Source: United Kingdom – Executive Government & Departments
A preprint (non peer-reviewed research), published on the German Research network Zoonotic Infectious Diseases website*, looked at the amount of virus (viral load) in patients with COVID-19 from different age groups.
Dr Alasdair Munro, a clinical research fellow in paediatric infectious diseases, said:
“This paper was released via social media in a format which has not yet been peer reviewed, and is unclear if has been submitted to a pre-print server. There are several methodological issues which compromise the ability of the paper to answer the question which it addresses, including some inappropriate statistical analyses which have resulted in an inaccurate conclusion.
“The methods of virological testing and interpreting viral loads are common and not-controversial.
“Some extremely important patient information is missing to enable any meaningful interpretation of the results. There is no indication as to how the population sample was acquired for testing. It is unclear if they are all symptomatic or not. It is not clear if they are derived from hospital inpatients or community testing. It is unclear the severity of illness of any of the patients. It is not clear if these factors are consistent across the arbitrary age categories or not. Without understanding the population tested it is not possible to conclude with any certainty what these results represent. The conclusions are drastically different depending on whether these are medically unwell population, or population screening, or a mix of both.
“In conclusion, this paper does not present the most important information necessary to draw meaningful conclusions from the data. Furthermore, the statistical approach was inappropriate, and the paper itself contains a more appropriate analysis which contradicts the main reported conclusion of the study – these results most likely do represent a statistically significant difference in viral loads depending on age, however, as previously mentioned it is impossible to draw meaningful conclusions from this without information on the population from which these samples were drawn.”
Dr Maarten van Smeden, Senior researcher, Leiden University Medical Center, said:
“I make note of a common method of inappropriate statistical analysis of the data. Instead of analysing the data according to age as a continuous variable, the authors have created arbitrary categories according to different age brackets. This has drastically reduced the statistical power of being able to determine differences in viral loads between age groups, which is the main hypothesis of the study. Furthermore, the closest thing to the most appropriate analysis which was reported in the paper was a Kruskal-wallis test which actually found a statistically significant difference (p=0.01) in viral loads according to age: the opposite of the stated conclusion of the paper. Despite this being statistically different, from their analysis it is not clear if this would be a clinically significant difference. There are many important factors in determining how infective an individual is beyond the detectable viral load, which are not addressed in this study.”
Dr James Gill, Honorary Clinical Lecturer, Warwick Medical School, and Locum GP, said:
“Things are very pressured in the time of COVID-19. Particularly when it comes to scientific journals and publications. As a result we are seeing a growing number of media reports based on preprint material. It should be noted that the preprint is one of the early stages in submission of an article to be peer reviewed before the work is potentially accepted as an academic paper to be published. As a result of this, a degree of caution should be taken with articles being based on preprint work.
“One way that a preprint article could be viewed, is looking at a car in a show room, yet not being able to test drive it, and you have to simply take the authors (sellers) word the car runs well. By contrast a journal article in a peer reviewed publication, to return to our car analogy, means an independent mechanic has kicked the tyres, taken it for a test drive, had a look under the bonnet and discussed their thoughts with colleagues, before recommending if you should buy into it. You can be confident that you know what you are buying / reading when the article has been peer reviewed and published, whereas you need to exercise a greater degree of caution with a preprint.
“With that in mind, one of the latest preprint articles by Jones et al is reporting that children may be as infectious as adults when it comes to carrying the COVID-19 causing Coronavirus. So far the evidence has suggested that children seem to be at lower risk of developing COVID-19 and those which do tend to have very mild symptoms.
“As a result of seeing fewer children with COVID-19 symptoms, we have even less paediatric data on asymptomatic children than we do on adults. There has been suggestions from early in the outbreak that the proportion of asymptomatic adult carriers may be between 18-30%. Jones et al make the reasonable extrapolation that given children have milder symptoms, the number of asymptomatic children may be similar, if not higher. Their concern that children may be as infective as adults, seems to be a little less robust however, given we are still determining the role of asymptomatic patients in the spread of coronavirus.”
Prof Martin Hibberd, Professor of Emerging Infectious Disease, London School of Hygiene and Tropical Medicine (LSHTM), said:
“This manuscript gives very good evidence that the amount of virus collected in respiratory swabs from people identified as COVID-19 positive is similar across all age groups. This suggests that children may be able to transmit the virus as easily as other age groups although that is not tested in this manuscript. Further research is required to understand other factors involved in viral spread, for example coughing, as children without symptoms may cough less and transmit less through this mechanism. Ideally this would take the form of actual transmission studies based on contact tracing. However, this new research is consistent with the concept that children can be infected and perhaps also transmit the virus, just as easily as other age groups, without suffering as many symptoms. With testing currently focussing mostly on those with COVID-19 symptoms, it may be difficult to quickly identify the full role of children in transmission.”
*‘An analysis of SARS-CoV-2 viral load by patient age’ by Jones et al is posted on the German Research network Zoonotic Infectious Diseases website.
This is not peer-reviewed work.
All our previous output on this subject can be seen at this weblink: www.sciencemediacentre.org/tag/covid-19