vladimir69
Aug12-04, 03:24 AM
i am required to show that the odds ratio, "OR", and relative risk, "RR", are related by:
OR = \frac{RR}{(1-p_{0})} \frac{1}{(1-p_{0} RR)}
where p_{0} is the probability of disease for a non-exposed person and p_{1} is the probability of disease for an exposed person.
from the notes i have that
RR = \frac{p_{1}}{p_{0}}
and
OR = \frac{p_{1} (1-p_{0})}{p_{0} (1-p_{1})}
so when i put these into the mixing pot out pops
OR = \frac{RR (1-p_{0})}{1-p_{1}}
but there is a troublesome p_{1} in there which i cant seem to easily get rid of. i have tried using bayes formula but it just gets messy
could someone pls show me how to get rid of the p_{1}
perhaps i am going the wrong way about it after doing a bit of backward engineering, but these are all the formulas i know
thnx,
vladimir
OR = \frac{RR}{(1-p_{0})} \frac{1}{(1-p_{0} RR)}
where p_{0} is the probability of disease for a non-exposed person and p_{1} is the probability of disease for an exposed person.
from the notes i have that
RR = \frac{p_{1}}{p_{0}}
and
OR = \frac{p_{1} (1-p_{0})}{p_{0} (1-p_{1})}
so when i put these into the mixing pot out pops
OR = \frac{RR (1-p_{0})}{1-p_{1}}
but there is a troublesome p_{1} in there which i cant seem to easily get rid of. i have tried using bayes formula but it just gets messy
could someone pls show me how to get rid of the p_{1}
perhaps i am going the wrong way about it after doing a bit of backward engineering, but these are all the formulas i know
thnx,
vladimir