Puzzling numbers

The differing regional approaches to testing make it hard to figure out what the "positive" rates mean. In some areas, there's almost random sampling going on, while in others, most people are unlikely to have access to a test unless they have clear symptoms plus a troubling contact or travel history. A few samples included nearly all of a more-or-less captive population, like the souls aboard the Diamond Princess or the U.S.S. Roosevelt. Until today, all the results I'd seen suggested that well under half of the ordinary closed population will test positive, while something close to half of detectable cases were asymptomatic. (Note that "asymptomatic" doesn't tell you anything about whether a case is contagious. Pre-symptomatic or permanently asymptomatic patients may be very contagious, barely contagious, or variably contagious depending on the patient, the severity of the case, or the days since exposure, or all three.  Not every asymptomatic patient is a Typhoid Mary.)

HotAir has a piece today that reports anomalous results: the infection rates in prisons are sky-high, nearly 90%, and the percentage of asymptomatic positive testers is even higher. The only similar result I'd heard rumors of so far is the puzzling lack of severe cases among the U.S. homeless, and in the entire population of Bali. In the former case, speculation included the possibility that life outdoors was protective, while in the latter case people bandied about the notion that the Balinese lifestyle confers special advantages for immune systems. Neither explanation leaps out as likely for the prison population.  I suppose it's possible that both the homeless and the prison population have led such rough-and-tumble lives that they've been exposed to everything under the sun and have robust immune systems.  Maybe they're poised to take over the world.

8 comments:

Christopher B said...

This falls into the homeless category but provides stats for what has usually been observational.

https://www.cnn.com/2020/04/17/us/boston-homeless-coronavirus-outbreak/index.html

When an outbreak of coronavirus in a Boston homeless shelter prompted officials to do more testing, the results caught them off guard. Of the 146 people who tested positive, all of them were considered asymptomatic.

"These are larger numbers than we ever anticipated," said Dr. Jim O'Connell, president of the Boston Health Care for the Homeless Program. "Asymptomatic spread is something we've underestimated overall, and it's going to make a big difference."

Gtingo said...

Related to the topic are some graphs from Noo Yawk.Coronavirus Statistics: Tracking The Epidemic In New York. Note the declining trend for New Cases and New Deaths- especially for 7-day averages.

David Foster said...

I don't understand why true random sampling isn't being done. It takes a remarkably small sample size to estimate the % occurrence of any phenomenon in a given population, regardless of how large that population size is. For getting an estimate which is accurate within plus or minus 5%, with a confidence level of 95%, all you need is a random sample of about 400 people. If you want a 99% confidence level, sample 800 people.

Larry said...

The thing that I have not heard discussed much is the seeming poor sensitivity of the test. Until the novel Coronavirus came along, I’d never heard of a PCR test having a 60-70% sensitivity. It does seem to be specific; the asymptomatic people who are testing positive are identified when the contacts are traced, and do actually have the disease. The low sensitivity though, is why some of our patients have to be tested 2-3 times. I haven’t trusted much of the Chinese data, but they mention repeatedly that CT scanning was much more accurate than the test for identifying people who have Covid, and I have found that to be true as well.

ymarsakar said...

How do they know if this is Covid 19 vs Covid 20 or Covid 1-18?

Even if they name it based on the year, that just means there are other versions. Did Gates forcibly upgrade them all to Covid 20? I don't think so.

It shouldn't be that easy for a general test to determine which strand of Covid it is.

So it just means that people who had been infected in the past, already had Covid.

ymarsakar said...

Also, they aren't poised to take over the world. Their elemental cold/heat conditioning is in no way equal to or superior to the Iceman's, my own, or Shaolin methods.

The Russians/Mongolians far beat them out.

As for prisons, that depends on how much time they spend in a yard vs isolation. It is also affected by electromagnetic interference and pollution.

When prisoners are able to bulk up and use weight lifting... they are being fed way too much.

ymarsakar said...

the infection rates in prisons are sky-high, nearly 90%, and the percentage of asymptomatic positive testers is even higher.

One of the exotic diseases that went through the US Mexican border, was most likely Covid. This happened during the Hussein years.

That means that Mexican populations, mixing with Asians or others in prisons, would likely have caused a "social network" spread.

Based on your social level, your associates and networks will be composed of those at a similar level. Prisoners are not necessarily in the same social network as the homeless, but they are in the same social level.

One should then check the deaths due to flu/pneumonia from 2012 onwards. But that takes work.

ymarsakar said...

Also, herd immunity mostly only applies to animals or humans that are soul evolution wise, about as mature as very intelligent pets.

The Homeless cannot easily be categorized as either or. Their reasons are too numerous. Prisons however... are full of less mature souls.

If a herd or 100th monkey effect applies to anyone, it would apply to that isolated population.

If the Wu WHO/CDC disease spread from China via air planes, then I don't see where the zero vector for prisoners started out given they aren't supposed to be flying in airplanes to begin with.