Special Pages

Tuesday, March 17, 2020

Remember The Flu Boat?

Remember the story of the Diamond Princess that we talked about here in mid-February?  I had said having all those people quarantined on a boat might be a good opportunity to study the virus.

It turns out that it really was quite a good place to study COVID-19.  Thanks to a post on Watts Up With That and hat tip to Borepatch for seeing it first, we get to learn some pretty remarkable facts about COVID-19.

To begin with, it really was a scenario almost designed to provide massive numbers of infected people and mass casualties; a way to observe the virus spread across the thousands on board, perhaps killing hundreds.  If the fatality rate was the kind of number people were talking about months ago (10% or more), that could have meant 400 killed.
As you might imagine, before they knew it was a problem, the epidemic raged on the ship, with infected crew members cooking and cleaning for the guests, people all eating together, close living quarters, lots of social interaction, and a generally older population. Seems like a perfect situation for an overwhelming majority of the passengers to become infected.
Patient zero was on board on January 20th to the 25th.  They man wasn't diagnosed until February 1st.  In those five days, he spread the virus.  In the next week, the situation was like that previous paragraph.  People unaware they were spewing virus working in close contact with each other. 
  • It was a "worst case" scenario, as people were confined for weeks with other infected people in very close quarters.  While this protected the general public from the infection, it likely maximized the spread of the virus on board the ship.
  • It was a "worst case" scenario as the population on board skewed dramatically to older - and thus, more vulnerable - people.
The reality turned out far less scary.  Obviously, if you're one of the people who dies, it's very serious to you and your family, so I'm talking in the “detached, passive, scientist voice.”  Despite the dire situation...
  • 83% (confidence interval of 82.7% – 83.9%) of the passengers never got the disease at all.  Said the other way, only 17% of the passengers were infected.
  • The oldest portion of the passengers, over 80 years old, had a higher infection rate: 25%, but that's not even twice the rate of the general population.
  • Wait - it gets better.  Slightly less than half the passengers (48.6% ± 2.0%) who would test positive for the disease did NOT get sick and showed NO symptoms.  They never knew they had it.
  • The young (under 20) and old (over 50) disproportionally showed no symptoms after being infected
  • The overall, age-adjusted death rate was 1.2% (7 cases total)
These two graphs go together.

This shows the percentage of each age group on board the Diamond Princess who did NOT contract the virus.  The youngest, birth to 9 years, looks to say 93% did not test positive for the virus.  Note that the lowest percentage who didn't test positive is the 80-89 year old cohort, which says the chance of contracting the virus seem so go up with age.  Note how the colored bars are shorter for ages 50-59 than the three groups above it, and then shorter every additional decade.

The next plot is percentage of each age group who showed no symptoms.

The birth to 9 years old group was 100% symptom free.  From the first plot, roughly 7% of that age group tested positive and that test was the only way it could be known they had the virus.  We can keep going on those; the second group in first plot shows 78% did not contract the virus, so 22% did.  Of the 22%, 60% showed no symptoms.  Even in the oldest group, with 25% testing positive for the virus, almost half of them (45%) had no symptoms.

The study itself is Estimating the infection and case fatality ratio for COVID-19 using age-adjusted data from the outbreak on the Diamond Princess cruise ship (PDF warning).  The study authors define an age-adjusted Infection Fatality Rate of 1.2% with a possible interval of 0.38%–2.7%. The wide uncertainty range is due to the small number of deaths.  Although this might not be relevant to an epidemiologist studying the disease, I see this as 7 deaths out of 3,711 passengers and crew on the Diamond Princess.  I don't see any reason to exclude the 83% who didn't get infected because they were exposed and could have been infected.  7 deaths out of 3711 people on board is a 0.19% fatality rate.  I hear the number 0.1% for the annual flu, but I don't know if that compares more directly to my raw numbers or their age-adjusted 1.2%.

These 3711 people were in a virtual incubator for the virus and there's every reason to think they were all at risk.  According to this article, a Japanese Infectious Diseases expert went onboard the Diamond Princess and was afraid of how poorly they implemented their quarantine.  The fact that 83% never got the disease hints that it may not be as communicable as everyone fears. 

It seems that the more I learn about COVID-19, the less concerned I am about it. 


  1. Two questions:

    1) And the study does...what, exactly...about the fact that infected persons were pulled off the boat as soon as they became symptomatic?

    2) Extrapolate the shipboard casualty figures nationwide.
    Get back to me on how this is therefore less concerning than before.
    Now, let's get serious: as any sort of "study", this was finger-banged from beginning to end, coming and going.

    Do a study, instead of on the MV Death Princess, on the Apocalypse Acres Old Pharts Retirement Home in Seattle, where 30-40% or more of the "young, healthy" staff were fired after showing signs of the virus, and half the U.S. Kung Flu deaths (at this early date) are currently clustered, and nobody's been pulled out among patients (except in body bags).

    I told people beforehand, after 0.2 seconds of thoughtful analysis, that keeping patients aboard ship was a perfect way to kill as many people as possible.
    And TPTB figured that out too - only about a week later.
    And ended the experiment for U.S. passengers.

    The only useful takeaway from the MV Death Princess experiment, exactly like the Stanford Prison Experiment, is that when you use medieval medical methods, you get medieval levels of medical outcome.


    Oh and the Japanese "expert" who was concerned about their quarantine procedures?
    This was the same soopergenius who went from room to room talking with all passengers with no protective equipment on, himself, wasn't it?
    That's some expertise he brought to that party. One Typhoid Mary honeybee cross-pollinating 3000 flowers, single-handedly.

    The prosecution rests, Your Honor.

    1. It's wrong because you say so!

      You've presented no counter facts or figures, so...

      Objection sustained.

    2. Aesop, 50% of those tested positive were asymptomatic. Pulling people off the ship when symptoms appeared left a large pool of infected people. The linked article goes into this in some depth.

      You were right that the plague ship was a worst case incubation scenario. The data don't seem to show that it was nearly as bad as we feared.
      As to the retirement home example, it doesn't look like observed infection was particularly different for >80. Not sure if this is an apples to apples comparison, as people in nursing homes are generally much less healthy than those who go on a cruise.

      But this just gives us a better set of data for more selective quarantine (retirement homes yes; bars and restaurants no)

    3. Angus,

      There are no figures to pull in, because the Japanese pulled everybody positive off the ship as soon as they became symptomatic. When you scrape all the visible infection off a petri dish as soon as you see it prout, and then at the end you only show a modest infection rate, that does not, ergo, prove you conducted a proper test. It shows you don't get how science works. The model for that is a beauty contest, where you screen out the ugly chicks. When you're doing that with a viral infection, as though it's statistically valid anyways, it's a turd that cannot be polished.

      The point of the Old Pharts home you're apparently missing is two-fold:
      1) The young, healthy people came down with the infection there at a 40-50% rate, not 17%. So clearly it's a helluva lot more infectious than that so-called study shows. In the actual wild, and with no careful plucking of them, because no one knew they were sick. It does demonstrate that for large numbers of the population, asymptomatic infection may be the rule. Nifty for them, but otherwise irrelevant in the grand scheme, unless one is cheerleading for euthanasia.

      2) Debilitated shut-ins didn't go out and get this, it was brought in by staff and/or visitors. If you have come up with a way to care for a million people in SNFs and convo homes in this country with no staff and no visitors that doesn't look like the BL-IV wards at Emory Medical Center, Nebraska, etc., please share.

      I'm just spitballing, but I'm thinking putting them all in a bubble will run a bit expensive, and their nexts-of-kin are liable to get a tad bit upset with just writing them off as "expendable".

      You might could get 49% of the country and the governor of VA to go for it, if you just call them late-late-late-term 320th trimester abortions, but I still think that will cause a bit more ruckus than you want.

      Your thoughts on that?

  2. "I hear the number 0.1% for the annual flu, but I don't know if that compares more directly to my raw numbers or their age-adjusted 1.2%."

    The annual flu death rate is from tested cases. It's from people who got sick and felt bad enough to see a doctor and then were tested and found positive for influenza. It's not going to count the people who got sick, grabbed an over-the-counter symptom treatment and gutted it out.

  3. https://www.cdc.gov/flu/about/burden/preliminary-in-season-estimates.htm
    US flu numbers are statistically adjusted to account for those who do not seek medical attention.

    1. That's good to know and probably means the 0.1% is the total population as opposed to the "Age Adjusted Infection Fatality Rate" that the article used. That makes it more like the 0.2% that comes from 7 deaths out of 3711 patients. That was the total population on the ship that walked around sharing virus for weeks before anyone left the ship.

  4. I'm pretty sure my wife, son and I just had the Kung Flu. We live on the outskirts of Middle Of Nowhere, Midwestia. I thought we were having a mild case of spring allergies after a series of warm days started everything greening up. It wasn't until we all had sore throats for a few hours on day 3 (of 5 total) that we started to suspect something was up.

    No, we haven't been tested. We're not going to waste the hospital's time.

    1. Without the tests, I don't think you can know you had anything in particular, whether this year's Plain Old Flu (POF) or Kung Flu. Or some cold virus. If you live in the Middle of Nowhere, the question is how you'd have come across it. Or anything, really.

      One thing I think that can be gotten out of this study comes from the 83% of people who were exposed and never got the disease. That means it's not jumping off door knobs or desktops and forcing its way into your lungs like so many people seem to think is going on.

    2. Well, testing has only recently been decently wide spread. If the virus does spread as quickly as people say, then it is probably everywhere now. I.e. mass quarantine will have precisely zero effect.

      If this is true then it is very good that mortality looks much lower than previously thought.

    3. I've been keeping a spread sheet of the numbers taken from the John's Hopkins data at 8:00AM daily since 10-MAR-20. Currently the U.S. death rate, as of today, is 1.44% decreasing by the day, down from 2.91% on 10-MAR.

      WW death rate is 4.08% increasing from 3.51% from same date. I suspect that growth is due to people in Europe, especially Italy, dying like flies in a snowstorm is September. I know that China's number are included in the WW totals and that they have and are continuing to lie about the numbers. The numbers are as they are. Nothing can be done about China, other than putting the whole country out of their collective misery, which is probably not going to happen, at least until the U.S. brings the bulk of its manufacturing back on shore and relearns how to make stuff inexpensively here.

      China needs to pay a heavy price for their perfidy, especially in light of their recent threat to cut off antibiotic shipments to us.


    4. Thanks for that interesting comment.

      When I think of handling data like this, I always think that we're not measuring what's going on so much as doing statistical sampling of it. The data is always suspect.

      The US death rate may be going down because of more testing going on that's raising the number of confirmed cases. If about 50% of the people on the boat that tested positive never had symptoms, we've got to figure some percentage of the US population also would test poisitve for the virus but never have symptoms.

      Then there's that whole subject of false positives and false negatives.

  5. Outside of the cruise ship petri dish, and it's results are a useful study, I've been speaking with friends in the UK and Italy who currently have COVID-19. None of them have been tested. In the UK, they stopped testing. In Italy, according to sources in the north and south (unscientific sample), "everyone has it". When this is the case, the fatality numbers are disproportionately high because the numbers of dead are reliable, while the numbers of those infected are not (because they're not all counted).

    I'm healthy, don't smoke, never have, no heart disease, etc. I'm not worried that I'll die if I catch it, but a lot of people who do catch it get quite ill and it tends to last a little longer than the annual flu. So I don't want to be exposed and I'm staying in and eating at home. But I don't fear the reaper with this one. However, if I had compromised lungs and kidneys, or had a compromised immune system, I might worry.

    1. Mark Steyn, sitting in for Rush yesterday, reminded me of something I had seen and forgotten. If I'm recalling correctly (it's before my second cup of coffee, so dangerous to think I recall) in the state where this all blew up in Italy, there was an attempt to show the world they weren't racist by a "Hug a Chinese man" campaign. They had people out in public with signs saying something like, "I'm a Chinese man. Come hug me" So determined to virtue signal that they're not bad people that they threw caution to the wind.

      That's a big difference between self-isolate/self-quarantine and go hug strangers. I've got to think that helped it spread. Add in there's a lot of old smokers with damaged lungs and it's a really deadly combination.

      I don't think you have anything to worry about. First off, you're not over 80 and the death rate curves upward with age. Second, you're not frail. Someone over 70 that's walking around with an oxygen generator had better be self-isolating. That's not you.

  6. I am not convinced that this virus is doing anything more the killing people who are old and compromised, just "pulling demand forward".
    We do have some good tests run, on the Diamond Princess I would be interested in the outcomes for the crew, who would be normally younger and healthy. Also, in the Kirkland nursing home, there were 20 odd firefighters and cops quarantined , some of whom contracted the virus- how did they fare?

    1. on the Diamond Princess I would be interested in the outcomes for the crew, who would be normally younger and healthy It's in those colored bars. The 20-50 demographics had over 90% not even contract the disease, and of the roughly 8% in each age group (except 15% in the 50-59 group) who were positive on a screening test, 10, 20, 30 (of 8%) and 55% (of 15%) showed zero symptoms. If you read that carefully, you'll notice that the younger they were, the more likely they were to have some sort of symptoms and know they had something.

      Also, in the Kirkland nursing home, there were 20 odd firefighters and cops quarantined, some of whom contracted the virus- how did they fare? I have no idea.

      At the risk of sounding snotty, I can't say that any time I've visited people in a nursing home I've been impressed with the staff. I don't think that's where the top echelons of nursing go to work, and I imagine it has been a long time since they had to deal with infectious diseases instead of caring for chronic conditions. In short, the staff screwed up and exposed everybody.

    2. 1) It may have been visitors, not staff, who brought it in.

      2) Those staff members deal with infectious diseases 24/7/365. (How well they do so is an open question.) I know this (and make the parenthetical comment as well) only by direct observation of the situation for my entire working career.

      3) Either way, however it got into the facility, if they were asymptomatic for Kung Flu, this only proves that people doing Standard Precautions (for staff) and Zero Precautions (for every Swinging Richard visitor/family member, like they do) will do nothing to stop a virus that requires Droplet and/or Airborne Precautions (both much higher levels than Standard, and Nothing) to prevent. I could amplify your opinions of staff there or anywhere similar, but it's doubtful they even knew they had anything, and if the Doc doing surveillance (against explicit FedGov orders) for Kung Flu with the Seattle Flu Project hadn't tripped over the cluster there, no one would have ever known this was Covid-19.

      The greater point, as noted above, is that 40-50% of the (younger, healthier) employees at large had Kung Flu symptoms, and were subsequently terminated. Not 17%. So this is evidently a lot more contagious than the cruise ship seems to indicate. And however bad a convo hospital's staff is, the cleanliness and sanitation practiced there is de facto likely to be orders of magnitude better than among the common spaces and hospitality staff on a multi-cultural cruise ship crew, which points to this being far more virulent in an environment that wasn't as finger-banged before and during "study", as was the Death Princess.

      Whatever merit there may be to the "data" (for some value of that word) from this, I think it's clearly overshadowed by being a rampant case of GIGO.

      This is the kind of massaged confirmation-bias nonsense we'd be laughing out of the room if the GloBull Warming fruitloops tried passing it off.

      Just saying.

  7. "Wait - it gets better. Slightly less than half the passengers (48.6% ± 2.0%) who would test positive for the disease did NOT get sick and showed NO symptoms. They never knew they had it."

    This is not necessarily a good thing, as it means that if people only pay attention to their symptoms, half of the infected people will think, "I'm fine," and go about their business spreading infection, until they cross paths with the spouse of a nursing-home care-provider, the child of someone in a nursing-home, or a nurse or a doctor or an EMT or a police officer or...

    And then you have an outbreak in a nursing home, a hospital, on the Fire Department or the Police Force, and some of those people will spread the disease far and wide.

    It is good that only a small percentage of people will die. But the dispersal of the disease will mean that the maximum number of people will die.

    1. The transmittal and symptom patterns show that it is probably already everywhere.

  8. Thanks for drawing attention to this report. There may be a very simple explanation for age-related differences in who shows symptoms. Look at the data:

    Ages 20-29 (mostly healthy crew, probably)
    347 people - 25 symptomatic infection - 3 asymptomatic

    Ages 70-79 (probably all passengers, many with existing medical issues)
    1,015 people - 95 symptomatic infection - 139 asymptomatic

    Simple explanation -- the old passengers mostly already had age-related problems and existing medical conditions. They already had aches, pains, coughs, wheezing. Many of them simply did not notice the additional mild issues from the Media Panic Virus. In contrast, the young & healthy crew members were more likely to notice some mild discomfort.

    Nobel Prize-winner Bragg is reputed to have said: "Scientific progress depends on knowing which half of the data to ignore". It is too easy to ignore some of the data and make oneself an alarmist.

  9. None of it matters. The virus was the convenient scapegoat for the collapsing bubble economy. It's not the virus that's going to kill the most people. The ensuing totalitarian government or social collapse is going to truly suck. Riding out the virus is not that bad for prepared and relatively healthy people. But what happens when we (citizens) say "Hey, the pandemic is obviously winding down. How about y'all back of on these lockdowns and emergency powers declarations."
    What happens then when we ask TPTB four our constitutionally recognized liberties?
    Hint: totalitarianism never backs off willingly...