One question that's been on my mind: The number of Coronavirus cases is likely far more than what's being reported. Apparently there are those who get coronavirus and are unaware that they have it before they heal. However everyone that dies of coronavirus is reported as a coronavirus case.
Does this mean that there's a good chance the death rate is lower than the current calculated 3% with so many recoveries going unreported?
> Does this mean that there's a good chance the death rate is lower than the current calculated 3% with so many recoveries going unreported?
Yes and no. I'm not a doctor so an actual MD will step in here and correct me I'm sure. My understanding from reading both the CDC literature and the WHO literature is that there are two things, the SAR_nCOV2 virus which is out there spreading from host to host, and the COVID19 disease which is one manifestation of the virus in people.
If my understanding is correct, the answer is that not everyone who gets infected by the virus will develop the disease. That can happen for a number of reasons apparently including but not limited to, a small enough sample got your immune system going fast enough that you didn't develop the disease, or you have previously been exposed to something similar enough that your immune system got ahead of the virus, or during the course of the infection your lungs didn't get involved before the immune system had begun producing anti-bodies.
To explain the numbers, the fatality rate is a function of only those people who develop the disease. Your chance of developing the disease is a function of how strong your immune system is and so it develops more commonly in people with weak or otherwise compromised immune systems.
But take this as just one opinion based on reading the literature. It sounds to me as if it is similar to the difference between HIV and AIDS, where you can become infected with the HIV virus but not develop AIDS if that virus is kept minimized with anti-virals.
So early on I was suspicious of the numbers from China, but since they have managed to control the spread (and it would be obvious if they didn’t and other cities started to look like Wuhan), that means they have done a good job of catching almost all of the cases, not just the symptomatic ones. Otherwise the ones they miss would continue to spread.
Therefore it feels like at this point their data would be pretty representative of what percentage of the cases are severe vs critical and what the death rate is (accounting for both the overwhelmed medical case in Wuhan and for the under control case in rest of China).
South Korea is also doing a pretty good job with testing to control the outbreak although they haven’t yet gotten it as under control as China does. So their numbers would be a good data point to cross-validate the China numbers.
In countries like Italy (or Wuhan early on) the number of confirmed cases wasn’t equal to total cases as evidenced by continued growth of the epidemic. So it’s reasonable in those cases to assume that more severe cases are being found and less severe ones are being missed.
I'm no epidemiologist, but I think we might have to wait another month or two to see if there are any resurgences in China before we can be certain the containment really worked, no? Infectious disease expert Michael Osterholm recently said on a podcast that he and his colleagues expect there to be another outbreak in China within the next few months, once people start fully co-mingling again.
I don't see why they would back off of the monitoring and fever clinics without some better measure in place (a vaccine or effective treatment or whatever). It's of course very likely there will be new cases. The likelihood of unchecked spreading seems much lower.
I know nothing about all of this, but I guess the idea is there are too many closely concentrated people and too many variables to control for, so if there are new cases, they could potentially turn into new outbreaks if they aren't detected and handled in time. And it's likely just a matter of probability whether or not they all get detected. Even the most advanced and restrictive surveillance state in the world can't possibly catch everything with 100% probability.
I think there's a decent chance the outcome won't be nearly as bad as the worst case scenario, but the worst case scenario is pretty scary. There is a non-zero probability that this wave is only beginning, that there are going to be future outbreaks and pandemic waves this year, and that by year's end there'll be millions of deaths globally.
In the video he talks about taking temperatures every time someone enters a building.
Anyone with a fever is screened for the infection using a portable CT scanner (they don't do a full image, they take a couple slices).
People with suspect CT results are tested and held in quarantine until the test result comes in.
The virus simply will not spread as fast in those conditions as it did when local officials were covering up, and people can start moving around more without backing off those conditions.
If you're suspicious of the numbers of China, why would you not be suspicious when they said they 'controlled' the spread? I mean the press has all but forgotten about China and focused on the global spread, so there's probably less information being reported from China. Unless I had some good evidence apart from China state news sources (which is pretty much where this info is coming from), I would remain suspicious.
Isn't also true for influenza? I had flu symptoms back in January but other than telling my manager that I was going to be out of the office due to illness, never reported my flu to anyone. Doubt I'm the only one who did this.
CDC estimates the total number of flu infections at about 10-20% of the population, depending on the year. The total number of hospital cases is usually ~1-2% of the number of infections.
The, e.g. “32 million cases of flu” reported by the CDC is not based on 32 million positive test results. It’s extrapolated based on a factor of how many cases require hospitalization, and how many people seek medical care, which itself is based on antibody and behavioral surveys.
> The numbers of influenza illnesses were estimated from hospitalizations based on how many illnesses there are for every hospitalization, which was measured previously (5).
> Some people with influenza will seek medical care, while others will not. CDC estimates the number of people who sought medical care for influenza using data from the 2010 Behavioral Risk Factor Surveillance Survey, which asked people whether they did or did not seek medical care for an influenza-like illness in the prior influenza season (6).
After reading more about it [1] you almost wonder how much better than a WAG (“Wild Ass Guess”) it really is.
Frankly I am finding this whole area of medical science entirely underwhelming relative to the potential impact to lives and economies of the world.
Yes this happens, but for the flu we have done large scale antibody studies, so we have an idea that for every x number of people who get treated, y number are actually infected.
The problem is we don't have the same data for this virus yet.
I don't think so. There was a surveillance study in I believe Guangdong province where they tested tons of random people, and compared the positive rate with cases/population, and found that they had detected most all of them. Don't have the link at hand, sorry.
There was also a statistical analysis of the Diamond Princess cases. That produced an estimate that about 17% of infections were asymptomatic. [1] It's possible that there are less asymptomatic cases among the older population, but between the Guangdong and the Diamond Princess data, I doubt it.
The 3.4% CFR that everyone is looking at is also dependent on some reasonably adequate level of medical care. If this thing blows up, most people are going to be on their own, and in the absence of a miracle drug that is widely available (maybe Chloroquine?), the fatality rate with home care I would expect to be something like 10-15%.
> However everyone that dies of coronavirus is reported as a coronavirus case.
That seems unlikely.
Testing capacity in lots of the world is extremely limited. The US is reserving test kits for the most sick/at risk; people at a low risk are unlikely to be tested.
It seems to me that the focus would be on testing alive people, and that we are subsequently undercounting deaths - at least, in these relatively early stages.
My impression was, that everybody with syptoms who ot hospitalized also got tested. So the number of undiagnosed death should be very low. And by an order of magnitude lower than the number of untested, but infected, cases out in the wild.
But as everybody points out, the correct number can only be claculated after the fact. and even the it might be of.
I would definitely suspect you are right, though: many of the reports we’ve seen so far indicates community spread was occurring before the first case was detected - consider the Ohio estimate yesterday that ~1% of their population is infected.
I would say the odds that the hospitalization rate or death rate are significantly overestimated is pretty small given what it has done to the hospitals systems in places where it has spread even a little bit.
The absolute number of severe cases to be hospitalized overwhelmed hospitals. The relative number of severe cases, and fatalities, depends on the denominator. In our case, positive test results. So both things are true.
I'm not sure where you're getting the 3% number from. If you read the New York Times maps page data[1] and just divide the total deaths by total cases you get 3.7%, but that isn't the death rate because it doesn't take in to account a lot of factors.
Instead, there's a good article[2] about how bad the virus will get, that says estimates put it in the range of 0.1% to 3%. (With the seasonal flu being somewhere around 0.1%, but the Spanish Flu of 1918 being somewhere around 3%.)
That said, the general intuition that less reporting of cases means that some part of the death rate is over-inflated makes sense. But also consider that there are factors that go in the other direction. Anyone who currently has the virus but has not recovered or died can be under-inflating that rate. And there are likely lots of other factors, which is why articles often mention that the "true" death rate cannot yet be calculated.
(Also the death rate varies a lot with age, and isn't the only factor in determining the "badness" of a virus. Which that second article goes into more depth on—things like how contagious it is, and how early a vaccine can be developed. So the average rate being half of the current numbers doesn't necessarily mean the concern is over-inflated as well.)
Yeah, I think that's referred to as "the denominator problem". Have a listen to Sam Harris' podcast on it this week - the expert on there sees 0.6% as an upper bound for mortality for coronavirus as that's the figure from South Korea, where the testing has been most intensive, and even that doesn't cover everyone.
But that's with the assumption the health system doesn't get overwhelmed. Once we run out of available oxygen, ventilators and ecmos that number starts to shoot up.
Do you mean a lower bound? If SK, which seems to be doing a relatively good job is experiencing a CFR of .6%, then countries performing worse should experience a higher CFT.
0.6% is way out of date, as it was measured when many of the active cases were very early. At that time, I think I saw that deaths/cases was 0.6% and deaths/(deaths+recoveries) was 28.5%. As of today, deaths/cases is 0.9% and deaths/(deaths+recoveries) is about 9% . (Going by the numbers at worldometers.info)
That's the upper bound on the fatality rate if it infected the whole population. In actuality not even Hubei had 0.1% of the population infected, so the numbers will be higher if it mostly infects older people (Italy), and lower if it mostly infects younger people (South Korea).
And I would like to know how many people have died not due to coronavirus but it was labeled so because they just happened to have it when they passed away.
From a Penn State epidemiologist (note that the quote below speaks of the "infection fatality rate", which is different from the "case fatality rate", the difference is explained in the article)[1]:
"Scientists working at the London School of Hygiene and Tropical Medicine, Imperial College London and the Institute for Disease Modeling have used these approaches to estimate the infection fatality rate. Currently, these estimates range from 0.5% to 0.94% indicating that COVID-19 is about 10 to 20 times as deadly as seasonal influenza. Evidence coming in from genomics and large-scale testing of fevers is consistent with these conclusions. The only potentially good news is that the epidemic in Korea may ultimately show a lower CFR than the epidemic in China.
...
"On balance, it is reasonable to guess that COVID-19 will infect as many Americans over the next year as influenza does in a typical winter -- somewhere between 25 million and 115 million. Maybe a bit more if the virus turns out to be more contagious than we thought. Maybe a bit less if we put restrictions in place that minimize our travel and our social and professional contacts.
"The bad news is, of course, that these infection numbers translate to 350,000 to 660,000 people dying in the U.S., with an uncertainty range that goes from 50,000 deaths to 5 million deaths. The good news is that this is not a weather forecast. The size of the epidemic, i.e., the total number of infections, is something we can reduce if we decrease our contact patterns and improve our hygiene. If the total number of infections decreases, the total number of deaths will also decrease."
Thanks for posting this. It's exactly the sort of analysis I would want to do if I had way more time, and he explains his reasoning completely in a way that a non expert can understand (and provides his models), perhaps because he is not a professional epidemiologist.
A general question to all. Disclaimer, I'm no epidemiologist and not even a statistician. I may totally misunderstand something here. That said, I copied the data from the US, UK and Italy on cases (go to https://en.wikipedia.org/wiki/2019%E2%80%9320_outbreak_of_no... and click on the relevant flag in the table, copy the data off the graph, scrub with regexes).
I dumped this into Tableau, graphed it and put an exponential trend line over the top. The UK and US data dropped almost perfectly onto an exponential (R-squared > 0.99 for both, P value better than 0.0001 for both, whatever those mean). The visual fit was remarkably good.
The italian data very obviously did not fit so well. It clearly was flatter than the best fit exponential (strange, given that covid is totally ablaze there) and it had an R-squared of 0.946, though it's P value was still less than 0.0001.
Can anyone comment? The raw data for italy is below.
There are some good charts at http://nrg.cs.ucl.ac.uk/mjh/covid19/ which show a change in growth rate in Italy near the start of March, possibly due to the lockdown in Lombardy a week earlier
A possible explanation is that their testing can't keep up with the spread. ~13,000 test per day for the last several days, with more positive results yesterday for the same number of tests.
If true that would apply much more to the US data, where I understand the amount of testing has been extremely low.
In any case the data is surely of confirmed cases, so these have got past the testing, no? And I'm not sure if that would produce a flatter-than-exponential curve, which seems to be what it is.
It depends on how many tests are needed. If they are running 13,000 and need 50,000, the growth will look flat.
And then for the US, if most places are doing a high percentage of the needed tests, then the test results should reflect the real growth. Given the limitations on testing, we likely won't have a good understanding of this for weeks to come (people on the internet saying they should have been tested is only worth so much).
This article leaves off about where a similar report on epidemiological models did a couple of weeks ago: Covid-19 will be very very hard to control if persons showing no symptoms spread the disease too often, but we do not know whether or not they do.
This article does a better job than most, but I think still only half-asses the thought process necessary to logically reason about COVID.
The naive or “known” CFR is the fatality rate to-date divided by the number of positive tests to date.
For example, China at one point decided to consider a large number of people as positive cases based on purely clinical symptoms because they didn’t have the test capacity or specificity to definitively diagnose them. Which is why their case count gapped up one day.
The CFR after all cases have concluded is an upper bound on the IFR. There seems to be no true consensus on the percentage of asymptomatic or very mild cases which would not seek treatment or even testing. This number (which is basically unknowable during an outbreak) is a multiplier on the case count, and acts to reduce the true IFR.
Lastly, as we’ve seen quite clearly... How bad it is depends almost entirely upon the person, or at a higher level, the demographics of the people who are infected.
Italy as of March 11;
> Of the patients who died, 42·2% were aged 80–89 years, 32·4% were aged 70–79 years, 8·4% were aged 60–69 years, and 2·8% were aged 50–59 years (those aged >90 years made up 14·1%). The male to female ratio is 80% to 20% with an older median age for women (83·4 years for women vs 79·9 years for men).
Likewise, the outbreak in a WA nursing home was devastating to an elderly population where they would see 5-10 patients die in an average month.
IMO the path forward is either successful containment—which seems extremely unlikely given the transmission characteristics, but apparently China was able to do it? But where you constantly have to be watching for pockets of reemergence and clamp them down.
Or, if we can keep the virus away from high-risk populations while it spreads “like a flu” through the majority of the population, then you will see herd immunity kick in which ultimately works like a vaccinated population to prevent further outbreak and protect the at-risk population from ever becoming infected.
"How bad it is depends almost entirely upon the person, or at a higher level, the demographics of the people who are infected."
There are other important factors to consider.
There will be a percentage of infected individuals who otherwise would have survived had they had access to respiratory assistance by trained medical staff at a hospital.
But as medical staff get overwhelmed, get sick, or die, as free room in Intensive Care Units is used up, as respiratory assistance technology becomes unavailable because it's busy supporting other patients, the death rate will rise beyond what looking at the demographic data alone might lead one to believe.
In addition, there are likely to be extra deaths from other, completely unrelated diseases and conditions for which people can no longer get treatment as hospitals and medical staff get overwhelmed, and as we run in to medicine and medical equipment shortages from lack of supply in China (and, I expect before long, from India and other countries on which the US medical system relies).
To make matters worse, the people sick with COVID-19 will likely infect a lot of vulnerable patients in the hospitals they go to, as most hospitals just simply aren't equipped to effectively contain such infections, especially when there are so many of them. This will increase the death toll even further.
There are many group living facilities such as prisons and nursing homes which will likely be affected by this as well. They will be especially vulnerable because the people there tend to already not be in the healthiest shape and have weakened immune systems.
China’s method of containment was to scan pretty much the entire population for fever and quarantine everyone who might have been infected immediately and involuntarily, including taking children from parents. Needless to say, that will never happen in the us or Europe.
They did extensive testing and containment but that won’t matter eventually. It will break out into unchecked community spread eventually unless they implement a similar quarantine.
At this point I’m more worried about food, water, and toilet paper shortages due to people hoarding 20 times more than what they need. The virus itself doesn’t sound too bad if you’re under 70.
That’s incredibly short-sighted. This is a lethal disease that’s spreading rapidly across the entire world. Even if you personally don’t die or even feel sick, you could easily pass it on to others who do.
Don’t understand the toilet paper shortage at all. Invest in a bidet or just use normal soap/water (it’s better for you anyway and less chance of hemorrhoids)
Maybe TMI, but I was about to go to the doctor for hemorrhoids a few years ago, instead I bought a bidet and life has been so much better. $70 for a cheap one and $400 for a fancy once with heating. My wife and I are quarantining as a precaution for our daughter who has lung issues and are so glad we have it.
Not bad based on likelihood of dying if the hospitals are functioning normally or not bad based on likelihood of becoming hospitalized and in potential trouble when hospitals are overwhelmed?
Toilet paper shortages are the least of our problems, hospital capacity to treat severe and critical cases is the issue at the moment.
People don't need numbers any more. It's only going to matter to some people if someone they view as a 'healthy, normal' person tests positive. Even more of a freak-out if someone like that dies. If a famous celebrity or athlete gets were to die from this, it would set off a lot more worrying
Does this mean that there's a good chance the death rate is lower than the current calculated 3% with so many recoveries going unreported?