An Identification of the Knowledge Structure on the Resilience of Caregivers of People with Dementia using a Text Network Analysis

Abstract

The purpose of this study is to identify the knowledge structure of the concept of resilience of dementia caregivers and to understand core topics and the trends in dementia caregiver resilience research over time. Methods: This study is a quantitative content analysis using text network analysis on the studies on resilience of caregivers of people with dementia. We searched the available literature in two scientific databases, PubMed and the Web of Science. Term frequency, the number of co-occurrence, and centrality were analyzed, and Latent Dirichlet Allocation (LDA) analysis was conducted with Net Miner program. Results: A total 187 articles were identified. The core keywords found were "care", "family", "support", "burden", "intervention", "health", "stress", "relationship", "symptom", "depression", "partner", and "quality of life". Four main topics were identified through the LDA analysis, such as "support from family and social resources", "hardships to overcome", "building strength to overcome adversity", and "coping with negative emotions". There was only one study before 2000, but the number of studies steadily increased to 50 for the years 2011~2015, and 111 for the years 2016~2020. Conclusion: This is the first study to analyze the knowledge structure and research trend of resilience research of caregivers of people with dementia through social network analysis and topic modeling. This study provides a scientific basis of the future studies in Korea. Moreover, it is meaningful in that it brought a comprehensive understanding of the studies by comprehensively analyzing the studies of the resilience of caregivers of people with dementia.

How about an Educational Framework for Nursing Staff in Long-Term Care Facilities to Improve the Care of Behavioral and Psychological Symptoms of Dementia?

“…As BPSD care covers an extensive range of disciplines, including medicine, psychology, and nursing, the traditional method of reviewing the relevant literature would require a great deal of time and labor. Because a machine learning-based literature review can be an efficient and appropriate alternative method [ 22 , 23 ], we used topic modeling. Topic modeling is a technique that can infer latent topics from a large set of documents using computer software that prevents any human biases [ 23 , 24 ], and the topics it extracts indicate the core insights and values of the comprehensive knowledge body [ 25 ].…”