SARS-CoV-2, 19 times less likely to infect people under 15

Link post

Is a small vol­ume of low qual­ity data suffi­cient for mak­ing mean­ingful claims about CFR and R0 ?

Epistemic Sta­tus: Wild off-the-cuff speculation

Note: This post was writ­ten on 24-03-2020 dur­ing the srars-cov-2 pandemic

DISCLAIMER !

This is NOT an ar­ti­cle claiming that we should stop car­ing about the covid-19 pan­demic, nor is it claiming that the num­ber pre­sented in the ti­tle is ac­tu­ally cor­rect, rather, it’s a punny at­tempt to show the kind of data-driven un­cer­tainty that leads to us not fully un­der­stand­ing the be­hav­ior of the dis­ease in terms of spread and fre­quency of symp­toms.

If you wish to cru­cify me for claiming “Maybe the cur­rent mea­sure we are tak­ing are not ideal based on what we know so far”, please do me the fa­vor of ac­tu­ally read­ing the whole thing first.

NOR is this an ar­ti­cle claiming that our cur­rent mea­sures are “kil­ling the econ­omy at the cost of sav­ing lives”… I find that po­si­tion al­most ab­surd and lack­ing in un­der­stand­ing about this whole “so­cial con­tract thing” we have go­ing that means we value life over ma­te­rial goods.

The very un­cer­tain hy­poth­e­sis I am pre­sent­ing here is that we are en­gag­ing in a “panic re­sponse” which might be un­jus­tified given cur­rent data and could lead to both a higher num­ber of peo­ple dy­ing from coivd-19, higher num­bers of peo­ple dy­ing over­all and worse eco­nomic dam­age than the proper re­sponse.

Lastly, and very im­por­tantly, I am not an MD nor an epi­demiol­o­gist and tak­ing a though piece as a guide for de­cid­ing to iso­late your­self or to break quaran­tine is pretty stupid. If this en­courages any ac­tion other than “Hmh, in­ter­est­ing, let me try to find some data and look at it”, then I will claim right now said ac­tion is mis­guided.

Are chil­dren less likely to be in­fected by srars-cov-2 ? (No)

It’s fairly hard to get raw data about covid-19 cases, the mos com­pre­hen­sive dataset of in­di­vi­d­ual cases I could find is com­piled by the med­i­cal com­mu­nity at DXY.cn (a Chi­nese med­i­cal dis­cus­sion board):

https://​​docs.google.com/​​spread­sheets/​​d/​​1jS24DjSPVWa4iuxuD4OAXrE3QeI8c9BC1hSlqr-NMiU/​​edit#gid=1187587451

I find it cu­rios that out of 397 cases with a re­ported age in the US, only 4 were un­der the age of 15.

I say I find this cu­rios be­cause based on the 2018 cen­sus data, 18.71% of the US pop­u­la­tion is un­der 15 years of age.

This leads me to the click­baity num­ber in the ti­tle (round(0.187/​(4/​397)) == 19).

How­ever, I did not pick the US speci­fi­cally to get this high of a num­ber, in­deed, China yields an even higher one. I sim­ply thought 19 had a nice ring to it con­sid­er­ing the dis­ease’s name… and the whole point of this ar­ti­cle is to ex­plain why this ob­ser­va­tion is blatantly wrong any­way.

While this par­tic­u­lar coro­n­avirus is in­deed novel, coro­n­aviruses in gen­eral are a pretty wide spread species and some of them have symp­toms we la­bel un­der “com­mon cold”, to­gether with the origi­nal SARS and the MERS coro­n­avirus, I think enough data to say with a very high cer­tainty:

  • The claim that it’s less likely for a kid to get in­fected by a coro­n­avirus is bul­lshit, most cer­tainly so when the num­ber claimed is as high as 19 times less likely.

Even more so, kids are par­tic­u­larly bad at not lick­ing ran­dom sur­face, touch­ing strangers, eat­ing ran­dom stuff off the ground and over­all do­ing stuff that is un­safe in terms of not get­ting in­fected by al­most any given virus or bac­te­ria.

So why are so few kids get­ting in­fected ?

Well, you could say:

  • It’s be­cause kids are less likely to be in tightly packed places (e.g. nurs­ing homes, planes)

  • It’s be­cause kids are less likely to be in places where in­fected peo­ple reside (i.e. hos­pi­tals and triage cen­ters)

  • It’s be­cause the data of pa­tients un­der 18 is less likely to be made fully available (how­ever, I could be mak­ing the same point for peo­ple un­der 65 or un­der 40 here, it’s just that the num­ber would be lower)

But I could say:

  • Kids are more likely to be in tightly packed places with some of the low­est in­ter­per­sonal san­i­tary norms on earth (e.g. kinder­gartens and schools)

  • New­borns kids are very likely to be in hos­pi­tals (all be it in the ma­ter­nity wards) and there doesn’t seem to be a bump for newborns

  • Kids, un­like adults, are al­most always nearby other adults. I can iso­late my­self alone in my house, but most kids in this de­mo­graphic would still need at least one guardian near them at all times.

I think the real ex­pla­na­tion is more along the lines of:

  • Kids have a very well func­tion­ing im­mune sys­tem. Our im­pres­sion to the con­trary comes par­tially from niche viruses ala in­fluenza who pro­voke symp­toms via caus­ing an over-re­ac­tion from the im­mune sys­tem. Par­tially from the fact that kids don’t have anti-bod­ies ready for a host of in­fec­tions to which we have built im­mu­nity. Par­tially be­cause kids are gross and don’t un­der­stand the con­cept of per­sonal space or not eat­ing dirt and are thus more prone to come into con­tact with large quan­tities of pathogens. But over­all I would trade and arm and a leg to have the lym­phatic sys­tem of a 14 year old again.

  • The tests used to mon­i­tor sars-covid-2 are usu­ally the cheap and quick kind, which have sen­si­tivi­ties as low as 70% in the first 4-10 days of the in­fec­tion. But of course, this mea­sure is mainly based on test­ing adults (mainly older adults), it might well be that a young per­son (read: an or­ganism with ex­tremely well func­tion­ing in im­mune sys­tem, cell MHC1 and MHC2 pro­tein pre­sen­ta­tion mechanism and cell apop­to­sis mechanism) is able to stop the in­fec­tion in it’s tracks much quicker be­fore IgM or IgG lev­els get no­tice­ably high.

  • Many of the peo­ple tested for the dis­ease are those pre­sent­ing to hos­pi­tals, those peo­ple are pre­sent­ing to hos­pi­tals be­cause they are hav­ing symp­toms that bother them. Even if some­one un­der 15 has a no­tice­able raise in IgM or IgG lev­els, the im­mune re­ac­tion might be enough in or­der for symp­toms not to go past a mild fever and some cough­ing… which are ba­si­cally om­nip­re­sent in young kids dur­ing the spring, so no­body will bring them to a hos­pi­tal for those rea­sons alone.

How­ever, if this is the case, this rea­son­ing need not ex­tent to chil­dren alone. All of the rea­sons listed above are not spe­cific to chil­dren, it just so hap­pens that well func­tion­ing im­mune sys­tem are overly rep­re­sented in chil­dren.

There’s plenty of older peo­ple with well func­tion­ing or­ganisms that could re­act to an in­fec­tion similar to how most chil­dren could. Although this breaks down past a cer­tain age, since our thy­mus is slowly re­placed by adi­pose tis­sue in the mor­bid, painful and not yet re­versible pro­cess we call “ag­ing”.

In some ways this ar­ti­cle is a “lay­man’s defense for the lay­man” of this anal­y­sis by Oxford’s cen­ter for ev­i­dence-based medicine: https://​​www.cebm.net/​​global-covid-19-case-fatal­ity-rates/​​. Which claims CFRs (case fatal­ity rates) way bel­low what’s be­ing cur­rently re­ported and places much more of the blame or co­mor­bidi­ties.

That is to say, this is hope­fully an in­tu­ition as to why “A lot of peo­ple in re­tire­ment homes dy­ing in Berg­amo” doesn’t trans­late to “This virus will kill 1-5% of Earth’s pop­u­la­tion”.

It’s also hope­fully an ex­pla­na­tion of why R0 (the ba­sic re­pro­duc­tive rate) ranges from 1.4 to 3.9 (which is huge, even com­pared to the ranges we get for most sea­sonal in­fluenza strains) and thus it could re­ally be any­where from “if left unchecked it will spread like a venge­ful me­dieval plague” to “it spreads more or less like the flu”.

Two things that keep me awake at night

How­ever, it does leave me cu­rios as to two things.

I’ve seen the CFR and R0, or var­i­ous facts de­rived from them, used by physi­ci­ans and au­thor­i­ties in terms of com­mu­ni­cat­ing in­for­ma­tion to the pub­lic and en­act­ing quaran­tine mea­sures.

Th­ese two things would heav­ily in­fluence how that in­for­ma­tion is com­mu­ni­cated and how said mea­sures are en­cated.

1.

If the data for this dis­ease is so poor (e.g. re­spectable in­sti­tu­tions have num­bers or­der of mag­ni­tude differ­ent for the es­ti­mated CFR), why go ahead and make claims based on R0 and CFR to be­ing with ?

I’m not claiming the dis­eases is not dan­ger­ous, on the con­trary, I am claiming that it could be po­ten­tially MUCH MORE dan­ger­ous than we think or MUCH LESS dan­ger­ous than we think.

That’s an is­sue, be­cause that’s how you get to un­der-re­act­ing and caus­ing mil­lions of deaths or over-re­act­ing and caus­ing mil­lions of death (due to a hor­rible re­ces­sion, even though they won’t be in Europe or the USA, they will be in Africa, as peo­ple start tight­en­ing the belt and dona­tions to char­i­ties are among the first inessen­tial things to go away).

Would the wise thing here not be to try and gather more data as ag­gres­sively as pos­si­ble ? Rather than go­ing on TV and mak­ing claims around num­bers for which no de­gree of cer­tainty ex­ists.

2.

If the cur­rent data in­di­cates that kids and most likely healthy young adults are not at risk, while old peo­ple and peo­ple with co­mor­bidi­ties are at very high risk. Why not act along those lines:

Allow young peo­ple that want to break quaran­tine to do so, mon­i­tor them closely and test them in or­der to figure out what the con­se­quences of spread­ing in a young de­mo­graphic is… af­ter all, it seems that they aren’t very sig­nifi­cant based on cur­rent num­bers.

If the con­se­quences seem wor­ri­some, re­vert the mea­sure be­fore hos­pi­tals get over­whelmed.

If the con­se­quences seem on par with a bad strain of rhinovirus or in­fluenza, pro­gres­sively ad­vise more and more peo­ple that they can break iso­la­tion (but make sure to in­form them that the risk of death ex­ists and that they will be as­sum­ing it), thus get­ting the econ­omy par­tially back on track and in­creas­ing the co­hort you are mon­i­tor­ing.

On the other hand, if we are petty cer­tain that old peo­ple and peo­ple with cer­tain co­mor­bidi­ties (e.g. di­a­betes and CVD) are at very high risk. Why not tar­get those peo­ple di­rectly in terms of aid, quaran­tine mea­sures and mes­sag­ing.

Why do we have broad sweep­ing mes­sages about “stay at home”, which seems to be mainly caus­ing a panic in young well-ed­u­cated adults (aka the kind of healthy de­mo­graphic that is prone to lead­ing an overly-healthy lifestyle which could lead them to not even notic­ing an in­fec­tion) whilst not at all dis­turb­ing grandma and grandpa (aka the kind of peo­ple you re­ally re­ally re­ally want to stay at home) ?

Wouldn’t a func­tion­ing econ­omy bet­ter al­low for long-term quaran­tine of those at risk ?

I’d per­son­ally be pretty glad to go around shop­ping for a few old peo­ple in my lo­cal com­mu­nity and pass on the sup­plies to a guy wear­ing a haz­mat suit that will dis­in­fect them to the best of his abil­ities and then hand them over.

I feel like this kind of “al­low young vol­un­teers to boost herd im­mu­nity, keep the econ­omy go­ing and provide data, at least un­til our data tells us oth­er­wise” ap­proach would be prefer­able to the pro­gres­sively wors­en­ing quaran­tine that is not dis­crim­i­nat­ing based on age or co­mor­bidi­ties.

But

Again, I don’t think based on cur­rent data I am qual­ified in terms of propos­ing the kind of mea­sure we should en­act.

I do how­ever strongly think that any mea­sures should re­volve around “how do we gather bet­ter qual­ity data and mak­ing it more widely available for peo­ple to build mod­els and hy­poth­e­sis based on”.

I also think that throw­ing any­one in the same bucket in terms of risk and thus throw­ing them in the same bucket in terms of the mea­sures we en­act is silly.

To me it seems that young healthy peo­ple are:

  • An is­sue be­cause of asymp­tomatic spread of the diseases

  • An as­set be­cause if they gain im­mu­nity they could help with blood dona­tions for plasma therapy

Whilst old or un­healthy peo­ple are:

  • An is­sue be­cause of the sever­ity of symp­toms and risk of death

  • An as­set be­cause they provide an easy way to track spread (since they are more likely to be symp­tomatic)

  • An ideal tar­get for pro­tec­tive mea­sures if our goals it to re­duce cases with se­vere symp­toms and fatality

Ob­vi­ously, “young” and “old” or “healthy” and “un­healthy” are pretty vague terms, but we can pick some ar­bi­trary limits, e.g. you are “young and healthy” if you are bel­low 25, ex­er­cise daily, have no di­ag­no­sis of metabolic dis­eases, CVD, CVD risk fac­tor or res­pi­ra­tory dis­ease… start from there, and broaden or tighten the defi­ni­tion based on the data we gather.

I think some peo­ple might read some form of “moral haz­ard”, af­ter all out of those peo­ple a few might still get se­vere symp­toms or even die from the dis­ease… but it’s not like the cur­rent ap­proach has stopped deaths and it’s not like there’s an end in sight. We can’t just quaran­tine peo­ple for­ever and we can’t hope that the kind of quaran­tine mea­sures we en­act are go­ing to com­pletely empty all hu­man and (po­ten­tial) an­i­mal reser­voirs.

At most, what we are do­ing right now is stal­ling for time in or­der to get a cure… but what is a cure ?

An­tivirals sel­dom have the effi­cacy we’d be look­ing for.

A vac­cine seems to be 12 months away un­der the most op­ti­mistic es­ti­mates and is far from a guaran­tee.

The two things that ev­ery­one agrees would help are:

  • Plasma from peo­ple with immunity

  • More data about the virus (sars-cov-2) and the dis­ease (cov-19)

Those are the very things the poli­cies in the US and in most Euro­pean coun­tries are not go­ing to help us with.

DISCLAIMER !

This is NOT an ar­ti­cle claiming that we should stop car­ing about the covid-19 pan­demic, nor is it claiming that the num­ber pre­sented in the ti­tle is ac­tu­ally cor­rect, rather, it’s a punny at­tempt to show the kind of data-driven un­cer­tainty that leads to us not fully un­der­stand­ing the be­hav­ior of the dis­ease in terms of spread and fre­quency of symp­toms.

If you wish to cru­cify me for claiming “Maybe the cur­rent mea­sure we are tak­ing are not ideal based on what we know so far”, please do me the fa­vor of ac­tu­ally read­ing the whole thing first.

NOR is this an ar­ti­cle claiming that our cur­rent mea­sures are “kil­ling the econ­omy at the cost of sav­ing lives”… I find that po­si­tion al­most ab­surd and lack­ing in un­der­stand­ing about this whole “so­cial con­tract thing” we have go­ing that means we value life over ma­te­rial goods.

The very un­cer­tain hy­poth­e­sis I am pre­sent­ing here is that we are en­gag­ing in a “panic re­sponse” which might be un­jus­tified given cur­rent data and could lead to both a higher num­ber of peo­ple dy­ing from coivd-19, higher num­bers of peo­ple dy­ing over­all and worse eco­nomic dam­age than the proper re­sponse.

Lastly, and very im­por­tantly, I am not an MD nor an epi­demiol­o­gist and tak­ing a though piece as a guide for de­cid­ing to iso­late your­self or to break quaran­tine is pretty stupid. If this en­courages any ac­tion other than “Hmh, in­ter­est­ing, let me try to find some data and look at it”, then I will claim right now said ac­tion is mis­guided.

Main references

The main refer­ences used for get­ting the num­bers and fac­tual in­for­ma­tion claimed in this ar­ti­cle are (in or­der of im­por­tance):

Gen­eral in­for­ma­tion about covid-19 (the dis­ease), cov-sars-2 (the virus) and the mea­sures pro­posed against the pan­demic:

https://​​id­pjour­nal.biomed­cen­tral.com/​​ar­ti­cles/​​10.1186/​​s40249-020-00646-x

https://​​www.el­se­vier.com/​​con­nect/​​coro­n­avirus-in­for­ma­tion-center

https://​​pe­ter­at­ti­amd.com/​​pe­ter­hotez2/​​

https://​​pe­ter­at­ti­amd.com/​​pe­ter­hotez/​​

The study on which I’m bas­ing my claims for age and co­mor­bidi­ties be­ing the main driver of death and se­vere symp­toms:

https://​​www.cebm.net/​​global-covid-19-case-fatal­ity-rates/​​

In­for­ma­tion about the COVID-19 test:

https://​​www.na­ture.com/​​ar­ti­cles/​​d41587-020-00010-2

In­for­ma­tion about DXY & DXY dataset:

https://​​docs.google.com/​​spread­sheets/​​d/​​1jS24DjSPVWa4iuxuD4OAXrE3QeI8c9BC1hSlqr-NMiU/​​edit#gid=1187587451

http://​​www.dxy.cn/​​

De­mo­graphic data used to ex­tract the origi­nal num­bers:

https://​​www.statista.com/​​statis­tics/​​270000/​​age-dis­tri­bu­tion-in-the-united-states/​​

https://​​www.statista.com/​​statis­tics/​​251524/​​pop­u­la­tion-dis­tri­bu­tion-by-age-group-in-china/​​

No comments.