Thesis
Mixture Hidden Markov Model for clustering of nurse fatigue patterns
University of Iowa
Master of Science (MS), University of Iowa
Summer 2020
DOI: 10.17077/etd.005568
Abstract
Workplace fatigue has implications for the safety and productivity of health care workers, of which nurses are a large part. Additionally, workplace fatigue is shown to have effects on patient outcomes as well. Excess workplace fatigue in nurses can lead to a loss in productivity and a rise in the number of accidents, near misses and injuries. Fatigue in nurses is affected by a number of factors such as length of shift, work schedules, quality and quantity of sleep, and other family-related and environmental variables. Many studies have been conducted to analyze the effects of these factors on fatigue. However, the existing studies of fatigues were mostly descriptive and based on self-report surveys.
This study quantitatively models the changes in fatigue levels of nurses over time and groups nurses with similar fatigue behaviors using a Mixture Hidden Markov Model. A six-cluster model is used to model the fatigue patterns in a data set collected from nurses at eight hospitals in Iowa. We propose a state emission matrix with a special structure that can improve interpretability and simplicity of the model. A genetic algorithm based search method is implemented to find the optimal model configuration of the Mixture Hidden Markov Model. For each cluster with at least two states, the effects of work-shift scheduling information on the state transition probabilities are studied by a within-cluster Hidden Markov Models with covariates. Based on this model, strategies for mitigating fatigue can be implemented such that they target nurses specifically based on clustering of their fatigue patterns.
Details
- Title: Subtitle
- Mixture Hidden Markov Model for clustering of nurse fatigue patterns
- Creators
- Aditya Gune
- Contributors
- Yong Chen (Advisor)Chao Wang (Committee Member)Amany Farag (Committee Member)
- Resource Type
- Thesis
- Degree Awarded
- Master of Science (MS), University of Iowa
- Degree in
- Industrial Engineering
- Date degree season
- Summer 2020
- DOI
- 10.17077/etd.005568
- Publisher
- University of Iowa
- Number of pages
- x, 55 pages
- Copyright
- Copyright 2020 Aditya Gune
- Language
- English
- Description illustrations
- color illustrations
- Description bibliographic
- Includes bibliographical references (pages 52-55).
- Public Abstract (ETD)
Fatigue in medical workers such as nurses can lead to accidents and near miss accidents. In environments like hospitals, this has the potential to have an adverse effect on the health and safety of the nurses as well as the patients. Recovery from fatigue is affected by various factors such as sleep, work schedules and family-related stress factors. The factors affecting fatigue and fatigue recovery in nurses have been studied in great detail in the past. However, the existing studies of fatigues were mostly descriptive and based on self-report surveys.
This study quantitatively models the changes in fatigue levels of nurses over time and groups nurses with similar fatigue behaviors together using daily fatigue data collected from nurses at eight hospitals in Iowa. Within each group, the effect of work-shift scheduling on nurse fatigue is further analyzed. With the results obtained from this study, it is possible to develop scheduling and intervention strategies that target nurses of high risk groups to effectively mitigate nurse fatigue.
- Academic Unit
- Industrial and Systems Engineering
- Record Identifier
- 9983987794502771
Metrics
54 File views/ downloads
60 Record Views