Graph - Cumulative Count of CDC confirmed S-OI A/H1N1 Cases in the USA, May 6, 2009

[Laidback Al]

Graph - Cumulative Count of CDC onfirmed S-OI A/H1N1 Cases in the USA, May 5, 2009

The following graph is based on the update of confirmed cased provided by the CDC on May 5, 2009 at: http://www.cdc.gov/h1n1flu/ and from past updates provided by the CDC at that same url.

It is important to note that this graph is still showing an exponential increase in the number of confirmed cases in the USA. It does not indicate that we have yet reached the peak and that the counts are decelerating.

[gsgs]

Re: Graph - Cumulative Count of CDC confirmed S-OI A/H1N1 Cases in the USA, May 5, 2009

the black curve is well above the red bars

[Laidback Al]

Re: Graph - Cumulative Count of CDC confirmed S-OI A/H1N1 Cases in the USA, May 5, 2009

the black curve is well above the red bars

The trend is still exponential rather than linear or declining. Tommorrow's numbers from the CDC may show that the infection rate has decelerated. If there is a deceleration it will show up in the graph but will be lagging.

[Laidback Al]

Re: Graph - Cumulative Count of CDC confirmed S-OI A/H1N1 Cases in the USA, May 5, 2009

Graph - Cumulative Count of CDC confirmed S-OI A/H1N1 Cases in the USA, May 5, 2009

The following graph is based on the update of confirmed cased provided by the CDC on May 5, 2009 at: http://www.cdc.gov/h1n1flu/ and from past updates provided by the CDC at that same url.

It is important to note that this graph is still showing an exponential increase in the number of confirmed cases in the USA. It does not indicate that we have yet reached the peak and that the counts are decelerating.

The graph based on today's official report is still showing an upward trend.

[cartski]

Re: Graph - Cumulative Count of CDC confirmed S-OI A/H1N1 Cases in the USA, May 6, 2009 (see last post)

Sorry, but are these data by:

- the date the test confirmed H1N1
- the date the sample was taken
- the date the person became sick?

Due the backlog of test samples, we would expect a daily increase if the graph is drawn based on the date the case is confirmed. But we really don't know if things are getting better or worse unless the sample is tracked back in time and the graph is restated.

[ Sorry, I know this should be a discussion, rather than news, but I find it very hard to make that happen.]

J.

[Laidback Al]

Re: Graph - Cumulative Count of CDC confirmed S-OI A/H1N1 Cases in the USA, May 6, 2009 (see last post)

Sorry, but are these data by:

- the date the test confirmed H1N1
- the date the sample was taken
- the date the person became sick?

Due the backlog of test samples, we would expect a daily increase if the graph is drawn based on the date the case is confirmed. But we really don't know if things are getting better or worse unless the sample is tracked back in time and the graph is restated.

[ Sorry, I know this should be a discussion, rather than news, but I find it very hard to make that happen.]

J.

Cartski, those are good questions. The data is based strictly on the numbers reported by the CDC on a daily basis. I believe that the numbers are tabulated based on the number of cases that the CDC confirms in the preceding 24 hour period. I am not aware of any publicly accessible database that has individual demographic or epidemiological data on individuals cases such as date of onset, age, sex, etc. We have to rely on the daily CDC briefings to interpret these numbers for us.

You can see there is a steady geometric progression in the numbers. At the beginning of the outbreak, it is understandable that there may have been a backlog of samples for testing. But we are now more than two weeks into this outbreak. If this is as serious as it seems to be, I would think that the CDC would have thrown all available resources into this and ramped up its personnel and testing activities to clear up any back logs. So, if they are still having a back log of samples to be tested, I can only conclude that the numbers of suspected cases are still rising across the USA.

If we have reached an infection peak or plateau, the numbers, no matter how they are measured, should no longer be rising.

[Pert]

Re: Graph - Cumulative Count of CDC confirmed S-OI A/H1N1 Cases in the USA, May 6, 2009

If these graphs existed outside this website, there would be no need to post them here.

I agree that WHO and CDC data should be regarded as a lower bound on the growth curve, as delayed/back reporting would only tend to raise the least-squares fit.

Predictions about when the curve will more closely approaches a logistic curve are welcome.

I projected forward 7 days from now and created this graph below. Discussion is welcome about the implications of whether WHO or CDC data fall above or below the present predicted curve.

[Laidback Al]

Re: Graph - Cumulative Count of CDC confirmed S-OI A/H1N1 Cases in the USA, May 6, 2009

Pert - Welcome to FluTrackers and thanks for posting the graphs.

The graphs you posted represent a hypothesis that the infection rate is still growing in both Mexico and the USA. Within a week we will see how closely the actual number of infections coincide with your projections.

[gsgs]

Re: Graph - Cumulative Count of CDC confirmed S-OI A/H1N1 Cases in the USA, May 6, 2009

it's also worth to look at Spain separately, that's presumably where WHO
decides to go to 6.

And world-North-America

(at least weekly)

[cartski]

Re: Graph - Cumulative Count of CDC confirmed S-OI A/H1N1 Cases in the USA, May 6, 2009 (see last post)

If this is as serious as it seems to be, I would think that the CDC would have thrown all available resources into this and ramped up its personnel and testing activities to clear up any back logs. So, if they are still having a back log of samples to be tested, I can only conclude that the numbers of suspected cases are still rising across the USA.

If we have reached an infection peak or plateau, the numbers, no matter how they are measured, should no longer be rising.

Thanks. But I don't follow this. The 35,000 testing backlog was posted when, late last week? And there was a capacity post of 100 test per day - not sure if that was CDC, Winnipeg, or local labs. Even at 300 tests per day, and lots of overtime, that's a lot of days before they're up to date. So in reality we may be at an infection peak, but by measurement, it may seem that the cases are still rising. There must be a significant lag.

Not that I believe that we are at an infection peak. I think Dr. N. has a better estimate, that there are 200k infections in Canada that are not reported, etc. Friends back from the Boston Marathon are sick; overheard conversations at coffee shops are about being bed-ridden with sickness; a minimum of 70:1 contact ratio, etc. And here in Winnipeg, the West Nile capital of North America, we are quite familiar with the fact that we all have antibodies, but are not clinically sick. The titer of the virus may be low, but it is pervasive.

J.

[cartski]

Re: Graph - Cumulative Count of CDC confirmed S-OI A/H1N1 Cases in the USA, May 6, 2009

But the figures in the latest MMWR report are "date of illness onset":

(sorry, don't know how to post the actual pic)

http://www.cdc.gov/mmwr/preview/mmwrhtml/mm5817a1.htm?s_cid=mm5817a1_e

FIGURE 1. Number of confirmed (N = 822) and suspected (N = 11,356) cases of novel influenza A (H1N1) virus infection, by date of illness onset --- Mexico, March 11--May 3, 2009

[Laidback Al]

Re: Graph - Cumulative Count of CDC confirmed S-OI A/H1N1 Cases in the USA, May 6, 2009 (see last post)

Thanks. But I don't follow this. The 35,000 testing backlog was posted when, late last week? And there was a capacity post of 100 test per day - not sure if that was CDC, Winnipeg, or local labs. Even at 300 tests per day, and lots of overtime, that's a lot of days before they're up to date. So in reality we may be at an infection peak, but by measurement, it may seem that the cases are still rising. There must be a significant lag.

Not that I believe that we are at an infection peak. I think Dr. N. has a better estimate, that there are 200k infections in Canada that are not reported, etc. Friends back from the Boston Marathon are sick; overheard conversations at coffee shops are about being bed-ridden with sickness; a minimum of 70:1 contact ratio, etc. And here in Winnipeg, the West Nile capital of North America, we are quite familiar with the fact that we all have antibodies, but are not clinically sick. The titer of the virus may be low, but it is pervasive.

J.

We have to turn to somewhere to get an "official" count. Anecdotal accounts of friends and relatives doesn't get the job done. Niman's estimates are most likely realistic, but don't count as "official" information.

I am more dismayed with the progress that is being made. I would hope that 2 weeks into a pandemic the CDC would be better staffed and prepared to eliminate the backlog in short order.

Isn't this the event that the CDC has been preparing for since 1976? When are we going to get some realistic numbers, weeks or months after it is over?

[Laidback Al]

Re: Graph - Cumulative Count of CDC confirmed S-OI A/H1N1 Cases in the USA, May 6, 2009

But the figures in the latest MMWR report are "date of illness onset":

(sorry, don't know how to post the actual pic)

http://www.cdc.gov/mmwr/preview/mmwrhtml/mm5817a1.htm?s_cid=mm5817a1_e

FIGURE 1. Number of confirmed (N = 822) and suspected (N = 11,356) cases of novel influenza A (H1N1) virus infection, by date of illness onset --- Mexico, March 11--May 3, 2009

Cartski -

If you believe their results, the graphs are reassuring because they show that there is fall off in cases based on the dates of onset. So maybe the worst is over.

This, however, can not be true if your claim that there are still thousands of backlogged samples that haven't been processed. In that case, the only thing those graphs are showing (without the backlogged samples being processed) is that they haven't finished processing samples with more recent onset dates.

[cartski]

Re: Graph - Cumulative Count of CDC confirmed S-OI A/H1N1 Cases in the USA, May 6, 2009

Cartski -

This, however, can not be true if your claim that there are still thousands of backlogged samples that haven't been processed. In that case, the only thing those graphs are showing (without the backlogged samples being processed) is that they haven't finished processing samples with more recent onset dates.

Unless there is something I've missed (likely) about the number of backlogged samples, I agree that the graphs show that they haven't finished processing. But it's not necessarily the case that the samples are from more recent onset dates - the samples could be older too. CDC may sort the samples by onset date, but the sending lab/country may not submit them in date order, and it's possible for the HCW to take a sample much later than the illness onset date, I assume, even post-mortem. Has CDC restated the number of earlier onset cases in subsequent versions of the graph?

I noticed also that CDC is calling this a "worldwide outbreak", rather than a pandemic, which I thought was the correct term upon reaching level 5.

J.

[Vibrant62]

Re: Graph - Cumulative Count of CDC confirmed S-OI A/H1N1 Cases in the USA, May 6, 2009



Text says. The figure above shows the 394 confirmed and 414 probable cases of novel influenza A (H1N1) virus infection in the United States with known dates of onset from March 28 through May 4, 2009. Both confirmed and probable cases rose sharply from April 21 to April 27, then decreased sharply.

We know that many areas are NOW not testing unless cases hospitalised, or severe, or individuals are in at risk groups. This was not always the case. We also know that there is a large backlog of tests.

Therefore this re-assuring trend indicated in this graph must be questioned until it enough time has passed to see new trends under the new rules.