Statistics... The real questions ???

[ Follow Ups ] [ Post Followup ] [ The Iguana & Water Dragon Forum ] [ FAQ ]

Posted by Julie :o) on April 11, 2000 at 21:32:19:

In Reply to: statistics posted by cwatbay on April 11, 2000 at 13:34:15:

Hummm... statistics... Not always easy to know when they are significant or not... But I agree with you... The numbers here mean nothing... But it is not that simple...

From a statistical point of view, the real question should be : What is the "relative risk" of developping a salmonellosis when you have herps? I explain myself... What is the ratio of salmonellosis in herps exposed persons vs non-exposed? Is this ratio significant? Meaning : Am I more at risk of developping salmonellosis than the general population if I have herps?

Let be hypothetical here and try the exercice with completely fake numbers ( so maybe if someone have access to the real numbers, they can tell us their conclusions...)

Let say Illinois population is 2 000000 and that 3% (60 000) of them have reptiles (wich I think is approx. the % of houses with herps in the USA)
Let say 1 500 salmonellosis is reported each year, 60 being herps related (let say here that they consider that every salmonellosis in person with herps is cause by the herps and nothing else, wich is not necessarely the reality...)
Let say that all salmonellosis not related with herps are in people not exposed at all to herps, making 1440 cases in non-exposed and 60 cases in exposed persons.

Thus, let say our control group are the 1 998 500 persons non-infected, 59 940 being exposed to herps and 1 943 560 being not exposed.

To calculate the relative risk here, we must use what we call the "odds ratio" wich is the odds that a case is exposed to the risk factor divided by the odds that a control is exposed.

1) Odds that a person with salmonellosis is exposed : (60 cases herps-related/1 500 cases) / (1440 cases/ 1500) = 0.042
2) Odds that a control without salmonellosis is exposed : (59 940 / 1 998 500) / (1 943 560 / 1 998 500) = 0.031
3) The odds ratio is : 0.042 / 0.031 = 1,35

Meaning that in that hypothetical Illinois, the risk of developping a salmonellosis would be ** 1,35 greater ** if you had reptiles (Wich finally, wouldn't be that much but still would mean that you would be more at risk...)

Hope it will help understand how the community health services asses risks.

Julie :o)

Cold Blooded : A lizards and amphibians web site

Follow Ups:

And DOUBLE RISK? Just for the fun of playing with numbers... Julie :o) 04/12/00 (0)

Post a Followup

Name:
E-Mail:

Subject:

Comments:
: Hummm... statistics... Not always easy to know when they are significant or not... But I agree with you... The numbers here mean nothing... But it is not that simple... : From a statistical point of view, the real question should be : What is the "relative risk" of developping a salmonellosis when you have herps? I explain myself... What is the ratio of salmonellosis in herps exposed persons vs non-exposed? Is this ratio significant? Meaning : Am I more at risk of developping salmonellosis than the general population if I have herps? : Let be hypothetical here and try the exercice with completely fake numbers ( so maybe if someone have access to the real numbers, they can tell us their conclusions...) : Let say Illinois population is 2 000000 and that 3% (60 000) of them have reptiles (wich I think is approx. the % of houses with herps in the USA) : Let say 1 500 salmonellosis is reported each year, 60 being herps related (let say here that they consider that every salmonellosis in person with herps is cause by the herps and nothing else, wich is not necessarely the reality...) : Let say that all salmonellosis not related with herps are in people not exposed at all to herps, making 1440 cases in non-exposed and 60 cases in exposed persons. : Thus, let say our control group are the 1 998 500 persons non-infected, 59 940 being exposed to herps and 1 943 560 being not exposed. : To calculate the relative risk here, we must use what we call the "odds ratio" wich is the odds that a case is exposed to the risk factor divided by the odds that a control is exposed. : : 1) Odds that a person with salmonellosis is exposed : (60 cases herps-related/1 500 cases) / (1440 cases/ 1500) = 0.042 : 2) Odds that a control without salmonellosis is exposed : (59 940 / 1 998 500) / (1 943 560 / 1 998 500) = 0.031 : 3) The odds ratio is : 0.042 / 0.031 = 1,35 : : Meaning that in that hypothetical Illinois, the risk of developping a salmonellosis would be ** 1,35 greater ** if you had reptiles (Wich finally, wouldn't be that much but still would mean that you would be more at risk...) : Hope it will help understand how the community health services asses risks. : Julie :o)

Optional Link URL:
Link Title:
Optional Image URL:

[ Follow Ups ] [ Post Followup ] [ The Iguana & Water Dragon Forum ] [ FAQ ]