Predicting In-patient Discharges to Identify Bed Availability for Housekeeping During Emergency Department Surge Using Machine Learning Algorithms

Prolonged patient waiting times for an inpatient bed in the Emergency Department (ED) is considered a global crisis and most cited reason for the ED crowding. Inability to move admitted patients from the ED to an inpatient bed due to capacity shortage and inefficiencies in patient-flow affects patient care and patient experiences. Housekeeping is crucial in patient-flow from ED to inpatient bed. When a patient leaves a bed (for example, discharge), housekeeping staff must clean and sanitize the bed. Knowing when the beds will be available during ED surge can facilitate the process of bed cleaning and hopefully reduce ED wait times. This research aims to apply machine learning algorithms to predict inpatient discharges during ED surge to identify the number of beds available in next four hours. This will enable housekeeping staff to plan ahead and reduce ED wait times.

Faculty Supervisor:

Michael Carter

Student:

Partner:

Sodexo Canada

Discipline:

Engineering

Sector:

Administrative and support, waste management and remediation services

University:

University of Toronto

Program: