select search filters
briefings
roundups & rapid reactions
Fiona fox's blog
The government and Office for National Statistics (ONS) have published the latest estimates for the COVID-19 R value and rate of infection in the UK respectively.
Prof Oliver Johnson, Professor of Information Theory, School of Mathematics, University of Bristol, said:
“This increase in SAGE’s estimate of R is probably not a surprise. Values close to 1 reflect the fact that the size of the epidemic is currently not changing much, which is consistent with the weekly ONS infection survey and with daily case numbers. The estimate contains a range of outcomes, and the daily growth estimates still tend to be negative, which is somewhat reassuring compared with the current position in many European countries. However, the major concern is that R values of this magnitude do not leave a significant margin before the epidemic starts to grow in size again, and raise the possibility that some re-openings may need to be reversed to allow schools to open safely.”
Dr Yuliya Kyrychko, Reader in Mathematics, University of Sussex, said:
“The latest estimates of R number appear to indicate that in all regions of England, the estimates of R number have either remained the same or have grown in the space of one week. Most worrying about this is the possibility that in several regions, including London, South West and North West, the R number exceeds 1, suggesting a possibility of a substantial growth in new infections in these regions in the near future. With ONS data estimating around 2,400 new infections in England and Wales in the week 7-13 August, this means that the infection is still very prevalent across all of the UK, hence it is of paramount importance that people continue to follow the guidance on protection and social distancing to prevent major local outbreaks.”
Dr Konstantin Blyuss, Reader in Mathematics, University of Sussex, said:
“In the last few weeks, the majority of European countries have experienced a resurgence of coronavirus cases, and it appears that the same is starting to happen in the UK. The fact that the latest estimates of R number show the growth of infections in different regions of the UK suggests that the level of infection is indeed increasing in all of those regions, and is most likely driven by a local transmission. The increased number of cases puts a higher burden on the Test and Trace programme, which results in delays in returning test results, which can potentially lead to a higher overall number of new infection, because people are simply not aware that they have to self-isolate.”
Prof Kevin McConway, Emeritus Professor of Applied Statistics, The Open University, said:
“As has become usual on a Friday, three different sources of information on trends in numbers of COVID-19 cases have been published today. These are the weekly report on the latest results from the Office for National Statistics (ONS) Infection Survey, the report on the latest estimates of the rate of growth (or decrease) in new COVID-19 cases in England (from the Government Office for Science and SAGE), and the weekly COVID-19 surveillance report1 from Public Health England (PHE). As always, all these sources of information have strengths and limitations, and have their own peculiarities, and there inevitably remains some uncertainty about what’s actually happening. I’ll say something about all of them, and then summarise. My overall feeling is that there is not a major cause for immediate concern about the national position on new infections and cases of COVID-19, but that we must continue to be very vigilant. (And you must remember that I’m a statistician, not an epidemiologist.)
“The ONS infection survey has the great advantage that the people it tests are a representative sample of the community population of England (and now Wales too), and are not tested because they are showing symptoms or are working in particular roles. So its estimates aren’t affected by changes in overall testing effort, or by things like Government advertising encouraging people to get tested if they have symptoms, which can make the interpretation of other sources of test results more awkward. It produces effectively three different sets of figures: estimates of the number of people who would test positive on a swab test (for current infection) even if they have been infected for some time, estimates of the rate of new infections (as represented by people testing positive for the first time), and estimates of the percentage of the population that had a previous infection, possibly a long time ago, as measured by blood tests for antibodies. There’s not much news on the antibody testing this week. On the swab testing, ONS continue to report that the number of people who would test positive (for a current infection) seems to be levelling off, after a small increase since its lowest point at the end of July, and that the same is true of the rate of new infections in England. That rate of new infections (the incidence rate) is perhaps the most important in considering how the epidemic is currently changing, and for the most recent week (ended 13 August) the main estimate is down quite a bit on the previous week. The ONS statisticians cannot be sure that there is a real decrease, however, because the estimates come from survey data and are subject to quite a lot of statistical uncertainty, but it’s some sort of indication that things may be moving in the right direction on average.
“The new growth rate and R estimates are less encouraging, but still encouraging to a certain extent anyway, I’d say. These estimates are based on a wider range of data than just the infection survey results, including data on hospital admissions, ICU admissions, deaths, and some surveys of people’s behaviours. An issue is that numbers of hospital admissions and of deaths don’t directly relate to rates of new infections in the last week or so, because people going to hospital, or dying, would have been infected some time before. So these growth rate and R estimates tend to be affected by what was going on a few weeks ago, not just recently, to a greater extent than the latest ONS infection survey or PHE surveillance figures, and, like all the estimates I’m describing, are subject to uncertainty. Because of the uncertainty they are reported as ranges. It’s true that the ends of the ranges have increased from last week, and that must give a certain amount of concern. The Government are now saying that the numbers of cases might be falling by 3% a day, or increasing by 1% a day, or anywhere in between, but that the most likely figure is somewhere in the middle of that range, which would correspond to a slow decrease in cases. That’s similar to the picture presented by the ONS figure for new infections – if anything, it seems to be falling, but not very fast, and the ONS and SAGE statisticians and modellers can’t be confident that it’s falling. The new range for the R figure shows the same; it could be that the epidemic is growing, but it probably isn’t, it’s probably declining, though not very fast. And it seems clear that, at this stage in the pandemic, the most important figures aren’t the overall rates of change at national level, but what’s going on in specific areas where there might be an outbreak, or where a previous outbreak has been contained and reversed.
“The PHE surveillance report presents data from many sources on patterns of new cases – swab testing (related to Track and Trace, and otherwise), incidents of acute respiratory infection related to different institutions and premises, contact tracing, calls to NHS111, Google searches, data on GP activity, attendances at emergency departments, hospitalisations, and more. All of these provide only partial information and are subject to differing amounts of uncertainty. But, in my judgement, overall they present a similar picture to the ONS infection survey and the new figures on growth rates and R. The balance of the evidence seems to indicate that the rate of new cases is decreasing, overall, though very slowly, and we can’t be confident it really is a decrease. And the PHE report provides a lot of detail on differences between different locations, including indicating which local authority areas are on the Watchlist (decided by the Health Secretary, ultimately), where specific actions such as restrictions on some activities might have been imposed, or where there is concern. Actually the current trend in infection rates in many of the areas on the watchlist is downwards, though in others the trend is upwards, and three new areas (Northampton, Birmingham, Slough) have either moved onto the watchlist this week, or have a higher (more serious) status on it than last week.
“All of this fits with my overall assessment (as a statistician, not an epidemiologist) that the current path of the pandemic in England, on average, is downward, though we can’t be confident of that because there’s too much unavoidable uncertainty and because, if new infection rates are indeed falling on average, they aren’t falling fast. But that overall pattern hides a lot of variation from one place to another, which is why I’m pleased that local actions to contain outbreaks are being taken. It would be much more pleasant for everyone, particularly for the people living in the areas in the watchlist, if those actions weren’t necessary – but in my view they very much are necessary.”
1. Linked from https://www.gov.uk/government/publications/national-covid-19-surveillance-reports
Dr Michael Head, Senior Research Fellow in Global Health, University of Southampton, said:
“We are seeing increases in new daily cases, and now we have an R number estimate that is approximately at the 1.0 threshold. The data suggests that as people mix more freely, there are probably increases in community transmission. This is to be expected, but highlights the problems in coming out of lockdown, particularly when there are many thousands of active cases in a population. Suppressing COVID-19 is extremely difficult. We know that most cases are transmitted within the indoor environment. It’s currently August, and so outdoor meetings are possible, but as we head towards the winter, fewer people will be inclined to meet outside. That raises difficult questions about how best to handle social contacts across the winter months, particularly with the increased potential for transmission from younger populations, who usually have relatively mild cases, on to vulnerable populations where the consequences can be more serious.”
R number and growth rates: https://www.gov.uk/guidance/the-r-number-in-the-uk)
ONS Infection Survey: https://www.ons.gov.uk/peoplepopulationandcommunity/healthandsocialcare/conditionsanddiseases/bulletins/coronaviruscovid19infectionsurveypilot/englandandwales21august2020
All our previous output on this subject can be seen at this weblink:
www.sciencemediacentre.org/tag/covid-19
Declared interests
Prof Kevin McConway: “I am a member of the SMC Advisory Committee, but my quote above is in my capacity as a professional statistician.”
None others received.