3 ms·

The article is interesting but it seems to be just a mathematical model rather than a biological study actually finding and studying people who had undocumented

The article is interesting but it seems to be just a mathematical model rather than a biological study actually finding and studying people who had undocumented infections. I know they have found people like that but I'd still be leery about going from data to a pronouncement since I assume the data just isn't extensive or reliable.

gwbas1c 7y ago
But you need actual testing in order to get real data
- Taek 7y ago
  South Korea has done a good job of getting enough tests done to have meaningful data
  - iso947 7y ago
    Self selecting tests. Need the types of control data that polling companies use.
    - akiselev 7y ago
      South Korea has been doing extensive contact tracing and quarantining from the start. They're as close to a statistical sample as we're going to get until large scale random antibody tests are done. At best the US could probably fit it into the census process.
      - spookthesunset 7y ago
        Until you get a good bit of random antibody tests, it is all speculation. Nobody really knows how widespread this thing is, what its actual death rate is or anything else. I personally wouldn't be surprised to learn the number of people who had it or have it now is hundreds of times higher than the "confirmed cases". I would also speculate that this virus has been in wide circulation for months now. But my guess is as good as anybody else. Until we get actual, statistically valid high quality data.... all we have is speculation. I sure hope we are making the right call shutting things down as much as we are. People have to remember that there are serious physical health consequences to the actions being taken. Suicide rates will climb, alcoholism & addiction will jump, crime will go up, etc. A few weeks of this kind of economic pause is one thing but at some point in the very near future people are going to want to know what the end game is.
        drivebycomment 7y ago
        As of today, Korea reports 274K tested, with 3.2% infected out of that. They also show strong signs of containment. Those two datapoints indicate that they didn't miss many infected cases.
        spookthesunset 7y ago
        Does that testing include people who had it before and just have antibodies? Or just active cases?
        dboreham 7y ago
        There is no deployed serological test at present.
- m0zg 7y ago
  People overestimate the accuracy of this test as well. There's a doctor in Seattle who tested negative, and a few days later ended up in the ICU. He's currently in the process of dying. Just because the test says "no virus" (or vice versa) does not mean you're free and clear. Especially early in the disease cycle RT-PCR is known to produce false negatives sometimes.
btilly 7y ago
The Seattle flu study is an example of what you want. As https://komonews.com/news/coronavirus/seattle-flu-study-allegedly-tested-samples-for-covid-19-against-federal-state-guidelines https://komonews.com/news/coronavirus/seattle-flu-study-alle... shows, the USA was refusing to test Americans in Seattle while COVID-19 was actually spreading. When the Seattle flu study tested they found that it had already been locally spreading for weeks. And in fact most people who think that they have COVID-19 in the USA are unable to verify whether or not they have it.

gwbas1c 7y ago

But you need actual testing in order to get real data

Taek 7y ago
South Korea has done a good job of getting enough tests done to have meaningful data
- iso947 7y ago
  Self selecting tests. Need the types of control data that polling companies use.
  - akiselev 7y ago
    South Korea has been doing extensive contact tracing and quarantining from the start. They're as close to a statistical sample as we're going to get until large scale random antibody tests are done. At best the US could probably fit it into the census process.
    - spookthesunset 7y ago
      Until you get a good bit of random antibody tests, it is all speculation. Nobody really knows how widespread this thing is, what its actual death rate is or anything else. I personally wouldn't be surprised to learn the number of people who had it or have it now is hundreds of times higher than the "confirmed cases". I would also speculate that this virus has been in wide circulation for months now. But my guess is as good as anybody else. Until we get actual, statistically valid high quality data.... all we have is speculation. I sure hope we are making the right call shutting things down as much as we are. People have to remember that there are serious physical health consequences to the actions being taken. Suicide rates will climb, alcoholism & addiction will jump, crime will go up, etc. A few weeks of this kind of economic pause is one thing but at some point in the very near future people are going to want to know what the end game is.
      - drivebycomment 7y ago
        As of today, Korea reports 274K tested, with 3.2% infected out of that. They also show strong signs of containment. Those two datapoints indicate that they didn't miss many infected cases.
        spookthesunset 7y ago
        Does that testing include people who had it before and just have antibodies? Or just active cases?
        dboreham 7y ago
        There is no deployed serological test at present.
m0zg 7y ago
People overestimate the accuracy of this test as well. There's a doctor in Seattle who tested negative, and a few days later ended up in the ICU. He's currently in the process of dying. Just because the test says "no virus" (or vice versa) does not mean you're free and clear. Especially early in the disease cycle RT-PCR is known to produce false negatives sometimes.

btilly 7y ago

The Seattle flu study is an example of what you want. As https://komonews.com/news/coronavirus/seattle-flu-study-allegedly-tested-samples-for-covid-19-against-federal-state-guidelines https://komonews.com/news/coronavirus/seattle-flu-study-alle... shows, the USA was refusing to test Americans in Seattle while COVID-19 was actually spreading. When the Seattle flu study tested they found that it had already been locally spreading for weeks. And in fact most people who think that they have COVID-19 in the USA are unable to verify whether or not they have it.