September 22, 2020

expert reaction to review study looking at the proportion of people infected with SARS-CoV-2 who develop symptoms

A review, published in PLOS Medicine, looked at the proportion of people infected with SARS-CoV-2 who develop symptoms.

Dr Julian Tang, Honorary Associate Professor in Respiratory Sciences, University of Leicester, said:

“I would interpret this paper’s findings with some caution – and use their estimate of a 20% asymptomatic infection rate as just a lower limit.

“Although it addresses an important question, it has several limitations, which are mainly due to the limitations in the source material.

“Many of the papers included in the review likely did not test for other respiratory viruses when they tested for SARS-COV-2.  If patients were infected with other respiratory viruses and as well as SARS-COV-2, then the COVID-19 symptoms may have been due to these other viruses (like flu, RSV, parainfluenza, adenovirus, human metapneumovirus, rhinovirus and seasonal coronaviruses).

“Exactly how large this potential confounder effect might be is uncertain – coinfection with more than one respiratory virus can be variable depending on the season.

“Also most of the papers included in the review would have been based on hospitalised patients.  Clearly such patients are more likely to have symptoms as this is why they attended hospital.

“In contrast, if you perform large-scale community surveillance with PCR and serology, you may find a very different picture – where a higher proportion of SARS-COV-2 PCR and/or antibody positive cases are more likely to be asymptomatic, as shown in the recent UK REACT studies¹.

“So it would have been clearer and more informative if the authors had separated their findings into hospitalised vs. community patients.

“Similarly, the authors could have also stratified these COVID-19 populations by age to show the age-related symptomatic vs. presymptomatic vs. asymptomatic rates of infection more clearly – as they do recognise that children and adults present with very different symptoms, and the proportion of cases showing symptoms in these different age groups does vary significantly.”

1 : REACT 1 (SARS-COV-2 PCR surveillance): 72% asymptomatic at the time of swabbing and in the previous week, ~600,000 participants.  https://www.imperial.ac.uk/media/imperial-college/institute-of-global-health-innovation/public/Resurgence-of-SARS-CoV-2-in-England–detection-by-community-antigen-surveillance.pdf

REACT 2 (SARS-COV-2 antibody surveillance): 32% asymptomatic, ~100,000 participants.

https://www.medrxiv.org/content/10.1101/2020.08.12.20173690v2

Dr Simon Clarke, Associate Professor of Cellular Microbiology at the University of Reading, said:

“This extensive overview of multiple studies comes to the conclusion that most people who are infected with the coronavirus causing Covid19 eventually develop symptoms.  Much has been made of the number of people with the virus, but without symptoms.  While this is very interesting, it remains the case that on any given day, the majority of people with the virus will not be displaying any symptoms and these findings should not in any way detract from current infection control advice.”

‘Occurrence and transmission potential of asymptomatic and presymptomatic SARSCoV-2 infections: A living systematic review and meta-analysis’ by Diana Buitrago-Garcia et al. was published in PLOS Medicine at 19:00 UK time on Tuesday 22 September 2020.

DOI: 10.1371/journal.pmed.1003346

Declared interests

None received.