Tutorial name:Tutorial #2: Missing Data in Clinical Trials
Day & Time:April 28, 1:30PM - 5:00PM
Instructor(s):Roderick J. Little, PhD
Dept of Biostatistics
Univ of Michigan, United States
Description:This short course will discuss methods for the statistical analysis of data sets with missing values, emphasizing applications to clinical trials. Topics include: Definition of missing data; assumptions about mechanisms, including missing at random; pros and cons of simple methods such as complete-case analysis and imputation; weighting methods; maximum likelihood and Bayesian inference with missing data; multiple imputation; computational techniques, included EM algorithm and extensions, and Gibbs sampler; software for handling missing data; missing data in common statistical applications, including regression, repeated-measures analysis, clinical trials. Selection and pattern-mixture models for nonrandom nonresponse. Sensitivity analysis for deviations from missing at random.
Prerequisites: Course requires knowledge of standard statistical models such as the multivariate normal, multiple linear regression, contingency tables, as well as matrix algebra, calculus, and basic maximum likelihood for common distributions, at the level of Statistical Inference, 2nd Ed. by G. Cassella and R. L. Berger
Recommended Texts:
Little, R.J. and Rubin, D.B. (2002), Statistical Analysis with Missing Data, 2nd edition, Wiley.
National Research Council (2010). The Prevention and Treatment of Missing Data in Clinical Trials. National Academy Press: Washington DC. A pdf copy can be obtained without charge from the list of publications at http://www7.nationalacademies.org/cnstat/
Learning Objective(s):At the conclusion of this tutorial, particpants should be able to:
- Discuss methods for the statistical analysis of data sets that have missing values
- Discussion application of missing data methods in the context of missing data in clinical trials
Target Audience:Statisticians, clinicians, and others working in clinical trials with an interest in how to address issues of missing data.