There are several types of R: the R0 that applies to a naive population with no immunity or interventions; and the “effective R” or Re (also called Rt) that the politicians are talking about, and that measures how we are doing in controlling the virus.
We calculate the R in several ways. What it was two to four weeks ago can be back-calculated by the changes in Covid-19 hospitalisations or deaths. The Office for National Statistics (ONS) is now doing national swab surveys to estimate the proportion of people who carry the virus and how that changes over time; but these take a week or two to process and report. Thus, we cannot be very confident about the precise value of the R day to day, and can be even less confident about regional variations that are inevitably based on less data than the national numbers.
The government’s science committee, Sage, estimates that the R value for the UK was between 0.7 and 1.0 on 22 May. This is an average value across the nation, and there is substantial regional variation – Re may be twice as high in the north-east than in London, the original centre of the UK’s epidemic. The R value varies across individuals, with so-called “super-spreading” events being associated with a high proportion of the infections, and many people, perhaps most, not passing the infection on to anyone. Infectious disease modelling and virus sequencing data suggest that, early in the epidemic, 10% or less of virus carriers led to about 80% of infections.
The amount of virus each individual carries may be partially responsible for this, but his or her opportunities to contact others – at parties, in the workplace, in restaurants or religious services – is important too. Thus, one man may have infected more than 90 clubgoers in South Korea, and one member of a choir in Washington State may have infected 52 out of 61 fellow chorists at a single rehearsal.
Early estimates of R0, the original infection rate, for Covid-19 were in the range of 2.0 to 3.0, but more recent estimates have come in at 4.0 or more. This is substantially higher than the usual estimate for influenza of about 1.4 to 2.0. The infection rate dropped incrementally with the series of recommendations that started on 12 March, but did not drop precipitously until the full lockdown was announced 11 days later. The epidemic grew much faster prior to the lockdown than it has diminished afterwards. To decline at the same rate, the R number would now have to be about 0.25 – and no one thinks it has declined to that level. As Harvard epidemiology professor Bill Hanage has put it: “A fire burns fast at first but the embers take a long time to die down.”
So where does this leave us over the next few months? With Re so close to 1.0 there are basically two options. First, we could try to accelerate the decline in the number of infected people – bringing Re down further – so that efficient test, trace and isolate measures combined with the quarantine of new arrivals might give us a chance of suppressing the virus to the point of virtual elimination, as has been done in South Korea, Taiwan, Australia, New Zealand and perhaps China.
It became clear from the prime minister’s announcement on 10 May on easing the lockdown that the government has chosen instead to open up the economy when there are still large numbers of infected people who do not know they carry the virus – guaranteeing ongoing infections and the inability to suppress the virus. As a consequence, the Covid-19 death rate will stabilise but may not drop much lower, and the threat to older people and those in care homes will remain serious.
Even to achieve stabilisation in the context of the “back to work and school” decisions already announced, three key control elements must be put in place. Indoor gatherings of any sort must remain limited in size to prevent super-spreading events, and those who encounter many people each day, such as shop workers, care workers and bus drivers, should be given protective equipment and be tested regularly. On trains and buses, there are obvious implications for how many people can be commuting. And the use of face coverings should be the new normal – including at government press briefings
The contact-tracking system must be up and running efficiently. Though a test and trace system is being introduced today, scientists have already warned that it will prevent less than 15% of infections, given the large numbers still infected.
And travel from one part of the country to another should be minimised, so infected travellers do not spark new epidemics in lower-risk regions. There are ominous signs that the mixed messages from the government have already led to more gatherings and crowded trains, and, in backing No 10 adviser Dominic Cummings, it has given the green light to long-distance travel.
Stabilising the epidemic only buys time to find an effective treatment or vaccine. Perhaps the warmer temperatures will help, if we responsibly socially distance outdoors while the epidemic rumbles on. But when the colder weather pushes us indoors, without a vaccine, and unless contact tracing is stepped up, there will be a second wave: the epidemic will resurge. Not because it had to, but because we did not push the virus closer to extinction, we did not plan properly for the rebound, and thus gave the virus a second lease of life. And all of this will guarantee more Covid-19 deaths in the UK.
• David Hunter is the Richard Doll professor of epidemiology and medicine in the Nuffield Department of Population Health, University of Oxford