Factors Associated with Healthcare-Acquired Catheter-Associated Urinary Tract Infections

Analysis Using Multiple Data Sources and Data Mining Techniques

Jung In Park, Donna Z Bliss, Chih-Lin Chi, Connie W Delaney, Bonnie L Westra

Research output: Contribution to journalArticle

2 Citations (Scopus)

Abstract

PURPOSE: The purpose of this study was to identify factors associated with healthcare-acquired catheter-associated urinary tract infections (HA-CAUTIs) using multiple data sources and data mining techniques. SUBJECTS AND SETTING: Three data sets were integrated for analysis: Electronic health record data from a university hospital in the Midwestern United States was combined with staffing and environmental data from the hospital's National Database of Nursing Quality Indicators and a list of patients with HA-CAUTIs. METHODS: Three data mining techniques were used for identification of factors associated with HA-CAUTI: Decision trees, logistic regression, and support vector machines. RESULTS: Fewer total nursing hours per patient-day, lower percentage of direct care RNs with specialty nursing certification, higher percentage of direct care RNs with associate's degree in nursing, and higher percentage of direct care RNs with BSN, MSN, or doctoral degree are associated with HA-CAUTI occurrence. The results also support the association of the following factors with HA-CAUTI identified by previous studies: Female gender; older age (>50 years); longer length of stay; severe underlying disease; glucose lab results (>200 mg/dL); longer use of the catheter; and RN staffing. CONCLUSIONS: Additional findings from this study demonstrated that the presence of more nurses with specialty nursing certifications can reduce HA-CAUTI occurrence. While there may be valid reasons for leaving in a urinary catheter, findings show that having a catheter in for more than 48 hours contributes to HA-CAUTI occurrence. Finally, the findings suggest that more nursing hours per patient-day are related to better patient outcomes.

Original languageEnglish (US)
Pages (from-to)168-173
Number of pages6
JournalJournal of Wound, Ostomy and Continence Nursing
Volume45
Issue number2
DOIs
StatePublished - Jan 1 2018

Fingerprint

Catheter-Related Infections
Data Mining
Information Storage and Retrieval
Urinary Tract Infections
Delivery of Health Care
Nursing
Nursing Specialties
Certification
Catheters
Midwestern United States
Urinary Catheters
Decision Trees
Electronic Health Records
Length of Stay
Logistic Models
Nurses
Databases
Glucose

Keywords

  • Catheter-associated urinary tract infections
  • Data mining
  • Urinary catheter
  • Urinary tract infections

PubMed: MeSH publication types

  • Journal Article
  • Observational Study

Cite this

@article{2802ffb9bff4474fa987bf962ce09763,
title = "Factors Associated with Healthcare-Acquired Catheter-Associated Urinary Tract Infections: Analysis Using Multiple Data Sources and Data Mining Techniques",
abstract = "PURPOSE: The purpose of this study was to identify factors associated with healthcare-acquired catheter-associated urinary tract infections (HA-CAUTIs) using multiple data sources and data mining techniques. SUBJECTS AND SETTING: Three data sets were integrated for analysis: Electronic health record data from a university hospital in the Midwestern United States was combined with staffing and environmental data from the hospital's National Database of Nursing Quality Indicators and a list of patients with HA-CAUTIs. METHODS: Three data mining techniques were used for identification of factors associated with HA-CAUTI: Decision trees, logistic regression, and support vector machines. RESULTS: Fewer total nursing hours per patient-day, lower percentage of direct care RNs with specialty nursing certification, higher percentage of direct care RNs with associate's degree in nursing, and higher percentage of direct care RNs with BSN, MSN, or doctoral degree are associated with HA-CAUTI occurrence. The results also support the association of the following factors with HA-CAUTI identified by previous studies: Female gender; older age (>50 years); longer length of stay; severe underlying disease; glucose lab results (>200 mg/dL); longer use of the catheter; and RN staffing. CONCLUSIONS: Additional findings from this study demonstrated that the presence of more nurses with specialty nursing certifications can reduce HA-CAUTI occurrence. While there may be valid reasons for leaving in a urinary catheter, findings show that having a catheter in for more than 48 hours contributes to HA-CAUTI occurrence. Finally, the findings suggest that more nursing hours per patient-day are related to better patient outcomes.",
keywords = "Catheter-associated urinary tract infections, Data mining, Urinary catheter, Urinary tract infections",
author = "Park, {Jung In} and Bliss, {Donna Z} and Chih-Lin Chi and Delaney, {Connie W} and Westra, {Bonnie L}",
year = "2018",
month = "1",
day = "1",
doi = "10.1097/WON.0000000000000409",
language = "English (US)",
volume = "45",
pages = "168--173",
journal = "Journal of Wound, Ostomy and Continence Nursing",
issn = "1071-5754",
publisher = "Lippincott Williams and Wilkins",
number = "2",

}

TY - JOUR

T1 - Factors Associated with Healthcare-Acquired Catheter-Associated Urinary Tract Infections

T2 - Analysis Using Multiple Data Sources and Data Mining Techniques

AU - Park, Jung In

AU - Bliss, Donna Z

AU - Chi, Chih-Lin

AU - Delaney, Connie W

AU - Westra, Bonnie L

PY - 2018/1/1

Y1 - 2018/1/1

N2 - PURPOSE: The purpose of this study was to identify factors associated with healthcare-acquired catheter-associated urinary tract infections (HA-CAUTIs) using multiple data sources and data mining techniques. SUBJECTS AND SETTING: Three data sets were integrated for analysis: Electronic health record data from a university hospital in the Midwestern United States was combined with staffing and environmental data from the hospital's National Database of Nursing Quality Indicators and a list of patients with HA-CAUTIs. METHODS: Three data mining techniques were used for identification of factors associated with HA-CAUTI: Decision trees, logistic regression, and support vector machines. RESULTS: Fewer total nursing hours per patient-day, lower percentage of direct care RNs with specialty nursing certification, higher percentage of direct care RNs with associate's degree in nursing, and higher percentage of direct care RNs with BSN, MSN, or doctoral degree are associated with HA-CAUTI occurrence. The results also support the association of the following factors with HA-CAUTI identified by previous studies: Female gender; older age (>50 years); longer length of stay; severe underlying disease; glucose lab results (>200 mg/dL); longer use of the catheter; and RN staffing. CONCLUSIONS: Additional findings from this study demonstrated that the presence of more nurses with specialty nursing certifications can reduce HA-CAUTI occurrence. While there may be valid reasons for leaving in a urinary catheter, findings show that having a catheter in for more than 48 hours contributes to HA-CAUTI occurrence. Finally, the findings suggest that more nursing hours per patient-day are related to better patient outcomes.

AB - PURPOSE: The purpose of this study was to identify factors associated with healthcare-acquired catheter-associated urinary tract infections (HA-CAUTIs) using multiple data sources and data mining techniques. SUBJECTS AND SETTING: Three data sets were integrated for analysis: Electronic health record data from a university hospital in the Midwestern United States was combined with staffing and environmental data from the hospital's National Database of Nursing Quality Indicators and a list of patients with HA-CAUTIs. METHODS: Three data mining techniques were used for identification of factors associated with HA-CAUTI: Decision trees, logistic regression, and support vector machines. RESULTS: Fewer total nursing hours per patient-day, lower percentage of direct care RNs with specialty nursing certification, higher percentage of direct care RNs with associate's degree in nursing, and higher percentage of direct care RNs with BSN, MSN, or doctoral degree are associated with HA-CAUTI occurrence. The results also support the association of the following factors with HA-CAUTI identified by previous studies: Female gender; older age (>50 years); longer length of stay; severe underlying disease; glucose lab results (>200 mg/dL); longer use of the catheter; and RN staffing. CONCLUSIONS: Additional findings from this study demonstrated that the presence of more nurses with specialty nursing certifications can reduce HA-CAUTI occurrence. While there may be valid reasons for leaving in a urinary catheter, findings show that having a catheter in for more than 48 hours contributes to HA-CAUTI occurrence. Finally, the findings suggest that more nursing hours per patient-day are related to better patient outcomes.

KW - Catheter-associated urinary tract infections

KW - Data mining

KW - Urinary catheter

KW - Urinary tract infections

UR - http://www.scopus.com/inward/record.url?scp=85045673639&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85045673639&partnerID=8YFLogxK

U2 - 10.1097/WON.0000000000000409

DO - 10.1097/WON.0000000000000409

M3 - Article

VL - 45

SP - 168

EP - 173

JO - Journal of Wound, Ostomy and Continence Nursing

JF - Journal of Wound, Ostomy and Continence Nursing

SN - 1071-5754

IS - 2

ER -