, Ylva Hård af Segerstad, Ramona Schenell, Jakob Wenzer, Joakim Öhlén, Emma Forsgren