ORAHS2024
Abstract Submission

216. Work time analytics in Emergency Departments using localisation sensors

Contributed abstract in session FA-5: Emergency Department /3, stream Regular talks.

Friday, 9:20-10:30
Room: Room S6

Authors (first author is the speaker)

1. Thierry Garaix
LIMS, EMSE
2. Marius Huguet
Ecole des Mines de Saint-Etienne
3. Canan Pehlivan
Industrial Engineering Center, IMT Mines Albi
4. François Ballereau
Centre Hospitalier Le Corbusier, Firminy
5. Antoine Dodane-Loyenet
Centre Hospitalier Le Corbusier, Firminy
6. Franck Fontanili
IMT Mines Albi
7. Youri Yordanov
APHP
8. Vincent Augusto
Mines Saint-Etienne
9. Karim Tazarourte
Université Lyon 1
10. Abdesslam Redjaline
Centre Hospitalier Le Corbusier, Firminy

Abstract

Emergency department (ED) crowding has become a major public health
issue in many countries worldwide. Crowding can be defined as a market
failure when the demand for emergency care outstrips the availability of
physical and human resources. In this study, we implemented an indoor
positioning system to track the activities of healthcare professionals
in an emergency department, aiming to gain a better understanding of the
emergency care production process. Healthcare professionals wore a
sensor to record their location within the emergency department. We
analyzed a substantial amount of quasi-real-time data to objectively
assess physicians’ time allocation and movement patterns and their
correlation with the emergency department’s occupancy. Additionally, we
developed a user recognition algorithm (i.e., random forest classifier)
capable of detecting the job category of the participant wearing the
sensor. We found that the proportion of time spent on care-related
activities ranged from 26% to 39% for doctors. The burden of
non-care-related activities appeared to be largely induced by the time
spent on administrative duties and transit, and to be correlated with ED
occupancy. Lastly, we developed a decision support tool to predict daily
and hourly loads of patients in the emergency department. Using a deep
learning approach, models show an accuracy greater than 90% in
predicting occupancy levels up to 7 days ahead.

Keywords

Status: accepted

Back to the list of papers