[BostonCard]

New paper from Chinese researchers.

https://www.nature.com/articles/s41591-020-0869-5

We observed the highest viral load in throat swabs at the time of symptom onset, and inferred that infectiousness peaked on or before symptom onset. We estimated that 44% (95% confidence interval, 25–69%) of secondary cases were infected during the index cases’ presymptomatic stage, in settings with substantial household clustering, active case finding and quarantine outside the home. Disease control measures should be adjusted to account for probable substantial presymptomatic transmission.

Almost half of secondary cases were infected while the index case was presymptomatic.  In other words, #&$@!

BC

[BostonCard]

New paper from Chinese researchers.

https://www.nature.com/articles/s41591-020-0869-5

We observed the highest viral load in throat swabs at the time of symptom onset, and inferred that infectiousness peaked on or before symptom onset. We estimated that 44% (95% confidence interval, 25–69%) of secondary cases were infected during the index cases’ presymptomatic stage, in settings with substantial household clustering, active case finding and quarantine outside the home. Disease control measures should be adjusted to account for probable substantial presymptomatic transmission.

Almost half of secondary cases were infected while the index case was presymptomatic.  In other words, #&$@!

BC

[2006alum]

New paper from Chinese researchers.

https://www.nature.com/articles/s41591-020-0869-5

We observed the highest viral load in throat swabs at the time of symptom onset, and inferred that infectiousness peaked on or before symptom onset. We estimated that 44% (95% confidence interval, 25–69%) of secondary cases were infected during the index cases’ presymptomatic stage, in settings with substantial household clustering, active case finding and quarantine outside the home. Disease control measures should be adjusted to account for probable substantial presymptomatic transmission.

Almost half of secondary cases were infected while the index case was presymptomatic.  In other words, #&$@!

BC

Yeah, that's terrifying, because short of randomized, nondiscriminatory rapid testing, how can you possibly stop infectious individuals from unknowingly spreading the virus? This certainly suggests that we're going to need to keep many social distancing measures in place for quite a while for anyone who is in a higher-risk group or does not want a non-zero chance of being laid up in the hospital for several weeks.

[2006alum]

New paper from Chinese researchers.

https://www.nature.com/articles/s41591-020-0869-5

We observed the highest viral load in throat swabs at the time of symptom onset, and inferred that infectiousness peaked on or before symptom onset. We estimated that 44% (95% confidence interval, 25–69%) of secondary cases were infected during the index cases’ presymptomatic stage, in settings with substantial household clustering, active case finding and quarantine outside the home. Disease control measures should be adjusted to account for probable substantial presymptomatic transmission.

Almost half of secondary cases were infected while the index case was presymptomatic.  In other words, #&$@!

BC

Yeah, that's terrifying, because short of randomized, nondiscriminatory rapid testing, how can you possibly stop infectious individuals from unknowingly spreading the virus? This certainly suggests that we're going to need to keep many social distancing measures in place for quite a while for anyone who is in a higher-risk group or does not want a non-zero chance of being laid up in the hospital for several weeks.

[BostonCard]

An alternative is aggressive tracking and tracing.  But it would have to be aggressive, and would require putting a lot of people in quarantine.

BC

[BostonCard]

An alternative is aggressive tracking and tracing.  But it would have to be aggressive, and would require putting a lot of people in quarantine.

BC

[M T]

I remember the much earlier results that suggested a serial interval of something like 2.5 days less than the incubation period, along with the results that the virus is heaviest early in the infection.  I am not surprised at a finding that a significant portion of the infections are from pre-symptomatic patients.

Look at the Santa Clara County Dashboard for Testing.  I see what appears to be multiple highs that are about 5 days apart: 
   March 27, March 31/April 1, April 6, April 11.

I believe we're seeing a wave of infections at roughly 5 days.  For whatever reason, there was a bump some weeks back.  A new wave of people are getting the disease, becoming infectious about the same time, and spreading it to a new generation of people.  Some of those people are symptomatic and are going in for tests.  What we are seeing is when those tests come back.

I would expect this bump to spread out and become less distinct over time because of variability in exposure to infectious time, exposure to symptom time, and symptom onset to test time.   If it doesn't, that would suggest that another factor is in play.

If this were a 7-day cycle, I'd believe it is more likely some day of the week influence.  I'm not aware of 5 day cycle that isn't related to the illness, but it could be, say, the cycle at which test materials are delivered or something hard for me to know.

Note that this cycle is the same as the incubation period.  A possibility:  when a person get the symptoms, someone goes to the drug store, some may go to a medical facility.   The increases the chances of exposure for both sick & not-yet-sick patients. Voila, a 5 day cycle of new infections.   The serial interval might be shorter (say, 3 days), but the activities that go one involved with symptom onset may perpetuate a 5-day cycle of infections.

This cycle might happen even if the infected person doesn't realize they have symptoms (or doesn't want to admit it).  "Honey, I've got a {headache, sore throat, stuffy nose}.  Can you go buy me some meds?  And pick me up some chicken soup while you're out, please."  At which point the spouse or parent (infected 2.5 days ago & newly infectious, but presymptomatic) goes out.


Santa Clara County's data can be shown as a table, but can't be downloaded  San Francisco's data can be downloaded.  I grabbed their test reports and summed all the results since the beginning for every 5th day (ie, one bucket is the sum of day 1, day 6, ...).
The last full set of buckets are  209, 232, 284, 243, 237.   That may be a bump for day 3 + 5*n.

[M T]

I remember the much earlier results that suggested a serial interval of something like 2.5 days less than the incubation period, along with the results that the virus is heaviest early in the infection.  I am not surprised at a finding that a significant portion of the infections are from pre-symptomatic patients.

Look at the Santa Clara County Dashboard for Testing.  I see what appears to be multiple highs that are about 5 days apart: 
   March 27, March 31/April 1, April 6, April 11.

I believe we're seeing a wave of infections at roughly 5 days.  For whatever reason, there was a bump some weeks back.  A new wave of people are getting the disease, becoming infectious about the same time, and spreading it to a new generation of people.  Some of those people are symptomatic and are going in for tests.  What we are seeing is when those tests come back.

I would expect this bump to spread out and become less distinct over time because of variability in exposure to infectious time, exposure to symptom time, and symptom onset to test time.   If it doesn't, that would suggest that another factor is in play.

If this were a 7-day cycle, I'd believe it is more likely some day of the week influence.  I'm not aware of 5 day cycle that isn't related to the illness, but it could be, say, the cycle at which test materials are delivered or something hard for me to know.

Note that this cycle is the same as the incubation period.  A possibility:  when a person get the symptoms, someone goes to the drug store, some may go to a medical facility.   The increases the chances of exposure for both sick & not-yet-sick patients. Voila, a 5 day cycle of new infections.   The serial interval might be shorter (say, 3 days), but the activities that go one involved with symptom onset may perpetuate a 5-day cycle of infections.

This cycle might happen even if the infected person doesn't realize they have symptoms (or doesn't want to admit it).  "Honey, I've got a {headache, sore throat, stuffy nose}.  Can you go buy me some meds?  And pick me up some chicken soup while you're out, please."  At which point the spouse or parent (infected 2.5 days ago & newly infectious, but presymptomatic) goes out.


Santa Clara County's data can be shown as a table, but can't be downloaded  San Francisco's data can be downloaded.  I grabbed their test reports and summed all the results since the beginning for every 5th day (ie, one bucket is the sum of day 1, day 6, ...).
The last full set of buckets are  209, 232, 284, 243, 237.   That may be a bump for day 3 + 5*n.

[Goose]

An alternative is aggressive tracking and tracing.  But it would have to be aggressive, and would require putting a lot of people in quarantine.

BC

As I see it, that is really the only alternative. An exposed person can test negative at day N and also become positive (and spreading the disease) at day N + 1 through probably day N + <incubation period>, give or take. The best approach is quarantine all the contacts of any positive case for at least <incubation period>, and then test them. If they are negative, assume they don't have it and let them go. The incubation period is on the order of two weeks, as I understand it. I see no other choice. The definition of "contacts" is a bit of a problem, but I am sure that will sort itself out. Start with a very inclusive one and see what happens. Better than a total lockdown of everybody.

[Goose]

An alternative is aggressive tracking and tracing.  But it would have to be aggressive, and would require putting a lot of people in quarantine.

BC

As I see it, that is really the only alternative. An exposed person can test negative at day N and also become positive (and spreading the disease) at day N + 1 through probably day N + <incubation period>, give or take. The best approach is quarantine all the contacts of any positive case for at least <incubation period>, and then test them. If they are negative, assume they don't have it and let them go. The incubation period is on the order of two weeks, as I understand it. I see no other choice. The definition of "contacts" is a bit of a problem, but I am sure that will sort itself out. Start with a very inclusive one and see what happens. Better than a total lockdown of everybody.

[M T]

It isn't sufficient to quarantine contacts of known cases and wait for symptoms to show.  Those contacts may have already infected others and may be pre-symptomatic or asymptomatic.   Either you need to test those contacts, or quarantine their contacts.

For each case, contact tracing for a case wants to discover both the infector (the source of the infection) and any infectees (of the case being tracked).


In particular, suppose person A gets infected by community spread (no contact with someone with the disease) on day 1 and that A never noticess any symptoms.   Now person A infects person B on day 3, and person C on day 7.   Person B gets symptoms on day 8.  Person C becomes infectious on day 9.
 
On day 9, person A is interviewed as a contact of person B and reports no symptoms.   If that were the extent of the contact tracing, person C is missed and goes on to infect others.

Suppose that all contacts (even one-time) are asked to get tests.
Person A may no longer be positive by a PCR test.  Hopefully, person A is also tested for IgM antibodies, which show they recently had the illness and may be an infector of B.    If neither test is positive, the contact with C is missed.   But if A tests as recently positive, person A's contacts should be traced.  Then C may be found.

--------

I'm not sure whether the authorities can force person A to take either test.  But, if the testing is available, A could be encouraged to take the test.
Even 2 months from now, using a manual trace & test approach, I'm not sure the authorities would encourage A to take the test if the only contact was that A and B had encountered each other in the grocery store one time.

With the contact tracing app, person A may be told (anonymously?) that they had contact with a known case.  Person A may shelter-in-place and consider that is sufficient for a quarantine, without ever taking a test.

----------

What is the motivation for a person to take either a PCR or antibody test given that someone says they were exposed but you don't recognize any symptoms?
1)  If you find that you do have the disease, you will take extra precautions to avoid giving it to someone around you.
     1a) If you do have the disease, those around you can get tested so they might not spread it to others.
2)  If you repeatedly take the test and get negative results, you might shorten a quarantine.
3)  If there are some treatment that can be done early in the course of the disease to improve the outcome, you might do that.
4)  You may wish to know early so you can get things in order (for illness or even death).

-----------

To more quickly reduce the infection rate, tests (supporting the trace & test effort) should be free.  Priority should be given for those with the most contacts in the last 2 weeks.  (For instance, I have no contact with anyone but my wife.  A grocery store clerk's test should get priority over my test.)

I am aware of a healthcare worker who had multiple symptoms in Texas.  She was advised to go to the ER because its results came back sooner, but walked out when told it would cost $75.  5 days later, her symptoms are going away, and she doesn't have any results yet.  The facility she works at (who did say they'd pay the ER once she told them about it) has been in limbo.   That's going to be pretty much a wasted test.

[M T]

It isn't sufficient to quarantine contacts of known cases and wait for symptoms to show.  Those contacts may have already infected others and may be pre-symptomatic or asymptomatic.   Either you need to test those contacts, or quarantine their contacts.

For each case, contact tracing for a case wants to discover both the infector (the source of the infection) and any infectees (of the case being tracked).


In particular, suppose person A gets infected by community spread (no contact with someone with the disease) on day 1 and that A never noticess any symptoms.   Now person A infects person B on day 3, and person C on day 7.   Person B gets symptoms on day 8.  Person C becomes infectious on day 9.
 
On day 9, person A is interviewed as a contact of person B and reports no symptoms.   If that were the extent of the contact tracing, person C is missed and goes on to infect others.

Suppose that all contacts (even one-time) are asked to get tests.
Person A may no longer be positive by a PCR test.  Hopefully, person A is also tested for IgM antibodies, which show they recently had the illness and may be an infector of B.    If neither test is positive, the contact with C is missed.   But if A tests as recently positive, person A's contacts should be traced.  Then C may be found.

--------

I'm not sure whether the authorities can force person A to take either test.  But, if the testing is available, A could be encouraged to take the test.
Even 2 months from now, using a manual trace & test approach, I'm not sure the authorities would encourage A to take the test if the only contact was that A and B had encountered each other in the grocery store one time.

With the contact tracing app, person A may be told (anonymously?) that they had contact with a known case.  Person A may shelter-in-place and consider that is sufficient for a quarantine, without ever taking a test.

----------

What is the motivation for a person to take either a PCR or antibody test given that someone says they were exposed but you don't recognize any symptoms?
1)  If you find that you do have the disease, you will take extra precautions to avoid giving it to someone around you.
     1a) If you do have the disease, those around you can get tested so they might not spread it to others.
2)  If you repeatedly take the test and get negative results, you might shorten a quarantine.
3)  If there are some treatment that can be done early in the course of the disease to improve the outcome, you might do that.
4)  You may wish to know early so you can get things in order (for illness or even death).

-----------

To more quickly reduce the infection rate, tests (supporting the trace & test effort) should be free.  Priority should be given for those with the most contacts in the last 2 weeks.  (For instance, I have no contact with anyone but my wife.  A grocery store clerk's test should get priority over my test.)

I am aware of a healthcare worker who had multiple symptoms in Texas.  She was advised to go to the ER because its results came back sooner, but walked out when told it would cost $75.  5 days later, her symptoms are going away, and she doesn't have any results yet.  The facility she works at (who did say they'd pay the ER once she told them about it) has been in limbo.   That's going to be pretty much a wasted test.

[Leftcoast]

Per several news reports today less than half of the aircraft carrier crew testing positive were symptomatic.  

https://www.google.com/amp/s/www.militar...s.html/amp

This is clearly a younger and fitter sample base than cruise ships - I foresee many published studies comparing these two sample types.

[Leftcoast]

Per several news reports today less than half of the aircraft carrier crew testing positive were symptomatic.  

https://www.google.com/amp/s/www.militar...s.html/amp

This is clearly a younger and fitter sample base than cruise ships - I foresee many published studies comparing these two sample types.

[Goose]

Per several news reports today less than half of the aircraft carrier crew testing positive were symptomatic.  

https://www.google.com/amp/s/www.militar...s.html/amp

This is clearly a younger and fitter sample base than cruise ships - I foresee many published studies comparing these two sample types.

Hopefully they will follow up on the asymptomatic cases for 3 weeks or so. We need to know how many of these positive cases become symptomatic. That will give us a better idea of the scope of this issue. One would assume that in 3 or 4 days most of the people who are going to have symptoms will have them, but given that this virus has properties we didn't expect, it is better to be sure, especially while we have a captive test group.

[Goose]

Per several news reports today less than half of the aircraft carrier crew testing positive were symptomatic.  

https://www.google.com/amp/s/www.militar...s.html/amp

This is clearly a younger and fitter sample base than cruise ships - I foresee many published studies comparing these two sample types.

Hopefully they will follow up on the asymptomatic cases for 3 weeks or so. We need to know how many of these positive cases become symptomatic. That will give us a better idea of the scope of this issue. One would assume that in 3 or 4 days most of the people who are going to have symptoms will have them, but given that this virus has properties we didn't expect, it is better to be sure, especially while we have a captive test group.