6 ms·

I don't understand why people are still taking Sunetra Gupta seriously, Oxford or not. She predicted that half the UK population may have already had COVID-19,

I don't understand why people are still taking Sunetra Gupta seriously, Oxford or not. She predicted that half the UK population may have already had COVID-19, back in March. Even as late as May, the true figure was less than 7%. I wrote about this in this post, which was discussed here on HN at the time: https://rssdss.design.blog/2020/03/31/all-models-are-wrong-but-some-are-completely-wrong/ https://rssdss.design.blog/2020/03/31/all-models-are-wrong-b...

ketamine__ 6y ago
People in public health have made a lot of bad calls and predictions as well. That's the nature of the virus. We just don't know.
nradov 6y ago
We don't actually know the true figure of how many people have been infected with SARS-CoV-2. The antibody assays used in most population studies miss many of the mild cases. This is naturally difficult to quantify. https://www.bmj.com/content/370/bmj.m3364 https://www.bmj.com/content/370/bmj.m3364
- IshKebab 6y ago
  You can infer it from models though. If you know the R number, fatality rate, and death rates you can fit it all to a model and estimate how many people have been infected. Not as simple as that, but my point is that testing people isn't the only method, so imperfect tests don't mean that we definitely don't know the answer. This is the best one I've seen for the UK - they just updated it today and estimate that 9-15% of people in England have been infected. https://joshuablake.github.io/public-RTM-reports/iframe.html https://joshuablake.github.io/public-RTM-reports/iframe.html
  - nradov 6y ago
    No you can't infer the infection rate from models. That's circular logic. You have to accurately measure the infection rate first before you can calculate the fatality rate. All of the existing tests have a fairly high false negative rate. So we can only calculate the lower bound on infections; the upper bound remains unknown.

ketamine__ 6y ago

People in public health have made a lot of bad calls and predictions as well. That's the nature of the virus. We just don't know.

nradov 6y ago

We don't actually know the true figure of how many people have been infected with SARS-CoV-2. The antibody assays used in most population studies miss many of the mild cases. This is naturally difficult to quantify. https://www.bmj.com/content/370/bmj.m3364 https://www.bmj.com/content/370/bmj.m3364

IshKebab 6y ago
You can infer it from models though. If you know the R number, fatality rate, and death rates you can fit it all to a model and estimate how many people have been infected. Not as simple as that, but my point is that testing people isn't the only method, so imperfect tests don't mean that we definitely don't know the answer. This is the best one I've seen for the UK - they just updated it today and estimate that 9-15% of people in England have been infected. https://joshuablake.github.io/public-RTM-reports/iframe.html https://joshuablake.github.io/public-RTM-reports/iframe.html
- nradov 6y ago
  No you can't infer the infection rate from models. That's circular logic. You have to accurately measure the infection rate first before you can calculate the fatality rate. All of the existing tests have a fairly high false negative rate. So we can only calculate the lower bound on infections; the upper bound remains unknown.