A Third Solution(paulbuchheit.blogspot.com)
paulbuchheit.blogspot.com
A Third Solution
https://paulbuchheit.blogspot.com/2020/04/a-third-solution.html
516 comments
Quotes from the article:
“It’s easy to fall into dystopian visions of the future — a world shut down by one virus after another”
“It doesn’t have to be that way. ..... Ubiquitous screening is the key.”
The approach is interesting and the possibility of eliminating large scale spread of covid, flu and others is attractive.
However the idea of requiring a saliva swab from every visitor to an office or event has the potential to create an equally terrifying dystopian future where those samples are used to collect and use other data (DNA for example).
How long before screening companies offers to provide free screening and access control systems in return for anonymised data?
This kind of solution needs to have very well thought out privacy rules supported by strong and enforceable legislation to protect the individuals rights.
“It’s easy to fall into dystopian visions of the future — a world shut down by one virus after another”
“It doesn’t have to be that way. ..... Ubiquitous screening is the key.”
The approach is interesting and the possibility of eliminating large scale spread of covid, flu and others is attractive.
However the idea of requiring a saliva swab from every visitor to an office or event has the potential to create an equally terrifying dystopian future where those samples are used to collect and use other data (DNA for example).
How long before screening companies offers to provide free screening and access control systems in return for anonymised data?
This kind of solution needs to have very well thought out privacy rules supported by strong and enforceable legislation to protect the individuals rights.
This isn't a unique idea. This is the mainstream view. Everyone knows we need more testing and that testing is the only way to effectively ease distancing rules. That was a pretty extensive writeup to say what we've been hearing from all rational information outlets for a month.
I really hope this works. Without some new testing technology, I just don't see how we can stop the spread of this disease. A month ago, the US did about 100,000 tests per day. Yesterday, the US did about 200,000 tests. Growth in testing started off as exponential but now it looks linear.[1] Even if testing capability doubled every month, it would take 8 months before we could test every American once a week. (200,000 * 2^8 == 51,200,000, which is 15% of the population.)
It only took a couple of months for 20% of New Yorkers to get infected.[2] If we assume that half the population will get this disease over the next two years, and we assume an infection fatality rate of 0.3%, that's around 500,000 deaths. (328,000,000 * 0.5 * 0.003 == 492,000). Those are optimistic projections. The IFR is likely higher and the R0 is somewhere between 3 and 9[3], so that means at somewhere between 60% and 90% of the population needs to be infected before we get herd immunity.[4]
Unless there are radical improvements in testing and/or treatment, I think we'll end up with at least 500,000 deaths in the US. That would mean we're about 10% of the way through this catastrophe. So strap in, it's gonna be a long ride.
1. https://twitter.com/COVID19Tracking/status/12538071759457443...
2. https://twitter.com/NYGovCuomo/status/1253353516803993600
3. https://wwwnc.cdc.gov/eid/article/26/7/20-0282_article
4. For the relationship between R0 and herd immunity, see figure 2 of this paper: https://academic.oup.com/cid/article-pdf/52/7/911/847338/cir...
It only took a couple of months for 20% of New Yorkers to get infected.[2] If we assume that half the population will get this disease over the next two years, and we assume an infection fatality rate of 0.3%, that's around 500,000 deaths. (328,000,000 * 0.5 * 0.003 == 492,000). Those are optimistic projections. The IFR is likely higher and the R0 is somewhere between 3 and 9[3], so that means at somewhere between 60% and 90% of the population needs to be infected before we get herd immunity.[4]
Unless there are radical improvements in testing and/or treatment, I think we'll end up with at least 500,000 deaths in the US. That would mean we're about 10% of the way through this catastrophe. So strap in, it's gonna be a long ride.
1. https://twitter.com/COVID19Tracking/status/12538071759457443...
2. https://twitter.com/NYGovCuomo/status/1253353516803993600
3. https://wwwnc.cdc.gov/eid/article/26/7/20-0282_article
4. For the relationship between R0 and herd immunity, see figure 2 of this paper: https://academic.oup.com/cid/article-pdf/52/7/911/847338/cir...
The post refers to an alternative method of testing for COVID-19 based on surface plasmon resonance that would have significant advantages, but unfortunately it provides absolutely no real substantiation that the test exists or works. The link about surface plasmon resonance goes to a generic Wikipedia page, the link about saliva is a small scale study that was conducted on RT-PCR not surface plasmon resonance, and the actual link to the team goes to a nothing more than splash screen with the company logo on it.
If there is substance to this then it would be massively in Paul's (and the company's) interest to better link to that in both the post and the company web site. At the moment, it looks like at worst vaporware or at best, something so early stage it's years out from viability.
If there is substance to this then it would be massively in Paul's (and the company's) interest to better link to that in both the post and the company web site. At the moment, it looks like at worst vaporware or at best, something so early stage it's years out from viability.
You can rank people by the likely benefit of testing them. Interestingly, it goes up as the square of the number of people they interact with daily. (Because their risk of having it increases, and also the number they are likely to spread it to.)
So you can allocate tests by sorting by (# of daily contacts in a closed space) ^ 2.
But as PB says, it should be practical to test everyone every day.
So you can allocate tests by sorting by (# of daily contacts in a closed space) ^ 2.
But as PB says, it should be practical to test everyone every day.
> If we were able to identify and quarantine everyone who is contagious, including those who are asymptomatic, then we could let everyone else out of lockdown and resume ordinary social and economic activity.
> Even with imperfect screening, if we are able to prevent 90% of disease transmission, then the virus’s reproductive number, or R0, will drop below one and the pandemic will quickly fade. There is no risk of reintroduction from the outside because any new outbreaks will quickly be caught and contained. If used consistently, there will be no second wave, ever.
I'm not sure this "test and release" strategy works unless absolutely everybody gets tested simultaneously.
Even if testing of the total population can be completed in a week (a highly ambitious timeframe), there's still time for people released on Day 1 to be reinfected by people who don't get tested until Day 6.
Then you have to go through who knows how many follow-up rounds of testing absolutely everybody not in quarantine to identify those people. When responding to new outbreaks involves re-testing large populations of people, you're going to run into many problems. Notification, compliance, testing fatigue, etc.
Sounds like a logistical nightmare.
> Even with imperfect screening, if we are able to prevent 90% of disease transmission, then the virus’s reproductive number, or R0, will drop below one and the pandemic will quickly fade. There is no risk of reintroduction from the outside because any new outbreaks will quickly be caught and contained. If used consistently, there will be no second wave, ever.
I'm not sure this "test and release" strategy works unless absolutely everybody gets tested simultaneously.
Even if testing of the total population can be completed in a week (a highly ambitious timeframe), there's still time for people released on Day 1 to be reinfected by people who don't get tested until Day 6.
Then you have to go through who knows how many follow-up rounds of testing absolutely everybody not in quarantine to identify those people. When responding to new outbreaks involves re-testing large populations of people, you're going to run into many problems. Notification, compliance, testing fatigue, etc.
Sounds like a logistical nightmare.
This is essentially what South Korea did. Granted it was done in a different way but everyone was “tested.” I was there just as covid-19 was on the rise and every shop, every station, and every high traffic area had people set up with thermal guns. Shop staff were having their temps checked before their shifts started. Everyone was gloved and masked. And once they had a proper test in place it was made easily accessible — even set up drive through testing.
And now they are pretty much back to normal while much of the world is still at a standstill.
And now they are pretty much back to normal while much of the world is still at a standstill.
Really interesting read and sounds like it could be a game changer for testing - I know in NY we've been clamoring for increased testing for a while to help us reopen. I've got two questions I'm curious about:
0) Sensitivity/specificity: Any data yet on what the sensitivity and specificity of this form of test for SARS-COV-2 will be? And, is work in characterizing all of that far enough along that we can expect to see emergency authorization by the FDA and scale up happening sooner rather than later?
1) Reagent supply: The biggest problem with PCR tests and all seems to have been reagent shortages and supply chains dependent on manufacturers not able to scale. Assuming the test is approved, are there any operational advantages to this approach in terms of securing reagents to overcome that problem?
Thanks for the fascinating read!
0) Sensitivity/specificity: Any data yet on what the sensitivity and specificity of this form of test for SARS-COV-2 will be? And, is work in characterizing all of that far enough along that we can expect to see emergency authorization by the FDA and scale up happening sooner rather than later?
1) Reagent supply: The biggest problem with PCR tests and all seems to have been reagent shortages and supply chains dependent on manufacturers not able to scale. Assuming the test is approved, are there any operational advantages to this approach in terms of securing reagents to overcome that problem?
Thanks for the fascinating read!
A very recent Yale study suggests promises for saliva-based detection:
“While saliva has shown promise for SARS-CoV-2 detection, very few studies have directly compared it to the current gold standard, nasopharyngeal (NP) swab. So, we compared NP and saliva samples from COVID-19 patients and self-collected samples from asymptomatic healthcare workers”
“COVID-19 patients: SARS-CoV-2 detection from saliva is comparable to (or better than!) NP swabs and more consistent over time ...”
“Plus, the detection of SARS-CoV-2 from the saliva of two asymptomatic healthcare workers (...and counting!) who tested negative from their NP swabs suggests that saliva could be a viable alternative for identifying mild or subclinical infections.”
https://twitter.com/awyllie13/status/1252996627217801218?s=2...
“While saliva has shown promise for SARS-CoV-2 detection, very few studies have directly compared it to the current gold standard, nasopharyngeal (NP) swab. So, we compared NP and saliva samples from COVID-19 patients and self-collected samples from asymptomatic healthcare workers”
“COVID-19 patients: SARS-CoV-2 detection from saliva is comparable to (or better than!) NP swabs and more consistent over time ...”
“Plus, the detection of SARS-CoV-2 from the saliva of two asymptomatic healthcare workers (...and counting!) who tested negative from their NP swabs suggests that saliva could be a viable alternative for identifying mild or subclinical infections.”
https://twitter.com/awyllie13/status/1252996627217801218?s=2...
They're in a "startup incubator" for bio firms in South San Francisco.[1][2] The incubator is run by a unit of Johnson and Johnson. Multiple companies (51 are listed, but some may no longer be there) share 30,000 square feet of workspace there. "We welcome new resident companies with the infrastructure and tools they need to get up and running on day one. This capital efficient model takes time and investment out of the equation—eliminating the normal setup typically required of a startup."
[1] https://businesssearch.sos.ca.gov/Document/RetrievePDF?Id=04...
[2] https://jlabs.jnjinnovation.com/locations/jlabs-ssf
[3] https://jlabs.jnjinnovation.com/JLABSNavigator#/location/Bay...
[1] https://businesssearch.sos.ca.gov/Document/RetrievePDF?Id=04...
[2] https://jlabs.jnjinnovation.com/locations/jlabs-ssf
[3] https://jlabs.jnjinnovation.com/JLABSNavigator#/location/Bay...
PB, can you elaborate on which other possible answers you've found for fast, easy, and abundant tests?
I'm working with a team that has a test that detects proteins associated with covid. It works like a pregnancy test and does not need a special scanner. Would love to discuss further.
I'm working with a team that has a test that detects proteins associated with covid. It works like a pregnancy test and does not need a special scanner. Would love to discuss further.
Solutions like this give me hope that we can actually return to something resembling normal life in the future. I hope Paul's got everything he needs in funding and resources to pursue all three of his goals.
I hope everyone has noticed that if this approach could wipe out COVID-19 it could wipe out influenza and even the common cold.
A fourth way: We throw as many resources as we can at sampling undiagnosed populations, like the recent NYC study that suggests 20% of the city (10% of the state) has antibodies already.
We could get real confidence that it's safe enough to return to normal, acceping that COVID is a new disease that's just going to be around, the 5th coronavirus that we deal with seasonally.
We could get real confidence that it's safe enough to return to normal, acceping that COVID is a new disease that's just going to be around, the 5th coronavirus that we deal with seasonally.
There is an episode of Sliders (Fever, Season 1, Episode 3), where they slide into a world affected by an infection with no cure, and scanners have been placed at the entrance to every store to detect if you have it.
In the show the disease is used as a classist thing or something. Anyways, its bacterial not viral, and they discover than antibiotics were never discovered so the Professor scrapes some fungus off some trash and takes it and is cured.
In the show the disease is used as a classist thing or something. Anyways, its bacterial not viral, and they discover than antibiotics were never discovered so the Professor scrapes some fungus off some trash and takes it and is cured.
Cheap & working saliva test would be such a tracking game changer that I am really hoping it works, so good luck!
A) Does he have any financial interest in PreDxion Bio? This sounds like another smaller startup claiming bullshit.
B) Testing for O2 levels using already cheap + widely available pulse oximeters is probably an insufficient but necessary measure.
B) Testing for O2 levels using already cheap + widely available pulse oximeters is probably an insufficient but necessary measure.
Thank you Paul. Two questions :
Does this approach bypass the reagent shortages ?
What are the specificity / sensitivity metrics ?
Does this approach bypass the reagent shortages ?
What are the specificity / sensitivity metrics ?
Paul, is there anything HN readers can do to help with this?
Ah, of course. We create a new type of [patentable] ubiquitous technology and sell our way out of this. Leave it to the entrepreneurial mind...
Sorry for being cynical -- I just feel suspicious of this particular tired mindset to addressing complex public health and social issues, ones that intersect with (and aggravate) many other pre-existing social dilemmas. There are a thousand other ways to look at this that don't involve a small cornered market, I just doubt the entrepreneurial mind knows how to parse for it on its own. When you have a hammer...
Wake me up when someone's talking about this sort of thing amongst members of a consortium building open patents, not from some guy with plain-as-day zero-to-one ambitions. I'll root for someone who sees the interlocking opportunities, not someone who speaks about personal aspirations to "wipe out COVID-19" in 2020
Sorry for being cynical -- I just feel suspicious of this particular tired mindset to addressing complex public health and social issues, ones that intersect with (and aggravate) many other pre-existing social dilemmas. There are a thousand other ways to look at this that don't involve a small cornered market, I just doubt the entrepreneurial mind knows how to parse for it on its own. When you have a hammer...
Wake me up when someone's talking about this sort of thing amongst members of a consortium building open patents, not from some guy with plain-as-day zero-to-one ambitions. I'll root for someone who sees the interlocking opportunities, not someone who speaks about personal aspirations to "wipe out COVID-19" in 2020
I’m having difficulty understanding why SPR would be more scalable than LFAs for this type of frequent screening? And what does the ROC look like for this startup’s SPR assay?
Frankly, I don’t understand how this test is supposed to work, and I’ve used a Biacore! It might be helpful to have a technical explanation available, for domain experts to evaluate.
Frankly, I don’t understand how this test is supposed to work, and I’ve used a Biacore! It might be helpful to have a technical explanation available, for domain experts to evaluate.
> Even with imperfect screening, if we are able to prevent 90% of disease transmission, then the virus’s reproductive number, or R0, will drop below one and the pandemic will quickly fade. There is no risk of reintroduction from the outside because any new outbreaks will quickly be caught and contained. If used consistently, there will be no second wave, ever.
This is dangerously wrong.
If you reduce R0 below 1, you may stop community spread. You will not eradicate it however, unless this is done globally for a prolonged period of time with no error. This cannot be done with the proposed solution.
We can (probably) stop this from hurting the vulnerable population while we find a long term solution like an inoculation, but we cannot just skip that and call everything good after some period of low / no new cases in a region.
This is dangerously wrong.
If you reduce R0 below 1, you may stop community spread. You will not eradicate it however, unless this is done globally for a prolonged period of time with no error. This cannot be done with the proposed solution.
We can (probably) stop this from hurting the vulnerable population while we find a long term solution like an inoculation, but we cannot just skip that and call everything good after some period of low / no new cases in a region.
The latest on the COVID-19 R0 is a median of 5.7, up from the previously thought range of 2.2 - 2.7.
https://wwwnc.cdc.gov/eid/article/26/7/20-0282_article
This means herd immunity kicks in at 82% of the population.
Antibody testing appears to be showing infection rates are a lot higher than previously thought as well.
Both of those things together mean that 1) "there's no way to stop it" 2) "it might not be as dangerous as we thought".
But who knows, right? There's a ton of science that needs to be done to find out what's really going on. Large-scale, accurate, randomized testing will hopefully fill out the data picture.
Because the US has an awful medical system where access to care comes through your job, it seems to me that more people will die (from non-virus causes) than from the economic damage than from the virus itself.
I guess we'll know more in a few years.
https://wwwnc.cdc.gov/eid/article/26/7/20-0282_article
This means herd immunity kicks in at 82% of the population.
Antibody testing appears to be showing infection rates are a lot higher than previously thought as well.
Both of those things together mean that 1) "there's no way to stop it" 2) "it might not be as dangerous as we thought".
But who knows, right? There's a ton of science that needs to be done to find out what's really going on. Large-scale, accurate, randomized testing will hopefully fill out the data picture.
Because the US has an awful medical system where access to care comes through your job, it seems to me that more people will die (from non-virus causes) than from the economic damage than from the virus itself.
I guess we'll know more in a few years.
Not sure if I'm missing something, isn't Saliva based testing almost/basically here?
1. https://finance.yahoo.com/news/orasure-technologies-receives...
2. https://www.politico.com/states/new-jersey/story/2020/04/23/...
1. https://finance.yahoo.com/news/orasure-technologies-receives...
2. https://www.politico.com/states/new-jersey/story/2020/04/23/...
After the whole Theranos debacle, I'm most curious how due diligence was done on this team and their technology.
The most important thing was mentioned at the top of that article: there is a very effective thing everyone can do right now and does not require any technology: keep your distance. Stay at home, if you can, and in the public, keep distances. Face masks are an important help in this distancing, as there are always situations, where you can't quite maintain the "safe" 2m distance. If we all kept perfect distancing, the virus would be erradicated in 4-8 weeks.
As this isn't always possible and mistakes are made, testing is indeed the other important part of fighting any spreadable disease. The comparison of the Covid-19 infections across different countries show clearly, how effective strict testing is. Any technology, which allows for frequent, wide-range testing is a big help in fighting diseases. If you could run a test when you are like feeling like getting a cold, the common cold and the flu would become much rarer diseases. (Especially if the west picks up the asian habit on wearing face masks, when you have the flu or a cold).
HIV could be erradicated quickly, if there was even a yearly test of the whole population and anyone tested positive would get treated with the antiviral medicine which already exists for quite a while. Once treated, the virus count decreases rapidly and there is very little risk of spreading the virus, especially when taking minimal precautions.
All above of course require for the tests to be available to literally everyone. So this should be a state run function, where you can get tested without any question asked about possible health insurance and also certain treatments should be given free of any charge.
As this isn't always possible and mistakes are made, testing is indeed the other important part of fighting any spreadable disease. The comparison of the Covid-19 infections across different countries show clearly, how effective strict testing is. Any technology, which allows for frequent, wide-range testing is a big help in fighting diseases. If you could run a test when you are like feeling like getting a cold, the common cold and the flu would become much rarer diseases. (Especially if the west picks up the asian habit on wearing face masks, when you have the flu or a cold).
HIV could be erradicated quickly, if there was even a yearly test of the whole population and anyone tested positive would get treated with the antiviral medicine which already exists for quite a while. Once treated, the virus count decreases rapidly and there is very little risk of spreading the virus, especially when taking minimal precautions.
All above of course require for the tests to be available to literally everyone. So this should be a state run function, where you can get tested without any question asked about possible health insurance and also certain treatments should be given free of any charge.
There’s not enough consumable tests, and there is currently no installed base of surface plasmon resonance machines, nor any of the other myriad types of constant-monitoring systems that have actually been proven to work for many years now in high-risk facilities (I developed some for the government 15 years ago and it wasn’t new).
If you want my opinion, the right way to approach this is using the consumable tests to maximum effect for mass viral surveillance by contact group hierarchies. For instance, pool an entire school district on a single test, and then hunt down positives by school then class etc. There won’t be enough tests to find every case. That’s okay; others in contact are suspect anyway even if their test would have been negative at that time. The contact group discovery is simple too: cell tower data (civil liberties notwithstanding).
This is pretty obvious, but it doesn’t work because our medical system is set up to charge individuals, and the highest priority will always be hospital admittances. This does little for the patient, but does protect others in the hospital.
If you want my opinion, the right way to approach this is using the consumable tests to maximum effect for mass viral surveillance by contact group hierarchies. For instance, pool an entire school district on a single test, and then hunt down positives by school then class etc. There won’t be enough tests to find every case. That’s okay; others in contact are suspect anyway even if their test would have been negative at that time. The contact group discovery is simple too: cell tower data (civil liberties notwithstanding).
This is pretty obvious, but it doesn’t work because our medical system is set up to charge individuals, and the highest priority will always be hospital admittances. This does little for the patient, but does protect others in the hospital.
An honest query by a non-medical professional as I'm sincerely curious...
Paul advocates daily saliva-based testing, but as an intermediary imperfect, but "better than nothing" measure, what are the benefits and drawbacks of requiring people entering public shared spaces to have their body temperatures taken via handheld temperature guns or infrared monitors, a measure that's already taking place in much East Asia (Greater China, Japan, Korea, etc.) in public shared spaces like malls, restaurants, office buildings? My understanding is that these methods are not as accurate as direct thermometers or Paul's saliva-based test; nonetheless, they would detect a good portion of mildly symptomatic people and also have the benefit of externally signaling to the populace to continue "sheltering-in-place" if they have a fever.
Is there any issue with supply chains? Or is there scientific evidence disproving the effectiveness of this precautionary measure that's already in place in so many regions that have already seemed to have crested the first wave of the pandemic?
Paul advocates daily saliva-based testing, but as an intermediary imperfect, but "better than nothing" measure, what are the benefits and drawbacks of requiring people entering public shared spaces to have their body temperatures taken via handheld temperature guns or infrared monitors, a measure that's already taking place in much East Asia (Greater China, Japan, Korea, etc.) in public shared spaces like malls, restaurants, office buildings? My understanding is that these methods are not as accurate as direct thermometers or Paul's saliva-based test; nonetheless, they would detect a good portion of mildly symptomatic people and also have the benefit of externally signaling to the populace to continue "sheltering-in-place" if they have a fever.
Is there any issue with supply chains? Or is there scientific evidence disproving the effectiveness of this precautionary measure that's already in place in so many regions that have already seemed to have crested the first wave of the pandemic?
I’m sure this is the last of their concerns, but the word “disposable” and “daily” immediately made me think, “great, more garbage.”
It would be nice if they could figure out a recycling system — or at least make sure the vials are made of somewhat environmentally neutral glass — out of the gate, rather than have another problem to solve after there are millions of these out there
It would be nice if they could figure out a recycling system — or at least make sure the vials are made of somewhat environmentally neutral glass — out of the gate, rather than have another problem to solve after there are millions of these out there
I guess it's a YC17 company. The founders are Caroline Landau, Tim Cornell, Walker McHugh. From 2016: Landau was an MBA candidate, the other two founders have biomedical research/medical backgrounds: Walker McHugh, Co-Founder, PreDxion Bio / Biomedical engineering candidate, University of Michigan Dr. Tim Cornell, Co-Founder, PreDxion Bio / Pediatric Critical Care Physician, University of Michigan (https://www.hbs.edu/openforum/openforum.hbs.org/goto/challen...)
At least until recently, the company focused on making diagnostics for immune disorders (microkine) for CAR-T patients which I can't find much detail on. I don't know if it's related to the SPR-based tests mentioned in the pb post.
They received a government business grant (SBIR) in 2018 and have some VC funding.
It looks like this post demonstrates their pivot to a specific infectious disease, and from a hospital provider setting to a public setting.
As an ex-advisor to a successful (in bio) VC fund, this is not something I would really spend a lot of time considering. There are too many non-technical hurdles that would need to be jumped before this was widespread, popular, effective, and profitable.