COVID-19: List of ideas to reduce the direct harm from the virus, with an emphasis on unusual ideas

This is a col­lec­tion of ideas on what in­ter­ven­tions could be in­ves­ti­gated, funded, or im­ple­mented to po­ten­tially re­duce the harms from COVID-19 by in­ter­ven­ing di­rectly on the causal chain lead­ing to in­fec­tion and death. In par­tic­u­lar, it has a lot of “un­usual ideas” that haven’t got­ten much at­ten­tion yet.

Caveats: Most of the “un­usual ideas” (see later sec­tion) were gen­er­ated or vet­ted only quickly. Be­fore im­ple­men­ta­tion, they will definitely need fur­ther in­ves­ti­ga­tion, and some are only meant to serve as in­spira­tion for bet­ter in­ter­ven­tions. But ca­pac­ity to im­ple­ment them if they do work can gen­er­ally be built in par­allel with in­ves­ti­gat­ing whether they work. Fur­ther, it’s also pos­si­ble that some things here could have con­se­quences worse than the dis­ease (stuff like surveillance states, long term harm from bad vac­ci­na­tions, or dual use dan­ger­ous biotech­nol­ogy). There­fore, the sug­ges­tion here is to pour a lot of re­sources into in­ves­ti­gat­ing those par­tic­u­larly risky op­tions, rather than to im­me­di­ately and unilat­er­ally im­ple­ment them.

As the sur­round­ing con­text to keep in mind, we want to:

  • Stop this pan­demic in its tracks

  • Min­i­mize lives lost, eco­nomic dis­rup­tion, psy­cholog­i­cal harm, and poli­ti­cal harm done di­rectly by the dis­ease or our re­sponses to it

  • Per­haps, if lucky, find a way to make hu­man­ity bet­ter off afterwards

For some re­lated re­search top­ics check out Coron­avirus Re­search Ideas for EAs.

A gen­eral list of things for in­terfer­ing with the causal chain of in­fec­tion and death

In this sec­tion, I’ll list cat­e­gories of in­ter­ven­tions to re­duce the harm from COVID-19. Th­ese should ideally be in­ves­ti­gated and im­ple­mented in par­allel at speed for some­thing as dis­rup­tive as this virus. Some of these are be­ing im­ple­mented by vary­ing de­grees in na­tions fight­ing the pan­demic.

  • Large scale pre­ven­tion of ex­po­sure: So­cial dis­tanc­ing and man­ag­ing pack­ages and surfaces

  • Small scale of pre­ven­tion of ex­po­sure: Phys­i­cally block­ing the virus be­fore it gets to the body

    • masks

    • sanitation

    • etc.

  • Chem­i­cal in­ter­ven­tions once it has en­tered the body: Vac­cines, antivirals

  • Phys­iolog­i­cal sup­port once it has en­tered the body: Stuff like ven­tila­tor pro­duc­tion and sup­ple­men­tal oxygen

  • Meta: In­ves­ti­gate the al­ter­na­tive causal path­ways to death and hy­pothe­ses about them; an­swer­ing ques­tions like: how much dam­age and death is due to lung dam­age, cy­tokine storms, dam­age to the ner­vous sys­tem, or in­terfer­ing with oxy­gen trans­porta­tion in the blood?

  • Mis­cel­la­neous “un­usual ideas” (see later sec­tion)

Policy for ac­cel­er­at­ing vac­cine de­vel­op­ment (and similar for an­tivirals and pos­si­bly some other po­ten­tial solu­tions)

  • Many can­di­dates can be tested in par­allel for speed

  • To speed up the test­ing of a spe­cific vac­cine we can skip di­rectly to later stage tri­als—ex­pos­ing hu­mans to the virus

  • We can ac­cel­er­ate test­ing for long-term nega­tive effects by hav­ing a large and di­verse set of vol­un­teers take it at once—effects that are rare or slow to show them­selves will be de­tected much faster this way

  • Slow and rare effects can plau­si­bly also be de­tected more eas­ily by se­lect­ing peo­ple based on weak­ness to the plau­si­ble rare or slow effect, or al­ter­na­tively ex­pos­ing peo­ple to con­di­tions that would cause a plau­si­ble rare effect or slow effect to show itself

  • In par­allel with test­ing, pro­duc­tion ca­pa­bil­ity and per­haps even stocks of plau­si­ble vac­cine can­di­dates can be built up and even dis­tributed so they are ready to use on a mo­ment’s no­tice if the vac­cine works. (Bill Gates is now do­ing this.)

  • See also the this EA fo­rum post

Note: Some of these ideas will face eth­i­cal ob­jec­tions or may be hard poli­ti­cally to im­ple­ment.

Unusual ideas

Min­i­miz­ing en­vi­ron­men­tal exposure

  • Cop­per­ing sur­faces to de­crease the virus lin­ger­ing time (this effect may not last much be­yond when the cop­per sur­face ox­i­dizes)

  • Leav­ing soap resi­dues on sur­faces to de­crease virus lin­ger­ing time (plau­si­ble be­cause covid-19 is sur­rounded by a lipid mem­brane, but this may not work at nor­mal tem­per­a­ture and hu­midity)

  • There may be other al­ter­na­tive bet­ter an­tiviral coat­ings that can be coated on surfaces

  • Hav­ing UVc lights of the ideal wave­length in peo­ple dense areas

    • Pos­si­bly hav­ing them on when peo­ple are not around

    • Maybe you could even build some­thing to shine in the lungs, un­likely though

  • As­sum­ing COVID-19 ex­hibits sea­son­al­ity: chang­ing the tem­per­a­ture, hu­midity, or wind pat­terns in­side peo­ple-dense ar­eas to in­duce ar­tifi­cial sum­mer and “out­side”

So­cial dis­tanc­ing mea­sures and management

  • Ap­ply­ing even more so­phis­ti­cated strate­gies for con­tact trac­ing and immunity

    • Re­move the small-world net­work hubs or per­haps just wait for them to be­come im­mune (or re­place them with with im­mune folks) to re­duce the viral re­pro­duc­tion number

    • Highly pri­vacy pre­serv­ing con­tact trac­ing: https://​​ and https://​​

    • Take into ac­count var­i­ance of R0 when do­ing con­tact trac­ing and lockdowns

    • Do pooled and ran­dom sam­ple test­ing to more effi­ciently use limited test­ing abil­ity and search for com­mu­nity spread

    • As an idea for in­spira­tion (not fea­si­ble as is), change the net­work topol­ogy of interactions

    • Another idea for in­spira­tion (not fea­si­ble as is), move the re­cov­ered peo­ple around, in­te­grate them into the so­cial web, and have them act as fire­breaks. You take re­cov­ered or oth­er­wise im­mune peo­ple and move them to places where there is an out­break in or­der to drop R0 be­low 1

  • Ap­ply a more syn­chronous form of lock­down. In the­ory if ev­ery­one was in lock­down for a month at the same time we’d be able to iden­tify es­sen­tially ev­ery­one who has it and have the mild cases non con­ta­gious by the end of that time pe­riod (we’d still have to be care­ful with con­tam­i­nated sur­faces (es­pe­cially frozen or re­friger­ated items)). Some­thing like this could very well be cheaper and eas­ier than an ex­tended par­tial lock­down.

  • In or­der to spread out the timing of when peo­ple are sick, a stag­gered form of quaran­tine where you ex­pose some peo­ple ear­lier than oth­ers could be im­ple­mented. This op­tion is not a good op­tion and more of a last re­sort to spread out when peo­ple are at the hos­pi­tals and give us more time to come up with some­thing bet­ter. It is bet­ter than just let­ting the epi­demic spread freely though. Per­haps im­ple­ment some­thing like this con­trol­led infection

  • Ap­ply com­pli­cated forms of the “dance”: things like vary­ing the de­gree and in­ten­sity of trac­ing and lock­down over time and lo­ca­tions dy­nam­i­cally in re­sponse to con­ta­gion (per­haps us­ing some­thing like a PID con­trol­ler). For some con­crete­ness check out:

  • Ubiquitous health mon­i­tor­ing (tem­per­a­ture sen­sors, cough sound de­tec­tors, …) - this can done us­ing ei­ther at­tached or far away sensors

  • Figure out how to make the symp­toms more ap­par­ent to peo­ple (by ed­u­ca­tion, cus­tomized bio­track­ing, some­thing that am­plifies symp­toms (stuff like the op­po­site of a cough sup­pres­sant or fever re­ducer?), …)

De­crease the ini­tial viral load

Chem­i­cal means of coun­ter­ing covid 19 infections

  • Put the en­tire pop­u­la­tion on weak im­mune boost­ers (stuff like vi­tam­ins D and C, se­le­nium, …)

  • Put the en­tire pop­u­la­tion on stronger forms of im­mune sys­tem boost­ers (I think there are some im­mune sys­tem boost­ers in cer­tain vac­cines). Alter­na­tively maybe just hand them out to be used at the first sign of illness

  • Prime the im­mune sys­tem be­fore or dur­ing the ear­lier stages of the disease

  • Put the en­tire pop­u­la­tion on other im­mune boost­ing strate­gies: in­ter­mit­tent fast­ing, cold show­ers, saunas, good sleep, get­ting to the ideal weight for res­pi­ra­tory dis­ease, oxy­tocin/​so­cial sup­port/​hugs, placebo pills, …

  • Avoid im­mune sys­tem sup­pres­sants like steroidal chem­i­cals, NSAIDs and Tylenol, and symp­tom sup­pres­sants (symp­tom sup­pres­sant use may also in­crease con­ta­gion by en­abling peo­ple to go out with­out visi­ble signs of ill­ness (to them­selves or oth­ers) and se­lect for more dan­ger­ous ver­sions of the virus)

  • Lots of spec­u­la­tive chem­i­cals like quercetin, rose­mary, …

  • Trans­fer blood plasma/​an­ti­bod­ies from the re­cov­ered to the ill (some peo­ple are im­ple­ment­ing this) - pos­si­bly boost this by hook­ing up cir­cu­la­tory sys­tems or boost­ing the re­cov­ered an­ti­body pro­duc­tion (per­haps by trans­fer­ring some more viral load later)

  • Trans­fer blood from the young to the old—prob­a­bly need im­mune sup­pres­sors for this be­cause you’d need to trans­fer im­mune cells and in any event this shouldn’t work be­cause you very likely need to gen­er­ate new im­mune cells to han­dle the disease

  • Disperse zinc lozenges to the pop­u­la­tion to use as soon as they even spec­u­late they are com­ing down with some­thing in or­der to kill the virus in their throat and per­haps up­per res­pi­ra­tory sys­tem to de­crease or per­haps elimi­nate their viral load. High per­centage ethanol va­por can also pro­duce this effect.

  • Per­haps the high tem­per­a­ture and hu­midity of saunas can kill the virus in one’s up­per res­pi­ra­tory sys­tem and per­haps help clear the lower lungs

  • Disperse a coat of an­tiviral and non-cy­to­toxic oil, do­cosanol, deep into the lungs—Danielle Fong had this idea but she wor­ries about it gum­ming up the lungs and it hasn’t gone any­where yet

  • See if chem­i­cals use­ful for deal­ing with high al­ti­tude or boost­ing oxy­gen lev­els help

  • Aer­o­bic ex­er­cise to boost lung ca­pac­ity and car­dio­vas­cu­lar fit­ness be­fore ex­po­sure (but not af­ter)

  • De­crease ex­po­sure to air pol­lu­tion and smoke

  • Look into what ath­letes use when dop­ing to boost red blood cell count and hemoglobin levels

  • Use chem­i­cals that de­crease cy­tokine storms (like pos­si­bly vi­tamin C, mela­tonin, tocil­izumab, and drugs that help with au­toim­mune di­s­or­ders)

  • Maybe some of the chem­i­cals used in pre­vent­ing au­tophagy, stroke dam­age, etc, can be used to pre­vent dam­age in this case

  • Maybe boost an­tiox­i­dants to in­hibit dam­age to the lungs and in­hibit cy­tokine storms

  • Maybe in­ject lots of ACE2 re­cep­tors into the body and lungs to di­vert the virus away from its target

  • Cas­tra­tion imi­ta­tion chem­i­cals, to re­gen­er­ate the thy­roid and pos­si­bly re­turn it to a state similar to that of younger peo­ple where the death rate is lower. I re­call read­ing a dis­cus­sion on this some­where but can­not find it again.

  • Have peo­ple take es­tro­gen or phy­toe­stro­gens, as­sum­ing that is rele­vant for the differ­ence be­tween the male and fe­male fatal­ity rates: Estro­gen re­cep­tor im­pairs in­ter­leukin-6 ex­pres­sion by pre­vent­ing pro­tein bind­ing on the NF-kap­paB site.

Non-chem­i­cal ways of coun­ter­ing covid 19 infections

  • Im­ple­ment so­phis­ti­cated forms of liquid breathing

  • Maybe there is a way to ul­tra­son­i­cally shake the lungs to clear them, similar to what is used for kid­ney stones

  • Maybe there is truth to the pos­si­bil­ity that some forms of near in­frared can boost heal­ing. Maybe bathe the pa­tient or their lungs (some­how) in it

  • See if in­creas­ing par­tial air pres­sure can help for blood oxygenation

  • Have peo­ple im­ple­ment prone po­si­tions in the ear­lier stages of the dis­ease or for self care if hos­pi­tals are overwhelmed


I again wish to em­pha­size that the ideas col­lected here range from strongly im­plied by sci­ence to be helpful (such as cop­per sur­faces) to the very spec­u­la­tive (such as an an­tiviral “oil” for the lungs), and that some have the po­ten­tial to cause more harm than the dis­ease it­self. Many of these ideas would need to be in­ves­ti­gated fur­ther be­fore im­ple­men­ta­tion, though ca­pac­ity to im­ple­ment could be built up at the same time. And hope­fully some of the “un­usual ideas” can in­spire more such ideas from oth­ers. In gen­eral, peo­ple need to move fast to com­bat COVID-19, but also need to be thought­ful, and to avoid in­for­ma­tion haz­ards and highly risky ac­tions.

What do you think of these ideas; how prob­a­ble, costly, and effec­tive would they be?

Do you have ad­di­tional ideas you think would be good to in­ves­ti­gate and im­ple­ment?

I hope that promis­ing ideas from here will in­spire ad­di­tional progress and will be fur­ther in­ves­ti­gated, for­warded to rele­vant par­ties, and acted upon if suffi­ciently vet­ted and de­vel­oped.

Thanks to David Kristoffers­son, Michael Aird, and Eliz­a­beth Van Nos­trand for edit­ing help, to Alexey Turchin for shar­ing some ideas, and to the many peo­ple who have been com­ing up with great ideas for fight­ing the pan­demic, many of which I’ve in­cluded here or have been in­spired by.