"Modeling & Simulation to Improve the Design of Clinical Trials"
Gary Rosner, ScD Professor and Director Division of Biostatistics and Bioinformatics Johns Hopkins University
Abstract: Clinical studies in cancer are becoming more complex, especially in early phase studies. The complexity arises because of many factors, such as the desire to be more efficient than in the past, the need to address endpoints other than tumor shrinkage with newer targeted agents, an interest in applying decision theoretic considerations to optimize the design, and the current interest in determining the best drugs for the specific molecular characteristics of each tumor. Many studies that incorporate Bayesian ideas in their design often require modeling and simulation to help determine the best design or to evaluate each study's operating characteristics. In this talk, I will discuss several examples of designs that benefitted from modeling and simulation.
"A novel kernel-based statistical approach to testing association in longitudinal genetic studies with an application of alcohol use disorder in a veteran cohort"
Zuoheng Wang, PhD Assistant Professor Department of Biostatistics Yale University
Abstract: Alcohol use disorder (AUD) is a major public health concern in the United States and contributes to the pathogenesis of many diseases. The risk for AUD is multifactorial including both genetic and environmental factors. Multiple measurements in longitudinal genetic studies provide a route to reduce noise and correspondingly increase the strength of signals in genome-wide association studies (GWAS). In this study, we developed a powerful kernel-based statistical method for testing the joint effect of gene variants with a gene region on disease outcomes measured over multiple time points. We applied the new method to a longitudinal study of veteran cohort with both HIV-infected and HIV-uninfected patients to understand the genetic risk underlying AUD. We found an interesting gene that may involve the interaction of HIV replication, suggestive of potential gene by environment effect in alcohol use and HIV. We also conducted simulation studies to access the performance of the new statistical methods and demonstrated a power gain by taking advantage of repeated measurements and aggregating information across a biological region. This study not only contributes to the statistical toolbox in the current GWAS, but also potentially advances our understanding of the etiology of AUD.
"Causal modeling under complex dependency in clustered and longitudinal observations"
Jiwei He, MS PhD Candidate Division of Biostatistics Department of Biostatistics and Epidemiology
Abstract: In assessing the efficacy of a time-varying treatment MarginalStructural Nested Mean Models (SNMMs) are useful in dealing with confounding by variables affected by earlier treatments. MSMs model the joint effect of treatments on the marginal mean of the potential outcome, whereas SNMMs model the joint effect of treatments on the mean of the potential outcome conditional on the treatment and covariate history. These models often consider independent subjects with noninformative time of observation.
We extend the two classes of models to clustered observations with time-varying treatments in the presence of time-varying confounding. We formulate models with both cluster- and unit-level treatments and derive semiparametric estimators of parameters in such models. For unit-level treatments, we consider both the presence and absence of interference, namely the effect of treatment on outcomes in other units of the same cluster. For MSMs, we show that the use of unit-specific inverse probability weights and certain working correlation structures can improve the efficiency of estimators under specified conditions. The properties of the estimators are evaluated through simulations and compared with the conventional GEE regression method for clustered outcomes. To illustrate our methods, we use data from the treatment arm of a glaucoma clinical trial to compare the effectiveness of two commonly used ocular hypertension medications.
We also extend SNMMs to situations with intermittent missing observations. In observational longitudinal studies, subjects often miss prescheduled visits intermittently. Previous literature has mainly focused on dealing with monotone censoring due to early dropout. Here we focus on intermittent missingness that can depend on the subjects' covariate and treatment history. We show that under certain assumptions the standard SNMMs can be used for situations where non-outcome covariates are missing intermittently. In situations where outcomes are also missing intermittently, we use a method that does not require artificially censoring the data, but requires a strict missing at random assumption. The estimators are shown to be consistent and achieve reasonable efficiency. We illustrate the method by estimating the effect of non-steroidal anti-inflammatory drugs (NSAIDs) on genitourinary pain using data from a study of chronic pelvic pain.
Dissertation Advisor: Marshall Joffe, MD, MPH, PhD, Alisa Stephens, PhD Committee Chair: Russell Taki Shinohara, PhD Committee Members: John Kempen, MD, Linda Zhao, PhD
“Temporal Dependencies with Time-Varying Exposures: Modelling Exposure-Lag-Response Associations”
Antonio Gasparrini, PhD
Imperial College London, UK
“Enhancing Comparative Effectiveness Research through Data Linkages: Examples from Rheumatology”
Jeffrey Curtis, MD, MS, MPH
University of Alabama at Birmingham, AL
“Bayesian Analysis of Multi-Type Recurrent Events and Dependent Termination With Nonparametric Covariate Functions”
Sheng Luo, PhD
School of Public Health, Johns Hopkins University
"Robust and Powerful Sibpair Test for Rare Variant Association"
Sebastian Zoellner, PhD Associate Professor Department of Biostatistics University of Michigan School of Public Health
"Social Determinants of Cancer Disparities: From Screening to Survivorship”
Lorraine Tiera Dean, ScD Instructor,
University of Pennsylvania School of Medicine Department of Biostatistics and Epidemiology