Date on Master's Thesis/Doctoral Dissertation


Document Type

Doctoral Dissertation

Degree Name

Ph. D.


Health Management and Systems Sciences

Degree Program

Public Health Sciences with a specialization in Health Management, PhD

Committee Chair

Karimi, Seyed

Committee Co-Chair (if applicable)

Little, Bert

Committee Member

Little, Bert

Committee Member

Egger, Michael

Committee Member

Antimisiaris, Demetra

Author's Keywords

causal diagrams; air pollution epidemiology; NSCLC; survival outcomes; treatment disparities; SEER


Background: Lung cancer is the second leading cause of death in the United States (U.S.). The most prevalent histological type of lung cancer is Non-Small Cell Lung Cancer (NSCLC), which has an overall five years survival of 5% if left untreated. Therefore, early treatment of NSCLC is vital to improve overall survival (OS) outcomes. Several factors affect survival outcomes, which can be categorized as modifiable or non-modifiable. The difference in timely receipt of guideline-concordant treatment affects the survival outcomes of patients with stage 1A NSCLC. This dissertation explored factors that affect treatment and survival outcomes among stage 1A NSCLC patients using a nationally representative cancer registry population (i.e., Survival, Epidemiology, and End Results (SEER) 18 plus cancer registry), air pollution and weather data, and local health resource information. Methods: The first chapter of this study utilized an evidence-based Directed Acyclic Graph (DAG) synthesis method to review the causal pathways explored in the literature for factors affecting treatment receipt and survival outcomes among stage 1A NSCLC patients in the U.S. Subsequently, guided by the theoretical frameworks proposed previously Andersen and Aday1 ; Shi and Steven2, DAGs for the second and third chapter analyses were proposed, respectively. The second chapter utilized logistic regression adjusted for the year of diagnosis and county-specific time-invariant factors with standard errors clustered at the state level. The third chapter employed Kaplan-Meier survival estimates and a Cox proportional regression model adjusted for the year of diagnosis and county-specific time-invariant factors to determine survival outcomes, allowing for right censoring. Findings: The DAGs identified several causal pathways that were accounted for in Chapters Two and Three analyses. The results of the Chapter Two analysis corroborated with the existing literature that there exists a difference in guideline-concordant treatment receipt. The findings of the Chapter Three analysis confirmed a difference in survival outcomes among stage 1A TN0M0 NSCLC first primary patients exposed to higher versus low levels of air pollution in the U.S. Policy implications: Black and Medicaid enrolees are less likely to receive guideline-concordant treatment than others are. This warrants future policy decision-making geared toward reducing the difference in treatment receipt, which ultimately improves survival outcomes. Additionally, the shortage of sufficient air pollution monitoring stations in non-metropolitan areas warrants an improvement in determining the health outcomes for non-metropolitan residents. Ambient air pollution control policies are required to improve the survival outcomes of patients with stage 1A TN0M0 NSCLC.