If you’re like me, you raise a skeptical eye when @apsmunro is happy. So I took a look at this study and I’m not embarrassed to say: this myth is far from busted. 🧵
The study compares symptoms recorded by an app from periods of Alpha dominance to the period of Delta’s dominance. It purports to show Delta is no worse than Alpha when it comes to kids: either in terms disease burden (number of symptoms) or duration. medrxiv.org/content/10.110…
Describing the symptoms tracked, the authors tell us they “do not include some common paediatric co-morbidities (e.g., neurological or neuro-disability disorders).”
This is worrying. If there's no tracking of common pediatric comorbities, that exclude the possibility of Covid exacerbating them leaves open whether Delta does this more than Alpha. Nonetheless, the authors go with “similar symptom burden”. I’m no MD, but that seems premature.
The authors then describe the study’s inclusion criteria.
“As previously, children were considered to have COVID-19 if proxy-reported with relevant symptoms between two weeks before and one week after SARS-CoV-2 infection confirmation.”
The study doesn’t include enough kids to really see cases of MIS-C, but would likely be miss MIS-C anyway given that MIS-C can come after asymptomatic infection. nytimes.com/2021/04/06/hea…
Further, MIS-C occurs weeks later so either by way of the time frame, or by way of requiring early symptoms, MIS-C gets excluded. But “similar symptom burden” is what we’re going with. cdc.gov/mmwr/volumes/6…
The authors tell us they had no information on variant strain, inferring it instead dominant strain at time of infection. Still, the authors go with “similar symptom burden” despite similarity of symptoms made more likely artificially by mixing. But who am I to question doctors?
Given that previously Alpha proxy-reporting was “assiduous”, the authors assume it continues to be.
I'm going to go out on a limb here but, if Delta were worse, parents might stop reporting. Mind you, this might be spitballing but, as a mom, I can tell you I would not be screwing w/an app if my kid were really sick. But maybe that’s just me. Anyway, “similar disease burden”.
Now, since I’ve gone out on the limb already, I’m just going to ask: Why is disease burden a function of the number of symptoms? For one, consider fever. Surely if Alpha causes low grade fevers but Delta causes fevers to the point of delirium, that would make Delta more severe.
In fact, it may well be that intensity of symptoms is what is behind the differences in hospitalization frequency: 2.2% of young children with Alpha are hospitalized vs. 3.5% with Delta.
Secondly, what if some symptoms are worse than others and the worse symptoms appear more frequently in Delta? The authors tell us “headache, rhinorrhoea sore throat, dysosmia, fever, dizziness, ‘chills or shivers’, eye soreness and hoarse voice" are more common with Delta.
Whether it’s worse to have some of these rather than some of the others on the list, I’ll leave the reader to judge. But as before, the authors go with "similar disease burden."
The authors tell us that depending on age, Delta adds 1 additional symptom to Alpha’s 3 for younger kids and 4 for older kids.
While the authors are unmoved from the refrain, “similar disease”, is it really? Ailing already with a selection from the listed medley of maladies, adding one more seems a significant step in the wrong direction and not “similar disease burden” which is what the authors go with.
The author’s final point on symptom burden is to let us know that “any novel symptoms unique to Delta variant would not be capured..”
Really; is everything so different in Medical school that you get to say the burden of Alpha is similar to Delta even if you don’t know whether Delta caused anything new or what it is?
I’m just saying that my PhD thesis advisor would have tossed me out of the room if I told him Aristotle was like Hegel if you just ignore things unique to Hegel.
Moving to symptom duration, we are told that rates of Covid lasting beyond 28 days are low.
But there are issues in concluding much about the relative risk of long covid from the study. The study began May 26 and ended 28 days after the last included patient, August 5. The authors admit this limits the study to children “still unwell at 28 days” but not beyond.
But in that case we unable to determine if Delta might be worse by causing a more peristent (beyond 28 days) form of Long Covid than Alpha.
In any event, as indicated at the bottom of in Table 1, there does seem to be a sight uptick in Long Covid from Alpha to Delta (1.7 vs. 2.1) largely due to the youger group going from an incidence rate ot 0.7 to 1.3.
At just over two months, this study also risks missing long covid cases that consist of periods of remission followed by relapse; a pattern that might be more common with Delta, though that wouldn’t be known given the parameters. vox.com/2020/6/4/21274…
Perhaps aware that the 28 day cut-off raises issues, the authors claim extending it would be a “bias”. Really; is that how this works?
It’s a bias if you extend the study to see if Delta is worse because it lasts longer, but it’s not a bias if you just cut both studies off at the same endpoint so you don’t know whether or not there are differences beyond 28 days?
Here’s my guess: this isn’t an issue of bias, but a shortcoming of the study. Admittedly, “bias” sounds better though.
The authors begin wrapping things up by telling us “the pandemic has been extremely disruptive for UK schooling” and catalogue lockdown dates and attendance compromised by “quarantines” studiously avoiding the possibility that being sick has some role to play in attendance.
In fact, the authors discussion of attendance differences between Alpha and Delta talks brings up that routines may have differed between study periods, along with exposures (presumably referring again to quarantines) and testing (that well known cause of absences).
So back on the limb I’ve gone out on, & repeating I'm not an MD, so this is a guess, but Delta being more transmissible probably means it also had a greater impact in terms of actual illness too, a fact these MDs seem intent on ignoring. But what do I know–I’m just a layperson.😉
• • •
Missing some Tweet in this thread? You can try to
force a refresh
.@ShiraDoronMD et al argue against universal masking in healthcare based on 2 claims: 1) transmission risk from HCWs w/out symptoms is low 2) vaccination is sufficienly protective against hospital-acquired infection.
These claims are both knowably, provably wrong.
@AnnalsofIM
🧵
To establish their first claim, that the risk of transmission from HCWs without symptoms is low, the authors cite two sources: a) Killingley and b) Tayyar.
But a) Killingley does NOT show presymptomatic transmission risk is small. What Killingley actually shows is that, what the authors pass off as "minimal presymptomatic shedding," is consistent with **44% of transmission occuring presymptomatically**.
In honor of @TracyBethHoeg’s new anti-mask propaganda/preprint, I compiled her Tweets on mask studies into a textbook I’m calling, “Confounders: A Matter of Convenience.” It’s an expose of Hoeg’s bad faith hypocrisy. 🧵
Here’s the title page with a table of contents.
CHAPTER I:
HOW TO PROTEST PRO-MASK STUDIES
It's easy! Just complain the conclusion doesn’t hold because the data is confounded! 1. Eg. The Boston Mask Study
The moral panic over school closures has left us with problems more intractable than they were pre-pandemic because now, to solve those problems, we first have to dispel lies. That school is a preventative to suicide is a reprehensible distortion, but it is not the only one. 🧵
There is the hysteria over how closures hurt minorities the most which obscures that, for minorities, school is the source of a problem: the school-to-prison pipeline. nytimes.com/2020/10/28/opi…
There is the hysteria that closures are causing obesity which obscures the problem of fatty, nutritionally bankrupt cafeteria food. nypost.com/2019/11/16/the…
Following a lead in @mehdirhasan's receipt-riddled expose, I looked into FL's deadly summers but in terms of excess deaths (Hasan uses C0VID deaths) in the 10 US states with the highest percentage of seniors. Tl;dr: DeSantis won't be using these stats on the campaign trail. 🧵
The long list of concerns downplayers coopted for the sake of opening schools and then quickly abandoned: learning loss that didn't carry over to C0VID related cognitive declines, newsinfo.inquirer.net/1639956/omicro…
Shenoy et al urge abandoning universal masking on the grounds masks have little benefit & some harm. Yesterday I showed they're wrong about benefits. Today I show they're wrong about harms. Their strongest evidence favors masks. The rest has little relation to their ambitions. 🧵
In making the case that masks harm, the authors use three sources.
In Cormier et al patients & providers rate their masked & unmasked encounters in terms of communication difficulties. Using scales from 1-5 (patients) or from 1-6 (providers) participants are asked about eg. listening effort, ability to connect, understand & recall.