| dc.description.abstract | This dissertation used multilevel regression methods to explore rural-urban disparities in stage at diagnosis of, and treatment of early-stage breast cancer among women diagnosed between 1995 and 2012, and whether other geographical characteristics explained these disparities. Trends in probability of late-stage diagnosis and treatment were explored.
The Surveillance, Epidemiology and Ends Results data was used to explore rural-urban disparities in stage at diagnosis of breast cancer and whether county poverty estimates explain these disparities. Compared to urban metropolitan residents, rural residents had higher odds of late-stage diagnosis (OR=1.09; CI= 1.03-1.15). County poverty explained the association between late stage diagnosis and urban nonmetropolitan (UNM) but not rural residence. Across all years, the probability of late-stage diagnosis was highest for residents of rural high poverty areas.
The Texas Cancer Registry (TCR) was utilized to explore rural-urban disparities in stage at diagnosis, and whether racial residential segregation, census tract (CT) poverty and travel distance to the nearest primary care provider (PCP) explained these disparities. Compared to urban metropolitan residents, residents of UNM residents had higher odds of late-stage diagnosis (OR=1.08; 95% CI=1.03 -1.13). Racial residential segregation, CT poverty and distance to the nearest PCP did not explain residential disparities. Residents of UNM high CT poverty areas had the highest probability of late-stage diagnosis across all years.
The TCR was also used to explore rural-urban disparities in surgical treatment, type of surgery and adjuvant radiotherapy, and whether CT poverty and residential segregation explained these disparities. Compared to urban metropolitan residents, UNM residents had higher odds of having surgery (OR=1.39; 95%=1.27-1.53) and mastectomy (OR=1.18; 95%=1.12-1.25); UNM (OR=1.16; 95%=1.01-1.34) and rural (OR=1.66; 95% CI=1.12-2.44) residents had higher odds of adjuvant radiotherapy. CT poverty and racial residential segregation explained the association between rural residence and surgical treatment. Across all years, residents of urban metropolitan high poverty CT had the lowest probability of being treated surgically and with adjuvant radiotherapy, and the highest probability of being treated using mastectomy.
In conclusion, interventions aimed at reducing disparities in breast cancer diagnosis and treatment should be guided by both residence and area poverty levels. | en |
