Miettienen-Nurminen confidence limit for stratified data

[Smile_to_you]

Hello, I have a result dataset from a clinical trial:

obj  stratum  trt  success

 1       1          A      1

 2       1         B       0

 3       2         B       1

  ......

I want to calculate difference of success rate between treatment A and B, as well as its CL using Miettienen-Nurminen

method stratified by the variable "stratum".

The code I used is

proc freq data=data order=formatted;
tables STRATUM*TRT*sucess/ riskdiff(cl=mn);
ods output RiskDiffCol2=pdiff;
ods output PdiffCLs=pdiffci;
run;

With this code I can only get MN CI within each stratum. I am wondering how I can get overall weighted CI accross stratum?

My SAS version is 9.4

Thanks!

[Ksharp]

It looks not so easy. Check:

http://support.sas.com/kb/37/228.html

data have;
input stratum  trt $  success;
cards;
1          A      1
1         B       0
1          A      0
1         B       0
1          A      1
1         B       1
1          A      1
1         B       0
1          A      0
1         B       1
1          A      0
1         B       1
1          A      1
1         B       0
1          A      0
1         B       0
2         B       1
2          A      1
2         B       1
2          A      0
2         B       0
2          A      0
2         B       1
2          A      1
2         B       1
2          A      1
2         B       0
2          A      0
2         B       0
2          A      1
2         B       0
2          A      1
2         B       0
;
run;

proc genmod data=have descend;
 class stratum trt;
 model success=stratum trt/dist=binomial link=identity;
 lsmeans trt/ diff cl ;
 effectplot /clm;
run;

OR

proc catmod data=have;
  response 0 1;
  model success = stratum trt / param=ref clparm;
  contrast "Prob Diff Trt A-B" trt 1  / estimate=parm;
 quit;

[Smile_to_you]

Thank you! But I am not sure if it gives us MN confidence limit.

It appears that Proc Freq can do it, but SAS does not provide detailed Syntax.

http://support.sas.com/documentation/cdl/en/procstat/67528/HTML/default/viewer.htm#procstat_freq_det...

[Ksharp]

As far as I know, riskdiff is only for 2x2 table, I am not sure if you can get it by including the third way/variable.

[Smile_to_you]

I am not sure how SAS can handle with statified CL with proc freq, but this is what it says on the website.