COVID deaths still declining

[mikewhitney]

They have the next flu season to look forward to.

Well. There is much effort to promote a second breakout of coronavirus to keep Eeyore happy during the summer.

[oxmixmudd]

Eeyore just never could be happy. Maybe you can find something else to make you happy when the coronaphobia season is over.

Mike - I'm going to take resorting to this sort of complaining and ad hom as a concession to the fact you've lost the scientific debate with me and move on.

[mikewhitney]

Mike - I'm going to take resorting to this sort of complaining and ad hom as a concession to the fact you've lost the scientific debate with me and move on.

I'm just irritated that you are falling for untrustworthy news and statistics. There is nothing left but humor to respond to what you are presenting. Though I should add, you are doing your task of promoting coronaphobia very well. It just does not work well for people who are informed of the facts.

[little_monkey]

Depends on the type of mask and if it is properly used. The overwhelming majority of masks in use in the public are the wrong type and won't do a thing to protect you or anybody else from the China flu, and the overwhelming majority of people who wear a mask do so in such a way that it confers zero benefits to the wearer anyway (leaving one's nose uncovered, touching the mask, removing the mask to talk on the phone, reusing the same mask throughout the day or over multiple days, etc.).

So, no, wearing a mask is in no way analogous to wearing a seat belt which is effectively foolproof to use and has been proven through decades of rigorous repeated testing.

The data shows otherwise. In one hospital using no mask had two and half times more cases among the staff than in another hospitable which rigorously used a mask among its staff. The data is indisputable - it has no respect for your opinion based on thin air.

[Ronson]

That graph doesn't tell the story. In fact, it is not what you say it is. It is the recovery RATE and the death RATE (currently 7.85%*), not the raw count, or even the daily count, nor does it show the daily trend.

Well, yes - the percentage rate, which is shown on the left variable.

This one, however, does: (same source - worldometers)

[ATTACH=CONFIG]46712[/ATTACH]

There does seem to have been a hiccup from the boneheaded switch in reporting destinations, in that the peak has typically been Tuesday for a given week, but this week Wednesday was higher than Tuesday. But BOTH where higher than the week previous, which was higher than the week previous to that.

The highest (1001) number of deaths since June 9 in fact.

The bottom line is that deaths are rising, just delayed relative to the rise in cases, as expected.

They must be rising in concentrated areas again where facilities are overburdened. Because the reports I've read indicate treatments are better across the board; fewer ventilators are required now versus the first wave.

I am concerned about the shift to a non-public database for housing the data. The administration is not above massaging the data for their own purposes. They should have left the system alone and let the CDC continue to be the primary reporting destination where doctors and scientists make the data public for other researchers and are not subject to the political goals of the current administration.

Both stink. The CDC admits that there is no methodology in place for the data being supplied, where a guess is as good as a test.

[Ronson]

They must be rising in concentrated areas again where facilities are overburdened. Because the reports I've read indicate treatments are better across the board; fewer ventilators are required now versus the first wave.

https://www.telegraph.co.uk/news/202...ew-cases-soar/

Why coronavirus deaths remain low in the US despite surge in new cases

Advances in a number of treatments appears to have contained the death rate as records continue to be broken for new infections

[oxmixmudd]

https://www.telegraph.co.uk/news/202...ew-cases-soar/

Why coronavirus deaths remain low in the US despite surge in new cases

Advances in a number of treatments appears to have contained the death rate as records continue to be broken for new infections

There is no doubt treatment has improved and can be credited with helping reduce overall death rates. The point is that with record surges in cases cause by unconstrained spread, the daily death count (not rate) will go up. That means more people will be dying. Right now its heading to 1000/day. Current mortality for cases with outcome rate for diagnosed patients is a bit less that 8%, but even if that gets cut to 4% or even 3%, with better treatment, that would mean with 70000 new cases /day, we are going to see 2100 deaths/day in about 3 to 4 weeks.

[Sparko]

There is no doubt treatment has improved and can be credited with helping reduce overall death rates. The point is that with record surges in cases cause by unconstrained spread, the daily death count (not rate) will go up. That means more people will be dying. Right now its heading to 1000/day. Current mortality for cases with outcome rate for diagnosed patients is a bit less that 8%, but even if that gets cut to 4% or even 3%, with better treatment, that would mean with 70000 new cases /day, we are going to see 2100 deaths/day in about 3 to 4 weeks.

Your latest graph shows the rise leveling off.

[oxmixmudd]

Your latest graph shows the rise leveling off.

Perhaps. This week we saw a change in reporting destination for the data, which also resulted in a change in the data pattern. Instead of a tuesday spike followed by a decline wed/thursday, we say nearly equal load on Tuesday/Wednesday/Thursday at nearly 1000 cases per day. So if we compare T/W/Th the week before with T/W/Th this week we see:

993+891+961 vs 936+1002+963 or a net rise of 2845 -> 2901. Not a huge amount, but still a rise. The 7 day average rose from 741 to 761 week to week.

The part that concerns me the most is the change in reporting destination and the removal of public access to the data. It will be important to test individual state reporting websites against the federal numbers going forward. Of course, we already know Florida was fudging their data:

https://www.floridatoday.com/story/n...te/5212398002/

The clear danger here is that with Trump flailing in the polls and the coronavirus response of the admin being so poor, they and his state level minions will be tempted to change reporting in ways that make things look better - as we saw in Florida. Without public access to the data, it will be easier to do that.

Fortunately, worldometers at least gathers their data from state level websites.

[Reepicheep]

A report produced for the White House Coronavirus Task Force recommends that eighteen states should roll back their reopening measures because infection rates are increasing. The states are:

Alabama
Arkansas
Arizona
California
Florida
Georgia
Iowa
Idaho
Kansas
Louisiana
Mississippi
North Carolina
Nevada
Oklahoma
South Carolina
Tennessee
Texas
Utah

In the above list, states which went Republican in 2016 are in red and states which went Democrat are in blue. It is interesting that there seems to be a strong correlation between a state's inability to get the coronavirus pandemic under control and support for Donald Trump.

[Cow Poke]

In the above list, states which went Republican in 2016 are in red and states which went Democrat are in blue. It is interesting that there seems to be a strong correlation between a state's inability to get the coronavirus pandemic under control and support for Donald Trump.

Interesting that the inverse seems to be true for riots and mayhem and chaos.

[mikewhitney]

A report produced for the White House Coronavirus Task Force recommends that eighteen states should roll back their reopening measures because infection rates are increasing. The states are:

Alabama
Arkansas
Arizona
California
Florida
Georgia
Iowa
Idaho
Kansas
Louisiana
Mississippi
North Carolina
Nevada
Oklahoma
South Carolina
Tennessee
Texas
Utah

In the above list, states which went Republican in 2016 are in red and states which went Democrat are in blue. It is interesting that there seems to be a strong correlation between a state's inability to get the coronavirus pandemic under control and support for Donald Trump.

We have to remember that this just means that they are testing a lot more people and that the tests "indicate" a positive result. We also have places in Florida, for example, where all the test results were positive. This means that fraud can be involved. The number of "cases" is an big unscientific number with very little value.

[DivineOb]

We have to remember that this just means that they are testing a lot more people and that the tests "indicate" a positive result. We also have places in Florida, for example, where all the test results were positive. This means that fraud can be involved. The number of "cases" is an big unscientific number with very little value.

That's why I think we should focus on hospitalizations. That is a totally unambiguous datapoint and one which is impossible to artificially inflate. The hospitalization data is grim.

[Mountain Man]

The data shows otherwise. In one hospital using no mask had two and half times more cases among the staff than in another hospitable which rigorously used a mask among its staff. The data is indisputable - it has no respect for your opinion based on thin air.

Yes, masks of the proper type and used correctly do offer some benefit in situations where you are in extended, close contact with an infected person, which would be the case in a hospital. A mask offers virtually no benefit in casual day to day activities such as walking through a grocery store.

[mikewhitney]

That's why I think we should focus on hospitalizations. That is a totally unambiguous datapoint and one which is impossible to artificially inflate. The hospitalization data is grim.

We seem to have fraud there to. This apparently was the case in NYC and some other localities where any minor flu-symptom was deemed a covid-19 case. Plus, afaik there still is a bigger payment by medicare for covid-19 cases. The Dallas County data seems decent since they show no presumed covid-19 but only counts for patients who were tested, at least for those who died. I also liked some details they provided on the percentage of deaths by ethnic groups which also included the percentage of those ethnic groups in the population of the county. This showed good correlation between the proportion of deaths and the population -- there was no big difference due to ethnicity.