A medical procedure-based patient grouping method for an emergency department

M. Xu, T.C. Wong, K.S. Chin

Research output: Contribution to journalArticle

12 Citations (Scopus)

Abstract

Variability and unpredictability are typical characteristics of complex systems such as emergency department (ED) where the patient demand is high and patient conditions are diverse. To tackle the uncertain nature of ED and improve the resource management, it is beneficial to group patients with common features. This paper aims to use self-organizing map (SOM), k-means, and hierarchical methods to group patients based on their medical procedures and make comparisons among these methods. It can be reasonably assumed that the medical procedures received by the patients are directly associated with ED resource consumption. Different grouping techniques are compared using a validity index and the resulting groups are distinctive in the length of treatment (LOT) of patients and their presenting complaints. This paper also discusses how the resulting patient groups can be used to enhance the ED resource planning, as well as to redesign the ED charging policy.
LanguageEnglish
Pages31-37
Number of pages7
JournalApplied Soft Computing
Volume14
Issue numberPART A
DOIs
Publication statusPublished - 1 Jan 2014

Fingerprint

Self organizing maps
Large scale systems
Planning

Keywords

  • medical procedure
  • patient grouping
  • emergency department

Cite this

@article{fe4259bc4ca34f74bc455995280464ad,
title = "A medical procedure-based patient grouping method for an emergency department",
abstract = "Variability and unpredictability are typical characteristics of complex systems such as emergency department (ED) where the patient demand is high and patient conditions are diverse. To tackle the uncertain nature of ED and improve the resource management, it is beneficial to group patients with common features. This paper aims to use self-organizing map (SOM), k-means, and hierarchical methods to group patients based on their medical procedures and make comparisons among these methods. It can be reasonably assumed that the medical procedures received by the patients are directly associated with ED resource consumption. Different grouping techniques are compared using a validity index and the resulting groups are distinctive in the length of treatment (LOT) of patients and their presenting complaints. This paper also discusses how the resulting patient groups can be used to enhance the ED resource planning, as well as to redesign the ED charging policy.",
keywords = "medical procedure, patient grouping, emergency department",
author = "M. Xu and T.C. Wong and K.S. Chin",
year = "2014",
month = "1",
day = "1",
doi = "10.1016/j.asoc.2013.09.022",
language = "English",
volume = "14",
pages = "31--37",
journal = "Applied Soft Computing",
issn = "1568-4946",
number = "PART A",

}

A medical procedure-based patient grouping method for an emergency department. / Xu, M.; Wong, T.C.; Chin, K.S.

In: Applied Soft Computing, Vol. 14, No. PART A, 01.01.2014, p. 31-37.

Research output: Contribution to journalArticle

TY - JOUR

T1 - A medical procedure-based patient grouping method for an emergency department

AU - Xu, M.

AU - Wong, T.C.

AU - Chin, K.S.

PY - 2014/1/1

Y1 - 2014/1/1

N2 - Variability and unpredictability are typical characteristics of complex systems such as emergency department (ED) where the patient demand is high and patient conditions are diverse. To tackle the uncertain nature of ED and improve the resource management, it is beneficial to group patients with common features. This paper aims to use self-organizing map (SOM), k-means, and hierarchical methods to group patients based on their medical procedures and make comparisons among these methods. It can be reasonably assumed that the medical procedures received by the patients are directly associated with ED resource consumption. Different grouping techniques are compared using a validity index and the resulting groups are distinctive in the length of treatment (LOT) of patients and their presenting complaints. This paper also discusses how the resulting patient groups can be used to enhance the ED resource planning, as well as to redesign the ED charging policy.

AB - Variability and unpredictability are typical characteristics of complex systems such as emergency department (ED) where the patient demand is high and patient conditions are diverse. To tackle the uncertain nature of ED and improve the resource management, it is beneficial to group patients with common features. This paper aims to use self-organizing map (SOM), k-means, and hierarchical methods to group patients based on their medical procedures and make comparisons among these methods. It can be reasonably assumed that the medical procedures received by the patients are directly associated with ED resource consumption. Different grouping techniques are compared using a validity index and the resulting groups are distinctive in the length of treatment (LOT) of patients and their presenting complaints. This paper also discusses how the resulting patient groups can be used to enhance the ED resource planning, as well as to redesign the ED charging policy.

KW - medical procedure

KW - patient grouping

KW - emergency department

UR - http://www.scopus.com/inward/record.url?scp=84888287306&partnerID=8YFLogxK

U2 - 10.1016/j.asoc.2013.09.022

DO - 10.1016/j.asoc.2013.09.022

M3 - Article

VL - 14

SP - 31

EP - 37

JO - Applied Soft Computing

T2 - Applied Soft Computing

JF - Applied Soft Computing

SN - 1568-4946

IS - PART A

ER -