Abstract
Massive generation of health-related data has been key in enabling the big data science initiative to gain new insights in healthcare. Nursing can benefit from this era of big data science, as there is a growing need for new discoveries from large quantities of nursing data to provide evidence-based care. However, there are few nursing studies using big data analytics. The purpose of this article is to explain a knowledge discovery and data mining approach that was employed to discover knowledge about hospital-acquired catheter-associated urinary tract infections from multiple data sources, including electronic health records and nurse staffing data. Three different machine learning techniques are described: decision trees, logistic regression, and support vector machines. The decision tree model created rules to interpret relationships among associated factors of hospital-acquired catheter-associated urinary tract infections. The logistic regression model showed what factors were related to a higher risk of hospital-acquired catheter-associated urinary tract infections. The support vector machines model was included to compare performance with the other two interpretable models. This article introduces the examples of cutting-edge machine learning approaches that will advance secondary use of electronic health records and integration of multiple data sources as well as provide evidence necessary to guide nursing professionals in practice.
Original language | English (US) |
---|---|
Pages (from-to) | 28-35 |
Number of pages | 8 |
Journal | CIN - Computers Informatics Nursing |
Volume | 38 |
Issue number | 1 |
DOIs | |
State | Published - Jan 1 2020 |
Bibliographical note
Publisher Copyright:Copyright © 2019 Wolters Kluwer Health, Inc. All rights reserved.
Keywords
- Catheter-associated urinary tract infections
- Data mining
- Machine learning