Hi,
I have the model below, for which I run a logistic regression
including the interaction term (NSAID*Diuretic)
------------------------
fit1=glm(resp ~ nsaid+diuretic+I(nsaid*diuretic), family=
binomial,data=w)

NSAID Diuretic Nausea+ve Nausea-ve
0 0 185 6527
0 1 53 1444
1 0 42 1293
1 1 25 253

Coefficients Std. Error z value Pr(>|z|)
(Intercept) -3.56335 0.07456 -47.794 < 2e-16 ***
NSAID 0.13630 0.17361 0.785 0.43242
Diuretic 0.25847 0.15849 1.631 0.10293
NSAID*Diuretic 0.85407 0.30603 2.791 0.00526 **
-------------------------------

Odds ratio of Nausea when on Diuretic is exp(0.25847)= 1.29
and the odds ratio of Nusea when on NSAID is exp(0.13630)=1.14
Normally when we want to see the odds ratio of Nausea when a patient
is on both drugs we multiply 1.29*1.14= 1.48 (is this correct? do we
multiply or do we add,)

But since the interaction term is significant then we take that into
account? Does that mean that the odds ratio of the interaction is
exp(0.25847)*exp(0.13630)*exp(0.85407)= 3.486297 ?

Or do we use additions?

Thanks.

OR for NSAID when Diuretic = 1 = (25*1444)/(253*53) = 2.6922
OR for NSAID when Diuretic = 0 = (42*6527)/(1293*185) = 1.1460

2.6922/1.1460 = 2.3492

Your b3 = 0.85407
Exp(b3) = 2.3492
....<Snipped for brevity

On Mar 14, 10:00 pm, Bruce Weaver <bwea...@lakeheadu.ca> wrote:
...<Snipped for brevity
OR for NSAID when Diuretic = 1 = (25*1444)/(253*53) = 2.6922
OR for NSAID when Diuretic = 0 = (42*6527)/(1293*185) = 1.1460

2.6922/1.1460 = 2.3492

Your b3 = 0.85407
Exp(b3) = 2.3492
...<Snipped for brevity

Thanks that was very useful. But what still puzzels me that:

OR for NSAID when Diuretic = 1 = (25*1444)/(253*53) = 2.6922
AND
OR for Diuretic when NSAID = 1 = (25*1293)/(253*42) = 3.0421
also verified by
exp(0.25847)*exp(0.85407)=3.0421

Why do we have 2 odds ratios (that are different) since when a patient
is taking the combination of the two drugs his odds ratio for nausea
should be fixed at one value?

On Mar 14, 10:00 pm, Bruce Weaver <bwea...@lakeheadu.ca> wrote:
...<Snipped for brevity
OR for NSAID when Diuretic = 1 = (25*1444)/(253*53) = 2.6922
OR for NSAID when Diuretic = 0 = (42*6527)/(1293*185) = 1.1460

2.6922/1.1460 = 2.3492

Your b3 = 0.85407
Exp(b3) = 2.3492
...<Snipped for brevity

Thanks that was very useful. But what still puzzels me that:

OR for NSAID when Diuretic = 1 = (25*1444)/(253*53) = 2.6922
AND
OR for Diuretic when NSAID = 1 = (25*1293)/(253*42) = 3.0421
also verified by
exp(0.25847)*exp(0.85407)=3.0421

Why do we have 2 odds ratios (that are different) since when a patient
is taking the combination of the two drugs his odds ratio for nausea
should be fixed at one value?

OR for NSAID when Diuretic = 1 = (25*1444)/(253*53) = 2.6922
AND
OR for Diuretic when NSAID = 1 = (25*1293)/(253*42) = 3.0421
also verified by
exp(0.25847)*exp(0.85407)=3.0421

Why do we have 2 odds ratios (that are different) since when a patient
is taking the combination of the two drugs his odds ratio for nausea
should be fixed at one value?

I'm not sure why you expect them to be the same. If those two odds
ratios were identical, then the odds ratios for NSAID when Diuretic = 0
and the odds ratio for Diuretic when NSAID = 0 would also have to be
identical, because the multiplicative factor has to be 2.3492 [your
Exp(b) for the product term] for both variables.

From what I understood from you I'd get two answers= one would tell me
the OR is 2.6922, and another would say the OR is 3.0421!

On 15 mrt, 02:03, Bruce Weaver <bwea...@lakeheadu.ca> wrote:
OR for NSAID when Diuretic = 1 = (25*1444)/(253*53) = 2.6922
AND
OR for Diuretic when NSAID = 1 = (25*1293)/(253*42) = 3.0421
also verified by
exp(0.25847)*exp(0.85407)=3.0421
Why do we have 2 odds ratios (that are different) since when a patient
is taking the combination of the two drugs his odds ratio for nausea
should be fixed at one value?
I'm not sure why you expect them to be the same. If those two odds
ratios were identical, then the odds ratios for NSAID when Diuretic = 0
and the odds ratio for Diuretic when NSAID = 0 would also have to be
identical, because the multiplicative factor has to be 2.3492 [your
Exp(b) for the product term] for both variables.

Maybe I didn't ask my question correctly, what I wanted to say was:
If I ask someone "what is the Odds Ratio of having Nausea when a
patient is taking both NSAIDS and Diuretics, compared to when he is
taking neither?"
From what I understood from you I'd get two answers= one would tell me
the OR is 2.6922, and another would say the OR is 3.0421!

On 15 mrt, 02:03, Bruce Weaver <bwea...@lakeheadu.ca> wrote:
OR for NSAID when Diuretic = 1 = (25*1444)/(253*53) = 2.6922
AND
OR for Diuretic when NSAID = 1 = (25*1293)/(253*42) = 3.0421
also verified by
exp(0.25847)*exp(0.85407)=3.0421

Why do we have 2 odds ratios (that are different) since when a patient
is taking the combination of the two drugs his odds ratio for nausea
should be fixed at one value?

I'm not sure why you expect them to be the same. If those two odds
ratios were identical, then the odds ratios for NSAID when Diuretic = 0
and the odds ratio for Diuretic when NSAID = 0 would also have to be
identical, because the multiplicative factor has to be 2.3492 [your
Exp(b) for the product term] for both variables.

Maybe I didn't ask my question correctly, what I wanted to say was:
If I ask someone "what is the Odds Ratio of having Nausea when a
patient is taking both NSAIDS and Diuretics, compared to when he is
taking neither?"
From what I understood from you I'd get two answers= one would tell me
the OR is 2.6922, and another would say the OR is 3.0421!

Again, look at the logs of the odds. The effect you are interested in,
log(1,1) - log(0,0) = 1.24883, is the sum of the two simple effects
and the interaction, .13630 + .25847 + .85407.