Are there good EA projects for helping with COVID-19?

post by Linch · 2020-03-03T23:55:59.259Z · score: 31 (17 votes) · EA · GW · 10 comments

This is a question post.

Contents

  Answers
    12 bwr
    10 Davidmanheim
    9 Nathan Young
    7 jaythibs
    5 ZachWeems
    1 Agrippa
    1 Peterslattery
    1 BenHoffman
None
10 comments

I think I (and some other EAs) are in a possibly rare category of being young, healthy, and having unusually high financial, etc stability so we can take time off our work/studies and other projects to fully focus on this.

In addition, some of us may have especially relevant skills. (For example, many of us have extensive amateur or professional experience on analyzing tail risks).

While I agree that there should be a strong prior against EAs working on disaster relief, I think this is plausibly different from other disaster relief situations because it's likely to be unusually bad and humanity is unusually ill-prepared relative to previous disasters.

I'm tentatively more optimistic about new projects than about donating money or joining existing orgs, since this seems like an important space to move fast on.

Answers

answer by bwr · 2020-03-04T08:19:22.133Z · score: 12 (6 votes) · EA(p) · GW(p)

I've seen concern that hospitals will run out of ventilators. Potential intervention: design a cheap machine to pump bag valve masks (which are ubiquitous and apparently do much of the same job as a ventilator, but currently require a human operator). I'd guess you could build something to perform this job for <$50; possibly very quickly if you had a team of competent engineers.

I don't know how you'd get them distributed though, and I'm skeptical that the FDA would make it easy to sell them to US hospitals. I'm interested in anyone with experience in the medical device space, or experience in the constraints on what devices hospitals are allowed to use, weighing in on that question.

comment by HaukeHillebrandt · 2020-03-04T08:47:49.249Z · score: 10 (4 votes) · EA(p) · GW(p)

Found this paper: "Optimizing respiratory management in resource-limited settings"
"Mechanical ventilation is an expensive intervention associated with considerable mortality and a high rate of iatrogenic complications in many LMICs. Recent case series report crude mortality rates for ventilated patients of between 36 and 72%. Measures to avert the need for invasive mechanical ventilation in LMICs are showing promise: bubble continuous positive airway pressure has been demonstrated to decrease mortality in children with acute respiratory failure and trials suggest that noninvasive ventilation can be conducted safely in settings where resources are low." ... "One of the most significant developments in acute care research in LMICs in recent years has been the publication of three trials demonstrating that continuous positive airway pressure (CPAP) can reduce mortality in children under 5 years of age, compared with oxygen delivered via standard low-flow nasal cannula [35▪,36,37▪]. CPAP can also decrease the need for invasive mechanical ventilation [38▪▪]. There are three main ways to generate CPAP: first, by using a pressure driver or a ventilator; second, using high flow nasal-cannula oxygen therapy (HFNC); or third, by submerging the expiratory limb of a breathing circuit in water to create so-called bubble CPAP. Traditionally bubble CPAP circuits also contain a driver, although some newer iterations only use the oxygen/air flow from an oxygen concentrator to generate CPAP [39].

All three trials used bubble CPAP as the intervention and together showed a risk ratio of survival of 0.58 [95% confidence interval (CI) 0.41–0.82] [38▪▪]. One study had an additional intervention arm using HFNC, but no conclusions were drawn regarding its efficacy as the study was terminated early due to increased mortality in the control group.

Nasal cannulae, used as the patient interface in all three trials, are an attractive option for understaffed environments because they generally require lower levels of nursing supervision to use safely [39]. The basic circuits and simplified care protocols meant that the equipment required few adjustments, especially when compared with invasive mechanical ventilation.

There are elements of each of these studies that epitomize context-appropriate innovation and research. The bubble CPAP circuit deployed in the Bangladesh study was fashioned out of readily available, cheap equipment (standard nasal cannula, a shampoo bottle and intravenous fluid tubing) so the cost of the circuit was approximately $3 per patient [35▪]. They used an oxygen concentrator and no driver in the circuit with additional cost savings." https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6319564/

here’s a build diagram for a bubble cpap

https://www.edjones.org/articles/bubble-cpap-in-resource-poor-settings/?fbclid=IwAR05oxQ2tPg1LDD6o73cBWOGKukaYBq8APcFpNmB1y900nPovTwV0yFBWBQ

are studies indicating it’s helpful for adults (despite its primary use in infants)

https://intjem.biomedcentral.com/articles/10.1186/s12245-019-0224-0

comment by bwr · 2020-03-04T09:31:30.267Z · score: 3 (2 votes) · EA(p) · GW(p)

This is super interesting - Some of the most interesting-sounding links seem broken, though [edit: fixed]

comment by tessa · 2020-03-04T10:02:39.892Z · score: 3 (3 votes) · EA(p) · GW(p)

A "portable, easy-to-use ventilator" was highlighted in the Johns Hopkins Center for Health Security report on Technologies to Address Global Catastrophic Biological Risks (press release, full PDF). Their write-up of this technology is on page 61-63 of the report.

One of the sources they link describes the OneBreath ventilator. Might be a good place to start looking!

answer by Davidmanheim · 2020-03-04T17:09:11.095Z · score: 10 (6 votes) · EA(p) · GW(p)

Based on a paper that was just accepted for publication, I outlined 4 areas where I think there is still critical work that can be done to prepare for wider-scale disruption, should it occur [EA · GW]. They are:

1) Enable people to stay isolated effectively.
2) Triage and manage medical care remotely.
3) Manage critical services through disruptions.
4) Ensure transport systems remain functional.

For details about what each means, see the linked post.

answer by Nathan Young · 2020-03-04T13:10:52.615Z · score: 9 (6 votes) · EA(p) · GW(p)

I have been leading on the Coronavirus Tech Handbook.

It's a collection of tools, websites and data relating to Coronavirus. If you see useful website, tools or datasets, please add them.

https://coronavirustechhandbook.com/

I imagine there are tech interventions people could make which would save lives, so connecting the community seems effective given how cheap it is.

comment by Nathan Young (nathan) · 2020-03-04T13:31:11.599Z · score: 1 (1 votes) · EA(p) · GW(p)

Retweets of this appreciated https://twitter.com/NathanPMYoung/status/1234908740803010562

answer by jaythibs · 2020-03-13T03:33:33.251Z · score: 7 (2 votes) · EA(p) · GW(p)

Reposting what I wrote on Facebook:

Young low-risk adults doing groceries and other errands for high-risk old adults.

I wonder if it is effective and if there is a way to scale it? I talked to one EA from Italy and they said a student union is also doing this there. I am looking into how to accomplish this in Canada.

We could potentially fund an app or something so that anyone who wants to volunteer can quickly take part and accept a request.

The request could be taken via telephone for example and then placed in the app.

Or we create a simple process without any apps. Google sheet?

Dealing with superspreaders: it’s crucial to give guidelines and make sure young people are much less likely to catch the virus than old people. I think this is doable.

answer by ZachWeems · 2020-03-04T07:05:42.973Z · score: 5 (4 votes) · EA(p) · GW(p)

Medicine isn't my area, but I'd guess the timelines for vaccine trial completion might be significantly accelerated if some trial participants agreed to be deliberately exposed to SARS-CoV-2, rather than getting data by waiting for participants to get exposed on their own. This practice is known as a "human challenge trial" (HCT), and is occasionally used to get rapid proof-of-concept on vaccines. Using live, wild-type SARS-CoV-2 on fully informed volunteers could possibly provide valuable enough data to reduce the expected development time of the vaccine by several weeks, with a large expected number of lives saved as a result.

Similar usage of HCT's seems to generally be permitted by the relevant ethics committees for low-risk diseases, such as dengue fever, but not high-risk ones, like Ebola or HIV. A brief look at a WHO document on these, and a longer look at relevant US federal law, didn't turn up any hard rules on how dangerous a disease can be before exposure to a "wild-type" virus is forbidden, and both at least mention considering societal benefits as a factor. However, sometimes HCT's for relatively minor diseases like Zika are refused.

The WHO document mentions that these sorts of tests are considered better for selecting between vaccine candidates or supporting evidence than as robust proof of effectiveness for general usage (see Section 5 of the linked document). The document seems to expect that most usages for preventing dangerous diseases will involve modified diseases. Using wild-type coronavirus would be both faster and stronger evidence of efficacy.

There are probably many other people on this forum who could address the expected value of such a trial better than I could, but my suggestion is that EA's engage with the relevant regulators to push for allowing such trials to take place if they would help. Basically, having volunteers put themselves at risk for a faster vaccine would be net positive; independent ethics committees might reject such a study anyways; generating regulatory or public support could make this less likely.

If this were to happen, it seems like a key narrative point would be that the government is allowing people to voluntarily take on risks to find a cure. I think that there would be plenty of volunteers if you asked right, and if some EA's were to do this, it would help their optics tremendously if several of them vocally volunteered.

comment by RoboTeddy · 2020-03-04T16:06:50.977Z · score: 1 (1 votes) · EA(p) · GW(p)

Software engineers could help conduct real-time outbreak response in Seattle: https://twitter.com/trvrb/status/1234931579702538240

comment by ishi · 2020-03-05T13:48:37.912Z · score: -2 (5 votes) · EA(p) · GW(p)

I have mixed feelings about this idea because

1) its still fairly early to know how big a problem this is (and I have heard or read expert opinions on both sides---some say it may be a big problem, while others say it most likely is not)

2) using the EA INT ( Impact-Neglectedness--Tractability) framework (though some use SNT (U) where S= scale=Impact and (U) is 'urgency' (a time discounting or triage factor ---i.e. there's no point in setting up a research program to find a cure if it is going to arrive too late ) I am not sure this issue is Neglected. (I think US govt just allocated 1/2 billion$ to work on this).

there are also already many people---epidimiologists, virologists, health care workers, and health management people ---working on this (internationally). Of course that doesn't mean they couldn't use some help, or that what they are doing is the most 'effective'.

I wouldn't dismiss what these professional people are doing (health departments, CDC, etc.) as innefective or in need of help or input than I would dismiss the efforts of the fire department for fires around here and just try to put the fire myself. But, its possible they could use help (even in a variety of ways---average people can just call the fire dept if they can't put a fire out, and help any people displaced by the fire.)


3) I have already seen a few theoretical epidemiological papers on this subject. https://infomullet.com has one which is not peer reviewed and less theoretical (its more a 'fermi ' or 'back of an envelope' set of calculations (though done on a computer ) than a fully fledged theoretical model. (I think its based on the standard SIR model in epidemiology , or an updated , more complex modification of that).

If one is doing a theoretical model , I think one has to try to review what people are doing or have done , though one can at the same time develop your own models and compare---there is no reason to reinvent the wheel. (While I have done a little labwork as an undergrad, and studied some molecular biology, I am not a 'wet lab worker', so I would be unable to judge anything regarding possible vaccines except from a theoretical POV---and I'm very limited even at that . )

4) My view is, if this turns out to be a big problem, the real thing is to deal with 'population heterogeneity' . I gather China took this route --- they sort of have quarantined entire cities, and have shut down transportation and trade routes . Also, within these cities, they have sort of quarantined subsections of the cities, homes and businesses . (So you have quarantines within quarantines.) (I imagine most people on this forum know that).

This is a preventetive intervention--no different than containing a wildfire (such as the ones that destroyed many homes and lives in California not long ago).

In lieu of a cure, thats the way to go. Just try to keep the problem localized.

In USA, CDC and others are already doing that.


It may be useful if people had an 'app' to know how close and where the virus (or fire) is --no different than an app for checking the weather.

Also, perhaps an 'app' which is a 'checklist' of things you should have at home, carry with you, and also where to go if there is a problem (health facility, or authorities who can tell you where to go if you need help---eg you know someone who i sick).

Also an 'app' which has a list of things you need to do to prepare for a case in which you lose electricity and internet service.


5) As an aside i'm sort of working (possibly with someone else though the project is in its infancy so we haven't agreed on exactly what the project is or what form it takes--we each have our own closely related but distinct approaches---so we are just seeing if collaboration is possible ) . Both of us are more theoretically oriented but want a 'fermi type' solution --something that could be widely useful and applicable, rather than buried in an academic journal written in language noone without a PhD can understand.

It was not oriented towards the corona virus---its a general formalism--but could be applicable to that issue as well. (The idea is really like creating a very small and short 'manual' or handbook, so people can make 'good decisions' based on the information they have---and also figure out what information is most relevant, and how to assign priorities or weights to different possible actions.)

(sorry for the length of this comment. i'm listening to the news right now and they are talking about how they are making emergency proclamations in Seattle and California---and some people are criticizing them because they say they don't have any teeth in them,--just rhetoric or hot air---and don't deal with 'population heterogeneity'. i.e. FB workers are told to stay home and telecommute, while people like post office and delivery workers are told to continue business as usual. )

comment by ZachWeems · 2020-03-25T04:04:07.168Z · score: 3 (2 votes) · EA(p) · GW(p)

I have read and reread this comment and am honestly not sure whether this was a reply to my answer or to something else.

On point 1, I think the past week is a fair indication that the coronavirus is a big problem, and we can let this point pass.

On point 2, as of my answer, there seemed to be no academic talk of human challenge trials to shorten vaccine timelines, regardless of how many were working on vaccines. The problem I see is that if a human challenge trial would shorten timelines, authorities and researchers might still hesitate to run one due to paternalistic attitudes in medical ethics. The problem not that authorities and researchers are not trying to make a vaccine or need amateurs to do their job for them. So, this problem in particular seemed neglected, and worth raising to their attention.

On point 3, I'm not sure if you intended to discuss the expected impact of speeding vaccine development, or if you were confused on what a human challenge trial is? I did not discuss making theoretical models of the impact of the coronavirus on the world.

Points 4 and 5 do not seem to engage with my answer at all.

If this was a mispost, no harm no foul.

Otherwise- I'm not opposed to having a respectful, in-depth discussion of this issue. But the majority of your reply was off-topic and the rest only vaguely engaged with what I wrote. If future replies are similar I'm not going to respond.

comment by ishi · 2020-03-31T22:18:12.098Z · score: -1 (2 votes) · EA(p) · GW(p)

https://sciencehouse.wordpress.com has a more recent study and discussion of 2 other studies at imperial college london and oxford. Science Magazine AAAS also has a whole issue (march 27) on topic. COVID-19 appears to be a real problem but time will tell. (My area has many scientists, but also many poor and uneducated people, so there are lots of 'conspiracy theories' floating around --'viruses of the mind' --there are academic papers on these as well, mostly written by physicists.)

My point 4 i actually view as the main one, unless you are actually developing vaccines in a laboratory or testing them in the field. I have done a tiny bit of lab biology and field biology as a student a but its not my area )

In that sense my comment was 'off topic'---i was talking about prevention, not cures. A term commonly used now is to avoid 'hot spots' --the temperature or incidence of the virus is not the same everywhere, so while it may seem biased, avoid the hot spots . You can say hi to your neighbor, but you cant hug them.

https://johncarlosbaez.wordpress.com may have more discussion that is more relevant to your post.

answer by Agrippa · 2020-03-14T00:40:48.354Z · score: 1 (1 votes) · EA(p) · GW(p)

Could be socially beneficial to start a project developing good online conferencing tech, the landscape is pretty limited ATM.

answer by Peterslattery · 2020-03-10T02:50:30.466Z · score: 1 (1 votes) · EA(p) · GW(p)

[Please disregard this comment for now - we are going to take more time to finalise the pledge before we seek support]

Helping to support and disseminate distribute a viral social contract (i.e., a pledge) to coordinate collective action against the Coronavirus might be one option to consider.

Please see this document for the project background and the current working draft for the pledge.

Please fill this form if you want to help us to distribute the pledge during and after launch (the 11th to 13th of March).

I really think that this is a case where leveraging EA networks can have a very positive social impact.

Taking and sharing the pledge will protect you, your family and your community but also the wider world and future generations. It will also help us to collect useful data about the best way to promote pledges and safety related behaviour.

answer by BenHoffman · 2020-03-04T16:24:11.672Z · score: 1 (1 votes) · EA(p) · GW(p)

This project seems relevant - an app to track COVID-19. Especially given the lack of testing in e.g. the US (and anecdotal evidence from my own social circle suggests it's already more prevalent than official statistics suggest) simple data-gathering seems relevant.

comment by Linch · 2020-03-04T20:20:37.118Z · score: 2 (1 votes) · EA(p) · GW(p)

Are you referring to https://www.covid19risk.com/ , or something else?

comment by tessa · 2020-03-06T00:57:45.246Z · score: 1 (1 votes) · EA(p) · GW(p)

Could be http://www.coepi.org/? (the two groups are currently talking to each other)

10 comments

Comments sorted by top scores.

comment by eca · 2020-03-04T23:11:49.378Z · score: 14 (5 votes) · EA(p) · GW(p)

If you're reading this and have wet lab biology experience (say, have run > 50 PCRs in your life) and would be interested in helping with a project please message me.

Likewise if you have experience making epidemiological models and/or stochastic process models (markov chain monte carlo etc).

I am considering starting 2 projects that require some work to design/ pitch and want to gauge skills/ interest before I invest that time.

A brief sentence about your background would be cool. Thanks!

comment by Linch · 2020-03-04T23:18:13.450Z · score: 4 (2 votes) · EA(p) · GW(p)

In case it helps, I know eca by reputation and would strongly encourage people with the relevant backgrounds to look into this.

comment by Sanjay · 2020-03-05T12:43:19.665Z · score: 1 (1 votes) · EA(p) · GW(p)

I have a maths background, qualified as an actuary in half the industry average time, and am comfortable with stochastic models, including markov chains and monte carlo methods. Are you able to provide more information? In case you don't want to do so in public, I have sent you a direct message via the forum.

comment by Linch · 2020-03-13T00:40:08.779Z · score: 8 (2 votes) · EA(p) · GW(p)

The prediction platform Metaculus is interested in a part or full-time project lead for their new pandemic site.

You get to administer and choose a lot of questions for Metaculus forecasters to predict things on, as well as decide things like question resolution.

I think this is a plausibly useful thing to do because expert forecasts are all over the place, so if you figure out a way to help shepherd/guide amateurs to forecast well, I see a plausible world where Metaculus increasingly gets used to help with more decisions (a la Good Judgement project).

I also generally think it might be valuable to have more superforecasters and similar people in our community.

comment by Linch · 2020-03-04T00:02:08.483Z · score: 4 (2 votes) · EA(p) · GW(p)

Prior work: https://forum.effectivealtruism.org/posts/8chk6DHZXctGHtNoz/covid-19-risk-assessment-app-idea-for-vetting-and-discussion [EA · GW]

comment by tessa · 2020-03-04T02:47:05.805Z · score: 2 (2 votes) · EA(p) · GW(p)

This now has a website: https://www.covid19risk.com/.

The (all-volunteer) team could use more help! You can ask for an invite to the Slack, or send tips about other related efforts, using contact@covid19risk.com.

comment by Tuukka_Sarvi · 2020-03-06T04:48:11.008Z · score: 3 (2 votes) · EA(p) · GW(p)

If it were possible to make a home test kit for COVID, I think that would be helpful. By home test kit I mean an arrangement where people could order these tests by phone or from Internet and they would be mailed to them. Then the person would proceed to take a test sample according to the instructions in the kit and mail it back to the laboratory. The laboratory would test the sample and give the person the results via phone or web.

This kind of arrangement would allow people suspecting that they might have COVID to self-quarantine until they have a certainty (or near certainty since no test is perfect) whether they have the disease or not. Thus, it would lead to less people being exposed to potential early-stage COVID carriers and hinder the disease from spreading. Importantly, it would also protect the health care workers as they would have to come to contact with so many potential COVID carriers.

comment by tessa · 2020-03-06T08:13:55.235Z · score: 3 (2 votes) · EA(p) · GW(p)

You might be interested in the Just One Giant Lab OpenCovid19 project. They just had their first conference call and their goal is to "develop an open source methodology to safely test for the presence of SARS-CoV-2 using tools as common as possible".

comment by tessa · 2020-03-04T02:51:07.366Z · score: 2 (2 votes) · EA(p) · GW(p)

Other prior work: Would activism to ensure local hospitals and health departments are adequately preparing for COVID-19 be high-leverage?

https://forum.effectivealtruism.org/posts/QmMZwYjNBFeMqWh5o/activism-for-covid-19-local-preparedness [EA · GW]

comment by Linch · 2020-03-04T03:15:52.621Z · score: 5 (3 votes) · EA(p) · GW(p)

I'm personally less optimistic about advocacy, because I think there's a strong worry that the professionals should do their jobs (especially since they're likely to be overtaxed very soon) and advocates can easily be more harmful than helpful.

ETA: Maybe I should mention this in that thread.