Nancy L. Keating is the principal investigator of this project.

Guidelines and consensus statements for the treatment of early-stage breast cancer state that (a) breast conservation offers survival rates equivalent to mastectomy, and (b) patients undergoing breast-conserving surgery (BCS) should also receive radiation. These statements are based on clinical trial data; however, only 2.5% of cancer patients enroll in clinical trials; findings from trials, therefore, do not always provide definitive guidance for the treatment of all patients. Those who enroll in clinical trials tend to be younger and healthier than other patients, and they are typically cared for by providers who practice at or are affiliated with cancer centers. Moreover, clinical trials may be underpowered to assess outcomes among subgroups of patients with certain demographic or clinical characteristics. Through careful design and analysis, observational data may address questions regarding treatment effectiveness that are not feasible to answer within the context of randomized trials.

This study will use propensity score techniques to examine survival for up to 10 years after local treatment of breast cancer among community-based cohorts of women living in various regions of the United States. The research will assess outcomes among all women and among subgroups of women—such as older women, women with more comorbid illness, and women with advanced tumors—for whom few clinical trial data are available to guide treatment decisions. The cohorts include 63,750 women living in a SEER region and diagnosed with stage I or II breast cancer during the period from 1991 to 2002 and included in the SEER-Medicare database, and 3,828 women in Minnesota or Massachusetts and diagnosed with stage I or II breast cancer during the period from 1993 to 1995. Specifically, we will:

• assess whether long-term survival for mastectomy versus BCS with radiation is equivalent across clinically distinct groups of women with early-stage breast cancer in community populations;
• assess whether long-term survival is equivalent among women with early-stage breast cancer who do and do not undergo radiation therapy following BCS in community populations;
• assess the robustness of our estimates of treatment effectiveness obtained for the previous aims using instrumental variable techniques.

By providing reliable estimates of treatment effectiveness, this work will help clinicians guide treatment choices for women with early-stage breast cancer, including women for whom few clinical trial data are available. Moreover, the work will provide insight into different techniques for the analysis of observational data.