From Flu Wiki 2

Forum: Need Attack Rate Formula

23 September 2006

Birdie Kate – at 16:37

If someone could please help me here… I have a chart for 14,000 people with the attack rates of 25 and 50%. I just want to double check them so if someone could do them up for me I would appreciate it. I also did not include where I got them from, so if you could tell me that also I would appreciate it.

Thanks

LauraBat 17:06

Birdie - do you mean 25/50% get ill - infection rate? 30% is kind of the commonly accepted infection rate the gov’t tosses around. It is loosely based on 1918, but in fact we really don’t know what hte actual infection rate was due to lack of record keeping. The other factor is CFR - that is of those who become ill, how many actually die. Right now that is over 50%.

Birdie Kate – at 19:08

Laura,

I would like to do a chart with a 25% infection rate showing ill, severe ill, and deaths and a 50% infection rate chart.

I do have one but for the life of me have no clue where I got the info - it’s been a long year LOL!

The one I have has a 12 week period and shows the numbers for week 1, week 2, etc.

Thanks

anon_22 – at 19:39

Birdie Kate,

If all you are doing is give a chart, then

14,000 at 25% clinical attack rate (ie people infected AND sick) = 3500

50% clinical attack rate = 7000

Also, remember that ‘infection attack rate’ is percentage of population infected, ‘with or without symptoms’, whereas clinical attack rate is percentage of population showing symptoms.

The sources are very illusive. If all you want is to give a credible presentation, then just say that even experts are just making very approximate guesses, cos it can vary by a lot, depending on the virus.

If you want source for your own understanding, they are very hard to find. These are the various sources and numbers that I found:

The only one from WHO site that I found was this set of slides from Klaus Stohr, WHO who used 35% clinical attack rate in his slide, quoting Meltzer (CDC)

PAHO quotes WHO as source for 15–35%, but I haven’t found any other WHO figure apart from the one above.

I think the 25% figure came from Ferguson ie 50% infected, of whom 50% are symptomatic.

However Ferguson, wrote 2 major papers related to pandemic mitigation that carried contradicting analysis. The first in Sep 05, Strategies for containing an emerging influenza pandemic in Southeast Asia came to 50% infection attack rate with half symptomatic ie 25% clinical attack rate. But the next one in July 06, Strategies for mitigating pandemic, which deals mainly with US and UK, he says, btw the figure of 50% used in my previous paper was wrong, it should be, erh.., 68% instead, but never mind, it doesn’t affect the result by a lot!!

(There is a more complex set of inconsistencies in the Ferguson figures which I won’t go into. If someone is interested, post and I will respond, so as not to confuse Birdie Kate more than is necessary.)


I find this most recent paper slightly more credible Reducing the impact of the next influenza pandemic using household-based public health interventions, Wu et al, using 1918 data, gives 50% infection attack rate, with 1/3 asymptomatic, ie 34% clinical attack rate.


For what it’s worth…

Birdie Kate – at 20:55

ok I am going to try and post this, it is excel. After I post it might just look like a bunch of mixed up numbers (so feel free to remove). If I am lucky it will come out in some legible form LOL

parameters population 14,000 attack rate 20% severity rate 2% case fatality 1% weeks percentage Ill Severe Deaths 1 5% 140 2.8 1.4 2 15% 420 8.4 4.2 3 45% 1260 25.2 12.6 4 28% 784 15.68 7.84 5 7% 196 4 1.96 6 7 8 9 10 11 12 total 100% 2800 56 28

parameters population 14,000 attack rate 50% severity rate 25% case fatality 10% weeks percentage Ill Severe Deaths 1 5% 350 87.5 35 2 15% 1050 262.5 105 3 45% 3150 787.5 315 4 28% 1960 490 196 5 7% 490 123 49 6 7 8 9 10 11 12 total 100% 7000 1750 700

Birdie Kate – at 21:00

oops, looks like it didn’t work I will try again - i won’t do the weeks this time

population 14,000 attack rate 20% severity rate 2% case fatality 1%

2800 ill 56 severely ill 28 dead

population 14,000 attack rate 50% severity rate 25% case fatality 10%

7,000 ill 1,750 severly ill 700 dead

i think I got these from flu surge though I am not sure

anon_22 – at 21:00

Birdie Kate – at 20:55

Well, I know exactly what you mean!

LOL

Now if you will kindly ask another mod who is more clued up than me <g> those numbers might come out right!

anon_22 – at 21:01

still doesn’t make sense

what is severity rate?

Birdie Kate – at 21:10

anon_22, that is the problem! I did it a while back and don’t have a clue where I got the info from. I might have even got it from someone here.

What I have is a spread sheet that shows 12 weeks. Each week the percentage level goes up.

Maybe I can do a search for my posts and see what the formula was. I just want to post a chart on a webpage I am creating and felt I needed to have some reference to where my numbers came from.

Thanks for your help. Avian flu drives me crazy sometimes!

Birdie Kate – at 21:17

found it - http://www.fluwikie2.com/pmwiki.php?n=Forum.NeedNumbers

anon_22 – at 21:55

Birdie Kate,

It looks like the original spreadsheet came from lugon. You might want to ask him his source for reference, or whether he means something else.

My opinion? Make it simple. At any time, do not have more than 2 things on a page that people won’t understand right away. Anything 3 or more, they become convinced either that it’s all rubbish or they won’t ever understand it.

anon_22 – at 21:55

Its the same as never having more than 2 font types in a document.

24 September 2006

TRay75at 00:45

I’m slugging my way through the Flu Surge software manual and have to have a report out by Wednesday morning using it. I’ll let you know if I come across anything that explains it. I’m pretty frustrated myself, because the results either come in dramatically small, or impossibly large with minor tweaks. No wonder no one will commit to numbers on this.

TRay75at 00:54

Actually, one item I forgot to mention. You have to reduce the population by the number of deaths the prior week, and factor in the duration of the clinically ill taken out of the available pool of new population that can be infected each week. Then you can total the number of ill or dead by adding each week to the week before (an array formula). I gave up and went to use Flu Surge, hoping it already had the factors taken into account to try and save time - plus it was written by CDC sources, so it was “credible”.

05 November 2006

Closed - Bronco Bill – at 22:21

Closed to maintain Forum speed.

Retrieved from http://www.fluwikie2.com/index.php?n=Forum.NeedAttackRateFormula
Page last modified on November 05, 2006, at 10:21 PM