According to MSNBC today, Oct23.09, the H1N1 body count in the US is “more than 1000.”
http://www.msnbc.msn.com/id/33449637/ns/health-cold_and_flu/

And here’s a headline from AP today: “US Swine Flu Deaths Surpass 1,000"
http://www.huffingtonpost.com/2009/10/23/us-swine-flu-deaths-surpa_n_332360.html

Where do these people get these numbers?? US surpassed 1000 deaths 5 weeks ago.

Today CDC reports death figures of 411 lab-confirmed and 2416 syndrome-based. To that you have to add the 593 they dropped on Aug30.09 when they reset the counter. Total = 3420. If you do the math, there were 506 new deaths reported by CDC for Wk 41. This weekly number has been pretty consistent for 5 weeks – average 493, range 405-572.

So the real body count to date is likely near 4000 because CDC reports a week behind.

Ok, Ok, 4000 really IS more than 1000, so they are technically right. But why would any reporter report more than 4x fewer deaths than actual?

1) Too lazy to get the right numbers, 2) Intentionally low-balling the numbers, or 3) ??...

The weekly death rate has plateau’d for the last 5 weeks. The fatality figures do not reflect the sharp rise in other indicators over this period, probably because 1) death comes after fever, etc. and 2) deaths likely take longer to report, so the fatality numbers will always lag other metrics.

BTW, could somebody look at that pediatric bar graph for Week 41 and tell me what these ding-dongs at the CDC are up to now.

You’ve got the top yellow part of the bar, which corresponds to “2009 Influenza A (H1N1) Deaths reported Current Week.” That’s OK.

But under that you’ve got a purple bar – “2009 Influenza A (H1N1) Deaths reported Previous Week.”

So, how did someone report in Week 40 pediatric deaths for Week 41?

Then you’ve got the blue bar at the bottom. In the past it has been used to show deaths during past weeks reported in the present week.

Looks like the bar for Wk 41, which should be all yellow, has been deleted.

2) Intentionally low-balling the numbers

has my vote

Sorry, RoRo, although you seem to have some of the best numbers going, I’m having more trouble following you than the CDC.

Confusion #1
“From August 30 to October 17 there were 411 confirmed and 2,416 probable deaths.”

These deaths weren’t “probable,” they were real, especially for the families. The diagnosis was what the CDC calls “symptomatic” or “syndromic-based,” which means not lab-confirmed. I think that’s what you mean by “probable deaths.”

Confusion #2
I don’t see how you connect the data on Table III of the MMWR to the mortality data on the weekly flu report. For week 41, MMWR has 684 and the flu weekly has 411 lab confirmed + 2416 syndrome-based = 2827. What connection do you see between the 684 and the 2827?

It appears that the weekly flu report is not limited to the 122 reporting sites that the MMWR is, but if you or someone else can point out where the CDC says otherwise, please do.

Clearly, the pediatric deaths are nationwide, not restricted to reporting sites.   

Confusion #3
“Prior to August 30 there were 593 confirmed, but CDC did not track probables. 411:2416 isabout 1:6. If we were to apply this same ratio to the pre August 30 confirmed fatalities, that suggests 3,000 additional probable fatalities prior to Aug 30.”

No, no, no – not close. In the early stages CDC did not distinguish between lab-confirmed and non lab-confirmed because they were lab-confirming virtually every suspicious death. Now they don’t confirm every death because they couldn’t keep up with 500 per week.

So you can’t take a ratio from Wk 41 and go back with it to the “early days” and say “Aha, there were really 3000 deaths, we gotta’ add those, too!!” No, no, no.   That would be like taking the ratio of hybrid to gas cars today and multiplying the ratio times the number of gas cars in 1960 to conclude there were millions of hybrids in 1960.

IslandCrow
Sure can’t argue with you that 593 + 411 = 1004, and that looks suspiciously like the 1000 number the media are latching on to.

But my guess is that no reporter has followed this thing closely enough to even realize that 593 figure is out there or what it means. Maybe the CDC is feeding them the total figure based on your calculation, which would only confirm what dolts they are.

I can’t agree with your argument that the current death figures include significant non-viral pneumonia deaths. Maybe in two months you’ll have a point. But if you look at the historic data and what the CDC has been saying all along, at this time of year anything that looks like a duck, walks like a duck, and quacks is H1N1. Or should I say, looks like a pig, walks like a pig, and oinks.

But it’s not strictly an either/or situation. In 1918, 80% of the “Spanish flu” deaths were actually due to bacterial pneumonia secondary to the flu, which should give us cause for hope because bacterial pneumonia is more easily controlled with antibiotics, which weren’t available in 1918.

Or alternatively syndromic based pediatric fatalities are not being reported as pediatric fatalities.

122 cities are 55 milion of people (my calculation with wikipedia). But 122 cities P&I mortality corresponds to about 60% of total mortality in USA.
In accordance with my calculations , from week 20 to week 39 there was about 1000 ( not 1500 as I previously wrote) deaths exceeding the expectation, expecially in the 45-64 age group)  So I argue about 1700 ( 1000 correspond to 60%) deaths were caused by flu in this period . But others deseases hide flu deaths, expecially cardiovascular one .
I observed that mortality in 45-64 age group is higher than expectation also  in 1-19 weeks

20-39 W	2009	2008	2007	2006	2005	2004	2003
TOT >65	134596	139835	134015	134852	139593	141546	139877
45-64	49602	50677	47815	47842	48145	47539	45103
25-44	13376	14276	14450	14543	14959	15288	15743
1-24	5005	5380	5464	5489	5729	5787	5843
0-1	4472	4860	4911	4631	4741	4638	4641
P&I  20-39	12986	13431	12176	12030	12894	13197	13042
P&I 31-39	5436	5517	5248	4879	5418	5749	5536
1-19 W
>65	151514	159020	149761	154995	163064	164743	151567
45-65	52049	52156	50153	50565	50399	49872	45521
P&I	16199	18793	15600	16821	19353	19184	16625

 

 

In 2006 year there were 56326 P&I total deaths ( http://www.cdc.gov/nchs/data/nvsr/nvsr57/nvsr57_14.pdf), in 122 cities they were 37377. Really it  corresponds to 66%.

But just adding up the CDC numbers alone and you get 3,420 deaths. That's nearly 3 and a half times the figure quoted in the press. Journo's seem to be clueless on this already.