select search filters
roundups & rapid reactions
before the headlines
Fiona fox's blog

expert reaction to R now being between 0.7 and 1.0

The government have released new figures for the R value of COVID-19 in the UK, now believed to be between 0.7 and 1.0


Dr Sebastian Funk, Associate Professor and Wellcome Trust Senior Research Fellow, London School of Hygiene & Tropical Medicine, said:

In the UK, there is perhaps a hint that R has recently edged up slightly, but no strong evidence.  The recent estimates all overlap, that is they are consistent with R staying the same as well as with an increase.  It is therefore too early to speculate whether recent changes in behaviour have increased transmission in the UK.

“Prevalence of infection tells us about the current state of the epidemic, R about the direction of travel.  They are therefore both important metrics to monitor.  In fact, the prevalence of infection affects the options for responding to R going above 1.  At a high level of infections R becoming greater than 1 has to be managed with a widespread measures such as the nationwide physical distancing implemented in many countries.  From a low level of infections it is possible to address settings with R>1 with a much more targeted approach, as seen in South Korea where entertainment venues were shut down and an enormous effort to trace contacts was initiated following a recent cluster in cases probably due to transmission in nightclubs.  But ultimately an R of greater than 1 indicates the potential for explosive spread and will therefore have to be managed in a timely fashion.”


Dr Konstantin Blyuss, Reader in Mathematics, University of Sussex, said:

“The values of R very close to 1 cannot be interpreted as any definitive guide for whether the epidemic is growing or subsiding due to the stochastic nature of the disease spread, time lags in data reporting, and statistical variation.

“It is very likely that R varies not only across the country, but also in various social settings, such as schools, care homes, hospitals, universities etc., because they have a very particular structure of social interactions that determines a faster/slower spread of infection.

“In light of recently announced nationwide roll-out of antibody tests, it should hopefully be possible in the near future to obtain some additional data that would provide accurate estimates of disease prevalence, which, in turn, is essential for correctly estimating R value, and hence, the subsequent course of the epidemic.

“As with many things, there is no one magic number or characteristic that would tell with certainty if the epidemic is growing or subsiding at any given time instance.  In this respect, despite all limitations, R is a very useful tool, but only when considered alongside accurate data on new cases, extended contact tracing, and antibody testing to determine if someone has already had the disease (even if the resulting immunity will not be life-long).”


Dr Yuliya Kyrychko, Reader in Mathematics, University of Sussex, said:

“The number R0 is a good and accurate estimate of how the infection spreads at the start of an epidemic, and is easy to calculate since everyone is susceptible, and no-one has had the disease yet.  In contrast, when the disease hits the pandemic level, especially, as in the case of COVID-19, there is a significant but unknown number of asymptomatic carriers, which makes the calculation of R much more challenging and prone to major statistical uncertainty.  To determine/estimate the value of R, one needs to know how many new infections are produced by a currently infected individual, and to determine this one needs to wait and see changes in the numbers of new cases, and this takes time, which explains the time lag.

“Since R quantifies the speed at which epidemic is progressing, and historic data over the last couple of years shows that the majority of influenza-like illness outbreaks in the UK arose at care homes due to many factors including close proximity of people, confined environment, and age, this strongly suggests that dynamics of COVID-19 in care homes has a profound effect on the overall level of disease prevalence, and the R number.

“Recent increases in R number can be attributed to the above-mentioned statistical variation in how it is calculated, but if adherence to lockdown measures had been waning that could also have played a part.”


Prof Karl Friston, Professor of Imaging Neuroscience, UCL, and panellist on the ‘Independent SAGE’ ( with special responsibility for modelling, said:

“An important alternative to calculating R – that provides a real time estimate within minutes – has been developed recently, using dynamic causal modelling1.  This evaluation of R is not based on curve fitting but on modelling the underlying causes of the reproduction rate, using state-of-the-art Bayesian methods.  This raises a pressing question: how are alternative estimators of R evaluated – and is the SPI-M considering all the alternatives?”


Prof Keith Neal, Emeritus Professor in the Epidemiology of Infectious Diseases, University of Nottingham, said:

How is R calculated?

“It uses a range of data of the number of new infections from as many sources as possible.

Does this new slightly higher R reflect the situation now or does it apply to a couple of weeks ago?

“It is too early for this increase in R to be a result of changes that started on Wednesday.  We have not seen the evidence the data that this conclusion has been reached.  It will most likely apply to things that happened a week ago.

Why is there a time lag?

“Cases are mainly diagnosed when they get symptoms or become ill.  There is lag of 5-7 days for symptoms and nearer 10 days for admissions.

Could this be explained by the situation in care homes?

“If the R is being pushed up by care homes this is not an issue for community transmission.  Care homes need to be controlled by infection control, the R in care homes does not influence your control measures in these circumstances.  The community R is what is crucial for determining what social distancing measures are needed to control transmission.

Can we tell why it might have gone up?

“This is difficult to judge as we don’t have data on how people are acquiring infection.  If we knew where transmission was occurring more targeted measures could be used.

What does an R of 0.7-1.0 mean for an epidemic?

“An R = 1.0 and the epidemic just keeps going, under 1 it declines and the more under 1 the faster the decline.

Is it likely R varies across the country?

“This is currently being documented with a much lower value on London.  It will also vary within regions.

Is it important to consider prevalence of disease as well as just the R?

“Yes, the R reflects transmission from the average case so more infections means more new cases for the same R.

Is R the be all and end all or is it not the only measure to judge the status of the epidemic on?

“No, in some places like hospitals and care homes R is a total irrelevance as the control strategy is infection control which does not depend on whatever R is.  It does remain the key parameter for what to do next.”


Prof Rowland Kao, Sir Timothy O’Shea Professor of Veterinary Epidemiology and Data Science, University of Edinburgh, said:

“The continued estimation of R being below one is a good sign.  Fluctuations in its value from week to week are only to be expected.  It could for example experience a temporary rise because of the introduction of a random case into a new area but then decline again as the R in that particular location starts to decline.  These fluctuations will likely become more pronounced as numbers go down, and so it is important to emphasise the overall trend, rather than a single point estimate.  The use of multiple approaches to get a joint estimate is therefore an important reassurance – it means that it is less likely to be due to the particular assumptions of a single model or approach to calculation.  However, that it remains close to one is an important reminder that even small changes in behaviour could make it shift above one again – should this happen for a prolonged period, case numbers could rise again, increasing burdens on the NHS, and risks to vulnerable demographic groups.

“Increases in R matter much more if the prevalence is high, as the overall number of new cases is a critical factor in determining, for example, additional burden on ICUs and logistical requirements.  Also important are considerations of geographical spread of new cases, and not just the overall R value.  High concentrations in one location may put local stresses on health care systems, while bursts of cases in new locations may put at risk previously unexposed individuals.”


All our previous output on this subject can be seen at this weblink:


Declared interests

Dr Sebastian Funk is a member of SPI-M (Scientific Pandemic Influenza Group on Modelling) for the UK government.

None received.

in this section

filter RoundUps by year

search by tag