Managing responsiveness in the emergency department: Comparing dynamic priority queue with fast track

• DOI: http://doi.org/10.1016/j.jom.2018.03.001
• Published date: 01 March 2018
• Author: Uday S. Rao,Todd F. Glass,Michael J. Magazine,Yann B. Ferrand
• Date: 01 March 2018

Contents lists available at ScienceDirect

Journal of Operations Management

journal homepage: www.elsevier.com/locate/jom

Managing responsiveness in the emergency department: Comparing

dynamic priority queue with fast track

Yann B. Ferrand

a,∗

, Michael J. Magazine

b

, Uday S. Rao

b

, Todd F. Glass

c,1

a

Department of Management, Clemson University, 374E Sirrine Hall, Clemson, SC 29634, United States

b

Department of Operations, Business Analytics, and Information Systems, University of Cincinnati, 501 Carl H. Lindner Hall, 2925 Campus Green Drive, Cincinnati, OH

45221, United States

c

Cincinnati Children's Hospital Medical Center, Division of Emergency Medicine, 3333 Burnet Avenue, ML 16022, Cincinnati, OH 45229, United States

ARTICLE INFO

Accepted by: Tyson Browning

Keywords:

Service system design

Focused and flexible resources

Customer prioritization

ABSTRACT

Emergency Departments (EDs) commonly face capacity imbalances and long wait times in a service system

handling patients with different priorities. These problems are particularly important for low-priority patients

who often remain in the queue for extended periods. We investigate two distinct approaches to address these

challenges: fast track (FT) and dynamic priority queue (DPQ). Traditionally, EDs have prioritized patients using

an Emergency Severity Index (ESI), in conjunction with FT, to strictly or partially dedicate resources to different

ESI patient classes. With our proposed DPQ, patients are prioritized using ESI and additional real-time opera-

tional information about the patient, specifically the amount of accumulated wait time and flow time. Using an

empirical simulation, we compare the impact of different resource allocation and prioritization approaches on

patient length of stay (LOS), including the existing system at the ED, FT with strict and partial dedication and the

possibility of shorter and less variable service times, and versions of the proposed DPQ using simple dynamic

prioritization. Our main results are that: (i) the DPQ approach dominates the other approaches tested; (ii) for

various ED sizes, FT with strict and partial dedication do not reduce average LOS of low-priority patients without

significantly increasing average LOS of high-priority patients, unless service time mean and variance are re-

duced; (iii) DPQ using accumulated wait time or accumulated flow time improves performance. The results are

robust to changes in the proportion of patients in each priority level. Overall, expanding decision making about

patient prioritization from only considering the patient's clinical condition to also including operational data can

improve performance dramatically, even without improved service times.

1. Introduction

Emergency departments (EDs) are service systems handling custo-

mers with different priorities. Historically, there has been an increase in

the annual patient visit rate to EDs (Tang et al., 2010) and a decrease in

the number of hospital EDs (Nawar et al., 2007), resulting in over-

crowding, and higher pressure to cost-effectively manage resources. In

EDs, after safety, timeliness and efficiency are the two most critical

criteria (Graffet al., 2002;Lindsay et al., 2002), yet EDs have shown

increasing difficulty in providing timely care (Horwitz and Bradley,

2009). This paper studies approaches to providing timely care, parti-

cularly for lower priority (non-urgent and simple) patients that re-

present the preponderance of patients seen by typical EDs.

A common approach is to prioritize patients as they arrive using a

triage protocol and the Emergency Severity Index (ESI, Gilboy et al.,

2012). The ESI accounts for patient acuity, which is a combination of

urgency and severity, and the anticipated amount of resources needed,

which is a proxi for complexity of the patient's case. ESI level is de-

termined at triage, and should not change once assigned. In general,

patients are prioritized from most severe (ESI 1) to least severe (ESI 5).

However, in this approach the low-priority ESI 4 and 5 patients can

wait a long time. To alleviate this wait time, a commonly-used ap-

proach, referred to as Fast Track (FT), proactively allocates these low-

priority patients some dedicated capacity (Gilboy et al., 2012).

The medical director of the Liberty Township Emergency

Department (Liberty ED), a branch ED of Cincinnati Children's Hospital

Medical Center (CCHMC), sought to investigate ways of improving

patient length of stay (LOS) in this pediatric ED, without adding re-

sources. A key goal of hospital administration was to significantly reduce

LOS for lower priority patients without increasing LOS for higher priority

https://doi.org/10.1016/j.jom.2018.03.001

Received 24 July 2016; Received in revised form 28 March 2018; Accepted 29 March 2018

∗

Corresponding author.

1

Permanent address: Nemours Children's Hospital, Division of Emergency Medicine, 13535 Nemours Parkway, Orlando, FL 32827.

E-mail addresses: yferran@clemson.edu (Y.B. Ferrand), mike.magazine@uc.edu (M.J. Magazine), uday.rao@uc.edu (U.S. Rao), todd.glass@nemours.org (T.F. Glass).

Journal of Operations Management 58–59 (2018) 15–26

0272-6963/ © 2018 Elsevier B.V. All rights reserved.

T