• Presentation Slides (PDF)

In research examining the effect of an intervention or exposure, a key secondary objective often involves evaluating differential effects of this intervention or exposure in subgroups of interest, often referred to as assessing effect modification or heterogeneity of treatment effects (HTE). Observed HTE can have important implications for policy, including intervention strategies (e.g., whether some patients will benefit more from intervention than others) and prioritizing resources (e.g., to reduce observed health disparities). Analysis of HTE is well understood in studies where the independent unit is an individual. 

This presentation discusses two issues when the independent unit is a cluster (e.g., a hospital, a school). First, in cases with a cluster-level outcome but individual-level characteristic (e.g., self-reported race), aggregating this individual-level variable to the cluster level results in reduced ability to detect HTE. Second, when mixed-effects models are used for analysis, a poor choice of correlation structure can result in severely inflated type I error rates.

Dr. Noorie Hyun is an Associate Biostatistics Investigator in the Biostatistics Division at Kaiser Permanente Washington Health Research Institute (KPWHRI). She earned her Ph.D. in biostatistics from The University of North Carolina at Chapel Hill and completed postdoctoral training in the Division of Cancer Epidemiology and Genetics at the National Cancer Institute. Prior to joining KPWHRI, she was an Assistant Professor in the Division of Biostatistics at the Medical College of Wisconsin. Dr. Hyun’s research has focused on methodological development in survival analysis, nonlinear modeling, complex probability sampling, and measurement error, with applications to cancer screening and risk prediction using electronic health records. Since joining KPWHRI, she has expanded her work to include pragmatic clinical trials and research on substance use disorders.

Dr. Brian Williamson is an Assistant Biostatistics Investigator at KPWHRI. He conducts statistical methods research on using machine learning to make more accurate and efficient use of data. He also collaborates on drug and vaccine safety and effectiveness research; pragmatic clinical trials; and mental health research.