Hacker News Reader

Comment by citilife | Hacker News Reader

It means their body responds to the virus, we aren't necessarily clear on why. Supposedly, this could be because they came in contact with SAR COV-2 OR some other virus which enabled the response.

In this case, 77% of participants were in the medical profession and were evaluated between May and June 2020. I suspect they came in contact with SAR COV-2 and didn't even realize it (body fought it off).

What this study is really suggesting is that your body either (a) has an immune response from previous contact (likely due to similarity with regular corona viruses) or (b) lots of people show no symptoms (we know this) and that they obtain an immune response from it.

In either case, the SAR COV-2 is likely far-less dangerous in the sense that more people have it and there are less deaths per infection OR many of us have a natural immune response (more than previously suspected).

View on HN

Well, we do know how dangerous this virus is. This new data doesn’t tell us anything new because mortality rate is mortality rate and that’s that. What this data might do is help point to what further research to conduct to try to figure out why some people have such strong reactions to COVID while others have few or no symptoms at all.

> mortality rate is mortality rate and that’s that

Not really. Mortality rate is number of deaths / number of infections.

We roughly know the total number of deaths. We have no idea what the total number of infections has been.

For practical purposes, do you want to count these symptom-free infections in that calculation? Knowing this, would you want to do anything differently in how we handle this pandemic? I don’t see how this changes much at all about our present situation.

What are you saying?

We should continue to make decisions based on the pretend mortality rate and ignore the true one?

Of course we "want to know it."

I am not against knowledge in any way. More research will only help. What I am saying is that in the people who get sick, the mortality rate is 1-3%. This data does not tell us whether that generalizes to everyone or not because merely having these antibodies does not guarantee that you actually won’t get deathly ill. It could be interpreted that you have full immunity or it could be interpreted that you have a weak immunity to COVID. More research is necessary.

However the “pretend mortality rate” you refer to is not a thing. It is 1-3% in people who test positive, which correlates pretty strongly with people who have the disease. Think about it this way: the yearly flu in 2018-2019 had a mortality rate of 1.3-48.7% according to the CDC. The denominator in that was 61,000 or so cases. If you found out that 100,000 additional people were exposed to the flu but had no symptoms and by any definition had no flu illness would that change your outlook on whether the flu vaccine or hand washing is worth it?

Or another example: there is surgery which has a mortality rate of 15% due to post surgery infection. Not everybody needs the surgery, only 100 people per year do. Does that mean the rate of mortality here is fake because the denominator should be 7.5 billion? Or should we only count the people who get the disease that necessitates the surgery?

> If you found out that 100,000 additional people were exposed to the flu but had no symptoms and by any definition had no flu illness would that change your outlook on whether the flu vaccine or hand washing is worth it?

Yes, of course my outlook would change. Not on the binary scale--i wouldn't suddenly think it's not worth it to wash hands-- but I would want to know when deciding on measures.

The mortality rate is a specific figure, it's not whatever figure makes your reaction most appropriate.

If some people get the virus and have no symptoms, that's an incredibly important piece of the puzzle when deciding how to proceed.

Using your exaggerated examples, what if 99% of the population didn't suffer any adverse effects? Wouldn't you reevaluate the current response in light of that?

> If some people get the virus and have no symptoms, that's an incredibly important piece of the puzzle when deciding how to proceed.

This study does not tell you that these people had COVID. That's the point: we don't know what this means. All we know, is that we need to study this further.

> The mortality rate is a specific figure, it's not whatever figure makes your reaction most appropriate.

From Wikipedia: "Mortality rate, or death rate, is a measure of the number of deaths (in general, or due to a specific cause) in a particular population, scaled to the size of that population, per unit of time." What is the "particular population" here is in question. I argue that "people who had the SARS-CoV-2 disease" are the relevant population. That is, people who had symptoms or tested positive after an exposure. This study does not change this because it shows no source of the antibodies: if I inject my COVID-19 antibodies into your arm, it does not mean that you had the disease and therefore should not be counted in the mortality rate, does it? How those antibodies got there is a question yet to be answered but the possible answers range from "those people got a mild case and didn't notice" to "those people developed low level antibodies due to incidental contact with COVID-19" without getting the the illness. Their immune system is not trained well enough to fight off a higher viral load."

> Using your exaggerated examples, what if 99% of the population didn't suffer any adverse effects? Wouldn't you reevaluate the current response in light of that?

Yes and no. First, this study does not tell me anything regarding suffering adverse effects. But if a future study did, I think the 3 million people that have died so far would have wanted us to take more precautions, not less, just because we, the living, won the antibody lottery. At the end of the day this is a highly infectious + somewhat deadly disease that adds up to one deadly pandemic. If your point is that only another few million will die if we let it run rampant vs all of humanity, I don't disagree with you but urge you to examine what it means to let millions of people die in this case.

> For practical purposes, do you want to count these symptom-free infections in that calculation?

well it possibly changes the calculus of how many people could potentially be a carrier of the virus. if that many people had the disease enough to have some antibodies even with all the extreme measures put into place it is possible this virus is even more contagious than we thought it could be. i don't know and i feel like we won't know because the study of transmission is woefully hard to do. going into this we didn't even know specifically how the flu was transmitted other than hand waving ways(maybe on the surface, maybe aerosols, maybe particles). our research has improved but it's still not great.

Agreed. This tells us that we need to do 100 more studies and not much else. It doesn’t conclude that the people with these antibodies are actually any better off and not getting sick or not transmitting the disease. As you point out they might be worse off. Everyone here is getting very excited about this study because I suspect people think this means that everyone already had COVID so there is less reason to worry. This study says nothing of the sort.

Mortality rate is the ratio of deaths to individuals in a specific population.

So there's the mortality rate for the overall population (what GP is referencing) vs mortality rate among people who've had COVID (what I believe you're referencing). Both are valid and useful in different circumstances: "How likely am I to die if I get COVID" changes as we discover a glut of asymptomatic cases, but "What % of the population will die of this disease?" stays the same

Ref: https://www.merriam-webster.com/dictionary/mortality%20rate

Most likely everyone in this thread wanted to talk about case fatality rate.

https://www.merriam-webster.com/dictionary/case%20fatality%2...

> "What % of the population will die of this disease?" stays the same

That doesn’t sound right — you can’t know the population death rate definitely without knowing the infection rate;

If no one gets the disease in your sample population of 100, your death rate says 0. If 1 person is infected and 1 dies, your population death rate says 1%. If 80 are infected, and 10 dies, your population death says 10%.

The “correct” number should be stable, but the population death rate should be changing as the disease spreads, approaching the correct number (but this is also true of liklihood of dying to COVID)

Your population death rate would only be static if your sample population is 100% infected, which is the same as “How likely am I to die if I get COVID”

You are getting a bit carried away. You aren’t wrong, the number of people who already got sick and either recovered or died affects future infection rates. So do other things. But I’ll use the example I used in another comment here: if you have 7.5 billion people in the world, and a rare disease that requires surgery where 100 people per year get sick, and all 100 undergo the surgery, and 15 of them die due to post surgery infections, is the mortality rate of the surgery 15/100 or 15/7.5 billion? If the latter, why?

As another example: lots of people have the tuberculosis infection which lies dormant in their lungs. TB infection is distinct from TB disease which is where you get symptoms. The former is only dangerous in that it can develop into the latter. The latter can kill you. How should you look at these numbers? Is it worthwhile to drop the number of people with TB disease and just divide deaths by TB infections or is it more informative to derive mortality as deaths / TB diseases while also using TB infections and the conversion rate from those to TB disease to calculate risk and transmissibility? I argue the latter is much more relevant to decision making.

Keep in mind that the death rate still has to exist in a given time frame so only can only be extrapolated backwards and has to be recalculated annually.

Strikes me as similar to those who predict x% returns on their s&p index funds based on past stock market performance.

The future is always unknown.

I don’t think you can directly compare disease to the stock market. The stock market is a level 2 chaos system: predictions about it are direct inputs into the system affecting outcomes. Disease is generally much closer to level 1: it is random but unaffected by predictions. COVID is possibly mildly affected by predictions in that people might make travel plans based on predictions but I don’t think the majority of people are acting in ways that seriously change their transmission risk based on predictions.

Mortality rate may be unknown. But we know it's killed at least a half million in the US alone, 3 million worldwide. There are also deaths not associated with Covid 19 but caused by it (they didn't get a test before or after death). That's still the highest rate of death we've seen since the last pandemic (spanish flu) caused in a single season by a virus. So yes it is dangerous.

Thanks for putting that so succinctly.

We can estimate the total number of infections through seroprevalence studies. The latest research shows a global infection fatality rate of about 0.15%, although there has been a wide variance between countries.

https://onlinelibrary.wiley.com/doi/10.1111/eci.13554

We know the mortality rate to a degree. Mortality rate is number of deaths over the number of instances. If you have limits on the accuracy of the denominator you have limits on the accuracy of mortality rate. Sero studies like this one suggest to how many infections of the virus we have been missing.

This study doesn’t suggest any number of infections. We don’t know if by any definition people who have these antibodies had a COVID infection or not.

We know deaths / symptomatic infections really well due to the large numbers involved.

Wouldn't mortality rate drop off a cliff once we have herd immunity?

The current answer to me seems to be that we do not know one way or the other. The hope is that it will but we are just now starting to get some data on how effective vaccines are at preventing transmission (Pfizer seems effective, others have no conclusive data AFAIK), we don’t know what herd immunity actually is (Fauci estimates between 70 and 95% vaccinated; 70% is possible without vaccinating children, 95% is not. In the US you have enough anti-vaxxers to prevent herd immunity from ever happening even at 70%), we don’t know how long the vaccine is good for (current estimate is at least 6 months but no clue if it is more), and we don’t know what other mutations will show up. The peachy scenario is that we vaccinate everyone super quickly while keeping social distancing and masks to not have any new variants develop and herd immunity can be achieved with very low percentages. The other scenario is that we take our time, dozens of vaccine resistant variants pop up in the meantime and it never goes away because a virus can mutate and spread more quickly than we can play whack a mole.

The conclusion the study authors came to is different than yours:

It is unclear whether this antibody reactivity may confer clinical benefits, for instance, modulating the severity of a SARS-CoV-2 infection.

Reply on news.ycombinator.com