SAS KM Survival Analysis | Survival at interest of Time

carl_miles — Sat, 11 Jan 2025 17:38:56 GMT

data Patients;
    format Diagnosis_Date Followup_Date date9.;
    input Patient_ID $ Diagnosis_Date :date9. Followup_Date :date9. Obs Age Gender $ Condition $ Treatment $ Event $ event_fl;
    Days = Followup_Date - Diagnosis_Date;
    Months = (Followup_Date - Diagnosis_Date)/30;
    datalines;
P001 05OCT2011 03JUL2015 1 45 Male Hyperten DrugA No 0
P002 13MAY2011 21MAY2011 2 60 Female Diabetes DrugB Yes 1
P003 17OCT2011 07NOV2015 3 50 Male HeartDis DrugA No 0
P004 08NOV2011 17NOV2016 4 39 Female Hyperten DrugC Yes 1
P005 16JUN2011 11FEB2015 5 48 Male Diabetes DrugB No 0
P006 11FEB2011 07NOV2014 6 55 Female Obesity DrugD Yes 1
P007 16JUN2011 10SEP2015 7 65 Male Hyperten DrugA Yes 1
P008 01NOV2011 09JAN2015 8 70 Female HeartDis DrugE No 0
P009 17MAR2011 11JUL2014 9 52 Male Diabetes DrugB Yes 1
P010 09NOV2011 05OCT2014 10 46 Female Obesity DrugC Yes 1
;
run;

proc print data=Patients;
    title "Patients Dataset with Days Between Diagnosis and Follow-up Dates";
run;

proc lifetest data = Patients method=km plots=survival (cb = hw test atrisk) timelist=25 30 35 40 45; 
              time Months*event_fl(0);
              title "Kaplan Meier Analysis on time to death";
              label Months = "Time to Death";

Hi Guys,

The attached code snippet is a sample data
this is KM survival analysis using proc lifetest

timelist=25 30 35 40 45

this is the timelist i using to find survival probability at 25th month, and 30th month ..... like that

along with that i need to find the survival at >50 months
i tried mentioning timelist=25 30 35 40 45 >50

this is not giving any results so i want to find a way to find survival probability at >50 (Note: I want to mention excatly >50)

Kindly give your inputs guys

Re: SAS KM Survival Analysis | Survival at interest of Time

FreelanceReinh — Sat, 11 Jan 2025 18:54:14 GMT

Hello @carl_miles,

Just omit the ">" sign in the TIMELIST= option. The survival function S(t), which you estimate using the Kaplan-Meier method, is defined as the probability of the survival time being greater than t. So the ">" sign is implied and applies also to 25, 30, ... anyway.

Re: SAS KM Survival Analysis | Survival at interest of Time

Ksharp — Sun, 12 Jan 2025 06:34:33 GMT

As FreelanceReinh said, the value at time=50 is SDF, SDF=1-CDF , it is maybe just what you are looking for >50.

proc lifetest data = Patients method=km  plots=survival (cb = hw test atrisk) timelist=25 30 35 40 45 50; 
              time Months*event_fl(0);
              title "Kaplan Meier Analysis on time to death";
              label Months = "Time to Death";
			run;

topic Re: SAS KM Survival Analysis | Survival at interest of Time in SAS Data Science

SAS KM Survival Analysis | Survival at interest of Time

Re: SAS KM Survival Analysis | Survival at interest of Time

Re: SAS KM Survival Analysis | Survival at interest of Time