Hacker News Reader

Comment by blauditore | Hacker News Reader

Testing more thoroughly would also provide us with a better understanding of the virus' properties, like the distribution of symptoms, severities, and mortality rate. I think one of the main reason for general fear and panic has been the alleged high mortality rate, with numbers up to 5% being mentioned. But there's a strong selection bias in those statistics (people with severe cases get tested more often), so the actual number is likely much lower.

View on HN

It's a small sample, but it's a comprehensive one as everyone was tested. On the Diamond Pricess, it only reached 20% of the people on board, and of those 60% had no medical symptoms. Overall 1% of people catching it died.

As a cruise ship the passengers tend to be older, but the crew tend to be younger.

Doesn't matter if you test 1 million people or 10,000 people, if you aren't doing statistically significant tests, it's meaningless to draw conclusions.

Statistics is not the only purpose of testing!

If physical distancing is successful in reducing the reproduction number (R0), you will then want to track, trace and isolate every case to prevent further outbreaks when you relax the rules.

That means a lot of testing.

OP said

"better understanding of the virus' properties, like the distribution of symptoms, severities, and mortality rate"

Random and self-selecting testing doesn't seem to offer much in that. Imagine going to a Nics game and asking everyone there who they support, chances are you'll get a lot of Nics fans.

How does that give you a better indication of mortality rate or severity if you have no idea how many people actually have it, because you only have a statistically insignificant tests.

Sure, if you're in the stage where you want to lock down specific people (like South Korea) then random widespread testing works to increase the number of locked down people.

But if you're in a country where you've locked down everyone already, then testing could well be increasing the risk.

If the population assumes they've tested positive, and acts like they are, then what does testing add?

> If the population assumes they've tested positive, and acts like they are

That is a fair point, and I agree that testing is most important when there are a limited number of cases - a stage which was essentially missed by Western governments but hopefully will have another opportunity.

I don't agree that populations are acting like they have tested positive, though, or can do that en masse.

If you know yourself to be positive, you should ideally be taking stricter measures - like having people leaving food at your doorstep.

Only people who are not infectious should be involved in food and resource distribution.

In order to enact that, you would need to actually test.

The purpose of random testing or testing of more people who have exposure but no symptoms would be to limit household spread by separating people from their households or having the households act differently knowing that they'll be constantly exposed.

If you're testing people without symptoms, you need to be testing large numbers otherwise you'll be ignoring 99% of houses.

Globally there have been fewer than 5 million tests performed since the start of January.

The trouble is, the mortality rate is highly dependent on the degree to which health services are overwhelmed, because the health services are playing such a critical role in keeping people alive for a few weeks until they get better.

To the extent that mortality rate may depend more on transmission rate, health service dynamics and availability of people and equipment than on the disease itself.

This is very different from the mortality rate in a non-pandemic situation.

Mortality rate has also shown to be highly dependent on comorbidity which varies country to country. Presumably a county where residents have easy, affordable access to health care will be healthier on average and have fewer comorbidity concerns that will lead to death from Covid19. Just speculating but perhaps this is one of the things going on in Germany? Germany has a much lower Covid19 mortality rate than other countries.

Reply on news.ycombinator.com