3 weeks ago
I am using data from the National Health and Nutrition Examination Survey (NHANES) which used a complex survey design to recruit participants. NHANES contains cause-specific mortality data. I am looking to use 1999-2010 NHANES data sets to analyze the relationship between an exposure and cardiovascular disease mortality using PROC SURVEYPHREG, but need to a do a competing risk analysis because participants may die from other causes before dying from cardiovascular disease and thus, from my understanding, should not be censored. It is fairly straightforward to do a competing risk analysis with PROC PHREG with the eventrate option, but noticed this approach is not possible with PROC SURVEYPHREG.
As such, is there a way to do competing risk analysis with PROC SURVEYPHREG? Any help would be great!
3 weeks ago
you could estimate the cause specific hazard by simply including the competing risk (death) as censored obs. Whether or not it's of interest is another question. See slides 26-29: https://www.sas.com/content/dam/SAS/en_ca/User%20Group%20Presentations/Vancouver-User-Group/Gondara_... - they use proc phreg here, but the code would be the same in surveyphreg i would assume