Low-complexity emergency cases at Arnau de Vilanova University Hospital have decreased by 11% thanks to the artificial intelligence-based self-triage tool

The Arnau de Vilanova University Hospital in Lleida (HUAV) has recorded an 11% decrease in low-complexity A&E visits, a year after introducing an artificial intelligence-based self-triage tool. This is demonstrated by the findings of a research study conducted by professionals from HUAV, the ERLab, recerca en urgències i emergències research group of the Lleida Biomedical Research Institute (IRBLleida), the Open University of Catalonia (UOC) and the University of Lleida (UdL), was recently published in the European Journal of Emergency Medicine.

The tool, available on the hospital's website and in the Salut Lleida app, offers the public an immediate medical guidance system based on symptoms. Upon completing a short questionnaire, the person receives a recommendation on which healthcare service is most appropriate: a Primary Care Urgent Care Centre (CUAP), their local health centre (CAP), or the hospital's Accident & Emergency department.

The study confirms that after three consecutive years of increases in low-complexity A&E visits, the 2024-2025 period has recorded a significant reduction. Level V (lowest complexity) A&E visits have decreased by 11.4% and level IV A&E visits by 4.5%. Overall, low-complexity visits (levels IV and V) have fallen by 6%, equivalent to 1,215 visits avoided in four months.

In total, the hospital attended to 109,000 A&E cases in 2024, of which 49% were low-complexity. The decrease is considered a key indicator of organisational efficiency.

The AI-based self-triage tool has been provided by Mediktor, in collaboration with the TIC Salut Social Foundation and the eHealth Center of the Open University of Catalonia (UOC).

Over a thousand people directed to the appropriate service

During the first year, 1,035 people used the tool, completing the self-triage process. Of these, 43% were redirected to a CUAP, avoiding an unnecessary trip to the hospital. People who are advised to go to A&E and arrive with the generated QR code are prioritised in face-to-face triage. At the same time, coordination with the CUAP has been strengthened to ensure consistency throughout the entire emergency care pathway.

The HUAV is the first hospital in the State to integrate an AI self-triage system into a winter contingency plan, with the aim of anticipating patient flows before they arrive at the Accident & Emergency department. This strategy contrasts with the conventional flow-inversion model (redirecting patients once they are already in A&E) and has proven to be more efficient, comfortable and sustainable.

This model of pre-triage has already been established in other European countries (the United Kingdom, France, Switzerland), but the Lleida experience stands out for incorporating adaptive AI, which is more flexible than the rigid algorithms commonly used in other systems.

The reduction in minor visits has allowed for less pressure on services and an improved ability to respond to serious cases, as well as offering the public more accurate guidance, avoiding unnecessary waiting and travel. The most common reasons for consultation among users have been respiratory symptoms, headaches and viral symptoms.

Information: HUAV Communication and Press