Hacker News Reader

Comment by throwawayswede | Hacker News Reader

Forgive me but isn't this a similar form of "bad assumptions" and so called "4chan level analysis" used to count Covid cases? You can look at the definition of a covid case from the European Centre for Disease Prevention and Control here https://www.ecdc.europa.eu/en/covid-19/surveillance/case-def....

View on HN

Sometimes it's better to have a consistent yardstick than an accurate one. Noise vs bias. This definition has the significant merit of having a lot lower noise than any clinical definition of a Covid death, where you'd have wildly different estimations from doctor to doctor, let alone from country to country.

It does have bias, true, but you can deal with bias. You do pinpoint studies to measure it and use them to correct bigger trends. You compare with the overall mortality change. And once you have an estimation of the bias, you can apply it back to the overall mortality numbers and get a much better result.

Aehm, can you explain what your critique is? How is the definition of a covid case flawed in your opinion?

CT > 40 then vs CT 28 now. I can't help but be reminded of "yellowcake."

One of the main criticisms in counting covid deaths is that any person who died and tested positive for covid was counted as a covid death, regardless of whether covid was the main cause of death.

This isn't the case for any country that I know of. Nearly every country uses the WHO standards for death certificates and related statistics, which can include multiple contributing factors toward death but only if they actually contributed.

So if someone is a heart patient, but they got COVID and their condition suddenly worsened, the death certificate might state the cause of death as heart failure from the existing condition with COVID as a contributing factor.

People who had COVID but died in completely unrelated ways, such as from car accidents, are not counted in COVID death statistics and don't have it listed as a contributory cause in their death certificates.

You're right, it is the same flawed assumption. Both approaches lead to problematic conclusions, both should face scrutiny.

Unfortunately, we mostly see one-sided criticism.

Edit: Thank you for downvoting without commenting!

Reply on news.ycombinator.com