Geographic Variation of Breast Cancer Survival

描述（由申请人提供）：乳腺癌是美国女性癌症死亡的第二大原因。大约80%的此类死亡发生在65岁或以上的妇女中。这些妇女的存活率也较低，接受推荐治疗的可能性也较小。确定65岁或65岁以上妇女乳腺癌存活率较低的地理区域以及造成这种情况的潜在因素，对于确保解决治疗差异以及能够在能够实施的水平上设计干预措施非常重要。本修订应用程序的目的是：1)检查66岁及以上妇女乳腺癌生存率的小区域地理聚类；2)确定66岁及以上妇女生存的地理变异在多大程度上可以用地区社会经济剥夺的地理变异来解释；3)找出可能的途径，通过该地区的社会和经济剥夺来解释66岁及以上女性乳腺癌生存率的任何地理差异。根据审稿人的意见，我们将概念模型和相关假设重新聚焦于描述区域社会和经济剥夺的地理变化，以及区域剥夺影响这一结果的途径。分析模型也被重新聚焦，并以更详细的统计描述。为了实现具体目标，将链接以下数据来源：1)1992-1999年监测、流行病学和最终结果（SEER）项目的数据（生存、人口统计学、诊断、治疗阶段、肿瘤生物学）；2) 1991- 1999年医疗保险数据及其与SEER数据的联系（合并症，治疗）；3) 1992-1999年医疗保险和医疗补助中心的数据（医疗服务的可获得性和接近性）和4)1992-1999年人口普查期间的估计（地区剥夺措施）。将进行乳腺癌生存的高级贝叶斯分析。拟议研究的结果将确定生存的地理差异和负责因素，可用于制定和实施未来的干预措施，以减少这些差异。

DESCRIPTION (provided by applicant): Breast cancer is the second leading cause of cancer death among women in the United States. About 80% of such deaths occur in women age 65 years or older. These women also have a lower survival rate and are less likely to receive recommended treatment. Identifying geographic areas where survival from breast cancer among women 65 years of age or older is lower and the underlying factors responsible is important to ensure that differences in treatment are addressed and that interventions can be designed at a level where they can be implemented. The aims of this revised application are to: 1) examine small-area geographic clustering of breast cancer survival among women 66 years of age and older; 2) Determine the extent to which geographic variation of survival can be explained by the geographic variation in area social and economic deprivation among women age 66 and older; and 3) identify potential pathways by which area social and economic deprivation explains any geographic variation of breast cancer survival among women age 66 and older. Based on the reviewers' comments, we have refocused the conceptual model and associated hypotheses on delineating the geographic variation of area social and economic deprivation on the geographical variation of breast cancer survival and the pathways by which area deprivation influences this outcome. The analytic models have also been refocused and are described in more statistical detail. To address the specific aims, the following data sources will be linked: 1) 1992-1999 data from the Surveillance, Epidemiology, and End Results (SEER) program (survival, demographics, stage at diagnosis, treatment, tumor biology); 2) 1991- 1999 Medicare data and its linkage to the SEER data (comorbidity, treatment); 3) 1992-1999 data from the Centers for Medicare and Medicaid (availability of and proximity to medical care) and 4) the 1992-1999 intercensal estimates (area deprivation measures). Advanced Bayesian analyses of breast cancer survival will be performed. The findings of the proposed study will identify geographic disparities in survival and factors responsible, which can be used to develop and implement future interventions to reduce these disparities.