Degree
Doctor of Philosophy (PhD)
Department
Industrial Engineering
Document Type
Dissertation
Abstract
There is increasing acknowledgment of the impact that Social Determinants of Health (SDOH) have on health outcomes, especially for chronic illnesses. Chronic conditions such as congestive heart failure (CHF) are the leading cause of hospitalization among older adults in the United States. Understanding the complex connections between SDOH, socioeconomic factors affecting health, and CHF can help reduce the high rates of morbidity, mortality, readmission, and costs associated with CHF. In this project, we aim to explore known and unknown SDOH influences on CHF by analyzing patient data from the electronic health records at Our Lady of the Lake Regional Medical Center (OLOL RMC), supported by a modern approach that uses unsupervised machine learning techniques to identify themes and hidden relationships from thousands of medical publications on SDOH and CHF. This approach is more efficient than time-consuming or labor-intensive traditional review methods. The specific goal of this study is to thoroughly examine the relationship between SDOH and CHF. To accomplish this, we will apply a topic modeling technique called Latent Dirichlet Allocation and hierarchical clustering using the statistical software R to identify specific trends, patterns, and correlations. Then, in collaboration with OLOL RMC, we will validate these relationships using SDOH information extracted from the hospital’s electronic health records. Finally, we will analyze the relationships between readmission rates, the Epic Risk Score, and sepsis OPA alerts. By identifying key SDOH factors that significantly impact CHF outcomes, we aim to guide the development of interventions and policies to reduce these disparities and promote health equity. The findings and connections established by this research will lay a foundation for future studies on CHF and SDOH, which are of interest to NIH and national nonprofits such as the American Heart Association.
Date
10-29-2025
Recommended Citation
Solaru, Ifeoluwa R., "A Multi-Level Analytics Framework for Understanding CHF and Readmissions: Integrating Natural Language Processing of Literature with EHR-Based SDOH Modeling" (2025). LSU Doctoral Dissertations. 6927.
https://repository.lsu.edu/gradschool_dissertations/6927
Committee Chair
Ikuma, Laura