Short Courses

Registration for Short Courses now open!

Four Short Course proposals have been selected for presentation just before the International Biometric Conference begins. These Short Courses will take place on August 26, 2012. Courses are all day and are taught by experienced professionals who are experts in their fields.
The cost to participate is the same for all courses. Registration fee includes lunch, coffee/tea breaks, and any notes provided by the instructors.

By April 30th After May 1st
IBS Member $250 $350
IBS SCC Member $125 $200
IBS Non-member $350 $450
IBS SCC Non-member $200 $250
Student $125 $200
SCC Student $75 $125
All prices are in USD.

For more information about these short courses, please see below introductions. To register for a short course, please click here.

Please note that in order to register for the short courses you must register for the IBC first. Register for IBC2012 here.

Course 1

Group Sequential and Adaptive Methods for the Design of Clinical Trials

Tutors: Bruce W. Turnbull, Cornell University, Ithaca, USA
Chris Jennison, University of Bath, Bath, UK
Length of the course: Full day
Level/prerequisites:
 This course is aimed at Masters level statisticians who have some familiarity with clinical trials but not necessarily with the aspects of sequential monitoring or adaptive trial design.
Summary:
 Formal data monitoring procedures are now a standard feature of the design and conduct of long-term clinical trials. A unified formulation of group sequential procedures allows a simple, powerful approach to their implementation with different types of stopping rule and a great variety of endpoints. We shall survey the main ideas of group sequential procedures including: one-sided, two-sided, non-inferiority and equivalence testing; normal, binary, survival, regression, longitudinal and multiple endpoints; estimation on termination; nuisance parameters; multiple arm trials and multiple endpoints.
 More recently, methods have been proposed to allow modification of a trial in mid-course while still protecting the type I error. Possible modifications include enlarging the sample size to increase power, changing the study population (for example, enrichment), modifying the treatment, or reducing the number of treatment arms. These adaptations may follow rigid rules, pre-specified in the protocol; more flexible approaches permit unplanned changes at unplanned interim analyses. We shall describe these procedures in detail and discuss their benefits and limitations.
 Statistical software will be used throughout the course to illustrate the methods and examples.
Textbook:
 "Group Sequential Methods with Applications to Clinical Trials" by C. Jennison and B. W. Turnbull (2000), published by CRC Press (Taylor and Francis Group).
 A Japanese translation of this book will be available shortly.
Outline:
  1. Principles of group sequential methods; underpinning theory and computation; the general framework, including normal, binary and survival endpoints
  2. Boundaries: efficacy, futility (binding and non-binding); error spending designs; the "pipeline" problem when there is a delayed response
  3. Information monitoring and nuisance parameters; estimation and stochastic curtailment
  4. From group sequential to adaptive designs: sample size modification to improve power
  5. General methodology for adaptive designs: combination tests and tests for multiple hypotheses.
  6. Adaptive enrichment designs: changing the target population
  7. Multiple treatments: seamless Phase II/III transition
  8. Case studies: examples of sequential and adaptive trials will be presented; participants will also be invited to discuss their experiences in implementing adaptive designs, interacting with regulators or serving on Data and Safety Monitoring Boards


Course 2

Joint Modeling Approaches in Longitudinal Studies Using Random Effects

Tutors: Dimitris Rizopoulos, Erasmus University Medical Center, Rotterdam, The Netherlands
Geert Molenberghs, Interuniversity Institute for Biostatistics and Statistical
Bioinformatics, Katholieke Universiteit Leuven and Universiteit
Hasselt, Belgium
Geert Verbeke, Interuniversity Institute for Biostatistics and Statistical Bioinformatics,
Katholieke Universiteit Leuven and Universiteit Hasselt, Belgium
Length of the course: Full day
Level/prerequisites:
 This course is aimed at MSc-level statisticians who are interested in getting familiar with recent advances in joint modeling approaches applicable to longitudinal studies.
 Course attendees should consider as a prerequisite for the course familiarity with the subject at the level of: Linear Mixed Models for Longitudinal Data, Chapters 1-7 (Springer-Verlag) Verbeke and Molenberghs; The Statistical Analysis of Failure Time Data, 2nd Edition, Chapters 1-4 (Wiley) Kalbfleisch and Prentice.
Summary:
 In longitudinal studies, measurements are often collected for different types of outcomes for each subject. These may include several longitudinally measured responses (such as blood values relevant to the medical condition under study) and the time at which an event of particular interest occurs (e.g., death, development of a disease or dropout from the study). These outcomes are often separately analyzed; however, in many instances, a joint modeling approach is either required or may produce better insight into the mechanisms that underlie the phenomenon under study. The aim of this course is to first identify the type of research questions that require joint modeling, and then present state of the art statistical models that are designed to optimally use the data to answer those questions. Emphasis is placed on three settings: (1) longitudinal studies with nonrandom dropout; (2) time-to-event analysis with time-dependent covariates measured with error; (3) multivariate longitudinal data scenarios where the aim is to study the association structure. These joint modeling approaches are presented within a unified framework that is based on the use of random effects to explain the interdependencies between the observed outcomes.
Outline:
Part I: Motivation for Joint Modeling
  1. Description of the types of data and research questions that necessitate joint modeling. In particular, missing data, time-to-event data and multiple longitudinal outcomes
  2. Presentation of motivating cases studies: Datasets with the features that we will illustrate
  3. Contrast different approaches to analyze multiple outcomes simultaneously:
    • Conditional approach versus random effects models (including shared parameter models)
    • Motivation for the focus on random effects models
    • General joint modeling framework using random effects
  4. Missing data mechanisms
    • Problems with nonrandom dropout (i.e., bias, loss of efficiency, etc.)
    • Various modeling frameworks to handle dropout: Selection, pattern mixture and shared parameter models
  5. A brief review of different approaches to analyze longitudinal responses:
    • Linear and generalized linear mixed effects models
    • Generalized estimating equations & inverse probability weighting

Part II: Joint Models for the Longitudinal and Dropout Processes
  1. Shared-parameter Models
    • Presentation of the shared parameter models framework
    • Continuous random effects & latent classes
    • Estimation & examples
  2. Sensitivity Analysis
    • Why is it needed?
    • Different approaches to sensitivity analysis
    • Examples

Part III: Joint Models for Longitudinal and Time-to-Event Data
  1. Motivation
    • Focus on time-to-event outcomes: A brief review of relative risk models
    • Different types of time-dependent covariates
    • Including time-dependent covariates standard relative risk models
    • The joint modeling approach
  2. Joint Models
    • Presentation of the joint modeling framework
    • Model building strategies: Different options for the survival submodel, the longitudinal submodel and the random effects distribution
    • Estimation: Basics of ML estimation for these models (e.g., requirement for numerical integration)
    • Examples
  3. Connection with Missing Data
    • How these models are related to the shared parameter models presented in Part II, and what are the implications

Part IV: Joint Models for Multivariate Longitudinal Data
  1. Motivation
    • Study association structure between different longitudinal outcomes
  2. Multivariate Joint Models
    • Definition
    • Estimation: Full likelihood versus the pairwise approach & pseudo-likelihood
    • Examples

Part V: Connections & Extensions
  1. A case study bringing together many different aspects of the previous parts


Course 3

Clinical Trial Data Analysis Using R

Tutors: Ding-Geng (Din) Chen, University of Rochester Medical Center, USA
Michikazu Nakai,National Cerebral and Cardiovascular Center, Japan
Length of the course: Full day
Level/prerequisites:
  We expect the students to have some basics in R and biostatistics even we will be carefully guide them in the learning process. The course is at the intermediate level.
Recommended textbook:
 Ding-Geng (Din) Chen, and Karl Peace, "Clinical Trial Data Analysis Using R", Chapman and Hall 2011, ISBN: 9781439840207 (will be available for purchase on site)
Summary:
 This short course is based on the book: "Clinical Trial Data Analysis Using R", by Chen and Peace, which uses R to design and analyze clinical trials. This short course provides a thorough presentation of biostatistical analyses of clinical trial data with detailed step-by-step illustrations on their implementation using R. Examples of clinical trials based on the authors' actual experience in many areas of clinical drug development are presented. After understanding the application, various biostatistical methods appropriate for analyzing data from the trials are identified. Then analysis code is developed using appropriate R packages and functions to analyze the data. Analysis code development and results are presented in a stepwise fashion. This stepwise approach should enable students to follow the logic and gain an understanding of the analysis methods and the R implementation so that they may use R to analyze their own clinical trial data.
 Since its creation in the mid1990s, R has become widely used in statistical modeling and computing, and it is now an integrated and essential software for biostatistical analyses. Assuming no prior experience in R, this tutorial will start with a basic introduction to the R system, where to get R, how to install R and how to upgrade R packages. Then we proceed on using R for design of clinical trials, analyses clinical trial data from treatment comparisons, time-to-event endpoints, longitudinal trials, bioequivalence trials, etc. Additional topics are covered, such as meta-analysis, Bayesian analysis and microarray data analysis in clinical trials.
Outline:
  1. Treatment Comparisons (Review Chapters 1 to 4)
    • R fundamentals associated with clinical trials (Chapter 1)
    • A simple simulated clinical trial (Chapters 1 and 2)
    • Statistical models for treatment comparisons (Chapter 3)
    • Incorporating covariates (Chapter 4)
    • Step-by-step implementations in R on analyzing treatment significance in clinical trials
  2. Survival Analysis (Chapter 5)
    • Time-to-event data structure
    • Statistical models for survival data
    • Right-censored data analysis
    • Interval-censored data analysis
    • Step-by-step implementations in R to analyze survival data from clinical trials
  3. Data from Longitudinal Clinical Trials (Chapter 6)
    • Trial designs and data structure
    • Statistical models and analysis
    • Step-by-step implementations in R in analyzing longitudinal clinical trials
  4. Sample Size Determination and Power Calculations (Chapter 7)
    • Review of sample size determination
    • Sample size determination in treatment comparison, time-to-event data, longitudinal trials with R
  5. Meta-analysis of clinical trials (chapter 8)
    • Data structures
    • Fixed-effects and random-effects models
    • Step-by-step implementations in R on meta-analysis with clinical trials data
  6. Bayesian methods in Clinical Trials (Chapter 9)
    • Review statistical models
    • Review R packages including WinBugs, R2WinBUGS, BRugs, rbugs and MCMCpack
    • Data analysis with R
  7. Analysis of DNA Microarrays in Clinical Trials (Chapter 12)
    • Statistical models in R/Bioconductor for microarrays
    • Breast cancer data analysis in R


Course 4

Identifying Genes for Complex and Mendelian Traits Using Next Generation Sequence Data

Tutors:

Suzanne Leal, Baylor College of Medicine, USA
Length of the course: Full day
Outline:
  1. Introduction to Genetic Concepts
  2. Quality control of Next Generation Sequence Data
  3. Gene discovery using Family Data
    • Mendelian Traits
    • Complex Traits
  4. Gene Discovery for Complex traits
    • Case-Control Data
    • Quantitative Traits
    • Extreme Sampling
  5. Estimating Sample Size and Power for Rare Variant Association Tests
    • Estimation of Effect Sizes
    • Replication of Gene Discoveries
    • Analysis of Exome Chip data