On Jun 27, 2012, at 3:11 AM, ericlsh wrote:
> I would like to compare the incidence rates of three groups. They are
> supposed to have different risks so I would like to test whether there is a
> increasing trend in the incidence rates. Does R or any packages provide a
> trend test for incidence rates? I checked epiR and epitools. It seems they
> do not have this function.
>
> Thank you for the help.
You can use ?prop.trend.test or you can also look at the ?independence_test
function with 'teststat = "quad"', which is in the coin
package on CRAN.
Regards,
Marc Schwartz