7 ms·

I'm always wary of any strategy that is fine with having less information. Of course cost of information is never zero but in this case the actual cost of tests

I'm always wary of any strategy that is fine with having less information. Of course cost of information is never zero but in this case the actual cost of tests is relatively low.

mynegation 7y ago
The cost as in “COGS + markup” may be low but the opportunity cost (cost of not doing a test when it is more important) might be higher when the supply of tests and lab time is limited. While it is absolutely important to ramp up both, it is a different story.
- groby_b 7y ago
  ^^^^^^^ This, right here. The entire world is struggling to get enough testing supplies. Testing somebody who's experiencing mild symptoms takes away from testing cases that are much more important - because part of what is happening right now is that everybody with a light fever is all "I GOT THE COVIDS" and wants a test. Fact of the matter: We don't have enough tests, worldwide. We can't produce them fast enough. We need to triage.
  - woodandsteel 7y ago
    Yes, but it is also true that the number of available tests is increasing rapidly, and also is partly in response to government demand. If the the available number increased to the level where everyone with symptoms could be tested, would you favor doing so? and do you think that governments should push til enough tests are available to do that?
    - groby_b 7y ago
      Late reply, but yes to the former, with the caveat that it's contingent on sensitivity and specificity. (I.e. false negatives/positives) Probably yes to the latter, unless it takes away resources we need more urgently elsewhere. But really, what we need to do is depress the count of infected people to a point where it's even feasible to test everybody with symptoms. Even if we rolled out testing to a point where we can test 1M/day (I think it's 200K/day right now?), it'd take a year to test all of the US. If we're in flu-prevalence territory, that's only 19M, and so suddenly much more doable. And at that level, we need the ability to both test anybody with symptoms and to conduct random samplings of asymptomatic people, because without that, we couldn't lift shelter-in-place until we have a vaccine. That's... a tad long. So, more tests, plus suppression doing its job. And it'd be lovely if our government would push to get anything done, really. They've wasted 8 weeks by now, at least. Coordinated response is allowed to kick in any day, by my book.
surfsvammel 7y ago
Cost isn’t zero though. Have a lot of people with mild symptoms come in puts strain on health care, and increases the risk that important health care workers get the affliction. If the test could be done yourself, at home, then I would agree with you.

I'm always wary of any strategy that is fine with having less information. Of course cost of information is never zero but in this case the actual cost of tests is relatively low.

mynegation 7y ago

The cost as in “COGS + markup” may be low but the opportunity cost (cost of not doing a test when it is more important) might be higher when the supply of tests and lab time is limited. While it is absolutely important to ramp up both, it is a different story.

groby_b 7y ago
^^^^^^^ This, right here. The entire world is struggling to get enough testing supplies. Testing somebody who's experiencing mild symptoms takes away from testing cases that are much more important - because part of what is happening right now is that everybody with a light fever is all "I GOT THE COVIDS" and wants a test. Fact of the matter: We don't have enough tests, worldwide. We can't produce them fast enough. We need to triage.
- woodandsteel 7y ago
  Yes, but it is also true that the number of available tests is increasing rapidly, and also is partly in response to government demand. If the the available number increased to the level where everyone with symptoms could be tested, would you favor doing so? and do you think that governments should push til enough tests are available to do that?
  - groby_b 7y ago
    Late reply, but yes to the former, with the caveat that it's contingent on sensitivity and specificity. (I.e. false negatives/positives) Probably yes to the latter, unless it takes away resources we need more urgently elsewhere. But really, what we need to do is depress the count of infected people to a point where it's even feasible to test everybody with symptoms. Even if we rolled out testing to a point where we can test 1M/day (I think it's 200K/day right now?), it'd take a year to test all of the US. If we're in flu-prevalence territory, that's only 19M, and so suddenly much more doable. And at that level, we need the ability to both test anybody with symptoms and to conduct random samplings of asymptomatic people, because without that, we couldn't lift shelter-in-place until we have a vaccine. That's... a tad long. So, more tests, plus suppression doing its job. And it'd be lovely if our government would push to get anything done, really. They've wasted 8 weeks by now, at least. Coordinated response is allowed to kick in any day, by my book.

surfsvammel 7y ago

Cost isn’t zero though. Have a lot of people with mild symptoms come in puts strain on health care, and increases the risk that important health care workers get the affliction. If the test could be done yourself, at home, then I would agree with you.