Spring 2010 Meeting

Held on March 11, 2010 at the Crowne Plaza Hotel in Northbrook, IL. 

The Program consisted of three presentations:

  1. Gatekeeping Procedures Ajit Tamhane, Ph.D., Senior Associate Dean & Professor, Northwestern University
  2. Power Analysis for Trials with Discrete Time Survival Endpoints Mirjam Moerbeek, Ph.D., Professor, Utrecht University & Katarzyna Jozwiak, Ph.D. Candidate, Utrecht University, The Netherlands
  3. Application of the Delta Method and Poisson Process on Relative Risk Chihche Lin, Ph.D., Astellas Pharmacueticals


Gatkeeping Procedures by Ajit Tamhane, Ph.D., Northwestern University

Biographical Background Ajit Tamhane is Senior Associate Dean of McCormick School of Engineering at Northwestern University and Professor of Industrial Engineering & Management Sciences (IEMS) with a courtesy appointment in the Department of Statistics. He was Chairperson of the Department of IEMS from 2001 to 2008. He received Ph.D. in Operations Research and Statistics from Cornell University and B.Tech. in Mechanical Engineering from I.I.T. Bombay. He has been with Northwestern University since 1975.  Professor Tamhane has authored three books: Multiple Comparison Procedures, with Yosef Hochberg (Wiley, 1987), Statistics and Data Analysis with Dorothy Dunlop (Prentice Hall, 2000), and Statistical Analysis of Designed Experiments (Wiley, 2009). He has edited two volumes of collected papers and chapters: Design of Experiments: Ranking and Selection with Tom Santner (Marcel-Dekker, 1984) and Multiple Testing Problems in Pharmaceutical Statistics (Taylor & Francis, 2009) with Alex Dmitrienko and Frank Bretz. He has published over 85 papers in the areas of multiple comparisons, multiple testing problems in clinical trials, design of experiments, chemometrics, quality control and clustering. His research is funded by NIH and by NSA. He is a fellow of ASA.

Abstract Gatekeeping procedures address the problems of testing hierarchically ordered and logically related null hypotheses that arise in clinical trials involving multiple endpoints, multiple doses, noninferiority-superiority tests, subgroup analyses, etc. while controlling the familywise type I error rate. Because of the practical importance of these problems, gatekeeping procedures have become an active area of research in the last decade. This talk will trace the developments in this field concluding with a summary of two current research topics.


Power Analysis for Trials with Discrete Time Survival Endpoints by Mirjam Moerbeek, Ph.D., and Katarzyna Jozwiak, Utrecht University, the Netherlands

Biographical Background Dr. Mirjam Moerbeek is associate professor at the department of Methods and Statistics, Utrecht University, the Netherlands. She specializes in statistical power analysis and optimal design, in particular for trials with hierarchical or longitudinal data, such as cluster randomized trials. She has received various research grants from the Netherlands Organization for Scientific Research and is supervisor of a project on "Improving statistical power analysis for trials on event occurrence by using an optimal design". She currently supervises three Ph.D. students. 

Katarzyna (Kasia) Jozwiak is a Ph.D. student at the department of Methods and Statistics, Utrecht University, the Netherlands. She studies optimal designs for trials with survival endpoints that are measured in discrete time. Relevant design issues are the optimal number of subjects, the optimal number of measurements and the optimal duration of the trial.

Abstract Studies on event occurrence aim to investigate if and when subjects experience a particular event. The timing of events may be measured continuously, using thin precise units, or discretely, using time periods. The latter metric of time is often used in social science research and the generalized linear model is an appropriate model for data analysis. While the design of trials with continuous time survival endpoints has been extensively studied, hardly any guidelines are available for trials with discrete time survival endpoints. This presentation will explore the relationship between sample size and power to detect a treatment effect in a trial with two treatment conditions. The exponential and Weibull survival functions will be used to represent constant and varying hazard rates, along with logit and complementary log-log link functions. It will be seen that for constant hazard rates the power depends on the event proportions at the end of the trial in both treatment arms and on the number of time periods. For varying hazard rates the power also depends on the shape of the survival functions and different power levels are observed in each time period in the case where the logit link function is used.


Application of the Delta Method and Poisson Process on Relative Risk by Chihche Lin, Ph.D., Astellas Pharmacueticals

Biographical Background Dr. Chihche Lin is a senior statistician at Astellas Pharma Global Development. Chihche is currently working in the Oncology group for late phase studies. Before joining Astellas in 2008, Chihche received statistics education and accumulated consulting experience in the School of Statistics, University of Minnesota, where he was also awarded his Ph.D. degree in 2007. His academia research interests include machine learning, model selection/combining and empirical processes. Along with his recent exposure to clinical trial studies, Chihche's research interests turn to Bayesian Analysis, Survival Analysis and Adaptive Design.

Abstract The concept of relative risk has been a prevailing way to describe the comparison of event rates, especially when the reference event rate is small. Although many statistical strategies have been approached to realize relative risk, there are pros and cons, as well as connections and disconnections among those approaches in practical studies. This presentation will review several methods for relative risk analysis and discuss the following topics: •Confidence intervals of relative risk by using the delta method •Application of the delta method: An example on constrained confidence interval for survival rate •Relative risk adjusted to risk factors. •Robust analysis of relative risk •Study design for relative risk analysis with the stopping criterion on the number of events. •An application of Poisson processes on the estimation of population size.