[Leftcoast]

Quick question, as California ends it’s first week of shelter in place, I see little change in the data. There are lots of variables here, the most important being the incubation period, sometimes reported as 1-14 days with a mean of 5-6 days, but also increasing compliance and delayed test availability skewing reporting.

So ... When should we see the leading edge of data movement?  When will we know if this is making an impact or if adjustments should be made?

[Leftcoast]

Quick question, as California ends it’s first week of shelter in place, I see little change in the data. There are lots of variables here, the most important being the incubation period, sometimes reported as 1-14 days with a mean of 5-6 days, but also increasing compliance and delayed test availability skewing reporting.

So ... When should we see the leading edge of data movement?  When will we know if this is making an impact or if adjustments should be made?

[Farm93]

Quick question, as California ends it’s first week of shelter in place, I see little change in the data. There are lots of variables here, the most important being the incubation period but also iincreasing compliance and delayed test availability skewing reporting.

So ... When should we see the leading edge of data movement?  When will we know if this is making an impact or if adjustments should be made?

Testing will change dramatically over the next 4 weeks, so known cases will be an awful metric.
Hospitals will be near capacity and that will alter the admission policies.   So hospitalizations will be a bad metric too.
Same would be true for number of patients in ICUs.
If that all happens deaths won't be the best metric either, since a greater percentage of those with the virus will die if the healthcare system is really beyond capacity.

Still, given all of those expected changes the best indicator will be the ultimate endpoint.   COVID-19 deaths.
First milestone will be a slowing of the weekly death rate.
If that happens then the health care system is likely treating a great percentage of the sick patients or the number of sick patients is slowing down.

[Farm93]

Quick question, as California ends it’s first week of shelter in place, I see little change in the data. There are lots of variables here, the most important being the incubation period but also iincreasing compliance and delayed test availability skewing reporting.

So ... When should we see the leading edge of data movement?  When will we know if this is making an impact or if adjustments should be made?

Testing will change dramatically over the next 4 weeks, so known cases will be an awful metric.
Hospitals will be near capacity and that will alter the admission policies.   So hospitalizations will be a bad metric too.
Same would be true for number of patients in ICUs.
If that all happens deaths won't be the best metric either, since a greater percentage of those with the virus will die if the healthcare system is really beyond capacity.

Still, given all of those expected changes the best indicator will be the ultimate endpoint.   COVID-19 deaths.
First milestone will be a slowing of the weekly death rate.
If that happens then the health care system is likely treating a great percentage of the sick patients or the number of sick patients is slowing down.

[2006alum]

Your question goes to the more immediate term rebound, but we also have to think medium-term (i.e. in 2-18 months out while we wait for a widespread vaccine.)

In my mind, the two best things we could do besides testing, tracing, and isolation is:

1. Get a serology test to market - unlike a vaccine, this doesn't need to go through clinical trials, and researchers at Mount Sinai produced one several days ago and are shipping it out to other labs:

[tweet]https://twitter.com/NickKristof/status/1240329111748448256?s=20[/tweet]

Here's the tweet from the research announcing the paper explaining their creation of such a test:
[tweet]https://twitter.com/florian_krammer/status/1240432285184405505?s=20[/tweet]
2. Confirm no possibility of reinfection. Thus far I've seen mixed answers to this question, but it seems like the trend in research is reaching the conclusion that reinfection is not possible without the virus mutating first. (Those with MD/immunology backgrounds please clarify if I'm misspeaking?)

See here for why this matters:
[tweet]https://twitter.com/NAChristakis/status/1240689939030790144?s=20[/tweet]
Once we can confirm no possibility of rapid reinfection and then be able to identify who has SARS-COV-2 antibodies (i.e., those who were exposed to SARS-COV-2 and have either recovered or were asymptomatic and outside the window of being contagious), those individuals can return to normal life with no need to shelter in place, etc., because they will be both immune from becoming infected and therefore also would not be able to transmit the virus to others through shedding. (They could still do so through a handshake in which they are carrying live virus, etc.) 

This would be the safest and easiest way for people to resume normal life. In theory, this was the UK's herd immunity strategy before the UK realized (a day later and a dollar short) that without spreading the curve, herd immunity requires culling a fair chunk of the herd in the process.

[2006alum]

Your question goes to the more immediate term rebound, but we also have to think medium-term (i.e. in 2-18 months out while we wait for a widespread vaccine.)

In my mind, the two best things we could do besides testing, tracing, and isolation is:

1. Get a serology test to market - unlike a vaccine, this doesn't need to go through clinical trials, and researchers at Mount Sinai produced one several days ago and are shipping it out to other labs:

[tweet]https://twitter.com/NickKristof/status/1240329111748448256?s=20[/tweet]

Here's the tweet from the research announcing the paper explaining their creation of such a test:
[tweet]https://twitter.com/florian_krammer/status/1240432285184405505?s=20[/tweet]
2. Confirm no possibility of reinfection. Thus far I've seen mixed answers to this question, but it seems like the trend in research is reaching the conclusion that reinfection is not possible without the virus mutating first. (Those with MD/immunology backgrounds please clarify if I'm misspeaking?)

See here for why this matters:
[tweet]https://twitter.com/NAChristakis/status/1240689939030790144?s=20[/tweet]
Once we can confirm no possibility of rapid reinfection and then be able to identify who has SARS-COV-2 antibodies (i.e., those who were exposed to SARS-COV-2 and have either recovered or were asymptomatic and outside the window of being contagious), those individuals can return to normal life with no need to shelter in place, etc., because they will be both immune from becoming infected and therefore also would not be able to transmit the virus to others through shedding. (They could still do so through a handshake in which they are carrying live virus, etc.) 

This would be the safest and easiest way for people to resume normal life. In theory, this was the UK's herd immunity strategy before the UK realized (a day later and a dollar short) that without spreading the curve, herd immunity requires culling a fair chunk of the herd in the process.

[Mick]

This would be the safest and easiest way for people to resume normal life. In theory, this was the UK's herd immunity strategy before the UK realized (a day later and a dollar short) that without spreading the curve, herd immunity requires culling a fair chunk of the herd in the process.

You don't think Boris & Co. realized that out of the gate?

[Mick]

This would be the safest and easiest way for people to resume normal life. In theory, this was the UK's herd immunity strategy before the UK realized (a day later and a dollar short) that without spreading the curve, herd immunity requires culling a fair chunk of the herd in the process.

You don't think Boris & Co. realized that out of the gate?

[burger]

Quick question, as California ends it’s first week of shelter in place, I see little change in the data. There are lots of variables here, the most important being the incubation period, sometimes reported as 1-14 days with a mean of 5-6 days, but also increasing compliance and delayed test availability skewing reporting.

So ... When should we see the leading edge of data movement?  When will we know if this is making an impact or if adjustments should be made?

Shelter in place didn't go statewide until Friday (or Thursday night, I guess), so use that as the starting point.

The incubation period isn't the important part.  It's the time to serious symptoms, which is several days after symptoms show up iirc.  That's when things get bad enough for many to seek care or testing.  And that's more like 10 days after infection.  So don't expect any changes for at least 8-9 more days.  Because testing capacity is still ramping up (a month after community spread began--thanks, Trump!), we'll probably see cases skyrocketing in the coming days.  That's not necessarily an indication of a faster growth rate than before--just testing catching up (though I think I saw that LA was stopping testing except of the worst cases, so that may influence reported rates).

Anyway, variability in testing levels makes counting numbers of cases kind of a fool's game.  I think looking at deaths is more useful.  But death from the virus doesn't happen until like 3 weeks or more after infection.  So, we may not see a decline in the growth of death rates for a month.  That is super bad for hospitals, etc. because the future number of serious cases is somewhat locked in already even if there aren't that many sick people yet.

[burger]

Quick question, as California ends it’s first week of shelter in place, I see little change in the data. There are lots of variables here, the most important being the incubation period, sometimes reported as 1-14 days with a mean of 5-6 days, but also increasing compliance and delayed test availability skewing reporting.

So ... When should we see the leading edge of data movement?  When will we know if this is making an impact or if adjustments should be made?

Shelter in place didn't go statewide until Friday (or Thursday night, I guess), so use that as the starting point.

The incubation period isn't the important part.  It's the time to serious symptoms, which is several days after symptoms show up iirc.  That's when things get bad enough for many to seek care or testing.  And that's more like 10 days after infection.  So don't expect any changes for at least 8-9 more days.  Because testing capacity is still ramping up (a month after community spread began--thanks, Trump!), we'll probably see cases skyrocketing in the coming days.  That's not necessarily an indication of a faster growth rate than before--just testing catching up (though I think I saw that LA was stopping testing except of the worst cases, so that may influence reported rates).

Anyway, variability in testing levels makes counting numbers of cases kind of a fool's game.  I think looking at deaths is more useful.  But death from the virus doesn't happen until like 3 weeks or more after infection.  So, we may not see a decline in the growth of death rates for a month.  That is super bad for hospitals, etc. because the future number of serious cases is somewhat locked in already even if there aren't that many sick people yet.

[2006alum]

This would be the safest and easiest way for people to resume normal life. In theory, this was the UK's herd immunity strategy before the UK realized (a day later and a dollar short) that without spreading the curve, herd immunity requires culling a fair chunk of the herd in the process.

You don't think Boris & Co. realized that out of the gate?

Based on how quickly the UK completely reversed strategies, as well as the available reporting about the process by which they did it, no, Mick, I don't:

From one story on that process:

Political aides tacitly criticised other countries who had taken more dramatic steps, claiming Britain was being “guided by the science” rather than politics.
Towards the end of last week, some ministers and political aides at the top of the government were still arguing that the original strategy of home isolation of suspect cases — but no real restrictions on wider society — was correct, despite almost every other European country taking a much tougher approach, and increasing alarm among SAGE experts.

The thought of months or even a year of social distancing was simply not feasible, some in Johnson’s team still thought at that point. They continued to privately defend the controversial “herd immunity” approach outlined to the media by Vallance, even as other aides scrambled to claim the UK had never considered it to be policy. And there was fury behind the scenes among members of Johnson’s team at the likes of Rory Stewart and Jeremy Hunt, who had been publicly saying the government had got it wrong.

But data from Italy — presented to the government before it was published by experts at Imperial College on Monday — changed all that. Their report confirmed the earlier fears of the epidemiologists who had been calling for more drastic action. It showed that the NHS would indeed soon be overwhelmed if the UK’s “mitigation” strategy continued. The country had to move right away to a “suppression” approach with much more social distancing and restrictions on normal life, Imperial concluded.

Vallance and Johnson accepted the chilling findings and the prime minister finally advised social distancing on Monday. The experts who had been demanding this for days wondered what the lost time would mean in terms of deaths.

Not to mention public reversals like this:
[tweet]https://twitter.com/devisridhar/status/1241089612296212493?s=20[/tweet]

[2006alum]

This would be the safest and easiest way for people to resume normal life. In theory, this was the UK's herd immunity strategy before the UK realized (a day later and a dollar short) that without spreading the curve, herd immunity requires culling a fair chunk of the herd in the process.

You don't think Boris & Co. realized that out of the gate?

Based on how quickly the UK completely reversed strategies, as well as the available reporting about the process by which they did it, no, Mick, I don't:

From one story on that process:

Political aides tacitly criticised other countries who had taken more dramatic steps, claiming Britain was being “guided by the science” rather than politics.
Towards the end of last week, some ministers and political aides at the top of the government were still arguing that the original strategy of home isolation of suspect cases — but no real restrictions on wider society — was correct, despite almost every other European country taking a much tougher approach, and increasing alarm among SAGE experts.

The thought of months or even a year of social distancing was simply not feasible, some in Johnson’s team still thought at that point. They continued to privately defend the controversial “herd immunity” approach outlined to the media by Vallance, even as other aides scrambled to claim the UK had never considered it to be policy. And there was fury behind the scenes among members of Johnson’s team at the likes of Rory Stewart and Jeremy Hunt, who had been publicly saying the government had got it wrong.

But data from Italy — presented to the government before it was published by experts at Imperial College on Monday — changed all that. Their report confirmed the earlier fears of the epidemiologists who had been calling for more drastic action. It showed that the NHS would indeed soon be overwhelmed if the UK’s “mitigation” strategy continued. The country had to move right away to a “suppression” approach with much more social distancing and restrictions on normal life, Imperial concluded.

Vallance and Johnson accepted the chilling findings and the prime minister finally advised social distancing on Monday. The experts who had been demanding this for days wondered what the lost time would mean in terms of deaths.

Not to mention public reversals like this:
[tweet]https://twitter.com/devisridhar/status/1241089612296212493?s=20[/tweet]

[BostonCard]

If we could rewind the clock to the end of January, and could have developed mass testing capability then, one could have had a process to contain the virus that involved mass test and trace, with strict quarantine of positive cases and isolation of contacts, and the world could have continued with little disruption.  Instead we waited until there were 1000’s of known cases and probably ten times as many unknown and untraceable cases.  Alas, that horse is out of the barn, and now we are playing catch up.  If you haven’t read it, please read “the hammer and the dance”, which I linked to in a separate post.  I am optimistic we can bring this under control, but there is also a chance that things will spiral out of control country-wide.

BC

[BostonCard]

If we could rewind the clock to the end of January, and could have developed mass testing capability then, one could have had a process to contain the virus that involved mass test and trace, with strict quarantine of positive cases and isolation of contacts, and the world could have continued with little disruption.  Instead we waited until there were 1000’s of known cases and probably ten times as many unknown and untraceable cases.  Alas, that horse is out of the barn, and now we are playing catch up.  If you haven’t read it, please read “the hammer and the dance”, which I linked to in a separate post.  I am optimistic we can bring this under control, but there is also a chance that things will spiral out of control country-wide.

BC

[BostonCard]

Your question goes to the more immediate term rebound, but we also have to think medium-term (i.e. in 2-18 months out while we wait for a widespread vaccine.)

In my mind, the two best things we could do besides testing, tracing, and isolation is:

1. Get a serology test to market - unlike a vaccine, this doesn't need to go through clinical trials, and researchers at Mount Sinai produced one several days ago and are shipping it out to other labs:

[tweet]https://twitter.com/NickKristof/status/1240329111748448256?s=20[/tweet]

Here's the tweet from the research announcing the paper explaining their creation of such a test:
[tweet]https://twitter.com/florian_krammer/status/1240432285184405505?s=20[/tweet]
2. Confirm no possibility of reinfection. Thus far I've seen mixed answers to this question, but it seems like the trend in research is reaching the conclusion that reinfection is not possible without the virus mutating first. (Those with MD/immunology backgrounds please clarify if I'm misspeaking?)

See here for why this matters:
[tweet]https://twitter.com/NAChristakis/status/1240689939030790144?s=20[/tweet]
Once we can confirm no possibility of rapid reinfection and then be able to identify who has SARS-COV-2 antibodies (i.e., those who were exposed to SARS-COV-2 and have either recovered or were asymptomatic and outside the window of being contagious), those individuals can return to normal life with no need to shelter in place, etc., because they will be both immune from becoming infected and therefore also would not be able to transmit the virus to others through shedding. (They could still do so through a handshake in which they are carrying live virus, etc.) 

This would be the safest and easiest way for people to resume normal life. In theory, this was the UK's herd immunity strategy before the UK realized (a day later and a dollar short) that without spreading the curve, herd immunity requires culling a fair chunk of the herd in the process.

Looks like the South Koreans have developed a serology test.

[tweet]https://twitter.com/vaitor/status/1241337295606657033?s=21[/tweet]

BC

[BostonCard]

Your question goes to the more immediate term rebound, but we also have to think medium-term (i.e. in 2-18 months out while we wait for a widespread vaccine.)

In my mind, the two best things we could do besides testing, tracing, and isolation is:

1. Get a serology test to market - unlike a vaccine, this doesn't need to go through clinical trials, and researchers at Mount Sinai produced one several days ago and are shipping it out to other labs:

[tweet]https://twitter.com/NickKristof/status/1240329111748448256?s=20[/tweet]

Here's the tweet from the research announcing the paper explaining their creation of such a test:
[tweet]https://twitter.com/florian_krammer/status/1240432285184405505?s=20[/tweet]
2. Confirm no possibility of reinfection. Thus far I've seen mixed answers to this question, but it seems like the trend in research is reaching the conclusion that reinfection is not possible without the virus mutating first. (Those with MD/immunology backgrounds please clarify if I'm misspeaking?)

See here for why this matters:
[tweet]https://twitter.com/NAChristakis/status/1240689939030790144?s=20[/tweet]
Once we can confirm no possibility of rapid reinfection and then be able to identify who has SARS-COV-2 antibodies (i.e., those who were exposed to SARS-COV-2 and have either recovered or were asymptomatic and outside the window of being contagious), those individuals can return to normal life with no need to shelter in place, etc., because they will be both immune from becoming infected and therefore also would not be able to transmit the virus to others through shedding. (They could still do so through a handshake in which they are carrying live virus, etc.) 

This would be the safest and easiest way for people to resume normal life. In theory, this was the UK's herd immunity strategy before the UK realized (a day later and a dollar short) that without spreading the curve, herd immunity requires culling a fair chunk of the herd in the process.

Looks like the South Koreans have developed a serology test.

[tweet]https://twitter.com/vaitor/status/1241337295606657033?s=21[/tweet]

BC

[magnus]

How long would it take for the FDA to allow any of these to be used?

[magnus]

How long would it take for the FDA to allow any of these to be used?

[BostonCard]

EUA could be granted in 24 hours. The FDA might want some certainty that the test is effective at diagnosing what it claims to diagnose.

BC

[BostonCard]

EUA could be granted in 24 hours. The FDA might want some certainty that the test is effective at diagnosing what it claims to diagnose.

BC

[Goose]

The next logical question is Do we have what is required to do the "dance" part of the equation? I don't have any idea, but I do hope that the powers that be are scrambling like mad to ensure we do in the interval where we are waiting for the curve to flatten.
I do not believe that even if we had no testing issues that we could have reacted with an epidemiological response. Certainly Spain, Italy, and France, who did not have the testing issues the USA did, were unable to do so. These nations also allow enforcement of quarantine without as much legal actibity as would be required in the USA. It is perfectly possible to defeat a disease via epidemiological methods in the absence of a good test for the disease. It has been done many times in the past. However, if the disease has reached the "community spread" point such that the resources of the authorities are overwhelmed, the actions taken by Italy, Spain et al are the only practical response.
China found that their test had a problem with "false negatives". This is somewhat of a misnomer, because in many cases the patient had not yet developed full scale disease and therefore tested negative even in the presence of the disease. The Chinese added a CT scan/chest X-ray as an additional method of diagnosing the disease. China also realized that if you have a symptomatic person who tests negative, you can't release him into the general population. You have to quarantine them anyway, because if you don't they can develop and spread the disease. It is worthy of note that the USA uses different tests than China did. Korea also uses a different test. The sensitivities may be different, we do not yet know.
Korea was fortunate that they had a residual epidemiological capability left over from the SARS outbreak that could be directed towards the problem. They also have a population that can generally be relied upon to do what they are requested to do (although the woman who tested positive and left the hospital to attend multiple religious events almost blew up the Korean response).
I just hope that if we "flatten the curve" we can respond adequately  with contact tracing, testing, and quarantine. Many people aren't going to like the fact they in effect need to be locked up, as the people at Travis were. If we are lucky, that is coming. I hope we are ready.

[Goose]

The next logical question is Do we have what is required to do the "dance" part of the equation? I don't have any idea, but I do hope that the powers that be are scrambling like mad to ensure we do in the interval where we are waiting for the curve to flatten.
I do not believe that even if we had no testing issues that we could have reacted with an epidemiological response. Certainly Spain, Italy, and France, who did not have the testing issues the USA did, were unable to do so. These nations also allow enforcement of quarantine without as much legal actibity as would be required in the USA. It is perfectly possible to defeat a disease via epidemiological methods in the absence of a good test for the disease. It has been done many times in the past. However, if the disease has reached the "community spread" point such that the resources of the authorities are overwhelmed, the actions taken by Italy, Spain et al are the only practical response.
China found that their test had a problem with "false negatives". This is somewhat of a misnomer, because in many cases the patient had not yet developed full scale disease and therefore tested negative even in the presence of the disease. The Chinese added a CT scan/chest X-ray as an additional method of diagnosing the disease. China also realized that if you have a symptomatic person who tests negative, you can't release him into the general population. You have to quarantine them anyway, because if you don't they can develop and spread the disease. It is worthy of note that the USA uses different tests than China did. Korea also uses a different test. The sensitivities may be different, we do not yet know.
Korea was fortunate that they had a residual epidemiological capability left over from the SARS outbreak that could be directed towards the problem. They also have a population that can generally be relied upon to do what they are requested to do (although the woman who tested positive and left the hospital to attend multiple religious events almost blew up the Korean response).
I just hope that if we "flatten the curve" we can respond adequately  with contact tracing, testing, and quarantine. Many people aren't going to like the fact they in effect need to be locked up, as the people at Travis were. If we are lucky, that is coming. I hope we are ready.