Abstract Using health care systems data for research and care improvement is of intense interest. The now widespread availability of electronic health records (EHRs) holds great promise for population sciences cancer research. Founded in 1999, the Cancer Research Network (CRN) is a consortium of a dozen research groups affiliated with US integrated health care systems. CRN health care systems are pioneers in use of data systems to document and inform administrative and clinical aspects of health care. In addition to administrative (e.g., claims) data, the clinical databases of CRN health systems capture details of care that have generally not been available until the advent of EHRs. In most CRN health systems, such databases have been in use since at least the mid-1990s, with adoption of EHRs in the mid-2000s. CRN researchers developed the Virtual Data Warehouse (VDW) common data model to facilitate collaborative, multi-institution research. The VDW, with standardized data tables, variable names, and formats implemented at each participating research group, enables efficient multi-institution research collaborations. From a cancer research perspective, the VDW is also one of the only EHR-based common data models with tumor registry data of comparable quality to that of the SEER Program. Infusion medication data are also available in the VDW, and inclusion of results of clinical genetic and molecular tests is in development. The VDW is updated regularly, monthly or more frequently in most CRN institutions for most data domains. Advantages of conducting cancer research in the CRN setting include the ability to identify a “denominator” population, and the capture of primary and specialty care. Retention of health plan enrollees is high in CRN health systems, enabling longitudinal research. In most US health care settings, the population eligible to receive care cannot be identified, and data may be available for some, but not all, episodes of care. For example, tertiary-care cancer centers may have details on cancer treatment, but not information from primary care or other specialty care encounters. However, it is important to use caution because EHR data result from clinical encounters and associated activity and are not collected for research purposes; missingness is unlikely to be random and confounding by indication is an analytic and interpretational issue. Collaborative, efficient use of these data also requires navigating regulatory concerns. The federated nature of the VDW minimizes HIPAA-related privacy concerns, while agreed-upon human subjects and data use procedures facilitate collaborations. The CRN provides unparalleled opportunities for research across the full range of the cancer experience, from primary prevention to end of life, incorporating data from EHRs and other clinical and administrative databases. Use of these data and adherence to epidemiologic principles in the design of studies and their analysis and interpretation can greatly advance knowledge to decrease the burden of cancer. Citation Format: Lawrence H. Kushi, Jeffrey S. Brown, Diana S.M. Buist, Jessica Chubak, V. Paul Doria-Rose, Kathleen M. Mazor. Opportunities and lessons learned from the Cancer Research Network [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 IA23.
Read full abstract