Abstract IA13: Use of administrative and electronic health records to identify cancer recurrences

Abstract

Abstract There are approximately 16 million cancer survivors in the United States, and outcomes of major concern for these survivors are their risks of developing a cancer recurrence, a second primary cancer, and dying of cancer. While cancer registries routinely collect data on second cancers and mortality, they do not ascertain data on recurrences. Gold-standard data on recurrences currently relies on review of medical records, which is costly, time consuming, and does not include events that occur after the review takes place. Extracting data from administrative and electronic health records provides an opportunity to ascertain information on recurrences in real time. We are using two approaches to address this issue that leverage gold-standard data on recurrences we have ascertained through one of our large population-based studies of breast cancer. The first involves the development and implementation of an algorithm for breast cancer recurrences using medical claims data, and the second applies machine learning and artificial intelligence to text in electronic health records. Methods and progress related to these two approaches will be presented. Citation Format: Christopher I. Li. Use of administrative and electronic health records to identify cancer recurrences [abstract]. In: Proceedings of the AACR Special Conference on Modernizing Population Sciences in the Digital Age; 2019 Feb 19-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Epidemiol Biomarkers Prev 2020;29(9 Suppl):Abstract nr IA13.