topic Re: Multivariable Cox proportional hazards regression. in Statistical Procedures

Multivariable Cox proportional hazards regression.

Camp — Sun, 17 Jul 2016 18:15:26 GMT

Hi Everyone,

Someone please explain me through your own example (data) the:-

Multivariable Cox proportional hazards regression models (procedure/fitting in SAS) - adjusting for baseline covariates in the model.

Thanks!

Re: Multivariable Cox proportional hazards regression.

Reeza — Sun, 17 Jul 2016 19:07:19 GMT

Have you looked at examples in documentation?

I think the third one is applicable.

http://support.sas.com/documentation/cdl/en/statug/68162/HTML/default/viewer.htm#statug_phreg_examples03.htm

Re: Multivariable Cox proportional hazards regression.

Camp — Sat, 30 Jul 2016 16:42:02 GMT

Hi Reeza,

Thanks for the link. Can you please clarify one more thing. Like we use class statement in our 'proc logistic' procedure (to specify our categorical variables used in model statement). Do we have to do the same in 'proc phreg' also? And, do it serves the same purpose 'specify our categorical variables used in model statement'?

Thanks!

Re: Multivariable Cox proportional hazards regression.

Reeza — Sat, 30 Jul 2016 19:23:14 GMT

@Camp wrote:

Like we use class statement in our 'proc logistic' procedure (to specify our categorical variables used in model statement). Do we have to do the same in 'proc phreg' also?

Yes, you use the CLASS statement in PROC PHREG similar to how you would in PROC LOGISTIC

@Camp wrote:

And, do it serves the same purpose 'specify our categorical variables used in model statement'?

Not sure what that means. You still have to list the categorical variables in both your CLASS and MODEL statements.

Re: Multivariable Cox proportional hazards regression.

Camp — Sun, 31 Jul 2016 16:44:44 GMT

Thanks! your answer clarifies my last question. Can you suggest/answer my one question (as I am not sure)

As per standard practice: Can we include continuous variable specially 'Age' on RHS of = sign (i.e. independent variable) in model statement for PHREG procedure?

I am asking this because I am not a statistician.

Re: Multivariable Cox proportional hazards regression.

Reeza — Sun, 31 Jul 2016 17:45:14 GMT

My personal suggestion, don't run a proc if you don't know what it does, especially the statistical procedures.

Yes, continuous and categorical variables can go into a model.