An ageing population means more people are living with multiple chronic conditions and have increasing needs for healthcare. To reduce the demand for extensive services, it is important to detect early signs of health decline. This allows for timely interventions—such as adjusting medication or preventing falls—before the situation becomes serious. Such measures can help reduce hospital admissions, which are both costly and stressful for patients.
The PAI project (Predicting Care Needs of Older Adults in the Healthcare System through AI-enabled Analysis of Patient-Monitoring Data) addresses this challenge by using artificial intelligence (AI) to analyse sensor data. The first step is to explore whether continuous measurements—such as physical activity and heart rate—can be used in an AI model to predict future health problems in older adults. The next step is to assess how such a model can be integrated into healthcare services in a way that saves time and resources, and fits into existing workflows.
The project is being carried out in Norway, Sweden, and Spain, and combines quantitative data (such as sensor readings and health registries) with insights from healthcare professionals and patients. Collaborations with local healthcare providers have been established, ethical approvals are in place, and a shared website and press releases have been published. A protocol paper describing the project’s methods and goals is nearing completion.
A key contribution to the development of the AI model comes from Norway. Here, activity data from the large population study HUNT4 (2017–2019) is being linked with national health registries (NPR and KPR). The study investigates whether variation in movement patterns is associated with healthcare use in the months following the measurement, regardless of diagnosis. Between 8,000 and 10,000 participants are expected to meet the inclusion criteria. The data has been received and analysis is underway.
Preparations for testing implementation in healthcare services are also in progress. This phase involves collecting data from individuals at high risk of acute health deterioration—typically older adults with multiple conditions and recent healthcare contacts. Different types of sensors have been evaluated for data quality and practical use. So far, the conclusion is to use commercially available smartwatches. Work is ongoing to determine whether medical device certification is required, and how data from the watches can be extracted and linked to health information.
As the population ages, the demand for healthcare services increases, including acute care. This puts a strain on healthcare resources, making efficient care management crucial at all levels - home, nursing homes, primary care, and hospitals.
The PAI project aims to address this challenge by using artificial intelligence (AI) to predict the care needs of older individuals. By analyzing patient monitoring data, the project can detect early signs of health or functional decline. This allows for preventive measures to be taken, reducing the need for costly acute care and aftercare. For example, hospitalizations can be prevented by avoiding falls in nursing homes or adjusting medication in primary care.
However, predicting the care needs of older patients is not straightforward. It involves dealing with varying data availability, quality, and potential interventions across different care levels and diagnoses. To tackle this, the PAI project uses a design science approach and a multiple case study design. It develops analytics in collaboration with care recipients and healthcare practitioners and implements them in real-life settings in Spain, Sweden, and Norway.
The project uses a mixed-methods strategy for data collection and analysis, combining quantitative and qualitative elements. This includes the thematic interpretative analysis of meeting notes, interviews, and observations, as well as the statistical longitudinal trend analysis of Key Performance Indicators in the respective healthcare systems.
The benefits of the PAI project are twofold. Older individuals receive adequate early care, preventing injuries and the worsening of diseases. Healthcare staff can use their resources more effectively, creating a sustainable and satisfying work environment. The outcomes of the project are scalable and can be applied in various contexts, making it a valuable contribution to the future of European healthcare.
