My talk @ Imperial Infectious Disease

Today I went to the Department of Infectious Disease at Imperial College, to give a talk on our work on HPV vaccination. Eventually we managed to published both the applied work (which focussed on a specific application to the Italian National Health Service — which have eventually changed their national vaccination programme, also in line with our recommendations) and a more methodological piece, in which we were presenting our sort-of-compromise.

I think that infectious disease modelling is very interesting, particularly when it’s embedded in the wider economic evaluation. From what I have seen (and I think confirmed talking to experts, after the talk, earlier today), usually the focus is pretty much on the transmission model, which is used to estimate the population dynamics (ie how the infection rates vary depending of the mix of the underlying population, which is of course affected by vaccination — the more people are vaccinated, in general, the less are prone to infection). But when it comes to translating this to the economic component of the model (eg extrapolating over a long time-horizon, valuing costs and utilities, etc), things are less-than-ideal, with models often build effectively on point estimates, with hacks to approximate PSA.

So what our model was trying to do was effectively to re-address this balance: we have used a lower-temporal resolution version of the transmission model (in which we approximated the continuous infection rates in discrete time intervals); but by doing so, we’re able to effectively construct both the transmission model and the economic evaluation in one single step (and based on a Bayesian dynamic Markov model structure, which of course, at least in its more standard version, is almost second nature to health economic modellers).

I think the talk went well and I got some very interesting questions afterwards. And, I got lunch…