Developing a predictive model for shift-level nurse staffing using routine data
(Workforce Effectiveness Research at the Inselspital- WER@INSEL)
National Institute for Health Research Collaboration for Leadership in Applied Health Research and Care (Wessex), University of Southampton, UK (Petter Griffiths) | Institute of Clinical Chemistry, Inselspital, Bern University Hospital, University of Bern, Bern, Switzerland & Laboratory of Biometry, University of Thessaly, Volos, Greece (Christos Nakas) | Institute of Clinical Chemistry, Inselspital, Bern University Hospital, University of Bern, Bern, Switzerland (Alexander Lechtle) | Medical Directorate, Inselspital, University Hospital of Bern, Bern, Switzerland (Olga Endrich)
Ort der Datenerhebung
– Institute of Clinical Chemistry, Inselspital, Bern University Hospital, University of Bern, Bern, Switzerland
– Medical Directorate, Inselspital, University Hospital of Bern, Bern, Switzerland
2014 bis 2018
Hospitals are challenged to determine optimal nurse staffing to balance cost and quality patient care. About one third of the operational costs in Swiss hospitals are nurses’ salaries, which make them a primary target for cost containment measures. On the other hand, hospitals face problems to maintain adequate staffing levels and unfilled nursing positions potentially intensify nurses’ workload. Part of the problem is that the mechanism of how nurses or nursing care influence or contribute to patient outcomes are unclear. Non-optimal nurse staffing potentially lead to adverse patient outcomes and increased healthcare costs.
The purpose of the proposed study is to develop a method to determine nurse staffing requirements from routine data on a shift-by-shift basis.
A retrospective observational study will be conducted exploiting routine data from several sources from the University Hospital of Bern (Inselspital): 1) tacs® nurse staffing system; and 2) hospital discharge data. Adult population from the medical, surgical or oncological departments of the University Hospital of Bern will be investigated. All the statistical analysis will be conducted with the free software R and employ statistical methods like generalized linear mixed models, machine learning and predictive modeling techniques.
Erwarteter Nutzen / Relevanz
This project will develop a predictive model for staffing requirements, which will improve patient outcomes. New methodologies will be used, such as machine learning.