Cancer Control Research5R01CA137750-04
GEOGRAPHIC DISPARITIES IN COLORECTAL CANCER SURVIVAL
ABSTRACT Colorectal cancer (CRC) is the fourth most common cancer in the United States and the second leading cause of cancer deaths. Despite declines in CRC incidence rates, survival following diagnosis has improved only modestly over the past few decades. Even though environmental contexts play an important role in health, disease, and behavior, most studies about CRC survival have largely ignored the geographic variation and importance of area-level socioeconomic conditions that have been associated with CRC screening, CRC stage at diagnosis, and with the survival of many other cancers. CRC patients who live in areas with worse socioeconomic conditions have decreased survival than those who live under more affluent conditions; however, the mechanisms by which these area-level factors exert their influence on CRC survival remain unclear. This amended application of the proposed population-based, prospective study has three specific aims. Aim 1) Determine the extent of the geographic variation of CRC survival across the United States based on small geographic areas at the sub-county (census-tract) level using the linked Surveillance, Epidemiology, and End Results (SEER)-Medicare data for over 100,000 men and women aged 66 or older diagnosed with CRC. Aim 2) Determine the extent to which lower CRC survival can be explained by higher area socioeconomic deprivation among persons age 66 and older. Aim 3) Identify potential mediating pathways by which higher area socioeconomic deprivation is associated with lower CRC survival among persons age 66 and older, namely a) patient characteristics, b) physician and hospital characteristics, c) tumor characteristics, d) type of treatment received, and e) surveillance for CRC after diagnosis to detect recurrence and metastases. A multilevel spatial model will be developed to address the specific aims of the proposed study. We will use the following existing data sources: 1) 1992-2005 data from NCI's SEER program (survival, patient characteristics, type of treatment, tumor characteristics); 2) 1991-2005 Medicare data (patient characteristics, type of treatment, surveillance after diagnosis) which is linked to the SEER data; 3) 1991-2005 data from the Provider of Services File (hospital and physician characteristics), 4) 1990-2005 census data (area deprivation measures), and 5) Medicare Current Beneficiary Survey data. Advanced Bayesian spatial analyses of CRC survival will be performed and a geographic information system will be used to display the results. Our study will increase understanding of and identify important mechanisms of the role of area-level socioeconomic deprivation on CRC survival. In addition, this study will help provide opportunities for targeting specific geographic areas to allocate resources and interventions locally to improve CRC survival using evidence- based approaches, thereby reducing health disparities associated with living in socioeconomically deprived areas.