Sign in to follow this  
Followers 0
mananiso

Here the confidence interval does not cross 1. (HELLLLLLP)

5 posts in this topic

i dont understand the meaning of this

how come the CI does not cross one when we have the following value

odd ration 2

CI 1.4 to 2.7

how can i estimate the CI here ??????

Share this post


Link to post
Share on other sites

I am uncertain whether I understood your question correctly.

However, I will try to answer.

Your CI is 1.4 to 2.7, then how can it cross one?

95% CI for Odds ratio is estimated by the following formula,

Upper limit InOR = InOR + 1.96 * SE(InOR)

Lower limit InOR = InOR - 1.96 * SE(InOR)

Upper limit OR = eupper limit lnOR

Lower limit OR = elower limit lnOR

Read the following links,

http://www.biostat.umn.edu/~susant/Fall10ph6414/Lesson14_complete.pdf

http://www.hutchon.net/confidor.htm

Good luck

Share this post


Link to post
Share on other sites

because they always bring a value of for instance 1.5 for Odd ratio, and then thet say since the odd ratio did not cross 1!!!

isnt 1.5 is more than 1??? or am i hallucinating!!

i am sure there is something i cant understand in here and this goes for all the value stated in spmm course for the point of no difeer3nce for odd ratios, relative risk, number needed to treat and so on.

Share this post


Link to post
Share on other sites

Dear Friend,

I get your point now.

The statistical significance of odds ratios cannot be interpreted without their 95% confidence intervals.

Assume that the OR is 1.5 and its 95% CI is 1.2-2.0,

The odds ratio for the study sample is 1.5.

OR for the whole population of interest may vary between 1.2 (lower limit) and 2.0 (upper limit).

In the most conservative scenario, the investigated exposure variable increases the odds of outcome by 1.2 times (lower limit of 95% CI of OR). So, it is a significant risk factor. Its p value should be less than 0.05.

Now assume that the OR is 1.5 and its 95% CI is 0.7-2.0

OR for the whole population of interest may vary between 0.7 (lower limit) and 2.0 (upper limit).

In the most conservative scenario, the investigated exposure variable decreases the odds of outcome by 0.7 times (lower limit of 95% CI of OR). However, at the other end, it increases the odds of outcome by 2 times. We are uncertain whether the exposure variable is a risk factor or a protective factor. So, this OR is not significant. Its p value should be above 0.05.

In other words, when the 95% CI does not cross the value one and it is always above one, the exposure variable is a significant risk factor.

When the 95% CI does not cross the value one and it is always below one, the exposure variable is a significant protective factor.

When the 95% CI crosses the value one, whatever the OR value, it is not significant.

Do I make any sense?

2 people like this

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!


Register a new account

Sign in

Already have an account? Sign in here.


Sign In Now
Sign in to follow this  
Followers 0