select search filters
briefings
roundups & rapid reactions
before the headlines
Fiona fox's blog
A preprint, an unpublished non-peer reviewed study, reports on the latest data from the REACT-1 study on COVID-19 spread across England.
This Roundup accompanied an SMC Briefing.
Prof Sheila Bird, Formerly Programme Leader, MRC Biostatistics Unit, University of Cambridge, said:
“REACT-1 has released an urgent interim report on the 6th round of its swab-testing (75% sensitivity assumed). In total, 86,000 persons aged 5+ years participated during 16th to 25th October 2020, when England might have had (but did not) a two to 3-week circuit-breaker to coincide with school holidays.
“Uncompromising message: This highly competent analysis sends an uncompromising message. The overall weighted prevalence of SARS-CoV-2 infection in the community in England was 1.28% (1.15% to 1.41%), or 12.8 per 1000 during 16-25 October: up from 0.6% (0.55% to 0.71%), or 6 per 1000, during the previous round (18 September to 5 October).
“During the 16 to 25th October, R was estimated at 1.6 (1.3 to 1.9) with swab-test-positive infections doubling every 9 days (95% uncertainty: 18 days to 6 days).
“Reason for weighting: Re-weighting adjusts for response-rates that vary across the population invited to take part. For example, the most deprived two quintiles contributed only 24% (20,670, nearest 10) of REACT-1’s 85,970 participants – not the 40% that should be expected; the middle quintile accounted for 18,290 (21%); while the least deprived two quintiles accounted for an impressive 47,010 participants (55%). Re-weighting does not correct for systematic bias within-deprivation-quintile between the quintile’s participants and non-responders.
“Re-weighting did, however, increase the estimated prevalence in the most deprived quintile from its unweighted estimate of 1.5% (uncertainty: 1.3% to 1.8%) to 2.2% (uncertainty: 1.7% to 2.8%) versus around 1% in the middle and least deprived pair.
“COVID case contact: Between Round 5 and interim Round 6, weighted prevalence doubled from 0.5% to 1.2% for persons who had no known contact with a COVID-case; remained around 9% for those in recent contact with a confirmed COVID case (such as, for example, members of the household of a confirmed cases whom Test & Trace asks to self-isolate); and more than doubled from 1.1% (uncertainty: 0.6% to 1.8%) to 4.2% (uncertainty: 2.25 to 8.2%) for those in contact with a suspected COVID case. The latter is probably explained by “suspect” being now more likely to herald imminent confirmation that was the case in Round 5.
“Asymptomatic infections: Across all rounds of REACT-1, at least half the people with detectable virus: will not report symptoms on the day of testing or in the week prior. The strong recommendation on 23 July 2020 by the Royal Statistical Society’s COVID-19 Taskforce was that persons whom Test, Trace and Isolate asks to self-isolate should be offered swab-testing on a random pair of days during their quarantine to learn efficiently about asymptomatic infections. Shamefully – in terms of infection control – Test, Trace and Isolate’s high-risk cohorts have been offered no such swab-testing.
“North East excepted: Exceptionally, REACT-1 illustrates a modest but convincing downturn in the rate of increase in SARS-CoV-2 infections is apparent in the North East.
“The North East has a long and proud tradition of contact tracing for infection control. Here, I declare an interest – the first HIV infections in the North East were diagnosed by my late husband, Dr A. Graham Bird, then consultant immunologist at Newcastle, who also instigated contact tracing and counselled HIV testing for potentially-infected partners and children.”
Prof Kevin McConway, Emeritus Professor of Applied Statistics, The Open University, said:
“It’s not entirely clear to me why the REACT-1 study has published interim results in the middle of this round of data collection, and indeed of the previous round, when they did not do so for the rounds before that. The results so far are based on only about half of the planned swab tests for round 6 (86,000 tests over 10 days from 16 to 25 October). That’s still a lot of tests, so their estimates of the average number of infected people in England over that period should be reasonably reliable, if the people tested in the first part of round 6 are roughly a random selection of the people who will eventually be tested in the whole round. On the last round, the estimate of the infection rate did change between the interim report and the final report on the round, though only by a fairly small amount that was within the calculated margin of error. Assuming something similar might happen in this round, I’m reasonably confident that the estimates of prevalence (the percentage of people who are infected) won’t be any less accurate than is given by the margin of error given in this interim report.
“The REACT-1 team estimate that, for this first part of this sixth round, 128 people in every 10,000 in the community in England are infected with SARS-CoV-2, with a confidence interval (reflecting the statistical uncertainty in this estimate) going from 115 to 141 per 10,000. That’s considerably more than their estimate from round 5 (18 September to 5 October), where the central estimate was 60 per 10,000 people, with a confidence interval from 55 to 71 per 10,000. So the proportion of infected people appears to have more then doubled, according to these estimates, over roughly three weeks between the middle of the round 5 to the middle of the period when these interim round 6 data were collected. These increases aren’t confined to particular groups of the English population. The report also gives estimates of the prevalence of infection for more than 40 groups within the population, defined by gender, age, the regions where they live, their type of employment, ethnicity, household size, and more. In all but one of these groups, the infection rate went up between round 5 and this part of round 6. (The one group where the rate fell rather than increasing was in the largest households, with 7 or more people, but there are few people in the REACT-1 samples that live in households that big, so that their infection rate is not estimated very accurately.) There are differences between the groups on how much it went up. Notably, it increased proportionally less in people aged 18-24 than in other age groups (where the infection rate was particularly high in round 5), and in the North East compared to other English regions – but it still has increased in these groups too. None of this is good news. Prevalence of infection remains higher in most of the North (and in the West Midlands) than in the rest of England, but it seems to be increasing at a slower rate in some of the North than elsewhere. That is slightly encouraging, but it doesn’t apply right across the North (in particular, not in the Yorkshire and the Humber region).
“The report estimates that 960,000 people in England were infected with SARS-CoV-2 on any given day, with an interval showing the statistical uncertainty from 860,000 to 1,050,000 people. Those are worryingly high numbers. They are also considerably higher than the estimates of the number of people who would test positive from the most recent report from the ONS Infection Survey (which had a range from 407,500 to 459,300). It’s not really surprising that the results differ. The ONS results are for an earlier period, 10-16 October, and data from all sources agree that numbers of infections are growing quite fast. Also, a technical point, the two surveys deal with potential false negative test results differently. The main ONS estimates are for the numbers who would test positive, whereas the REACT-1 estimates make and adjustment for inaccuracy in the test results and estimate the number of people who are actually infected. Making such an adjustment on the basis of figures in the ONS report, their range for the number of infected people might run from 473,000 to 585,000 – still considerably lower than the REACT-1 figures, but remember they are for an earlier week. We might be able to make a more appropriate comparison after this week’s ONS results come out on Friday, but, honestly, it is probably safer to wait till after the full REACT-1 results come out before comparing the two sets of survey results.
“The REACT-1 estimates for the R number and the daily growth rate of infections are, on the face of it, perhaps even more concerning than the increase in the numbers of infected people. But I’d urge considerable caution about those estimates. The estimates are based only on 8 days’ data. The range of statistically plausible values for R is pretty wide, from 1.27 to 1.88, and the same goes for the growth rate (from about 4% to about 11% increase in infections per day) and the doubling time for infections (from 6 days to 18 days). Also, it’s likely that these ranges do not capture all the uncertainty in the estimates. That is because estimating R and the growth rates requires a more complex statistical model than estimating the current prevalence of infections. In Round 5, there were reasonably large changes in the estimates of R and the growth rate between the interim and the final report. If infections really are growing at 8% a day, as the central estimate from this interim report suggests, then that’s concerning, and it’s a higher growth rate than last week’s Government estimate for England, but I really do think we need to be careful about these estimates based just on the first few days of round 6.
“The estimates based on round 5 of REACT-1 together with round 6 so far are likely to be more reliable, and they are considerably lower. The range of statistically plausible values for R is from 1.18 to 1.23, rather lower than the most recent Government estimate for England (1.2 to 1.4). The range for the growth rate from these rounds of REACT-1 is from 2.7% to 3.4% per day, which is again similar to last week’s Government estimate (3% to 4%). The estimate of doubling time from REACT-1 round 5 and round 6 so far has a range from 20 to 26 days.
“REACT-1, like the ONS Infection Survey, draws its results from swab tests from a reasonably representative sample of the English community population. Unlike the daily figures for new confirmed infections, the REACT-1 and ONS survey estimates are therefore not affected by changes in test availability or in the places where tests are concentrated.
“The REACT-1 researchers conclude by saying that “it is now time-critical to control the virus and turn R below one.” I quite agree. It may well be this time-criticality that led them to publish another interim report. However, they give no details on exactly how this might be done, and their survey cannot really throw direct light on that, so there are still very important questions to be answered.”
Prof Paul Hunter, Professor in Medicine, UEA, said:
“The recent report from the REACT-1 study by Riley and colleagues include results from samples taken between 16th to 25th October so give very close to up-to-date data. The REACT-1 study is in fact very similar in its design to the ONS Coronavirus Infection survey which was last reported on the 23rd October.
“The most obvious finding in the latest REACT-1 report is that the weighted prevalence of swab positivity is now 1.28% (1.15%, 1.41%), which is more than double that of 0.60% (0.55%, 0.71%) obtained in round 5. This compares to a weighted prevalence of just 0.16% (95% CI: 0.13%,0.19%) from 1st May to 1st June 2020. By contrast the ONS prevalence data for the two weeks ending 16th October was 0.88% (0.83% – 0.94%) and for the first two weeks in May was 0.37% (0.26%-0.5%).
“Although we do not know for certain what the prevalence of infection was at the April peak, it is likely that infection rates now are very similar to and maybe even higher than at the peak in April.
“The data from REACT-1 and from the ONS survey are prevalence statistics and do not directly give a figure for the number of new numbers of cases occurring each day, they do nevertheless give a good indication of the trend in infections. Both studies essentially confirm what we believe to be the state of the epidemic based on the daily statistics of new cases produced by the UK. All three sources of data indicate a rapidly increasing epidemic in the UK. But because of the lag between becoming infected and developing a positive test such data does not necessarily tell us what may be happening now with recent changes in restrictions.”
Prof Igor Rudan, Joint Director of the Centre for Global Health and WHO Collaborating Centre, University of Edinburgh, said:
“This is a timely and much needed epidemiological analysis of the prevalence of SARS-CoV-2 infection in representative samples of the population in England. The study is based on a very large sample size, i.e. 86,000 individuals. The PCR-based test based on self-administered swabs that were used to establish the status of infection should be considered accurate. Although the study based on the swabs taken between Oct 16-25 implies that only 1.28% of the population in England tested positive and carried a virus in that period, this is, in fact, very concerning. Comparatively, this is about twice as many infected individuals per 100,000 people than in the previous round, which took place between 18 September and 5 October. Also, the national reproduction number parameter R is estimated to be 1.56, which is much higher than in the previous round, when it was 1.16.
“The numbers of infected people are relatively the highest in Yorkshire and The Humber and the North West. The speed of spread of the infection is the most concerning in the South East, East of England, London and South West, although the estimates of that parameter are somewhat less certain. It is particularly concerning that the fastest increase in the number of infections is being seen in the age groups of 55-64 years and 65 or more years of age. This study should be considered very accurate and reliable scientific evidence that shows that a very large second wave of COVID-19 pandemic is underway. It will inevitably lead to a very large number of infections, severe episodes and deaths in the coming weeks and months. Efforts will be required to reduce the national reproduction number below 1.0 again. The measures that were in place over the past two months across most of Europe were clearly insufficient to prevent the new large growth of infected cases and fast spread of the virus. I commend the authors for this excellent effort.”
Dr Simon Clarke, Associate Professor of Cellular Microbiology at the University of Reading, said:
“This latest REACT study paper covering 16th to 25th October contains some sobering numbers. 128 people per 10,000 were infected with the coronavirus in England, which is up from 60 per 10,000; the virus was doubling every 9 days, a substantial increase in the tempo of spread. The national R number increased to 1.6. These are robust findings and should set alarm bells ringing if they’ve not already, they indicate a significantly deteriorating situation which needs an urgent response. 18-24 year olds remain the age group with the highest level of infections, but in the 55-64 age group, the number infected people has tripled, demonstrating how the virus spreads into more at risk people if left unchecked. We can expect this situation to continue to deteriorate if authorities remain slow to react.”
Preprint (not a paper): ‘High prevalence of SARS-CoV-2 swab positivity and increasing R number in England during October 2020: REACT-1 round 6 interim report’ by Steven Riley et al. was posted online at 00:01 UK time on Thursday 29 October 2020. This work is not peer-reviewed.
Declared interests
Prof Kevin McConway: “I am a Trustee of the SMC and a member of the Advisory Committee, but my quote above is in my capacity as a professional statistician.”
None others received.