Time based clustering for analyzing acute hospital patient flow.

Area of Science:

• Health Informatics
• Data Science
• Hospital Operations

Background:

• Analyzing patient flow is crucial for hospital efficiency.
• Traditional methods may not capture dynamic patient movement effectively.
• Understanding patient journeys requires robust analytical tools.

Purpose of the Study:

• To present a novel time-based clustering approach for health data.
• To visualize and analyze patient flow in hospitals.
• To demonstrate the technique's utility in addressing patient flow queries.

Main Methods:

• Clustering patient episodes into hourly slots using timestamps.
• Grouping clustered data based on relevant parameters.
• Applying time-based clustering to inpatient and emergency department data.

Main Results:

• The approach enables visualization of patient flow dynamics.
• It facilitates analysis of interactions between patient flow parameters.
• Demonstrated efficacy in answering typical patient flow questions.

Conclusions:

• Time-based clustering is a powerful tool for hospital patient flow analysis.
• The technique offers insights into interdependencies within patient pathways.
• This method enhances the understanding of healthcare system operations.