Automating the identification of patient safety incident reports using multi-label classification

Ying Wang, Enrico Coiera, William Runciman, Farah Magrabi

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Automated identification provides an efficient way to categorize patient safety incidents. Previous studies have focused on identifying single incident types relating to a specific patient safety problem, e.g., clinical handover. In reality, there are multiple types of incidents reflecting the breadth of patient safety problems and a single report may describe multiple problems, i.e., it can be assigned multiple type labels. This study evaluated the abilty of multi-label classification methods to identify multiple incident types in single reports. Three multi-label methods were evaluated: binary relevance, classifier chains and ensemble of classifier chains. We found that an ensemble of classifier chains was the most effective method using binary Support Vector Machines with radial basis function kernel and bag-of-words feature extraction, performing equally well on balanced and stratified datasets, (F-score: 73.7% vs. 74.7%). Classifiers were able to identify six common incident types: falls, medications, pressure injury, aggression, documentation problems and others.

LanguageEnglish
Title of host publicationMEDINFO 2017
Subtitle of host publicationPrecision Healthcare through Informatics - Proceedings of the 16th World Congress on Medical and Health Informatics
EditorsZhao Dongsheng, Adi V. Gundlapalli, Jaulent Marie-Christine
PublisherIOS Press
Pages609-613
Number of pages5
ISBN (Electronic)9781614998297
DOIs
Publication statusPublished - 1 Jan 2017
Event16th World Congress of Medical and Health Informatics: Precision Healthcare through Informatics, MedInfo 2017 - Hangzhou, China
Duration: 21 Aug 201725 Aug 2017

Publication series

NameStudies in Health Technology and Informatics
Volume245
ISSN (Print)0926-9630
ISSN (Electronic)1879-8365

Other

Other16th World Congress of Medical and Health Informatics: Precision Healthcare through Informatics, MedInfo 2017
CountryChina
CityHangzhou
Period21/08/1725/08/17

Keywords

  • Machine learning
  • Patient safety
  • Risk management

ASJC Scopus subject areas

  • Biomedical Engineering
  • Health Informatics
  • Health Information Management

Cite this

Wang, Y., Coiera, E., Runciman, W., & Magrabi, F. (2017). Automating the identification of patient safety incident reports using multi-label classification. In Z. Dongsheng, A. V. Gundlapalli, & J. Marie-Christine (Eds.), MEDINFO 2017: Precision Healthcare through Informatics - Proceedings of the 16th World Congress on Medical and Health Informatics (pp. 609-613). (Studies in Health Technology and Informatics; Vol. 245). IOS Press. https://doi.org/10.3233/978-1-61499-830-3-609
Wang, Ying ; Coiera, Enrico ; Runciman, William ; Magrabi, Farah. / Automating the identification of patient safety incident reports using multi-label classification. MEDINFO 2017: Precision Healthcare through Informatics - Proceedings of the 16th World Congress on Medical and Health Informatics. editor / Zhao Dongsheng ; Adi V. Gundlapalli ; Jaulent Marie-Christine. IOS Press, 2017. pp. 609-613 (Studies in Health Technology and Informatics).
@inproceedings{a7440c0ec3104600b9919a9885c2a292,
title = "Automating the identification of patient safety incident reports using multi-label classification",
abstract = "Automated identification provides an efficient way to categorize patient safety incidents. Previous studies have focused on identifying single incident types relating to a specific patient safety problem, e.g., clinical handover. In reality, there are multiple types of incidents reflecting the breadth of patient safety problems and a single report may describe multiple problems, i.e., it can be assigned multiple type labels. This study evaluated the abilty of multi-label classification methods to identify multiple incident types in single reports. Three multi-label methods were evaluated: binary relevance, classifier chains and ensemble of classifier chains. We found that an ensemble of classifier chains was the most effective method using binary Support Vector Machines with radial basis function kernel and bag-of-words feature extraction, performing equally well on balanced and stratified datasets, (F-score: 73.7{\%} vs. 74.7{\%}). Classifiers were able to identify six common incident types: falls, medications, pressure injury, aggression, documentation problems and others.",
keywords = "Machine learning, Patient safety, Risk management",
author = "Ying Wang and Enrico Coiera and William Runciman and Farah Magrabi",
year = "2017",
month = "1",
day = "1",
doi = "10.3233/978-1-61499-830-3-609",
language = "English",
series = "Studies in Health Technology and Informatics",
publisher = "IOS Press",
pages = "609--613",
editor = "Zhao Dongsheng and Gundlapalli, {Adi V.} and Jaulent Marie-Christine",
booktitle = "MEDINFO 2017",

}

Wang, Y, Coiera, E, Runciman, W & Magrabi, F 2017, Automating the identification of patient safety incident reports using multi-label classification. in Z Dongsheng, AV Gundlapalli & J Marie-Christine (eds), MEDINFO 2017: Precision Healthcare through Informatics - Proceedings of the 16th World Congress on Medical and Health Informatics. Studies in Health Technology and Informatics, vol. 245, IOS Press, pp. 609-613, 16th World Congress of Medical and Health Informatics: Precision Healthcare through Informatics, MedInfo 2017, Hangzhou, China, 21/08/17. https://doi.org/10.3233/978-1-61499-830-3-609

Automating the identification of patient safety incident reports using multi-label classification. / Wang, Ying; Coiera, Enrico; Runciman, William; Magrabi, Farah.

MEDINFO 2017: Precision Healthcare through Informatics - Proceedings of the 16th World Congress on Medical and Health Informatics. ed. / Zhao Dongsheng; Adi V. Gundlapalli; Jaulent Marie-Christine. IOS Press, 2017. p. 609-613 (Studies in Health Technology and Informatics; Vol. 245).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

TY - GEN

T1 - Automating the identification of patient safety incident reports using multi-label classification

AU - Wang, Ying

AU - Coiera, Enrico

AU - Runciman, William

AU - Magrabi, Farah

PY - 2017/1/1

Y1 - 2017/1/1

N2 - Automated identification provides an efficient way to categorize patient safety incidents. Previous studies have focused on identifying single incident types relating to a specific patient safety problem, e.g., clinical handover. In reality, there are multiple types of incidents reflecting the breadth of patient safety problems and a single report may describe multiple problems, i.e., it can be assigned multiple type labels. This study evaluated the abilty of multi-label classification methods to identify multiple incident types in single reports. Three multi-label methods were evaluated: binary relevance, classifier chains and ensemble of classifier chains. We found that an ensemble of classifier chains was the most effective method using binary Support Vector Machines with radial basis function kernel and bag-of-words feature extraction, performing equally well on balanced and stratified datasets, (F-score: 73.7% vs. 74.7%). Classifiers were able to identify six common incident types: falls, medications, pressure injury, aggression, documentation problems and others.

AB - Automated identification provides an efficient way to categorize patient safety incidents. Previous studies have focused on identifying single incident types relating to a specific patient safety problem, e.g., clinical handover. In reality, there are multiple types of incidents reflecting the breadth of patient safety problems and a single report may describe multiple problems, i.e., it can be assigned multiple type labels. This study evaluated the abilty of multi-label classification methods to identify multiple incident types in single reports. Three multi-label methods were evaluated: binary relevance, classifier chains and ensemble of classifier chains. We found that an ensemble of classifier chains was the most effective method using binary Support Vector Machines with radial basis function kernel and bag-of-words feature extraction, performing equally well on balanced and stratified datasets, (F-score: 73.7% vs. 74.7%). Classifiers were able to identify six common incident types: falls, medications, pressure injury, aggression, documentation problems and others.

KW - Machine learning

KW - Patient safety

KW - Risk management

UR - http://www.scopus.com/inward/record.url?scp=85040512077&partnerID=8YFLogxK

U2 - 10.3233/978-1-61499-830-3-609

DO - 10.3233/978-1-61499-830-3-609

M3 - Conference contribution

T3 - Studies in Health Technology and Informatics

SP - 609

EP - 613

BT - MEDINFO 2017

A2 - Dongsheng, Zhao

A2 - Gundlapalli, Adi V.

A2 - Marie-Christine, Jaulent

PB - IOS Press

ER -

Wang Y, Coiera E, Runciman W, Magrabi F. Automating the identification of patient safety incident reports using multi-label classification. In Dongsheng Z, Gundlapalli AV, Marie-Christine J, editors, MEDINFO 2017: Precision Healthcare through Informatics - Proceedings of the 16th World Congress on Medical and Health Informatics. IOS Press. 2017. p. 609-613. (Studies in Health Technology and Informatics). https://doi.org/10.3233/978-1-61499-830-3-609