Insights from actuarial science into HTA: Building joint models of random quantities - why correlation is not enough

Thursday 19 April 2018, 12.15PM to 1.15pm

Speaker: Professor Alexander McNeil, The York Management School, University of York

Background: In recent years in the UK there has been an interest in applying actuarial approaches to the modelling and mitigating of uncertainty in HTA. So far progress has been limited, which might result from actuaries and health economists not fully understanding and appreciating the perspectives and methods of each other. This seminar aims to build this understanding and appreciation. In addition to the main topic of this seminar, below, there will be discussion of other research areas of mutual interest.

Building joint models of random quantities - why correlation is not enough:

If X and Y are two random outcomes and we know something about their marginal distributions and have an estimate of their correlation/covariance, how do we say something about their joint distribution and, in particular, the joint probability that both X and Y are large? Can we say something about the uncertainty in the distribution of their sum X+Y or any other functional f(X,Y) of interest? The issue is that the joint model is not fully specified and there will be infinitely many joint models that satisfy the constraints. How do we pick one of these joint models to obtain a conservative assessment of risk? Much insight can be provided by considering different copulas and the topic lends itself to visualisation.

Location: ARRC Auditorium A/RC/014

Who to contact

For more information on these seminars, contact:

Thomas Patton
Dina Jankovic

Economic evaluation seminar dates

  • Thursday 17 May
    Rita Faria, CHE,
    University of York
  • Thursday 14 June
    Dr Claire Rothery, CHE, University of York
  • Friday 13 July
    Julie Ratcliffe, Adelaide
  • Thursday 19 July
    David Tordrup, Triangulate Health Ltd
  • Wednesday 19 September
    Dr Gianluca Baio, UCL
  • Thursday 18 October
    Dr Han-I Wang,
    Health Sciences, University of York