The ‘Cambridge’ Model
An eminent team of statisticians at the MRC Biostatistics Unit (BSU),
University of Cambridge, are collaborating with Public Health England
(PHE) regularly to ‘nowcast’ and forecast COVID-19 infections and deaths
(4). This information feeds directly to the Special Advisory Group for
Emergencies (SAGE) sub-group, Scientific Pandemic Influenza sub-group on
Modelling (SPI-M), and to regional PHE teams,
The group have estimated the total infected population at 6,540,000
(12%). The findings also indicate an transmission rate R=0.4 in London
likely due to many more cases of infection at around 20% of the
population - in other words the London results appear to lead the
national trend. In contrast, the South-West of England with R=0.76 only
5% of that population are reported as having been infected.
As the Cambridge MRC BSU based their model on deaths which occur around
3 weeks after the MRC Biostatistics Unit (the report is dated
10th May 2020), the originating infections were
occurring around the middle of April 2020. This time period coincides
with the period to which our own published model applies.
The low R-value in now being seen in London likely relates to the
conurbation being at the forefront of the pandemic with a large number
of confirmed cases even prior to UK Lockdown on 23 March 2020.
Nevertheless, it is reasonable that there are a large number of
unreported cases in London, sufficient to reduce the R-value to the
level now being seen.