Time based clustering for analyzing acute hospital patient flow.

This paper describes a novel approach employing time based clustering of health data for visualization and analysis of patient flow. Clustering inpatient and emergency department patient episodes into hourly slots based on recorded timestamps, and then grouping them on required parameters, the technique provides a powerful tool for visualizing and analyzing interactions and interdependencies between hospital patient flow parameters. To demonstrate the efficacy of the approach, we employ time based clustering to address some typical patient flow related queries and discuss the findings.