absolute risk difference CIs and pvalue

kc — Tue, 07 Sep 2021 16:11:09 GMT

I am reporting absolute risk difference and CI's from proc freq below.

I am wondering which CIs and p-value should be reported in the results.

There are 2 treatment groups and both Ns > 100. But the events of interest (=1) are < 5.

proc freq data=data;
table trt*event/ chisq riskdiff(cl=( newcombe exact wald) ) pdiff;
run;

Re: absolute risk difference CIs and pvalue

gcjfernandez — Mon, 13 Sep 2021 03:44:44 GMT

Please refer this paper for more information of this topic:
https://www.lexjansen.com/wuss/2016/127_Final_Paper_PDF.pdf