> Studying changes in death rate is tricky because although the overall U.S. death rate for COVID-19 seems to be dropping, the drop coincides with a change in whom the disease is sickening.
> So have death rates dropped because of improvements in treatments? Or is it because of the change in who's getting sick?
> To find out, Horwitz and her colleagues looked at more than 5,000 hospitalizations in the NYU Langone Health system between March and August. They adjusted for factors including age and other diseases, such as diabetes, to rule out the possibility that the numbers had dropped only because younger, healthier people were getting diagnosed. They found that death rates dropped for all groups, even older patients by 18 percentage points on average.
> The study, which was of a single health system, finds that mortality has dropped among hospitalized patients by 18 percentage points since the pandemic began. Patients in the study had a 25.6% chance of dying at the start of the pandemic; they now have a 7.6% chance.
> mortality has dropped among hospitalized patients by 18 percentage points since the pandemic began. Patients in the study had a 25.6% chance of dying at the start of the pandemic; they now have a 7.6% chance.
It's strange to phraseit like that; I would have phrased it as "mortality has dropped by a factor of 3"
I prefer the original - it tells me that 18% of all hospitalised patients are alive today where they would not have been before.
Yours tells me 2/3 of them that were dying are no longer, but it puts me on edge like I might be being baited into clicking - a 2/3 drop isn't necessarily impressive, like in normal times if 0.5% of ED patients are dying, dropping by a factor of three might just be the difference between a good and bad month.
18pp is always a big chunk (~1/5) of the total, it's notable whether the starting point was 90% or 20%.
hmm..I see your point but I'm not sure I agree, saying 2/3 of people that otherwise would have died are no longer is tailored to the problem. E.g. if mortality rate from all covid infections went from 0.1% -> 0.001%, thats incredibly impressive and important since the mortality is now 1% of what it was, meaning 99% of deaths are being avoided but "the mortality rate went down by 0.099%" sounds small and insignificant.
Which is why it is good to give absolute risks before and after a change. That way you can easily compute and judge both absolute risk reductions and risk ratios.
With the 18% figure I need to know the starting point to get the picture. It could still be 82% or all the way down to 0%. Three fold increase in survivability is more informative.
So the remaining question is, has the virus changed (mutated) to be less deadly, or have masks lessened the viral load on initial infection enough to give the immune system a better chance of fighting it off, or have treatments become that much more effective? Or a combination of the above?
We see the same drop in death rate across the board, by state, by country, regardless of policy choices. Regardless of mask wearing. Sweden's daily death count has flatlined at near zero per day for over three months, Florida, Wyoming, Idaho, South Dakota, etc all seeing similar drop in hospitalizations and deaths.
You say "death rate" in the first sentence, but then talk of "death count" and "deaths" in the rest.
Did you mean "death rate" in the rest too? If you meant "death count" or "deaths", then it is not the cast for South Dakota. Their deaths have gone way up per capita over the last few weeks.
There are two big questions most people would like answers too:
1. What are my chances of getting COVID?
2. How screwed am I if I get COVID?
There are a couple of different numbers that can be used for the second.
First, the death rate among people who have COVID can be given. Second, the deaths per capita due to COVID can be given.
If you have the answer to the first, then either form of answer to the second can be computed from the other form of answer.
What's going on in South Dakota is that they are doing very badly on the first half (getting COVID). By "badly" I mean that they make the worst in Florida, back when Florida was the epitome of mishandling COVID, look great.
So even if the death rate among people who have COVID has gone down a little, the deaths per capita has gone way up. The death rate among people with COVID would have to go down way more to offset the way the number of people who get COVID is increasing in South Dakota.
You can see mortality rate plummeted shortly after the pandemic started and that it has stayed flat or dropped even in the face of the recent rise in cases.
While cases have gone up finally in South Dakota, the mortality rate is still less than a 10th of what it was in the beginning.
So meanwhile, they were open for 9 months. That is well beyond the incubation period for this virus so what changed to cause the spike?
They are still ranked 40th in terms of deaths per million by state. The worst states have been New Jersey and New York, and Massachusetts, by far, who failed to protect their nursing home and assisted living patients early on in pandemic.
Florida is completely open right now. Not only is there no mask orders, there are no restrictions on populations or businesses.
I wonder if there are also statistics on permanent damage caused by the virus because death isn't the only risk - I've heard of 30-somethings being told they'd never be able to ski again after getting it.
Aside from other factors like better treatments, it seems logical that mutations of the virus that make it less severe and therefore less detectable would spread more easily. Social distancing could inadvertently be breeding the virus towards banality
That might happen eventually but so far there is no evidence that any of the newer strains are less lethal than the original one. They seem to be clinically equivalent. Coronaviruses mutate slowly.
Its crazy they didn't talk about testing. In April if you were a 25 year old with a mild cough you couldn't get a test even if you begged for one. Now you can get a test easily, its little wonder the denominator has gone up a lot.
Edit: Yeah I see its hospitalized and they tried to control. I'm not convinced its a useful study but at least dont upvote this.
The studies here attempt to sidestep the question of testing by measuring only hospitalized patients, adjusting for age and health conditions to try and rule out the "otherwise healthy people couldn't get into the hospital before" hypothesis.
It could also be that back in April not all patients that should have been hospitalised were indeed hospitalised (for lack of beds/doctors), presumably the threshold for "you should be admitted into the hospital" was higher in April compared to lets say July or August. As such, a lower "hospitalisation threshold" for July or August (compared to April) should have indeed caused a lower death-rate.
I don't see why a drop death rate of people getting submitted to hospital is not a useful study?
It is not measuring mild cases, its measuring people who are severe enough to require hospitilasation. This is clearly showing we are improving in the treatment of the disease.
One person's experience doesn't really demonstrate efficacy; in his age group, the survival rate is still ~94%.
The costs of Trump's extremely aggressive treatment course - a dedicated staff of 12 doctors, a private suite, experimental expensive therapies, etc. - aren't feasible for the general population, either.
Great news! What about long-term impacts, like damage to the lungs or heart?
We'll likely have a cohort of patients, like "Americans alive during 2020-2021", that'll be greatly impacted by health issues as they get older if we don't figure out how to mitigate long-term impacts.
You're probably right, but I think you're getting downvoted because you're not providing a source.
Polio -
"Poliovirus causes acute, nonpersistent infections, virus is transmitted by infectious humans or their waste, survival of virus in the environment is finite, humans are the only reservoir, and immunization with polio vaccine interrupts virus transmission."
"Research so far suggests many species can be infected. In lab experiments, cats, fruit bats (Rousettus aegyptiacus), ferrets, rhesus macaques and hamsters have been shown to be susceptible to SARS-CoV-2. Outside the lab, animals including pet cats and dogs, tigers and lions at zoos, and farmed mink have also caught the virus — probably from people."
So, coronaviruses can hang out in animals, which makes it really hard to eradicate because it will just keep showing up. With Polio, however, if you kill it in humans, it's not hanging out in animals, and it's just goooone.
Plus, with Poliovirus, vaccine adoption was high because the thing paralyzes children. As a result, parents who were considering vaccination were sort of faced with a very straightforward question: "Do you want your child to be in a wheelchair for the rest of their life like the neighbor's kid, or do you want to give them a quick vaccine?." With COVID, we don't have such a clear-cut causal relationship in terms of "give your children this vaccine or XYZ really bad things will happen to your kid." If you don't vaccinate them now, as a parent, you won't have to explain to them 10-years down the road why you're the reason they have to spend the rest of their life in a wheelchair. So, although it's a bit premature to say anything, I think it's reasonable to expect that vaccine adoption will be a little lower.
Haha, I'm getting downvoted too, so I guess it wasn't because you didn't provide a source.
Oh well, love to hear anybody with a counter-argument. Is there something I missed?
It's hard to eliminate a virus that hides in reservoirs.
> "Practical disease control requires answers to two questions: 1) Can an acceptable level of control be accomplished without consideration of a reservoir? 2) If not, what populations constitute the reservoir? "
> "Given a target-reservoir system, policies to manage infection may contain elements of three broadly different tactics:
1) target control: directing efforts within the target population with no reference to the reservoir (e.g., human vaccination against yellow fever [23]);
2) blocking tactics: directing control efforts at blocking transmission between source and target populations (e.g., game fences to control FMDV in cattle); and
3) reservoir control: controlling infection within the reservoir (e.g., culling programs, vaccination, or treatment of reservoirs)."
> "These three approaches require progressively increased levels of understanding of reservoir structure and function."
Haydon DT, Cleaveland S, Taylor LH, Laurenson MK. Identifying reservoirs of infection: a conceptual and practical challenge. Emerg Infect Dis. 2002;8(12):1468-1473. doi:10.3201/eid0812.010317
Coronaviruses have a lot of reservoirs.
Shi Z, Hu Z. A review of studies on animal reservoirs of the SARS coronavirus. Virus Res. 2008;133(1):74-87. doi:10.1016/j.virusres.2007.03.012
Scope of the question was whether or not it is possible to eliminate the virus entirely a la efforts with poliovirus. I looked at the literature, and cited it. Doesn't seem possible at the moment, unless we decide to make all of the confirmed and theoretical animal reservoirs extinct (e.g. kill ALL pet cats and dogs).
Doesn't mean you can't vaccinate people so they don't catch it.
I'm not aware of any epidemiologist who believes Covid-19 can be globally eliminated without decades of work. (This shouldn't be surprising, given how few diseases we've managed to eliminate in the past.) By all means let's work on it, but we're going to need to do some management in the meantime.
"A longitudinal study of SARS1 patients from 2003 to 2018 found that more than a third had reduced lung capacity. Similarly, with MERS, a third of survivors had long-term lung damage."
> Horwitz believes that mask-wearing may be helping by reducing the initial dose of virus a person receives, thereby lessening the overall severity of illness for many patients.
Now if only we could drill this lesson into everyone's head.
As I write this the post I reply to is downvoted, but it seems to be true in Norway were there's hardly a mask to be seen outside of the big cities.
Full disclosure: An alternative explanation here in Norway seems to be that the second wave mostly hits younger people.
Edit: there's a dead comment saying "citation needed".
As for the "hardly a mask to be seen" that is my eyewitness account based on what I see in shops and gas stations I visit as well as reading and watching the news.
Point to me a single country or state "with no mask wearing". There are sane people (and believe it or not, a lot of them) everywhere, even where politicians are insane.
As I said, politicians asking people not to wear masks != no people wearing masks. And here's a Reuters report from two months ago saying mask sales soared, picturing at least one person wearing mask: https://www.reuters.com/article/us-health-coronavirus-sweden...
According to YouGov the latest numbers are 9% wearing masks when in public, it has become a bit higher lately (was 7% just a couple of weeks ago). But I think 9% is far too high. I very rarely encounter a mask while taking a walk in the city closest to where I live. Although that might be due to those wearing masks not being as active in public.
I'm pro mask, but I don't get wearing a mask if someone is just walking outside with no one around them. It's not like the whole world turned to Pripyat next to Chernobyl...
Do you recommend 14% mask wearing in the US? What's the current rate in the US? Is there a big difference between the outcomes of 14% and whatever the rate is in the US?
Sweden, along with the other Scandinavian examples, are also way less densely populated, particularly compared to a place like NYC.
There's also completely different base-levels for access to healthcare, different climates, a ton of factors besides masks.
One might as well point at Asian countries like Taiwan, Vietnam or Korea. Those are densely populated countries were mask-use is wide-spread and generally considered to have done best in their response, with the numbers to back it up.
Florida, South Dakota, Wyoming, Idaho, etc. It's not mandated my state law, and it's frequently disregarded in many jurisdictions that do. You would expect that states with stricter restrictions and higher compliance would have better results but the data does not show that.
tl;dr nursing homes need special care during this event and it has been shown that doing so does reduce infection rates if not deaths
As to present numbers...
Well we stopped forcing nursing homes from accepting sick people during the event and not exposing those most at risk does tend to result in fewer deaths.
A lot of deaths could have been avoided if we went out of way to not expose those most at risk but some states actually did the exact opposite.
Years of studying pointed in different directions. Now all of a sudden they work. Same with vaccinations - if you shortcut the process the acceptance rate will drop significantly
You mean the heads of the US health officials who advised not to wear masks at the start of the pandemic because they don’t help? Yeah, I wonder why people distrust them now...
This may be uncomfortable to point out, but at some point, the virus has to run out of susceptible people to kill. The real chance of dying from COVID for an individual is not the average of ~0.4%. It's 0.0% for the vast majority of people, but for people whose immune system fails them, it's closer to 100%. Such a condition usually takes years to build up. Excess mortality graphs show that a lot of people dodged the Influenza bullet in the past years, those were highly susceptible to die of COVID-19 this year.
It’s a satisfying rationale, but it doesn’t take into account a range of unknown factors. What about different strains and variability in how deadly they are? What about the difference that the size of the initial dose of the virus makes to the end outcome? What about the effect of chance in how quickly and effectively a person’s immune system deals with it?
The rationale relies on a set of constants which are... probably not constant.
Of course there's lots of factors, but this is one obvious factor that must not be ignored.
> What about different strains and variability in how deadly they are?
Is there any evidence that the variations spreading now are different from those spreading earlier this year?
> What about the difference that the size of the initial dose of the virus makes to the end outcome?
There was a study done on hamsters that only showed a minimal impact of initial dose on disease severity.
On the other hand, you have increased mortality among healthcare worker, which could be caused by higher exposure. However, there's also the impact of such a stressful situation on the immune system.
Classic problem with derivatives: when you have a "rate" of something, it's with respect to a second thing. Most often it's time, which is what people usually default to if you leave out that piece of information. And that's what your chart shows: steady number of deaths per week. But the article is talking about number of deaths per person hospitalized. That is the rate that is decreasing... with respect to time.
"So have death rates dropped because of improvements in treatments? Or is it because of the change in who's getting sick? "
No, it's likely dropping because the virus is changing. There were several hundred variants of the virus by May (in the UK, most of which were found worldwide). It is a completely normal and expected event that a virus new to humans, will evolve to become less likely to either kill the host, or stir up a counter-response from the immune system, since variants that do either of those have a harder time spreading than variants that don't (or do them less).
John Ioannidis (https://www.who.int/bulletin/online_first/BLT.20.265892.pdf) has demonstrated that the virus' mortality rate is around 0.39% among those under 70. The study here is looking at the mortality rate among those hospitalized, which could be changing simply because those who get sick are more or less willing to go to the hospital, or for half a dozen other reasons that have nothing to do with either who's getting sick or how they are being treated.
Sure, evolution pressure favors easy spreading, but if sick people stay home or go to the hospital, and non-sick people go out, then that pressure is to make the host sick less often.
Early estimates for mortality in Wuhan were 3-5%. Later estimates kept dropping, and are now at 0.39%.
Most viruses do not make the host sick, and the longer it's been in humans, the more likely it is that this is so. If the mortality rate is dropping (the article's assertion), but the WHO says none of the existing treatments has been shown to work, then the simplest explanation is that the virus has evolved in that direction. Given that the virus is very new in humans, there's nothing surprising about that.
Coronaviruses mutate quite slowly, this one has been around for less than a year, you would not expect to see much change in virulence over that timespan. Any lowering of mortality is best explained by more thorough testing.
"There are too many rapid mutations to neatly trace a COVID-19 family tree. We used a mathematical network algorithm to visualise all the plausible trees simultaneously," said geneticist Dr Peter Forster, lead author from the University of Cambridge.
"They identified 198 mutations that appear to have independently occurred more than once, which may hold clues to how the virus is adapting..."
It may "mutate quite slowly" compared to, say, influenza, but compared to anything else it is mutating quite a lot, in only a matter of months. Virus generations are pretty fast.
Interesting article! But, from that very article:
"Despite the virus’s sluggish mutation rate, researchers have catalogued more than 12,000 mutations in SARS-CoV-2 genomes."
Now, that doesn't tell us that these mutations matter, but that's because we don't know which ones matter. But even a rate of two single-letter mutations per month could theoretically result in 2^10 different variants in less than a year. It almost certainly won't, because many of the mutations will be the same (some parts of the genome almost certainly are more likely to mutate than others). But a rate of 2 single-letter mutations per month, is not much of a limit.
The nature article you reference (for which thanks, btw) says its mutation rate is about half of influenza's. Which is, yes, less, but that's still really fast compared to almost anything that isn't influenza. Even a slow-mutating virus is still mutating fast.
What about the vitamin D hypothesis that HN loves so much? I'm far from sold on it yet, but summer in the northern hemisphere should correlate with increased Vitamin D levels, which is hypothesized to be protective from severe results in COVID-19. The concern here then of course is that the death rate is about to start going back up.
Vitamin D is protective against all viruses, not just Covid. But we did see spikes in cases and deaths during the summer months in sun belt states. However, many of the worst cases did come from patients in long-term care. It remains to be seen whether the seasonal flu and pneumonia or Covid will be more deadly this winter.
Here in the UK the deaths are climbing back up again as the second wave comes in. Yesterday we had 241 Covid-19 fatalities, a number not seen since the start of June.
Additionally, the way the UK is counting deaths changed a few months ago. Originally we were counting everyone that had been diagnosed with Covid, now if the death doesn't occur within 28 days of infection the death isn't counted as a Covid-19 death.
And finally, we're testing a hell of a lot more now so the alleged case numbers back in April (~5000 a day) don't marry up at all with case numbers now (~20,000 a day). I'm guessing back in April it was closer to 80,000 cases a day in reality, which would make the disease just as deadly now as it was back then.
That fatalities are going up as cases are going up is expected, but compare the rate of increase now with those during the first wave in March: significant quantitative difference.
People are quite clearly dying at slower rate than in March, while the weekly rate of new cases is growing faster now than they did in March.
Yesterday, 241 Covid-19 fatalities were reported, with the actual dates of death spread out over the previous days/weeks.
Edit: Just for extra clarity - we don't know how many died of Covid yesterday, due to reporting delays. There is data for "Covid deaths by date of death", but it always has a slight drop off due to these reporting delays.
Medicine is basically process engineering: try some stuff and keep knocking back the worst thing on the pareto chart. Make a note of what worked and do more of that but also try other stuff.
The other thing I keep wondering about is whether people that are both sensitive and vulnerable have already caught Covid and have either died or recovered. Since we don't know everything about how you catch it it would be hard for the statisticians to control for that effect.
"Dry tinder" hypothesis: the massive (but hidden) early spread exposed vulnerable people, whereas now they are protecting themselves and possibly receiving smaller inoculations.
The magnitude of the early spread in cities like NY is difficult to estimate. Antibodies does not tell the full story.
88 comments
[ 8.8 ms ] story [ 153 ms ] thread> Studying changes in death rate is tricky because although the overall U.S. death rate for COVID-19 seems to be dropping, the drop coincides with a change in whom the disease is sickening.
> So have death rates dropped because of improvements in treatments? Or is it because of the change in who's getting sick?
> To find out, Horwitz and her colleagues looked at more than 5,000 hospitalizations in the NYU Langone Health system between March and August. They adjusted for factors including age and other diseases, such as diabetes, to rule out the possibility that the numbers had dropped only because younger, healthier people were getting diagnosed. They found that death rates dropped for all groups, even older patients by 18 percentage points on average.
> The study, which was of a single health system, finds that mortality has dropped among hospitalized patients by 18 percentage points since the pandemic began. Patients in the study had a 25.6% chance of dying at the start of the pandemic; they now have a 7.6% chance.
> mortality has dropped among hospitalized patients by 18 percentage points since the pandemic began. Patients in the study had a 25.6% chance of dying at the start of the pandemic; they now have a 7.6% chance.
It's strange to phraseit like that; I would have phrased it as "mortality has dropped by a factor of 3"
Yours tells me 2/3 of them that were dying are no longer, but it puts me on edge like I might be being baited into clicking - a 2/3 drop isn't necessarily impressive, like in normal times if 0.5% of ED patients are dying, dropping by a factor of three might just be the difference between a good and bad month.
18pp is always a big chunk (~1/5) of the total, it's notable whether the starting point was 90% or 20%.
Did you mean "death rate" in the rest too? If you meant "death count" or "deaths", then it is not the cast for South Dakota. Their deaths have gone way up per capita over the last few weeks.
There are two big questions most people would like answers too:
1. What are my chances of getting COVID?
2. How screwed am I if I get COVID?
There are a couple of different numbers that can be used for the second.
First, the death rate among people who have COVID can be given. Second, the deaths per capita due to COVID can be given.
If you have the answer to the first, then either form of answer to the second can be computed from the other form of answer.
What's going on in South Dakota is that they are doing very badly on the first half (getting COVID). By "badly" I mean that they make the worst in Florida, back when Florida was the epitome of mishandling COVID, look great.
So even if the death rate among people who have COVID has gone down a little, the deaths per capita has gone way up. The death rate among people with COVID would have to go down way more to offset the way the number of people who get COVID is increasing in South Dakota.
You can see mortality rate plummeted shortly after the pandemic started and that it has stayed flat or dropped even in the face of the recent rise in cases.
While cases have gone up finally in South Dakota, the mortality rate is still less than a 10th of what it was in the beginning.
So meanwhile, they were open for 9 months. That is well beyond the incubation period for this virus so what changed to cause the spike?
They are still ranked 40th in terms of deaths per million by state. The worst states have been New Jersey and New York, and Massachusetts, by far, who failed to protect their nursing home and assisted living patients early on in pandemic.
Florida is completely open right now. Not only is there no mask orders, there are no restrictions on populations or businesses.
Deaths per day are down and mortality rate is down: https://floridahealthcovid19.gov/#latest-stats. New cases flat.
Edit: Yeah I see its hospitalized and they tried to control. I'm not convinced its a useful study but at least dont upvote this.
It is not measuring mild cases, its measuring people who are severe enough to require hospitilasation. This is clearly showing we are improving in the treatment of the disease.
There was no sacrifice at the expense of other people that needed a scarce resource more, only complete mismanagement and disorganization
If it’s as effective as Trump’s experience has shown, this could have the potential of saving many or most of the US deaths going forward.
The costs of Trump's extremely aggressive treatment course - a dedicated staff of 12 doctors, a private suite, experimental expensive therapies, etc. - aren't feasible for the general population, either.
We'll likely have a cohort of patients, like "Americans alive during 2020-2021", that'll be greatly impacted by health issues as they get older if we don't figure out how to mitigate long-term impacts.
https://eurekalert.org/pub_releases/2020-09/elf-cps090320.ph...
Most people didn't die with Polio... they were somehow crippled for life... I suspect the same is true with Covid-19.
We need to eliminate it, not manage it.
Polio -
"Poliovirus causes acute, nonpersistent infections, virus is transmitted by infectious humans or their waste, survival of virus in the environment is finite, humans are the only reservoir, and immunization with polio vaccine interrupts virus transmission."
Dowdle WR, Birmingham ME. The biologic principles of poliovirus eradication. J Infect Dis. 1997 Feb;175 Suppl 1(Suppl 1):S286-92. doi: 10.1093/infdis/175.supplement_1.s286. PMID: 9203732; PMCID: PMC7110371.
https://pubmed.ncbi.nlm.nih.gov/9203732/
Sars-CoV -
"Research so far suggests many species can be infected. In lab experiments, cats, fruit bats (Rousettus aegyptiacus), ferrets, rhesus macaques and hamsters have been shown to be susceptible to SARS-CoV-2. Outside the lab, animals including pet cats and dogs, tigers and lions at zoos, and farmed mink have also caught the virus — probably from people."
https://www.nature.com/articles/d41586-020-01449-8
So, coronaviruses can hang out in animals, which makes it really hard to eradicate because it will just keep showing up. With Polio, however, if you kill it in humans, it's not hanging out in animals, and it's just goooone.
Plus, with Poliovirus, vaccine adoption was high because the thing paralyzes children. As a result, parents who were considering vaccination were sort of faced with a very straightforward question: "Do you want your child to be in a wheelchair for the rest of their life like the neighbor's kid, or do you want to give them a quick vaccine?." With COVID, we don't have such a clear-cut causal relationship in terms of "give your children this vaccine or XYZ really bad things will happen to your kid." If you don't vaccinate them now, as a parent, you won't have to explain to them 10-years down the road why you're the reason they have to spend the rest of their life in a wheelchair. So, although it's a bit premature to say anything, I think it's reasonable to expect that vaccine adoption will be a little lower.
Oh well, love to hear anybody with a counter-argument. Is there something I missed?
It's hard to eliminate a virus that hides in reservoirs.
> "Practical disease control requires answers to two questions: 1) Can an acceptable level of control be accomplished without consideration of a reservoir? 2) If not, what populations constitute the reservoir? "
> "Given a target-reservoir system, policies to manage infection may contain elements of three broadly different tactics:
1) target control: directing efforts within the target population with no reference to the reservoir (e.g., human vaccination against yellow fever [23]);
2) blocking tactics: directing control efforts at blocking transmission between source and target populations (e.g., game fences to control FMDV in cattle); and
3) reservoir control: controlling infection within the reservoir (e.g., culling programs, vaccination, or treatment of reservoirs)."
> "These three approaches require progressively increased levels of understanding of reservoir structure and function."
Haydon DT, Cleaveland S, Taylor LH, Laurenson MK. Identifying reservoirs of infection: a conceptual and practical challenge. Emerg Infect Dis. 2002;8(12):1468-1473. doi:10.3201/eid0812.010317
Coronaviruses have a lot of reservoirs.
Shi Z, Hu Z. A review of studies on animal reservoirs of the SARS coronavirus. Virus Res. 2008;133(1):74-87. doi:10.1016/j.virusres.2007.03.012
Scope of the question was whether or not it is possible to eliminate the virus entirely a la efforts with poliovirus. I looked at the literature, and cited it. Doesn't seem possible at the moment, unless we decide to make all of the confirmed and theoretical animal reservoirs extinct (e.g. kill ALL pet cats and dogs).
Doesn't mean you can't vaccinate people so they don't catch it.
"A longitudinal study of SARS1 patients from 2003 to 2018 found that more than a third had reduced lung capacity. Similarly, with MERS, a third of survivors had long-term lung damage."
Now if only we could drill this lesson into everyone's head.
Full disclosure: An alternative explanation here in Norway seems to be that the second wave mostly hits younger people.
Edit: there's a dead comment saying "citation needed".
As for the "hardly a mask to be seen" that is my eyewitness account based on what I see in shops and gas stations I visit as well as reading and watching the news.
The numbers of deaths / infections can be derived from here: https://www.vg.no/spesial/2020/corona/?utm_source=corona-wid...
https://today.yougov.com/topics/international/articles-repor...
So telling people to wear what they couldn’t get was awkward.
https://time.com/5899432/sweden-coronovirus-disaster/
There's also completely different base-levels for access to healthcare, different climates, a ton of factors besides masks.
One might as well point at Asian countries like Taiwan, Vietnam or Korea. Those are densely populated countries were mask-use is wide-spread and generally considered to have done best in their response, with the numbers to back it up.
As to present numbers...
Well we stopped forcing nursing homes from accepting sick people during the event and not exposing those most at risk does tend to result in fewer deaths.
A lot of deaths could have been avoided if we went out of way to not expose those most at risk but some states actually did the exact opposite.
The rationale relies on a set of constants which are... probably not constant.
> What about different strains and variability in how deadly they are?
Is there any evidence that the variations spreading now are different from those spreading earlier this year?
> What about the difference that the size of the initial dose of the virus makes to the end outcome?
There was a study done on hamsters that only showed a minimal impact of initial dose on disease severity.
On the other hand, you have increased mortality among healthcare worker, which could be caused by higher exposure. However, there's also the impact of such a stressful situation on the immune system.
https://ourworldindata.org/covid-deaths#what-is-the-daily-nu...
First graph is global deaths daily. Pretty linear to me.
Second grapn is per country/continent.
Third graph plots cases vs deaths.
From this it seems that we detect more cases but same amount of people are dying.
We are testing more. Thats it sadly.
All
No, it's likely dropping because the virus is changing. There were several hundred variants of the virus by May (in the UK, most of which were found worldwide). It is a completely normal and expected event that a virus new to humans, will evolve to become less likely to either kill the host, or stir up a counter-response from the immune system, since variants that do either of those have a harder time spreading than variants that don't (or do them less).
John Ioannidis (https://www.who.int/bulletin/online_first/BLT.20.265892.pdf) has demonstrated that the virus' mortality rate is around 0.39% among those under 70. The study here is looking at the mortality rate among those hospitalized, which could be changing simply because those who get sick are more or less willing to go to the hospital, or for half a dozen other reasons that have nothing to do with either who's getting sick or how they are being treated.
Oh, come on, the pox had mortality > 30 %. Evolutionary pressure favours easy spreading.
Early estimates for mortality in Wuhan were 3-5%. Later estimates kept dropping, and are now at 0.39%.
Most viruses do not make the host sick, and the longer it's been in humans, the more likely it is that this is so. If the mortality rate is dropping (the article's assertion), but the WHO says none of the existing treatments has been shown to work, then the simplest explanation is that the virus has evolved in that direction. Given that the virus is very new in humans, there's nothing surprising about that.
"There are too many rapid mutations to neatly trace a COVID-19 family tree. We used a mathematical network algorithm to visualise all the plausible trees simultaneously," said geneticist Dr Peter Forster, lead author from the University of Cambridge.
Also here: https://www.sciencedaily.com/releases/2020/05/200505190550.h...
"They identified 198 mutations that appear to have independently occurred more than once, which may hold clues to how the virus is adapting..."
It may "mutate quite slowly" compared to, say, influenza, but compared to anything else it is mutating quite a lot, in only a matter of months. Virus generations are pretty fast.
"A typical SARS-CoV-2 virus accumulates only two single-letter mutations per month in its genome."
Now, that doesn't tell us that these mutations matter, but that's because we don't know which ones matter. But even a rate of two single-letter mutations per month could theoretically result in 2^10 different variants in less than a year. It almost certainly won't, because many of the mutations will be the same (some parts of the genome almost certainly are more likely to mutate than others). But a rate of 2 single-letter mutations per month, is not much of a limit.
The nature article you reference (for which thanks, btw) says its mutation rate is about half of influenza's. Which is, yes, less, but that's still really fast compared to almost anything that isn't influenza. Even a slow-mutating virus is still mutating fast.
Pneumonia deaths without Covid have exceeded those with Covid in the United States since August: https://www.cdc.gov/nchs/nvss/vsrr/covid19/index.htm
We're also at less than the number of typically expected weekly deaths across the board for the past couple of weeks.
Here in the UK the deaths are climbing back up again as the second wave comes in. Yesterday we had 241 Covid-19 fatalities, a number not seen since the start of June.
Additionally, the way the UK is counting deaths changed a few months ago. Originally we were counting everyone that had been diagnosed with Covid, now if the death doesn't occur within 28 days of infection the death isn't counted as a Covid-19 death.
And finally, we're testing a hell of a lot more now so the alleged case numbers back in April (~5000 a day) don't marry up at all with case numbers now (~20,000 a day). I'm guessing back in April it was closer to 80,000 cases a day in reality, which would make the disease just as deadly now as it was back then.
People are quite clearly dying at slower rate than in March, while the weekly rate of new cases is growing faster now than they did in March.
Yesterday, 241 Covid-19 fatalities were reported, with the actual dates of death spread out over the previous days/weeks.
Edit: Just for extra clarity - we don't know how many died of Covid yesterday, due to reporting delays. There is data for "Covid deaths by date of death", but it always has a slight drop off due to these reporting delays.
https://www.cebm.net/covid-19/covid-19-florence-nightingales...
The ONS actually shows COVID deaths as now lower than flu deaths:
https://www.ons.gov.uk/peoplepopulationandcommunity/birthsde...
The other thing I keep wondering about is whether people that are both sensitive and vulnerable have already caught Covid and have either died or recovered. Since we don't know everything about how you catch it it would be hard for the statisticians to control for that effect.
The magnitude of the early spread in cities like NY is difficult to estimate. Antibodies does not tell the full story.