All courses are limited to 50 participants
Course1 (Full day)
Bayesian Statistics for Clinical Trials and Drug Regulation
| This workshop will give an introduction to Bayesian methods and their applications in medical research, and in particular in clinical trials. Bayesian methods are increasingly being used across many stages of drug development, and this workshop is aimed to provide a grounding in the key aspects of these methods – it will also provide a refresher to those who may have studied Bayesian methods in the past and now need to revisit them. | | | Topics covered will include Bayes theorem, likelihood and the key principles behind Bayesian analysis, the choice of prior distributions and inference from Bayesian methods. Examples will be used throughout, showing how Bayesian methods have already been used in practice, and where they may be used increasingly in the future. Although the primary focus will be on clinical trials, the utility of Bayesian approaches in integrating other sources of information for more complex regulatory decision-making will also be illustrated.
| | | Course leader: Professor of Medical Statistics and Clinical Trials
Department of Epidemiology and Public Health,
Imperial College London,
Faculty of Medicine, St Mary's Campus,
Norfolk Place, Paddington,
London W2 1PG UK |
Course 2 (Half day)
Inference and Prediction using Longitudinal Data:
analysis of time-dependent data for inference and prediction.
Longitudinal studies allow investigators to correlate changes in time-dependent exposures or biomarkers with subsequent health outcomes. However, there are two key statistical challenges associated with the use of time-dependent predictive information.
First, inference regarding the impact of time-dependent covariates on subsequent repeated measures outcomes requires consideration of the factors that drive the change in covariates. Statistical methods appropriate for analysis of time-dependent covariates has expanded to include causal inference methods derived from from both biometry and from econometrics. Second, the use of time-dependent markers to predict a subsequent change in clinical status such as transition to a diseased state require the formulation of appropriate prediction error concepts.
The first part of this course will review longitudinal data analysis methods recently developed for valid inference regarding time-dependent covariates. Methods include likelihood-based approaches as well as specific structural approaches such as G-estimation, G-computation, marginal structural models, and instrumental variable methods. We will ilustrate and compare the analysis options using an example from surgery where there is time-dependent non-compliance, and an example from the treatment of anemia among chronic kidney disease subjects.
The second part of this course will introduce predictive accuracy concepts that allow evaluation of time-dependent sensitivity and specificity for prognosis of a subsequent event time. We will overview options that are appropriate for both baseline markers and for longitudinal markers. Methods will be illustrated using examples from HIV and cancer research and will highlight R packages that are currently available. | | | Course leader:
Professor
Department of Biostatistics
University of Washington
Seattle WA 98195-7232 |
Course 3 (Full day)
Analysis of high-throughput SNP association studies
| Understanding of the genetic epidemiology of complex diseases has advanced considerably since high-throughput single nucleotide polymorphism (SNP) genotyping arrays became available. As costs fall, these arrays are being used increasingly for a wide number of discrete and quantitative traits (including gene expression traits, or eQTLs). This short course will give an overview of statistical problems in the design and analysis of such studies and will include practical sessions using the snpMatrix R package (available from the Bioconductor project). | | | Course leader:
Diabetes and Inflammation laboratory
Cambridge Institute for Medical Research
Wellcome Trust/MRC Building
Addenbrooke's Hospital
Hills Road
Cambridge |
Course 4 (Full day) Causal inference
Recent developments in causal inference within the statistical and artificial intelligence literature have led to important new insights on how to address problems of confounding and selection bias in a wide variety of settings. The aim of this course is to review these developments and to provide state-of-the-art statistical solutions for dealing with these problems. | | |  The first half day of the course will focus on probabilistic graphical models to express causal background knowledge including: (i) ways for reading off such graphs whether a given data situation suffers problems of confounding and selection bias (ii) and whether/how this can be accommodated, for instance via advanced identification results (G-computation). | | | | The second half day of the course will focus on statistical techniques to adjust or measured confounding. We will discuss limitations of ordinary regression adjustment, especially with view to estimation of direct effect and adjusting for time varying covariates. Successful alternatives, such as inverse probability weighting estimators in marginal structural models and G-estimators will be explained. As motivating examples we will use cross sectional and longitudinal studies, as relevant to epidemiological and medical research, throughout. | | | Course leaders: Professor
Dept. Applied Mathematics and Computer Science
Ghent University
Belgium
Department of Mathematics
University of Bristol
University Walk
Bristol BS8 1TW (UK) | | |
|
|
Last Updated on Wednesday, 07 July 2010 10:21 |
Short Courses
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 