Programme

Tuesday 11 June 2019

10:00 - 10:30 h Preliminary - Introduction to (Bayesian) modelling in HTA
  • Standard (two-stage) analysis of HTA data vs Bayesian (one-stage, full propagation of uncertainty) modelling
  • The role of Probabilistic Sensitivity Analysis
10:30 - 11:00 h Introduction to missing data
  • What's difficult about it?
  • What's so special about it in HTA?
  • Mechanisms for missingness (Rubin's categorisation)
  • Graphical representation of mechanisms
11:00 - 12:00 h Missing data in HTA
  • State of the art (literature review)
  • Correlation between outcomes
12:00 - 13:30 h Lunch
13:30 - 14:30 h Examples
  • Problems with HTA data (asymmetry, non-normality, correlation and “structural” values)
  • Example 1. MenSS trial (small pilot study)
  • Example 2. PBS trial (larger study, with longitudinal aspects)
14:30 - 16:30 h Practicals
  • R and JAGS
  • Basic models
  • Potential extensions

*The working language will be English.