Projecting the Current Case Fatality Rate for Egypt

Laidback Al · #1 February 28th, 2007, 12:09 AM

Projecting the Current Case Fatality Rate in Egypt
by Laidback Al

Recent information published about the number of infections and testing among suspected H5N1 human cases in Egypt provides a basis for estimating the current attack rate (AR) and case fatality rate (CFR) of H5N1 in Egypt.

Quote:

Originally Posted by Google Translation of French

....
To limit the pressure of selection

On February 18, 2007, so only 22 cases had been confirmed, 2,365 people had, since the beginning of epizooty in this country, profited from a treatment by oseltamivir. "They were suspect cases (1991 in 2006, 207 in January 2007 and 147 in February) at which, for lack of means, a test serologic was carried out less once on three, but which were treated while following the recommendations of WHO", explains Pr Khattab. For the experts French infectiologists, "this type of use of antiviral can create a pressure of selection and contribute to the emergence of resistances to the currently available drugs. Under these conditions, it is imperative to place at the disposal of the local hospitals diagnostic tests fast which could make it possible to locally make a first sorting of the suspect cases"....

posted by lili61: http://www.flutrackers.com/forum/sho...8&postcount=20

Data
1. Since the first case of human infection of H5N1 in Egypt in March of 2006, a total of 2365 individuals had flu/influenza-like symptoms and were suspected of being infected with H5N1.
2. All of these 2365 individuals received oseltamivir (Tamiflu).
3. Only about 33% (about 780 individuals) were actually tested for H5N1.
4. About 66% (about 1585 individuals) were NOT tested for H5N1, but experienced flu/influenza-like symptoms and were treated with oseltamivir (Tamiflu).
5. Of the approximately 780 individuals tested, 22 tested positive and were confirmed for H5N1. These data imply a attack rate (AR) of 2.8% among all 780 tested individuals. Thirteen of the 22 confirmed individuals died, a case fatality rate (CFR) of 59%.

Assumptions
A. Assume that the Tamiflu was effective in treating all H5N1 infected individuals among the group of 1585 that were not tested.
B. Assume a similar AR of 2.8% for the other 1585 individuals that were not tested but given Tamiflu.

Discussion
Then it would follow that as many as 44 individuals in the non-tested group of 1585 individuals could have been infected with H5N1, but recovered after receiving Tamiflu.

Adding 44 potentially recovered cases to the 13 dead and 9 recovered confirmed cases would give a CFR of about 20%.

These data suggest, based on a generous interpretations of the number of infected, but recovered, unconfirmed human cases in Egypt, that the minimum CFR could be about 20%. If, in fact, none of the 1585 non-tested individuals were infected with H5N1, then the CFR in Egypt is 59% based on all 22 confirmed cases. Assuming a new strain of H5N1 is infecting humans this flu season, then the CFR could be as high as 87.5% (7 deaths among 8 individuals, since October 2006).

The CFR in Egypt at this time could range between a high of 87.5% to a low of about 20% based on the speculative discussions here. For the CFR to be any lower, there would have to be very many infected, but asymptomatic, human cases of H5N1 in Egypt.

---------------------------
Thanks to lili61 for posting the article and for Anne for providing the initial translation, and to NS1 for additional comments.

Theresa42 · #2 February 28th, 2007, 12:19 AM

Wow! Thanks Al, lili61 and Anne!

Theresa42 · #3 February 28th, 2007, 12:26 AM

Quote:

2. All of these 2365 individuals received oseltamivir (Tamiflu).

Do we know this to be true? I had a message from someone who said that (at least some of) the hospitals are reluctant to give out Tamiflu to just anyone with flu-like symptoms 'cause they are afraid of depleting their stocks....

Laidback Al · #4 February 28th, 2007, 12:36 AM

Quote:

Originally Posted by Theresa42

Do we know this to be true? I had a message from someone who said that (at least some of) the hospitals are reluctant to give out Tamiflu to just anyone with flu-like symptoms 'cause they are afraid of depleting their stocks....

I am just using the numbers from the article. Even if all were given anti-virals, who knows what kind of compliance they had for a full course of treatment. Besides if the attack rate is correct, as long as the potentially infected 44 individuals were in the group that received the anti-virals, distribution or lack of distribution of anti-virals to the remaining individuals would be irrelevant to estimating the CFR.

The complete article (in french) posted by lili61 was taken from:

The Daily newspaper of the Doctor of: 21/02/2007
French medical review

http://www.quotimed.com/journal/inde...ewsid=20070221

The link doesn't seem to be working.

Anne · #5 February 28th, 2007, 01:53 AM

thank you laidback : good spoken ( bien parlé !)
very clear.
the link don't work because it s a medical paper ( must pay ), and lili make a copy.

Theresa42 · #6 February 28th, 2007, 12:48 PM

Quote:

Originally Posted by LaidbackAl

I am just using the numbers from the article....

Duh! Thanks, Al -- just ignore me -- I was tired yesterday when I read your post and didn't see the numbers in the original article. Found them!

Well, it would be terrific (in a way) if the CFR was really "only" ca. 20%. That's not good -- but sure is a helluva lot better than 59% or 87.5%...!

hawkeye · #7 February 28th, 2007, 01:15 PM

Al, thank you. 2.8% with 20% CFR is getting us closer to the assumed 1918 numbers..which could, in our current situation, be good news (as odd as that sounds).

HenryN · #8 February 28th, 2007, 01:27 PM

Quote:

Originally Posted by hawkeye

Al, thank you. 2.8% with 20% CFR is getting us closer to the assumed 1918 numbers..which could, in our current situation, be good news (as odd as that sounds).

There have been many deaths in negative patients.

sharon sanders · #9 February 28th, 2007, 01:34 PM

Nice piece Al.

The influence of tamiflu treatment (even partial as a prophylatic) is unknown on the reported rates of infection. It may be large and lead to the more likely 20% CFR that you calculate.

In the case of the Egypt stain that is not tamiflu resistant this is significant.

Unfortunately one of the Egyptian strains is tamiflu resistant.

According to the study Antiviral Resistance and the Control of Pandemic Influenza by Marc Lipsitch, Ted Cohen, Megan Murray, Bruce R. Levin, if a drug resistant mutation emerges during a pandemic that reduces the virus' strength by 20% or less, the resistent strain will have a strong advantage over non-resistant virus' and will propagate until approx. 1/3 or more of all cases are drug resistant.

Also, since the resistant strain is less "fit" and takes time to get traction, its emergence may delay the pandemic for some period of time.

On the 1 hand we are potentially delaying a pandemic, on the other hand an anti-viral strain may becoming stronger.

gsgs · #10 February 28th, 2007, 02:25 PM

they might test only the more severe cases

Niko · #11 February 28th, 2007, 03:10 PM

Thanks, Laidback Al!

Lots of unknown here but I think it's worth a look and discussion.

Given the limited availability (non-existent for some) of Tamiflu stocks and the likelihood of resistance developing quickly even where resistance is not currently detected... shouldn't more emphasis be placed upon rapid testing in suspected cases? Does anyone know the price of a rapid PCR test versus a course of Tamiflu?

Ref. hawkeye posted this recently:
http://www.flutrackers.com/forum/sho...8&postcount=65

Laidback Al · #12 February 28th, 2007, 04:32 PM

Quote:

Originally Posted by niman

There have been many deaths in negative patients.

With over 2300 people in the sample, there would indeed have been deaths among those testing "negative" (about 750+) as well as deaths among the 1500 with flu symptoms that were never tested. Can we assume that there were no follow-up testing among the dead individuals in either of these two groups to determine cause?

Laidback Al · #13 February 28th, 2007, 04:47 PM

Quote:

Originally Posted by gsgs

they might test only the more severe cases

You may be correct. But in order to arrive at an accurate AR and CFR, there has to be consistent testing and epidemiological analysis. We know that people die from H5N1, we just don't know how deadly it really is without accurate data.

The application of a Tamiflu "blanket" may be obscuring our understanding the AR and CFR at the very least, and at the very worst, could be ultimately selecting for a strain that is not contained by an antiviral as noted by Florida1.