02-02-2026
The results of the HUMAI project – High Users Management with Artificial Intelligence – show that the use of Artificial Intelligence through machine learning, combined with multidisciplinary teams, can help reduce the overuse of emergency departments, improving the delivery of healthcare and generating cost benefits for Portugal’s National Health Service (SNS).
The project was developed by the Almada-Seixal Local Health Unit (ULSAS), the NOVA School of Science and Technology (NOVA FCT), and the Value for Health CoLAB (VOH.CoLAB), which leads the consortium. It is based on the work of the High Users Resolution Group (GRHU), a ULSAS team that brings together hospital professionals and primary healthcare providers to deliver coordinated and tailored responses to patients’ needs.
Within the scope of HUMAI, a platform was created that allows clinical teams to continuously monitor patients’ progress by integrating clinical, demographic, and healthcare utilization data. This enables the identification of patterns in emergency department use, the monitoring of intervention outcomes, and the support of clinical decision-making based on concrete data.
One of the most significant advances was the use of Machine Learning and Deep Learning models to predict which patients are at higher risk of returning to the emergency department. These models support decisions on which cases to follow more closely and help define personalized strategies for each patient.
“This project demonstrates how combining expertise in artificial intelligence with clinical intervention can transform the way we care for patients who most frequently use emergency services. We believe that this collaborative approach, in addition to empowering healthcare professionals to incorporate AI, made it possible to jointly develop digital tools that support more informed decisions, more personalized care, and more efficient management of SNS resources,” says Ana Rita Londral, Director of VOH.CoLAB and Professor at NOVA FCT.
“With the support of Artificial Intelligence and this platform developed under HUMAI, the GRHU will be able to act more preventively, further personalizing the care provided to our patients and optimizing the response of the General Emergency Department at Hospital Garcia de Orta, making the system more efficient and sustainable,” states Pedro Correia Azevedo, President of the Board of Directors of ULSAS, highlighting the synergy with academia.
Exploratory results show that there are patient profiles with a significant reduction in emergency department visits after the GRHU intervention, as well as profiles that require longer or more tailored follow-up. The economic analysis also reveals a positive financial benefit from the intervention, particularly among frequent users.
The HUMAI project was based on clinical and sociodemographic data from the General Emergency Department of Hospital Garcia de Orta (ULSAS), covering 8,250 patients previously classified as frequent or very frequent users between April 2022 and May 2025.
The first line of work focused on anticipating the expected impact of the GRHU intervention for each patient, supporting clinical decision-making and contributing to a more efficient allocation of resources.
The HUMAI project included an evaluation involving clinicians, managers, and decision-makers, aligning clinical, operational, patient experience, and cost indicators. This approach helped strengthen the commitment of the different stakeholders and ensure the sustainability of the intervention in the medium and long term.
In addition, the project stands out for its participatory approach: the development of the platform, indicators, and Artificial Intelligence models was carried out through co-creation with researchers, healthcare professionals, and managers, ensuring solutions that address real needs and integrate seamlessly into existing workflows. The partnership between ULSAS, the collaborative laboratory, and NOVA FCT demonstrates how Artificial Intelligence can support clinical decisions, personalize care, and make the healthcare system more efficient, sustainable, and people-centered.
HUMAI is funded under the European Union’s NextGenerationEU program, through the RE-C05-i08 investment – “More Digital Science” – of the Recovery and Resilience Plan (PRR), within the Call “Artificial Intelligence, Data Science and Cybersecurity of relevance to Public Administration,” promoted by the Portuguese Foundation for Science and Technology (reference 2024.07543.IACDC/2024).
