“Addressing Biases in Real World Data: The Role of "Meta"-Datasets”
Advances in data storage, transfer, and analysis have increased the use of retrospective real-world data to answer questions that could previously be answered only in prospective clinical trials. As researchers wrangle with data quality issues while designing and executing studies with real world data, new approaches for identifying and handling bias in retrospective datasets need to be developed. One such approach is the use of external datasets that contain medication approvals, treatment guidelines, and insurance formulary decisions to better understand which patients could have received the intervention of interest. This presentation will expand on the importance of identifying these patients, explore current datasets that could be used to characterize eligible patients, and discuss an approach to integrate these data into patient-focused real-world datasets and analyses.
Zach Rivers, PharmD, PhD, is a Senior Manager of Health Economics and Outcomes Research at Tempus Labs, Inc., a molecular diagnostics and artificial intelligence company. His research focuses on harmonizing patient journeys across multimodal real world datasets and leveraging this information to explore the impact of testing and treatment interventions in retrospective studies and economic simulations. Prior to joining Tempus, Zach received his PharmD and PhD at the University of Minnesota College of Pharmacy, where he practiced as a clinical oncology pharmacist, and completed a postdoctoral fellowship at the Hutchinson Institute for Cancer Outcomes Research at Fred Hutch.
