Coronavirus Outbreak...

[oxmixmudd]

I would love it if someone could explain to me why we are seeing so many calls to restart the economy based on the claim that we have "flattened the curve." Although we have significantly bent the logarithmic curve for infections as well as deaths, the linear curve is still a straight line with a steep slope. We have seen an average of almost 30,000 new cases every day for the last two weeks with little sign of slow down. Yesterday was 37,725 new confirmed cases. During that same period we have seen an average of almost 2,400 deaths each day, again with little sign of abatement. Yesterday was 2,577. On 911, 2977 people lost their lives, not including the highjackers. This means for the last 14 days we have had over 80% of a 911-scale event every single day!

How on earth can anyone be arguing "we're ready to re-open the economy?"


ETA: BTW, in the worst flu season we see 339 deaths per day if they are averaged over the entire flu season, and about 1,017 per day if we consolidate the deaths to the two-month peak of the season. The current rate of deaths is 2.4 times that rate. The latest models I saw said 66,000 deaths by August. Unless there is a downturn soon, we'll hit that number in 7 days.

And yet you have breitbart (as posted my MM) claiming it's all over hyped as proven by noting that projections based on current rate of growth were always 'too high' as the social distancing and then shelter in place lowered that trend over time. And this sort of drivel is feeding a non trivial number of heads in this country. I have to deal with it constantly from friends that tend to watch fox over other outlets.

[carpedm9587]

And yet you have breitbart (as posted my MM) claiming it's all over hyped as proven by noting that projections based on current rate of growth were always 'too high' as the social distancing and then shelter in place lowered that trend over time. And this sort of drivel is feeding a non trivial number of heads in this country. I have to deal with it constantly from friends that tend to watch fox over other outlets.

But that can't BE! Fox News is NEUTRAL!

[Sparko]

I would love it if someone could explain to me why we are seeing so many calls to restart the economy based on the claim that we have "flattened the curve." Although we have significantly bent the logarithmic curve for infections as well as deaths, the linear curve is still a straight line with a steep slope. We have seen an average of almost 30,000 new cases every day for the last two weeks with little sign of slow down. Yesterday was 37,725 new confirmed cases. During that same period we have seen an average of almost 2,400 deaths each day, again with little sign of abatement. Yesterday was 2,577. On 911, 2977 people lost their lives, not including the highjackers. This means for the last 14 days we have had over 80% of a 911-scale event every single day!

How on earth can anyone be arguing "we're ready to re-open the economy?"


ETA: BTW, in the worst flu season we see 339 deaths per day if they are averaged over the entire flu season, and about 1,017 per day if we consolidate the deaths to the two-month peak of the season. The current rate of deaths is 2.4 times that rate. The latest models I saw said 66,000 deaths by August. Unless there is a downturn soon, we'll hit that number in 7 days.

Here is a chart of the daily cases in the US, from the Johns Hopkins map

daily.jpg
https://coronavirus.jhu.edu/map.html
So it looks like it has flattened out to me.

[carpedm9587]

Here is a chart of the daily cases in the US, from the Johns Hopkins map

[ATTACH=CONFIG]44198[/ATTACH]
https://coronavirus.jhu.edu/map.html
So it looks like it has flattened out to me.

That might explain part of it. I've been tracking the data from the worldometer map. I find the data a little better organized. Their numbers don't 100% align. But then again, I don't believe "flattening the curve" refers to the number of daily incidents, which is what this map shows. I think it refers to the logarithmic representation of the total cases.

Be that as it may, even your chart shows news cases per day to be in the upper 20s range for the last two weeks.

[Sparko]

That might explain part of it. I've been tracking the data from the worldometer map. I find the data a little better organized. Their numbers don't 100% align. But then again, I don't believe "flattening the curve" refers to the number of daily incidents, which is what this map shows. I think it refers to the logarithmic representation of the total cases.

Be that as it may, even your chart shows news cases per day to be in the upper 20s range for the last two weeks.

um total cases will always go up because it is cumulative. It is the sum of all of the new daily cases. It may bend as new daily cases slow down, but it can never go down because cumulative cases don't go away. So you have no bell curve like they are talking about flattening. The curve they are talking about flattening is active cases, or deaths (which lag by about two weeks).

[oxmixmudd]

um total cases will always go up because it is cumulative. It is the sum of all of the new daily cases. It may bend as new daily cases slow down, but it can never go down because cumulative cases don't go away. So you have no bell curve like they are talking about flattening. The curve they are talking about flattening is active cases, or deaths (which lag by about two weeks).

Flattening the curve can refer to several things, but in our context it typically refers to the peak load on hospitals. We want the case load spread out over time so that we don't overload our hospital capacity.

But that is virtually identical to flattening the logarithmic curve, because a log curve is whose slope trending closer to 0 indicates constant or declining growth in the number of cases. A log curve that looks like a diagonal line of positive slope, or worse, bending up, is very bad, as it means exponential growth or worse, which is going to create a massive peak demand for hospital care.

right now the curve is trending the right way, and we are actually seeing a slow trend toward declining growth in the number of active cases. But it is bumpy, going up sometimes, going down others and yielding a very shallow decline on average which will likely mean it will be a while longer before we can safely open up.

The other thing that a shallow decline can signal is some areas of the country that are declining, while others are rising. and this is the real problem we have in that many states did not shelter in place and most of them are seeing increasing growth in the number of new cases per day - the wrong direction, especially for considering a 'reopen' strategy. New York exploded and to the point it more than dominates the US picture. So what that means is that what is going on in the other 50 states is mostly hidden from view in the aggregate data for the entire US. To see what is actually happening in terms of potential resurgence of exponential growth we need to look at individual state data.

[carpedm9587]

um total cases will always go up because it is cumulative. It is the sum of all of the new daily cases. It may bend as new daily cases slow down, but it can never go down because cumulative cases don't go away. So you have no bell curve like they are talking about flattening. The curve they are talking about flattening is active cases, or deaths (which lag by about two weeks).

No - it will not always go up. If it is increasing exponentially, the logarithmic graph can be a straight line sloping upwards or it can actually bend upwards, depending on the degree of exponentiality. The linear graph will definitely bend upwards. If the growth is linear, the exponential curve will bend downwards and the linear graph will be a straight line with an upward slope. If the growth begins to taper off, both the exponential and linear graphs will bend downward. When there are no new cases, the graphs of both will be flat at the highest point.

Right now, the logarithmic graph is bending downwards, but is not flat. The linear graph is almost a straight line with a steep slope, so we are seeing linear growth and it is averaging around 24,000 cases per day. For U.S. infections, the logarithmic graph looks like this:

US_Infections_Logarithmic.PNG

The linear graph looks like this:

US_Infections_Linear.PNG

The little hook at the end is because I put in today's numbers already, but it is not the end of the day yet. The logarithmic graph can make us feel like we are really making progress, but basically what it tells us is whether we have managed to stop exponential growth. THAT we have managed to do. But the linear graph is what tells us we are by no means out of the woods yet. THAT curve is not flattening or even curving.

[carpedm9587]

Flattening the curve can refer to several things, but in our context it typically refers to the peak load on hospitals. We want the case load spread out over time so that we don't overload our hospital capacity.

But that is virtually identical to flattening the logarithmic curve, because a log curve is whose slope trending closer to 0 indicates constant or declining growth in the number of cases. A log curve that looks like a diagonal line of positive slope, or worse, bending up, is very bad, as it means exponential growth or worse, which is going to create a massive peak demand for hospital care.

right now the curve is trending the right way, and we are actually seeing a slow trend toward declining growth in the number of active cases. But it is bumpy, going up sometimes, going down others and yielding a very shallow decline on average which will likely mean it will be a while longer before we can safely open up.

The other thing that a shallow decline can signal is some areas of the country that are declining, while others are rising. and this is the real problem we have in that many states did not shelter in place and most of them are seeing increasing growth in the number of new cases per day - the wrong direction, especially for considering a 'reopen' strategy. New York exploded and to the point it more than dominates the US picture. So what that means is that what is going on in the other 50 states is mostly hidden from view in the aggregate data for the entire US. To see what is actually happening in terms of potential resurgence of exponential growth we need to look at individual state data.

Alas, I have not been tracking "recovered" for the U.S data, so I can't produce a graph of active cases. Worldwide, the active cases logarithmic graph is bending downwards, and the linear graph is beginning to show a downward bend.

[carpedm9587]

So the antibody testing in NYC is suggesting that 1 in 5 in the city may have had the virus at some point. That is about 3M people. SInce NYC has had 16,388 deaths, that would put the bottom edge of the mortality rate at 0.54% (about 5 times as deadly as the flu), and that number can only go up as the active cases resolve and people either die or recover.

I cannot tell how they selected their random sample, however. If they did what LA did, we again my have an inflated infection rate, making the mortality even higher.

[oxmixmudd]

So the antibody testing in NYC is suggesting that 1 in 5 in the city may have had the virus at some point. That is about 3M people. SInce NYC has had 16,388 deaths, that would put the bottom edge of the mortality rate at 0.54% (about 5 times as deadly as the flu), and that number can only go up as the active cases resolve and people either die or recover.

I cannot tell how they selected their random sample, however. If they did what LA did, we again my have an inflated infection rate, making the mortality even higher.

Mortality for diagnosed cases - and this is simply a measurement - is running at 5% in Germany where 100,000+ have recovered and >5000 have died, with 50,000 left to recover and 3000 serious to critical. That - unfortunately - implies that of the 225,000 cases remaining in NYC, with a naive mortality of 7.5% (we won't even consider worst case) we will likely see somewhere between 12,000 and 18,000 additional deaths putting the total at or near 35,000 to 40,000. or a real mortality between 1 and 1.3%. These are the numbers that keep coming up, low is usually higher than your .5%, (like .7%) and the high rarely goes above 2%.

[carpedm9587]

Mortality for diagnosed cases - and this is simply a measurement - is running at 5% in Germany where 100,000+ have recovered and >5000 have died, with 50,000 left to recover and 3000 serious to critical. That - unfortunately - implies that of the 225,000 cases remaining in NYC, with a naive mortality of 7.5% (we won't even consider worst case) we will likely see somewhere between 12,000 and 18,000 additional deaths putting the total at or near 35,000 to 40,000. or a real mortality between 1 and 1.3%. These are the numbers that keep coming up, low is usually higher than your .5%, (like .7%) and the high rarely goes above 2%.

Yeah - that's why I noted that the 0.54% was a floor - because we have so many active cases remaining. It's not a pretty situation.

[little_monkey]

Studies coming from S. Korea and China are showing that those who recovered from the coronavirus have permanent damage not only to their lungs but also to their heart, kidney and liver. In some cases up to a 50% of functional reduction. These people will have health issues for the rest of their lives. For those who think this is just another flu virus, think again.

[Starlight]

The linear graph looks like this:

[ATTACH=CONFIG]44201[/ATTACH]

The little hook at the end is because I put in today's numbers already, but it is not the end of the day yet. The logarithmic graph can make us feel like we are really making progress, but basically what it tells us is whether we have managed to stop exponential growth. THAT we have managed to do. But the linear graph is what tells us we are by no means out of the woods yet. THAT curve is not flattening or even curving.

That linear graph is a worrying one. Someone asked me yesterday what the total US deaths might be by the end of August, and I looked at that linear graph and said, "well if that continues it works out at about 500k dead in the US by the end of August". Hopefully I made a math boo-boo because that's a scary number.

[carpedm9587]

That linear graph is a worrying one. Someone asked me yesterday what the total US deaths might be by the end of August, and I looked at that linear graph and said, "well if that continues it works out at about 500k dead in the US by the end of August". Hopefully I made a math boo-boo because that's a scary number.

The key is "if that continues." Hopefully, the bend in the logarithmic graph will be echoed by a similar bend in the linear one. We're just not seeing it yet.

[oxmixmudd]

Studies coming from S. Korea and China are showing that those who recovered from the coronavirus have permanent damage not only to their lungs but also to their heart, kidney and liver. In some cases up to a 50% of functional reduction. These people will have health issues for the rest of their lives. For those who think this is just another flu virus, think again.

link? all cases, or just those that had serious complications?