Grant #: U01
PI Name: Smith-Bindman R
Title: The Sensitivity of Medicare Billing Claims for Monitoring Mammography use by Elderly Women
Institute: National Cancer Institute

Using prospectively collected data from the San Francisco, New Mexico, and Washington State Breast Cancer Surveillance Consortium (BCSC) registries, linked with data from SEER-Medicare for the same geographical regions from 1992–1996, the investigators will assess whether Medicare physician claims can be used to accurately distinguish screening from diagnostic mammography among 4140 elderly women with breast cancer. Specifically, they will:

1. Compare the accuracy of an algorithm developed from Medicare claims for distinguishing whether mammography is obtained for screening or diagnostic purposes by comparing the algorithm's classification of type of examination with the corresponding information from three BCSC mammography registries.
2. If the algorithm for differentiating type of mammographic examination based on the SEER-Medicare data is valid, i.e. at least 90% accurate, the SEER-Medicare database will be used to evaluate screening utilization among Medicare recipients, and will evaluate how utilization varies by age and racial/ethnic group.
3. If the algorithm for differentiating type of mammographic examination based on the SEER-Medicare data is valid, i.e. at least 90% accurate, the SEER-Medicare database will be used to evaluate differences in breast cancer tumor attributes including size and stage, and differences in breast cancer treatment rates including mastectomy, lumpectomy and radiation between women who undergo routing screening mammography and women who do not.