Medical News Observer

• The research used data from the Swedish Trauma Registry (2013-2020) to create AI models that better assess injury severity and appropriate transport destinations compared to current methods.
• The AI models, considering factors like respiratory rate, age, and injury type, accurately determine injury severity, aiding in timely and accurate emergency decisions.
• Despite positive outcomes, challenges remain in integrating AI into ambulance services, such as developing user-friendly interfaces and ensuring effective interaction with healthcare professionals.

The research conducted by Chalmers University of Technology, the University of Gothenburg, and the University of Borås in Sweden highlights the potential of AI in assisting healthcare professionals with rapid, life-critical decisions in trauma care. 

This study evaluates the potential of Artificial Intelligence (AI) in improving prehospital care for trauma patients, a leading cause of death in young adults. Trauma care faces challenges like field triage limitations, necessitating effective methods to assess injury severity and determine appropriate transport destinations.

AI Model Research in Prehospital decision making

The study developed five mathematical models based on data from over 47,000 cases between 2013 and 2020 from the Swedish Trauma Registry. The study found that about 40% of patients were undertriaged and 46% overtriaged. These AI models, considering variables like respiratory rate, blood pressure, age, gender, and injury type, performed better than the transport decisions made by ambulance staff at the time of the incident. The results using AI models, utilizing 21 predictors including injury location, effectively determine the severity of injuries, which is crucial for making accurate and timely decisions in emergency medical situations.

The study concludes that AI-based On Scene Injury Severity Prediction (OSISP) models can enhance injury severity assessment in the prehospital setting, potentially improving trauma care and reducing morbidity and mortality.

Challanges and Future

Despite the promising results, there are several challenges before this AI technology can be implemented in ambulance services. A significant part of the development involves finding methods for quick and easy information input into the AI tool and ensuring effective interaction with users.

Reference

Bakidou, Anna, Eva-Corina Caragounis, Magnus Andersson Hagiwara, Anders Jonsson, Bengt Arne Sjöqvist, and Stefan Candefjord. 2023. “On Scene Injury Severity Prediction (OSISP) Model for Trauma Developed Using the Swedish Trauma Registry.” BMC Medical Informatics and Decision Making 23 (1): 206. https://doi.org/10.1186/s12911-023-02290-5.