Analyse the care setting
Analyse the monitoring and record systems of a long-term care facility in the province of Jaén, the data available and how caregivers use them.
Objectives
Main objective
To develop and validate an artificial intelligence agent that fits into the long-term care workflow and processes residents' record sheets to generate recommendations and prioritise incidents, acting as a clinical decision support system to improve care responsiveness and reduce avoidable adverse events.
Specific objectives
Design the AI agent
Design a clinical decision support agent that processes staff observations, prioritises risk and generates daily summaries, contextualised alerts and personalised care plans.
Seamless integration
Evaluate the integration of the agent into the facility's workflow without substantial changes to its technology infrastructure.
Measure the impact
Assess the impact on clinical safety, workload and perceived stress among staff (MBI-HSS and PSS-14), using a pre-post design and qualitative methods.
Measure usability and acceptance
Measure the usability and acceptability of the system (SUS scale and interviews), identifying barriers and enablers.
Good Practice Guide
Produce an evidence-based Good Practice Guide to improve observation and the prioritisation of clinical alerts.
Working hypothesis
Integrating an AI agent into the care routine will reduce caregivers' cognitive and operational load, improve care traceability and increase responsiveness to incidents, contributing to the early detection of clinical risks and the reduction of avoidable adverse events.
Expected results
Technological
A working prototype of the agent validated for at least two months in a long-term care facility in Jaén, with a replicable modular architecture.
Scientific
Evidence on the impact on cognitive load, efficiency and responsiveness, with conference papers and open data.
Social and clinical
An estimated 15–25% reduction in administrative workload and an open-access Good Practice Guide.