Providing rapid and targeted emergency care requires complex decisions to be made under considerable time pressure. Emergency call handlers in integrated control centres must therefore assess medical, operational and organisational factors simultaneously. To improve the confidence and quality of decision-making by emergency call handlers, the PRIORITAS joint research project (Pre-clinical Rescue Resource Optimisation through an Intelligent, Optimising, Resource-Oriented IT-Supported Analysis System) is researching an approach to adaptive emergency call handling supported by Artificial Intelligence (AI). PRIORITAS distinguishes itself from the status quo, amongst other things, through a holistic view of rescue chains within the context of a clearly structured digital situational picture. This is intended to lay the foundation for more consistent alerting procedures, which take into account both the provision of first aid at the scene of the emergency as quickly as possible and the minimisation of pre-hospital time.
By analysing a wide range of information in parallel, emergency call handlers gain, amongst other things, an overview of the medical situation, available treatment capacity in hospitals, and flight conditions for air ambulance services. The approach being researched by PRIORITAS addresses this by using AI to augment human information processing. In this context, targeted yet consistent human–AI interaction is essential for providing effective support in highly critical situations. PRIORITAS meets this requirement by drawing on the fundamental principle behind the established approach of a structured or standardised questioning scheme. However, whilst a questioning schema focuses primarily on promoting consistency and quality in interpersonal dialogue, PRIORITAS develops an adaptive schema that aims to achieve similar outcomes with regard to human-AI interactions. It explicitly defines under which circumstances which AI-based interventions are permissible. The reliability associated with the execution of such a schema makes an important contribution to consistent, controllable and, ultimately, high-quality human–AI interactions.
The PRIORITAS collaborative project is funded by the DRF Foundation and is being carried out under the leadership of the DFKI over a period of three years. As a partner, DRF Luftrettung is contributing its practical experience and network. The framework for AI-supported, adaptive emergency call handling developed in PRIORITAS will be implemented in an emergency call handling system as part of the project. The resulting framework will be released as an open-source solution. In collaboration with selected control centres in Baden-Württemberg, the project will define the requirements and framework conditions for a trial run of PRIORITAS.
The enhancement of decision-making powers for emergency call handlers brought about by PRIORITAS holds the potential to deliver added value across the wider healthcare sector: Whilst earlier and more targeted alerts optimise on-site medical care, they also result in more stable and successful operational procedures for emergency services. Health insurance funds and funding bodies also benefit from increased potential for cost savings, as suitable hospitals are contacted from the outset, thereby avoiding sub-optimal initial clinical care and unnecessary secondary transfers.
