Date of Award

Fall 1-10-2020

Degree Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Public Health

First Advisor

Lee Mobley, PhD

Second Advisor

Harry Heiman, MD

Third Advisor

May Kuo, PhD

Abstract

Background: Approximately 51% of women with cervical cancer (CVC) are diagnosed at a late stage (regional or distant), an outcome associated with increased morbidity and mortality.African American, and Hispanic women,and women residing in specific geographic regions of the (US) are among those most heavily burdened by late stage CVC.The cause(s) of these disparities are multifaceted and not well understood. However, the most significant predictor of late stage CVC diagnosis and disparities is current screening, which is largely impacted by access to care. Thus, the goal of this study was to identify access-related predictors of late stage CVC and develop a comprehensive understanding of where and why racial and geographic disparities in late stage CVC occur. Methods: This three-part study examined primary CVC cases diagnosed between the years of 2005-2014, from the United States Cancer Statistics (USCS) database. The final sample included 120,325 CVC cases within 43 states and their 2,357constituent counties. First, Empirical Bayes LISA clustering methods were applied to identify clusters of counties considered to be high risk for late stage CVC “hotspots” during two 5-year time periods (pre- and post-2010). Second, a series of T-tests were conducted to determine whether various contextual and compositional factors were significantly different in hotspots versus other places. Third, two Generalized Linear Mixed Models (GzLMM), using data from person and county levels, were estimated to identify predictors of late stage CVC diagnosis and racial or ethnic disparities among women with CVC in the US. Lastly, a General Linear Mixed Model (GLMM) using data from county- and state –levels was estimated to examine predictors of higher proportions of late stage CVC among counties. Results: Primary care physician shortage areas, Planned Parenthood (PPH) clinics, area-level poverty rates, area-level uninsured rates, percent of immigrants from other countries, state CVC screening mandates and prevalence of self-insured employer health plans were all statistically significant predictors of access to care associated with late stage CVC diagnoses and geographic disparities. We also found that PPH clinics play an important role in reducing the odds of late stage CVC among Hispanic women with CVC. Conclusion: Access to CVC screening plays a significant role in the etiological pathway to late stage CVC diagnoses and disparities. Given that significant access barriers occurred at various ecological and geographical levels, it is recommend that future research and intervention efforts begin to focus on multilevel and/or spatial approaches. Without further exploration of the factors impacting late stage CVC diagnoses, CVC mortality rates will remain high and at a disproportionately higher rate for women in various geographical areas and among African American and Hispanic women.

File Upload Confirmation

1

Available for download on Friday, October 02, 2020

Share

COinS