The challenge of personalized medicine is knowing when and how to apply screening and diagnostic tests and targeted treatments so that they result improved outcomes and cost-effective investment of resources at a population level. To date, genomic tests have been introduced into practice with scant empirical knowledge about their actual impact on health and economic outcomes, or use in diverse patient subgroups by varying physicians in multiple health care systems. No study has estimated the cost-effectiveness of genomic tests as they are used in the community, or used primary data in a robust US population model to compare real-life cost-effectiveness to analyses using trial data or hypothetical cohorts. Our project will use an innovative multi-criteria decision analysis framework to evaluate the impact of physician, patient, and system factors on the prioritization of competing strategies for the delivery of personalized cancer care in the community. We concentrate on cancer because of the large and growing number of genomic applications for this disease, the potential for major public health impact, and our ability to leverage specialized resources based on NCI's investment in several large research programs we lead. Our approaches are designed to be generalizable to personalized health interventions for other chronic diseases. We will (1) Identify critical factors that influence the cost-effectiveness of genomic testing based on guideline- recommended care vs. usual care without genomic testing, using our validated CISNET multi-cohort population model to evaluate an exemplar gene expression profile test in breast cancer; (2) Describe actual community care, and measure criteria important for decisions about use of genomic tests, using existing electronic records linked with registry data from diverse structures of care in two defined geographic regions, and surveys of a nested sample of 800 patients and their physicians; and (3) Test whether population cost-effectiveness based on actual use and outcomes of genomic testing in community practice is different from that predicted assuming all receive guideline care, and evaluate how the use of multi-criteria decision analysis compared with traditional cost-effectiveness analysis affects the prioritization of strategies. This research will provide a generalizable framework and tools for obtaining evidence on factors affecting the cost-effective use of genomic testing in real world community practice. Our results will add an important foundation in data and methods for future economic analyses that inform that the translation of NIH-supported research into practice and address priorities to use personalized medicine to better prevent, screen, diagnose, and treat cancer.
The challenge of personalized health care is knowing when and how to apply screening and diagnostic tests and targeted treatments in diverse groups of patients in real-life settings. This project will identify ways of enhancing personalized health cae by evaluating the impact of physician, patient, and healthcare system factors using a genomic test for breast cancer prognosis as the exemplar and applying an innovative multi- criteria decision analysis approach that incorporates physician and patient preferences for important psychological and qualitative factors, as well as health outcomes and costs. Our work will develop a generalizable framework and tools useful for other emerging technologies and for other chronic diseases.
