UC Irvine

Not only is ovarian cancer the 5th leading cause of cancer death in American women, it is also the deadliest of the gynecological cancers and among the malignancies of which disparities in care and outcomes in underserved populations are prominent. The aims of this dissertation were to investigate the association between residential geographic location and ovarian cancer outcomes, while considering the relationship between race and socioeconomic factors. Incident ovarian cancer cases were ascertained from the California Cancer Registry for women diagnosed between 1996 and 2014. Adherence to the National Comprehensive Cancer Network (NCCN) guidelines for treatment was used as a binary measure of receipt of quality care. Two geographic variables assessing access to care were also considered: distance traveled to receive care and the proximity of the closest high-quality-of-care (QOC) hospital. Spatial analyses using generalized additive models found geographic location to be an independent predictor of NCCN treatment adherence for women in California. Women of lower socioeconomic status and minority race/ethnicity were found to receive less quality care and to be disproportionately affected by geographic barriers. While spatial analyses identified location as an independent predictor of ovarian cancer survival, location no longer had an effect on survival after adjusting for sociodemographic variables, receipt of NCCN care, and geographic access to care. To assess the impact of air pollution, ozone, particulate matter with diameter less than 2.5 microns (PM2.5), and nitrogen dioxide (NO2) data was extracted from California Air Resources Board’s (CARB) online database, Air Quality and Meteorological Information System (AQMIS). Monthly averages of each pollutant were linked to women’s residential address and calculated over their survival period. Distance to nearest major roadway was determined to account for local traffic. The analyses are suggestive of a potential association between ovarian cancer survival and NO2 and PM2.5 exposure in California, independent of sociodemographic and treatment factors. The impact of these pollutants was greatest among women in early stages. Null associations were observed for ozone and distance to road when examined alone, although they had marginal effects when examining them in multipollutant models.