New real-world data released last week has allowed Sage to improve the assumptions which underpin their models on both vaccine effectiveness and rollout. Crucially, a PHE study released last week showed for the first time that vaccines cut “breakthrough transmission” of the virus by about half after a single shot.
“If you look at where they were in early April, compared to where they were in early February, they moved a huge distance,” said James Ward, a mathematician and insurance risk manager, who runs his own Covid model which closely shadows the official ones.
“So actually, it’s not very far for them to move now, from predicting an exit wave of 15,000 to 20,000 deaths to them predicting an exit wave of zero to 5,000, or maybe nothing at all.”
Mr Ward is not critical of the Sage modellers, who he notes are obliged to follow the published evidence when deciding on the assumptions to use in their models in a way that he is not.
His own model – which has for some time predicted only a slight uptick in cases in the autumn – already includes the now substantiated assumptions on vaccine effectiveness.
Sage will look at modelling provided by Imperial College London, the University of Warwick and the London School. All three provide a range of projections but it is their “central scenarios” on which Sage places most weight.
In March, the central scenarios of the Imperial and Warwick models projected the virus would die off by September if the exit plan was stopped at phase two, but that it would spike again in the late summer if phases three and four were implemented.