Exclusive: UK’s first ‘real-time’ R number shows sharp decline
The UK’s first daily calculation of the coronavirus “R rate” in real time shows a sharp fall in the past two weeks, a slight increase in recent days and a marked peak in early October – all missed by the government’s official numbers, which are produced only weekly and with data that are weeks out of date.
The calculations, carried out exclusively for the Telegraph by an Italian consultancy, also suggest that the UK’s R rate is at its lowest level since at least July last year.
The official R number- which shows how many people each infected person passed the virus onto – published by the government today is between 0.7 and 0.9. However, this number is not calculated in real time and does not represent the state of the epidemic now: instead it represents the situation “over the past few weeks”, in the government’s words.
The UK’s up-to-date R number, based on yesterday’s official infection data, is between 0.78 and 0.92, according to TomorrowData, a Turin-based consultancy that has been calculating Italy’s R rate in real time on a daily basis since last summer. Its numbers are in pink in the graph below while the Government’s numbers are in blue.
The apparent agreement between the two numbers is misleading: they refer to different stages of the epidemic. TomorrowData estimated the R rate of two weeks ago, roughly the date to which today’s official numbers relate, at about 0.93-1.03, whereas the official figures for that date put the R number lower at 0.7-0.9.
Comparing TomorrowData’s numbers for today with those of two weeks ago suggests a marked decline in the spread of the virus, albeit with a slight reversal in the past few days. The official R rate, based as it is on data from about two weeks ago, says nothing about this period.
The consultancy’s graph also suggests the existence of a pronounced peak in the R rate in late September, which, although marked by large uncertainty, was not picked up by the official figures.
TomorrowData’s co-founder, Stefano Terna, has been campaigning for the Italian authorities to use a real-time R number and not the one calculated by the country’s official bodies – which, like Britain’s, is based on data several weeks old.
“For making decisions about lockdown restrictions, old data are worse than useless,” he told the Telegraph. “But the researchers who compile the numbers have an academic mindset ill suited to dealing with a fast evolving crisis. Governments should recognise this and switch to the use of real-time figures.”
A spokesman for the Department of Health said: “The time delay between initial infection, having symptoms, and the need for hospital care means that epidemiological data usually takes up to three weeks to reflect changes in the spread of disease.”
He added that, while the most recent estimates of the R number were based on the latest data, available up to March 1, including hospitalisations and deaths, as well as symptomatic testing and prevalence studies, these estimates represented the transmission of Covid-19 “over the past few weeks”.
“As R is a lagging indicator, these estimates cannot account for the most recent policy changes, nor changes in transmission that have not yet been reflected in epidemiological data,” the spokesman said.