Abstract
To improve the work environment, nursing managers need to understand the actual conditions and analyze objective data on nurses’ workloads and movement lines obtained from long-term surveys. In this paper, we aim to validate an accuracy of a method for estimating nurses’ movement lines in an unattended time study using beacons and mobile devices and discuss the practicality of the visualization software framework using the estimation method. To do so, a pilot experiment was conducted in a surgical ward of a large acute care hospital. To evaluate the impact of parameter settings on estimation accuracy, we proposed a new approach for determining optimal thresholds based on the F-measure. By comparing the proposed optimization-based and conventional heuristic approaches, we demonstrate the effectiveness of the heuristic approach in achieving the high accuracy of position estimation. After that, we validated the high accuracy of movement lines estimated by the heuristic approach. Error evaluation for entering/leaving timing and staying periods showed that the method has a time resolution of a few seconds. Results suggest that the framework enables nursing managers to easily capture nurses’ movement lines throughout the ward and analyze them qualitatively to improve nursing workflows.
