On 2021-08-07 15:14:47, user A Call for Honesty wrote:

What a number of critics ignore in their comments is that there are really poor countries that have a very, very limited supply of expensive medicines. If they have a good supply of Ivermectin, like one I know well, a caring doctor would not hesitate to try this medicine. I have had family that were quite ill with covid and responded very well to ivermectin. If their GP was in the US, UK or some other EU country, he would quite likely have his license withdrawn at the behest of politicians and academic experts who are not working with patients at the coal face. Less doctors in these situations would mean even more suffering and deaths.

On 2021-08-08 14:35:35, user Yaakov Kranz wrote:

Hi I'm curious was this study taking into acount the Delta variant?<br /> Did anything cahnge since then with regards to the Delta variant?<br /> Thanks!

On 2021-12-02 21:55:33, user Volker wrote:

The main message taken from this "paper" is "we estimate that unvaccinated individuals are involved in 8–9 out of 10 new infections.". This message is as wrong as it seems to be in comparison with the real world.<br /> One must add "that vaccinated individuals are involved in 5–6 out of 10 new infections.", which means "in more than half of new infections vaccinated individuals are involved". This is exactly what is shown in the graphics, too. <br /> The most dangerous thing about this paper is, that it is used as a base for political decisions only in a single part, not in complete.

On 2021-12-15 11:10:40, user DatenNarr wrote:

The main Shortcoming of this publication is that it accentuates the participation of unvaccinated individuals on new infections (91.1%, or 8–9 of 10), but does not mention the participation of vaccinated (49% or 5-6 of 10). 51 + 25 + 15 = 91 and 9 + 25 + 15 = 49 according to picture "FIG. 1". You see, the portion of vaccinated is not as small as someone would image (9%).

Because of the withholding of the participation of vaccinated individuals, the result description is insufficient and misleading. It gives people a wrong illusion as if the unvaccinated would be responsible for 91% infections. Such illusion leads to the effect that people fail to see the necessity to reduce the contacts of vaccinated persons in fight against the pandemic disease., with wrong politics about Covid-19 as result.

My proposal to the publication:<br /> 1) Change please the publication title to "participation of vaccinated and unvaccinated individuals on new Covid infections"<br /> 2) Add please the statement "vaccinated are involved in 5-6 of 10 new infections" after the statement "unvaccinated are involved in 8–9 of 10 new infections".<br /> 3) Add please the statement "the vaccinated population plays a role in 49% of cases" after the statement "the unvaccinated population plays a role in 91.1% ... of cases".<br /> 3) Take please real data about symptomatic cases from the RKI's weekly report of KW 41-44 for checking your modeling result, from which one can get that the vaccinated population plays at least a role in 40% of infection cases.

On 2021-12-04 08:24:25, user Benjamin George wrote:

I am skeptical about a paper that was designed with the alpha and delta variants in mind and reinfection rates. <br /> How does a paper that reviewed data from March 2020 to November 27 2021, suddenly include with great conviction that Omicron picked up on Nov 16 and 17 samples and announced on November 25 as a new variant show higher reinfection rates. <br /> All based on a few days of studies. <br /> And Omicron dropped into the summary to add credibility to the paper!

One should question the validity and professionalism of the writers.

On 2021-12-04 16:11:11, user Liz Jenny wrote:

Would be fascinating to further dissect the data with inclusion of vaccination status and age. Our initial wave in NYC was children (what I call the invisible leading edge of school absenteeism seen in early March 2020), then relatively young "out and about types". Implication of this article seems to suggest that COVID is undergoing antigenic drift leading driven by ambient antibody in COVID experienced population.

thanks to The NY Times for picking this up.

On 2021-12-07 07:25:37, user Atharva Gurav wrote:

How can I request for the data. I need it for research purpose.

On 2021-12-14 17:22:40, user Ronnie wrote:

they refuse to properly interpret the data ...?

Reaching the unvaccinated with current vaccines remains a priority in order to reduce transmission levels and reduce the potential for severe disease in the immunologically naïve.

On 2021-12-14 20:23:52, user John wrote:

You have tons of anti-vaxxers citing this article. Might want to have a better title and summary.

On 2021-12-18 19:05:20, user Ben Veal wrote:

Looks likely we'll be facing this individual freedom vs collective safety dilemma on a regular basis from now on, omicron is unlikely to be the last variant, and the inevitable flu pandemic hasn't struck yet.

The kinds questions we all want answers to are: if I go the cinema in an area with a population density of X, infection rate of Y, and an immunity level of Z, how much extra risk am I incurring on myself and others? <br /> How does this compare with other risks such as smoking in a public space, or driving above the speed limit? <br /> The comparison ought to be done in terms of life expectancy rather than lives, e.g. using microlives rather than micromorts, to ensure a fairer comparison. <br /> How do the corresponding costs of safety compliance compare?

On 2021-12-21 15:47:29, user Yingying Wang wrote:

A typo: "The group differences in fluency were not significant, and the CI group scored marginally lower in fluency tasks than their TH peers (g = - .67, p < .001)." should be "The group differences in fluency were not significant, and the CI group scored marginally lower in fluency tasks than their TH peers (g = - .67, p =0.054)."

On 2021-12-22 19:07:52, user Pedro Abdalla wrote:

The complete paper was published on: https://rdcu.be/cDtQS or https://doi.org/10.1186/s13...

On 2021-12-22 22:18:39, user Le Bon wrote:

Interesting work.<br /> I suggest to the authors to add another analysis grouping all vaccinated patients 3 month after dose 2, including thoose having received the dose 3 and simulating an ITT analysis.

There are several possibilities here.<br /> 1)The vaccine efficacy of 2 doses is really negative for Omicron 3 months after vaccine, and the third dose is really efficient against Omicron, at least during one month.<br /> 2)The negative efficacy is due to a temporal bias, or a selection bias (for example the more careful persons get vaccinated first) ins this case the 2 doses could be without a negative efficacy, but the efficacy of the third dose could be an artefact.

Those 2 groups have to be merged, and analyzed in time.

Typically, if the merged group has zero efficacy, an artefact is likely.<br /> If the merged group has globally a positive efficacy, a real effect of the the third dose is likely. (But no conclusions on the negative effect of 2 doses)

If the merged group has globally a negative effect, it means a real negative effect of 2 doses after 3 months (but no conclusions on the positive effect of 3 doses)

On 2021-12-29 01:23:50, user lowell2 wrote:

The negative estimates in the final period arguably suggest different behaviour and/or exposure patterns in the vaccinated and unvaccinated cohorts causing underestimation of the VE. --uh, evidence that suddenly at 91 days people started behaving differently than they did the previous 90 days? this conclusion in the discussion has no substantiation whatsoever. Maybe the vaccine just didn't work after that regardless of what people did or didn't do.

On 2021-12-25 08:38:40, user Eslam Maher wrote:

The authors investigate whether Machine Learning (ML) algorithms fare better compared to traditional Cox models in big data. They selected Glioblastoma and gliosarcoma from SEER as the basis of their data set. There are two main points that are worth considering here, (1) statistical, and (2) clinical.

(1) a- Glioblastomas are relatively rare diseases, therefore, readers need to bare in mind that the hypothesis studied here may not be relevant to their work that is usually mono-institutional or multi-institutional. Unlike the huge SEER database, we never actually have such numbers at hand to analyze in survival models.

There is no doubt that Cox would outperform ML models in smaller samples. ML is gaining popularity in the medical community that is hugely inflated and unnecessary.

b- Unlike ML approaches, the performance of Cox models is heavily dependent on its assumptions. This includes the proportionality of hazards between levels of a given variable, which the authors do not seem to have investigated this assumption before running the model.

Another assumption is how the model was selected in the first place. The authors say they have run Cox univariably to decide upon the variables that would be used in the final mode. It is unclear whether a "significant" variable is considered as such at 5% alpha. Regardless of the alpha level, automated stepwise methods are notorious, this is because they are very popular among physicians and not professional statisticians and epidemiologists. Stepwise methods do not allow modelers to think about the model at hand. Plus, some causal variables may not be statistically significant, while some nuisance variables may be coincidentally significant due to high N. Automated regression using p-values is a bad idea because it also ignores multiplicity problems.

(2) a- 22.6% of the cases included had no surgery, how then were they diagnosed as glioblastomas if no tissue samples were available? It is unclear if surgeries comprised craniotomies and biopsies or the former alone.

b- All glioblastomas and gliosarcomas are grade IV tumors, however, for some reason, grade is a variable included in the models with levels of grade I, II, III, and IV!

c- Reference categories in the authors' models were selected alphabetically rather than clinically. For Site, there are 14 levels using ICD-O classifications. Such classifications are not meant for clinical correlations. For example, all Lobar sites (frontal, pariental, occipital etc) are part of the Cerebrum. There are only 2 cases available for cauda equina glioblastomas, which is nonsensical to include as a separate level in the model (which puts more constraints in the model's degrees of freedom while also resulting in unstable ratios).

d- Finally, the median survival for glioblastoma patients as noted by the authors was eight months. Looking for model accuracy at 120 months is just insane.

This would have a been a neat paper had the authors run a proper Cox model rather than run a straw man, and designed their study with a neuro-oncologist. Even then, please note that this preprint is concerned with the performace of these models IN BIG DATA only, so do not extrapolate to the data you are routinely working with.

On 2021-12-25 11:00:44, user Jeffrey_S_Morris wrote:

These results are strange indeed.

The fact that this time period is so short, and that delta was still so dominant during this period, makes these results very difficult to interpret given the competition between delta and omicron

Given the already well documented immune escape properties of omicron vs vaccination, previous infection and monoclonal antibodies, yet the strong transmission advantages of delta, it is possible (likely?) that the infections from those with immune protection are predominantly omicron, but for those without immune protection it is predominantly delta. This effect could strongly reduce omicron specific infection rates in unvaccinated and have a strong effect on VE estimates.

Too bad they removed previously infected from these data and didn’t consider them a stratified cohort. If this competition effect was a major factor here, then one would also expect the cases among previously infections would also be dominated by omicron, and previous infection would also be deemed to have “negative effectiveness” vs omicron. These data would give us an idea of whether the negative effect is a component of vaccination or an artifact of this competition effect.

If omicron eventually dominates and delta disappears, then this competition effect would disappear and we’d have a clearer sense of VE of vaccines vs omicron outside of this competition effect.

But there is also a chance that delta has an inherent transmission advantage that will cause it to remain dominant in unvaccinated while omicron’s immune escape makes it dominate among vaccinated and previously infected.

In that case we might see delta and omicron coexist in some sense.

We will have to watch these data unfold in the coming weeks and months

On 2021-12-25 19:28:38, user Nate Dyer wrote:

Decades of coronavirus research warned us about this.

Why was it ignored?

Why did scientists choose to make vaccines that specifically target the spike protein when so many studies made the warning below?

"We suggested that antibodies against spike proteins of SARS-CoV may cause ADE effect. This data raises reasonable concern regarding the use of SARS-CoV vaccine and shed light on some roles in SARS pathogenesis."

https://www.ncbi.nlm.nih.go...

On 2022-01-08 03:26:25, user Neven Karlovac wrote:

Interesting article but the author's explanation of the negative infectivity seems arbitrary speculation and the article would be better without it.

On 2022-01-24 21:23:22, user KBNJ wrote:

Am I wrong in thinking it's entirely possible that the negative effectiveness of the vaccine for recovered people is real (biological), and not behavioral? Our bodies don't have unlimited immune resources. If vaccines induce an immune response that is less effective than recovered immunity at fighting an evolved variant (which various studies have shown at least for Delta), I would think this would be expected, and not a surprise.

On 2021-12-26 04:38:16, user anon wrote:

12 (95%CI 1, 7) Is this an error?

On 2021-12-26 08:33:43, user Jack Bean wrote:

"asymptomatic transmission" Really. No evidence for asymptomatic transmission. https://www.nature.com/arti...

On 2021-12-31 02:47:03, user Robert Crombie wrote:

Is the same true for Astrazeneca ?<br /> If Astrazeneca is less protective, why aren't the authorities warning those that have been given it ?

On 2021-12-31 07:01:37, user Georg Neeby wrote:

Hi, I was surprise to see your paper, it was mentioned in a conspiracy blog, https://childrenshealthdefe...

ITs being used as a reason to not get vaccinated, and for children specifically to not get vaccinated. I took a look at your data, and there is nothing to your paper. Its all just noise, nothing would be reproducible. Your claims massively overstate your data, which basically shows no difference, and you havent controlled for batch effects. Given how this information is being used by nefarious players, you should repeat the entire study with adequate power, I bet none of your new data will show the same trends as your old data.

On 2022-01-02 21:25:37, user madmathemagician wrote:

In its 4th revision, part of the title changes from an unsubstantiated suggestive question "Are COVID-19 data reliable?" which the article fails to answer, to "Applying Benford’s law to COVID-19 data: ...". The article mentions the conditions of applicability of Benford’s law, but does not test these assumptions on the source data, and "compensates" by making weak but suggestive claims, and recommending further research.

I'd recommend the author not doing any further research.

On 2022-01-02 22:34:18, user madmathemagician wrote:

The article does not discuss substantial differences in reporting by country in the used data set: some countries are only reporting every week, or not in weekends.

I guess the author could add another "disclaimer" about this issue, and creating a new revision. Perhaps even discussing how it would affect the fitness of a Benford's distribution.

However even when this issue is properly addressed, this article is about applying non-applicable statistics, concluding nothing, but making weak but suggestive claims.

On 2022-01-04 00:38:33, user Leicesterboy wrote:

This study looks at 29,000 Omicron patients, but the authors seem unable to find the courage to look at the statistics for “risk of death” independently, preferring to conflate it with other outcomes, like ICU or hospitalizations. Why? Isn’t that the outcome most of us are interested in and fear? By the way, I’d also be keen to see people be more definitive about the word “Hospitalizations” with COVID. We now know that hospitalizations arising from COVID are conflated with Hospitalizations with COVID, meaning those admitted due to an ankle sprain and discharged quickly are included in the “Hospitalizations” numbers. Clearly this is meaningless if the purpose of the analysis is to look at risk from COVID. The fact that someone with a sprained ankle had asymptomatic COVID has absolutely no interest for me. It Carrie’s zero information about health risk and potential outcomes for me.

On 2022-01-05 09:52:18, user Zacharias Fögen wrote:

Dear Authors,

Thank you for this in-depth analysis, yet you draw conclusions that lack physiological plausibility.

First, you draw conclusions from serum neutralisation to transmissibility. As any SARS-COV-2 variant replicates in nasopharyngeal epidermal tissue, and protection from these kind of viruses is solely reliant on sIgA or IgM but not on monomeric IgG, you cannot draw this conclusion.<br /> This has been well researched in animal coronavirus vaccinations and is also the reason why Polio vaccines are given p.o. instead of s.c. in endemic regions.<br /> Furthermore, you are describing the variant as hyper-transmissible, yet again, there is no physiological explanation for this. Entry into epithel cells is only a micrometer of the distance needed to move from one person to another. Changes in spike protein have no effect on the men-to-men transmission. <br /> Furthermore, you are relying on studies which do not control for differences in contacts between the group now infected with Omicron variant and those previously infected with Delta variant. There are less restrictions now in England and Denmark then when Delta variant arrived. Statistics also show that age group 20-30 is primarily infected with Omicron, and as those have the most variable contacts, there is clear indication of bias.

Best,

Zacharias Fögen

On 2022-01-05 16:22:39, user Ryan Anderson wrote:

How long until the booster protection equals zero?

On 2022-01-08 01:52:21, user Danuta Skowronski wrote:

When reporting paradoxical findings (e.g. negative vaccine effectiveness (VE)) based upon an observational study design, the first explanation to be considered is methodological bias (i.e. study weakness). Only after due diligence investigation of that most likely hypothesis can a possible biological effect (e.g. increased vaccine-associated risk) then be considered. The pre-print posted here does not provide that due diligence check.

An underlying requirement for valid VE estimation by any observational study (including the test-negative design) is comparable exposure risk between vaccinated and unvaccinated participants. However, vaccine passports have permitted broader social mixing by vaccinated compared to unvaccinated people. There is thus good reason to suspect that the vaccinated and unvaccinated are no longer at comparable likelihood of exposure. Higher exposure risk and therefore spuriously increased likelihood of vaccinated individuals contributing to the case series would negatively affect VE estimates due to behavioural rather than biological differences.

Another underlying requirement for valid VE estimation is comparable case ascertainment between vaccinated and unvaccinated participants. The test-negative design standardizes for the likelihood of being tested as an advantage over other observational (e.g. cohort) study designs but the likelihood of being found a case is not the same across multiple different reasons for being tested (i.e. testing indication), which may differ between vaccinated and unvaccinated people. Testing indications with different pre-test likelihoods of being positive include symptomatic illness vs. asymptomatic exposure vs. being part of an outbreak vs. routine pre-travel, workplace or pre-hospital admission screening etc. The recent deployment of rapid antigen testing, followed by confirmatory PCR testing, also affects VE estimates in uncertain ways. The pre-print posted here provides overall VE estimates against any infection in any age group, pooling these multiple testing indications. As such, selection bias remains one of the foremost explanations for their paradoxical findings.

We urge extreme caution before accepting paradoxical negative VE estimates at face value based on any observational study that has not addressed the above methodological issues.

Danuta M Skowronski MD, FRCPC<br /> BC Centre for Disease Control<br /> Vancouver, British Columbia<br /> Canada

and

Gaston De Serres MD, PhD<br /> Institut national de santé publique du Québec<br /> Quebec City, Quebec<br /> Canada

On 2022-01-05 19:42:22, user Christopher Hickie wrote:

Sorry, your paper is not valid. Omicron not dominant in US until week of Christmas so your 2-week sampling ending on Christmas week is way too premature. Please retract your preprint and wait several months to do this analysis.

On 2022-01-08 00:34:01, user AC wrote:

If ~40% of prevalent cases during the “omicron emergent” window are delta, why would severe outcomes be less than 40% what they were in the previous time window? Even if the severe effects were concentrated entirely among delta patients and none from omicron, these rates are still lower than expected. This suggests to me there could be an unidentified confounding factor.

On 2022-01-07 16:35:19, user Toby Koch wrote:

No adjustments for age, vaccination status, and variant for risk of hospitalization and death?

On 2022-01-08 08:27:05, user Menno Schaap wrote:

Significant contribution! Self-tests are promoted for testing oneself before an event or visiting relatives. But in case the subject has no symptoms, reliability is only 22.6%. Therefore perceived feeling of safety is questionable. People should realize that.

On 2022-01-08 18:02:39, user Eithan Galun wrote:

We are waiting for comments and critics,<br /> Eithan Galun in the name of all authors

On 2022-01-10 09:16:10, user Roger Helgesen wrote:

How accurate are self-reported measures in determining actual illness?

Those who know that they have had a positive Covid test might report symptoms more often than those that know they have not had a positive Covid test due to confirmation bias.

On 2022-01-10 09:51:12, user RBNZ wrote:

typo : "Berlina and Bill Gates Foundation"

On 2022-01-13 17:45:28, user T S wrote:

There is a calculation error in Table S8 for the Hispanic noSGTF. The percentage listed in parenthesis is incorrect and appears to be a transposition of the line above it for Black non-hispanic.

Also, I would hate for this article to appear biased upon publication but listing the actual value for increased risk of Omicron infection as compared to Delta for people with prior Covid diagnosis "4.45 (3.24-6.12) fold higher" yet listing a general statement of just "higher" when presenting that same data for prior vaccines leads an astute reader to assume the authors are using framing to present a bias to the data in favor of vaccines. With our top health officials playing politician by misrepresenting studies and dodging questions it is ever more important that actual scientists and peer reviewed studies are above reproach or we risk further deteriation of the rapidly declining public trust in scientists and studies.

"Among cases first ascertained in outpatient settings, adjusted odds of documented prior SARS-CoV-2 infection >=90 days before individuals’ first positive test during the study period were 4.45 (3.24-6.12) fold higher among cases with Omicron variant infections than among cases with Delta variant infections infection (Figure 1; Table 2). <br /> Similarly, adjusted odds of prior receipt of each vaccine series (1, 2, or 3 doses of BNT162b2/mRNA-1973, or Ad.26.COV2.S with or without a booster dose of any vaccine) were higher among cases with Omicron as compared to Delta variant infections."

On 2022-01-16 04:06:03, user FreedomForEvar wrote:

I would like to know how many of these people in the Cohorts had previous infections le Natural immunity The Hospitals can tell and which did these people have Delta or Omicron?<br /> https://www.medrxiv.org/con...<br /> Death rate Los Angeles County Omicron 1 died out of 57,000 that is .0018% <br /> I recommend you take a look at this study that actually includes the Naturally immune START Acknowledging What has been around Since the Very beginning of Human life. <br /> Also <br /> Death rates per WHO<br /> 7 days ending Jan 11 <br /> World wide Death rate is .24% USA same time Death rate is .24% Same period For California Death rate is .09% <br /> 1st week of December USA Death rate is .90% per WHO<br /> Data from England and other countries Show that the Vaccinated are Catching Omicron/Delta virus 80% of the time<br /> it's time to figure out why the vaccine for Covid 19 Is no longer working<br /> What in Delta and what in Omicron Is causing this? What in the Vaccine Is causing this?

On 2022-01-19 19:56:15, user Victor Yman wrote:

This is a preprint of an article published in Nature Communications. The final authenticated version is available online at: https://doi.org/10.1038/s41...

On 2022-01-23 20:43:42, user ELSA GRENMYR wrote:

Did you verify that all patients infected with VOC-Delta and VOC-omicron were SARS-COV2 naive, or could there be a mix of convalescent and naive individuals? How would the infectious viral titres look in a re-infected cohort?

On 2022-02-01 12:17:22, user Daniel Anthony wrote:

While the primary endpoint may have been missed, the shorter illness course and attenuation of loss of smell and taste is very interesting and warrants further investigation. It also suggests that the mode of action is unlikely to be related to the inhibition of TMPRSS2 to block SPKIE activation.

On 2022-02-01 17:13:38, user Ilya Gordeychuk wrote:

A disclaimer. I'm employed at the Chumakov Center, the developer and manufacturer of CoviVac.

First of all, thank you for your work. Clinical description and assessment of cases of symptomatic SARS-CoV-2 infection in vaccinated people during the circulation of emerging virus variants are essential both for the general public and for the healthcare system.

Still, I think some interpretations require further clarification. The first two questions coming in mind after reading the paper are:

1. Did you do any genome sequencing of the virus isolates during your work? It looks now that you assume that those cases were all caused by the delta variant based only on the general epidemiological data saying that delta was predominant in St. Petersburg at the period of observation. If so, the title of the study may be a bit misleading, as it states that those cases were all delta.

2. You state throughout the manuscript that you see significant differences between the effectiveness of the three vaccines, but these interpretations are not supported by the data. <br /> Namely, you state in the abstract that "In contrast to other Russian vaccines, Gam-COVID-Vac is effective against symptomatic SARS-CoV-2 infection caused by Delta VOC", on Page 8 that "CoviVac usefulness is also doubtful" etc. At the same time there is no data in the paper supporting this statement. There is no statistical comparison between the CoviVac group and other vaccine groups. Moreover, I performed a statistical assessment of the data presented in Table A1 and there is no statistically significant difference between the CoviVac group and the Gam-COVID-Vac group, so the data presented in this table directly contradicts your interpretation of the data.

Best regards,<br /> Ilya Gordeychuk

On 2022-02-06 23:31:27, user Danilo Vieira wrote:

I consider that the lack of education in PGx among clinicians makes implementation difficult, but I don't think this barrier is 'extremely' relevant.

On 2022-02-21 11:05:51, user diveoceanos wrote:

Studies 4 through 6 are doing a matched-cohort analysis of Ct values between group 2 (unvaccinated and reinfected) and unvaccinated and infected individuals, individuals with breakthrough infections after BNT16262 vaccine and individuals with breakthrough infection after mRNA-1273 vaccine respectively.

Based on the data the mean Ct value is higher for the unvaccinated and reinfected individuals in all studies compared to the matched-cohort, with studies 4 and 5 reaching statistical significance, while in study 6 the P-value is at 0.104 indicating not a statistically significant difference.

In the text the authors are ranking the infectiousness in order of decreased magnitude in line with their findings i.e.

“The different comparisons suggest an overall hierarchy, present for both asymptomatic and symptomatic infections, where primary infections in unvaccinated persons are most infectious, followed by BNT162b2 breakthrough infections, mRNA-1273 breakthrough infections, and finally reinfections in unvaccinated persons.”

Figure 2 is clearly showing that reinfections are associated with higher Ct compared to all other studied groups.

However there is misleading information on tables 4 and 5. Specifically tables 4 and 5 are showing in the last two rows that infectiousness of breakthrough infections is less compared to infectiousness of reinfections in unvaccinated individuals:

• Infectiousness of BNT162b2-vaccine breakthrough infections relative to reinfections in unvaccinated individuals<br /> • Infectiousness of mRNA-1273-vaccine breakthrough infections relative to reinfections in unvaccinated individuals

Either the line descriptions should change to reflect the correct ratio (i.e. infectiousness of reinfections in unvaccinated individuals over the breakthrough infections or the relative infectiousness should be recalculated to reflect the line description.

On 2022-03-07 04:32:04, user Bidossessi Wilfried Hounkpe wrote:

Preprint status: This study is under review at Experimental Biology and Medicine.

On 2022-04-25 15:00:55, user Nevo Itzhak wrote:

This manuscript has been published in the Journal of Urban Health (DOI: https://doi.org/10.1007/s11524-021-00601-7). <br /> Please cite only the published version!

On 2020-03-29 18:14:17, user Sinai Immunol Review Project wrote:

Main findings:

This study examined<br /> the incidence of diarrhea in patients infected with SARS-CoV-2 across three<br /> recently published cohorts and found that there are statistically significant differences<br /> by Fisher’s exact test. They report that this could be due to subjective<br /> diagnosis criterion for diarrhea or from patients first seeking medical care<br /> from gastroenterologist. In order to minimize nosocomial infections arising<br /> from unsuspected patients with diarrhea and gain comprehensive understanding of<br /> transmission routes for this viral pathogen, they compared the transcriptional<br /> levels of ACE2 of various human tissues from NCBI public database as well as in<br /> small intestine tissue from CD57BL/6 mice using single cell sequencing. They<br /> show that ACE2 expression is not only increased in the human small intestine,<br /> but demonstrate a particular increase in mice enterocytes positioned on the<br /> surface of the intestinal lining exposed to viral pathogens. Given that ACE2 is<br /> the viral receptor for SARS-CoV-2 and also reported to regulate diarrhea, their<br /> data suggests the small intestine as a potential transmission route and<br /> diarrhea as a potentially underestimated symptom in COVID19 patients that must<br /> be carefully monitored. Interestingly, however, they show that ACE2 expression<br /> level is not elevated in human lung tissue.

Limitations of the Study:

Although this study demonstrates a statistical<br /> difference in the incidence of diarrhea across three separate COVID19 patient<br /> cohorts, their conclusions are limited by a small sample size. Specifically,<br /> the p-value computed by Fisher’s exact test is based on a single patient cohort<br /> of only six cases of which 33% are reported to have diarrhea, while the<br /> remaining two larger cohorts with 41 and 99 cases report 3% and 2% diarrhea incidence,<br /> respectively. Despite showing significance, they would need to acquire larger<br /> sample sizes and cohorts to minimize random variability and draw meaningful conclusions.<br /> Furthermore, they do not address why ACE2 expression level is not elevated in<br /> human lung tissue despite it being a major established route of transmission<br /> for SARS-CoV-2. It could be helpful to validate this result by looking at ACE2<br /> expression in mouse lung tissue. Finally, although this study is descriptive<br /> and shows elevated ACE2 expression in small intestinal epithelial cells, it<br /> does not establish a mechanistic link to SARS-CoV-2 infection of the host.<br /> Overall, their claim that infected patients exhibiting diarrhea pose an increased<br /> risk to hospital staff needs to be further substantiated.

Relevance:

This study provides a possible transmission route and a potentially underappreciated<br /> clinical symptom for SARS-CoV-2 for better clinical management and control of<br /> COVID19.

On 2020-04-01 13:40:40, user Sinai Immunol Review Project wrote:

Title: Correlation between universal BCG vaccination policy and reduced morbidity and mortality for COVID-19: an epidemiological study

Keywords: BCG vaccine – epidemiology – vaccination policy

Main findings:The authors compared middle and high income countries that never had a universal BCG vaccination policy (Italy, Lebanon, Nederland, Belgium) and countries with a current policy (low income countries were excluded from the analysis as their number of cases and deaths might be underreported for the moment). Countries that never implement BCG vaccination have a higher mortality rate than countries which have a BCG vaccination policy (16.38 deaths per million people vs 0.78). Next, the authors show that an earlier start of vaccination correlates with a lower number of deaths per million inhabitants. They interpret this as the vaccine protecting a larger fraction of elderly people, which are usually more affected by COVID-19. Moreover, higher number of COVID-19 cases were presented in countries that never implemented a universal BCG vaccination policy.

Limitations:While this study aims to test an intriguing hypothesis unfortunately, the data is not sufficient at this time to accurately make any determinations. Several caveats must be noted including: not all countries are in the same stage of the pandemic, the number of cases/deaths is still changing very rapidly in a lot of countries and thus the association may only reflect exposure to the virus. This analysis would need to be re-evaluated when all the countries are passed the pandemic and more accurate numbers are available. Additionally, very few middle and high-income countries ever implemented universal BCG vaccination, which can be a source of bias (5 countries, vs 55 that have a BCG vaccine policy). Effective screening and social isolation policies also varied considerable across the countries tested and may reflect another important confounder. The authors could consider analyzing the Case Fatality Rate (CFR, % of patients with COVID-19 that die), to more correct for exposure although testing availability will still bias this result. Variability in mortality within countries or cities with variable vaccination and similar exposure could also be appropriate although confounders will still be present.

Relevance:BCG vaccine is a live attenuated strain derived from Mycobacterium bovis and used for a vaccine for tuberculosis (TB). This vaccine has been proven to be efficient in preventing childhood meningitis TB, but doesn’t prevent adult TB as efficiently. For this reason, several countries are now only recommending this vaccine for at-risk population only.<br /> This study shows that there is a correlation between BCG vaccination policy and reduced mortality for Covid-19. Indeed, BCG vaccine has been shown to protect against several viruses and enhance innate immunity1, which could explain why it could protect against SARS-CoV-2 infection, but the exact mechanism is still unknown. Moreover, the efficiency of adult/older people vaccination and protection against Covid-19 still needs to be assessed. Regarding this, Australian researchers are starting a clinical trial of BCG vaccine for healthcare workers2, to assess if it can protect them against Covid-19.

1. Moorlag SJCFM, Arts RJW, van Crevel R, Netea MG. Non-specific effects of BCG vaccine on viral infections. Clinical Microbiology and Infection. 2019;25(12):1473-1478. doi:10.1016/j.cmi.2019.04.020
2. BCG vaccination to Reduce the impact of COVID-19 in Australian healthcare workers following Coronavirus Exposure (BRACE) Trial | Murdoch Children’s Research Institute. https://www.mcri.edu.au/BRACE. Accessed March 31, 2020.

Review by Emma Risson part of a project of students, postdocs and faculty at the Immunology Institute of the Icahn school of medicine, Mount Sinai.

On 2020-04-08 06:20:46, user Vincent Hare wrote:

Logarithmic differences in death rates are also partially explained by the fact that UM and H income countries had the virus first; and there is a clear lag between infections and deaths - of several weeks. On top of this, lower income countries with mandatory BCG programs have also pursued more aggressive lockdowns. Both biases - lag, and lockdown - need to be factored into this analysis BEFORE the effect is tested for statistical significance. Otherwise, the comparison is more or less meaningless.

On 2020-04-08 07:30:20, user Lara wrote:

It does not appear that the authors adjusted for number of tests conducted. There is a significant difference in the number of tests conducted in high income like the US (>2 million) versus LMIC like South Africa (1700 tests). Right now it can't be assumed that BCG is protective, when the full scope of the problem is not unknown, or in other words true case load is presented.

On 2020-06-30 16:50:34, user Rhyothemis wrote:

Was lysophosphatidylinositol (LPI) measured in this study? I don't see it in the text or figures. A study of SARS patients found elevated LPI:<br /> https://pubmed.ncbi.nlm.nih...

If anyone has information on LPI in covid, please post a reply.

On 2020-07-05 14:07:29, user mike marchywka wrote:

I noted my comments on twitter re this paper were piciked up here. You can see it referenced in context here, fwiw, <br /> https://www.linkedin.com/po...

On 2020-07-09 18:37:18, user K dawg wrote:

Nobody cares about the CFR because it is arbitrarily based on testing availability.

What is the Covid IFR? Looks to be around 0.1% from what I've seen... about like influenza.

On 2019-07-09 23:24:44, user Guyguy wrote:

EVOLUTION OF THE EBOLA EPIDEMIC IN THE PROVINCES OF NORTH KIVU AND ITURI

Tuesday, July 9, 2019

The epidemiological situation of the Ebola Virus Disease dated July 8, 2019:

Since the beginning of the epidemic, the cumulative number of cases is 2,428, of which 2,334 are confirmed and 94 are probable. In total, there were 1,641 deaths (1,547 confirmed and 94 probable) and 683 people healed.<br /> 322 suspected cases under investigation;<br /> 10 new confirmed cases, including 8 in Beni, 1 in Vuhovi and 1 in Oicha;<br /> 11 new confirmed case deaths:<br /> 5 community deaths, including 3 Beni, 1 in Vuhovi and 1 in Oicha;<br /> 6 deaths in Ebola Treatment Center including 3 in Beni, 1 in Mabalako, 1 in Butembo and 1 in Katwa.

The cumulative number of confirmed / probable cases among health workers is 128 (5% of all confirmed / probable cases) including 40 deaths.

NEWS<br /> Ebola Virus Disease in Uganda<br /> The Ministry of Health of the Republic of Uganda announced that all index case contacts have completed their mandatory 21-day follow-up period without developing signs of the disease. As a result, Ebola transmission in Kasese District was interrupted. As a reminder, the index case was a 5-year-old boy who had traveled with his mother to the burial of his grandfather who died of Ebola in Aloya, in the health zone of Mabalako.<br /> Uganda has strengthened its border surveillance system. Thus, all travelers coming from the DRC or having traveled to the DRC during the last 21 days must go through the sanitary control at Entebbe airport and at the various road and sea entry points of the country.<br /> Source: Ministry of Health press team on the state of the response to the Ebola epidemic in the Democratic Republic of Congo.

On 2019-07-23 17:47:15, user GuyguyKabundi Tshima wrote:

EVOLUTION OF THE EBOLA EPIDEMIC IN THE PROVINCES OF NORTH KIVU AND ITURI

Monday, July 22, 2019

The epidemiological situation of the Ebola Virus Disease dated 21 July 2019:

Since the beginning of the epidemic, the cumulative number of cases is 2,592, of which 2,498 are confirmed and 94 are probable. In total, there were 1,743 deaths (1,649 confirmed and 94 probable) and 729 people healed.<br /> 272 suspected cases under investigation;<br /> 14 new confirmed cases, including 10 in Beni, 2 in Mandima, 1 in Oicha and 1 in Mutwanga;<br /> 6 new confirmed case deaths:<br /> 3 community deaths including 1 in Beni, 1 in Mandima and 1 in Mutwanga;<br /> 3 deaths at the Ebola Treatment Center of Beni.

Change in coordination of the response to Ebola Virus Disease<br /> A new arrangement of the Presidency of the Republic announced this Saturday, July 20, 2019, the establishment of a technical secretary under the direct supervision of the Head of State to coordinate the response against the Ebola Virus Dpsease in North Kivu and Ituri. This technical secretary is headed by Professor Jean-Jacques Muyembe, who was also chairman of the laboratory committee in coordinating the current response since August 2018.<br /> As a result, all communications related to the response will now be managed directly by the Presidency.

Source: The press team of the Ministry of Health.

On 2020-02-10 08:17:02, user zjuliu wrote:

Firstly I think this should be a milestone for 2019 NCov research because of the work from Academician Zhong and his colleagues. But one thing I need to point out: this paper included patients from Wuhan, Hubei (except Wuhan) and other cities except Hubei, but as we know, it is very different on epidemiological trend, mortality etc when compared with these cohorts. Therefore, I think it is worthy to share this trend on public database so that the scholars can better use these data for further study.

On 2020-02-10 20:08:48, user Marc Bevand wrote:

93.6% of cases (1029 out of 1099) are still in the hospital. Their outcome (death or recovery) is not known yet. This is why the case fatality rate observed (1.4%) so far is low..

For comparison, the two other studies with 41 and 99 cases had only 17% and 58% cases still in the hospital at the time of their writing. More cases had resolved, this is why their case fatality rate was higher (15% and 11%).

On 2020-02-11 06:09:14, user ) Bill Ford wrote:

They do not know how long it remains potent only they have found 24 days. What is the total? No one yet knows

On 2020-02-12 07:04:04, user Marc Bevand wrote:

Table S3 in supplemental data says there are 3665 confirmed patients with 2019-nCoV infections. All the numbers in the table add up to 3665. However the rest of the preprint claims 4021 confirmed patients. What explains this discrepancy?

On 2020-02-13 16:20:21, user Xiaolin Zhu wrote:

We are the authors. We have retrained our model with confirmed cases by Feb. 11. We updated our prediction results. The total infections in mainland China would be 72172 by March 12, 2020 under current trend. It will be 149774 in the worst situation.

On 2020-02-17 09:19:59, user Ellie_K wrote:

Keep this assessment of R0 values when making comparisons to other contagious disease, via the CDC in Atlanta https://wwwnc.cdc.gov/eid/a...

On 2020-02-20 09:21:38, user Linh Ngoc Dinh wrote:

Thanks for sharing your research.<br /> Just a small comment: In an introductory graph, you said "In jurisdictions outside China (and excluding Hong Kong, Macao and Taiwan) the CFR as detailed in the 13 FebruaryWHO Report [3] was 1/447 = 0.22% (95% confidence interval (CI) = 0.40% to 1.26%)."<br /> This is quite misleading statement, WHO has never mentioned an estimate of CFR outside China until now. I think what cite here is the information that there was 1 death and 447 confirmed cases outside China. You should make this point clear, because as a reader, I feel like the number 0.22% (95%CI: .4%-1.26%) is what WHO said. <br /> Also, I wonder how you arrived to that 95%CI as we have only 1 point estimate.

Thanks much!

On 2020-02-21 00:19:53, user Leon Yuan wrote:

The link for Supplementary material is not available, please update it, thanks!

On 2020-02-25 12:11:41, user Igor Nesteruk wrote:

Dear colleagues,

Unfortunately, the coronavirus epidemic in Italy is developing very much like we have seen in mainland China (details in my preprint).

http://dx.doi.org/10.13140/...

Only very strict quarantine and safeguards can stop the spread of the infection throughout Europe.

May be, this information could be useful for your investigations.

I have found today the accumulated number of cases in Italy – 229 - on the official site of Italian Health Ministry.

http://www.salute.gov.it/po...

This point was already in the figure from

http://dx.doi.org/10.13140/...

We need the correct and reliable information about the accumulated number of cases. Do you have any links?

Be careful and healthy!

Sincerely yours,

Igor Nesteruk,

On 2020-02-26 19:26:13, user ricwerme wrote:

Did I miss the exported case counts the paper used to determine the internal # of cases, or is it just "three" with assumed missed cases?

The abstract says "suggesting a underlying burden of disease in that country than is indicated by reported cases." Should that be "a greater underlying..."?

On 2020-02-28 02:55:16, user Art Enquirer wrote:

Will it be possible for an AI startup (actually a hackathon team) to secure access to your data? Will you share for the sake of the world? We are also running servers with AI algorithms and wish to pre-test your conclusion.

On 2020-02-28 19:16:00, user Antoine Jomier wrote:

Hello, i am running an ai algorithm start up company. Would it be possible to share your model or data set so that we distribue it in France. We would not do any commercial exploitation but make it available widely to the community. My contact antoine.jomier@incepto-medical.com<br /> Thanks

On 2020-02-29 18:15:11, user Per Carlbring wrote:

For reference #21 a correction was published in 2016: https://journals.plos.org/p...

On 2020-03-05 09:13:52, user Jørgen K. Kanters wrote:

Important paper but needs to be improved to be a High flier. First around 50 % had a hypertensive history. In an American population that would mean hypertension is protective. You need an age gender matched control population from the same area to compare with. Furthermore you miss a very important point. Which medicine prescriptions had the patients before admission? ACE Inhibitors and A2 antagonists as the most interesting. Again compared to a control population

On 2020-03-05 16:05:56, user Erik Kulstad wrote:

Thank you for this data. You mention that you excluded patients with mild symptoms who had been transferred to mobile cabin hospitals (as well as patients who had been transferred to other hospitals for advanced life support), but were any of the patients with mild symptoms then re-admitted, to then become patients that were included in your 109 total (or are you able to track)?

On 2020-03-06 15:54:48, user Steven Ge wrote:

Please let us me know if you have any questions or suggestions. Twitter @StevenXGe

On 2020-03-13 11:54:03, user Murat Von Marit wrote:

For what temperature is the time you set?<br /> 4C? 20C? 40?<br /> And what about Textile,Fabric,Clothes?

On 2020-03-14 08:04:28, user Stefano Gaburro wrote:

Minimal viral titer for infection: thanks for this great piece of work that allows governmental bodies to give suggestion. One question: the viral titer decreases over time meaning it could be detected but no longer infectious. Have you determined the minimal titer of virus to determine an infection?

On 2020-03-14 19:27:30, user Halmartin Brown wrote:

Covid-19 Question: What is the risk of the virus being transmitted on paper and mail in general and packages? Are mail and package carriers being tested and are they using gloves? I read it can survive up to a day on cardboard and cash doesn't allow viruses to survive as long.

On 2020-03-15 18:34:58, user tusitw wrote:

Are you also going to study as a function of humidity?<br /> Below LOD, do we know it is still capable of infecting? a question in the same theme as Stefano Gaburro...

On 2020-03-25 17:42:51, user Rudolf Brüggemann wrote:

It is a bit irritating that version1 and version2 give different values for the half-lifes. Is there an error based on a factor ln10 somewhere? The half-lifes in Table 1 of the supplement of the published version are much smaller than those in Table 1 in the preprint version. E.g., half-life (median) for steel 13.1 hours in the preprint, median of 5.63 hours in Table 1 of the Supplement of the published version.

On 2020-03-17 23:37:34, user RunningThrough wrote:

Given the study cohort of patients are all hospital admitted patients there in Wuhan, presumably are all in the 'severe' and 'critical' category of all COVID-19 patients per admission policies that we read, so does this present data suggests that the SARS-CoV-2 virus has a higher infectivity amongst blood Gp A patients or that blood Gp A patients are more likely to develop a more severe disease?

On 2020-03-20 17:02:20, user Kevin Hamill wrote:

An editing suggestion:<br /> This manuscript will be read by new media/journalists therefore I would encourage more careful use of the term "significant".

For example where it says "blood group A had a significantly higher risk for COVID-19 " if that was quoted then people would hear this as "blood group A had a much higher risk for COVID-19."

In the same sentence, I would write:<br /> blood group A had approximately 20% higher risk for COVID-19 (odds ratio-OR, 1.20; 95% confidence interval-CI 1.02~1.43, P = 0.02). <br /> [and equivalent changes with the other phrases].

Note that as you have already stated the p values, using the word significant has no added value, it only provides a source for ambiguity.

On 2020-03-26 07:52:06, user M.E.Valentijn wrote:

Has anyone been able to verify their source claiming 33% prevalence of Type O in the general population? I can't find the journal that's cited for that, and a newer article says 30.2% for Han Chinese, not the nearly 34% claimed here. Though that's Han Chinese in general, not just in Wuhan. Can't find their other sources for normal blood types in the area either.

On 2020-04-15 18:32:13, user Jaime Navarro wrote:

There is a significant flaw in this paper's claim that Type A blood types are more susceptible to CoViD-19, and type O are less. In that the paper does not address the susceptibility of those with type B or AB blood. If as the paper suggests type O blood sees the virus as a type A antigen and so attacks the virus. Shouldn't the same happen in patients with Type B or AB? After all they would have antibodies to type A the same as type O people would.

On 2020-03-20 20:57:29, user Sylvie Vullioud wrote:

Could authors provide information to dissipate high risks of bias:

1. Manuscript was first published on mediterranee-infection.com website, not on medRxiv. On the manuscript on the website on mediterranee-infection.com, I can read 'In Press 17 March 2020 – DOI : 10.1016/j.ijantimicag.2020.105949'. It means that manuscript was already accepted by International Journal of Antimicrobial Agents at the time when the manuscript was deposit on the 20.03.2020 on medRxiv.

-> Pre-print on medRxiv is not a real pre-print to collect feed-back for manuscript improvement, as originally designed for. Moreover, medRxiv states: 'All preprints posted to medRxiv are accompanied by a prominent statement that the content has not been certified by peer review'.

-> There is an obvious potential conflict of interest, because last author Raoult is editor of the article collection COVID-19 Therapeutic and Prevention in International Journal of Antimicrobial Agents.

-> International Journal of Antimicrobial Agents is runned by Elsevier, suggesting 'If accepted for publication, we encourage authors to link from the preprint to their formal publication via its Digital Object Identifier (DOI)'.

1. Discussion on the controversy of main cited Chinese paper, ref 8 ?

2. According to paper, allocation of patients group was random but treated group is 51.2 years average and control group 37.3 years?

3. Article describes 3 conditions of patients: asymptomatic, low and high symptoms. Why?

4. Care to patients, biological and physiological sampling and analyses, and statistical analyses were not blinded. Why?

5. I think that no placebo was used. Why?

6. 6 patients on total of 42 were excluded from study: three patients were transferred to intensive care unit, 1 stopped because of nausea, 1 died. One left hospital. <br /> It is written :'study results presented here are therefore those of 36 patients (20 hydroxychloroquine-treated patients and 16 control patients). Why were dead, intensive care, and nausea patients not included in statistical treatment? <br /> -> This may be a selection bias? <br /> -> What about unwanted very worrying effects of the treatment?

7. 'The protocol, appendices and any other relevant documentation were submitted to the French National Agency for Drug Safety (ANSM) (2020-000890-25) and to the French Ethic Committee (CPP Ile de France) (20.02.28.99113) for reviewing and approved on 5th and 6th March, 2020, respectively'. Pre-print was posted on 20.03.2020. Time points on day 14 on patients.<br /> -> So recruitment and study started before approval of ANSM and French Ethic Committee? How is it possible?

8. How is it plausible that numerous authors (18!) participated equally to the work? Is it possible to add their respective contributions?

Thank you in advance for considering my questions. <br /> Regards, <br /> Sylvie Vullioud

On 2020-03-22 04:52:08, user Juan B. Gutierrez wrote:

In summary, provided that our Ro is correct, and we are certain it is, as we reused very long results from our recent peer-reviewed result, https://doi.org/10.1007/s11... Bulletin of Mathematical Biology (the premier venue for the discipline), then with the information that we have today, Ro cannot be close to 3.

By a suggestion of Dr. Jeremy Faust, MD, Brigham and Women's Hospital, @jeremyfaust, I modified the most uncertain parameters to produce an Ro of 3.These parameters are the mean infectious periods for symptomatic (lambda_yr) and asymptomatic (lambda_ar) subjects. If we consider the median of the other parameters to be correct (there is more data), then the mean infectious period of a symptomatic patient should be 4.9 days, and the mean infectious period of an asymptomatic should be 4.1 days. These numbers do not match what is happening on the ground. If we reduce alpha, the probability of becoming asymptomatic upon infection, to something less than 0.86, e.g. alpha = 0.5, then the mean infectious period of a symptomatic patient should be 3.7 days, and the mean infectious period of an asymptomatic should be 3.1 days.

The reality is that patients are infectious before the onset of symptoms, and the disease lasts more than 3 days in symptomatic patients. The necessary conclusion is that via a computational reductio ad absurdum, and with the information we have today, Ro cannot be close to 3.

On 2020-03-23 21:39:01, user Shayan wrote:

wondering what the 4000+ test results refers to with there only being 28 patients? looking at the distribution plots, there seem to be more than 28 data points per biomarker

On 2020-03-24 14:55:05, user syed aleem wrote:

Excellent paper! Any particular reason to express RBD as a secreted protein?

On 2020-03-25 22:43:03, user Sinai Immunol Review Project wrote:

Title: A serological assay to detect SARS-Cov-2 seroconversion in humans

Immunology keywords: specific serological assay - ELISA - seroconversion - antibody titers

Note: the authors of this review work in the same institution as the authors of the study<br /> Main findings: <br /> Production of recombinant whole Spike (S) protein and the smaller Receptor Binding Domain (RBD) based on the sequence of Wuhan-Hu-1 SARS-CoV-2 isolate. The S protein was modified to allow trimerization and increase stability. The authors compared the antibody reactivity of 59 banked human serum samples (non-exposed) and 3 serum samples from confirmed SARS-CoV-2 infected patients. All Covid-19 patient sera reacted to the S protein and RBD domain compared to the control sera.<br /> The authors also characterized the antibody isotypes from the Covid-19 patients, and observed stronger IgG3 response than IgG1. IgM and IgA responses were also prevalent.

Limitations of the study:The authors analyzed a total of 59 control human serum samples, and samples from only three different patients to test for reactivity against the RBD domain and full-length spike protein. It will be important to follow up with a larger number of patient samples to confirm the data obtained. Future studies will be required to assess how long after infection this assay allow to detect anti-CoV2 antibodies. Finally, while likely, the association of seroconversion with protective immunity against SARS-Cov-2 infection still needs to be fully established.

Relevance: <br /> This study has strong implications in the research against SARS-Cov-2. First, it is now possible to perform serosurveys and determine who has been infected, allowing a more accurate estimate of infection prevalence and death rate. Second, if it is confirmed that re-infection does not happen (or is rare), this assay can be used as a tool to screen healthcare workers and prioritize immune ones to work with infected patients. Third, potential convalescent plasma donors can now be screened to help treating currently infected patient. Finally, the recombinant proteins described in this study represent new tools that can be used for further applications, including vaccine development.

Review part of a project by students, postdocs and faculty at the Immunology Institut of the Icahn School of Medicine, Mount Sinai.

On 2020-03-24 18:03:39, user Sinai Immunol Review Project wrote:

This study is a cross-sectional analysis of 100 patients with COVID-19 pneumonia, divided into mild (n = 34), severe (n = 34), and critical (n = 32) disease status based on clinical definitions. The criteria used to define disease severity are as follows:

1. Severe – any of the following: respiratory distress or respiratory rate >= 30 respirations/minute; oxygen saturation <= 93% at rest; oxygen partial pressure (PaO2)/oxygen concentration (FiO2) in arterial blood <= 300mmHg, progression of disease on imaging to >50% lung involvement in the short term.

2. Critical – any of the following: respiratory failure that requires mechanical ventilation; shock; other organ failure that requires treatment in the ICU.

3. Patients with pneumonia who test positive for COVID-19 who do not have the symptoms delineated above are considered mild.

Peripheral blood inflammatory markers were correlated to disease status. Disease severity was significantly associated with levels of IL-2R, IL-6, IL-8, IL-10, TNF-?, CRP, ferroprotein, and procalcitonin. Total WBC count, lymphocyte count, neutrophil count, and eosinophil count were also significantly correlated with disease status. Since this is a retrospective, cross-sectional study of clinical laboratory values, these data may be extrapolated for clinical decision making, but without studies of underlying cellular causes of these changes this study does not contribute to a deeper understanding of SARS-CoV-2 interactions with the immune system.

It is also notable that the mean age of patients in the mild group was significantly different from the mean ages of patients designated as severe or critical (p < 0.001). The mean patient age was not significantly different between the severe and critical groups. However, IL-6, IL-8, procalcitonin (Table 2), CRP, ferroprotein (Figure 3A, 3B), WBC count, and neutrophil count (Figure 4A, 4B) were all significantly elevated in the critical group compared to severe. These data suggest underlying differences in COVID-19 progression that is unrelated to age.

Given the inflammatory profile outlined in this study, patients who have mild or severe COVID-19 pneumonia, who also have any elevations in the inflammatory biomarkers listed above, should be closely monitored for potential progression to critical status.

On 2020-03-26 16:57:03, user S Weeth wrote:

How about soap? Any testing on with soap?

On 2020-03-27 15:01:22, user Sinai Immunol Review Project wrote:

These authors looked at 17 hospitalized patients with COVID-19 confirmed by RT-PCR in Dazhou, Sichuan. Patients were admitted between January 22 and February 10 and the final data were collected on February 11. Of the 17 patients, 12 remained hospitalized while 5 were discharged after meeting national standards. The authors observed no differences based on the sex of the patients but found that the discharged patients were younger in age (p = 0.026) and had higher lymphocyte counts (p = 0.005) and monocyte counts (p = 0.019) upon admission.

This study is limited in the sample size of the study and the last data collection point was only one day after some of the patients were admitted.

These findings have been somewhat supported by subsequent studies that show that older age and an immunocompromised state are more likely to result in a more severe clinical course with COVID-19. However, other studies have been published that report on larger numbers of cases.

On 2020-03-27 20:04:02, user Sinai Immunol Review Project wrote:

The authors present a digital PCR (dPCR) diagnostic test for SARS-CoV-2 infection. In 103 individuals that were confirmed in a follow-up to be infected, the standard qPCR test had a positivity rate of 28.2% while the dPCR test detected 87.4% of the infections by detecting an additional 61 positive cases. The authors also tested samples from close contacts (early in infection stage) and convalescing individuals (late in infection stage) and were able to detect SARS-CoV-2 nucleic acid in many more samples using dPCR compared to qPCR.

The authors make a strong case for the need for a highly sensitive and accurate confirmatory method for diagnosing COVID-19 during this outbreak and present a potential addition to the diagnostic arsenal. They propose a dPCR test that they present has a dramatically lower false negative rate than the standard RT-qPCR tests and can be especially beneficial in people with low viral load, whether they are in the earlier or later stages of infection.

On 2020-03-28 18:07:46, user Ian Timaeus wrote:

I may be being very stupid, but isn't the ACFR formula given in the preprint wrong? Aren't you simply averaging the age-specific CFRs? So don't you want to multiply their sum by n/100, i.e. divide by the number of age intervals, not multiply by the width of those intervals? As an alternative, you could standardise the age-specific CFRs on the age-sex distribution of Italy, rather than on a uniform age distribution, so that the adjusted CFR equated to the CFR if incidence were constant by age and sex.

On 2020-03-30 15:27:50, user Sinai Immunol Review Project wrote:

Summary and key findings: Summary of clinical trials registered as of March7, 2020 from U.S, Chinese, Korean, Iranian and European registries. Out of the 353 studies identified, 115 were selected for data extraction. 80% of the trials were randomized with parallel assignment and the median number of planned inclusions was 63 (IRQ, 36-120). Most frequent therapies in the trials included; 1) antiviral drugs [lopinavir/ritonavir (n-15); umifenovir (n=9); favipiravir (n=7); redmesivir (n=5)]; 2) anti-malaria drugs [chloroquine (n-11); hydroxychloroquine (n=7)}; immunosuppressant drugs [methylprednisolone (n=5)]; and stem cell therapies (n=23). Medians of the total number of planned inclusions per trial for these therapies were also included. Stem cells and lopunavir/ritonavir were the most frequently evaluated candidate therapies (23 and 15 trials respectively), whereas remdesivir was only tested in 5 trials but these trials had the highest median number of planned inclusions per trial (400, IQR 394-453). Most of the agents used in the different trials were chosen based on preclinical assessments of antiviral activity against SARS CoV and MERS Cov corona viruses.

The primary outcomes of the studies were clinical (66%); virological (23%); radiological (8%); or immunological (3%). The trials were classified as those that included patients with severe disease only; trials that included patients with moderate disease; and trials that included patients with severe or moderate disease.

Limitations: The trials evaluated provided incomplete information: 23% of these were phase IV trials but the bulk of the trials (54%) did not describe the phase of the study. Only 52% of the trials (n=60) reported treatment dose and only 34% (n=39) reported the duration. A lot of the trials included a small number of patients and the trials are still ongoing, therefore no insight was provided on the outcome of the trials.

Significance: Nonetheless, this review serves as framework for identifying COVID-19 related trials, which can be expanded upon as new trials begin at an accelerated rate as the disease spreads around the world.

On 2020-03-31 19:02:12, user earonesty wrote:

It's an immunomodulator, it prevents some of the inflammation issues associated with COVID-19. Not a surprising result. There may be better ones, but since this is used to treat asthma, and other issues pulmonary inflammation, it's a good choice.

On 2020-04-07 23:44:19, user Ronaldo Wieselberg wrote:

I see a lot of problems with this study.

First of all, it does not mention whether risk factors - such as hypertension or diabetes - were taken into account. The table provided for randomization doesn't show them either. Having a difference about risk factors could play a huge part in the difference - let's consider that control group had more people with pre-existing conditions, for instance, thus, the risk of evolving to a severe disease would be improved, independently of HCQ. Moreover, there is no mention of whether the four individuals who progressed to severe disease had any pre-existing conditions, needed mechanical ventilation or any other details of the "severity" - could it be a SpO2 of 92% only, accordingly to inclusion criteria.

The paper does not describes clearly the evaluation criteria. How was measured the cough? Was it dicotomic (have cough/don't have any cough)? How could you determine whether the pneumonia "improved"? Was it accordingly to the presence/absence of infiltrates on X-Rays, or % of lungs compromised in CT-scan? Thus, calculating a significative p value for subjective criteria is really, really a difficult point.

It does not states, as well, the symptoms duration prior to admission and start of the intervention. People who had, for instance, 10 days of symptoms before looking for a physician could have fewer days of symptoms than another individual who looked for medical assistance with two days of symptoms - and there is no mention about this.

On 2020-03-31 22:51:51, user Ruth Etzioni wrote:

When will the actual description of the model be available?

On 2020-03-31 22:59:25, user Whiskers wrote:

Even more worrying if it is air spread, we have been led to believe that it is only really contact spread unless someone coughs directly over you.<br /> Perhaps this accounts for the prolific spread of this disease.

On 2020-04-01 14:34:11, user Sinai Immunol Review Project wrote:

Summary: ?Retrospective study on 97 COVID-19 hospitalized patients (25 severe and 72 non-severe) analyzing clinical and laboratory parameter to predict transition from mild to severe disease based on more accessible indicators (such as fasting blood glucose, serum protein or blood lipid) than inflammatory indicators. In accordance with other studies, age and hypertension were risk factors for disease severity, and lymphopenia and increased IL-6 was observed in severe patients. The authors show that fasting blood glucose (FBG) was altered and patients with severe disease were often hyperglycemic. Data presented support that hypoproteinaemia, hypoalbuminemia, and reduction in high-densitylipoprotein (HDL-C) and ApoA1 were associated with disease severity. ?

Limitations: ?In this study non-severe patients were divided in two groups based on average course of the disease: mild group1 (14 days, n=28) and mild group 2 (30 days, n=44). However mild patients with a longer disease course did not show an intermediate phenotype (between mild patients with shorter disease course and severe patients), hence it is unclear whether this was a useful and how it impacted the analysis. Furthermore, the non-exclusion of co-morbidity factors in the analysis may bias the results (e.g. diabetic patients and glucose tests) It is not clear at what point in time the laboratory parameters are sampled. In table 3, it would have been interesting to explore a multivariate multiple regression. The correlation lacks of positive control to assess the specificity of the correlation to the disease vs. correlation in any inflammatory case. The dynamic study assessing the predictability of the laboratory parameter is limited to 2 patients. Hence there are several associations with disease severity, but larger studies are necessary to test the independent predictive value of these potential biomarkers.?

Findings implications:? As hospital are getting overwhelmed a set of easily accessible laboratory indicators (such as serum total protein) would potentially provide a triage methodology between potentially severe cases and mild ones. This paper also opens the question regarding metabolic deregulation and COVID-19 severity.

On 2020-04-03 05:01:21, user Jacob G Scott wrote:

Please find our update, with HIGHER recommended exposure times for porous PPE, on our github repo: https://github.com/TheoryDi...

We expect another update in the coming days with filtration/fit testing results at these exposures, as well as biologic validation.

Please also see recent CDC guidelines: https://www.cdc.gov/coronav...

and a cooperative groups recommendation for N95 decontamination: https://www.n95decon.org/

Please stay safe and healthy.

On 2020-04-17 19:08:16, user Jacob G Scott wrote:

See github repo for live updating and opportunities to contribute data: https://github.com/TheoryDi...

On 2020-04-04 19:53:38, user VirusWar wrote:

It is interesting study, but I'm surprised you don't talk about level of Potassium in the blood ? Did you check it ? Was there any Magnesium given to correct level of Potassium. Especially, you noted heart troubles with people having renal disease, but it is well known they have excess of Potassium which creates such heart troubles. Also, was treatment H+A stopped when QTc >=500ms ?

On 2020-04-07 19:27:46, user Archisman Mazumder wrote:

Indian study showing COVID-19 affects the 20-39 yrs age group most in India. This really has to be studied further. Even the Health Ministry of India corroborated the findings.

On 2020-04-15 18:26:55, user Archisman Mazumder wrote:

These predictions actually matched.

On 2020-04-12 05:17:40, user John Roberts wrote:

Cytodyn’s drug Leronlimab is proving effective in treating the cytodyne storm and getting severe Covid patients off of ventilators.

On 2020-04-12 08:33:12, user tsuyomiyakawa wrote:

Thanks, everyone, for your precious comments.

1. We are examining the potential confounders, which includes the ones mentioned here.

2. As Rosemary mentioned, BCG is an attenuated version TB and, indeed, big protective effect of TB prevalence against COVID-19 exists. We will incorporate the data in the next version.

3. We obtained the data from the web site of European Centre for Disease Prevention and Control, and are re-analyzing the growth of spreading in a more quantitive manner. Basically, there are significant effects of BCG/TB against COVID-19 growth, which will replace the data shown in Figure 3.

4. Regarding the tourists from China, according to a survey, the top 10 destination countries of China’s out bound countries are Japan, Thailand, South Korea, Indonesia, Singapore, Malaysia, Australia, UK, New Zealand, and Maldives, and 9 out of 10 of them are the ones with extremely low COVID-19 cases and deaths (4 or lower deaths per million) , as of April 13th, which makes it unlikely that the Traveling activity from China matters. This will be added to the discussion. Also, we evaluated the number of international arrivals in each country and it did not essentially affect the results (almost at all).

5. As for masks and green tea, they cannot explain 1) the differences between Eastern Europe and Western Europe and 2) low COVID-19 indices in Africa, South America and South East Asia. We may consider their potential effect, once we can get any good statistics representing those things, but so far, we set priority low for these potential confounders.

Anyway, we will upload next version sometime in next week and it will be appreciated if you could keep providing us critical comments, which will greatly improve our manuscript. Thank you!

On 2020-04-12 13:23:51, user Joe Gitchell wrote:

Thank you to the authors for taking the time and effort to report these findings in the midst of confronting the challenges from managing the COVID19 pandemic. Please stay safe!

And it is with humility that I make a request to them to do two things with their data on tobacco use. The first I think should be pretty straightforward, the second will depend on the specificity available within the Epic records:

1) In Tables 1 and 2, can you please break out "Never" and "Unknown" in to separate categories; and

2) In these tables and in other analyses, can you please break out "tobacco use" at least in to combustible (cigarettes, cigars, cigarillos, hookah, etc) and noncombustible (smokeless tobacco, snus, vaping) categories?

Thank you. I also found your use of CART/decision-tree analysis really helpful, btw.

Joe

Disclosures:<br /> My employer, PinneyAssociates, provides consulting services on tobacco harm minimization on an exclusive basis to JUUL Labs, Inc., a manufacturer of nicotine vaping products. I also own an interest in an improved nicotine gum that has neither been developed nor commercialized.

On 2020-04-13 00:38:02, user Craig wrote:

The safety of HCQ alone has already been proven. It's the efficacy, both as prophylaxis and as treatment, that needs to be studied.

Studying HCQ in combination with a drug that is already known to have adverse cardiac effects seems like a study designed to produce data with a negative bias.

On 2020-04-13 13:45:53, user Rosemary TATE wrote:

Hi, I dont see the STROBE (for observational studies) guidelines checklist uploaded, although you ticked yes to this<br /> "I have followed all appropriate research reporting guidelines and uploaded the relevant EQUATOR Network research reporting checklist(s) and other pertinent material as supplementary files, if applicable. "<br /> A lot of people seem to ignore these but they are important and any good journal will require them.<br /> Can you please upload? Many thanks.

On 2020-04-14 08:27:08, user Lisa Kane wrote:

Can the authors comment on the role air conditioning and/or building/residence heating may have played in the cooler cities?

That is, are cooler, measured city temperatures actually proxies for warm, indoor temperatures?

On 2020-04-14 14:03:10, user Chris Pericone wrote:

The death data in table 3 are the reverse of what is described in the results section. I assume the high and low dose columns were mistakenly switched in that table?

On 2020-04-14 18:23:20, user Phyllis Bergiel wrote:

Here's a dissociation statement form Rockelfeller: https://www.rockefeller.edu...

On 2020-04-15 14:34:21, user Bio wrote:

I have several issues with this study:

1. I find not including time as a factor in the model bewildering. After all, time is the single most important factor for the number of cases for most of the countries in the model. The model is only log(cases) ~ population + temperature. But for example, in half a month's time, population and temperature won't change much, while the number of cases could increase several fold for some countries. Time is a critical factor to model and is more important than temperature and population. Not including time, the model in the paper cannot be stable. Basically as time changes, your conclusions likely will change.

2. Many other important factors were not considered. For example, at what point of COVID-19 growth is each country at? If one compares March 14 to March 27, China's numbers are not much different while USA and many other countries have quite different numbers on those 2 days. The model cannot be stable due to this as well (time + point during growth). Also what about containment policy causing slower growth? Effects from such important confounding factors were not considered in the model.

3. There are many other smaller issues such as USA cases were mostly in Northeast, with latitude clearly higher than the one used to represent USA. In China, the cases happened in many provinces with vastly different latitude/temperature but all cases counted at one latitude/temperature. Moreover, the vast majority of the cases happened in China during Jan/Feb while you used late March temperature to represent them. Inaccuracies like these seriously impact modeling latitude/temperature as a continuous variable. Excluding countries with small population but high case rate as outliers is also questionable, given that you modeled population size already.

Fig. 7 could benefit from being plotted also for 2/29, 3/14, 3/21. Interpretation of Fig. 7 could instead be that it showed 2 groups of countries/latitudes that reflects the temporal sequence of events rather than temperature: COVID-19 started in China, spread to South Korea and Italy, then Europe and America as they trade/travel often and happen to all be cold countries; then in March COVID-19 picked up in southern hemisphere and tropical area and still going. It's likely time and policies that helped Australia case rate be relatively low instead of temperature, because the spread of COVID-19 is still early there and they learned from other countries to control the spread from early on.

Therefore I do not believe the paper provided convincing evidence for temperature-dependency of COVID-19.

On 2020-04-15 18:51:04, user Greg Lambert wrote:

The study seems to incorrectly use a UVAB meter instead of a UVC meter to measure the exposure from their UVC lamp, consequently their UV exposure readings are wrong(low).

From the study:

"Ultraviolet light. Plates with fabric and steel discs were placed under an LED high power UV germicidal lamp (effective UV wavelength 260-285nm) without the titanium mesh plate (LEDi2, Houston, Tx) 50 cm from the UV source. At 50 cm the UVAB power was measured at 5 u W/cm2 using a General UVAB digital light meter (General Tools and Instruments New York, NY)."

Their lamp emits effective UV wavelength of 260-285nm but a General UVAB meter only measures from 280 to 400 nm with a calibration point of 365nm.

A

On 2020-04-16 01:47:35, user 777Rampage wrote:

I have the following questions and issues on their testing of UVC LED.<br /> 1. In the article it states that General Tools UVAB digital light meter was used. Is that the UV513AB meter? If so, that monitoring probe is only able to measure UVA&B, not C. <br /> 2. If using General Tools UV512C meter, that probe's spectral range is 220 to 275 and it is used to measure low pressure mercury UVC lamp, not UVC LED.

On 2020-04-15 20:08:45, user davidmeiser wrote:

No Evidence of Rapid Antiviral Clearance or Clinical Benefit with the Combination of Hydroxychloroquine and Azithromycin in Patients with Severe COVID-19 Infection https://www.sciencedirect.c...

On 2020-04-15 23:59:04, user Mark .Minnery wrote:

The average time between symptom onset and randomisation was 16.6 days. Could the authors discuss the implications of this potential confounder. Was the long time before treatment because of delay between symptoms and presentation at hospital?

On 2020-04-16 09:05:11, user Stef Verlinden wrote:

Please study the paper thorougly before jumping to conclusions. Standard of care means that (most of the) patients were also treated with lopinavir-ritonavir, arbidol, oseltamivir, virazole, entecavir, ganciclovir and/or interferon-alpha.

The autors also did a post-hoc subanalyses with patients who did not get other medications other than HCQ in the treatment group or nothing in the SOC group. And, for what it is worth, here they found ‘a significant efficacy of HCQ on alleviating symptoms’.

What also could be of interest is that patients treated with HCQ showed a significantly higher reduction of CRP. One of the proposed MOA for HCQ is an anti-inflammatory one.

All in all a very weak study from wich not much can be concluded

On 2020-04-15 22:33:38, user suradip das wrote:

Very interesting work. I have some questions -<br /> 1. Page 15 (Table 1): C-Reactive Protein is an indicator of cardiovascular disease. It is interesting that the authors chose to conduct the study in a population where 86% of all the patients had high CRP.<br /> 2. Out of the 84 patients receiving HCQ (90.5% having CRP>40mg/l and 45.2% having cardiovascular diseases) only 3 patients died (3.6%). In comparison, the group which did not receive HCQ but had similar weight proportions of high CRP and CVD saw 4.1% mortality.<br /> To summarize, there appears to be no significant difference in mortality rates when patients with CVD and COVID-19 are treated with HCQ versus a placebo.

On 2020-04-16 11:23:18, user Rajesh Ranjan wrote:

Please read the abstract in the original paper (PDF). There is a slight misprint in the web content.

On 2020-04-16 21:55:50, user Sinai Immunol Review Project wrote:

Title:<br /> Immunopathological characteristics of coronavirus disease 2019 cases in Guangzhou, China<br /> The main finding of the article: <br /> This study analyzed immune cell populations and multiple cytokines in 31 patients with mild/moderate COVID-19 (ave. 44.5 years) and 25 with severe COVID-19 (ave. 66 years). Samples from patients with fever and negative for the SARS-COV-2 test were used as control. At inpatient admission, total lymphocytes number was decreased in severe patients but not in mild patients, whereas neutrophils were increased in severe patients. CD4+ and CD8+ T cells were diminished in all COVID-19 patients. CD19+ B cells and NK cells were decreased in both mild and severe patients, however, severe patients showed a notable reduction. These data might suggest a profound deregulation of lymphocytes in COVID-19 patients. Further analysis showed significant increases of IL-2, IL-6, IL-10 and TNF? in blood of severe patients at the admission. Sequential samples revealed that IL-2 and IL-6 peaked on day 15-20 and declined thereafter. A moderate increase of IL-4 was seen in mild/moderate patients. Thus, elevation of IL-2, IL-6 can be indicators of severe COVID-19.<br /> Critical analysis of the study: <br /> There is no information on when the patients were assessed as severe or mild/moderate, at inpatient admission or later. The authors could have analyzed the correlation between immune cell population and cytokine levels to see, for example, if severe lymphopenia correlated to higher elevation of IL-2.<br /> The importance and implications for the current epidemics:<br /> While similar findings have already been shown, the data corroborates alterations in circulating adaptive and innate immune cell populations and cytokines, and its correlation to disease severity. The increase of IL-2 and IL-6 at the admission might an indicator to start intensive therapies (like convalescent serum) at an early time.

On 2020-04-16 23:52:26, user Geoff wrote:

Why were they bronching these patients?? Sounds incredibly dangerous.

On 2020-04-18 19:50:29, user Oliver Van Oekelen wrote:

"Raw data will be available in GEO."

When will the data be uploaded? This preprint was posted almost two months ago. Would be amazing to fuel collaborative efforts across the globe and increases the impact of this work...!<br /> Thanks

On 2020-04-17 18:50:31, user thecity2 wrote:

The sample could be biased by a self-selection effect. People on FB who wanted to be tested because they think they had the virus at some point.

On 2020-04-17 20:13:28, user AM wrote:

thanks for the information. you did not specify if you found the IgM or the IgG antibody which would allow us to know what stage they are in on the time frame of when someone sero-converts. it would be great to conduct the same study now, or even 2 weeks from now to see the changes and the transition to herd immunity. ideally, you have kept track of all the people you tested. we could derive a wide range of conclusions from this one time test. nonetheless, thanks for taking the time to conduct these tests.

https://uploads.disquscdn.c...

On 2020-04-17 22:27:44, user Julie Larsen Wyss wrote:

Have the participants been informed if they were positive yet?

On 2020-04-18 02:30:01, user Don Phan wrote:

This does not appear to be random sampling. If you are using volunteers, then the sampling is not representative of Santa Clara country. What have you done to correct this? And I am not talking about demographics. The people who suspected that they had the virus would be the most likely to risk the shelter at home, drive to location, and participate.

On 2020-04-18 14:03:13, user Richard Davis wrote:

Could you guys make the data and code available for other researchers to try and replicate, now that the paper is out there?

On 2020-04-18 15:48:14, user Jingeol Lee wrote:

I hope this method extends to general application on forecasting the spread!

On 2020-04-18 20:15:11, user Marya Lieberman wrote:

table 1 lists one entry for NPV as 32/25 and gives a value of 91%. Looks like a transcription or math error.

On 2020-04-18 20:25:12, user Scott Howell wrote:

Conceptually this study and its conclusions are patently false. Internal endogenous hormone levels do not equate to exogenously administered androgens. Perhaps a read of Morgentaler's saturation model would enlighten the authors. A statement that long-term elevated free testosterone levels causes prostate cancer does not align to either the saturation model or the use of bipolar androgen therapy at supraphysiologic doses to treat prostate cancer. There is a big jump from Mendelian randomization to nuances in physiology and what occurs in practice. Just like secondary data analysis of insurance claims to establish risk should be banned or at least relegated to their limitations, these type of studies drawing conclusions outside of the scope of the data should be relegated to their limitations or even less.

On 2020-04-21 08:35:42, user Shrisha Rao wrote:

See https://www.theguardian.com...

On 2020-11-19 20:34:46, user Hilja Gebest wrote:

Thank you for this study. The serum level aimed for of 30ng/ml is sufficient for bone health but the immune system needs higher levels, at least 40 if not 60ng/mL. More importantly, as the dose was late in the illness, in addition to being quite low. Single bolus doses of vitamin D3 are rarely effective as an intervention particularly when administered without the cofactors: (Mag, zinc, boron, vitamin B + K2 complex and Omega-3). Bolus doses of vitamin D3 start becoming effective up around 500,000 IU and there must be followup with a maintenance dose of at least 10,000 IU/day vitamin D3. The D3 group was disadvantaged by means of many values and risk factors, the three main ones known to us - hypertension, Diabetes II and COPD - by a factor of more than 4:3 vs the Placebo group.

On 2020-11-22 01:11:32, user Mahan Ghafari wrote:

Your phylogenetic analysis is flawed: you cannot estimate a unique TMRCA for two independent introductions like this. Your constructed phylogenetic tree (fig.4) is blatantly incorrect (what's going on with the branches on the red clade B1??). I suggest you retract the preprint immediately and correct the fatal flaws in your analysis. Also, you are not the first group to study the phylogenetics of Iran and you should appropriately acknowledge earlier contributions.

On 2020-11-23 07:29:44, user Andreas Haas wrote:

The peer-reviewed article has been published in JIAS: https://onlinelibrary.wiley...

On 2020-11-25 02:49:25, user Dr Bishnu Mohan Singh wrote:

This preprint has been published in peer-reviewed journal Hindawi: Advances in Preventive Medicine. DOI: https://doi.org/10.1155/202...

On 2020-12-07 14:47:02, user Mark S Perry wrote:

Although you looked for any correlation with self reported BMI I’m wondering if the possible F/M difference might be down to a weight related threshold for a nutrient/drug? As with low dose Aspirin, effective in 1ry prevention of IHD in women - but not (heavier) men

On 2020-12-08 21:22:40, user Michal Piják wrote:

DOUBTS ABOUT THE EFFECTIVENESS OF MASS TESTING OF ASYMPTOMATIC POPULATION FOR CORONAVIRUS (SARS-CoV-2) IN SLOVAKIA

Indeed, it might seem that the number of positive PCR tests / per day, per million inhabitants two weeks after the nationwide testing of the whole country in Slovakia has started to slowly decrease. However, this declining trend may be skewed by significantly less testing. For example data from Monday 9.11.20 show that if as many tests were performed in Slovakia as on Thursday 29.10.20 (when the highest number of positives in the second wave was reached), we should have about 3x times higher number of positives on Monday 9.11.20. cases, i.e. about 3150, instead of 1050.

The cause of the lower number of tests is not known and one of the reasons could be the lack of RT PCR tests or staff in other days. After extensive testing with antigen tests, we had a big problem in Slovakia. This is that so far we have evaluated the situation according to the positivity of PCR tests. However, antigen testing made this situation unclear to us because people tested positive for antigens fell out of the statistics. It should also be borne in mind that lower numbers of positive cases could also be explained by the tightening of epidemiological measures and also because most of the persons with positive antigen tests were quarantined and did not undergo PCR testing.

There is evidence that strategies based on a large number of tests may not produce the expected results. A good example is a comparison of the strategies used by New Zealand and Iceland.1-2 In both of these island countries, the first cases were identified at the end of February 2020, but each country took a different path. New Zealand was one of the few countries that openly announced a COVID-19 elimination strategy right at the beginning of the epidemic. This included a gradually strengthened system for monitoring and isolating contacts with the timely and consistent use of lockdowns and border controls. It should also be recalled that some EU countries, such as Belgium, the Czech Republic, Switzerland, France, Slovenia and the Netherlands, have had a progressive decline in the number of positives, despite the fact that they did not have any comprehensive testing of the entire country.

Unlike New Zealand and many other countries, Iceland's strategy did not include any lockdown period, no official border closure for non-residents and negligible use of quarantine facilities. The cornerstone of Iceland's strategy was easy access to testing and mass screening, along with quarantine and contact tracking. According to data from October 21, New Zealand had 6 times fewer deaths, despite 4.5 times fewer tests than Iceland. Similarly, Slovakia, despite more than 8 times lower number of tests, had half less deaths per million inhabitants than Iceland. It should be recalled that, despite the large number of tests in Iceland, this was not a full-scale test and PCR tests were used. Taken together these findings are further evidence that nationwide antigen testing in a country with low prevalence is ineffective.

References<br /> 1. Jefferies S, French N, Gilkison C. COVID-19 in New Zealand and the impact of the national response: a descriptive epidemiological study. Lancet Public Health. 2020;5:e612-e623

1. Murdoch, D, Gottfreðsson M. COVID-19 and small island nations: what we can learn from New Zealand and Iceland., The conversation, published, September 23, 2020, https://theconversation.com...

On 2020-12-16 21:08:32, user Dieudonné Balike wrote:

This is a preprint, not yet accepted for publication

doi: https://doi.org/10.1101/202...

On 2021-07-02 14:30:43, user Joe Psotka wrote:

I don't understand why they do not report vaccination rates. Did no one participate in a clinical trial?

On 2020-09-25 02:24:05, user Robert Stephens wrote:

Perhaps it is that young children rarely get lung disease with this virus, possibly on account of having fewer pulmonary ACE2 receptors. <br /> As such, they cannot produce aerosolised virus, hence they struggle to spread virus to the lungs of others. <br /> When young children do manage to infect, it is a "safer" transmission. "Non-aerosol" transmission results in virus deposited in upper respiratory/ oral mucosa, not lungs.

Dr Robert Stephens MB BS FACD

On 2020-09-25 03:59:30, user Eitan wrote:

The results are interesting and promising. However, the fact that they are statistically significant does not mean that they are statistically "strong" as long as the R^2 of the linear regressions is not presented. The plots are very scattered and it seems that the R^2 value is much smaller than 0.5. If this is the case, the readers should be very cautious when drawing consequences. If for example the R^2 is 0.3, the statistical meaning is that only 30% of the variance is significant and can be explained by the vitamin D values. But 70% of the variance is affected by other factors. Can you present the R^2 values?<br /> Thanks

On 2020-09-26 00:39:42, user Robert Stephens wrote:

In hospitalised cases, the viral measurement in the pharynx ("pharyngeal load") perhaps reflects a transfer of virions from the lower respiratory tract /lungs (i.e. "ascended" virions).

In mild, non-hospitalised cases (including children), infection is perhaps localised to the upper respiratory tract. The "pharyngeal load" may be high, but disease is mild as there is no involvement of lungs. <br /> Without lung involvement, there is no aerosolisation of virus, hence infectivity will be low as well, despite the high "load".

Robert Stephens MB BS FACD

On 2020-09-28 06:11:14, user Johann Holzmann wrote:

The study certainly provides an interesting perspective on the dynamics of SARS CoV2 transmission in a heavy dense population and on the infection fatality rate. However, it will be crucial to reproduce the findings in a different chohort using a different test to ensure representativeness. <br /> Takita et al, July 2020 provided seroprevalence data in 2 primary care clinics in Tokyo for which the most common cause for patients visits are respiratory infections. They found an appr 5% seroprevalence in their cohort during the outbreak in March/April and concluded that the number of cases corresponded to the cumulative number of confirmed COVID-19 patients by PCR test reported by the Tokyo Metropolitan Government. <br /> PCR testing capacity in Tokyo was significantly increased to 4000-5000 tests per day resulting in about 300 cases per day on average (~7% positive rate on average). It would be highly interesting to read a discussion on how the determined seroprevalence rate of 46.8% agrees with the number of PCR positive cases in the Tokyo metropolitan area.

On 2020-09-29 08:19:46, user Alan Tomalty wrote:

"Allowing for heterogeneity reduces the estimate of "counterfactual" <br /> deaths that would have occurred if there had been no interventions from <br /> 3.2 million to 262,000, implying that most of the slowing and reversal <br /> of COVID-19 mortality is explained by the build-up of herd immunity."

Since the number of counterfactual deaths (no lockdowns) is still over 2x the expected deaths with lockdowns under the heterogeneity model, I don't understand why you can claim that

" implying that most of the slowing and reversal of COVID-19 mortality is explained by the build-up of herd immunity." RATHER THAN BY LOCKDOWNS. My caps is what you actually meant but didn't say. Or am I misunderstanding what you said and what you meant to say?

On 2020-09-30 18:53:22, user James Rubin wrote:

Please note the authors have identified an issue with the underlying dataset that was analysed for this pre-print. Specifically, around 3% of respondents in the dataset reported having been contacted by NHS contact tracers and asked to quarantine. According to official data from NHS Test and Trace, this should be less than 1%. Given this difference, it is likely that we will revise our interpretation of the quarantine data in the final peer-reviewed paper. Until then, as noted in the manuscript, the data relating to quarantine should be treated with caution.

On 2020-10-02 20:40:57, user Juan Mejía Vilet wrote:

This paper is now available at Salud Publica de Mexico 2020; https://doi.org/10.21149/11684 Link: https://saludpublica.mx/ind...

On 2020-10-05 14:16:29, user Julii Brainard wrote:

Our article got kicked back for flaw= reporting 'suspected' not only confirmed cases in same period. I will be curious how Ian's team's work gets received using same case criteria.

On 2020-10-08 05:54:37, user Gennadi Glinsky wrote:

It will be interesting to see if these initial observations could be confirmed and expanded by the worldwide prospective studies precisely mapping the population-scale levels of pre-existing immune cross-reactivity against SARS-CoV-2 to the clinical course and outcomes of the pandemic.

On 2020-10-09 22:47:46, user BannedbyN4stickingup4Marjolein wrote:

OK so this is a theoretical, mathematical model of the spread of the SARS-Cov-2 virus.

The methodology is explained in the paper. There are several aspects of the spread which the model make no attempt to capture.

-The infection is repeatedly seeded by exogenous inputs from outside the population (except perhaps in New Zealand).

-The mode of infection is not homogenous, subject to susceptibility, but appears to be very random, predicated on super-spreading events from which c. 80% of infections derive, such events affecting the very susceptible and hardly susceptible alike.

• The network along which infection spreads is not a constant one but a temporal one. Those who are most susceptible by way of connectivity are not the same from one day to the next. In particular, throughout the pandemic, there have been significant behavioural changes, both in retreat (sometimes officially mandated "lockdown") and "opening up" (including both government sanctioned behaviour and the voluntary behaviour of individuals). Thus the temporal network is not so easily modelled as in the paper.

These are all fairly important effects on the spread of infection, yet none of them are incorporated in the mathematical model.

Each would have the effect of increasing the effective "herd immunity".

Hence I would be very cautious of using the paper, albeit a valuable academic/theoretical study, to inform public policy. To do so would be precipitous and perhaps the authors should recognise this in their discussion.

On 2020-10-14 05:33:25, user Gennadi Glinsky wrote:

Different interpretation of these analyses suggest that preexisting T cells cross-reactive against SARS-CoV-2 are more likely to affect diseases severity because high levels of pre-existing immunity in uninfected individuals appears associated with lower mortality (https://www.bmj.com/content... ). The significant direct impact on the innate herd immunity against COVID-19 and effect on populations’ susceptibility to the infection seems less likely because no association was observed between levels of preexisting immunity and prevalence of the infection.