From Policy to Practice: Navigating Statistical Challenges of using RWE to Support Drug Approval

Speaker: Kun Wang, PhD, FDA

Dr. Kun Wang, with over a decade of experience in clinical trials and observational studies, currently serves as a Senior Reviewer in the Hematological Malignancy field. Her expertise lies in leveraging Real-World Data (RWD) and Real-World Evidence (RWE) for drug development. Prior to her role at the FDA, Dr. Wang led the Biometrics team at a Roche Company, overseeing the liquid biopsy development through the integration of clinical trials and RWE databased (e.g. CGDB genomic-clinical database, Flatiron database). At Johnson & Johnson, her leadership in RWE research span five therapeutic areas. At Yale University's Center for Outcome Research and Evaluation, she was a Senior Statistician, involved in several CMS outcome measure developments and research. Dr. Wang is a distinguished author, with her research in RWE field featured in top medical journals like NEJM, BMJ, and JNCI. She holds a PhD in Biostatistics from New York University.

Unlocking Social Determinants of Health Factors Through Geocoding of EHR Data: Measuring Structural Racism and Its Influence on Patient Outcomes from Bench to Bedside

Speaker: Xiaoliang (Wendy) Wang, PhD, Flatiron Health

Dr. Xiaoliang (Wendy) Wang is a Senior Quantitative Scientist at Flatiron Health, Inc. Dr. Wang received her PhD in Epidemiology, with a focus on cancer epidemiology and statistical genetics, from the University of Washington, and an MPH in Epidemiology from Columbia University. Her research ranges from gene-environmental interactions, Mendelian randomisation to enhanced survival extrapolation methods and various areas in health outcomes, quality of care and health equity research in oncology. Her work has been published in leading academic journals including Nature Genetics, the Journal of Clinical Oncology, JAMA Oncology, Cancer Epidemiology, Prevention & Biomarkers, and Blood Cancer Journal.

Abstract

Structural racism (SR), a socio-structural determinant of health (SDOH), are macro-level forces that perpetuate racial/ethnic inequities beyond individual-level factors. SR measures are associated with cancer outcomes, but often lack documentation in EHR. Most efforts to link EHR to area-level SDOH use less granular and unidimensional measures. Here, we link census tract and block group level data from the American Community Survey to patients’ residential addresses, and systematically evaluate area-level SR measures as predictors of accessibility to investigational drugs and overall survival using Cox proportional hazards models among patients with cancer. We also perform mediation analysis to elucidate which measures explain racial/ethnic inequities in trial participation, using nonlinear multiple additive regression tree models. Our study demonstrates that incorporating statistically meaningful and comprehensive area-level measures is feasible, and most SR measures are statistically significantly associated with patient outcomes. We also show that SR measures explained a substantial proportion of racial/ethnic inequities in trial participation among patients with cancer.

Machine Learning-Based Trial Emulation to Dissect Real-World Generalizability of Oncology Clinical Trials

Speaker: Ravi Parikh, MD, University of Pennsylvania Perelman School of Medicine

Ravi B. Parikh, M.D., M.P.P., FACP, is an Assistant Professor of Medicine and Health Policy and a genitourinary and thoracic medical oncologist at the University of Pennsylvania and Corporal Michael J. Crescenz VA Medical Center. He is an Associate Director and Director of the Program in Augmented and Artificial Intelligence at the Penn Center for Cancer Care Innovation (PC3I) at the Abramson Cancer Center. In addition, he serves as Director of the Human-Algorithm Collaboration Lab, a multi-disciplinary laboratory focusing on developing and testing AI-driven clinical decision support interventions in cancer care.

Strategies for Utilizing External Controls in Randomized Trials while Mitigating Bias in Treatment Effect Estimation

Speaker: Shu Yang, PhD, North Carolina State University

Dr. Shu Yang is Associate Professor of Statistics and a University Faculty Scholar at North Carolina State University. She received her Ph.D. in Applied Mathematics and Statistics from Iowa State University and postdoctoral training at Harvard T.H. Chan School of Public Health. Her primary research interest is causal inference and data integration, particularly with applications to comparative effectiveness research in health studies. She also works extensively on methods for missing data and spatial statistics. She has been a Principal Investigator for several U.S. NSF, NIH, and FDA research projects.