Ricochet is the best place on the internet to discuss the issues of the day, either through commenting on posts or writing your own for our active and dynamic community in a fully moderated environment. In addition, the Ricochet Audio Network offers over 50 original podcasts with new episodes released every day.
Day 58: COVID-19 “Shelter-in-Place”
170 countries and territories affected worldwide. Few jurisdictions left not dealing with the virus.
But focusing today on the US: New York had more new cases yesterday than California’s total accumulated cases to date? The Times Union is placing the New York number at 1374, 1106 fewer cases than the Worldometer.com table. According to the Times Union
Here are the latest virus numbers mapped in New York state. Please note, the figures are based New York Department of Health confirmed cases and Times Union reporting. The counts represent where the affected person is a permanent resident, not where they caught the virus.
So, what is true? Maybe someone from the Northeast can tell us what is going on. [Update: Times Union has now updated their count to 2382 which is now only 99 below the Worldometer.com tabulation.]
There was a kerfuffle between Gov Cuomo and NYC Mayor de Blasio: The Mayor said NYC residents should prepare for “shelter in place”, but the Governor said that was his call and he had no intention of ordering it.
Meanwhile, the Bay Area has solved its traffic problem.
[Note: Links to all my COVID-19 posts can be found here.]
Published in General
By liking this, I mean I agree.
Sorry, my phrasing was imprecise. COVID19 is more deadly on a percapita basis than the seasonal flu. So if we let it spread as far and wide as we allow the seasonal flu to all the data indicates it will be worse. The question is how much worse. In Italy it is proving to be something like 80 time more deadly at least according to the number above.
I guess these numbers are a bit lagging, maybe by one day? Here in GA, our numbers were updated to 197/4 today (https://dph.georgia.gov/covid-19-daily-status-report?today). This is the first time I’ve compared.
I really value your daily updates. Thanks for doing that.
Tell me again how Gin & Tonics can be employed to prevent CoronaVirus.
Make a solution in a 1:2 ratio, and apply liberally to the back of the throat.
Is Italy an outlier? I think it’s the textbook definition of an outlier. Why should we use the outlier as the comparison point? I don’t think we should.
Also, when you say it’s more deadly, what do you mean by that? The rate for flu seems to be about 1-2% of infections. WuFlu seems to range between 1-8%. Yes that’s more deadly, as far as it goes, but how far does that really go? Other viruses of recent years have been mopre deadly than flu too yet we haven’t reacted the way we’re reacting now. Shouldn’t we consider that? Is it more deadly enough to justify all of the measures being taken?
Lost a fellow in my area. 80, suffered a stroke a couple of years ago and was wheelchair bound. The widow is in forced quarantine in their house, no deliveries, unable to arrange a funeral, no one to help out. That was early, hopefully some of those issues are getting worked out.
Lord, have mercy.
Again, that rate in itself doesn’t get you very far in the current situation. The other unanswered question is: how infectious is the virus? In other words, what percentage of a community could it infect if left unchecked, and how quickly?
Seasonal influenza typically only infects a small minority of any given community in any given year. But because there is no pre-existing immunity to the present coronavirus, it has the theoretical potential to infect a much higher percentage of a population. So in that case, a similar case fatality rate * much higher infection rate = much higher overall community mortality.
And the other factor is speed: the dynamics of coronavirus spread within communities does appear to be more rapid than influenza. Even if the coronavirus were no more deadly in a given community than the common flu on an annual basis, if it afflicts its patients within a 3-week span instead of over a 3-month span, that’s a severe strain on health care systems.
None of these are firm predictions, just theoretical possibilities with enough evidence to date that they need to be taken into serious consideration.
To date, the virus seems to be behaving similarly in northern Italy as it did in China and South Korea, locations of two other major outbreaks for which we have reasonably good data. Clinical presentation, viral sequence, transmission rate, age distribution, etc. are all reasonably similar, suggesting that there is indeed a standard viral profile.
The simplest explanation as to why the situation in Italy differs so greatly than other countries (except China and S. Korea) is because the infection has been established there longer. In any exponential-growth scenario (not limited to infectious diseases), changes can occur so rapidly that the situation appears to be of a completely different nature before or after only a short shift in time.
Another way of putting it: when the Germans invaded Sedan in 1940, a Parisian could have easily brushed it away as Sedan just being different.
Again, this is not any firm prediction. The uneven spread of the virus is certainly noteworthy. While it may well be simply due to the dynamics of its transmission to date, there is certainly not enough evidence available to rule out other underlying factors like environmental conditions, population density, lifestyle factors, and so on.
To shut down the American economy. Is this not obvious?
What a joke it will be if Global Warming defeats the Corona Virus!
Why is that? Aren’t the relevant rates the most important bits of information? That includes seriousness. Everyone gets a cold, it’s hardly ever serious, so we keep French kissing and shaking hands (probably after not having washed hands in a while).
This is not true according to the charts that Rodin has been posting. There is significant variance in the rates of seriousness. Perhaps you can pick a cluster with similar rates, but you can also pick a cluster with disparate rates. The rate should be the same.
Also, as the number of total cases goes up with testing and such, the rate of seriousness is likely to go down. Right now the serious cases are bubbling to the top. Isn’t this correct?
We fought several wars and spent trillions of dollars over those deaths.
The Italian Association of Biotechnologists wants to make clear that Covid-19 is not like a flu. To make this even more clear we have compared the available data on flu (season 2018-2019) available here with the Coronavirus data collected until March 15th 2020.
In the season 2018-2019, flu was responsible of 812 cases that required Intensive Care Unit (ICU) hospitalization and 205 deaths in a span of 33 weeks (8th October 2018 – 20th May 2019). The peak of ICUs was in week 4 (2019) with 93 cases and the peak of deaths in week 5 (29 deaths).
The data available for SARS-CoV-2 in Italy are on a completely different scale. The epidemic stated in week 9 of 2020, and the week after (10) registered 351 cases in ICUs and 131 deaths. This week (11) there are already 1672 cases in ICUs and 1661 deaths, and the situation is getting critical for the health system of many Italian regions.
The main issue is that for a large number of cases (up to 10%) there is the need for an Intensive Care Unit (ICU) hospitalization, also for young people, and the therapy requires up to 2-3 weeks before passing the critical phase. This requires a tremendous effort by the health system to guarantee adequate care to all in need, and for such a long time, draining resources also for those with other pathologies.
It is true that most of the deaths are elderly people, but only because the elderly often bear other chronic illnesses and do not tolerate invasive treatments well, but the hospitalization does not spare any age.
So, get ready, assist your country to correctly inform its people and help them to respect the rules to stop the contagion. Do it for yourself, for the people you care, for your fellow citizens, especially the weakest.
Again, the problem with the rates is that they all use the total number of cases as the denominator.
But each country (and in some cases each jurisdiction) is testing to a different extent and using different factors to determine who gets tested. This means that the method used to determine the total number of cases differs greatly from one country/jurisdiction to the next. Combined with the generally low rate of testing, and the total numbers of cases are inconsistent and unreliable.
And since the total number of cases determines the rates of serious illness and mortality, those rates are also suspect. Garbage in, garbage out.
The numbers of patients with serious disease and death are likely much more reliable, since reports are that almost every patient hospitalized for severe symptoms is being tested, and because these patients are much better characterized.
Hence, at the time being, the raw numbers of serious cases and deaths are likely much more informative than any rates based on all infections.
Again, the rates are unreliable. The data I was referring to are clinical reviews of cohorts of hospitalized patients and/or specific populations with well-documented testing. This data can’t be found at sites like Worldometer or the Johns Hopkins dashboard but rather in the scientific/medical literature.
And in case series and other cohort studies, disease characteristics like age distribution of serious cases, comorbidities, disease onset and duration, and mortality are quite consistent between China, South Korea, and Italy.
I think this type of presentation is much more useful than arguing about rates. That being said: like any snappy infographic circulating online, caution is always warranted.
The caption mentions a total of 205 influenza deaths in all of Italy during the last flu season. Another scholarly article estimates about 18,000 excess deaths each year (on average) due to influenza in Italy. That’s an enormous difference, and decreases the severity of the Covid-19 outbreak. However, without a detailed analysis of the individual data sources it’s hard to know which is the genuine apples-to-apples comparison.
And in the opposite direction, the infographic shows the data for all of Italy. But the vast majority of Covid-19 deaths have been in the province of Lombardy, which represents one-sixth of the total population of Italy. Viewing it through the more accurate lens of a community outbreak and not a national outbreak yields a much greater threat by Covid-19.
Bottom line: epidemiology is very difficult even when lots of data is available. And currently very little data is available. But the data that is available still provides ample reason for grave concern.
For a country 1/5th the size of the US, that seems like a large number.
Yes, and flu kills 22,000-55,000 each year per the cdc; that’s seven(!) 9/11’s on the low end. Yet no comparisons to 9/11 and shutting down the economy over regular flu. Maybe covid-19 is worse. It’s just that the comparisons that people make (like the one above) are contradictory to how we respond to other things that have significant tolls on the population. That comparison to 9/11 either justifies reacting the same way to regular flu or it justifies us reacting to wuflu the same way we react to regular flu (i.e. hardly at all). Otherwise, we need to address the disparity.
What does this same comparison look like for other countries (including our own)? In 2017/2018 we had 61,000 flu associated deaths according to the CDC. If this chart were to replicate here in the US then we should see (even if we use a low end of flu deaths of 22k) then we should have seen 44,000 WuFlu deaths in the US after a few weeks, according to the comparison you’re posting for Italy. As of 3/19 CDC is reporting only 10,442 total cases and only 150 total deaths. We’re not seeing what is being presented in that chart.
Maybe, but number tested (beyond some statistical minimum) is irrelevant. If only 100 people get tested then the rate should begin to assert itself; if the population of those being tested is skewed to serious cases then that means the rate is actually likely lower than reported.
Regarding criteria for including a case in the denominator (or in the numerator for that matter), are you saying that the criteria are too lax? Otherwise we might see some variance but I suspect that the criteria aren’t vastly different (yes that is speculation on my part). If we see outliers, then perhaps the problem is with the outlier’s criteria.
So the rate calculation isn’t perfect, but it’s not necessarily garbage either.
As far as raw numbers, if Italy were really the model then the US should already have seen tens of thousands of deaths (as opposed to 10,442 total cases and only 150 total deaths according to the CDC). Besides, if raw numbers of serious cases and deaths are a better indicator then our response to other viruses doesn’t make sense – or our current response doesn’t make sense.
I’m not locked into WuFlu being something to take lightly. I don’t take it lightly. It’s just that so far some arguments I see don’t make sense on their own terms according to the actual numbers we have. It’s also that I think the rate numbers, even imperfect, are a better indicator of risk whereas raw death numbers probably can never be an indicator of risk – they only show end effect.
Again, I disagree that the rates are unreliable, or, at least that they are unreliable in the sense that they are just as likely to increase as they are to decrease. I think the problem with the rate is that it’s likely the denominator is far too small – inflating the rate artificially.
Looking at hospitalizations as a data set is the definition of skewed isn’t it? What could that tell us about covid19 generally and the risk calculations we should make?
Apparently, it’s been showing some really good results, since they just approved it for use.
The point of that comment as that Italy is 1/5th the size of the US and has had as many deaths as the US did on 9/11.
And now add another 427 today. And more tomorrow, and the next day.
I guess you won’t be ready to act until we have what , 30,000 deaths? will that finally get your attention?
Again, step back a bit. Address what I’m actually saying. I didn’t say any of that. Obviously this all has my attention. The points remain – Italy appears to be an outlier. If we were to mirror Italy then we should have already seen tens of thousands of deaths. In fact we regularly see that many deaths – and we don’t do anything about it in comparison to what we’re doing now. Why not then? Why now? There is still a big range of reactions between doing nothing and shutting things down to the extent that we have. I suspect that somewhere in the middle would be a wiser reaction, but I could be wrong. I’m not callous to death and suffering, I don’t suggest we stick our heads in the sand. I just don’t understand why doing nothing in the face of tens of thousands of annual deaths is ho hum while suggesting that we slow down now is denialism.
FWIW, the Chinese government has approved chloroquine phosphate for treating future COVID outbreaks.
Not necessarily. Italy may have been seeded with the virus sooner and more heavily than the US. There are several times as many Chinese nationals per capita in Italy vs the US (due to One Belt One Road, among other things) and a lot of them are in northern Italy which has been particularly hard hit.
Variables include R0 (how many people does each case infect) and number of infected people. And how sick people get and how fast they get sick. You can affect R0 if you can identify carriers or likely carriers and isolate them as well as by isolating the patients. If you have enough patients, you may have to isolate them in mass settings.
The public health authorities must have draconian powers to deal with lethal pandemics or mass casualty situations. The problem is the temptation to apply these powers too soon.
The CDC estimates that in this flu season so far, in the US there have been 36-51 million flu illnesses which have resulted in 22,000 – 55,000 deaths. Let’s take the worst case combination, the top mortality figure and the smallest number of cases: 55K/36M. Rough mortality per case is 0.15% These are estimates, not confirmed cases.
Today’s John’s Hopkins numbers for Kung Flu in the US: a total of 13,159 confirmed cases and 176 confirmed deaths. That’s over 1.3%. Naturally, since there is not randomized testing in the US for coronavirus, this underestimates asymptomatic cases and individuals incubating the virus, destined to become ill but not yet symptomatic; these factors will tend to increase the denominator.
Comment from Aesop:
The implications of #1 are not yet clear. The implications of #2 depend on how universal COVID-19 testing is for pneumonia mortality in cases not yet tested for coronavirus.