topic Re: Survival Analysis using Kaplan-Meier Survival function in Statistical Procedures

Survival Analysis using Kaplan-Meier Survival function

Anita_n — Thu, 29 Apr 2021 08:46:43 GMT

Dear all,

I am tring to do an overall survival analysis using the variable tumor staging (Kaplan-Meier function)

I have attached a sample data. Please, I need some help in doing this. I know I have to use proc lifetest but I want to do that correctly. thanks

Re: Survival Analysis using Kaplan-Meier Survival function

OsoGris — Thu, 29 Apr 2021 12:18:29 GMT

Take a look at the Getting Started example in the PROC LIFETEST documentation:

https://go.documentation.sas.com/doc/en/statug/15.2/statug_lifetest_gettingstarted.htm

Re: Survival Analysis using Kaplan-Meier Survival function

FreelanceReinh — Thu, 29 Apr 2021 19:17:51 GMT

Hello @Anita_n,

Before you can run PROC LIFETEST on your data, you need to supply survival times (i.e., populate variable time_in_days) for the censored observations (i.e., those with censor=0). For example, if a censored patient was followed up until date d, set time_in_days = d - date_of_diagnosis for that patient. Then a generic PROC LIFETEST step might look like this:

proc lifetest data=have plots=(s);
time time_in_days*censor(0);
strata tumor_stage;
run;

Re: Survival Analysis using Kaplan-Meier Survival function

Anita_n — Sat, 01 May 2021 07:20:42 GMT

@FreelanceReinh @OsoGris thankyou, Is it allowed for variable time_in_days to have missings?

Re: Survival Analysis using Kaplan-Meier Survival function

FreelanceReinh — Sat, 01 May 2021 08:47:40 GMT

@Anita_n wrote:

Is it allowed for variable time_in_days to have missings?

Observations with missing time_in_days will automatically be excluded from the analysis. The same holds for observations with missing censor or missing values of a STRATA variable. So PROC LIFETEST will run without errors, but with notes like

NOTE: 3 observations with invalid time, censoring, or strata values were deleted.

in the log and the output. The results will be the same as if you had omitted those incomplete observations in the first place. The information contained in their non-missing values is not used.

Re: Survival Analysis using Kaplan-Meier Survival function

Anita_n — Mon, 03 May 2021 15:22:49 GMT

@FreelanceReinh okay,thanks