I have readmission rates for hospitals and I am looking to calculate a confidence interval for the readmission rates. So I have patients with an infection and patients without.
There were 66 patients with an infection and those 66 patients had 41 readmissions (Readmission Rate=.6212).
There were 1,922 patients without an infection those 1,922 patients had 227 readmissions (Readmission Rate=.1181)
I used a 2-sample Poisson Rate test to determine if patients with the infections had a greater readmission rate, but I need to know the confidence interval for each measure.
To get the confidence interval for the readmission rate should I use a 1-sample Poisson Rate? Should the confidence interval be asymetric?
It sounds to me like what you are working with are what we'd usually call proportions. Calculating confidence intervals for a proportion is quite simple: http://stattrek.com/lesson4/proportion.aspx
Reply With Quote
Posting Permissions