.@NYTimes has done great data journalism on #SARS_CoV_2, filling gaps left by an incompetent US federal response but their recent coverage re: schools is tipping towards sensationalism & divergence from data... nytimes.com/interactive/20…
This coverage is totally freaking people out once again with the idea that preschool-aged children are super-spreaders, which is NOT TRUE! 2/
There’s a lot we don’t know but the overwhelming weight of global data from household, school, & pop-surveillance studies indicates that small children are less likely to contract & spread this virus. Not zero risk but surprisingly low. dontforgetthebubbles.com/the-missing-li…
The study used data from March/April (remember trying to get tested then?) & compared #SARSCoV2-positive mildly *symptomatic* kids from a *tertiary care* children’s hospital to symptomatic adults also identified at that *children’s* hospital 🤔 jamanetwork.com/journals/jamap…
To totally rip off @nataliexdean THINK LIKE AN EPIDEMIOLOGIST
If I’m examining a disease where kids tend to have mild or no symptoms, & sampling kids from a tertiary care children’s hospital (even excluding the n=7 sickest kids), I expect kids to be the sickest in the city 5/
If I’m examining a disease where kids tend to have mild or no symptoms, & sampling kids from a tertiary care children’s hospital (even excluding the n=7 sickest kids), I expect kids to be the sickest in the city 5/
And if I’m comparing them to the tiny number of adults also tested at my *children’s* hospital (n=48 aged 18-65 yo), I’m going to think these adults are a weird sample (I’d love to see age distribution & other data) and probably not as sick as others 6/
In other words, I’d be all up in an #epitwitter snit about selection bias (forcing my students to draw more DAGs) and would NOT make the claim that “Infected children have at least as much of the coronavirus in their noses and throats as infected adults” as @nytimes did... 7/
This is NOT any a representative study of infected children. This is NOT a representative study of infected adults. This a selected study of probably high-risk symptomatic kids and an I-DONT-KNOW-WHO sample of few adults tested at a children’s hospital. 8/
And this line galls me: “on Thursday, a study introduced an unwelcome wrinkle into this smooth narrative” [that young children are mostly spared by #SARSCoV2 and don’t spread very often] *sigh* #Epitwitter is not just making up data for a “smooth narrative”! This is serious!
I’m not going to even get into the fact that viral load is not a specific marker of infectiousness so of limited utility for making conclusions about how infectious kids are ... medrxiv.org/content/10.110… 10/
I suspect author pushing back against US Right’s politicization of school reopening, e.g., risks negligible (FALSE), desperately needed extra federal $$ for #PPE, ventilation improvement not imp (FALSE). But amplifying risks past data also polarizes.Getting this right is too imp!
Now on to today’s headline which got me into a snit... First, this is stunning data journalism and #scicomm. @nytimes is doing US a real service... AND YET... nytimes.com/interactive/20…
“The estimates assume that children are as likely to carry and transmit the virus as adults — “a large assumption, given the unknowns about children,” said Spencer Fox, a member of the research team.” *sigh* 13/
Props to @UTAustin’s Spencer Fox for trying to be clear about this big limitation (This is meant to be a rough guide, a first step,” Dr. Fox said.) But these graphics are going to be read by admininstrators and parents as actual numbers... 14/
And some are going to decide that 5- and 6-yos will miss a year of in-person schooling bc of numbers thst might be vastly overstating risk among this group. 15/
People in US who care about children’s well-being realize that most places have no good options. Bc of our nat’l leadership, we are choosing among bad options. But to be clear, depriving young children of public education is also an awful, awful option (hard to overstate) 16/
If 5 yos truly are highly infectious, then that awful sacrifice could be warranted. And we as a society will have to gird ourselves for the longterm consequences. 17/
But if #SARSCoV2 has shown us the tiny mercy of lower infectiousness & illness among young children (which high-quality epi studies seem to indicate), then we should marshall our financial and scientific resources to make sure our 5, 6, & 7 yos get safe, quality in-person ed
Sigh. Again, there are no good choices here. While the #SARSCoV2 lit is still evolving, the literature on longterm consequences on kids of interruptions in schooling and economic & social stresses on families is depressingly clear. It breaks my heart every day /19
I am going to stop writing now. I’m late getting my kids to daycare. End/
