Artificial intelligence (AI) is gaining attention and importance within the healthcare system. AI can help to make tedious routine tasks easier - including forecasting staffing requirements in hospitals? Fraunhofer IKS is addressing this question in a current project – and will soon present its findings at the Healthcare Hackathon in Mainz, Germany.
© iStock/alvarez
Staffing in hospitals: Behind a complex term lies a complex process. Experienced specialists determine and plan the ideal staffing of hospital wards, considering regulatory requirements, to ensure the best possible patient care while reducing the workload of nursing staff. A demanding and time-consuming task that could be supported by artificial intelligence (AI) in the future.
In a joint project with ATOSS Software, the University Medical Center Mainz and the ecosystem for innovative healthcare Flying Health, the Fraunhofer Institute for Cognitive Systems IKS is developing an AI-based solution approach to predict future staffing requirements on a ward at the University Medical Center Mainz. State-of-the-art time series forecasting models are used to ensure the accuracy and reliability of the predictions. The focus of the AI-based prediction is the consideration of the Nursing Personnel Regulation (PPR) 2.0, a new instrument for staffing and part of the Hospital Relief Act (KHPflEG) of December 2 2022. Since January 1 2023, the trial phase of the PPR 2.0 has been running in some selected hospitals in Germany, including the University Medical Center Mainz.
The project idea originates from the AI Innovation Days 2022 in Berlin, jointly organized by Fraunhofer IKS and Flying Health. Experts from various areas of the healthcare sector discussed and developed AI-based solution approaches for use cases related to patient well-being, including AI-based staffing requirement forecasts in hospitals. What is special about this use case is that, in addition to providing the best possible patient care, the focus is also on reducing the workload of the nursing staff themselves.
Despite careful planning by experienced professionals, unpredictable peaks in occupancy and shortages of skilled staff pose major challenges for hospitals and other healthcare facilities. AI can help identify patterns and trends in retrospective hospital occupancy and staffing data and use this to predict occupancy peaks and staffing needs in the coming months, considering complex regulations such as PPR 2.0.
AI-based staffing, i.e., supporting demand-based planning with calculated suggestions, could improve hospital staffing management, the quality of patient care, and the satisfaction of highly stressed healthcare professionals. Nevertheless, it is important to realistically evaluate the expectations for AI-based staffing demand forecasting in the context of the project from a scientific point of view: an AI can provide suggestions for staffing as an additional technological tool, but it should not and cannot replace a staffing management system and a final assessment and decision by experienced professionals.
More information about the Hackathon can be found here:
The AI prototype for staffing requirement forecasting with integration of PPR 2.0 will be presented and further developed by the project team of Fraunhofer IKS, ATOSS, Mainz University Hospital and Flying Health together at the Healthcare Hackathon Mainz 2023. The project team will also discuss the possible transferability of the AI-supported staffing to other wards or hospitals by exchanging ideas with other innovation teams and healthcare representatives on site.
Based on the new findings, the AI prototype will be further improved, for example, by adding more decision parameters to the AI model, such as major events taking place, to further increase the accuracy and reliability of staffing demand forecasting. However, this requires the availability of sufficient quality training and test data - a well-known challenge in many AI use cases in healthcare and beyond.
With the project on staffing demand forecasting in hospitals using artificial intelligence, Fraunhofer IKS and the participating project partners are demonstrating the potential of AI for the future of healthcare. However, it is crucial to focus on the trustworthiness and reliability of AI.
