缅北强奸

Title:聽Power Calculation for Detecting Interaction Effect in Cross-Sectional Stepped-Wedge Cluster-Randomized Trials.

Abstract:Deukwoo Kwon is an Associate Professor at the Institute for Healthcare Delivery Science in the Department of Population Health Science and Policy and a biostatistician at the Biostatistics Shared Resource Facility, Tisch Cancer Institute (TCI) at Icahn School of Medicine at Mount Sinai. He earned his Master and Ph.D. degrees in statistics from Texas A&M University and worked at the National Cancer Institute (NCI) for six years. At NCI, Dr. Kwon worked on various epidemiologic studies including radiation exposure assessment, uncertainty analysis, and measurement error models in dose-response relationship. Before joining Icahn School of Medicine at Mount Sinai in February 2022, he worked at University of Miami over 10 years and gained extensive experience in developing optimal statistical design and conducting analysis for cancer clinical trials and observational studies. He has utilized survival analysis, longitudinal data analysis, cancer registry data analysis, Bayesian inference, and high-dimensional data analysis for his collaborative work. He is a member of Protocol Review and Monitoring Committee at TCI where he promotes use of emerging approaches to design and analysis of phase I and phase II cancer clinical trials.

Madhu Mazumdar is Director of the Institute for Healthcare Delivery Science at the Mount Sinai Health System and is a Professor of Biostatistics at the Center of Biostatistics, Department of Population Health Science and Policy. She also directs the Biostatistics Core of Tisch Cancer Institute. Website:

Stepped-Wedge Cluster-Randomized Trials (SW-CRTs) are increasingly utilized for evaluating complex healthcare delivery interventions where simple CRTs are not feasible. Appealing features of SW-CRTs include having each cluster acting as their own control, not needing to withhold the intervention from any patient, and having time to prepare clusters for administration of intervention while collecting baseline information. However, the design and analysis of SW-CRT is complex and methodology is not available for many scenarios including detection of interaction effects. Detecting interaction effect is important for a variety of research scenarios. We present four ways of computing power and showcase their comparative performance through simulation. We then apply the methodology to a published SW-CRT with binary outcome. Extension to continuous and censored outcomes are underway.