The Diamond Princess stories are not anecdotal. That is still the best hard data we have on a fixed group.

It's literally the best experiment that could be conducted short of deliberately infecting a captive population.

Keegan99 said:

---

Also, if it's merely an artifact of some mysterious change in collective behavior, why haven't infections reignited after the threshold is reached and behavior reverts?

---

Glad you brought that up
Dr Marcus Aurellius who posts here often is suffering through a second wave in his Birmingham Area ER
There was a post Mardi Gras spike that they suffered through and at the end of the first lockdown infection rates and ER utilization was manageable but after the lockdowns expired his patient census is back at full ICU capacity

I will bump that thread

None of those populations in AL or LA reached ~20%, so I'm not sure why that's relevant?

I'd also point out that on the USS Roosevelt, only about 25% of the sailors ever tested positive. On the Diamond Princess, the figure was just under 20%.

That's two closed environments where avoiding exposure was all but impossible, yet the portion infected remained bizarrely low.

Keegan99 said:

---

None of those populations in AL or LA reached ~20%, so I'm not sure why that's relevant?

I'd also point out that on the USS Roosevelt, only about 25% of the sailors ever tested positive. On the Diamond Princess, the figure was just under 20%.

That's two closed environments where avoiding exposure was all but impossible, yet the portion infected remained bizarrely low.

---

You presume that once a problem was discovered large numbers of people were interacting freely with the entire population of the ship. Again we do not know what sort of mitigation strategies were adopted when they became aware there was a covid outbreak onboard I am certain is was not business as usual.

hope you are right I just can't see how an infectious disease would burn out at 20%

Last point is the UK BJ thought he could go the Sweden route the numbers got too bad and he had to institute a lockdown Again IMO when the % gets over 5% behaviors change the % continues to climb because it takes awhile for the new behaviors to be effective % tops out near 20 and starts going down

Keegan99 said:

---

Except Sweden (and perhaps more strikingly, the Diamond Princess cruise ship) stand to refute the theory that it's just some consequence of behavioral happenstance.

---

The Diamond Princess departed from Japan on January 20th.

On January 25th a passenger disembarked for medical care in Hong Kong due to symptomatic illness.

On February 3rd the ship was notified that the passenger who was ill and dropped off in Hong Kong had tested positive for COVID-19. The crew and passengers were notified of the situation that same day and given counsel about importance of social distancing and how to monitor for symptoms.

On February 5th the passengers were quarantined to their cabins but the crew had to continue work the ship and were not quarantined.

"On the Diamond Princess, transmission largely occurred among passengers before quarantine was implemented, whereas crew infections peaked after quarantine (6)." https://www.cdc.gov/mmwr/volumes/69/wr/mm6912e3.htm#:~:text=The%20Diamond%20Princess%20and%20Grand,time%20(10).

So behavioral happenstance does appear to have changed the rate of transmission amongst passengers thus lowering the number of overall passengers that would have been infected.

The crew unfortunately didn't have the opportunity for the same level of mitigation measures and so their rate of transmission did not go down after February 5th.

The crews rate of transmission instead went up, and as you will see below far exceeded the 10-20% "Burn Out Theory" even though going forward from February 5th they were essentially their own community.

Of the total 712 that ultimately tested positive on the Diamond Princess, the highest total I can find for passengers was 218 on February 13th (8 days after quarantine and 10 days after social distancing pep talk). This was per a W.H.O. News Release on February 13th. https://www.who.int/news-room/detail/13-02-2020-remarks-by-dr-michael-ryan-executive-director-who-health-emergencies-programme-at-media-briefing-on-covid-19-on-13-february-2020

The ship had 1,045 person Crew.
The ship had 2,666 passengers.

The first crew member became ill February 2nd and tested positive and disembarked. By February 9th 20 symptomatic crew members had tested positive. And the rate of transmission continued grow while they worked on the ship.

I'd like to have final exact numbers of crew infected and passengers infected, but I'm not seeing them other than what's noted above. The CDC does say that the time of transmission to passengers was largely prior to February 5th passenger quarantine. February 13th, 8 days later,10 days after social distancing started, for an illness with a 3-14 day incubation period and with natural course data showing 85% of those who would be symptomatic should have already had been having symptoms and tested when the WHO reported 218 passengers had COVID. If we add another 15% to that number we get 250 sick passengers.

Estimated 250/2666 passengers = 9.4% infected with Behavioral Modification
Estimated 462/1045 crew members = 44% Without significant Behavioral Modification

That 44% of crew infected is FAR ABOVE the 10-20% "Burn Out" Immunity... and that's with all of the crew members working around each other and living on the same ship decks with each other, away from passengers.




Bottom line... I really don't think this thing is going to just burn out when we get to 10-20% of the population infected. It's not just "gonna disappear" any time soon.

Wear a mask in public settings at all times when indoors.
Wear a mask in public settings when not able to physically distance from others outside.
Avoid large indoor gatherings.
Give schools the support they need to be able to follow CDC guidelines.

Under that model, that implies their coefficient of variation was much lower than the value of 2.5 that produced a 20% herd immunity threshold. Lower CV means higher HI per the model.

That would make sense, since the CV measures variation in susceptibility, and the crew likely had very similar probabilities of exposure and thus low variation.

In fact, stands to reason that the CV would be lower for the passengers as well, which highlights the successful impact of the NPI.



Just goes to show, though, that their model is very sensitive to the CV assumption, as Fitch pointed out in the first page. And, its important to note that, in the linked document, the authors noted a couple interesting things:

• Other studies showed CV < 1 (though this may have been arrived at using a methodology that buckets people into groups and inherently underestimates the CV)
• The author notes that their own estimates are likely too high

Given the fact that, in the models, CV is squared, a range which starts under 1 and ends at 2.5 is really wide. Which makes me personally pretty nervous in interpreting 20% to be the actual threshold.

(But, i don't know anything about any of this and this is not my field. Just found the study really interesting)

Quote:

---

Bottom line... I really don't think this thing is going to just burn out when we get to 10-20% of the population infected. It's not just "gonna disappear" any time soon.

Wear a mask in public settings at all times when indoors.
Wear a mask in public settings when not able to physically distance from others outside.
Avoid large indoor gatherings.
Give schools the support they need to be able to follow CDC guidelines.

---

Agreed!

So is it your contention that somehow ~80% of the people on the Diamond Princess were never exposed? (This curiously included couples where only one partner tested positive?)

And that similarly, ~75% of the crew on the Roosevelt were never exposed? Despite no real ability for an aircraft carrier crew to quarantine?

They both put significant portions of people into isolation

thanks for the facts Ag E Doc
the crew infection like prison populations and meat packing facilities the covid does not burn out at 20%

Keegan99 said:

---

So is it your contention that somehow ~80% of the people on the Diamond Princess were never exposed? (This curiously included couples where only partner tested positive?)

And that similarly, ~75% of the crew on the Roosevelt were never exposed? Despite no real ability for an aircraft carrier crew to quarantine?

---

not all exposures lead to an infection and yes a cruise ship and an air craft carrier are large places and if the problem was detected when 5% of the crew is infected it would be possible to keep the total infected rate at 20 % through mitigation strategies

We'll see if Levitt is right.

He's calling a total of 170k excess deaths in a month, at which point COVID, as an event causing excess mortality, will burnout.

This would be roughly a 0.05% population fatality rate, generally in line with Europe.

if you are using mortality rates as your metric and trying to tie reverse engineer overall infection rate
how are you accounting for break throughs in treatments?
Milan's fatality rate was probably worse than NYC
and NYC will probably be worse than other major metro areas in the US

bay fan said:

---

Extremism regarding things I never mentioned is always a great fall back.

Let me speak more slowly, masks are working to a greater extent in Northern California where compliance is greater. You may not like that but the state numbers verify it. Masks worked when everybody wears them. The more that don't wear them, the less they work. It's pretty simple.

W

---

you don't have data for this. You are just hoping and wishing for it to be true.

there are many other possible explanations for why 1 area has more cases than another area.

Compare Northern California to wyoming

I mean, if we're going by anecdotes, I was in San Diego for the entirety of last week and the devotion to mask wearing was WELL beyond what I've observed in both DFW and Houston.

San Diego has had a mask ordinance since May 1st.

Their detected infection graph is below.

Thanks Keegan
I'm with you and I'll chose to believe this Nobel Prize winning Dr. (Video is from June 6th)
BTW he was saying in the Video that Sweden's initial projection was 60.000 deaths but was predicted when the video was shot was 6,000. Today it's at 5700

Keegan99 said:

---

We'll see if Levitt is right.

He's calling a total of 170k excess deaths in a month, at which point COVID, as an event causing excess mortality, will burnout.

This would be roughly a 0.05% population fatality rate, generally in line with Europe.

---

With the expanded US testing capabilities we have seen in the summer months the last 6 weeks we have been pretty consistently seeing a 3 week after case diagnosis death rate of about 2.2%.

This week's 4-Day daily average for the United States (Tues-Fri) was 71,717 x 2.2 = 1,578 dead per day 3 weeks from now. Even if we assume that this past week was our peak thanks to wide spread mitigation mandates in our country's most populous and hard hit areas, then we still have to come back down from that peak which took us 7 weeks to reach.

I'm estimating now, but will know better tomorrow, that Texas will drop about 10-15% lower on cases next week compared to this past week. This past week in Texas was a drop of 11.1% from about 2 weeks ago.

Let's say the US decline in cases mirrors that of Texas, then the United States would be back down to around 20,000 cases per day in about 6 weeks. Which would project to still produce about 400 deaths per day 3 weeks after that.



I don't see how we are done with COVID in 4 weeks with 170,000 total deaths.

Unfortunately there will be some who will think that and because there is this hope that Mr. Levitt is right they won't practice the same degree of mitigation efforts that they would if they knew that what they do today will determine where we are in the future with this pandemic.

It is nothing personal Keegan, but this is what frustrates me about Mr. Levitt and others that discount the importance of mitigation efforts because it can lead to school closures, no MLB attendance, potentially limited college football attendance, school athletic department financial problems. If we will do the little things I think we can have the bigger things... and yes that also means people not losing loved ones prematurely.

I wonder if Mr. Levitt would recommend all nursing homes to drop their visitation shut downs in 4 weeks since it is all going to be over in the United States based on his math.

Since he thinks that we will have reached a critical number of infections to cause a pandemic burn out and a subsequent very low incidence of new infections or deaths then I don't see why he would object, since I'm sure he has faith that his math has taken into consideration all of the science, medicine and mitigation factors that have created the curves on which he bases his math.

His claim was not that no one could or would die, but that the virus would no longer be producing excess death throughout the population.

Just as people still die from the flu after the when flu season ends. It just no longer causes excess death. And one would not recklessly allow a flu infected individual to enter a nursing home in June just because "flu season is over".

But if you want to attribute some disingenuous strawman to the Nobel Laureate, have at it, I guess.

Keegan99 said:

---

Please tell me how detected infections and subsequent fatalities are related?

---

If your threshold for who gets tested is consistent and there is a consistent access to be able to get people tested (even if results are too slow to aid with contact tracing) you will capture a certain picture or snap shot of who all is infected. Now as you mentioned in the past you won't catch all infections. And as you also mentioned depending upon what point in time you are looking at and what the testing threshold and capabilities were at the time you have to use different multipliers to estimate how many people were truly infected at that time.

Well when things are steady which they have been for 2+ months in terms who qualifies to get tested and what our testing capabilities are then you will start seeing a more consistent snap shot that allows you to make predictions based on natural course of the illness. The treatments with remdesivir, dexamethasone, convalescent plasma and prone positioning with intubation/ventilation have also been pretty steady for 2+ months. So in this window of time in Texas & the United States we are seeing pretty consistent percent of deaths that lag daily new cases by three weeks which is common with the natural course of this disease.

This is the chart I've posted on previous threads with Texas data...

This past week (July 21-24) Texas ended up averaging 174 deaths per day which was 2.3% of the daily deaths from three weeks before.

You could not do this with early New York Data compared to current data because we were missing too many cases because of our botched early efforts to get a workable test up to scale. We were only testing people who had significant risk factors based on travel, or only people with known contacts who had already tested positive, or only people sick enough to be in the hospital, or only people who had all of the most classic symptoms. The testing thresholds and capabilities were changing so much in March and April it would make it hard to have consistently captured the same percent of total cases and so much harder to predict future deaths.


Sweden is similar because they have changed their testing philosophy and capabilities also.

It's harder to show now because you don't find many graphs that show testing per capita over time.
But Sweden initially wasn't testing hardly anybody, unless you were very sick and essentially needing to consider hospitalization. Compared to Norway and their other comparable nordic neighbors they were testing less than half per capita.

Those other countries like Norway got their pandemics under control and haven't had to test at nearly as high a volume because when you have few cases you have few contacts to trace and test. (to illustrate you could use the comparison Norway is to Sweden, as South Korea is to the USA -- earlier adequate testing leads to less overall testing).

In Sweden they did continue have a significant number of cases and at one point late in the spring/early summer they decided that they were going to try and really ramp up their testing efforts beyond just testing people who were sick enough to consider hospitalization. This was done in an effort to try and suppress their death's per capita which is the 7th worst in the world and 5th worst in the world for countries with greater than 10 million people. Sweden to Ramp Up Coronavirus Testing (Link). [url=https://medicalxpress.com/news/2020-05-sweden-ramp-coronavirus.html][/url]

As they tested more, starting in late May and Early June they saw their number of documented cases sky rocket.
Coronavirus Cases Hit Daily Record in Sweden as Testing Ramps Up (Link from Mid June). I would argue that they didn't have more cases in late May and June than they did before or after their "pseudo surge" they just documented more cases and subsequently the deaths that were to trail 3 weeks later weren't going to go up because the actual level of infection had not changed substantially.

What ultimately happened is their increased testing helped them contact trace and isolate the more mild cases and instead of having a more slow burn like they were on before they were doing adequate testing, they actually now were able to identify and isolate more mild disease and subsequently slowed their rate of transmission. They saw their cases decline and subsequently their deaths have declined as well.

Keegan99 said:

---

But if you want to attribute some disingenuous strawman to the Nobel Laureate, have at it, I guess.

---

He is a Nobel Laureate in physics/chemistry. He isn't a medical professional. He isn't an infectious disease specialist. He isn't an epidemiologist. He isn't trained in public health.

Just because someone is intelligent and very accomplished in their field doesn't mean they are an expert in someone else's field of expertise.

I wouldn't pretend to tell an auto mechanic or mechanical engineer what is going to happen with a new problem that arises in their field. The Dunning Krueger Effect can still happen regardless of level of intelligence or how accomplished someone is in another unrelated field.

Levitt also predicted that Italy was half way through their deaths when they were at 2,503 deaths, they are currently at north of 35,000 deaths.

He also predicted Israel would only have 10 deaths from COVID-19, and they are now at 453 and potentially just now reaching the peak of their second wave (after lifting mask requirements temporarily as the weather got much hotter in May).

Article with Levitt Predictions (LINK)[url=https://www.jpost.com/israel-news/nobel-laureate-israel-will-have-no-more-than-ten-coronavirus-deaths-621407][/url]

With all due respect, when it comes to potential life and death predictions that affect the behavior of thousands/millions he really needs to stick to his field of expertise.

Keegan99 said:

---

Just as people still die from the flu after the when flu season ends. It just no longer causes excess death. And one would not recklessly allow a flu infected individual to enter a nursing home in June just because "flu season is over".

---

People can be significantly contagious for two days prior to ever showing symptoms, so we wouldn't knowingly be letting a contagious family member into the nursing home. Obviously if some one is overtly symptomatic they can't visit. But if this thing will truly be "Over" to use Mr. Levitt's word in 4 weeks and the incidence of infections in 4 weeks is going to be very low, then do we have to restrict visitors to the nursing homes.

If the rates of new infections is very low and this thing is over then the answer is no we don't have to restrict them because there will be an abundance of testing available since hardly anyone is getting infected anymore and we can make sure every nursing home has a rapid point of care COVID testing capability and then we can simply screen family coming into the nursing home for a visit with a rapid 15 minute test to make sure they aren't the very rare pre-symptomatic COVID patient that will be circulating in our communities after 4 more weeks once "Corona Season is Over".









It's late. I'm getting tired. I'm probably coming off cranky in my last few posts and if so, I apologize .That's not my intention, and as I mentioned in a previous post none of this is personal. I think we all want to be doing what we feel the right thing is for public health. It's been a good day discussing this with you all. Thank you for starting this thread. Goodnight!

Great discussion both Keegan and AgE Doc. I learned a lot and really glad you both maintained composure so the rest of us could learn something. I know you guys do it because of your own intellectual curiosity, but super beneficial for the rest of us that don't have the time, or in my case, the intellectual capacity to decipher these research papers and relevance to current situation. Very much appreciated

Considering NY has 20%+ antibodies, and has been steady at 800 cases / day and 40 deaths / day since June 8, I'd say there's something to this theory.