Colloquium: Comparative Effectiveness Research in Precision Medicine (PCER)
Abstract: In this talk, we will begin with defining PCER, which is a special case of comparative effectiveness research (CER). CER is a type of clinical studies that aim to compare evidence to identify effective intervention/treatment option to help clinical decision, and one usually compares evidence in a population- or subpopulation-level. Instead, in a PCER study, one compares evidence in an individual-level using prediction, optimization, and other data-driven approaches. Three studies will be discussed to demonstrate the PCER examples. They are: (1) Personalize selection of the warfarin treatment protocol to minimize two-sided risks (internal bleeding and stroke) based on one’s clinical and genetic characteristics (Aurora Health Care clinical trial simulation studies). (2) Predict personalized statin prescription to prevent or minimize the risks of therapy discontinuation and statin-associated symptoms (OptumLabs studies). (3) Predict an individual patient’s cognitive status and changes in 8 years to help early detect dementia risk in long-term future (Nation Alzheimer’s Coordinating Center and Alzheimer’s Disease Neuroimaging Initiative studies). All PCER examples require interdisciplinary team efforts and data-driven evidence discovered in big data.
Bio: Dr. Chi studied Health Informatics, Computer Science, and Operations Research. He is particularly interested in understanding the influence of multiple individual characteristics on the outcome of multiple interventions to best guide how an individual clinician, care team, and hospital network can select and implement intervention decisions to maximize overall healthcare outcomes. His research agenda starts from knowledge discovery in diverse type of healthcare data (data used in the previous projects including electronic health records, claims data, clinical trial data, Omaha system data, long-term patient cohort, omics data, virtual patient data, and other types of data that fit the research purpose) to clinical implementation. He has developed several integrative approaches (including clinical trial simulation, machine learning, optimization, artificial intelligence, and clustering analysis) to extract knowledge and evidence from data to support personalized healthcare decisions.
In 2013, Dr. Chih-Lin Chi joined the University of Minnesota as core faculty in the Institute for Health Informatics and Assistant Professor in the School of Nursing. He received his BS and MBA in Taiwan, and PhD in Health Informatics in the University of Iowa. He completed his postdoctoral training at the Center for Biomedical Informatics, Harvard Medical School in 2013 focusing on translational research.