On 2021-08-30 20:34:53, user Jason Anderson wrote:

I am from the opinion that this type of article being available before being peer-reviewed is slightly irresponsible due to the amount of news coverage it is likely to receive. After reading the manuscript, if I were reviewing, it would be a strong reject or major revisions (depending on the opinion of the handling editor). My expertise is certainly not medicine, but it is on data science and advanced statistical/econometric methods - the precise methodology the authors used here. To keep it short, splitting the data and generating separate models is not appropriate in this context based on the discrete outcomes the authors are modeling. IF, and big if, the authors are going to defend having separate models, there are a series of tests that need to be done to show that this is the appropriate approach. This is lacking. Also quickly, the authors have done their best at controlling for what they can, but there are still numerous unobservables that are not accounted for. Why is this important - it can bias parameter estimates, which leads to ORs (calculated from parameter estimates) that are not true representations of the population parameters. ORs can also be misleading; hence, the preferred inference is based on marginal effects.

If anybody, including the authors, are interesting in additional, more detailed comments, I'd be happy to discuss.

On 2021-09-02 18:28:43, user Kostas Damdas wrote:

Who were the participants? people vaccinated in january vs people covid positive recently? immunocompromised vs healthy?

On 2021-11-09 10:52:01, user Harry Wood wrote:

The Israeli study is flawed on many levels. This dissects the errors.

https://www.quora.com/A-rec...

On 2021-08-26 04:22:25, user brisalta wrote:

The paper does not clearly state which variant the subjects were previously infected with. If that data is available it may be useful to include that information.

On 2021-08-26 04:25:01, user glit00 wrote:

Perhaps I'm overlooking something obvious, but why is there no table that shows hospitalisation OR for model 3?

Also, should table 4b be labelled as "model 3" and not "model 2"?

Finally, I think drawing conclusions on "model 3"'s p-values is a bit cheeky...

On 2021-08-27 13:14:22, user Ash Guirgues wrote:

To which journals is this article being submitted

On 2021-08-28 22:44:31, user Business wrote:

Have you considered comparing Covid-19 naïve vaccinated vs unvaccinated and Covid-19 previously infected vs Covid-19 naïves?

Also, is the data available for further analysis?

On 2021-09-11 16:32:55, user Chadwick wrote:

Red Flags all over the place, like there are 2.5x more immunocompromised in normalized comparison groups, being immunocompromised makes re- or breakthrough infection LESS likely, Economic class swings between comparison groups, being wealthier makes infections more likely...

On 2021-08-31 14:22:02, user Shungo Y@JOJOLANDS(?)????????? wrote:

The standardized mean difference is the difference in proportions between groups divided by the standard deviation for a binary variable. <br /> https://handbook-5-1.cochrane.org/chapter_9/9_2_3_2_the_standardized_mean_difference.htm

The authors seem to use SMD simply as the difference in proportions. Therefore, we do not know if the two groups are well balanced.

On 2021-09-01 09:50:45, user Till Bruckner wrote:

This paper usefully highlights and quantifies the scarcity of randomised trials of NPIs. Providing a precise definition of NPIs and more details on inclusion/exclusion criteria might add value.

A potential weak point is the claim that "it is unlikely that we have been unaware of pertinent results of further NPI trials, given their substantial impact on current debates and scarcity of the evidence." This appears to assume that all NPI trials were either (a) registered in a trial registry or (b) reported in the academic literature.

There may have been experiments meeting the inclusion criteria that were run by government bodies and research units such as "nudge units" that were neither registered nor made public in academic formats.

Performing a grey literature search and/or reaching out to key informants outside academia who may be able to comment on the likelihood of such research having been performed would help to provide assurance that no relevant studies have been missed, and strengthen the conclusions of the paper.

Till Bruckner

On 2021-09-03 13:39:22, user rbrine@msn.com wrote:

Since “each mRNA-1273 dose provides three times more mRNA copies of the Spike protein than BNT162b2”, why do recipients of mRNA-1273 require two doses for “full vaccination”, like recipients of BNT162b2, especially if the first mRNA-1273 dose caused a prolonged adverse reaction?

On 2021-08-11 15:23:39, user Rene Reeves Brandon wrote:

The study included a cohort of unvaccinated individuals, but only reported on outcomes of individuals fully vaccinated with either of the two vaccines. What did the data on the vaccinated reveal in comparison to the vaccinated regarding previous infection, illness, hospitalization, and death? That data is necessary to share, especially as mandatory vaccines are being discussed in several states.

On 2021-09-03 20:53:44, user Cal Research wrote:

This article has been published in Frontiers in Medicine: https://www.frontiersin.org...

On 2021-12-01 13:58:36, user Nudnik_de wrote:

I'm missing one parameter in the study. It seems there is no differentiation made under which condition people interact with each other. In other words, whats the impact of 3G and 2G rules? Vaccinated but not tested people meet Unvaccinated but tested folks... I'm concerned that the lack of considering such aspects could have a severe impact on the results and therefor lead to improper measures.

On 2021-12-01 21:40:48, user anedabei wrote:

The statement of the paper "unvacs drive it" ist not grounded in reality.

The weekly report of the RKI Report from Nov 25 compares vacs and unvacs in "Tabelle 3"

Unfortunately, it is in German, so some help: The row "Symptomatische COVID-19-Fälle¹ " shows symptomatic cases for the prior 4 weeks.<br /> Adding them up results in 289.953 cases for all age groups. 139.856 of them or 48 % are vac breakthru.

So both vacs und unvacs contribute about the same to drive the pandemia. However, vacs are of course somewhat better protected.

For VE, the vac rate needs to be considered. It is, taken from page 24, 12-17 years 43.0 %, 18-59 years 75.0 % and for 60+ years 87.8 %, resulting in an average rate of 68.1 % for the entire population.

On 2021-12-01 22:43:19, user Tom wrote:

Is the use of a 2005 Contact-Model feasable? It does not take the "2G"-Rules and the general fear of Covid into account. I Assume that stadiums full of vaccinated people thinking they are safe while the unvaccinated are not allowed to enter would skew the contact-matrix.

On 2021-12-02 08:51:52, user koen wrote:

This publication makes a number of hard claims, with a title that insinuates as such. These claims are based on a model that is proposed by the authors without proper validation and verification of the model. One of your claims is that your models shows that with a vaccination uptake of 80% of the total population the reproduction number r remains at 0.86 in the current situation in Germany. These claims could be verified by applying the model to the COVID situation in different countries (with higher and lower vaccination uptake). Furthermore, contact tracing results should be used in part to validate claims about the source of infection. Based on the comments above and the discussion in the article the subjective title seems inappropriate and suggestive to person viewpoints of the authors.The best of luck in publishing this article in the current state!

On 2021-12-02 13:52:27, user Thomas Binder wrote:

With all due respect, this modelling study is based on totally wrong assumptions thus is just GIGO: Garbage In, Garbage Out!

On 2021-12-08 20:50:16, user doc_fishoil wrote:

The theme of this paper is the poor attempt of turning absurd assumptions into golden scientific insights by algebraic mumbo-jumbo: <br /> Just take formula (7) to see that "base transmissibility" for the vaccinated and the unvaccinated (that represents their behaviour) produces any proportion of contributions of vaccinated and unvaccinated, as all other parameters are gauged somehow on data. <br /> However, the authors want to blame the unvaccinated, hence they chose to set them as equal although rather harsh testing rules only for unvaccinated were in place in Germany during the referred time ("3G"). Without any reason, comment, validation estimation of real word data, just by assumption in obvious and absurd contrast to everyday life experience. <br /> The result is delivery as ordered.

On 2021-12-01 22:20:28, user Kevin J. Black, M.D. wrote:

One more you may want to look at:<br /> Vitale C, Pellecchia MT, Grossi D, et al. Unawareness of dyskinesias in Parkinson's and Huntington's diseases. Neurol Sci. 2001;22(1):105-106.

On 2021-12-03 05:20:19, user Alberto wrote:

"see Figure 1(b). The plots show the dramatic situation that would have occurred in the case of the lack of vaccines. Indeed, by looking at Figure 1 (a) and (b), we observe an increase of a factor 10 in severe infections. This scary increase would have generated a serious crisis in the Israeli health system."

A 10 times increase in severe cases and (therefor, presumably) deaths is indeed a scary scenario. So much so that it's incompatible with the reality we see everywhere (including, for example, Palestine), and incompatible with the previous year's numbers, when 0% of the population was vaccinated.

There is obviously a very strong confounding factor that must have not been taken into account in these calculations of vaccine efficacy. Finding that confounding factor would be essential for this and all other studies to give us correct estimations. Otherwise we're just speculating with unrealistic numbers.

On 2021-09-29 01:16:54, user Alberto wrote:

So on 2 September the total severe cases was 629 (hospitals and ICUs at the verge of collapse), and the estimation if 0% of the population had been vaccinated is that there would have been 5182 severe cases in the country. And this during the summer. I don't have the figure at hand of how many severe cases were on 2 September 2020, when 0% of the population was vaccinated, but it may have been around 250-300? It's hard to know what exact effect mass vaccination is having that leads to these kind of absurd results, but it would be worth looking at it in detail.

On 2021-12-03 21:49:24, user gwern wrote:

An incorrect result from the first version of this paper (about PhDs being the most reluctant to get vaccines, when really they are probably the least) is still being very widely shared on social media (I can see several instances on Twitter today alone). The error should be discussed explicitly, in more detail, not buried in a vague throwaway comment about some categories being 'higher'; not just so people reading it will understand it, but as an instructive lesson to other researchers about the perils of mischievous responders in surveys, particularly online ones.

On 2021-12-06 07:07:21, user neil Muller wrote:

While this study may raise important questions it is being interpreted in ways that are not justified by the analysis. The paper answers a very narrow technical question as to whether there is an increase in the hazard ratio of primary infection versus reinfection compared to the first wave. Given that the risk profiles of the groups subject to the risk of primary infection and reinfection are so different (by definition the group at risk of primary infections now consists of only 30 to 40 percent of the population who have either adopted behaviour that is less risky, live in communities that were bypassed by the previous waves, or are in the 26 million people vaccinated so far) while the previously infected include the population at higher risk of infection by definition amounting to as much as 70 percent of the population) one would simply expect this.

As the study notes, to date of the possibly 42 million South Africans who have survived Covid infection 36 000 of these people have been identified as reinfections. Naturally as only 3 million infections have been identified by a test so this will be a dramatic under-estimate. But even if it is off by a factor of 15 which identified cases may be this is still only about 500 000 reinfections from 40 million infections.

Natural immunity is highly effective against reinfection.

It is unclear how the estimated change in the hazard ratio changes the projected number of reinfections.

In addition as no information is provided on the risks of hospitalisation and death based on the 36 000 identified reinfections to date we don’t even know whether this has any meaningful policy implication.

But it is the use of this paper in the framing of social and health policy that suggests that if these implications are not spelled out that makes this article misinformation.

The analysts cannot be naive about the debate on vaccine mandates in South Africa. There is clearly a concerted push to demonise the unvaccinated and to make the path to Vaccine Mandates for Covid Acceptable.

The headlines in the popular press focus on the apparent implications of this paper for natural immunity. The claim is that it will not hold up for infection under omicron.

This is clearly NOT what the paper says. The authors need to take responsibility for the way in which this research is being presented and clarify exactly what the paper says about the likely number of people who will be infected by omicron and if so, the number that are likely to require hospitalisation and run the risk of death.

The fact of the matter is that the authors can’t say anything about this as we don’t know about omicron. They admit this.

But they can indicate the number of 42 million South Africans that have natural immunity are likely to be reinfected. They can say the number of these people who are likely to be hospitalised. They can say the number of reinfected people who are likely to die.

They can say that there is no evidence that vaccination will provide any more immunity against infection than previous infection. They can say that there is no evidence that vaccination will lead to less hospitalisations or deaths than natural immunity.

Absent this they remain silent on the calla to reimpose apartheid era strategies such as the population registration act, separate amenities act, and all the hate speech and violation of rights guaranteed in our constitution. This time it is not based on race but on the equally socially constructed and unscientific concept of the unvaccinated.

To sensitise oneself simply replace the term unvaccinated with the k or n word and see if the statements that are made so easily are acceptable.

On 2021-12-06 16:24:28, user Hank Black wrote:

What cycle levels were used in the PCR tests that were used to determine infections? If those levels we’re above 24, then this paper is irrelevant. If the authors can identify people who had symptomatic Covid disease, and compare that number to current persons symptomatic with Covid disease from the new variant, then this paper might have merit.

On 2021-12-06 18:05:54, user FACAGIRL wrote:

What were the CT values for the PCR confirmed cases. I ask because PCR test only test for presence of virus and not infectivity - yes? I found the following CEBM/Oxford systematic review on this and the detail suggesting a lower CT value is better to use as proxy for infection - was based on cultured and PCR samples. Reinfection would be subject to the same limitations associated with PCR tests.

Thanks

On 2021-12-06 15:18:05, user Jens Happel wrote:

Dear Robert,

thanks for the study. Is it possible to differentiate the group of the unvaccinated in unvaccinated and vaccinated between 1st dose and 2 weeks after infection?

In some studies they found the effect that between 1st and 2nd jab the likelihood of infection is significantly increased.

For example here

https://www.researchgate.ne...

see figure 2

Would be intressting to see what happens in this group.

Kind regards<br /> Jens Happel

On 2021-12-08 20:43:25, user Peka Bali wrote:

page 15 of the full text report reads: "Symptom probability time courses for participants with confirmed COVID-19 (n=1020, RT-PCR, antigen, or antibody tests) overlapped significantly with probability estimates from the whole population (Figure 7), except for “changes in sense of smell/taste."

How does this coincide with the conclusion of the report on the first page?!

"Conclusions. Patients with Long COVID report prolonged multisystem involvement and significant disability. Most had not returned to previous levels of work by 6 months. Many patients are not recovered by 7 months, and continue to experience significant symptom burden. "

I am simply flabbergasted by this pseudo-scientific conclusion, not to mention giving it a collective name!<br /> If anything, the only rational conclusion that can be drawn is that other than the altered sense of taste/smell there is NO correlation or causation whatsoever between Covid-19 and the other 65 symptoms as described!<br /> The insinuation as posed above in the publication, stating negative PCR and antibody tests as "suspected cases" is just absurd. What do you base this assumption on?!

This report makes no sense: when you have a control group and baptizing your control group to "suspected cases" to justify the conclusion, which is that these symptoms are Covid related when they are clearly not.

On 2021-12-13 14:29:16, user vepe wrote:

it looks like this study has a major flaw in the calculation of the covid cases

for example their data set contained 6846 cases in the cohort 12-17 (they applied the same logic for the other cohorts)

The 6846 number of covid cases for 12-17 was 2.5% of the total covid cases in their data set.<br /> Then they assumed the same infection rate as the adults at the time, 9.2% and normalized their total number of cases for 12-17 based on that:<br /> adjusted number covid cases = 6846*9.2/2.5 = 6846*3.7 = 25193

then they almost doubled the number of myocarditis cases on the premise that there would be cases that they would miss (e.g. people receiving care outside the TriNetX system)

so they end up with about ~12 myocarditis cases per 25193

so the biggest problem is that their estimated number of covid cases, is essentially the number of covid cases they were expecting to see in their data set and not the total number of covid cases associated with their data. Even if they were meant to estimate the number of covid cases they were expecting to see, this estimation is not accurate since the probability of a younger person ending up in the hospital is way smaller than adults.

In practice, based on that estimation of covid cases, the authors implicitly say that 2.5/9.2=27% of young people that get coronavirus, end up diagnosed/treated by health care provider. This looks like a big overestimation.

In practice, hospitalization rate for younger people looks like is closer to 2% as indicated below:<br /> https://www.aap.org/en/page...<br /> https://covid.cdc.gov/covid...

I think a more accurate estimation would have been to skip the normalization based on infection rates and estimate based on the probability a young person has to end up to a health care provider.<br /> example, covid cases = 6846*100/2 (instead of 6846*9.2/2.5)

Based on this estimation of covid cases, myocarditis risk would be higher in vaccination instead of infection for young people

On 2021-11-24 21:54:37, user Jens Happel wrote:

If the calculation and assumptions would be correct there would be a huge surge of Myocarditis during the Covid19 waves.

But that is clearly not tbe case.

https://jamanetwork.com/jou...

During the Covid19 waves the number of Myocarditis and Pericarditis was more or less constant.m, compared to 2019.

The surge started according to cited paper above in February, when most of the wave was over but vaccination rate started to pick up speed and was changing from elderly to the next younger groups where Myocarditis is more likely.

I guess your assumption about not detected Myocarditis is terrible over estimating that factor.

The charts in cited paper above show clearly that your paper has substantial flaws.

On 2021-09-04 04:21:48, user lifebiomedguru wrote:

Please describe how the groups "vaccinated" and "unvaccinated" defined? Were patients who were vaccinated consider "unvaccinated" until 14 days after their second dose for Moderna of Pfizer products, per CDC's definition? This would clearly bias the main result in favor of your conclusion. The fact that NYT cited this work as a subtext and the Editor chose your conclusions as their title confirms to me that this publishing preprints prior to peer review may be doing some damage to the long-term credibility of science.

On 2021-12-13 19:29:22, user Surya wrote:

Dear researchers,

It's stated in the text that : "However, a 1-42 day risk interval was also used, since this interval is often used in vaccine safety studies of GB S and other outcome."; also the text states "Results were similar when excluding Brighton level 4 cases and when using a 42-day risk window, with incidence rates ranging from 1.1 t o 2.1."

I'm wondering why the results of SCRI are not shown for the 42 days window at risk and mRNA vaccines.

On 2021-12-13 19:29:49, user Joseph Psotka wrote:

The study fails a basic test of good design: the HCW were only described as over 18. That's ridiculous! Full age and gender details should have been provided. Seems like a crummy study.

On 2023-08-30 17:56:46, user Caroline Lima wrote:

This review resulted from the graduate-level course "How to Read and Evaluate Scientific Papers and Preprints" from the University of São Paulo, which aimed to provide students with the opportunity to review scientific articles, develop critical and constructive discussions on the endless frontiers of knowledge, and understand the peer review process.

In this pre-print, the authors discuss the growing demand for ultra-processed foods and their harmful effects on human health. The presence of different oxidized substances and the low nutritional value is associated with chronic cardiometabolic diseases such as cancer, diabetes, Parkinson's, and Alzheimer's. This makes ultra-processed foods a subject of great interest and widely studied. This observation reinforces how important it is to study possible causes for the development of the aforementioned diseases and how research should be conducted to identify and possibly prevent them. It is also important to emphasize that the specific description of what leads these foods to develop oxidized substances is necessary in order to make a correct judgment of the causes and not classify all foods that have undergone some processing as equally containing oxidative substances.

Comments and questions:<br /> The authors prove that oxidative dietary substances and phytosterols are found in ready-to-eat foods and fast foods including those of animal or vegetable origin if preservatives/dyes were used, when high temperatures during the preparation process were used, and in a manner related to forms of storage and distribution.<br /> The use of different biomarkers has been suggested for both ready-to-eat foods and fast foods. Why use brassicasterol biomarkers for ready-to-eat foods and biomarkers (7?-OH and 7?-OH) for fast foods? Is there any specific reason for using these biomarkers? Are there other biomarkers that could be used?<br /> The use of different biomarkers for each food category is reccomended: dairy products (brassicasterol), eggs and derivatives (stigmasterol and ?-sitosterol), meat and poultry (7?-OH), seafood and baby food (?-sitosterol) and others (campesterol). What can each biomarker reveal for each food?<br /> How can the assessment of exposure to oxidative substances be established and what criteria should be considered and disregarded in this assessment? Would these values/results be enough for possible preventions and diagnoses?<br /> For biomarkers, is there any factor that interferes with this measurement and evaluation?

On 2021-12-20 07:43:09, user A. Mesut Erzurumluoglu wrote:

Now published at: https://www.sciencedirect.c...

On 2021-12-20 23:18:35, user Nico wrote:

One more comment - it seems the survey was originally designed to look at impacts of covid itself on menstrual cycles - has that analysis been done? It would be useful to mention in this paper as well. If not already done - that seems like a good control: how do the effects of vaccination on menstrual cycles compare to covid itself? People get so focused on effects of vaccination, forgetting that in many cases effects of covid are far worse. Thanks. (Going to go and search now to see what I can find!)

On 2023-10-23 04:42:47, user CDSL JHSPH wrote:

Dear Dr. Bi et al,

This is a valuable paper that examines the potential influence of prior-season vaccination on the risk of clinical influenza infection. You recognized that past research has shown that prior-season influenza vaccination is associated with an increased risk of clinical influenza infection among vaccine recipients. A key limitation of these previous studies is their reliance on a test-negative design, which fails to consider the intra-season timing of vaccination and the individual's history of clinical infection in the preceding season.

A noteworthy finding in this paper is that individuals who receive repeat vaccinations tend to get their vaccines earlier in the season compared to non-repeat vaccinees. Remarkably, even when after adjusting for this discrepancy in timing, it does not significantly alter the observed higher probability of clinical infection in repeat vaccinees.

Clinical infection seems to play a dual role in influencing vaccination behavior. First, it serves as a motivator, prompting individuals to get vaccinated in the following season. Second, it also provides some degree of protection against clinical infection of the same subtype. However, even after accounting for recent clinical infections, the effect of prior-season vaccination on the current season's clinical infection risk remains not significantly different.

A potential mitigating factor, subclinical infection, is theoretically posited to attenuate the effect of prior-season vaccination. However, you were clear in the paper that this aspect is still largely theoretical and necessitates further investigation to determine its actual impact on vaccine efficacy.

The primary contribution of this pre-print lies in its careful consideration of confounding factors, specifically the intra-season timing of vaccination and the history of clinical infection in the previous season. By addressing these variables, it challenges the established findings of prior research, which suggest an elevated risk of clinical influenza infection associated with prior-season vaccination. These insights carry significant public health implications, particularly in the realm of vaccine policy and compliance.

The paper is methodologically robust, particularly in the sections that explore the impact of timing and previous clinical infection. However, the discussion of subclinical infection is less conclusive, as it relies on a theoretical model and a pseudo-population. The exact details are not in the main body of the paper and was referred to the supplemental section. As the explanation for the main findings of the paper is hinged on subclinical infection, it may be helpful to develop this idea further in the main text.

In terms of its presentation, the paper is well-structured with clear delineation of sections, and the text is appropriately complemented by the figures. The inclusion of the "infection block hypothesis" in the discussion aids in facilitating a deeper understanding of the research.

Overall, this paper marks a significant breakthrough by challenging the conventional approach to assessing vaccine efficacy, incorporating the roles of vaccination timing and previous clinical infection. It also highlights the potential importance of subclinical infections, opening important conversations and may lead to enhanced strategies for data collection in this context.

We truly appreciate you sharing your pre-print with us.

On 2021-12-22 17:55:54, user Thomas Gade Koefoed wrote:

Awesome work! I would perhaps consider rephrasing the sentence in the abstract: "However, the VE is significantly lower than that against Delta infection and declines rapidly over just a few months", since it can be read ambigously providing two opposite meanings. (Depending on what "that" refers to; the VE or the the statistics just mentioned in the previous sentence.)

On 2021-12-24 07:45:25, user Jeff H wrote:

So assume the results you like (high VE for recent vaccination) are causal, but hand wave confounders at results you don't like (negative VE for distant vaccination)? Science?

On 2021-12-24 21:34:35, user Robert Parker wrote:

So, these vaccines are, essentially, not effective against Omicron. The upside is that Omicron seems, at the moment, to be like getting a really bad cold. Very little hospitalization, and no deaths as far as I can find. This may be a Godsend. It is highly transmissible, with few bad effects. It may actually serve as a means to herd immunity, with few deaths. Hope springs eternal.

On 2021-12-28 14:49:56, user Bob Horvath wrote:

There is a typo in the confidence interval reported here: "36.7% (95% CI: 69.9 to 76.4%)", since the confidence interval needs to incorporate the value of 36.7%.

Also, this paper defines vaccine effectiveness (VE) as protection against infection. As can be seen from some of the tweets on this study, that is confusing readers who don't realize the EUA granted in the U.S., for example, is based on definitions of effectiveness related to hospitalization and/or death. It would be very helpful to many to have this even very briefly clarified in your paper, that, for example, even if the VE was 0% (or even negative, as some of the threads here claim is being shown after 90 days) according to the definition used in this study, as long as it had more than 50% effectiveness against hospitalization and death, that it would still be used in the U.S.

On 2022-01-13 13:10:50, user David Knight wrote:

Scotland's latest official public health real world data tallies up with the negative effectiveness found by the scientists that carried out this study.

https://publichealthscotlan...

See Table 15.

People who had only 2 jabs were almost 3 times more likely to catch Covid in the week 25th Dec-31st Dec than the unvaccinated (who were similar to the 'boosted')

Unvaccinated 1,555,449, cases 20,276, 1.3%<br /> 2 Doses 1,522,961, cases 54,727, 3.59%<br /> Boosted 2,429,498, cases 30,222, 1.24%

But if you are boosted you appear to be at least 4 times less likely to be hospitalised or worse from Covid, than the 2 jabbed/unvaccinated. See tables 16 and 17. So there still is a case for the vaccines

On 2021-12-22 22:56:01, user Richmond Heath wrote:

Have any of the authors considered a possible restorative role & purpose of the tremors & vibrations rather than simply seeing them as a pathaological? Could they be efforts to help down-regulate the ANS from the chronic hyper-arousal associated with long Covid similar to the 'neurogenic tremors' associated with the recovery cascade after shock & trauma seeking to restore the systems of the body to natural states of flexibility & variability?

On 2021-12-26 20:00:42, user Lee Jimmy wrote:

I read the preprint and could not find any mention of mask use/non use nor was the type of "activity" at this gathering spelled out. Anybody know anything about these details ? Do I need new reading glasses? Etc.

On 2021-12-28 09:32:57, user Joe Random User wrote:

Table B clearly says that some / many of the participants had only just received their booster jab on December 2nd. The date of the private gathering was "early December" and the genetic sampling results were received by December 8th. So most likely the private gathering took place between December 2nd and December 7th.

Everbody knows that it takes 2 weeks for the antibodies to develop to their full potential after the third booster jab.

Since this article does not specify how many of the participants had not been fully vaccinated with the 3rd booser jab the data in the article is insufficient to learn anything about the omicron variant's resistance to the booster jab.

I recommend that the authors produce a revised paper where they more carefully describe the vaccination dates for the 3rd jab for all participants.

On 2022-01-06 14:06:14, user Maker innovations Tom maker wrote:

Notice they do not list the non pharmaceutical therapies

On 2022-01-09 16:15:33, user Greg wrote:

Here is my big bone with the study. The OR given for Omicron susceptibility is 1.04 for the unvaxxed, suggesting that the double-vaxxed were pretty much equally susceptible. Reading the Method, however, it stated that they counted the vaxxed with one dose as unvaxxed. What?! Would that not mean the true unvaxxed were less susceptible to Omicron? Likely!

Also, with vaccine protection waning rapidly, and even for the boosted, it would've been nice to know how vaccination timing was affecting susceptibility. This study did not consider that. Interestingly, the other prior Danish study suggested that after a few months of being double-vaxxed, there was a net negative protection against Omicron.

On 2022-01-31 21:33:01, user Anonymous wrote:

What do the "ref" mean in table 2. Trying to see the figures for fully vaccinated individuals.

On 2022-01-06 17:05:32, user SE Justus wrote:

Where/how do those who have recovered from the disease (naturally immune) fit into this study?<br /> Thank you.

On 2022-01-06 22:05:45, user Faithkills wrote:

Failure to include those with acquired immunity from recovery makes this of little use and exposes a likely bias of the researchers.

On 2020-05-01 11:05:45, user Robin Whittle wrote:

Please see this report from Dr Mark Alipio, Davao Doctors College; University of Southeastern Philippines: Vitamin D Supplementation Could Possibly Improve Clinical Outcomes of Patients Infected with Coronavirus-2019 https://papers.ssrn.com/sol... . Hospitalised COVID-19 patients were classified into Mild (without pneumonia), Ordinary (CT confirmed pneumonia with fever and respiratory symptoms), Severe (hypoxia and respiratory distress) and Critical (respiratory failure).

Of the 55 patients with greater than 30ng/ml (20nmol/L) 25OHD, 47 had Mild symptoms, 4 Ordinary, 2 Severe and 2 Critical. Of the 157 patients with 30ng/ml or less, 2 had Mild symptoms, 55, Ordinary, 54 Severe and 46 Critical.

On this basis, if everyone had more than 30ng/ml 25OHD, very few people would be dying from COVID-19 and there would be no need for lockdowns, with their extremely high social, health and economic costs.

In this research, Gallagher et al. 2014 “Vitamin D supplementation in young White and African American women” https://www.ncbi.nlm.nih.go... , almost all the White women had less than 30ng/ml 25OHD. Those who took 2500IU vitamin D3 raised their levels significantly, but about 16% of them were still below 30ng/ml. 4000IU a day would improve on this considerably. African American women generally had lower levels.

4000IU is 0.1 milligrams a day. A gram would last for 27 years. The ex-factory price of vitamin D is USD$2.50 a gram, so the cost of this good, healthy, level of vitamin D supplementation is 9 cents a year, plus the cost of making and distributing and selling capsules. D3 need only be taken every week or two. My wife and I take a 50,000IU capsule three times a month.

Figure 3 at https://www.ncbi.nlm.nih.go... shows that normal weight people taking 4000IU a day will, on average, reach 47ng/ml (117nmol/L) which is about the average level of African herders and hunter gatherers reported in https://www.ncbi.nlm.nih.go... . Toxicity (https://www.ncbi.nlm.nih.go... "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6158375/)") may occur at levels three times this.

More links to research are at my page: http://aminotheory.com/cv19/

On 2021-10-20 16:49:53, user Emily Wood wrote:

This paper has been accepted for publication in SAGE Open Medicine.

Please cite as:<br /> Wood, E., Bhalloo, I., McCaig, B., Feraru, C., & Molnar, M. (2021). Towards development of guidelines for virtual administration of paediatric standardized language and literacy assessments: Considerations for clinicians and researchers. SAGE Open Medicine. https://doi.org/10.1177/205...

On 2021-10-21 01:57:10, user Sook Wah Yee wrote:

Great work @Eric_Fauman: I am so excited but I did not see that Supplemental Tables are attached with this paper? Or did I miss it? I only see one pdf file attached with Supplemental Figures. Thanks for checking.

On 2021-10-26 17:10:43, user Stephane wrote:

Could you please explain why the effectiveness is lower between fully vaccinated people ? "Effectiveness of full vaccination of the index against transmission to fully vaccinated household contacts was 40%"

On 2021-11-09 13:00:01, user ingokeck wrote:

Dear Authors, two Questions:

(1) You state: "Partly vaccinated was defined as having received the first dose of a <br /> 2-dose schedule with a time since vaccination of at least 14 days." So you counted freshly vaccinated persons as not-vaccinated? IMHO this is a bad idea, because in the first 14 days after the 1. dose it is well known that the immune system is impacted by the vaccination and a high risk of testing positive for Covid19 exists. If you count these cases as not-vaccinated, this will skew your results towards higher vaccine effect.

(2) Thanks for plotting the case counts in figure 1. Did you check if there is some temporal imbalance in the cases? It seems the second part of your data interval has a substantial lower infection risk and may have higher vaccination numbers, i.e. you may have data that skews towards vaccinated in the lower risk time, also accounting for part of the measured vaccination effect. Could you please have a look at this as well? Thanks.

On 2021-11-02 05:39:54, user Athanasios Chalkias wrote:

The manuscript has been published in Surgery:<br /> https://www.surgjournal.com...

On 2021-11-02 11:57:57, user guy wrote:

Hopefully the reviewers will insist that the serological data and explicit discussion of their assumptions are brought into the main body of the text, most importantly this means table S4.<br /> From the abstract “We then use probabilistic risk assessment and data on [.., ]human SARSr-CoV seroprevalence, [..] to estimate that ~400,000 people (median: ~50,000) are infected with SARSr-CoVs annually in South and Southeast Asia. “ appears to be incorrect as the dataset used to approximate a distribution of SARSr-CoV seroprevalence, only 4% (1/27) of the positives are from this viral grouping (70% are from Nipah or Ebola viruses). The only data on SARSr-CoV comes from a single study (by the same authors (https://doi.org/10.1016/j.b...) "https://doi.org/10.1016/j.bsheal.2019.10.004))"), which if exclusively used in this study --as the abstract would imply-- would likely dramatically reduce striking numbers in the abstract. Using distributions to allow for uncertainty is a good approach if the data used to approximate them are valid, in this case any justification appears lacking.

The fact that the above cited article concluded that “Direct contact with bats was not identified as a risk factor [ in the transmission of coronaviruses to humans ]” should also be discussed in the current article, given that they now explicitly assume the opposite .

On 2021-11-07 19:23:55, user Eleutherodactylus Sciagraphus wrote:

It is relevant to note that this preprint (along with other two from the same group) includes data from human subjects that are under ethical scrutiny. The majority of patients enrolled were not informed nor agreed on participating in the study. The Brazilian National Comission for Research Ethics (CONEP) has been bypassed and is now investigating this case.

https://brazilian.report/li...<br /> https://www.emergency-live....<br /> https://www.dire.it/14-10-2...<br /> https://www.matinaljornalis...<br /> https://g1.globo.com/rs/rio...

On 2021-11-09 07:26:10, user Željko Serdar wrote:

Janssen<br /> March, 92%<br /> August, 3%

Moderna<br /> March, 91%<br /> August, 64%

Pfizer<br /> March, 95%<br /> August, 50%

On 2021-11-09 21:39:32, user Andy Bloch wrote:

Conclusions are contradicted by this paper: Waning of SARS-CoV-2 antibodies targeting the Spike protein in individuals post second dose of ChAdOx1 and BNT162b2 COVID-19 vaccines and risk of breakthrough infections: analysis of the Virus Watch community cohort.

On 2021-11-11 01:06:23, user transplantee wrote:

This paper is now published at https://pubmed.ncbi.nlm.nih...

On 2021-11-12 17:30:08, user Andre Bafica wrote:

Thanks for sharing this very interesting data! Have you guys seen our eLIFE paper with a similar IFN/IL6/CEBP signature in human TB? https://elifesciences.org/a...

On 2021-11-15 08:07:51, user kdrl nakle wrote:

Awesome, but now you need manufacturer.

On 2021-11-15 08:53:39, user kdrl nakle wrote:

Very high HR, I am really surprised.

On 2021-11-20 23:32:38, user Gordon V. Cormack wrote:

Were the previously infected also vaccinated, either before or after their infection?

Edit: I think I answered my own question: <br /> When we examined HCWs (n=423) with infections occurring before vaccination, no re-infection was observed, accumulating 74,557 re-infection-free person-days (starting 10 days after initial infection and censoring at the date of receiving their first vaccine dose). Further, after vaccination, previously infected HCWs did not contribute any breakthrough infection events among the vaccinated HCWs.

On 2021-11-26 03:56:31, user mike wrote:

I's really like to see this study continued until there is say 10 people who have been re-infected or a conclusion to say a million days without re-infection. 75k days is a lot, but they may be much more educated and a significant percent of these people do not want to get the virus a second time, therefore creating a dramatic improvement rate comparatively otherwise.

On 2020-04-22 08:46:06, user jaxthots wrote:

With self-selected subjects, sampling bias is always a major limitation for generalizing study findings. However, anxiety is probably a common emotional denominator in this sample over the focal existential issues of health and employment that are widely shared by the larger population. More significant bias, perhaps, are the unique characteristics of affluent, high-tech dominated Santa Clara County which, with median household incomes over $100,000 and median SFR values in the $1M range, is unrepresentative of America at large. But with a major airport and lots of foreign professionals visiting Silicon Valley, it has undoubtedly had more exposure to hitch-hiking viruses than most other US populations, which would predict above average infection rates. And similar findings from LA County, Germany and other populations sampled irrespective of varying subject samples and testing methods are providing a clear enough picture to establish a realistic denominator in rate calculations.

Unfortunately, the nominators are far less clear since US hospitals and doctors nationwide have been instructed and incentivized to identify covid-19 as cause of death irrespective of serious co-morbidities, which Italian doctors report are near-universal among the seriously ill and the deaths of presumed covid-19 patients, with definitive test results typically not returned until 2 weeks after the death certificates are signed and recorded. With no possibility before now of valid rate calculations, why are the media inflaming public panic that support draconian protective measures with severe economic consequences by reporting fabricated data?

In addition, many US doctors are reporting peculiar symptoms of "oxygen starvation" without the expected fluid lung congestion of pneumonia, suggesting a different, yet-unidentified disorder. These might possibly involve EMF interacting synergistically with the infection since the pandemic epicenters are in cities that rolled out 5G last year and EMF has been shown to alter voltage-gated calcium channels in hemoglobin molecules that deliver oxygen from the lungs into the circulatory system, as well as other compromising effects on pulmonary functioning.

There are one or more very serious, troubling and suspicious agendas at work here that beg more than perfunctory investigation.

On 2021-11-23 10:17:56, user Yixiang wrote:

Do you assume vaccine efficacy wane over time since the 2nd jab -- e.g. Antibody titer drops by 75% 6-12 months after 2nd jab? What about natural immunity?

On 2021-05-19 18:42:43, user Fred Bass wrote:

Were the patients randomized into those getting usual and those getting 12 week delay? Having 99 in one group and 73 in the other does not seem like a random split of 172 people! A bias toward giving healthier seniors the longer interval might account for some or even all of the results.

On 2021-01-21 16:36:21, user Michael wrote:

DL+DiReCT including the trained model is also available on github: https://github.com/SCAN-NRA...

This method is mentioned in the manuscript and refers to https://doi.org/10.1002/hbm...

On 2021-01-27 15:14:20, user Florence Paré wrote:

How do you account for the possibility of COVID infections disproportionately occurring later in the period under study (due to rapidly rising numbers of infections), whereas influenza and respiratory tract infections may tend to slightly go down over the period due to distancing measures? This seems to risk introducing a confounding variable - mental health deterioration due to social distancing and pandemic-related anxiety. Did you or do you intend to make adjustments to the control cohorts to match the distribution of events over the period under study?

On 2021-01-28 19:01:30, user lbaustin wrote:

This leaves out two simple blood tests that are more predictive than any of the parameters on the list: initial blood sugar of >140 and 25(OH)D of less than 20ng/ml. Please add these to the model prior to publication.

On 2021-01-29 12:29:23, user stephan walrand wrote:

Nice correlation with the cloudiness and sun light insolation, but which is also compatible with vitamin D production!!! However, it is obvious that when comparing deaths from March to July, it is impossible to see any latitude correlation, because sun elevation averaged between March-July is almost equal for all countries.

On 2021-01-31 18:31:27, user Graeme Ackland wrote:

The statement

"we showed approximately 51% effectiveness of BNT162b2 COVID-19 vaccine against PCR-confirmed SARS-CoV-2 infection 13-24 days"

Is highly misleading. The data suggests more like "15% effectiveness 13-18 days, 85% effectiveness 19-24 days.". The most relevant day is day 21, when the second dose is meant to be given.

So their conclusion is that someone else should be deprived of 85% protective first dose, in order to give an 10% uplift with a second dose.<br /> I find that logic debatable

On 2021-01-31 21:31:27, user Ilya Zakharevich wrote:

The last two columns in the tables do not match each other (as they probably “should” for all developed countries, if one wants to get “meaningful comparison”; look for Lithuania vs Liechtenstein). I think that this is due to very different strategies to count child mortality.

Is it possible to replace the last column, dividing by the mortality (say) after age 1 year? As I said, it may be a “more interesting” number. (Less dependent on arbitrary accounting policies…)

On 2021-01-31 22:01:02, user Pablo Olavegogeascoechea wrote:

I have read this trial with great interest and I have some worries about some detalles: fist of all, the absolute risk reduction is quite low (1.4%) and the NNT for the primary outcome is 70 as it is for hospitalization. On the other hand there were more patient who developed pulmonary embolism in the Colchicine group (may be this issue needs more infromation)

On 2021-02-01 11:20:07, user Fjortoft9 wrote:

Given that the study is assuming the rate of vaccinations will be around 1m a week in January, rising to 2m by February I’m afraid it doesn’t seem to be very useful. <br /> We know now that the actual rate of vaccinations in January was more like double that and the rate in the last week is well over 2.5m. That difference would completely change the modelling and it’s disappointing that you didn’t model the impact of a faster vaccination rollout.

On 2021-02-01 15:11:35, user Alessandro Soria wrote:

Very interesting paper. To my knowledge, there are at least three other papers which look at the same topic (the effect of healthcare strain on COVID-19 mortality) from other perspectives: <br /> 1. doi.org/10.1371/journal.pon.... This is our recently published work, in which we tried to assess the impact of patient load on in-hospital mortality from COVID-19 based on hospital stress variables, such as the number of daily admissions, the number of total daily census, and the period before the peak, and we did find an independent harmful impact on mortality.<br /> 2. doi:10.1001/jamainternmed.2020.8193. In this analysis on the variation of COVID-19 mortality over 6 months in the US, the authors found that increased mortality reflects increasing numbers of cases in the community, possibly reflecting hospital burden.<br /> 3. doi:10.1001/jamanetworkopen.2020.34266. In this report on ICU in the US, there is a clear association between exceeding bed occupancy and increased mortality.

On 2021-02-01 21:30:19, user Igi Dano wrote:

As a Slovak citizen, I agree with most comments/notes presented here. I could as well add my own experience with "following testing procedure recommended by manufacturer..", where this testing procedure was conducted outside (of any premise, just an open tent) with temperature well below recommended range.

But that is not the point of my post here. The point is that Slovakia is currently (1.2.2021) ending the second round of another population-wide screening.<br /> I am desperately waiting for another study from the authors, confronting the newest results with original ones. <br /> Without that I would recommend potential readers of this study to use extreme carefulness with interpretations of it..

On 2021-02-02 12:13:10, user Miriam wrote:

Nobody in Slovakia was informed about this research. And it was not voluntary as they signed. There was and there is still strictly prohibited to go at work and to the nature if we are not tested. The final result of this mass testing is, that numbers of covid positive strongly increase. That is all. I am really afraid about my human rights in future.

On 2021-02-19 01:42:51, user Oliver Cudziš wrote:

Voluntary? "All necessary patient/participant consent has been obtained and the appropriate institutional forms have been archived - Yes" What is this, areu all blind or what. Slovak nation was like experimental rabbit without knowing, congratulations you just made stage for Slovak national uprising 2, good luck.

On 2021-02-04 21:24:38, user Charles wrote:

I am a bit unsettled by the days they decide to pick to get their best (around 90%) estimate.

They use the daily rate from day 1 to 12. Day 1 it's .028% and average until day 12 is around .041% (there is a spike in infection from day 1 to 12). Day 21 is .004%, day 22 is .011%, day 24 .006%, i.e. there is some standard-error. <br /> Now, it's all about which days one picks. <br /> - if one calculates Expected as being day 1, the best effectiveness rate is 86% on day 21, but on day 22 it dwindles at 61%...<br /> - if you use day 1 to 12 as Expected, effectiveness rate is around 85% on day 24 and 73% on day 22. <br /> - to get the 90% in the paper, you need to pick day 21 (the lowest incidence, that went up again the next days) and the Expected as days 1 to 12 (the highest incidence).

It seems that efficiency estimate does improve over time, but reaching 90% depends on which days one picks, both in term of "actual" and "expected". This choice might very well explain fluctuations between 60% and 90%, i.e. the estimate is very sensitive to small numbers and differences. Differences with previous estimate might be methodological (no proper control group).

On 2021-02-10 19:35:03, user philippeboucher wrote:

Any link or conflict of interest?

On 2021-02-11 16:06:09, user David McAllister wrote:

Congratulations on this excellent work. The potential for ICS therapy to improve outcomes for intermediate risk individuals not yet vaccinated is tantalising.

No doubt the paper is currently under peer-review, but if the authors have time it would be great to know the following:-<br /> 1. How many of the primary endpoint events included hospitalisation.<br /> 2. How was such a high proportion of positive tests for SARS-CoV-2 obtained? Was this based on subjective clinical judgement, or was there some other factor driving the high pre-test probability ?<br /> 3. How difficult was it to teach adequate inhaler technique?<br /> 4. Did any of the participants have wheeze or other signs of reversible airflow obstruction?<br /> 5. Were any steps taken to exclude participants who might have had a lobar pneumonia (eg by excluding individuals with purulent sputum)?<br /> 6. In the Guardian interview it was mentioned that at least 5 other trials were investigating this use of ICS. Is it possible to say when these are due to report?

On 2021-02-16 20:52:40, user Chris Cappa wrote:

Very interesting study. Interesting to see that exercise doesn't appear to increase the smaller particles but does the larger particles. In any case, two factors you might consider in revision. First is the differential dilution that will occur between different activities. Breathing and talking expiratory airflow rates differ substantially from coughing, from the various ventilatory therapies, and importantly from the OPC. Thus, there will be different levels of dilution associated with each activity that you might factor in to facilitate comparison between activities. It doesn't appear this was done (although I could be wrong). Or, at least note that this likely had an influence. The second issue relates to the comparability between the different activities. For example, talking was continuous whereas coughing was just 6 times in a minute. If a person had (for example) been asked to cough twice as often the number of particles measured would have doubled. Or, if there were more breaks in speech the number of particles would have differed. You might consider normalizing to per second of activity to allow for greater comparability.

On 2021-02-19 10:08:04, user Javier Mancilla-Galindo wrote:

This study is interesting, with robust analyses and a great effort to adequately report the model. Including predictors like S/F ratio, frailty score, and acidosis clearly differentiates this model from others and would make it a highly clinically relevant model. However, I am afraid it may lack any real clinical utility as long as the authors do not clearly explain in a simple way to clinicians how this model should be used in real-world settings (unless I somehow missed it).

Dichotomization of age (i.e. greater than cut-off age) may have led you to loose discrimination ability since too many studies have already shown that age is the main risk factor for mortality in patients with COVID-19. This may, however, not be an issue for such a shot-term (48-hour) mortality prediction, although I do strongly believe this model would have had a better mortality discrimination had you evaluated age differently (i.e. multiple age categories could be included with different weighted risks or coefficients, or perhaps allow age to be inputted as a continuous variable if at all compatible with your model).

The model shown in Supplementary Table 4 that includes CRP and not IL-6 could have a greater potential to be widely used even in moderately resource-strained hospitals. Thus, I found it more useful from a global perspective. Even when the model including IL-6 is better at predicting the outcome, it could have limited clinical applicability as correctly stated in the manuscript.

Lastly, you have adequately reported your manuscript according to the TRIPOD statement. However, the RECORD statement may also apply to this particular study since you have used routinelly-collected data in an observational study design. You could consider including this checklist, too, for the peer-review process.

Congrats for such a great work!

On 2021-02-20 03:01:31, user kdrl nakle wrote:

Believable but samples too small and Fisher Exact is not a reliable test.

On 2021-02-21 14:31:34, user DMac wrote:

Good day. I've found this work immensely valuable as a reference for discussions in our office. With new variants developing and particularly the "UK" B.1.1.7 and "South African" or B.1.351 variant spreading, I wonder to what extent the changes they reflect would impact modeling results. I expect most variables are the same, but wonder if the added efficacy of transmission can be accounted for with the model. As an interim approach, might one adjust downwards the risk tolerance or other variable to approximate adjustment for the variants?

On 2021-02-23 23:14:07, user phil wrote:

Fig 1I - the plot is piecewise linear. Shouldn't it be a step function? The key dates mark the point where presumably R_t^eff changes, which should then be constant until the next key date?

On 2021-02-24 02:58:40, user Eric O'Sogood wrote:

1. The trial was stopped early and did not enroll enough subjects to meet its own initial power calculations. 2. Single dose ivermectin at this stage is not the recommended regimen. 3. Ivm arm had the highest d dimer (p 0.01) and I do not see any discussion of anticoagulant beyond thromboprophylaxis. 4. Absorbtion of ivm with food rises ~4 fold, was it given on an empty stomach or with food? 5. The authors write that this is the first trial of ivm vs placebo. There are already 5.

On 2021-02-24 22:37:41, user Sócrates Brasileiro wrote:

There are not two waves. It is the same pandemic, reaching different people. Countries population are more or less constant in one year. This means that infection and fatality rates should be computed by summing up (in the numerator) respective cases during the whole period. And not by splitting the numerator into two waves as if they were cases from different pandemics. If this was done, previous statements by one of the authors, such as "covid is as deadly as driving your car to work", would clearly be wrong, as they are indeed.

On 2021-02-25 19:02:42, user Lisa Mair wrote:

I'm so reassured that others are noticing that their conclusion does not match what their data showed. I've seen this in several of the pro mask studies. Like in the Lancet mask study, authors admit that the data is low certainty of evidence and that there were confounding variables, but they still strongly recommend masks. The WHO recommends masks but then admits data is weak. It's very common. Do you think it's because of encouragement of a specific conclusion due to funding? It is well known that research usually favors the desired result of the funder.

On 2021-02-28 00:56:39, user Kevin wrote:

Still, the vast majority of studies have shown significant increases in survival and with a drug generally as safe as Ivermectin waiting for perfect evidence is deadly and foolish. Remdisivir was approved with much less efficacy and much more side effects (many severe). I find it laughable that we are tip-toeing around with ivermectin but there was no problem at all pushing a drug through approval that hadnt shown a significant increase in survival but, hey, atleast it will help you get out of the hospital faster! - If your lucky enough to survive that is.

On 2021-02-28 12:37:41, user micro dentist wrote:

Many thanks for your effort. Very useful data, yet requires cautious interpretation.<br /> It is important not to aggrandise conclusions when the sample population is skewed due to disproportionate under-representation.

Such an aggrandisement potentially occurs here:<br /> “The observation that the seroprevalence amongst dental practice receptionists, who have no direct patient contact, was comparable to the general population, supports the hypothesis that occupational risk arose from close exposure to patients.’

Whilst in comparison to 16% of clinical staff 6% of receptionists were seropositive, it is important to also acknowledge that 21.6% of practice managers (also non-clinical) were seropositive.

Where significant conclusions may be derived through occupational comparisons, the effect of disproportionality should also be independently validated through careful examination of the internal validity of any inferred conclusions.

Here this would show lack of consistency with the derived conclusion. Should there still be a requirement a desire for an assumption, it may be worth considering combining any smaller similar samples (such as receptionists and practice managers in this case). In this study such combined group would show a seropositivity of 12.2% (n=131).

Through erroneously overlooking disproportionate occupational representation, there is the real potential of developing ludicrous conclusions: the most obvious being that seroprevalence is related to the amount of occupational administrative paperwork completed by each member of the team: practice managers>dentists>receptionists.

Clearly such a conclusion is neither desirable or valid.

On 2021-03-03 00:42:19, user James Gorley, PhD wrote:

In this ambitious study, the authors set out to show histological safety of low intensity FUS. A few key questions should be addressed by the authors. Namely, if the EEG was not usable, how is the claim of "temporal slowing" of one participant justified? Was any statistics or rigor applied to support this claim? Furthermore, two participants are excluded from the analysis, but the data is analyzed later anyway in the psych testing. Interested to see how this manuscript will evolve!

On 2021-03-05 15:23:14, user Martijn Hoogeveen wrote:

Research is picked up by De Telegraaf, largest Dutch newspaper https://www.telegraaf.nl/ni...

On 2021-03-06 04:23:48, user Larisa Tereshchenko wrote:

Now published in Heart Rhythm https://www.heartrhythmjour...

On 2021-03-09 14:22:05, user Alexis Alomoto wrote:

I would like to know, in which countries or where were the participants chosen to carry out this study?

On 2021-03-09 21:34:52, user Marm Kilpatrick wrote:

Thank you for this important study.<br /> Could you please upload all the supplementary materials as a single file? Thanks!

On 2021-03-10 12:58:49, user Zed wrote:

So it's not working in early treatment, and during hospitalisation (cf: https://www.clinicalmicrobi... "https://www.clinicalmicrobiologyandinfection.com/article/S1198-743X(20)30505-X/fulltext)")

On 2021-03-11 21:24:49, user disqus_foVd2sEK3I wrote:

Thank you for this important work. I was hoping to take a closer look at the model, only to find out that it was not included. It would be useful to people like me to include the new model's equations for reproducibility.

On 2021-03-12 16:18:56, user NickArrizza wrote:

Are you aware that up to 80% of the co-morbid conditions associated with<br /> 94% of all deaths from COVID-19 are totally preventable (and reversible<br /> within weeks) with a whole plant based diet that lowers inflammatory <br /> markers and hypercoagulability thought to be highly correlated with <br /> severity of illness in COVID-19?

On 2021-03-13 07:53:06, user Motaz wrote:

Thank you for such a great work. I just want to point out in the list of figures at the end, "Fugure 4" is misspelled.

On 2021-03-15 17:04:19, user Eli Yazigi wrote:

Decoding Distinctive Features of Plasma Extracellular Vesicles in Amyotrophic Lateral Sclerosis

Key main ideas in the paper:<br /> • Nickel-Based Isolation (NBI) of extracellular vesicles (EVs) is an effective technique that both preserves the integrity of EVs and easy carry out in a clinical setting.<br /> • Extracellular vesicles in Amyotrophic Lateral Sclerosis (ALS) have distinctive features—in terms of size distribution and protein composition—that are different from EVs of patient with other muscular degenerative diseases (MD).<br /> • The amount of accumulated TDP-43 is indicative of the pace of progression of ALS. Increased accumulation of TDP-43 indicates faster progression of ALS in patients.

Main contribution to the field: The paper established that size distribution and composition of plasma extracellular vesicle can be reliably used to distinguish ALS from other muscular degenerative diseases.

On the scale of 5 (breakthrough) to 1 (no contribution to the field) I would rate the contribution of this paper at 4. The paper provides fast, reliable, and easy technique for isolation of EVs in clinical settings. Using this technique to analyze composition and size of EVs helps in making differential diagnosis.

The conclusions the authors draw in this paper follow experiments performed. And the assumptions made by authors are reasonable and well-thought. However, I think expanding the age range for participants to include younger patients would enhance the credibility of the data and provide for crucial insights.

One disadvantage of using NBI, is that it does not allow for isolation and distinction of extracellular vesicles that are generated through different biological processes (i.e., exosomes vs. microvesicles). These different types of vesicles are regulated in different manner and contain different cellular components.

On a scale of 5 (great) to 1(muddled), I would rate the writing in the paper at 4. There are few typos and grammatical errors. But for the most part the writing was clear and concise. I had to re-read the discussion section couple of times to understand to various conclusions and connect them together. Overall, algorithms are clearly explained in the paper. The logical follow in the paper is smooth and relatively easy to follow. Nonetheless, I think that the connection among various conclusions in the paper could better emphasized.

I think this paper will have a profound, lasting impact in clinical settings. It outlines the use a creative method to draw differential diagnosis among ALS and other MD diseases. The reliability and ease of method presented in the paper along with the data will prove to be revolutionary in the field of medicine.

On 2021-03-16 00:48:29, user Brian wrote:

The main conclusion is driven by a particular 14 day past 2nd dose counterfactual which does not seem realistic in the context of other data. These are the are the graphs in the supplementary material. It makes VE look higher than it likely is during that timeframe. Otherwise results inline with other papers.

On 2021-03-16 18:23:19, user Zhe Zheng wrote:

It's a really interesting finding! Will be nice to know what kinds of NPI have been implemented in Lyon. This will enable a comparison between different countries.

On 2021-03-18 06:40:29, user CD wrote:

I have not read the full paper. Cautious comment: Recruitment between July and December is too large an interval. For example, if in one region most of the recruitment was done in July and in another most aas in December, this will affect the results.

On 2021-03-19 14:03:00, user Wendy Verelst wrote:

Aren't people living with -12y just younger than people without -12y old? -> less risk of hospitalisation .

On 2021-03-21 03:04:06, user Rick Shalvoy wrote:

Very encouraging data. This appears to be the textbook definition of a successful screening tool. Now that the U.S. FDA has finally released a template for device developers to use for EUA submissions when the developer is seeking to obtain a screening authorization, FDA authorization for OTC use of any properly validated device that screens for olfactory dysfunction should, and hopefully will, be granted relative soon after submission.

On 2021-03-24 00:09:50, user Elle wrote:

Also, how is it that patients were not broken out into smokers/non-smokers? All of these symptoms I think would be exacerbated by a smoking habit.

On 2021-03-24 15:21:57, user evacguy wrote:

I am pleased to annouce that this paper was accepted by the Journal of Travel Medicine following peer review on 11/02/21. It is noted that none of the findings, results or conclusions from the first draft have changed. The authors thank the reviewers for their insightful comments and suggested changes which improved our paper. The peer reviewed paper can be freely downloaded using the following link: https://doi.org/10.1093/jtm...

On 2021-04-05 17:44:22, user Armchair Hydrogeologist wrote:

Graph labels are swapped in Figure 1. Should be PepcidAC + Aspirin is 50% not other way around.

On 2021-04-06 17:16:29, user Rick Clem wrote:

I was infected in December along with my whole family. Loss of smell and<br /> a little lethargy was all we experienced. I have wondered if our luck <br /> was attributed to low loading factor or other. So I wonder on the <br /> degree of anitbody presence I attained from the infection. I received <br /> my 1st Moderna shot three weeks ago. Hit me like a freight train after <br /> 10 hours. Extreme fatigue, some headache. My thought is now directed <br /> to skipping my 2nd shot. Reading in the current studies on the <br /> necessity of a second shot, I hope they consider intensity of the <br /> previous infection in their studies. It would help folks like me to <br /> make a more informed decision on whether or not to ignore Fauci and the <br /> CDC's generalisms on needing a second shot.

On 2021-04-12 13:32:42, user H Arnold wrote:

Fantastic paper! What makes me a bit wonder is the discordance to the publications by Yost et al 2019 and Wu et al. 2020. Both report the replacement of T cells in the tumor (different entities) from external sources upon successful ICI.

Yost KE, Satpathy AT, Wells DK, Qi Y, Wang C, Kageyama R, McNamara KL, Granja JM, Sarin KY, Brown RA, Gupta RK, Curtis C, Bucktrout SL, Davis MM, Chang ALS, Chang HY. Clonal replacement of tumor-specific T cells following PD-1 blockade. Nat Med. 2019 Aug;25(8):1251-1259. doi: 10.1038/s41591-019-0522-3. Epub 2019 Jul 29. PMID: 31359002; PMCID: PMC6689255.

Wu TD, Madireddi S, de Almeida PE, Banchereau R, Chen YJ, Chitre AS, Chiang EY, Iftikhar H, O'Gorman WE, Au-Yeung A, Takahashi C, Goldstein LD, Poon C, Keerthivasan S, de Almeida Nagata DE, Du X, Lee HM, Banta KL, Mariathasan S, Das Thakur M, Huseni MA, Ballinger M, Estay I, Caplazi P, Modrusan Z, Delamarre L, Mellman I, Bourgon R, Grogan JL. Peripheral T cell expansion predicts tumour infiltration and clinical response. Nature. 2020 Mar;579(7798):274-278. doi: 10.1038/s41586-020-2056-8. Epub 2020 Feb 26. PMID: 32103181.

On 2021-04-12 15:56:55, user Philip Machanick wrote:

I am also curious how in a double-blind trial it comes about that someone in the control is given the test drug.

On 2021-04-26 03:17:28, user YingYing Irene Wang wrote:

The Journal of Deaf Studies and Deaf Education - Accepted for publication, DOI: http://doi.org/10.1093/deaf... --- online publication is under process. It will be online shortly. Thanks.

On 2021-04-27 09:15:58, user Ramy Ghazy wrote:

This manuscript describe the geospatial distribution of under-five mortality in Alexandria Egypt, moreover, we identified the main determinant of under-five mortality. We hope to help the health authority and stakeholders to decrease future increase in U5M.

On 2021-04-29 18:57:59, user Kirsty Short wrote:

Yes it has been accepted at CID https://pubmed.ncbi.nlm.nih...

On 2021-04-30 14:31:06, user Gustavo Bellini wrote:

congratulations on the study! it would be interesting if the dose of cholecalciferol and calcifediol used was reported. patients supplemented with Colecalciferol may have had less protection because they were supplementing with low doses, which were not sufficient to raise the levels of 25OHD to the ideal range, so that vitamin D performs its immunomodulatory functions at maximum level. it would also be very interesting if 25OHD levels were reported in the supplemented groups and in a sample from the control group.

it is also important to note that a daily dose of around 5,000 IU (person weighing> 50 kg) of cholecalciferol will cause the 25OHD levels to gradually increase and stabilize at around 50ng / ml only after 4 months. on the other hand, an attack dose of 600,000 IU of cholecalciferol in people with low levels causes the 25OHD levels to rise in 3 days to the optimum range. the level starts to drop after 15 days, and in order to stay in the ideal range, a daily (5,000 IU) or weekly (35,000 IU) supplementation with realistic doses should be started. if supplementation is not done continuously, the 25OHD levels fall back to around 20ng / ml in a 2-month interval.

• Daily oral dosing of vitamin D3 using 5000 TO 50,000 international units a day in long-term hospitalized patients: Insights from a seven year experience<br /> https://doi.org/10.1016/j.j...

• Effect of a single oral dose of 600,000 IU of cholecalciferol on serum calciotropic hormones in young subjects with vitamin D deficiency: a prospective intervention study<br /> https://doi.org/10.1210/jc....

On 2021-05-11 13:08:35, user Tomas Hull wrote:

There was no placebo group... <br /> If the same study was among the unvaccinated frontline health care workers, dealing with SARS-CoV2 patients, wouldn't most of them have at least detectable IgG and IgM titers??? <br /> Why not test the same group of people again 2-3 months later and see what the antibody titers are, if detectable at all...

On 2021-08-11 14:46:17, user Richard Bruce wrote:

This is a very informative study. The methods do not say how testing for infection was handled to ensure uniformity of testing frequency between the different cohorts. Given the retrospective nature, there may be a selection bias. If we assume that vaccination reduces symptoms (which is reasonable given many data points including this paper), we can also then assume that subjects will seek testing more frequently when symptoms are present than in the absence of symptoms. Therefore, given that unvaccinated will be more likely to experience symptoms following infection, unvaccinated subjects are more likely to receive testing when infected. This will bias the infection rates but should leave the hospitalization/ICU rates unchanged.

On 2021-08-12 21:07:51, user Sam Wheeler wrote:

What about Johnson Johnson = Janssen vaccine? And Sputnik V vaccine?<br /> Janssen recently said their vaccine gives better protection after 8 months compared to 1 month after vaccination. And Janssen was approved based on efficacy after only 2 weeks. So Janssen may be much more efficient than previosly thought!

On 2021-08-11 20:10:04, user nullcodes wrote:

State of residence seems like too big of a geographic area to use as a match criteria. Cities and rural may have been a better criteria. The rollouts within states were quite varied.

On 2021-08-14 17:36:32, user Matthias von Davier wrote:

Overclaiming, or the use of straight-lining is another option, as are other types of response biases.

In addition, the level of hesitancy for self-reported PhD (doctorate in the questionnaire) is at the same level of hesitancy seen in the group that chose to not answer the education question (missing education information also has 23.9% vaccine hesitancy).

On 2021-08-17 09:28:07, user One bird one cup wrote:

"Additionally, we assume the survey was completed in good faith." .... The assumption is what bothers me here. The people responding to a survey on Facebook aren't necessarily representative, as they're self-selected. This does not inspire confidence. Who's to say the respondents answered honestly about their education level? In addition -- apparently those who live in communities who were largely for Trump in 2020 appear to have been more vax-hesitant. I'm not a statistics person so I can't judge how the authors adjusted for this. But I feel hinky about it.

On 2021-08-14 17:57:29, user Hemaka Lakmal wrote:

How did you calculate the sample size required? How did you fix 282 individuals are enough for the test?

On 2021-08-16 05:24:54, user GT250ZX wrote:

I would take this research with a bit of grain of salt given the sample size of 282 observations. A much better study has been performed in Peru using 400,000 samples; the alleged level of efficacy is 50% per : https://www.reuters.com/wor...

On 2021-08-15 17:21:01, user carbsane wrote:

Can someone PLEASE explain to me how there can be 850 cases of COVID among the placebo group through March 13th if most of that group was subsequently vaccinated?? <br /> According to Pfizer's website they began unblinding and vaccinating in December (pretty much after the EUA), as they reported that as of Jan 29th 3,624 placebos had been FULLY vaxxed. Their last reported numbers (before dropping the information from their weibsite) were on Feb 24th by which time 16,904 had received at least one dose of vaccine.

On 2021-08-18 18:30:23, user Steve Kirsch wrote:

There were two people in the placebo group who got the drug after the unblinding. The paper never talks about the cause of death from those two people. This is EXTREMELY important. Does anyone know?

On 2021-08-22 00:51:09, user Anthony Loera wrote:

Supplementary data shows deaths between Vaccine people and controls are close to even<br /> https://www.medrxiv.org/con...

On 2021-08-06 22:23:54, user Ewin Barnett wrote:

The government of Scotland reported that 5,522 have died as a result of being vaccinated. No other data released like what percentage had comorbidities or were low on vitamin D at their time of admission to hospital.. No data released as to the appropriate percentage of the national population had been vaccinated. For a nation of about 5.5 million, this represents at least 0.1% risk.

On 2021-08-07 20:35:21, user vinu arumugham wrote:

Table S4 shows 4 deaths in the vaccine arm and 1 death in the placebo arm due to cardiac arrest. <br /> The probability that this outcome is a chance occurrence is 1.5%.

(((21999÷22000)^21996)×((1÷22000)^4)×(22000!))÷(21996!×4!) =0.0153 or 1.5%. <br /> So 98.5% chance that the vaccine CAUSED the excess cardiac arrest deaths.

Table S4 also shows 1 excess COVID-19 related death in the placebo arm.<br /> So to prevent 1 COVID-19 related death, the vaccine causes at least 3 deaths due to cardiac arrest.

On 2021-08-16 19:12:41, user Paul Wang wrote:

Abstract(results) needs to be fixed: first sentence gives the exact same name to 3 of the 4 subgroups

On 2021-08-18 14:12:25, user Jessica Lyne wrote:

I am curious if Dymista alters the testing results.

On 2021-08-20 22:49:31, user kdrl nakle wrote:

Poorly written paper, in bad need of rewriting.

On 2021-08-21 12:15:04, user muhamad khairulbahri wrote:

This preprint is now published at:<br /> https://wjst.wu.ac.th/index...

On 2021-08-22 14:07:20, user JB wrote:

have been approved for use in humans since December 2020

They have been authorized, not approved.

On 2021-08-22 19:34:47, user ingokeck wrote:

There is another issue with the article. You do not describe the data collection for the control group. Has this been done with the same preprocessing, using the same PCR arrays, i.e. are the values that the RT-PCR generates comparable even though there is one year difference in the sampling? Did you use an internal control for human DNA so you know that the sample collection was the same every time to get normalized RNA loads independent of the sampling procedure?

Even if the value generation is comparable, why do you think you can compare the Ct values from the D614G linage to the delta linage, given that it seems delta is more infectious than previous variants?

On 2021-08-23 10:16:41, user David States wrote:

Figure 1, panel C is key to much of the discussion. I’d like to see the actual data points as well as the fit curves. Also the units on the x-axis, genome equivalents per mL, are calculated from Ct using a proprietary undocumented formula and are not used elsewhere. I’d like to see a second x-axis labeled in Ct.

On 2021-08-23 21:24:22, user Stuart wrote:

Published in JCI Insight, here: https://insight.jci.org/art...

On 2021-08-04 02:30:29, user Deplorably Black wrote:

Interesting. Does this still apply considering<br /> the new variants?

Has a study been conducted as to the vaccines effect on long COVID?

I suffered from daily headaches post COVID for 8 months. They stopped immediately after vaccination.

I know several others with the same experience.

That in itself made vaccination post COVID worthwhile for quality of life.

On 2021-08-09 20:32:15, user KS wrote:

I haven't scrutinized this paper but even if all the results are accepted, the one-line "Conclusions" at the beginning is highly problematic without qualifiers. "Unlikely to benefit"? This study is limited, so the conclusion can't be so broad. Are elderly or long-haulers unlikely to benefit? You've got a 42 day window (seems arbitrary) so if you got the disease 6 months prior are you unlikely to benefit from the vaccination? The study doesn't address any of these things, yet makes a huge leap in its conclusion. This is a pre-print so PLEASE make a conclusion that fits your experiment and data. Because it's likely the only thing the general public will read and it will become the basis for more misinformation.

On 2021-09-11 12:19:37, user William Brooks wrote:

This study finds similar results to studies looking at infections among South Asians in England [1] and foreign workers in Kuwait [2]: lockdown heightened the curve for groups with more crowded living conditions. The results also agree with those of the nearest thing we have to a lockdown RTC: higher secondary attack rates in asylum centres that mass-quarantined all residents in Germany [3].

Despite this, the authors claim lockdowns work. Like a pharmaceutical intervention, for a non-pharmaceutical intervention to be said to work, the intervention group (e.g. NY, CA) has to show significantly lower mortality and morbidity than the control group (e.g. FL, SD), which isn’t the case [4]. Also, for extremely authoritarian interventions to justify their many negative side-effects, hospitals in the control group would need to be overflowing like the models predicted, which has never come close to happening.

[1] https://doi.org/10.1016/j.e...<br /> [2] https://doi.org/10.1186/s12...<br /> [3] https://doi.org/10.1101/202...<br /> [4] https://doi.org/10.1101/202...

On 2021-09-12 05:10:38, user kdrl nakle wrote:

This is really a weak study. It jams together 4 vastly different vaccines and then reports without differentiation. So what do we know out of that? It could be one has 100% protection and another 50% and we would claim it is 75% overall?

On 2021-09-12 06:41:02, user kdrl nakle wrote:

Air outlets versus inlets. Very important though not surprising information. The rest of this paper is also important info.

On 2021-09-14 06:20:47, user Mike Hawk wrote:

A friendly grammar edit to the study, in the abstract section. Instead of, "One possibility is that such negative outcomes...while some tend respond with empathy (feeling what others feel), others tend respond with compassion (caring about what others feel)", I suggest that it should have been "One possibility is that such negative outcomes ..: while some tend to respond with empathy (feeling what others feel), others tend to respond with compassion (caring about what others feel)."

On 2021-09-14 08:36:03, user Dharshana Kasthurirathne wrote:

one of the conclusions is that the younger people are 17 times more likely to get hospitalized if they are not vaccinated, compared to those that are fully vaccinated. however, if you think of the time period compared (jan-jul), those who were unvaccinated may have had much higher chance of encountering the virus (simply cos they were in that status for a longer time) compared to those who are fully vaccinated (who were in that status for a much smaller time period). it's safe to assume that those who are fully vaccinated (particularly younger people) changed to that status quite recently. so is it correct to do such a population wide comparison without normalizing for the time since acquiring the vaccinated or unvaccinated status?

On 2021-09-17 16:24:24, user Thom Davis wrote:

"Antibody neutralization titers against B.1.351 and P.1 variants measured by SARS-CoV-2 pseudovirus neutralization (PsVN) assays before the booster vaccinations, approximately 6 to 8 months after the primary series, were low or below the assay limit of quantification" is the key "real information" in this synopsis. All others are conjecture. If it isn't measurable, it isn't there.

On 2021-09-21 11:11:33, user 4qmmt wrote:

Would any of you agree that rate of myocarditis/pericarditis due to the vaccine in youth is a) unknown and b) higher than the available data suggest?

On 2021-09-10 16:16:11, user bee researcher wrote:

To clarify, this work is not comparing the risk of myocarditis in vaccinated individuals with the risk of hospitalization in similarly aged COVID-positive individuals, but rather an age-matched demographic regardless of COVID infection status. Is that correct?

This seems misleading in terms of risk assessment, because it's comparing the risk after a specific event (vaccination) with the background level of risk over certain periods of time. Yet active spread of COVID appears likely to continue for at some level for years, and the risk of hospitalization in COVID-positive individuals in this age group is much higher than the risk of vaccine-related myocarditis. Indeed the risk of COVID-related myocarditis is higher in this age group than the risk of vaccine-related myocarditis. If eventual infection by a now-endemic COVID-19 is incredibly likely, than it seems more informative to compare the risks associated with such an infection with the risks of vaccination.

On 2021-09-10 16:30:40, user Roger Seheult wrote:

Why did you choose to compare vaccine related CAE with COVID hospitalizations? Why not compare Vaccine related CAE with COVID CAE? Are we comparing apples to apples? Hospitalizations in pediatric population (especially before delta) was rare from COVID-19 for any reason and most vaccine CAEs do not require hospitalization as I understand.

On 2021-09-10 19:11:30, user David Goldberg, MD, MSCE wrote:

Although the scientific question that is being address is an important one, I have concerns about the methodology used to adjudicate the outcome. In similar circumstances (e.g., the FDAs Mini-Sentinel Initiative), complex clinical outcomes like this (e.g., acute liver failure) were adjudicated independently by two experts, with a third person serving to break any ties. That seems not to have been done in this study, as there was only one cardiologist involved. Secondly, the clinical data to adjudicate the outcome of myocarditis seems to be insufficient in many cases. Although one could argue "this is the best data we have" sometimes that is not good enough. When the question is so important and politically charged, incomplete/invalid data is sometimes worse than no data. Unless the authors can have two-party adjudication with record review, and classification using standard techniques (e.g., definite vaccine-induced myocarditis, highly likely, probable, possible, not) then there are major methodological concerns with the outcome, and the overall validity of the study.

On 2021-09-10 21:38:17, user anime profile picture wrote:

This study completely misses the point of young kids getting vaccinated. COVID is infectious. Meaning when someone gets the virus, it can be passed on. Whether or not they are at high risk relative to the adverse side effects, they should be vaccinated to reduce the probability of older, more at-risk people from getting it. In short, young boys should get vaccinated to protect them, their parents, their teachers, and their grandparents. Consult with your doctor of course. I am no medical professional, but I understand that a vaccine does more than protect the person being vaccinated.

On 2021-10-14 23:01:32, user doseofcommonsense wrote:

Public Health has estimated Delta's CFR to be 0.096 %.

On 2021-09-02 12:51:16, user David Curtis wrote:

I have a few comments.

Figure 2B suggests there is quite a lot of inflation of the test statistic.

Some genes will have an excess of variants in controls rather than cases. This means it makes sense to plot a signed log p (SLP), rather than a minus log p (MLP), in which a negative sign is given if there is a an excess of variants in controls. This is what I did in my study of the first 200.000 exome sequenced subjects:<br /> https://journals.lww.com/ps...

Plotting the SLP rather than the MLP makes it easier to detect possible problems with the analyses, such as inflation of the test statistic in one direction.

The result for SCL2A1 is based on a total of 52 carriers. As far as I can work out, 10 of them are cases and 42 of them are controls. So the claim that SCL2A1 is involved in depression aetiology is really based on the fact that damaging missense variants are observed in 10 cases. With such small numbers, regression analyses may give unreliable p values. In fact, I did a simple Fisher's exact test (not including any covariates) and this yields a p value of 4.027e-05, which falls just short of exome-wide significance.

I wonder if the test statistic is inflated because there are genes in which there is a slight excess of variants in cases but the methods used tend to produce p values which are too low, because of small numbers, as seems to be the case for SCL2A1.

The other thing I would say is that the estimated OR for SCL2A1, 6.01, does seem to be surprisingly high. I would not have expected that damaging missense variants, grouped together as a class, would have such a large effect.

On 2021-09-05 16:07:08, user JimmyJoe6000 wrote:

Someone posted an inception to date chart using daily deaths for the two groups of countries. I can't see to find in in any of the articles like this. It mentioned John Hopkins along with this link. <br /> Anyone have the link to the chart?

On 2021-09-07 01:37:42, user Simon Turner wrote:

This paper has now been peer reviewed and published at BMC Medical Research Methodology:

Turner, S.L., Forbes, A.B., Karahalios, A. et al. Evaluation of statistical methods used in the analysis of interrupted time series studies: a simulation study. BMC Med Res Methodol 21, 181 (2021). https://doi.org/10.1186/s12...