topic Re: How do you get the likelihood ratio LR+, LR- with their 95% CI in SAS Programming

How do you get the likelihood ratio LR+, LR- with their 95% CI

jeka1212 — Fri, 29 Nov 2019 14:29:53 GMT

data FatComp1;
         input Test Response Count;
         datalines;
         0 0  17
         0 1  21
         1 0  3
         1 1 39
         ;
      proc sort data=FatComp1;
         by descending Test descending Response;
         run;
      proc freq data=FatComp1 order=data;
         weight Count;
         tables Test*Response;
         run;

Hi,

Can anyone help?

In the example above I need to calculate the likelihood ratio LR+, LR- with their 95% CI and was wondering if you can help?

Many thanks

Re: How do you get the likelihood ratio LR+, LR- with their 95% CI

ballardw — Fri, 29 Nov 2019 16:55:40 GMT

Not familiar with the 'LR+' and 'LR-' notation.

I suspect that you may want, at least partially

proc freq data=work.FatComp1 order=data;
   weight Count;
   tables Test*Response /or;
   run;

proc freq data=work.FatComp1 order=data;
   weight Count;
   tables Test*Response / cmh1;
   run;

Re: How do you get the likelihood ratio LR+, LR- with their 95% CI

jeka1212 — Fri, 29 Nov 2019 20:01:51 GMT

Thanks Ballardw. The LR+ and LR- are positive and negative likelihood ratio respectively. That is what I am looking for.

Re: How do you get the likelihood ratio LR+, LR- with their 95% CI

FreelanceReinh — Sat, 30 Nov 2019 20:49:15 GMT

Hi @jeka1212,

SAS Usage Note 24170 "Estimating sensitivity, specificity, positive and negative predictive values, and other statistics" contains a worked example where these statistics are calculated and a SAS macro NLEstimate is used to obtain 95% confidence intervals for them. Please apply this to your data and get back to us if you get stuck.