Comment by estaseuropano | Hacker News Reader

Comment by estaseuropano | Hacker News Reader

Parent Full thread

There is a lot to unpack but I doubt you are as right as you think you are. For instance in new variant outbreaks there ate usually harsher responses than for known variant outbreaks, so it is very difficult to identity the real infectivity.

But more importantly it's well established that diseases that those diseases spread most widely that don't kill rapidly (or that more generally don't have severe symptoms) and that have a longer period between the start of infectivity (= shortly after infection) and the onset of identifiable symptoms. So you would indeed anticipate that any disease that runs sufficiently long in the population becomes more mild and has a later onset of symptoms as variants with these features outcompete thr others.

In other words: it is expected that covid mutates over time to become more mild on average. But that doesn't mean that everyone survives. Infection rates in many countries are higher than ever, which even if the % of severe cases is lower than with other variants, delta and omicron will still kill and have severe long-term effects on many.

I can’t tell what your point is but nothing here seems to contradict anything I said. It’s just some facts about diseases.

I didn’t say every one was going to survive omicron. I said the spread was unlikely to be materially higher than its predecessor and likely have a lower negative impact.

A disease that spreads twice as fast is unlikely to infect twice as many people (over the long term). A disease that causes fewer adverse effects, let’s say half as often, needs to, by definition, infect twice as many people over the long run in order to cause the same amount of harm.

As a dumb mental model, if omicron is 30% less dangerous, then omicron needs to infect 42% more people to be as harmful to society. That is very difficult to achieve, especially given our collective immunity achieved so far. I would say there is little chance of this occurring.

Is anyone even seriously claiming that a disease that spreads twice as fast is going to infect twice as many people? As long as the effective R0 of a disease remains greater than 1, it's eventually going to infect the same amount of people — everyone (over the long term). Greater R0 rate just means it spreads faster.

The real concerns about omicron are that: 1) A rapid initial spread will overwhelm the healthcare system and 2) that if its basic R0 rate is sufficiently high, then even after remedial measures are taken, we will still be left with an effective R0 rate above 1 for quite some time.

That's not how it works. Diseases dont infect everyone. Assuming no reinfection, at R0 of 1.3 a disease will only infect around 30% of the population.

If I conveyed the impression that I thought the effective R0 could remain above 1 indefinitely, that is my error. But why assume no reinfection, that's not what we're seeing so far with SARS-CoV-2.

That is not quite what R0 means. R0 is the reproductive rate in a naive population. I believe the variable you are looking for is R(t), which is the effective reproductive rate at any point of time.

Thank you. I've seen the phrase "effective R0" used, for example, here [0], but I can see that you are correct that R(t) is the generally accepted epidemiological term.

[0] https://labblog.uofmhealth.org/rounds/how-scientists-quantif...

this is a mathematical convention, not simply an epidemiological term. the 0th term of any function, like R0, is generally understood as the initial condition constant (at time t=0). R(t) is a function of time and R0 is the value of R at time 0, the initial condition. R0 tells you exactly nothing about the future, it only pins down the starting point (which is important information, but not the be-all-and-end-all of it).

i’ve been mildly amused/dismayed by the constant misplaced focus on R0 over the past 2 years by the media/laypeople, given this essential context.

That’s not correct. While R0 is indeed R at time 0, time is not the important part of that. It’s that at time 0 everyone is presumed susceptible so the growth is going to be as exponential as it’s going to get. R0 is a strong indicator of the limits of how the disease will spread and is very useful for predicting the future, both in terms of the speed of the initial exponential spread, but also the upper bound on spread that the disease is likely to encounter.

no, the epidemiological model in question has two degrees of freedom, and one of those degrees is stipulated by the assumed shape of the curve (exponential, even though that's a poor fit overall, logistic being a better fit). R0 does not constrain the other degree of freedom; it only gives you the initial condition. it requires knowing that other degree of freedom, which you're making implicit assumptions about, that governs what you can say about the rapidness of spread and makes some statements about bounding. this is rudimentary calculus and linear modeling.

Nope. You’re still making the same mistake. That is not how R0 works. You are falsely assuming that the exponential growth period of the virus stays the same forever. The virus does not reach 100%. R0 as a general trait of a disease assumes everyone is susceptible. That is not how diseases work. There are many diseases with R0 > 1 and guess what? Not everyone got them.

I know you’re caveating “effective R0” but frankly that’s stupid. Saying as long it keeps growing it will keep growing purposefully ignores the point I am making which is that it won’t keep growing.

The effective R0 falls on its own. That is the point being made. It will absolutely not reach 100% of the population. Not even close.

> the point I am making which is that it won’t keep growing.

And what I'm saying to you is that nobody believes otherwise. It's all known that diseases are self-limiting, and infection rates follow an s-curve and eventually drop off. This isn't a groundbreaking revelation.

In my experience the majority of people assume that exponential growth means a small increase in spread rate means and exponentially larger number of people will be infected, and that “everyone” will get omicron. I’m not claiming to have a novel epidemiological insight. I’m claiming that most people get this wrong and are panicking about omicron because they don’t get this.

COVID isn't particularly deadly. The selection pressure that can cause a virus to become weaker is mediated by the virus dying off in the corpses it creates. People aren't dropping like flies from COVID, so that selection pressure isn't there, and the virus is free to mutate into more virulent forms. There's no reason to expect a relatively mild, yet still deadly, virus to become weaker, especially if our medical infrastructure keeps people alive that would've otherwise died from infection.

Virus evolution tends to optimize for replication and transmission, and not reduced severity. If a particular mutation causes a virus to replicate better than other variants, but it ends up killing hosts more often, it doesn't really matter what happens to the host afterwards, evoluntionarily, as the virus has already spread it genes more than it did before it mutated.

Covid-19 doesn't kill rapidly, so there is no pressure for it to mutate to become milder.

The milder it gets, the less noticeable it is. As it gets less noticeable, the chance of exposure to new hosts increases. Seems reasonable.

the real question is, can we pivot faster than evolution can...

The milder it gets, the less it is able to spread because disease symptoms are what help spread the disease in the first place. Also, there are selection pressures that select for increased virulence, otherwise all viruses would have evolved to be completely benign billions of years ago.

That makes a lot of sense too! I guess the sweet spot is where you'd find the persistent seasonal flues. Mild enough for people to be reckless, but virulent enough to cause sufficient spread.

I think the effect is the same once the patient is quarantined. If they survive two days or two weeks the virus is unable to spread either way.

There's also no evolutionary pressure to conserve its lethality.

What appears to have happened with Omicron is that it has mutated in a way that increases transmission at the expense of lethality:

https://www.med.hku.hk/en/news/press/20211215-omicron-sars-c...

> There's also no evolutionary pressure to conserve its lethality.

That depends on whether the mutations that cause it to replicate better also make it more virulent. The symptoms that help spread Ebola, for example, are the same symptoms that kill those who become infected with it.

Reply on news.ycombinator.com