11 ms·

I’ve been tanking some downvotes for the past few days for saying this. Disease spread is only exponential for a brief initial period. It is not a good mental o

I’ve been tanking some downvotes for the past few days for saying this. Disease spread is only exponential for a brief initial period. It is not a good mental or even general model for how disease spreads. We have two years of observational covid data that shows this to be true. Spread slows down rapidly long before reaching 100%. People hear “5x as infectious” and reason that due to the nature of exponential models, that much more than 5x people will be infected. This is extremely incorrect. In truth, far fewer than 5x people will be infected over the long term. And again, no, this is not because it’s hitting the upper limits of 100% of the population or anywhere near that. I won’t be so bold as to say it’s probable, but given this is not a novel virus, it’s entirely believable to say that omicron could go on to infect fewer people than delta due to the past two years of vaccination and immunity and die off. Presuming data about lower severity holds, it would be surprising to me if hospitalization or deaths aren’t noticeably lower than delta; which, in turn, was noticeably lower than the original.

jonnycomputer 5y ago
Won't downvote you, but I think that this is mistaken. I think that everyone will get omicron eventually.
- xyzzy21 5y ago
  And it will generate herd immunity...
  - jonnycomputer 5y ago
    Maybe. There is a lot of reinfection/breakthroughs going on from alpha/delta to omicron. Does conferred immunity to omicron protect from alpha/delta? I don't know enough to say.
- bufferoverflow 5y ago
  You're mistaken. There's no disease on this planet that infects 100% of people, even when constantly exposed. Our immune systems are quite amazing.
  - jonnycomputer 5y ago
    Who hasn't had a cold?
    - oneoff786 5y ago
      “A cold” is not a single virus.
    - bufferoverflow 5y ago
      Many people.
snet0 5y ago
On the subject of oversimplified exponential models, we don't even need covid data to know this to be true. The fact that there is a duration of an individual case being contagious is enough to make that obviously incorrect.
- svara 5y ago
  That's not right, you can build a maximally simplistic SIR model with a finite infectious period. That will initially display approximately exponential growth, and slow down as acquired immunity increases.
donquichotte 5y ago
Thank you. It is mind boggling to me how politicians and scientists (well, mostly MDs) keep talking about exponential growth and make horrible predictions, without realizing that exponential growth is never sustainable - a logistic growth is what often happens in reality if resources are limited.
- Talanes 5y ago
  Asking politicians to not over-estimate exponential growth is like asking water not to be wet.
  - lb0 5y ago
    Not understanding what you are both getting at ... the situation has maybe changed now. But before, here at least, it was repeatedly underestimating the in the aftermath clearly visible exponential growth, exactly as the r factor is calculated? Which scientist/MD doesn't understand there is a saturation, in the extreme even schoolers understand you cannot have more than 100% infected. Yes, at some point it is becoming logistic, but the previous waves were not about getting anywhere there, but going from 0.1% to 1% or more, and there it was truly exponential growth and overwhelming the health care system, what the start of that logistic curve is about..?!
  - snth 5y ago
    Isn't the usual complaint that people (politicians included) _under_estimate exponential growth?
    - hsbauauvhabzb 5y ago
      Depends where you are, some parts of australia have certainly been fairly conservative. Opinions aside, they certainly estimated on the upper bounds of severity much more than the US. I personally am grateful that the aus gov did this, saving countless lives in the process.
- II2II 5y ago
  You're right, but there are a number of factors at play here. First and foremost, we can only make predictions based upon the data we have. The data we have is mostly based upon people going for voluntary testing. Given the variability in symptoms that motivate testing, varying levels of awareness to potential exposure, varying willingness to get tested, and the availability of testing, the number of known exposures will make it look like we're in an exponential growth phase even when things are tapering off. It is hard to blame public health organizations for accepting this data at face value since the potential consequences of not doing so are extremely bleak. The other consideration is that people will modify their behaviour based upon perceived risk, whether or not restrictions are imposed. People will tend to comply with restrictions, go about life as normal, follow some sort of middle path, or impose tighter restrictions upon themselves. It takes a truly special person to put themselves into a position of greater risk. Even then, there is a good chance that they are doing little more than translating one high risk circumstance into another (more or less balancing out the growth). We are facing a situation where scientists can make predictions based upon what is known, e.g. the outcome of restrictions, or making predictions based upon anticipated behaviour. Given there is not enough data to model anticipated behaviour, the natural response is to rely upon models that use expected behaviour (e.g. compliance with restrictions or no response). Since the average actual response will prefer over compliance, more bleak predictions are made. Finally, everyone has a different understanding of life now as compared to life before the pandemic. I remember when the coronavirus first caught my attention: it was when major U.S. universities started shutting down. I remember when I first started taking it seriously: it was when my provincial government issued a shelter in place style order. Since that time, I have paid attention to what is happening and thinking about how I should respond. Sometimes it is in an acute manner. Sometimes is in a cursory manner. Either way, I am more likely to notice and respond preemptively to outbreaks. I suspect that many, if not most, people are the same even if their only actions are stocking up to prepare for the worse. Again, this will affect outcomes.
stefan_ 5y ago
It's a bit silly to go "it's less severe" based on death statistics when after two years of Corona waves, there are plain less individuals around for it to kill.
- oneoff786 5y ago
  Death of the vulnerable is a thing, which I referenced, but it’s a relatively minor factor compared to vaccinations and immunity build up from past infections. Most people who would die from covid if they caught it did not catch it. I think it’s a bit silly to fixate on the base stats of the virus rather than it’s actually efficacy against the human population.
  - midasuni 5y ago
    The problem is by the time you see the effect (hospitilations) it’s too late to do anything about it. It’s like driving with a 20 second lag in your vision.
- dnautics 5y ago
  no viral infection happens in a vaccuum. Each year's seasonal and nonseasonal flu comes on the backs of years of "less people" for it to kill, and people who have gained immune responses to any of its closest relatives, and people who have been immunized, and even humans who have been culled by similar viruses milennia ago during migratory bottlenecks.
orblivion 5y ago
If you're right, that would be disappointing, because it means there's less immunity gained than I was hoping, for the (relatively small) amount of hospitalization we've seen. And I (uneducated on the matter) would expect that if it's not really getting around that much, it would not displace delta. But I've heard some claims that it already has started to displace Delta, though not from a source I'd feel confident citing. But just looking at it, if you believe the CDC estimate of 73% Omicron the other day, Delta must have dropped a lot despite it being winter. Appears like displacement happening.
- adzm 5y ago
  The CDC estimate was a projection, not an actual recording of data.
renewiltord 5y ago
Personally, I'm just waiting for the day when the 2^64th death happens.
specialp 5y ago
A place where it hit exponential levels with really bad results is here in NY at the beginning. We had no precautions, testing, or vaccines in place yet, and hospitals were overwhelmed. It was honestly a scary time. That was where we started "flattening the curve". Now with vaccines available if there are surges and they are not literally building hospitals on fields like they were near me that is something we have to just live with. COVID will never hit 0 cases. People will still be hospitalized and occasionally die of COVID as was always the case with winter flu seasons. Now we are at the point where fully vaccinated people and people with immunity from a previous infection have a low rate of death or serious illness. People who are very old or have a compromised immune system should lay low and take precautions now and during every flu season when it spikes. Those who choose not to get vaccinated and die, that is still on them.
- BrazzVuvuzela 5y ago
- Retric 5y ago
  We are currently sitting at at total of over 800,000 COVID deaths in the US and 1,500 COVID deaths per day, which is roughly where it’s been since September 1st. That’s approximately an additional half a million dead people per year. So yes the vaccine has already saved million of lives, but I sure hope that this is not the new normal this winter or we are going to cross 1 million deaths fairly soon.
  - PaulDavisThe1st 5y ago
    Based on excess mortality numbers, it seems fairly likely that we're already over a million deaths that would not have happened without COVID19.
    - refurb 5y ago
      What will be interesting is to see the long term mortality rate. My hypothesis is we’re just accelerating the deaths of people over 70 years old (70%+ of the cases). If that’s true, you’d see a decrease in mortality over the next few years. Covid is basically doing what the flu does every winter (but on a larger scale) - it pushes people close to death over the edge.
ses1984 5y ago
Spread doesn’t have to get close to 100% for things to get really bad.
- oneoff786 5y ago
  Indeed. That’s true. But thankfully that’s not in the table, or anything close to it. I only bring that up because it’s the natural conclusion of what will happen if you misunderstand how the exponential growth works.
rightbyte 5y ago
I wonder how many of these increasing share variants you would expect to observe from just random walk. I.e. how much more infectious does a strain need to be to be measurable.
rich_sasha 5y ago
This is of course true, but I'd imagine the counterargument and concern is that healthcare runs out of capacity well before the growth slows to a sub-exponential run.
joshocar 5y ago
What about the population of the unvaccinated? What percentage should we expect of the unvaccinated to get omicron given how contagious it is?
- oneoff786 5y ago
  It depends on many things and will vary wildly by geography. I’m not suggesting this is a population level trend one can safely piggyback on, unvaccinated. If you are unvaccinated in a community with spread, your odds of getting it are likely comparable, plausibly higher, to previous periods.
estaseuropano 5y ago
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.
- oneoff786 5y ago
  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.
  - zuminator 5y ago
    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.
svara 5y ago
Not sure what you're trying to say. Of course there is a limited pool of people to infect. And various effects can cause infections to spread in multiple waves even far from herd immunity. And yet, you can go from a few rare cases in your country to hospitals overflowing in a couple of weeks. That's because of exponential growth, obviously.
- oneoff786 5y ago
  The point is that 1) exponential growth has consistently ceased before the hospitals overflowed too bad 2) exponential growth confuses readers into thinking that a small boost in spread rate means many more will be infected. This isn’t true because the exponential phase is brief and is a progression towards a ceiling defined by the graph, not an arbitrary period of time. Like what we are seeing in South Africa right now. The big problem is that people think an increase in the spreading rate will cause an exponentially higher number of infected. But it won’t. The model is not appropriate.
  - svara 5y ago
    That's fair, yes. High exponential growth rates make the system much harder to control though, and will easily overshoot the 'set point' (maximal hospital capacity). At a 2-3 day doubling time that overshoot could look pretty drastic.
  - jamesrcole 5y ago
    > 1) exponential growth has consistently ceased before the hospitals overflowed too bad I believe quite a few hospitals around the world overflowed pretty badly.
    - tasogare 5y ago
      Yes, you believe. The reality is that most healthcare systems, at least in developed countries, handled the patients well enough.
      - 5y ago
erichocean 5y ago
> Presuming data about lower severity holds, it would be surprising to me if hospitalization or deaths aren’t noticeably lower than delta; which, in turn, was noticeably lower than the original. Well, that should always be true: the people that are most at risk are already dead from earlier variants, and our treatments are light years ahead of where they were 18 months ago, reducing hospitalizations (and death) for everyone else.
- oneoff786 5y ago
  I’ve seen a lot of people saying this. Again, it’s not wrong that deaths are a factor here. But by far, FAR, the largest factor is vaccinations. We should acknowledge this.
raverbashing 5y ago
Yes, I don't think the SIR models work well in predicting the dynamic of waves The exponential part is easy and that works, but it's more complicated with isolation of the infectious, refractory periods, etc
amelius 5y ago
Can you back this up with some math?
- oneoff786 5y ago
  Look into random graph models. Or exponential random graph models. Mathematical models are difficult because we don’t know the real inputs and fitting a curve in retrospect is easy to get a compelling looking answer which is wrong. You can gain an intuition for it just imagining a random walk on a social network graph though. Just jump from friend to friend randomly on Facebook. Early on it is easy to spread to new people. Later on it’s very difficult to find new people. You get stuck in the same cohorts. More simply just look at past covid outcomes. Or pretty much any epedemiological model. All of them claim only the initial period is exponential. It’s the issue of how to determine the slowdown period that’s tricky and frankly impossible without more data than we have. But assuming that the slowdown state will look similar to previous slowdown states is a good idea.
- School-Cotton 5y ago
  It's an empirical observation, not the output of a mathematical model. The graph of human social contacts is not even close to uniformly random, so it makes sense that simplistic formulas would not work.
- kazinator 5y ago
  https://en.wikipedia.org/wiki/Logistic_function#In_medicine:_modeling_of_a_pandemic https://en.wikipedia.org/wiki/Logistic_function#In_medicine:...
analog31 5y ago
At the start of the pandemic, when downloadable data became available, I started charting it every day. The curves for many countries were uncannily exponential. Of course I knew that those curves had to tip over at some point, but I decided that I would use the exponential growth curve as a baseline and watch for deviations from it. I told my friends: "I'll breathe a huge sigh of relief if this thing turns over before it's infected most of us." In fact it's probably hard to speculate why it turned over, but behavioral measures are probably an important factor.
- oneoff786 5y ago
  It’s not uncanny, it is, literally, how disease spreads. It ceases to be exponential when the people infected people are exposed to are no longer all susceptible because they themselves had been infected. At least in simplest terms. That’s the big assumption in R0.
timdaub 5y ago
You're wrongly assuming that all incidence data of COVID-19 accurately represents its real spread. E.g. it may just be that beyond a point of exponential growth testing ceases to be accurate. The German Robert Koch institute has pointed out many times that there's a hidden incidence number of all actual cases and the official one that represents everyone tested and recorded.
- oneoff786 5y ago
  This assumption isn’t critical to anything I have said
moksly 5y ago
> it would be surprising to me if hospitalization or deaths aren’t noticeably lower Here on vaccinated Denmark this is the truth. But we’re actually approaching last years levels regardless because of just how many people are getting covid this year. It’s more manageable thanks to the very high vaccination status, but we seem to have been too slow with the 3rd hit for a major part of the population. I have two shots myself, and my family is all in covid isolation all tested positive and “looking forward” to spend Christmas with ourselves and not our families. It’s not too bad for any of us, it’s not pleasant either, but the biggest thing for me is how much we’re having to shut down despite the high vaccination percentage. You can’t go to a movie or actual theatre. Bars close at 22:00. Most major Christmas parties (this is a big thing here) have been cancelled. But the biggest impact is on culture business like the theatres, concert houses, Christmas markets, museums and so on. If covid is going to be a recurrent thing every winter, then I think that we’re going to see some drastic changes to those aspects of society.
- trhway 5y ago
  what is the point of those lockdowns when 1. practically everybody is vaccinated and thus gets no symptoms or only light disease (and according to the latest science - superimmunity after that) and 2. the delta's and now omicron's transmissibility is so high among both - vaccinated as well as unvaccinated - that everybody susceptible to it (i.e. 30-70% of population, and anecdotally it seems to me that for covid it is on the lower end, ie. like 1/3rd) will get it (or omicron++) in the near future anyway (and the thing becoming endemic like flu). I mean technically the lockdowns possibly make sense as a way to control the spread until a spread controlling vaccine (the existing mRNA ones aren't such) is widely applied - ie. that seems to be the case in China where initial spread was effectively controlled and they use inactivated covid virus vaccine instead of mRNA - though we don't know for sure because Chinese government info can be very different from reality.
  - oneoff786 5y ago
    Both if your premises are not accurate
  - 5y ago
stefanpie 5y ago
I don't know how well this relates, but the topic of logistic growth (exp. growth with a limit) is common in undergraduate differential equations classes. I guess it can work well for simpler models. https://en.wikipedia.org/wiki/Logistic_function#Modeling_early_COVID-19_cases https://en.wikipedia.org/wiki/Logistic_function#Modeling_ear...
rory 5y ago
What you're saying is supported by the data, but do you (or does anyone) know why infections slow down when they do? I can't think of any intuitive reason for it, and haven't been able to find an explanation why the waves crash, so to speak.
- School-Cotton 5y ago
  My guess is the non-uniform nature of the graph of human contacts, which is probably made up of several densely-connected clusters each with relatively weaker links to other clusters. E.g. let’s say there is an office where everyone works in person and a bar with a group of regular customers. If someone in the office gets Covid, everyone in there has some decent chance of getting it. Similarly, if a bar regular gets Covid, each of the others has a decent chance to get it. But if only one person who works at the office is also a regular at the bar, then for the infection to hop from one cluster to the other, that person needs to get it from the initial outbreak, and they need to continue going into the office during their infectious period, and folks need to catch it from them, none of which is certain. So, my guess is that after enough of these clusters get seeded to start a wave, “R” is initially high, but R decreases massively once enough of the infected clusters are saturated, possibly low enough to make the growth visibly non-exponential, even if the entire population infected rate is nowhere near the point where growth rate would decline in a simple logistic model.
  - rory 5y ago
    Thank you for explaining, that sounds highly plausible!
    - School-Cotton 5y ago
      For what it’s worth, this is just my personal intuition — I’m not an expert in any of the relevant fields.
checkyoursudo 5y ago
> it’s entirely believable to say that omicron could go on to infect fewer people than delta due to the past two years of vaccination and immunity and die off While this does appear plausible, doesn't this depend to some degree on what exactly is the immunity conferred by vaccination or past infection? If vaccines and past infections are effective against severity of case/symptoms but ineffective against new/re-infection, then you could still see greater numbers with omicron than even with delta. No? Still not forever exponential, of course.
- oneoff786 5y ago
  Yes that’s correct
Jansen312 5y ago
In real world I got critized badly even for pointing out media only showing "numbers of infected" and walk into any hospitals now one will see significant less patients than during the peak. We all will get sick from flu. Do we keep harping the potential of Spanish flu return? If we treat everything with utmost cdc-lab4 kind of safety, a lot of things we can't do and even earn a living. Up to a point, people need to realize we have to live with Covid just like everyone living with the virus descendent of Spanish flu. And Omicron should be viewed as blessing to some extend as it is less severe and is one way nature provide immunity to us. A lot of people would want natural immunity than constantly getting jab every couple of months (Israel going for 4th right now).
- sjg007 5y ago
  Hospitals are packed... idk what you are talking about.
  - afrodc_ 5y ago
    Where are they packed? Genuinely curious to see some data about this.
    - sjg007 5y ago
      I don't think you are genuinely curious at all. And if you were actually genuinely curious you could do a cursory Google search or even read a major newspaper. Instead you sit waiting on someone else to deal with your asinine request.
      - afrodc_ 5y ago
        Man this is the most non constructive response I've heard. I'm looking for a resource of data showing hospital utilization and you come say me with this nonsense.
    - chaostheory 5y ago
      West coast, Midwest, and north east: Ohio, Rhode Island, Alaska, Iowa Providence hospitals on the west coast are overwhelmed Hennepin hospitals are overwhelmed in Minneapolis Trinity Health with hospitals in 25 states is also getting overwhelmed https://www.npr.org/sections/health-shots/2021/12/21/1066093265/with-omicron-now-dominant-depleted-u-s-hospitals-struggle-to-prepare-for-the-wor https://www.npr.org/sections/health-shots/2021/12/21/1066093... “ The troops deployed to hospitals will include doctors, nurses, paramedics and other personnel during January and February as needed, the senior administration officials said. The White House is also deploying six emergency response teams to Michigan, Indiana, Wisconsin, Arizona, New Hampshire and Vermont, the officials said.” https://www.google.com/amp/s/www.cnbc.com/amp/2021/12/21/omicron-us-to-deploy-troops-to-hospitals-purchase-500-million-covid-tests.html https://www.google.com/amp/s/www.cnbc.com/amp/2021/12/21/omi...
SergeAx 5y ago
Omicron easily breaks through immunisation (see Norway superspreading event: 60 out of 120 fully vaccinated negative tested people in 3 hours indoors). Question of severity is still open, let's hope for the best.

jonnycomputer 5y ago

Won't downvote you, but I think that this is mistaken. I think that everyone will get omicron eventually.

xyzzy21 5y ago
And it will generate herd immunity...
- jonnycomputer 5y ago
  Maybe. There is a lot of reinfection/breakthroughs going on from alpha/delta to omicron. Does conferred immunity to omicron protect from alpha/delta? I don't know enough to say.
bufferoverflow 5y ago
You're mistaken. There's no disease on this planet that infects 100% of people, even when constantly exposed. Our immune systems are quite amazing.
- jonnycomputer 5y ago
  Who hasn't had a cold?
  - oneoff786 5y ago
    “A cold” is not a single virus.
  - bufferoverflow 5y ago
    Many people.

snet0 5y ago

On the subject of oversimplified exponential models, we don't even need covid data to know this to be true. The fact that there is a duration of an individual case being contagious is enough to make that obviously incorrect.

svara 5y ago
That's not right, you can build a maximally simplistic SIR model with a finite infectious period. That will initially display approximately exponential growth, and slow down as acquired immunity increases.

donquichotte 5y ago

Thank you. It is mind boggling to me how politicians and scientists (well, mostly MDs) keep talking about exponential growth and make horrible predictions, without realizing that exponential growth is never sustainable - a logistic growth is what often happens in reality if resources are limited.

Talanes 5y ago
Asking politicians to not over-estimate exponential growth is like asking water not to be wet.
- lb0 5y ago
  Not understanding what you are both getting at ... the situation has maybe changed now. But before, here at least, it was repeatedly underestimating the in the aftermath clearly visible exponential growth, exactly as the r factor is calculated? Which scientist/MD doesn't understand there is a saturation, in the extreme even schoolers understand you cannot have more than 100% infected. Yes, at some point it is becoming logistic, but the previous waves were not about getting anywhere there, but going from 0.1% to 1% or more, and there it was truly exponential growth and overwhelming the health care system, what the start of that logistic curve is about..?!
- snth 5y ago
  Isn't the usual complaint that people (politicians included) _under_estimate exponential growth?
  - hsbauauvhabzb 5y ago
    Depends where you are, some parts of australia have certainly been fairly conservative. Opinions aside, they certainly estimated on the upper bounds of severity much more than the US. I personally am grateful that the aus gov did this, saving countless lives in the process.
II2II 5y ago
You're right, but there are a number of factors at play here. First and foremost, we can only make predictions based upon the data we have. The data we have is mostly based upon people going for voluntary testing. Given the variability in symptoms that motivate testing, varying levels of awareness to potential exposure, varying willingness to get tested, and the availability of testing, the number of known exposures will make it look like we're in an exponential growth phase even when things are tapering off. It is hard to blame public health organizations for accepting this data at face value since the potential consequences of not doing so are extremely bleak. The other consideration is that people will modify their behaviour based upon perceived risk, whether or not restrictions are imposed. People will tend to comply with restrictions, go about life as normal, follow some sort of middle path, or impose tighter restrictions upon themselves. It takes a truly special person to put themselves into a position of greater risk. Even then, there is a good chance that they are doing little more than translating one high risk circumstance into another (more or less balancing out the growth). We are facing a situation where scientists can make predictions based upon what is known, e.g. the outcome of restrictions, or making predictions based upon anticipated behaviour. Given there is not enough data to model anticipated behaviour, the natural response is to rely upon models that use expected behaviour (e.g. compliance with restrictions or no response). Since the average actual response will prefer over compliance, more bleak predictions are made. Finally, everyone has a different understanding of life now as compared to life before the pandemic. I remember when the coronavirus first caught my attention: it was when major U.S. universities started shutting down. I remember when I first started taking it seriously: it was when my provincial government issued a shelter in place style order. Since that time, I have paid attention to what is happening and thinking about how I should respond. Sometimes it is in an acute manner. Sometimes is in a cursory manner. Either way, I am more likely to notice and respond preemptively to outbreaks. I suspect that many, if not most, people are the same even if their only actions are stocking up to prepare for the worse. Again, this will affect outcomes.

stefan_ 5y ago

It's a bit silly to go "it's less severe" based on death statistics when after two years of Corona waves, there are plain less individuals around for it to kill.

oneoff786 5y ago
Death of the vulnerable is a thing, which I referenced, but it’s a relatively minor factor compared to vaccinations and immunity build up from past infections. Most people who would die from covid if they caught it did not catch it. I think it’s a bit silly to fixate on the base stats of the virus rather than it’s actually efficacy against the human population.
- midasuni 5y ago
  The problem is by the time you see the effect (hospitilations) it’s too late to do anything about it. It’s like driving with a 20 second lag in your vision.
dnautics 5y ago
no viral infection happens in a vaccuum. Each year's seasonal and nonseasonal flu comes on the backs of years of "less people" for it to kill, and people who have gained immune responses to any of its closest relatives, and people who have been immunized, and even humans who have been culled by similar viruses milennia ago during migratory bottlenecks.

orblivion 5y ago

If you're right, that would be disappointing, because it means there's less immunity gained than I was hoping, for the (relatively small) amount of hospitalization we've seen. And I (uneducated on the matter) would expect that if it's not really getting around that much, it would not displace delta. But I've heard some claims that it already has started to displace Delta, though not from a source I'd feel confident citing. But just looking at it, if you believe the CDC estimate of 73% Omicron the other day, Delta must have dropped a lot despite it being winter. Appears like displacement happening.

adzm 5y ago
The CDC estimate was a projection, not an actual recording of data.

renewiltord 5y ago

Personally, I'm just waiting for the day when the 2^64th death happens.

specialp 5y ago

A place where it hit exponential levels with really bad results is here in NY at the beginning. We had no precautions, testing, or vaccines in place yet, and hospitals were overwhelmed. It was honestly a scary time. That was where we started "flattening the curve". Now with vaccines available if there are surges and they are not literally building hospitals on fields like they were near me that is something we have to just live with. COVID will never hit 0 cases. People will still be hospitalized and occasionally die of COVID as was always the case with winter flu seasons. Now we are at the point where fully vaccinated people and people with immunity from a previous infection have a low rate of death or serious illness. People who are very old or have a compromised immune system should lay low and take precautions now and during every flu season when it spikes. Those who choose not to get vaccinated and die, that is still on them.

BrazzVuvuzela 5y ago
Retric 5y ago
We are currently sitting at at total of over 800,000 COVID deaths in the US and 1,500 COVID deaths per day, which is roughly where it’s been since September 1st. That’s approximately an additional half a million dead people per year. So yes the vaccine has already saved million of lives, but I sure hope that this is not the new normal this winter or we are going to cross 1 million deaths fairly soon.
- PaulDavisThe1st 5y ago
  Based on excess mortality numbers, it seems fairly likely that we're already over a million deaths that would not have happened without COVID19.
  - refurb 5y ago
    What will be interesting is to see the long term mortality rate. My hypothesis is we’re just accelerating the deaths of people over 70 years old (70%+ of the cases). If that’s true, you’d see a decrease in mortality over the next few years. Covid is basically doing what the flu does every winter (but on a larger scale) - it pushes people close to death over the edge.

ses1984 5y ago

Spread doesn’t have to get close to 100% for things to get really bad.

oneoff786 5y ago
Indeed. That’s true. But thankfully that’s not in the table, or anything close to it. I only bring that up because it’s the natural conclusion of what will happen if you misunderstand how the exponential growth works.

rightbyte 5y ago

I wonder how many of these increasing share variants you would expect to observe from just random walk. I.e. how much more infectious does a strain need to be to be measurable.

rich_sasha 5y ago

This is of course true, but I'd imagine the counterargument and concern is that healthcare runs out of capacity well before the growth slows to a sub-exponential run.

joshocar 5y ago

What about the population of the unvaccinated? What percentage should we expect of the unvaccinated to get omicron given how contagious it is?

oneoff786 5y ago
It depends on many things and will vary wildly by geography. I’m not suggesting this is a population level trend one can safely piggyback on, unvaccinated. If you are unvaccinated in a community with spread, your odds of getting it are likely comparable, plausibly higher, to previous periods.

estaseuropano 5y ago

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.

oneoff786 5y ago
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.
- zuminator 5y ago
  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.

svara 5y ago

Not sure what you're trying to say. Of course there is a limited pool of people to infect. And various effects can cause infections to spread in multiple waves even far from herd immunity. And yet, you can go from a few rare cases in your country to hospitals overflowing in a couple of weeks. That's because of exponential growth, obviously.

oneoff786 5y ago
The point is that 1) exponential growth has consistently ceased before the hospitals overflowed too bad 2) exponential growth confuses readers into thinking that a small boost in spread rate means many more will be infected. This isn’t true because the exponential phase is brief and is a progression towards a ceiling defined by the graph, not an arbitrary period of time. Like what we are seeing in South Africa right now. The big problem is that people think an increase in the spreading rate will cause an exponentially higher number of infected. But it won’t. The model is not appropriate.
- svara 5y ago
  That's fair, yes. High exponential growth rates make the system much harder to control though, and will easily overshoot the 'set point' (maximal hospital capacity). At a 2-3 day doubling time that overshoot could look pretty drastic.
- jamesrcole 5y ago
  > 1) exponential growth has consistently ceased before the hospitals overflowed too bad I believe quite a few hospitals around the world overflowed pretty badly.
  - tasogare 5y ago
    Yes, you believe. The reality is that most healthcare systems, at least in developed countries, handled the patients well enough.
    - 5y ago

erichocean 5y ago

> Presuming data about lower severity holds, it would be surprising to me if hospitalization or deaths aren’t noticeably lower than delta; which, in turn, was noticeably lower than the original. Well, that should always be true: the people that are most at risk are already dead from earlier variants, and our treatments are light years ahead of where they were 18 months ago, reducing hospitalizations (and death) for everyone else.

oneoff786 5y ago
I’ve seen a lot of people saying this. Again, it’s not wrong that deaths are a factor here. But by far, FAR, the largest factor is vaccinations. We should acknowledge this.

raverbashing 5y ago

Yes, I don't think the SIR models work well in predicting the dynamic of waves The exponential part is easy and that works, but it's more complicated with isolation of the infectious, refractory periods, etc

amelius 5y ago

Can you back this up with some math?

oneoff786 5y ago
Look into random graph models. Or exponential random graph models. Mathematical models are difficult because we don’t know the real inputs and fitting a curve in retrospect is easy to get a compelling looking answer which is wrong. You can gain an intuition for it just imagining a random walk on a social network graph though. Just jump from friend to friend randomly on Facebook. Early on it is easy to spread to new people. Later on it’s very difficult to find new people. You get stuck in the same cohorts. More simply just look at past covid outcomes. Or pretty much any epedemiological model. All of them claim only the initial period is exponential. It’s the issue of how to determine the slowdown period that’s tricky and frankly impossible without more data than we have. But assuming that the slowdown state will look similar to previous slowdown states is a good idea.
School-Cotton 5y ago
It's an empirical observation, not the output of a mathematical model. The graph of human social contacts is not even close to uniformly random, so it makes sense that simplistic formulas would not work.
kazinator 5y ago
https://en.wikipedia.org/wiki/Logistic_function#In_medicine:_modeling_of_a_pandemic https://en.wikipedia.org/wiki/Logistic_function#In_medicine:...

analog31 5y ago

At the start of the pandemic, when downloadable data became available, I started charting it every day. The curves for many countries were uncannily exponential. Of course I knew that those curves had to tip over at some point, but I decided that I would use the exponential growth curve as a baseline and watch for deviations from it. I told my friends: "I'll breathe a huge sigh of relief if this thing turns over before it's infected most of us." In fact it's probably hard to speculate why it turned over, but behavioral measures are probably an important factor.

oneoff786 5y ago
It’s not uncanny, it is, literally, how disease spreads. It ceases to be exponential when the people infected people are exposed to are no longer all susceptible because they themselves had been infected. At least in simplest terms. That’s the big assumption in R0.

timdaub 5y ago

You're wrongly assuming that all incidence data of COVID-19 accurately represents its real spread. E.g. it may just be that beyond a point of exponential growth testing ceases to be accurate. The German Robert Koch institute has pointed out many times that there's a hidden incidence number of all actual cases and the official one that represents everyone tested and recorded.

oneoff786 5y ago
This assumption isn’t critical to anything I have said

moksly 5y ago

> it would be surprising to me if hospitalization or deaths aren’t noticeably lower Here on vaccinated Denmark this is the truth. But we’re actually approaching last years levels regardless because of just how many people are getting covid this year. It’s more manageable thanks to the very high vaccination status, but we seem to have been too slow with the 3rd hit for a major part of the population. I have two shots myself, and my family is all in covid isolation all tested positive and “looking forward” to spend Christmas with ourselves and not our families. It’s not too bad for any of us, it’s not pleasant either, but the biggest thing for me is how much we’re having to shut down despite the high vaccination percentage. You can’t go to a movie or actual theatre. Bars close at 22:00. Most major Christmas parties (this is a big thing here) have been cancelled. But the biggest impact is on culture business like the theatres, concert houses, Christmas markets, museums and so on. If covid is going to be a recurrent thing every winter, then I think that we’re going to see some drastic changes to those aspects of society.

trhway 5y ago
what is the point of those lockdowns when 1. practically everybody is vaccinated and thus gets no symptoms or only light disease (and according to the latest science - superimmunity after that) and 2. the delta's and now omicron's transmissibility is so high among both - vaccinated as well as unvaccinated - that everybody susceptible to it (i.e. 30-70% of population, and anecdotally it seems to me that for covid it is on the lower end, ie. like 1/3rd) will get it (or omicron++) in the near future anyway (and the thing becoming endemic like flu). I mean technically the lockdowns possibly make sense as a way to control the spread until a spread controlling vaccine (the existing mRNA ones aren't such) is widely applied - ie. that seems to be the case in China where initial spread was effectively controlled and they use inactivated covid virus vaccine instead of mRNA - though we don't know for sure because Chinese government info can be very different from reality.
- oneoff786 5y ago
  Both if your premises are not accurate
- 5y ago

stefanpie 5y ago

I don't know how well this relates, but the topic of logistic growth (exp. growth with a limit) is common in undergraduate differential equations classes. I guess it can work well for simpler models. https://en.wikipedia.org/wiki/Logistic_function#Modeling_early_COVID-19_cases https://en.wikipedia.org/wiki/Logistic_function#Modeling_ear...

rory 5y ago

What you're saying is supported by the data, but do you (or does anyone) know why infections slow down when they do? I can't think of any intuitive reason for it, and haven't been able to find an explanation why the waves crash, so to speak.

School-Cotton 5y ago
My guess is the non-uniform nature of the graph of human contacts, which is probably made up of several densely-connected clusters each with relatively weaker links to other clusters. E.g. let’s say there is an office where everyone works in person and a bar with a group of regular customers. If someone in the office gets Covid, everyone in there has some decent chance of getting it. Similarly, if a bar regular gets Covid, each of the others has a decent chance to get it. But if only one person who works at the office is also a regular at the bar, then for the infection to hop from one cluster to the other, that person needs to get it from the initial outbreak, and they need to continue going into the office during their infectious period, and folks need to catch it from them, none of which is certain. So, my guess is that after enough of these clusters get seeded to start a wave, “R” is initially high, but R decreases massively once enough of the infected clusters are saturated, possibly low enough to make the growth visibly non-exponential, even if the entire population infected rate is nowhere near the point where growth rate would decline in a simple logistic model.
- rory 5y ago
  Thank you for explaining, that sounds highly plausible!
  - School-Cotton 5y ago
    For what it’s worth, this is just my personal intuition — I’m not an expert in any of the relevant fields.

checkyoursudo 5y ago

> it’s entirely believable to say that omicron could go on to infect fewer people than delta due to the past two years of vaccination and immunity and die off While this does appear plausible, doesn't this depend to some degree on what exactly is the immunity conferred by vaccination or past infection? If vaccines and past infections are effective against severity of case/symptoms but ineffective against new/re-infection, then you could still see greater numbers with omicron than even with delta. No? Still not forever exponential, of course.

oneoff786 5y ago
Yes that’s correct

Jansen312 5y ago

In real world I got critized badly even for pointing out media only showing "numbers of infected" and walk into any hospitals now one will see significant less patients than during the peak. We all will get sick from flu. Do we keep harping the potential of Spanish flu return? If we treat everything with utmost cdc-lab4 kind of safety, a lot of things we can't do and even earn a living. Up to a point, people need to realize we have to live with Covid just like everyone living with the virus descendent of Spanish flu. And Omicron should be viewed as blessing to some extend as it is less severe and is one way nature provide immunity to us. A lot of people would want natural immunity than constantly getting jab every couple of months (Israel going for 4th right now).

sjg007 5y ago
Hospitals are packed... idk what you are talking about.
- afrodc_ 5y ago
  Where are they packed? Genuinely curious to see some data about this.
  - sjg007 5y ago
    I don't think you are genuinely curious at all. And if you were actually genuinely curious you could do a cursory Google search or even read a major newspaper. Instead you sit waiting on someone else to deal with your asinine request.
    - afrodc_ 5y ago
      Man this is the most non constructive response I've heard. I'm looking for a resource of data showing hospital utilization and you come say me with this nonsense.
  - chaostheory 5y ago
    West coast, Midwest, and north east: Ohio, Rhode Island, Alaska, Iowa Providence hospitals on the west coast are overwhelmed Hennepin hospitals are overwhelmed in Minneapolis Trinity Health with hospitals in 25 states is also getting overwhelmed https://www.npr.org/sections/health-shots/2021/12/21/1066093265/with-omicron-now-dominant-depleted-u-s-hospitals-struggle-to-prepare-for-the-wor https://www.npr.org/sections/health-shots/2021/12/21/1066093... “ The troops deployed to hospitals will include doctors, nurses, paramedics and other personnel during January and February as needed, the senior administration officials said. The White House is also deploying six emergency response teams to Michigan, Indiana, Wisconsin, Arizona, New Hampshire and Vermont, the officials said.” https://www.google.com/amp/s/www.cnbc.com/amp/2021/12/21/omicron-us-to-deploy-troops-to-hospitals-purchase-500-million-covid-tests.html https://www.google.com/amp/s/www.cnbc.com/amp/2021/12/21/omi...

SergeAx 5y ago

Omicron easily breaks through immunisation (see Norway superspreading event: 60 out of 120 fully vaccinated negative tested people in 3 hours indoors). Question of severity is still open, let's hope for the best.