Sepsis is a common and severe condition that affects up to 40 000 people in Sweden every year w ith a mortality of up to 20%. Timely identification and treatment are crucial for the outcome. The majority of the patients have their first contact w ith the prehospital care, and should ideally be identified already in the ambulance. Screening tools that are used today rely on vital signs and are criticized for low precision. A prehospital AI based clinical decision support system (CDSS) have great potential to increase the accuracy of the early assessment and thus shorten the time to treatment, increase the chance of survival and reduce long-term complications.
Patient records from earlier clinical research w ill be used to develop an AI based CDSS.Selected AI methods w ill be evaluated in terms of performance and user experience. The CDSSw ill be incorporated in an existing ambulance IT-support system. The system should use realtimedata and notify the user if there is a risk of sepsis. This approach differs from the currentsituation in several w ays. Firstly, the paramedics do not have to suspect sepsis and secondly,also the patient’s symptom description can be included. The latter has proved to be of greatprognostic value. Visualization and utilization plays a significant role in the project.Presentation of the results from the CDSS as w ell as impact on the care process w ill beevaluated together w ith target users. Another important part of the project is to investigateregulatory issues and long term management of the system.