25 June 2026
University of Deusto
As part of the Bachelor’s Degree in Nursing at the University of Deusto, an innovative pilot experience based on generative artificial intelligence (AI) has been tested experimentally this academic year, with the aim of enabling students to develop their clinical competences and communication skills in a digital environment before facing real patient scenarios.
At present, this is an initial, local trial limited to two subjects, with the aim of verifying that the tools work and thus considering the possible development of a teaching innovation project which, if approved by the University’s Educational Innovation Unit, could be launched in the future and implemented, for the time being, in the Nursing Methodology subject.
To this end, the capacity of AI to simulate human conversation has been harnessed, using retrieval-augmented generation (RAG) technology, whereby the AI is fed exclusively with closed, pre-validated clinical data in order to ensure the rigour of nursing practice.
Fictitious patients
The first phase of the pilot experience focused on third-year students through a conversational chat in which the AI always assumed the role of a fictitious patient, while the student took on the role of nurse.
To replicate the complexity encountered in healthcare settings, the behaviour of the virtual patient was predefined in detail. Students were randomly presented with a wide range of profiles, from more aggressive or difficult-to-manage patients to others with a high level of knowledge about their own condition.
As this was an initial introduction, the activity was offered on a voluntary, individual basis within a single subject and at a specific point in the course. This approach is intended to promote personalised learning, as each student independently decided which strategy and conversational pathway to follow in order to resolve the clinical case. The teaching staff then collected the chat transcripts in order to analyse the optimal interview approach for obtaining the necessary clinical information.
Clinical case records for second-year students
In addition, second-year students took part in another pilot experience as part of one of their simulation workshops. In this case, a bespoke software programme was developed to recreate an electronic clinical records environment.
In this case, the challenge was analytical in nature, as students were required to apply their clinical reasoning to review the medical history and complementary test results of the virtual patient, identifying which data were truly relevant for resolving the case.
An evolving patient community
Given the positive response to this teaching methodology among students, the next objective is to integrate the AI-powered conversational chat with the electronic health record system to create a constantly evolving community of fictional patients. The future aim is for interactions across different years of the Nursing programme to affect the same patient. For example, if a third-year student indicates a treatment in the interactive chatbot, a second-year student will be able to see that change reflected, along with its effects, in the patient’s clinical record, thereby closely simulating the continuity of care that may occur in a real healthcare centre or hospital setting.
As this is a pilot experience limited to two subjects, the technological infrastructure for this project was developed in-house within the Nursing degree itself, using the Python and R programming languages, driven by the use of open-source tools.
