Exploring Glioblastoma Caregiving: From Personal Narratives to AI-Driven Insights and Real-Time Data
Publication Date
2025
Document Type
Dissertation/Thesis
First Advisor
Hughes, M. Courtney
Degree Name
Ph.D. (Doctor of Philosophy)
Legacy Department
School of Health Studies
Abstract
Glioblastoma multiforme (GBM) is an aggressive primary brain tumor with a rapid disease trajectory, and it creates profound challenges for family caregivers (GBM-FCs). Despite their critical role in patient care, GBM-FCs often face substantial physical, emotional, and logistical demands with minimal guidance from the medical community. This three-paper dissertation investigates GBM caregiving needs and explores innovative technological approaches to better understand and address GBM-FCs’ challenges. Paper one utilized a qualitative design to examine the lived experiences of GBM-FCs through semi-structured interviews. Findings revealed that GBM-FCs frequently lack assistance and guidance for practical, day-to-day caregiving tasks including mobility, medication management, and communication with healthcare providers. Additionally, hospice, palliative, and rehabilitation services were consistently underutilized, mostly due to lack of awareness or misconceptions about their purpose. These findings highlight the need for earlier and individualized supportive care interventions throughout the GBM disease course. Paper two evaluated the use of large language models (LLMs) as an adjunct to traditional thematic analysis of brain tumor support forum data. Analyses compared GPT-4, GPT-3.5, and Llama outputs to human-coded themes. GPT-4 demonstrated the highest performance in terms of accuracy, efficiency, and similarity with human-led analysis. Results indicate that LLMs may be able to serve as a valid and efficient complement to human analysis. Paper three piloted a novel ecological momentary assessment (EMA) protocol with health sciences students simulating the role of GBM-FCs. The protocol included daily text- and image-based prompts, which were analyzed using LLMs. Results demonstrated overall feasibility and usability of the EMA protocol, with most participants completing the majority of prompts and a small number reporting difficulty with full participation. Students also provided increased empathy ratings for GBM-FCs at the end of the study, potentially indicating that simulation experiences are helpful in fostering empathy in health sciences students. This approach illustrated that EMA protocol could be successful in a population with competing demands, thereby providing direction for future caregiver-based EMA studies. Collectively, these studies contribute to the literature by identifying unmet needs among GBM-FCs, demonstrating the utilizing of large language models in qualitative research, and validating an innovative EMA methodology for future caregiver studies. Together, they provide a foundation for the development of individualized, technology-enhanced interventions to support GBM-FCs across the disease continuum.
Recommended Citation
Muasher-Kerwin, Christina, "Exploring Glioblastoma Caregiving: From Personal Narratives to AI-Driven Insights and Real-Time Data" (2025). Graduate Research Theses & Dissertations. 8169.
https://huskiecommons.lib.niu.edu/allgraduate-thesesdissertations/8169
Extent
117 pages
Language
en
Publisher
Northern Illinois University
Rights Statement
In Copyright
Rights Statement 2
NIU theses are protected by copyright. They may be viewed from Huskie Commons for any purpose, but reproduction or distribution in any format is prohibited without the written permission of the authors.
Media Type
Text
