4 ms·
The Santa Clara study has gotten a fair bit of press and attention but it's not clear whether the results will hold up. The antibody test is new and they're ass
by skybrian 7y ago
The Santa Clara study has gotten a fair bit of press and attention but it's not clear whether the results will hold up. The antibody test is new and they're assuming very few false positives, and it's a non-random study (recruitment via Facebook).
Here's some more criticism: https://medium.com/@balajis/peer-review-of-covid-19-antibody-seroprevalence-in-santa-clara-county-california-1f6382258c25 https://medium.com/@balajis/peer-review-of-covid-19-antibody...
- sergiotapia 7y ago1. That's just one persons opinion. Don't treat it as fact. 2. Read the comments in that article for some counter-points. The point is, this can't go on and thank god we're opening up everything again.
- adventured 7y agoIt's why we should be doing massively increased blood testing and test manufacturing to resolve that question. For the last few months it has been the single most important question and they're putting very few resources behind trying to answer it. We can't test blood properly? Of course we could. They are choosing not to, is the actual context. It's a large mistake. How many infections have you already had? How widespread is it really? No answer, no attempt to answer at scale. At its best, it's immense incompetence in action by the authorities in charge of dealing with this mess.
- snowwrestler 7y ago> It's why we should be doing massively increased blood testing and test manufacturing to resolve that question. Yes, literally everybody agrees with this. > For the last few months it has been the single most important question and they're putting very few resources behind trying to answer it. This is not true, every available resource is being deployed to expand testing capability. Again, literally everyone wants to. > We can't test blood properly? Of course we could. They are choosing not to, is the actual context. It's a large mistake. No one is choosing against testing today. The large mistake was 3 months ago when the federal government should have started activating all of this. > How many infections have you already had? How widespread is it really? No answer, no attempt to answer at scale. At its best, it's immense incompetence in action by the authorities in charge of dealing with this mess. Totally agree on the incompetence, but again, the critical moment passed months ago. Everyone is desperately trying to catch up now.
- hcknwscommenter 7y agoYou seem to conflate antibody testing with viral rna testing. The former is somewhat difficult because you can't make the reagents necessary overnight, and the reagents are unique for the virus. Essentially, you have to clone the gene for an immunogenic portion of the virus (the spike protein most likely), express it in a recombinant host (which may require extensive optimization), purify it, and then validate that it properly detects blood of people who were confirmed positive for viral RNA and then recovered. That all takes time, and then you have to scale up manufacturing. In contrast, the the viral rna (RT_PCR) test takes about a day to design and the reagents are synthesized overnight and then you can run the test in the morning of the second day. Scaling that up to 1 million viral RNA tests a day should add at most a month, but our country already runs this number of tests a day if you count every academic, research, hospital, and CLIA lab, so the manufacturing capacity already exists.
- SpicyLemonZest 7y agoIt's important to acknowledge, though, that most of the criticisms in this article apply even more strongly to the official reporting model. (For example, the study might have accidentally skewed towards people with symptoms, while the official numbers are intentionally and maximally skewed towards them.) So the criticisms don't argue that the real number is closer to the 1,870 on the county's website; they argue that we really have no idea how many people have been infected.
- skybrian 7y agoYes, we have little idea about non-symptomatic carriers. However, combining a prescreen with a test does significantly reduce the chance of false positives.