Death Rates by Population

[JimO]

Some of the state death rates based on population are close to or above 1 death per million, which is equal to Mexico's. Here are my numbers (this includes FT's confirmed deaths plus suspected deaths).

Numbers are expressed as population/deaths/deaths per million

Mexico 108,700,891/113/1.04

Arizona 6,500,180/7/1.08

Connecticut 3,501,252/3/0.86

Illinois 12,901,563/10/0.78

New York 19,490,297/30/1.54

Utah 2,736,424/8/2.92

Mexico's death rate based on reported cases is 1.8%. I realize that their case reporting is biased toward hospitalized cases, but there are also indications that many more died in Mexico and were never tested. Let's assume an 8% hospitalization rate and that all of Mexico's cases were hospitalized cases (meaning that for every 8 cases reported, there were 92 "mild" cases not tested). If the infection rate is one in three, then total deaths would exceed 50,000. In the U.S. it would be 3x that even without any mutation of the virus. Here's my equation:

Potential Mexican deaths = 0.018 x 0.08 x 0.33 x 108,70,891 = 52,176

Potential U.S. deaths = 52,176/108,700,891 x 308,013,761 = 147,845

In other words, if this virus becomes more widespread or returns even unchanged in another wave or two over the next year, and we assume an overall infection rate of 1 in 3 people, we could lose over 4x the number we lose annually with the seasonal flu. And, most of those deaths would be relatively young and healthy people.

Please feel free to check my math, but if this is true, then I don't want to hear "this flu is no different than your garden variety of seasonal flu" quoted in the media anymore.

 

[AlaskaDenise]

Re: Death Rates by Population

Another factor, which we cannot count yet, but will influence even these more indicative numbers is.....whether the virus has completed its spread within a given area. I would expect that to take as long as 6 months (with a rerun in another 6 months), so we can only strive for a better estimated death/million count.

Thanks for compiling the stats.

.

 

[Snicklefritz]

Re: Death Rates by Population

Please feel free to check my math, but if this is true, then I don't want to hear "this flu is no different than your garden variety of seasonal flu" quoted in the media anymore.

JimO, welcome to Flutrackers. I want to give you big kudos for taking the time to do the math and present your ideas in a very clear way. I've been a long-time 'Flubie' and, believe me, I've played with numbers on H5N1 and with this novel H1N1 till I see them in my sleep. So, please don't take the following as a put down. But, I have questions about the 'certainty' of your final answers based upon the following observations. These observations are based upon my understanding of calculatiing probabilities, and not on your actual numbers. My understanding of probabilities may be flawed, and if anyone on this forum can correct my understanding, I'd be happy to be corrected.

In your calculations, each of the numbers you have used has a probability of being correct. The number of confirmed cases has a certain probability of representing an actual number of confirmed cases. And, the number of deaths has a probability of representing the actual number of deaths attributable to this virus, and so on through each of the numbers you use, including population figures which are also estimates.

When you have multiple variables, each with only a probability of being a correct number, the total probability of the final number being correct is the product of each individual probability.

If the probability of 'A' being correct is 90%,
And, the probability of 'B' being correct is 90%,
And, the probability of 'C' being correct is 90%,
Then the probability of A, and B, and C, is the product of the probabilities of each. And is lower than the probability for each.

.90 x .90 x .90 = .729

Likewise:

If the probability of 'A' being correct is 50%,
And, the probability of 'B' being correct is 50%,
And, the probability of 'C' being correct is 50%,
Then the probability of A, and B, and C, is the product of the probabilities of each. And is lower than the probability for each.

.50 x.50 x .50 = .125

Another way to look at this is that the 'uncertainty' is the reciprocal of the 'probability'. If the probability is 75%, the uncertainty is 25%.

So, as the number of factors grows, each with it's reciprocal uncertainty, the uncertainty grows!

Now, let's assign a probability for the accuracy of each of the numbers you have used for Mexico, and derive from that the uncertainty of accuracy of your final answer.

Confirmed cases = (as of now) 7083. What is the probability that that represents the true number of cases?

Number of deaths = 113. What is the probability that that represents the true number of deaths due to this virus.

CFR (based on reported figures) is 1.6% . What is the probability that the that the number of deaths predicted is correct? Let's say the figures are 70% accurate for both confirmed cases and deaths.

.70 x .70 = .49

In this case the uncertainty is greater (51%) than the certainty (49%).

Taking the uncertain CFR (49% = 1.6%) times a third of the population 33% (assumed attack rate for a novel virus or .33), and assume the population figure is completely accurate (giving a probabilty of accuracy of 1.0) the probability is .70 x .70 x.1.0 = .49 certain. But, if the probability of the accuracy of the population figure is 90% (.90) then the total probability for an accurate solution is .70 x .70 x .90 = .44, and the uncertainty is now .56 or 56%.

So, the total deaths is plus or minus 56%.

If the numbers are 100% accurate you get 1.6% of 33% of 108,700,791 (reported population of Mexico) = 573,940 fatalities.

The uncertainty is 56%. So, the margin of error is 56% of 573,940, or 321,406 plus/minus. The range then is 252,535 to 895,345 fatalities.

That's not a useful range of numbers.

My point is not that computations of this sort are incorrect, they actually could be accurate, but that they are full of uncertainty.

I hope, JimO, that you don't take this personally, and as I said, if others who are more astute in mathematics can correct my thinking, then so be it -for the good of all.

Peace and Health - Snick

 

[The Mountains Voice]

Re: Death Rates by Population

The range then is 252,535 to 895,345 fatalities.

That's not a useful range of numbers.

It is VERY useful to me... says, "Get The Heck Prepared!!!"
It says, "A Heck of a Lot of People Are Going to DIE!"

The bottom line is that ALL numbers are suspect. In much of this you can barely even get close to a reasonable estimate without having a lot of disagreement, even from experts & friends here on FluTrackers. What we HAVE been able to do is agree upn "General Ranges" and "Directional Trends" and "Historical Comparisons". Sometimes, in the absence of perfect date - Ya Just Gotta Wing It As Best You Can!!!

This isn't a science / math contest. It's a game of reality - one of life and death. The only right or wrong answer is the one that saves you - or gets you dead.

PS.... and that's not a personal comment directed at you either Snick. Just throwing in my two cents / perspective more than anything.

Peace

 

[AlaskaDenise]

Re: Death Rates by Population

It says, "A Heck of a Lot of People Are Going to DIE!"

from WHO regarding 1957 and 1968 moderate pandemics:

Subsequent pandemics were much milder, with an estimated 2 million deaths in 1957 and 1 million deaths in 1968

.

 

[Snicklefritz]

Re: Death Rates by Population

"What we HAVE been able to do is agree upn "General Ranges" and "Directional Trends" and "Historical Comparisons".

No offense taken, MV, but please don't include me in 'we'. I'm more circumspect, and I suspect there are others as well.

:tiphat:

 

[JimO]

Re: Death Rates by Population

I hope, JimO, that you don't take this personally, and as I said, if others who are more astute in mathematics can correct my thinking, then so be it -for the good of all.

Peace and Health - Snick

Not at all. I appreciate the response. I have a co-worker who is a statistician who would knock me in the head if he saw my post. Fortunately, he's not interested in epidemics. It took me a couple of readings, but I do understand what you are saying.

My concern with the way this flu is being portrayed by the CDC, WHO, and other government agencies, both in the U.S. and abroad, is that they keep saying that this is no worse than a typical seasonal flu. I still keep hearing reports that the high risk groups are considered the very young and the elderly, and yet, based on the data, those groups are at the ends of the bell curve with respect to numbers of infections.

Based on what we are seeing, IMHO (as an amatuer), I don't think the virus has to mutate into a more dangerous form to kill far more people than the public expects. Even if the death rate was no worse than the seasonal flu, the demographics of the deaths will likely be very different and a lot of young, healthy people, who thought they would live long lives, will die.

I really hope I'm wrong. I have four sons smack in the middle of the age group at the top of the curve.

 

[Snicklefritz]

Re: Death Rates by Population

Based on what we are seeing, IMHO (as an amatuer), I don't think the virus has to mutate into a more dangerous form to kill far more people than the public expects. Even if the death rate was no worse than the seasonal flu, the demographics of the deaths will likely be very different and a lot of young, healthy people, who thought they would live long lives, will die.

Thanks, JimO. As another amateur, I agree that a pandemic of seasonal flu virulence will cause a lot of pain, simply because of naivete of the population. But, I believe that WHO and CDC have been pretty straightforward on the uncertainty of how this will all evolve in time. And, their actions speak louder than their words - in my opinion.

But, reasonable people can disagree, reasonably. Keep up the good thinking :applause:

 

[Laidback Al]

Re: Death Rates by Population

Welcome to FluTrackers and thanks for the discussion, JimO.

One thing your ratios shows, but was not discussed, is that Utah has the highest death rate in the USA. It has the lowest population of the five states and Mexico. As I noted in the another thread, Utah has lowest median (mean?) age in the USA, a lot of children and young people.

I have not looked in detail at the ages in Utah, but it is interesting (even with cytokine storm) that most of the deaths are older individuals rather than children. The age breakdown of deaths in Utah, and around the the world as whole, deserve more epidemiological analysis.

 

[hongkonger]

Re: Death Rates by Population

It would be interesting to add Canada to the comparison of death rates by population. I know Canada's not exactly a state ...

 

[JimO]

Re: Death Rates by Population

It would be interesting to add Canada to the comparison of death rates by population. I know Canada's not exactly a state ...

I don't have the data from Canada broken down by Province yet, but the death rate based on population is 0.36 per million, with 13 deaths and a population of 33,390,141. This tracks closely with the overall U.S. death rate of 0.37 per million. I suspect as in the U.S., the provincial death rates will vary such that where the flu has spread widely, such as Quebec, it exceeds 1 per million residents.

I'll check and report back shortly.

 

[JimO]

Re: Death Rates by Population

After some quick searches, here are the deaths rates/million of Quebec and Manitoba, the two hardest hit provinces:

Quebec - 0.92 deaths per million (7 deaths/7.52M)
Manitoba - 1.79 deaths per million (2 deaths/1.12M)

I realize that this is very limited because the number of deaths are not statistically significant, but it might be worth tracking.

 

[JimO]

Re: Death Rates by Population

Here are some population death rates in deaths per million citizens:

Highest U.S. States

Arizona - 1.38
Connecticut - 2.00
Illinois - 1.01
New York - 2.21
Utah - 4.02

Highest Countries

Argentina - 0.94
Canada - 0.90
Chile - 0.98
Mexico - 1.09
United States - 0.52

 

[JimO]

Re: Death Rates by Population

Here are the population-based death rates (per million) for the U.S., rounded to the nearest whole number.

Alabama 7
Alaska 9
Arizona 13
Arkansas 5
California 10
Colorado 7
Connecticut 4
Delaware 5
Florida 8
Georgia 4
Hawaii 9
Idaho 8
Illinois 3
Indiana 4
Iowa 5
Kansas 6
Kentucky 5
Louisiana 8
Maine 2
Maryland 3
Massachusetts 2
Michigan 4
Minnesota 3
Mississippi 5
Missouri 1
Montana 10
Nebraska 7
Nevada 11
New Hampshire 1
New Jersey 2
New Mexico 14
New York 5
North Carolina 5
North Dakota 2
Ohio 2
Oklahoma 6
Oregon 11
Pennsylvania 2
Rhode Island 7
South Carolina 6
South Dakota 17
Tennessee 6
Texas 6
Utah 12
Vermont 2
Virginia 3
Washington 7
West Virginia 4
Wisconsin 4
Wyoming 15
Guam 11
Puerto Rico 10
Virgin Islands 10

Total U.S. 5.8

 

[alert]

Re: Death Rates by Population

0.01% is one in every 10,000, which is 100 per million.

 

[JimO]

Re: Death Rates by Population

0.01% is one in every 10,000, which is 100 per million.

Thanks. I just deleted my attempt at math in public and just left the data. It's a good thing I don't work for NASA.

 

[kiwibird]

Re: Death Rates by Population

In discussion with others - I compiled my own little version. Feel free to correct:

Fatalities by State;

Alabama 1/155,396
Alaska 1/114,382 (Mainly around Fairbanks which has a huge nitrate and arsenic problem)
Arizona 1/100,002
California 1/120,513
Florida 1/122,188
Kentucky 1/304,946
Michigan 1/434,931
New York 1/209,573 (which has already had round one of the flu and is on the second)
Pennsylvania 1/691,571
Texas 1/211,538

Of course a breakdown of these figures by county gives a much clearer picture. With Texas
Hidalgo County 1/35,086
Harris 1/212,536
Bexar 1/162288
Nueces County 1/39,205 Nitrate and Arsenic problems
El Paso 1/94,385 (Which has 4 times the national rate of hep A because of the insanitary conditions - not running water in many areas - water will certainly be contaminated)
All these counties have nitrate problems

Oregon
Umatilla County 1/23,516 (has a nitrate and perchlorate problem)
Lane County 1/69065
Deschutes 1/57683
There are 3 areas with nitrate in Oregon. One in Umatilla, one in the Willametta Valley which flows through Deschutes County, Klamath and part of Lane County.

California
Orange County 1/130511

Florida
Pinellas County 1/92148 (very old septic tanks leaching)
Duval County 1/63572
Polk County 1/112321

Figures taken from 2008 census, where available and cross referenced with Flutrackers tallies.

 

[Zac]

Re: Death Rates by Population

Here are the population-based death rates (per million) for the U.S., rounded to the nearest whole number.

Total U.S. 5.8

I thought CDC was now reporting that there were about 4000 deaths from swine flu.

With a US population of about 308 million, wouldn't that be about 13 deaths per million?

Is there a spreadsheet with the hospitalizations and fatalities being maintained somewhere? I recall seeing one that was set up on google apps, but don't remember where it was.

Thanks for all your efforts.