Abstract
Background and Aims: Breast cancer surgery is the most common cancer operation in the United States. While well established clinical trials performed more than 20 years ago have clearly established the equivalency of partial mastectomy and mastectomy regarding the endpoint of overall survival for breast cancer, controversies in the surgical management of breast cancer remain. Few guidelines exist regarding performance metrics and quality outcomes specific to breast cancer surgery. In order to identify more meaningful breast cancer surgery quality measures, the Vermont Breast Cancer Surgery Outcomes (VBCSO) database was initiated at the University of Vermont in 2003 and has been maintained continuously and successfully tracked outcomes of initial surgical treatment for over 1000 incident breast cancers. The database was specifically designed to assess measures of quality for a low risk cancer surgery that might impact patient treatment decisions, requirements for additional surgery, and perhaps ultimately cancer recurrence. We have recently been funded to expand this database to three CRN sites to evaluate surgical outcomes from a larger number of surgeons, hospitals and geographic regions. Methods: We will create a database of all women diagnosed with a primary breast cancer undergoing breast cancer surgery between 2003–2008 at three CRN sites (Kaiser Permanente Colorado, Group Health Cooperative, and Marshfield Clinic) using the Virtual Data Warehouse and medical record review. This database will be combined with the existing VBCSO database to evaluate potential measures of surgical quality. Results: To demonstrate the value of establishing this multicenter database, and the potential impact it could have on the quality of surgical treatment provided to breast cancer patients, the data will be analyzed initially to identify sources of (1) variation in mastectomy rate, (2) variation in initial partial mastectomy positive margin rates, and (3) variation in re-excision rates following initial partial mastectomy. Conclusion: An electronic clinical network that has detailed data on breast cancer diagnoses, surgical procedures, and surgical outcomes will allow for several comparative effectiveness research hypotheses to be tested. The development of data protocols and data capture tools as part of this proposal will facilitate participation by a larger network of hospitals in the future.