- Mark as New
- Bookmark
- Subscribe
- Mute
- RSS Feed
- Permalink
- Report Inappropriate Content
Sanjay,
Thank you for trying to help!
The SAS I'm using is in version 9.4.
Regarding to the data, it's a traditional survival data which contains
#1 event variable: allcause, 0 as not dead, 1 as dead
#2 follow-up time: followtime
#3 main independent variable: NewSurgeIndex: 1, 2, and 3 represent 3 different groups
#4 other confounders: sex age bmi_1 smoke alcohol ldl_c meansbp nightdaysbpratio
Again, what I'm trying to do is to estimate survival probability for NewSurgeIndex with adjusting for confounders.
Moreover, in order to make these three groups comparable, so I need to draw these three survival curves on the same graph.
Actually, I've came out one way to achieve this:
ODS GRAPHICS ON;
PROC PHREG DATA=one PLOTS(OVERLAY)=SURVIVAL;
CLASS NewSurgeIndex;
MODEL followtime*allcause(0)= sex age bmi_1 smoke alcohol ldl_c
meansbp nightdaysbpratio NewSurgeIndex;
BASELINE COVARIATES=one OUT=Pred1 SURVIVAL=_ALL_/ DIRADJ GROUP=NewSurgeIndex;
ODS GRAPHICS OFF;
However, very welcome if you have any better idea!
Accepted Solutions
- Mark as New
- Bookmark
- Subscribe
- Mute
- RSS Feed
- Permalink
- Report Inappropriate Content
Using a categorical variable in the CLASS statement is the usual way to compare groups. As you've seen, this also overlays the various groups on the ODS graphics that the procedure creates.
- Mark as New
- Bookmark
- Subscribe
- Mute
- RSS Feed
- Permalink
- Report Inappropriate Content
Using a categorical variable in the CLASS statement is the usual way to compare groups. As you've seen, this also overlays the various groups on the ODS graphics that the procedure creates.
- Mark as New
- Bookmark
- Subscribe
- Mute
- RSS Feed
- Permalink
- Report Inappropriate Content
Now I can use it more confidently.