A62

Breast cancer, the most commonly diagnosed cancer and the second leading cause of death among women, disproportionately affects certain areas of the United States, including many areas within Kentucky. We had previously found evidence that longer travel distance may adversely affect early cancer diagnosis for the population living in a rural state, indicating a need for more targeted prevention policy related to access to health care services. Because availability and accessibility of health care services that may lead to effective prevention of disease is of particular interest, we were further interested in identifying such geographic areas with excess proportion of late stage breast cancer cases so that we may further evaluate the effectiveness of prevention program based on the results. Here we particularly focus on breast cancer disparities in Appalachian Kentucky that includes most medically underserved and distressed, and economically disadvantaged areas within the state. This study utilizes GIS-based spatial statistical methods for the identification of late stage breast cancer clustering in the study area. Data for female breast cancer cases (n=14130) diagnosed in Kentucky between 1999 and 2003 were obtained from Kentucky Cancer Registry, and cases with unknown staging information (n=693) were excluded from the analysis. Geographic difference in late stage breast cancer (TNM III & IV, n=4853) throughout the state was examined in two geographic scales (county and census tract) under the null hypothesis that proportion of women diagnosed with late stage breast cancer is same across the study area. Geographic variation in the ratio of early to late stage and in the most advanced form (distant stage only, n=625) were also examined for further analysis. Those statistically significant geographic areas were identified and mapped using GIS. Preliminary analyses suggest that there are geographic differences in late stage breast cancer in Kentucky and in Appalachian Kentucky; we found geographic areas with high proportion of late stage breast cancer cases, but also identified those geographic areas with excessive numbers of distant stage cases. In addition, we explored such geographic difference in two geographic scales and found similar results. Our findings indicate that there are geographic areas where more targeted education and screening program should be directed to reduce the burden of breast cancer. Because early detection is a key to the prevention of breast cancer, these findings may indicate an urgent need to evaluate the effectiveness of screening programs in these geographic areas, thus reducing breast cancer disparities related to access to health care services.

First AACR International Conference on the Science of Cancer Health Disparities-- Nov 27-30, 2007; Atlanta, GA