Screening for breast cancer reduces mortality. Following the implementation of screening programs, European studies have shown that the incidence of early-stage and in situ breast cancers increased substantially. This was followed by a decline in the incidence of advanced-stage breast cancers, which approximated the size of the reduction in mortality rates. This pathway of effects towards mortality reduction resulting from screening would also be expected in the United States, but studies examining such reduction are currently lacking. Monitoring the effects of breast cancer screening in the United States among counties is clearly beneficial since this will facilitate local health planning and allocation of screening resources. Although screening use varies geographically, the extent to which changes over time in the effects of screening have varied among different counties in the United States is unclear. One explanation for the geographic variation in screening is that women from socioeconomically deprived areas are less likely to be screened. Because of the importance of physician recommendation to breast cancer screening and the mediating effect of primary care physician and mammography facility availability in the association between area deprivation and various health outcomes, we hypothesize that the differences in effect of breast cancer screening among counties with varying socioeconomic conditions will be mediated by the availability of primary care physicians and mammography facilities locally. In the second revision of this proposal, we will use an ecological study to: (1) Investigate the spatio-temporal variation of early stage breast cancer, advanced stage breast cancer, and breast cancer mortality at the county-level using 1988-2005 data from nine SEER programs and 200 counties; (2) Determine the association of the county-specific variation of breast cancer screening prevalence with the county-specific incidence of advanced stage breast cancer and with breast cancer mortality; (3) Determine if temporal changes in early stage breast cancer, advanced stage breast cancer, and breast cancer mortality vary by county socioeconomic conditions; and (4) Determine if the availability of primary care physicians and mammography facilities at the county level explains the temporal differences between affluent and socioeconomically deprived counties in the incidence of early stage breast cancer, advanced stage breast cancer, and breast cancer mortality. Bayesian methods will be used as part of spatio-temporal models.
Monitoring the effects of breast cancer screening in the United States among counties is clearly beneficial since this will facilitate local health planning and allocation of screening resources. We will use an ecological study to: (1) Investigate the spatio-temporal variation of early stage breast cancer, advanced stage breast cancer, and breast cancer mortality at the county-level using 1988-2005 data from nine SEER programs and 200 counties; (2) Determine the association of the county-specific variation of breast cancer screening prevalence with the county-specific incidence of advanced stage breast cancer and with breast cancer mortality; (3) Determine if temporal changes in early stage breast cancer, advanced stage breast cancer, and breast cancer mortality vary by county socioeconomic conditions; and (4) Determine if the availability of primary care physicians and mammography facilities at the county level explains the temporal differences between affluent and socioeconomically deprived counties in the incidence of early stage breast cancer, advanced stage breast cancer, and breast cancer mortality. ? ? ?
