" The Randomized Placebo-Phase Design: Evaluation, Interim Monitoring and Analysis"
Stephanie Pugh, PhD Senior Statistician CCOP, Symptom Management American College of Radiology
Abstract: The randomized placebo-phase design, also known as the randomized delayed-start design, has been proposed as an approach to circumvent the reluctance of patients and physicians to participate in trials with a placebo control. Although there is some practical appeal to the design and it has been used in an increasing number of active and ongoing trials, there are often overlooked issues relative to statistical power, estimating sample size and determining plans for interim analysis that may limit its usefulness. We developed a general model for describing the pattern of treatment response and based on the specified parameters of this model, derive and compare the necessary methods for estimating sample size. We also develop and compare different strategies for interim monitoring. In addition to statistical power considerations, we also provide results from extensive simulations investigating the robustness of the proposed procedures since the efficiency of the randomized placebo-phase design is highly dependent on the assumptions made about the form of the alternative hypotheses.
"Assessing the Mediating Effect of a Biomarker: a Nonparametric, Matching-based Approach"
Julian Wolfson, Ph.D. Assistant Professor Division of Biostatistics University of Minnesota
Abstract: The problem of identifying biomarkers which may mediate treatment effects in clinical trials is made challenging by the fact that biomarker values are measured after randomization, and are therefore subject to selection bias due to unmeasured confounding. As a result, statistical methods from the sub-field of causal inference are usually applied to account for this possible confounding. Since the estimands these methods rely on are typically not statistically identifiable unless strong assumptions are made about both the study design and data generating process, sensitivity analyses are often recommended in the literature. However, many of the sensitivity analyses proposed are challenging to implement in practice, either because several sensitivity parameters must be varied or the meaning of the parameters themselves is difficult to understand. We propose a nonparametric, matching-based sensitivity analysis approach to assessing the mediating effect of a biomarker on a binary outcome. The method makes no parametric assumptions about the biomarker distribution or the biomarker-outcome relationship, and easily accommodates information from potentially high-dimensional baseline covariates.
"A Constrained Mixed Effects Model Based on Semilinear Differential Equation for Cell Polarity Signaling in Tip Growth of Pollen Tubes"
Xingping Cui Associate Professor Department of Statistics University of California, Riverside
Abstract:The key of tip growth in eukaryotes is the polarized distribution on plasma membrane of a particle named ROP1. This distribution is the result of a positive feedback loop, whose mechanism can be described by a Differential Equation parametrized by two meaningful parameters kpf and knf . In this paper, we introduce a mechanistic Integro-Differential Equation (IDE) derived from a spatiotemporal model of cell polarity and we show how this model can be fitted to real data i.e ROP1 intensities measured on pollen tubes. At first, we provide an existence and uniqueness result for the solution of our IDE model under certain conditions. Quite interestingly, this analysis gives a tractable expression for the likelihood, and our approach can be seen as the estimation of a constrained nonlinear model. Moreover, we introduce a population variability by introducing a constrained nonlinear mixed model. We then proposed a constrained Least Squares method to fit the model under single subject case, and two methods, constrained Methods of Moments and constrained Restricted Maximum Likelihood (REML) to fit the model under the multiple subjects case. The performances of all the three methods are studied in a simulation example and are used on a real multiple subjects dataset.
"From Humans to Monkeys and Back: Physical Activity Patterns in Humans and Primates"
Vadim Zipunnikov, Ph.D Assistant Professor Department of Statistics Johns Hopkins Univeristy
Abstract: I will illustrate key statistical challenges of analyzing physical activity data by giving snapshots of three recent projects. First, I will talk about analysis of data collected on 700+ subjects wearing an Actiheart device that collects minute-by-minute activity counts and heart rate for one week as a part of the Baltimore Longitudinal Study of Aging. Secondly, I will talk about an experiment examining changes in activity in a monkey model of Parkinson’s disease that can be used to guide studies of activity changes in human Parkinson’s disease, as well as other related disorders associated with dopamine depletion. Finally, I will talk about “Home” and “Away” models for physical activity and inactivity in elderly adults. These models use both second-by-second activity and GPS data to identify activity “hotspots”, outside locations where alleviated physical activity is recorded.
"Community-based Participatory Research to Reduce Disparities in Access to Kidney” Transplantation “
Rachel Patzer, PhD, MPH
Assistant Professor, Division of Transplantation, Department of Surgery, Assistant Professor, Department of Epidemiology,
Rollins School of Public Health, Emory University School of Medicine
"Estimation and Inference of the Three-Level Intraclass Correlation Coefficient"
Matthew D. Davis, MS PhD Candidate Division of Biostatistics Department of Biostatistics and Epidemiology
Advisors: Warren B. Bilker, PhD and J. Richard Landis, PhD Committee Chair: Sharon X. Xie, PhD Committee: Robert DeRubeis, PhD
“Semi-parametric regression models for longitudinal data with outcome-dependent observation times”
Kay See Tan, MS
PhD Candidate Division of Biostatistics Department of Biostatistics and Epidemiology
Advisors: Andrea B. Troxel, ScD and Benjamin C. French, PhD Committee Chair: Sarah J. Ratcliffe, PhD Committee: Dylan Small, PhD and Kevin G. Volpp, MD, PhD
Abstract:Conventional longitudinal data analysis methods typically assume that outcomes are independent of the data-collection schedule. However, the independence assumption may be violated when an event triggers outcome assessment in between prescheduled follow-up visits. For example, patients initiating warfarin therapy who experience poor anticoagulation control may have extra physician visits to monitor the impact of necessary dose changes. Observation times may therefore be associated with outcome values, which may introduce bias when estimating the effect of covariates on outcomes using standard longitudinal regression methods. We consider a joint model approach with two components: a semi-parametric regression model for longitudinal outcomes and a recurrent event model for observation times. The semi-parametric model includes a parametric specification for covariate effects, but allows the effect of time to be unspecified. We formulate a framework of outcome-observation dependence mechanisms to describe conditional independence between the outcome and observation-time processes given observed covariates or shared latent variables.
We generalize existing methods for continuous outcomes by accommodating any combination of mechanisms through the use of observation-level weights and/or patient-level latent variables. We develop new methods for binary outcomes, while retaining the flexibility of a semi-parametric approach. We extend these methods to account for discontinuous risk intervals in which patients enter and leave the at-risk set multiple times during the study. Our methods are based on counting process approaches, rather than relying on possibly intractable likelihood-based or pseudo-likelihood-based approaches, and provide marginal, population-level inference. In simulations, we evaluate the statistical properties of our proposed methods. Comparisons are made to ‘naïve’ approaches that do not account for outcome-dependent observation times. We illustrate the utility of our proposed methods using data from a randomized trial of interventions designed to improve adherence to warfarin therapy and a randomized trial of malaria vaccines among children in Mali.
"A flexible framework for differential isoform expression analysis in RNA-seq"
Yang Yang MS Candidate Division of Biostatistics Department of Biostatistics and Epidemiology