Assessing patient risk of central line-associated bacteremia via machine learning

Cole Beeler, Lana Dbeibo, Kristen Kelley, Levi Thatcher, Douglas Webb, Amadou Bah, Patrick Monahan, Nicole Fowler, Spencer Nicol, Alisa Judy-Malcolm, Jose Azar

Research output: Contribution to journalArticle

3 Citations (Scopus)

Abstract

Background: Central line-associated bloodstream infections (CLABSIs) contribute to increased morbidity, length of hospital stay, and cost. Despite progress in understanding the risk factors, there remains a need to accurately predict the risk of CLABSIs and, in real time, prevent them from occurring. Methods: A predictive model was developed using retrospective data from a large academic healthcare system. Models were developed with machine learning via construction of random forests using validated input variables. Results: Fifteen variables accounted for the most significant effect on CLABSI prediction based on a retrospective study of 70,218 unique patient encounters between January 1, 2013, and May 31, 2016. The area under the receiver operating characteristic curve for the best-performing model was 0.82 in production. Discussion: This model has multiple applications for resource allocation for CLABSI prevention, including serving as a tool to target patients at highest risk for potentially cost-effective but otherwise time-limited interventions. Conclusions: Machine learning can be used to develop accurate models to predict the risk of CLABSI in real time prior to the development of infection.

Original languageEnglish (US)
JournalAmerican Journal of Infection Control
DOIs
StateAccepted/In press - Jan 1 2018

Fingerprint

Bacteremia
Infection
Length of Stay
Resource Allocation
Hospital Costs
ROC Curve
Machine Learning
Retrospective Studies
Morbidity
Delivery of Health Care
Costs and Cost Analysis

Keywords

  • Central line-associated bloodstream infection (CLABSI)
  • Infection control
  • Infection prediction
  • Machine learning
  • Nosocomial infection
  • Quality improvement

ASJC Scopus subject areas

  • Epidemiology
  • Health Policy
  • Public Health, Environmental and Occupational Health
  • Infectious Diseases

Cite this

Assessing patient risk of central line-associated bacteremia via machine learning. / Beeler, Cole; Dbeibo, Lana; Kelley, Kristen; Thatcher, Levi; Webb, Douglas; Bah, Amadou; Monahan, Patrick; Fowler, Nicole; Nicol, Spencer; Judy-Malcolm, Alisa; Azar, Jose.

In: American Journal of Infection Control, 01.01.2018.

Research output: Contribution to journalArticle

Beeler, Cole ; Dbeibo, Lana ; Kelley, Kristen ; Thatcher, Levi ; Webb, Douglas ; Bah, Amadou ; Monahan, Patrick ; Fowler, Nicole ; Nicol, Spencer ; Judy-Malcolm, Alisa ; Azar, Jose. / Assessing patient risk of central line-associated bacteremia via machine learning. In: American Journal of Infection Control. 2018.
@article{1e3ba79996e24f0d9734e0be1332b527,
title = "Assessing patient risk of central line-associated bacteremia via machine learning",
abstract = "Background: Central line-associated bloodstream infections (CLABSIs) contribute to increased morbidity, length of hospital stay, and cost. Despite progress in understanding the risk factors, there remains a need to accurately predict the risk of CLABSIs and, in real time, prevent them from occurring. Methods: A predictive model was developed using retrospective data from a large academic healthcare system. Models were developed with machine learning via construction of random forests using validated input variables. Results: Fifteen variables accounted for the most significant effect on CLABSI prediction based on a retrospective study of 70,218 unique patient encounters between January 1, 2013, and May 31, 2016. The area under the receiver operating characteristic curve for the best-performing model was 0.82 in production. Discussion: This model has multiple applications for resource allocation for CLABSI prevention, including serving as a tool to target patients at highest risk for potentially cost-effective but otherwise time-limited interventions. Conclusions: Machine learning can be used to develop accurate models to predict the risk of CLABSI in real time prior to the development of infection.",
keywords = "Central line-associated bloodstream infection (CLABSI), Infection control, Infection prediction, Machine learning, Nosocomial infection, Quality improvement",
author = "Cole Beeler and Lana Dbeibo and Kristen Kelley and Levi Thatcher and Douglas Webb and Amadou Bah and Patrick Monahan and Nicole Fowler and Spencer Nicol and Alisa Judy-Malcolm and Jose Azar",
year = "2018",
month = "1",
day = "1",
doi = "10.1016/j.ajic.2018.02.021",
language = "English (US)",
journal = "American Journal of Infection Control",
issn = "0196-6553",
publisher = "Mosby Inc.",

}

TY - JOUR

T1 - Assessing patient risk of central line-associated bacteremia via machine learning

AU - Beeler, Cole

AU - Dbeibo, Lana

AU - Kelley, Kristen

AU - Thatcher, Levi

AU - Webb, Douglas

AU - Bah, Amadou

AU - Monahan, Patrick

AU - Fowler, Nicole

AU - Nicol, Spencer

AU - Judy-Malcolm, Alisa

AU - Azar, Jose

PY - 2018/1/1

Y1 - 2018/1/1

N2 - Background: Central line-associated bloodstream infections (CLABSIs) contribute to increased morbidity, length of hospital stay, and cost. Despite progress in understanding the risk factors, there remains a need to accurately predict the risk of CLABSIs and, in real time, prevent them from occurring. Methods: A predictive model was developed using retrospective data from a large academic healthcare system. Models were developed with machine learning via construction of random forests using validated input variables. Results: Fifteen variables accounted for the most significant effect on CLABSI prediction based on a retrospective study of 70,218 unique patient encounters between January 1, 2013, and May 31, 2016. The area under the receiver operating characteristic curve for the best-performing model was 0.82 in production. Discussion: This model has multiple applications for resource allocation for CLABSI prevention, including serving as a tool to target patients at highest risk for potentially cost-effective but otherwise time-limited interventions. Conclusions: Machine learning can be used to develop accurate models to predict the risk of CLABSI in real time prior to the development of infection.

AB - Background: Central line-associated bloodstream infections (CLABSIs) contribute to increased morbidity, length of hospital stay, and cost. Despite progress in understanding the risk factors, there remains a need to accurately predict the risk of CLABSIs and, in real time, prevent them from occurring. Methods: A predictive model was developed using retrospective data from a large academic healthcare system. Models were developed with machine learning via construction of random forests using validated input variables. Results: Fifteen variables accounted for the most significant effect on CLABSI prediction based on a retrospective study of 70,218 unique patient encounters between January 1, 2013, and May 31, 2016. The area under the receiver operating characteristic curve for the best-performing model was 0.82 in production. Discussion: This model has multiple applications for resource allocation for CLABSI prevention, including serving as a tool to target patients at highest risk for potentially cost-effective but otherwise time-limited interventions. Conclusions: Machine learning can be used to develop accurate models to predict the risk of CLABSI in real time prior to the development of infection.

KW - Central line-associated bloodstream infection (CLABSI)

KW - Infection control

KW - Infection prediction

KW - Machine learning

KW - Nosocomial infection

KW - Quality improvement

UR - http://www.scopus.com/inward/record.url?scp=85045336476&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85045336476&partnerID=8YFLogxK

U2 - 10.1016/j.ajic.2018.02.021

DO - 10.1016/j.ajic.2018.02.021

M3 - Article

C2 - 29661634

AN - SCOPUS:85045336476

JO - American Journal of Infection Control

JF - American Journal of Infection Control

SN - 0196-6553

ER -