David Spiegelhalter – A rambling mind

Tuning the age and vulnerability model

Chart 19 for the grouped model UK957: modelled 800 day outcome compared with reported data to April 21st for active cases and deaths, with 2021 NPI relaxations up to June 21st 2021

In my latest post on March 26th I described a new Coronavirus group model, based on work I had done as a UK case study in support of Prof. Alex de Visscher’s paper, in conjunction with Dr. Tom Sutton, on “Second-wave Dynamics of COVID-19: Impact of Behavioral Changes, Immunity Loss, New Strains, and Vaccination” which has now been published for peer review as a pre-print on Springer’s site at https://www.researchsquare.com/article/rs-195879/v1. I have now added the latest UK vaccination progress figures, and the UK Government’s announced intentions for the near future regarding Non Pharmaceutical Interventions (NPIs). I have also updated mortality and infection characteristics for the four different population groups in the model.

Age and vulnerability related Coronavirus modelling

Chart 209 showing modelled and reported cumulative cases and deaths to 800 days in the groups model

In my most recent post on February 12th, I described modelling work I had done in support of Prof. Alex de Visscher’s paper, in conjunction with Dr. Tom Sutton, on “Second-wave Dynamics of COVID-19: Impact of Behavioral Changes, Immunity Loss, New Strains, and Vaccination” which has now been published for peer review as a pre-print on Springer’s site at https://www.researchsquare.com/article/rs-195879/v1. I have now added vaccination and multiple variants I had already added to our previous model into the new grouped population model, and this blog post reports on progress with that new model.

Model updates for UK lockdown easing points

Charts 4 and 13 for the comparison of cumulative & daily reported & modelled deaths, plus trend line, on the basis of 83.5% effectiveness, modified in 4 steps by -1%, -5% -10% and +2% successively

As I reported in my previous post on 31st July, the model I use, originally authored by Prof. Alex de Visscher at Concordia University on Montreal, and described here, was to be updated to handle several phases of lockdown easing, and I’m glad to say that is now done. Alex has been kind enough already to adopt a method I had been considering, of introducing an array of dates and intervention effectiveness parameters, and I have been able to add the recent UK Government relaxation dates, and the estimated effectiveness of each into a new model code. I have run two sets of easing parameters as a sensitivity test.