February 1, 2022

expert reaction to the addition of ‎SARS-CoV-2 reinfections to the UK Coronavirus Dashboard

The UK government have updated their Coronavirus Dashboard to included cases of SARS-CoV-2 reinfection as well as new cases in the UK.

Dr Julian Tang, Honorary Associate Professor/Clinical Virologist, Respiratory Sciences, University of Leicester, said:

“The use of the term ‘reinfection’ has been a bit variable during the COVID-19 pandemic – let me give an example using influenza:

“Let’s say I have 100 people who catch seasonal influenza A(H3N2).

“Of these, 10 had a influenza B (Yamagata) in the last 6 months (if you check their influenza B-specific antibodies); similarly, 15 had influenza B (Victoria), 7 had A(H1N1), 5 had influenza A(H3N2). The other 63 had no prior infection to any influenza subtype/type – as indicated by antibody testing (as indicated by HAI titres >40, for example).

“So what are the reinfection rates here? Reinfection relative to what?

– relative to influenza A (either H1N1 or H3N2) – you could say there were 12 cases of reinfection with influenza A (ignoring the specific subtypes H1N1 or H3N2)

– relative to influenza B (Yamagata or Victoria) – none

– relative to influenza A(H1N1) – none

– relative to influenza A(H3N2) – 5

“For those with prior infection to a specific influenza subtype/type – the immunity to that specific subtype will last longer than for other others, e.g. the immunity to A(H3N2) will last longer against A(H3N2) than to A(H1N1) – though some degree of cross-reactive antibody B/T-cell protection exists between these two subtypes.

“OK, so let’s try this for SARS-CoV-2/COVID-19:

“Let’s say I have 100 people who catch Omicron.

“Of these, 10 had Alpha infection within the last the last 3 months (from prior PCR records – the standard SARS-COV-2 antibody tests cannot distinguish between the different variants at the moment); similarly, 15 had Delta, 7 had Gamma, 5 had Omicron. The other 63 had no prior infection with any SARS-COV-2 variant – as indicated by previous PCR records or standard antibody testing.

“So what the reinfection rates here? Reinfection relative to what?

– relative to SARS-COV-2 – you could say there were 37 cases of reinfection (ignoring the specific variants)

– relative to Alpha – none

– relative to Delta – none

– relative to Gamma – none

– relative to Omicron – 5

“For those with prior infection to a specific SARS-COV-2 variant  – the immunity to that specific variant will last longer than for other others, e.g. natural immunity to Omicron (through past infection – but not via the current vaccines) will last longer against Omicron than natural immunity to a prior Delta infection.

“So the numbers of ‘reinfections’ reported depends on how these are defined – and the severity of reinfection will depend to some extent on how this is defined also, e.g.

– infection with Omicron after a prior Delta infection does not count as a ‘variant-specific reinfection’ because it is a different variant; but could count as a SARS-COV-2 reinfection because the variants are still SARS-COV-2 viruses

“Some people may have more severe infection with ‘reinfection’ and others more mild illness – and this may relate to the previous infecting variant – depending on the host response (we see this with other viruses like influenza) – and we saw these different responses during the first COVID-19 wave with the original Wuhan variant:

– Both Fred and Ginger had Omicron infections 2 months ago (confirmed by PCR/sequencing) – and they both catch it again, but Fred is more sick than before, whereas Ginger’s reinfection illness is milder 

– Abbott and Costello both had Delta 3 months earlier, and both get infected with Omicron now (not a reinfection but an infection with a different variant) – both are asymptomatic.

– Eric and Ernie both catch Omicron having had Alpha 6 months ago, but Eric is asymptomatic and Ernie only has mild symptoms

“It is difficult to predict how sick a reinfection will be (however you define it) because some people will have a strongly primed immune response to any previous variant infection that could either protect them from more severe disease – or occasionally – overreact if they see the same variant or a different one again.

“Again, there were case reports of various scenarios during the early pandemic with the original Wuhan virus:

https://www.nature.com/articles/d41586-020-02506-y

 and reiterated more recently:

https://jim.bmj.com/content/69/6/1253

“So to really understand reinfections – with any virus – you need to define what you mean by this – and even when you do, the responses are variable – in terms of longer term immunity and severity of disease – because the disease is largely just-mediated and all of us have unique host immune responses due to genetics, sex, ethnicity, nutritional status, obesity, comorbidities, socio-economic status, prior exposure experience, diet, exercise, sunlight (Vitamin D) exposure/levels, etc.

“So making sense of any reinfection data can be very tricky.”

Prof Paul Hunter, Professor of Health Protection, The Norwich School of Medicine, University of East Anglia, said:

“Yesterday, the DHSC started publishing data on episodes of infection for England. Until now only the first infection in a person was recorded so that any reinfection was not reported on the daily dashboard. The fact that reinfections have become much more common since the emergence of the omicron variant meant that not reporting such reinfections was no longer tenable.

“No infectious disease surveillance system has 100% sensitivity in that they do not record every single infection. Comparing the DHSC dashboard with the estimates from the ONS infection survey suggested the dashboard previously reported about 1 in 2.5 infections, maybe fewer since the emergence of omicron. The surveillance system for covid is better than for most infectious diseases. For example, the UK Second Study of Infectious Intestinal Disease in the Community (IID2 Study) estimated that only 1 in about 147 IID episodes were actually recorded in the national surveillance system.  Nevertheless, given the increased rate of reinfections with the omicron variant this meant that the ratio between reported cases (DHSC dashboard) and estimated total infections (ONS survey) was changing making it difficult to fully understand the direction of travel of the epidemic in England over recent weeks, at least for the recent two weeks before the ONS estimates are produced.

“The new definition requires 90 days between the last positive sample and the next positive sample for this to be classed as a reinfection. Anecdotally we are hearing of reports of reinfections occurring less than 90 days apart but proving these to be true reinfections without sequencing every positive would be difficult. Reducing this time limit would increase the risk of primary infections that shed virus for a long time being recorded as reinfections. Primary infections remain the most clinically relevant infection as reinfections are usually though not always much less severe than primary infections.

“Up to Sunday evening there were a total of 14,257,268 first episodes (equivalent to cases under the prior definition) in England and 588,114 possible reinfections. This represents an overall increase of 4.1% in headline reported numbers. But the number of reinfections relative to first infections has increased considerably in the past two months with omicron. Prior to 1st December 2021 there were 8,872,932 primary and only 82,080 (0.9% increase) possible reinfections. In December there have been 2,616,148 primary, 200,176 (7.7% increase) reinfections reflecting the much greater risk of reinfections associated with omicron. So far in January there have been 2,768,188 primary and 305,858 (11.0% increase) reinfections continuing this upward trend in proportion of infections due to reinfection.

“The addition of reinfections to the DHSC Dashboard data has increased the headline number of infections marginally. Nevertheless, although the proportion of all infections that are due to reinfection is still relatively small, this proportion is increasing, and we can expect this to grow. So far, the addition of reinfections have not much changed estimates of the speed of increase or decrease in the epidemic over recent weeks. Fortunately, reinfections are usually, though not always, less severe than primary infections and consequently we can expect the severity of covid to decrease further as more and more of the daily infections are reinfections.” 

Comments sent out on 31/01/2022 before the dashboard change:

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

“The additional data on reinfections will be useful information, particularly at this time, when it is likely that re-infections will be occurring with increasing frequency and therefore the rate of new infections occurring decreasingly representative of the true course of the epidemic. If there are high numbers of reinfections then this will be due to a combination of the sheer number of people already infected, and the fact that, as immunity wanes, the likelihood is that protection is also waning. Thus having an estimate of the rate of re-infection (and remember it is only an estimate as many people will not know they previously had covid) will help to determine how quickly immunity is waning across the country and therefore crucially, when further boosters may need to be considered.

“Of course the addition of re-infection data will mean that the number of infections recorded is likely to go up substantially, without the situation itself having changed – preferably both figures will remain available, both presenting the number of new infections as recorded up to now, as well as the total. What is important to remember is that, while the total number of infections recorded will change, the numbers we should be most concerned about are the numbers of individuals with severe consequences and the number of deaths – so long as these continue to go down, it will be broadly good news, while noting the reverse – should those numbers go up substantially, whatever happens with the number of cases, this will be a matter of some concern.”

Dr James Doidge, Senior Statistician, Intensive Care National Audit & Research Centre (ICNARC); and Honorary Associate Professor, London School of Hygiene and Tropical Medicine, said:

“Omicron has proved capable of escaping immunity that prior infection provide against re-infection. However, current evidence suggests that reinfections with Omicron, as with previous variants, pose very little risk of experiencing severe outcomes [1,2]. As COVID-19 becomes endemic, it is crucial that we shift to focusing on severe outcomes and away from simply counting cases, including cases of reinfection. Will the dashboard be extended to include hospitalisation and death associated with reinfection?”

[1] https://www.medrxiv.org/content/10.1101/2022.01.05.22268782v1

[2] https://www.nejm.org/doi/full/10.1056/NEJMc2108120

https://coronavirus.data.gov.uk/

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

www.sciencemediacentre.org/tag/covid-19

Declared interests

None received.