Longitudinal Rise in Seroprevalence of SARS-CoV-2 Infections in Children in Western Germany—A Blind Spot in Epidemiology?
Round 1
Reviewer 1 Report
This article occupies the superiority that the investigation of COVID-19 infection is hot recently. I think readers will be interested in similar studies. Nevertheless, some contents of this article are unclear. Therefore, I suggest this manuscript can be accepted after a minor revision. Unclear contents are as follows:
- The authors employed a logistic function to fit data; however, this logistic function was not clearly defined. What is the LL2.2? What is the θ in this logistic function?
- The authors applied the drc-package of the R programming language to implement the mathematical analysis in this manuscript. Please briefly describe this R package. Which functions of this drc-package were used in editing this manuscript?
- This manuscript looks like a government report. Are there studies similar to this article? The authors may review such works.
- Check the format of reference lists. Some journal references were edited in different formats.
Author Response
Dear reviewer,
thank you very much for considering our manuscript for publication and providing helpful comments.
Comments and Suggestions for Authors
This article occupies the superiority that the investigation of COVID-19 infection is hot recently. I think readers will be interested in similar studies. Nevertheless, some contents of this article are unclear. Therefore, I suggest this manuscript can be accepted after a minor revision. Unclear contents are as follows:
- The authors employed a logistic function to fit data; however, this logistic function was not clearly defined. What is the LL2.2? What is the θ in this logistic function?
Thank you for pointing out the missing information. LL2.2. is a logistic function with minimum 0 and maximum 1 tailored to model binomial response. Parameters b and e serve as dummy parameters without significance for the purpose of total curve evaluation. This information has been added to the text in line 78-84.
2. The authors applied the drc-package of the R programming language to implement the mathematical analysis in this manuscript. Please briefly describe this R package. Which functions of this drc-package were used in editing this manuscript?
Core function of the drc-package is the so called drm fitting routine, which, in essence, is a least-squares-based curve fitting routine. The drc package offers a set of in-build functions to be fitted, like the LL2.2 logistic function used in our analysis. Furthermore, this routine offers a most elegant and easy to handle way of constructing confidence bands. Please see the revised methods section for a detailed description of the package, the logistic function, and criteria for their choice (line 78-84).
3. This manuscript looks like a government report. Are there studies similar to this article? The authors may review such works.
Government reports on SARS- CoV- 2 infection rates are published by the Robert Koch Institute (RKI) on a weekly basis. These reports, however, focus on acute SARS- CoV- 2 infections rates based on positive SARs- CoV-2 PCRs. There is no governmental monitoring regarding antibody-based seroconversion rates in Germany. The data of RKI reports show that the seroconversion number is two to fourfold higher than the number of acute PCR confirmed SARS- CoV-2 infections. These results are discussed in the results section (116-130).
4. Check the format of reference lists. Some journal references were edited in different formats.
The reference list has been updated and formatted
Reviewer 2 Report
The manuscript describes an epidemiological study on the seroprevalence of children to SARS-CoV-2 in Germany. The findings are interesting and relevant, and provide further evidence that children and adolescents should not be excluded from data collection or consideration of disease. There are some major limitations of the manuscript that need to be addressed and rectified prior to publication.
Major comments:
The article does not describe the serological methods being used to detect antibody response in children, and while it is briefly mentioned at the end of the discussion, there is no comment regarding whether or not the results varied based on test (IgG, IgM, neutralization), or if the tests were standardized across clinics or even validated. Given the timeframe of the study (6/2020 – 2/2021), significant improvements in antibody tests over time are an important consideration for showing an increase in seroconversion. This needs to be addressed.
There are 9 references for the entire manuscript, which, considering the amount of information available on COVID-19, is low. In the introduction alone, several statements are made regarding under-diagnosis of children and seroconversion in adolescents that are not properly referenced.
It would be good to comment on whether or not vaccination had started in Germany during the study timeframe as most people don’t know the specifics from country to country and certainly in some parts of the world, vaccinations were happening as early as January/February 2021, which could skew the results of the study.
In Tables 1 and 2, the number of study participants does not match. Table 1 shows a total of 2145 over the two columns, Table 2 shows a total of 2124. Why are these numbers different? And then in the text (line 79) it says 2184 children were tested. Also in line 97, the ratio states 55/2091, which in not what’s shown in Table 1. Throughout the figures and the text, there are numerous inconsistencies in the sample size and percentages. These all need to be rectified prior to publication.
Minor comments:
The abstract states that 41% of the respondents were symptomatic, but the study explicitly states “asymptomatic children” on line 33 and elsewhere. Please clarify.
Line 33: change “was” to “were”
Lines 46-47: This is an incomplete sentence, please modify to make it a complete sentence.
Line 50: clarify that the age groups are broken down by years
Line 88: X-mas should be Christmas
Line 97: Please explain the drop-out due to failures of testing. What does this mean?
Line 109: What is a factor of “roughly two?” Two-fold increase? Please clarify
Line 151: Remove the comma after “country”
Author Response
Dear reviewer,
thank you very much for considering our manuscript for publication and providing very concise and helpful comments.
The manuscript describes an epidemiological study on the seroprevalence of children to SARS-CoV-2 in Germany. The findings are interesting and relevant and provide further evidence that children and adolescents should not be excluded from data collection or consideration of disease. There are some major limitations of the manuscript that need to be addressed and rectified prior to publication.
Major comments:
The article does not describe the serological methods being used to detect antibody response in children, and while it is briefly mentioned at the end of the discussion, there is no comment regarding whether the results varied based on test (IgG, IgM, neutralization), or if the tests were standardized across clinics or even validated. Given the timeframe of the study (6/2020 – 2/2021), significant improvements in antibody tests over time are an important consideration for showing an increase in seroconversion. This needs to be addressed.
The assay used for antibody detection in the study is the commercially available Roche “quantitative Elecsys SARS- CoV-2 ®” IgM and IgG assay (Roche Diagnostics, Laval Quebec) detects antibodies to the N (Nucleocapsid) domain and is therefore only positive after infection, not vaccination. The same test was used in the same reference lab for the entire period of the study. It was not formally evaluated for this study. Neutralizing antibodies were measured on top of the commercial assay but are referred to in a different publication. This additional information has been added in
There are 9 references for the entire manuscript, which, considering the amount of information available on COVID-19, is low. In the introduction alone, several statements are made regarding under-diagnosis of children and seroconversion in adolescents that are not properly referenced.
The references have been updated and cover the statements mentioned in the introduction as well as additional information in the discussion section.
It would be good to comment on whether vaccination had started in Germany during the study timeframe as most people don’t know the specifics from country to country and certainly in some parts of the world, vaccinations were happening as early as January/February 2021, which could skew the results of the study.
All participants in the study were unvaccinated at the time of the examination. Until February 2021 SARS- CoV- 2 vaccinations were only available for the eldest (>80 years.) and few health care workers. This information has been added in line 41/42.
In Tables 1 and 2, the number of study participants does not match. Table 1 shows a total of 2145 over the two columns, Table 2 shows a total of 2124. Why are these numbers different? And then in the text (line 79) it says 2184 children were tested. Also in line 97, the ratio states 55/2091, which in not what’s shown in Table 1. Throughout the figures and the text, there are numerous inconsistencies in the sample size and percentages. These all need to be rectified prior to publication.
Table 2 has been deleted as it was not referenced in the text. The discrepancy can be explained by a different mode of selection. For this study focusing on seroconversion rates only, 2184 children were tested for SARs- CoV- 2 antibodies, but only 2045 could be included in the analysis. Dropouts were predominantly caused by an insufficient amount of serum for antibody analysis. Of those 2045, initially 52 had SARS- CoV- 2 antibodies 2090 had not while later three additional patients (in total 55) of the cohort developed antibodies. This was included in the text in line 112+115 as well as in the tables and figures.
Minor comments:
The abstract states that 41% of the respondents were symptomatic, but the study explicitly states “asymptomatic children” on line 33 and elsewhere. Please clarify.
The children were asymptomatic at the time of examination, but 41% of the participants with antibodies to SARS- CoV- 2 recalled a recent infection within the preceding three months before the study. We included this information in the text in line
Line 33: change “was” to “were”
That has been changed.
Lines 46-47: This is an incomplete sentence, please modify to make it a complete sentence.
The sentence has been completed.
Line 50: clarify that the age groups are broken down by years
This information has been added
Line 88: X-mas should be Christmas
That spelling mistake has been corrected.
Line 97: Please explain the drop-out due to failures of testing. What does this mean?
The drop out rate was caused by insufficient amounts of blood for antibody testing. Therefore, these patients could not be included in the analysis.
Line 109: What is a factor of “roughly two?” Two-fold increase? Please clarify
The ratio describes seroprevalence in our to officially registered cases of acute SARS- CoV-2 infection (governmental data) in the region. We clarified this in line 121/22.
Line 151: Remove the comma after “country”
Removed
Reviewer 3 Report
In this manuscript, Folke Brinkmann et al analyzed seroprevalence of SARS-CoV-2 in 2124 children and their parents from Ruhr region of Western Germany. The results of their study suggested a three to four-fold higher in seroconversion rates compared with officially registered PCR results and 59% among positive cases were asymptomatic. Moreover, they also found the 51% of infected children have at least one of their parents developed SARS-CoV-2 antibodies, which indicate household transmission is a major risk factor.
Due to the existence of asymptomatic or oligosymptomatic infections, the seroprevalence of SARS-CoV-2 in children and adolescents is easily to get underestimated. With routine testing samples from children and their parents, the study in this manuscript revealed real seroprevalence of SARS-CoV-2 in children and indicated household transmission as major risk factor for spreading infection. The work reported here is very interesting and important to understand seroprevalence of SARS-CoV-2 in children and improve infection control measures.
The manuscript was clearly written, provided necessary information, and discussed conclusions/limitations of this study.
Specific comment:
Because of lower sensitivity of antibody tests compared with PCR tests, there is risk that the result in this study cannot precisely detect seroprevalence for SARS-CoV-2 infections. It would be important and helpful if the authors could provide information about sensitivity (comparison with PCR test) of the antibody tests used in this study.
Author Response
Dear reviewer,
thank you for considering our manuscript for publication and adding useful suggestions.
Because of lower sensitivity of antibody tests compared with PCR tests, there is risk that the result in this study cannot precisely detect seroprevalence for SARS-CoV-2 infections. It would be important and helpful if the authors could provide information about sensitivity (comparison with PCR test) of the antibody tests used in this study.
Thank you for this comment. The antibody assay used (Roche Elecsys Anti- SARS-CoV-2 assay ®, (Roche Diagnostics, Laval Quebec)) has been described to have a sensitivity of 99,5 % in SARS-CoV- 2 PCR positive individuals in a small study by the company. In another cohorts with PCR confirmed SARS- CoV-2 infection 79% of the infected developed SARS- CoV- 2 antibodies (Pflueger et al. J Clin Virol. 2020). Therefore, seroprevalence rates could be lower than the number of actual infections. This limitation has been added in line 180.
Round 2
Reviewer 2 Report
The author's have adequately addressed my concerns. There are a few minor spelling/grammar edits to be made in the added text (i.e. line 13 should read "drop-out," add a comma after "seroconversion" Line 102).
Author Response
Dear Reviewer,
thank you very much for reading our paper very carefully and pointing out the spelling mistakes. We corrected them (and a couple more). Your input has been very helpful.
Best wishes,
Folke Brinkmann on behalf of the authors