I am trying to find some measurements based on disease adjusting for age.
2 groups - has disease and healthy. The mean difference in age is about 10 years.
I have 4 different types of measurements but I don't know to do an AUC adjusting for age.
Please help.
%modstyle(name=linestyle, parent=statistical, type=CLM, linestyles=Solid MediumDash Solid) ods listing style=linestyle; proc template; define statgraph Graphics.AUC; notes "Receiver Operating Characteristic Overlaid Curves"; dynamic _byline_ _bytitle_ _byfootnote_; BeginGraph / designwidth=defaultDesignHeight; entrytitle 'Age Adjusted AUC of vessel parameters'; entryfootnote '*from a logistic regression controlling for inter-eye correlation*'; layout overlayequated / equatetype=square yaxisopts=(gridDisplay=auto_on label="Sensitivity" shortlabel="TPF" offsetmin=0.05 offsetmax=0.05) xaxisopts=(gridDisplay=auto_on label="1 - Specificity" shortlabel="FPF" offsetmin=0.05 offsetmax=0.05) commonaxisopts=(tickvaluelist=(0 .25 .5 .75 1) viewmin=0 viewmax=1); lineparm x=0 y=0 slope=1 / clip=true extend=true; seriesplot y=_SENSIT_ x=_1MSPEC_ / tip=(group y x) group=_ROC_ index=_GROUP_ name="Step" primary=true; discretelegend "Step" / title="ROC Curve (Area)"; endlayout; if (_BYTITLE_) entrytitle _BYLINE_ / textattrs=GRAPHVALUETEXT; else if (_BYFOOTNOTE_) entryfootnote halign=left _BYLINE_; endif; endif; EndGraph; end;run; proc genmod data=rr_g; class PID; model age=VAD_A; repeated subject=PID; output out=out1 predicted=pre1;run; proc genmod data=rr_g; class PID; model age=Flux_a; repeated subject=PID; output out=out2 predicted=pre2;run; proc sort data=out1;by PID; proc sort data=out2; by PID; run; data pregee;merge out1 out2 ;by PID;if not missing(pre1) and not missing(pre2);run; ods graphics on; ods trace on; ods select ROCOverlay; ods output ROCOverlay=ROCOverlaydata;proc logistic data=pregee;model age(event='1')=; roc 'VAD' pred=pre1; roc 'Flux' pred=pre2; roccontrast/estimate=allpairs; run; data rocoverlaydata2;set rocoverlaydata;if _ROC_='Model (0.5000)' then delete;run; proc sgrender data = rocoverlaydata2 template = Graphics.AUC; run; ods graphics off; ods html close;
I guess my model is wrong. The outcome variable should measured parameters.
Should it be switched?
I have used Proc Genmod in the past do to a logistic regression with repeated data. But I have never got a formal training in SAS, mostly reading and forums.
Are you ready for the spotlight? We're accepting content ideas for SAS Innovate 2025 to be held May 6-9 in Orlando, FL. The call is open until September 25. Read more here about why you should contribute and what is in it for you!
Learn how use the CAT functions in SAS to join values from multiple variables into a single value.
Find more tutorials on the SAS Users YouTube channel.