What we found
In establishing an RADIR value based on cases we were
helped, during the period that we evaluated, by the UK National Health
Service (NHS) in England operating on a testing strategy to confirm a
diagnosis in symptomatic individuals. With a third of tests in
symptomatic individuals being positive over the period studied, we made
the assumption that positive tests reflected a small consistent more
severe proportion of the total infected cohort.
We showed an independent relation over time between our calculated
RADIR value and the number of cases identified as
definitely positive in the population. We then inferred by extrapolation
using a linear regression model that by late April, 26.8% of the total
population were possibly infected (1). We did not seek to establish an
absolute proportion of the total infected number for the population.