Two Day Course
WinBUGS in Health Economic Evaluations
MRC Biostatistics Unit, Cambridge, UK
Wednesday 23, Thursday 24 September 2009
http://www.mrc-bsu.cam.ac.uk/bayescost/
* This course is intended to provide an introduction to Bayesian
analysis and MCMC methods using WinBUGS, as applied to
cost-effectiveness analysis and typical models used in health
economic evaluations.
* The emphasis throughout will be on practical examples: software and
code to carry out all the analyses will be provided.
* We shall assume a basic knowledge of standard methods in health
economics, and familiarity with a range of probability
distributions, regression analysis, Markov models and random-effects
meta-analysis.
* No knowledge of WinBUGS will be assumed.
Day 1: Wednesday 23 September, 10.00 to 17.15
1. The Bayesian paradigm. Prior distributions. Predictions.
Introduction to distributions to be used: beta, normal, uniform,
gamma, log-normal, Poisson. Forward sampling. Introduction to
Monte Carlo (MC) in WinBUGS.
2. Probabilistic sensitivity analysis in WinBUGS: Probabilistic
sensitivity analysis (PSA). Cost-effectiveness
plane. Incremental net benefit. Cost-effectiveness acceptability
curves. PSA in WinBUGS. Simple Markov models.
3. Introduction to MCMC in WinBUGS: Bayes theorem for learning from
data. Introduction to Markov chain Monte Carlo (MCMC) and WinBUGS.
Inference on any functions of parameters.
4. Parametric models for estimating expected cost. Advantages over
transforming data / non-parametric methods. Model fit. Predictions.
Inference on mean. Sensitivity to tail-area assumptions.
Day 2: Thursday 24 September, 9.30 to 17.00
5. Cost-utility data: Relating costs to effects using
regression. Bivariate posterior distributions. Baseline
adjustment. Subgroups.
6. Markov models: Estimating transition probabilities. Setting up a
Markov model in WinBUGS. Use in cost-effectiveness analysis. One
step model fitting and PSA.
7. Evidence synthesis in WinBUGS. Meta-analysis. Random-effects.
Indirect comparisons. Multiple sources of evidence. Evidence
synthesis applied to Markov models. Quality weights. Model fit:
DIC. Model uncertainty and sensitivity to assumptions.
8. Advanced topic: Running WinBUGS with R:
OpenBUGS. R2WinBUGS. BRugs. CODA. Contour plots.
Course instructors
* Dr Chris Jackson (MRC Biostatistics Unit)
* Dr Richard Nixon (Novartis, Basel)
Costs: £150 (student), £260 (public sector), £500 (commercial)
For registration and further information, please visit
http://www.mrc-bsu.cam.ac.uk/bayescost/
--
Christopher Jackson <[log in to unmask]>
Research Statistician, MRC Biostatistics Unit, Institute of Public
Health, Robinson Way, Cambridge, UK, CB2 0SR. +44 (1223) 330381
|