topic Re: ROC Curve Comparison using ROC Contrast in Statistical Procedures

ROC Curve Comparison using ROC Contrast

Le_h — Mon, 18 Mar 2024 20:21:11 GMT

Hello

I want to compare two ROC Curves via cross logistics using the ROCCONTRAST pathway.

This is the Code:

ods graphics on;

proc logistic data=work.import plots=roc(id=prob);

class SCAI_Erstmessung SCAI_Tag_1_ SCAI_min_num SCAI_max_num;

model EntlassartKH_Binary = SCAI_Erstmessung SCAI_Tag_1_ SCAI_min_num SCAI_max_num/ nofit;

roc 'SCAI_Erstmessung' SCAI_Erstmessung;

roc 'SCAI_Tag_1_' SCAI_Erstmessung;

roc 'SCAI_Tag_min_' SCAI_min_num;

roc 'SCAI_Tag_max_' SCAI_max_num;

roccontrast reference('SCAI_Tag_1_') / estimate e;

run;

Does SAS automatically integrate some adjustments to multivariable testing in the code? And if so, how can I identify them?

Thankful for any suggestions! 🙂

Re: ROC Curve Comparison using ROC Contrast

StatDave — Mon, 18 Mar 2024 20:42:05 GMT

It's not clear what you want. What "adjustments" do you mean? If you mean that you want to adjust for performing multiple tests, then you can do that by saving the table of the comparisons and using an p-value adjustment method in PROC MULTTEST. You can save the comparisons table by adding this ODS OUTPUT statement in the PROC LOGISTIC step:

ods output roccontrastestimate=roccon;

and then do the comparison in MULTTEST using the INPVALUES= option like below.

proc multtest inpvalues(probchisq)=roccon holm;

See the example titled "Inputting Raw p-Values" in the MULTTEST documentation.

Re: ROC Curve Comparison using ROC Contrast

Ksharp — Tue, 19 Mar 2024 01:58:33 GMT

I think answer is no .
For your code , there are four X variables correspond to four ROC curve separately , you can not adjust some multivariable in it.
If you want "adjust multivariable " ,then you need to output these ROC curve one by one and compare them afterwards.

Check this :
https://support.sas.com/kb/45/339.html