Virtual simulation models enable preparation of healthcare teams working in emergency and disaster responses, by providing practice of leadership and communication in decision making. The statistical functions are also suitable for assessment of team performance. However, developing virtual simulated scenarios focused on team training is time-consuming, expensive, and consists of complex developmental processes. This feasibility study aimed to explore if application of AI on trauma registry would support automated creation of virtual simulated scenarios based on real patient data.

To determinate design and effects of an automated system converting real patient data into virtual simulated scenarios.

Mixed methods with two data sets. The first data set was extracted from trauma patients records for the development of a system converting real patient data into virtual simulated scenarios. The second data set consisted of focus group interviews.

The end product consisted of a Python-based program for automating virtual simulation scenario creation and a graphical user interface (GUI) displaying the scenarios. Further improvements were needed in efficiency and correlated to the quality of data derived from patients’ records. Incorporating functions such as time as stress factor, integration of decision-making components based on a decision-making tree would also contributes to usefulness and acceptance of the system.

The potential of the system is cost efficient and beneficial for healthcare teams and educational bodies by its ability to provide great numbers of emergency and disaster scenarios, and the access and presentation of real patient data in situations with limitations, such as during the pandemic.