Two weeks ago Ben Pace and I asked for practical advice on what to do about coronavirus, with the requirement that advice be justified in some way- ideally full blown models informed by empirical data, but at a minimum, an explanation of why it was helpful. The resulting thread had wonderful advice but makes for, uh, inefficient reading, at best. This post is an attempt to take the best of the Justified Practical Advice Thread and present it in a clear, easy-to-digest format.
It’s important to note that neither the original post nor this one are attempting to give comprehensive advice. If you’re looking for that, I suggest checking out the Guide/FAQs/Intros tab on the LessWrong Coronavirus Link Database. This is for advice that fell through the cracks.
Though I will briefly mention the standard advice. From what I can tell, the most important (and luckily widely publicized advice) is the following:
Wash your hands a lot, and wash them to a hospital standard (spend 20 seconds and follow these instructions).
Socially isolate yourself, and don’t be around other people unless really necessary. Avoid groups of people, especially large gatherings. Really don’t go to bars, clubs, restaurants or other places that lots of people pass through (like public transportation).
If you are sick, use a mask. They are probably also effective at preventing the disease when you are not sick, but most of the western world currently has massively limited supply, so leave the masks to health workers and sick people.
The rest of this post will summarise the interesting ideas that went beyond these basic recommendations. But if you haven’t done (or at least seriously considered doing) all of the above, I would recommend you prioritise the things listed above.
Many of the below recommendations involve buying things. When possible I’ve included a link to a particular amazon page. This is not a strong recommendation for that particular product: it’s an attempt to lower activation energy for acting on a recommendation. If you have reason to believe a particular model is better than what I linked to, please comment and I’ll update if appropriate. Speaking of which, if you think of anything else I missed or got wrong, please comment with that too.
That said, here is the top advice from the Coronavirus Justified Practical Advice thread.
Cover your high-touch surfaces with copper tape.
Treat newly delivered packages as contagious for 48 hours.
This was the clear winner of JPA Olympics– it’s cheap, effective, and very few people had heard of it before. In fact we were hoping for a highly polished, intensely researched post on just this; unfortunately no one has the time right now.
In its place, please enjoy these links:
This pre-print showed that this coronavirus in particular had a half life of 2.4-5.11 hours on copper, in contrast to 10.5-16.1 on steel or 13-19.2 on plastic
This review showed H1N1 decreased by 4 logs (a factor of 10^4) in 6 hours;
This study showed vaccinia and monkeypox viruses reduced by 6 logs (a factor of 10^6) in 3 minutes
This study showed murine norovirus was destroyed in 30 minutes, though it doesn’t work very well at 4C;
This review says that copper oxide filters neutralize all of “bacteriophages [58-62], Infectious Bronchitis Virus [63], Poliovirus [61,64], Junin Virus [59], Herpes Simplex Virus [58,59], Human Immunodeficiency Virus Type 1 (HIV-1) [11,65-67], West Nile Virus [11], Coxsackie Virus Types B2 & B4, Echovirus 4 and Simian Rotavirus SA11 [68]. More recently, the inactivation of Influenza A [55,65], Rhinovirus 2, Yellow Fever, Measles, Respiratory Syncytial Virus, Parainfluenza 3, Punta Toro, Pichinde, Adenovirus Type 1, Cytomegalovirus, and Vaccinia [65]”.
Benefits
Reduces surface-to-hand-to-face transmission
Costs
$11, half an hour to apply tape (this may be a recurring cost, but we don’t know at what interval)
Risk of cuts while applying tape (this happened to me once)
Risk of cuts from applied tape (0 for me, but other people have reported lots). This is not just a convenience issue- small cuts are breaks in your defenses that microbes can enter.
Why you might not do this
You are worried about the risks of exposure to copper
You might be are allergic
Some people have reported hand irritation and discoloration when they put it on their phones and laptops
I poked at this and my takeaway was you’d need to be hooked up to skin infusion IV to get as much copper as you consume via food. Additionally copper IUDs are a common form of birth control, so it can’t be *that* dangerous.
You think surface-to-hand-to-face transmission is not significant.
You don’t think the results of tests on other viruses apply to coronavirus
The 2’’ is from personal experiences: I got a variety of sizes up to 1’’, and even that was often inconveniently small. I basically never wanted smaller except to patch up gaps, and they’re not so much better than just tearing a small piece off of 2’’ for that.
Treat newly delivered packages and mail as contagious for 48 hours (Elizabeth)
At the time I posted this, we had very little information on the risks of contamination via packages. I based my recommendation on studies showing other coronaviruses survived for a very long time on other surfaces. Since then someone has released a pre-print on this specific coronavirus on cardboard in particular, which found a half-life of ~2-5 hours. What that means your safety depends on exactly how much your mailman sneezes on it and what the infectious threshold is, which we don’t actually know. However at the concentrations that paper was using, concentration dipped below detection by 48 hours.
Here are the specific recommendations for handling potentially infected packages:
Do not touch your face between touching a package and washing your hands or taking off gloves.
Spray deliveries with disinfectants.
Note that disinfectants are typically less effective on porous material like cardboard. I know of no hard numbers on this.
Leave packages exposed to the sun for 1-3 days (this can probably be shortened given the newest information) or a UV lamp for shorter
Just don’t touch them for several days.
Benefits
Reduces package-to-hand-to-face transmission
Costs
Adds time and annoyance to handling packages
UV requires a sterilizer
Leaving outside may risk theft
Letting packages sit takes time and space.
Disinfectants are pennies/spray.
Why you might not do this
You think surface-to-hand-to-face transmission isn’t important.
You’re extremely confident your delivery person is not contagious.
You think there is little enough initial contamination pre-delivery that it will have degraded by the time the package reaches you.
Vitamin D deficiency has a surprisingly strong link with respiratory infection.
Study says 4x rate of respiratory infection in the very deficient, but doesn’t see an obvious effect in the partially deficient, so slightly weird statistics.
COVID-like illnesses frequently cause loss of electrolytes through sweating and diarrhea, which are not replenished from food due to appetite suppression. Insufficient electrolytes can lead to digestive and neurological problems.
The exact form of this is up for debate: some people feel that salt, maybe with sugar, honey, or molasses, is sufficient. Others thought that the magnesium and potassium from a dedicated electrolyte powder/concentrate/drink, or the ease of consumption while sick, was beneficial. Once you’ve gone that road, there are lots of electrolyte supplements with additional vitamins and minerals you might also be missing, leading to something of a paradox of choice for me.
No one disagreed that some form of electrolyte is good to drink to once sick, just what exact form that should take. Choosing any one of these is more important than which one you choose.
Forms you might take, with approximate cost:
A homemade version of WHO’s Oral Rehydration Salts ($5 in cheapest form).
Coronavirus moves from unpleasant to dangerous when your blood oxygen saturation level is too low. A pulse oximeter can tell you when that is happening, providing a clearer line for when to seek medical attention, which is especially important as the risks to contact with the medical system rise. However many people felt that you should ignore a low pulse ox if you felt fine or a high one if you felt short of breath, so it didn’t add anything to the process, an argument I find pretty compelling.
Benefits
Certainty in when to seek medical attention, especially if you can’t trust your internal sense of being out of breath (due to e.g. panic attacks).
Costs
$30
Why you might not do this
You trust your internal sense of being out of breath or not.
You expect medical care to be unavailable or net-negative no matter what.
Thank you to Raemon and Ben Pace for comments on this document.
Thank you very much to Julia Wise, Finan Adamson, Connor Flexman and Connor Flexman for this justified practical advice.
Coronavirus Justified Practical Advice Summary
Two weeks ago Ben Pace and I asked for practical advice on what to do about coronavirus, with the requirement that advice be justified in some way- ideally full blown models informed by empirical data, but at a minimum, an explanation of why it was helpful. The resulting thread had wonderful advice but makes for, uh, inefficient reading, at best. This post is an attempt to take the best of the Justified Practical Advice Thread and present it in a clear, easy-to-digest format.
It’s important to note that neither the original post nor this one are attempting to give comprehensive advice. If you’re looking for that, I suggest checking out the Guide/FAQs/Intros tab on the LessWrong Coronavirus Link Database. This is for advice that fell through the cracks.
Though I will briefly mention the standard advice. From what I can tell, the most important (and luckily widely publicized advice) is the following:
Wash your hands a lot, and wash them to a hospital standard (spend 20 seconds and follow these instructions).
Socially isolate yourself, and don’t be around other people unless really necessary. Avoid groups of people, especially large gatherings. Really don’t go to bars, clubs, restaurants or other places that lots of people pass through (like public transportation).
If you are sick, use a mask. They are probably also effective at preventing the disease when you are not sick, but most of the western world currently has massively limited supply, so leave the masks to health workers and sick people.
The rest of this post will summarise the interesting ideas that went beyond these basic recommendations. But if you haven’t done (or at least seriously considered doing) all of the above, I would recommend you prioritise the things listed above.
Many of the below recommendations involve buying things. When possible I’ve included a link to a particular amazon page. This is not a strong recommendation for that particular product: it’s an attempt to lower activation energy for acting on a recommendation. If you have reason to believe a particular model is better than what I linked to, please comment and I’ll update if appropriate. Speaking of which, if you think of anything else I missed or got wrong, please comment with that too.
That said, here is the top advice from the Coronavirus Justified Practical Advice thread.
Cover your high-touch surfaces with copper tape.
Treat newly delivered packages as contagious for 48 hours.
Take vitamin D supplements daily.
Buy electrolytes to drink if you get ill.
Maybe: Buy a pulse oximeter.
Practical Advice
Get 2-inch copper tape and cover high-touch surfaces with it, especially shared high-touch surfaces (Connor Flexman)
This was the clear winner of JPA Olympics– it’s cheap, effective, and very few people had heard of it before. In fact we were hoping for a highly polished, intensely researched post on just this; unfortunately no one has the time right now.
In its place, please enjoy these links:
This pre-print showed that this coronavirus in particular had a half life of 2.4-5.11 hours on copper, in contrast to 10.5-16.1 on steel or 13-19.2 on plastic
This review showed H1N1 decreased by 4 logs (a factor of 10^4) in 6 hours;
This study showed vaccinia and monkeypox viruses reduced by 6 logs (a factor of 10^6) in 3 minutes
This study showed murine norovirus was destroyed in 30 minutes, though it doesn’t work very well at 4C;
This review says that copper oxide filters neutralize all of “bacteriophages [58-62], Infectious Bronchitis Virus [63], Poliovirus [61,64], Junin Virus [59], Herpes Simplex Virus [58,59], Human Immunodeficiency Virus Type 1 (HIV-1) [11,65-67], West Nile Virus [11], Coxsackie Virus Types B2 & B4, Echovirus 4 and Simian Rotavirus SA11 [68]. More recently, the inactivation of Influenza A [55,65], Rhinovirus 2, Yellow Fever, Measles, Respiratory Syncytial Virus, Parainfluenza 3, Punta Toro, Pichinde, Adenovirus Type 1, Cytomegalovirus, and Vaccinia [65]”.
Benefits
Reduces surface-to-hand-to-face transmission
Costs
$11, half an hour to apply tape (this may be a recurring cost, but we don’t know at what interval)
Risk of cuts while applying tape (this happened to me once)
Risk of cuts from applied tape (0 for me, but other people have reported lots). This is not just a convenience issue- small cuts are breaks in your defenses that microbes can enter.
Why you might not do this
You are worried about the risks of exposure to copper
You might be are allergic
Some people have reported hand irritation and discoloration when they put it on their phones and laptops
I poked at this and my takeaway was you’d need to be hooked up to skin infusion IV to get as much copper as you consume via food. Additionally copper IUDs are a common form of birth control, so it can’t be *that* dangerous.
You think surface-to-hand-to-face transmission is not significant.
You don’t think the results of tests on other viruses apply to coronavirus
The 2’’ is from personal experiences: I got a variety of sizes up to 1’’, and even that was often inconveniently small. I basically never wanted smaller except to patch up gaps, and they’re not so much better than just tearing a small piece off of 2’’ for that.
Treat newly delivered packages and mail as contagious for 48 hours (Elizabeth)
At the time I posted this, we had very little information on the risks of contamination via packages. I based my recommendation on studies showing other coronaviruses survived for a very long time on other surfaces. Since then someone has released a pre-print on this specific coronavirus on cardboard in particular, which found a half-life of ~2-5 hours. What that means your safety depends on exactly how much your mailman sneezes on it and what the infectious threshold is, which we don’t actually know. However at the concentrations that paper was using, concentration dipped below detection by 48 hours.
Here are the specific recommendations for handling potentially infected packages:
Do not touch your face between touching a package and washing your hands or taking off gloves.
Spray deliveries with disinfectants.
Note that disinfectants are typically less effective on porous material like cardboard. I know of no hard numbers on this.
Leave packages exposed to the sun for 1-3 days (this can probably be shortened given the newest information) or a UV lamp for shorter
Just don’t touch them for several days.
Benefits
Reduces package-to-hand-to-face transmission
Costs
Adds time and annoyance to handling packages
UV requires a sterilizer
Leaving outside may risk theft
Letting packages sit takes time and space.
Disinfectants are pennies/spray.
Why you might not do this
You think surface-to-hand-to-face transmission isn’t important.
You’re extremely confident your delivery person is not contagious.
You think there is little enough initial contamination pre-delivery that it will have degraded by the time the package reaches you.
Take vitamin D supplements (Connor Flexman)
Vitamin D deficiency has a surprisingly strong link with respiratory infection.
Study says 4x rate of respiratory infection in the very deficient, but doesn’t see an obvious effect in the partially deficient, so slightly weird statistics.
Study says very large effects in children
WHO says vitamin D supplementation may reduce respiratory infections in children
Study says no effect from supplementing after already sick, so get on this before infection
While this hasn’t been verified for COVID-19 in particular, it does seem plausible that it applies, and that a supplement can treat it.
Benefits:
Reduces respiratory infections, via unclear mechanism.
Costs:
2.5 cents/day using the brand I recommended (recommended to me by a doctor but not otherwise verified), cheaper probably available.
Why you might not do this:
You’ve tested your vitamin D recently and levels were normal or high (as a fat soluble vitamin, it is possible to overdose on)
You don’t think the results from general respiratory infections transfer to COVID-19 in particular.
You think you get enough vitamin D from the sun, and will continue to do so during periods of isolation.
You don’t believe supplements can convey health benefits.
Have electrolytes on hand and use once ill (tragedyofthecomments)
COVID-like illnesses frequently cause loss of electrolytes through sweating and diarrhea, which are not replenished from food due to appetite suppression. Insufficient electrolytes can lead to digestive and neurological problems.
The exact form of this is up for debate: some people feel that salt, maybe with sugar, honey, or molasses, is sufficient. Others thought that the magnesium and potassium from a dedicated electrolyte powder/concentrate/drink, or the ease of consumption while sick, was beneficial. Once you’ve gone that road, there are lots of electrolyte supplements with additional vitamins and minerals you might also be missing, leading to something of a paradox of choice for me.
No one disagreed that some form of electrolyte is good to drink to once sick, just what exact form that should take. Choosing any one of these is more important than which one you choose.
Forms you might take, with approximate cost:
A homemade version of WHO’s Oral Rehydration Salts ($5 in cheapest form).
That, replacing sugar with molasses or raw honey
Powder ($30)
Tablets ($25)
Electrolyte drops ($20) (my choice, for esoteric reasons)
Benefits
Prevent electrolyte insufficiency, which can lead to neurological and digestive issues
Why you might not do this
Personal experience hating electrolytes
Inability to distinguish when you’ve had “enough” (since too much is also dangerous)
Maybe: Buy a pulse oximeter (juliawise)
Coronavirus moves from unpleasant to dangerous when your blood oxygen saturation level is too low. A pulse oximeter can tell you when that is happening, providing a clearer line for when to seek medical attention, which is especially important as the risks to contact with the medical system rise. However many people felt that you should ignore a low pulse ox if you felt fine or a high one if you felt short of breath, so it didn’t add anything to the process, an argument I find pretty compelling.
Benefits
Certainty in when to seek medical attention, especially if you can’t trust your internal sense of being out of breath (due to e.g. panic attacks).
Costs
$30
Why you might not do this
You trust your internal sense of being out of breath or not.
You expect medical care to be unavailable or net-negative no matter what.
Thank you to Raemon and Ben Pace for comments on this document.
Thank you very much to Julia Wise, Finan Adamson, Connor Flexman and Connor Flexman for this justified practical advice.