Area-based measures of socioeconomic status (SES) are widely used in studies of cancer survival. There is consistent evidence that survival varies by SES for many malignancies and that this relationship is resilient to the choice of area-based measure. The specific role of SES in population-based survival statistics and the mechanisms by which SES affects cancer survival are not fully understood.
Three separate studies were conducted utilizing population-based cancer registry data from the Surveillance, Epidemiology, and End Results (SEER) Program of the U.S. National Cancer Institute. The percent of the census tract population living below the poverty level was chosen as the area-based SES measure of interest based on previous research of health disparities using linked census data. The goals of these studies were: 1) to evaluate the validity of one area-based measure of SES; 2) to explore the effect of SES-specific mortality in the calculation of relative survival; and 3) to examine the relationships between SES and survival from non-localized prostate cancer.
Study I identified practical steps to improve geocoding outcomes of registry data and revealed substantial misclassification of SES when geocoding at the level of the ZIP code. Study II demonstrated that the use of SES-specific background mortality in the calculation of relative survival produced a reduction in the survival disparity observed when using national data that do not take SES into consideration. This observation was most pronounced in SES-stratified analyses. Study III, utilizing SEER-Medicare linked data, identified a significant interaction between SES and stage of disease at diagnosis and showed that while survival by SES does differ among study individuals with equal eligibility for care, much of the SES related disparity is explained by factors for which SES may serve as a surrogate.
Area-based measures of SES play an important role in population-based studies of cancer survival and need to be more fully utilized in the calculation of relative survival. It is important to understand and evaluate the completeness and accuracy of the geocoded data from which area-based measures are obtained. Further research is needed to elucidate the specific, complex pathways by which SES affects cancer survival.
Table of Contents
Chapter 1. Population-Based Cancer Registries
Chapter 2. Relative Survival
Chapter 3. Geocoding and Area-Based Measures of Socioeconomic Status
Chapter 4. Geocoded Registry Data and the Use of Area-Based Measures of Socioeconomic Status: Misclassification and Steps to Minimize its Effects
Chapter 5. Determining Population-Based Relative Survival: The Importance of Using Local Area and SES-Specific Measures of Background Mortality
Chapter 6. Examining the Role of Area-Based Poverty in Survival from Non-Localized Prostate Cancer in the Medicare Population
Chapter 7. Conclusions
About this thesis
|Committee Chair / Thesis Advisor|
|Population-Based Cancer Registries: The Role of Area-Based Measures of Socioeconomic Status and Cancer Survival ()||2018-08-28||