Proactive and patient centered resource planning in emergency care using AI

Important results from the project

The purpose of the project has been to improve resource planning at hospitals by better prediction of the flow of acute patients. The project has developed a digital tool based on artificial intelligence that predicts the acute care need at hospitals, on several different time horizons, by processing and analyzing the interaction between external factors and care need. The tool is complete and ready to be implemented as a decision support in resource planning at hospitals.

Expected long term effects

An activity plan to distribute the digital tool to users of the Sveus Analysis Platform is under development and will hopefully lead to wide usage of the tool, whose predictions performs considerably better than commonly used methods today. The tool is, when implemented, expected to improve patient safety, work environment, economy, and equal-opportunity care through reduced overcrowding, improved expectations of workload, reduced cancellations of care, and needs-based resource planning.

Approach and implementation

Automatic data collection has been established with the Sveus platform as starting point, adding on extra emergency care data and several external open data sources. The project has proceeded through a cyclical process of development and evaluation. Ivbar has performed development of algorithms and user interfaces (UI), and VGR (with clinical representatives from the Skaraborg Hospital) has evaluated the results and directed the development of prediction targets and UI to comply with the different goals of the organization.