10.08.2013 Views

Introduction to Stata 8

Introduction to Stata 8

Introduction to Stata 8

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

13. Survival analysis and related issues<br />

st [ST] manual<br />

The st family of commands includes a number of facilities, described in the Survival<br />

Analysis manual [ST]. Here I describe the stset and stsplit commands and give a few<br />

examples. The data is cancer1.dta, a modification of the cancer.dta sample data<br />

accompanying <strong>Stata</strong>.<br />

The observation starts at randomization (agein), the data set includes these variables:<br />

. summvl // summvl is a summarize displaying variable labels.<br />

// Get it by: findit summvl<br />

Variable Obs Mean Std.Dev Min Max Label<br />

------------------------------------------------------------------------<br />

lbnr 48 24.5 14 1 48 Patient ID<br />

drug 48 1.875 .841099 1 3 Drug type (1=placebo)<br />

drug01 48 .583333 .498224 0 1 Drug: placebo or active<br />

agein 48 56.398 5.6763 47.0955 67.8746 Age at randomization<br />

ageout 48 57.6896 5.45418 49.0122 68.8737 Age at death or cens.<br />

risktime 48 1.29167 .854691 .083333 3.25 Years <strong>to</strong> death or cens.<br />

died 48 .645833 .483321 0 1 1 if patient died<br />

stset [ST] stset<br />

stset declares the data in memory <strong>to</strong> be survival time (st) data. I create two versions: In<br />

st.cancer1.dta time simply is risktime, age not taken in<strong>to</strong> consideration. In<br />

st.cancer2.dta time at risk is defined by age at entry (agein) and exit (ageout)<br />

enabling <strong>to</strong> study and control for the effect of age.<br />

Simple analysis – age not included<br />

stset data with risktime as the time-of-exit variable:<br />

. // c:\dokumenter\proj1\gen.st.cancer1.do<br />

. use c:\dokumenter\proj1\cancer1.dta , clear<br />

.<br />

stset risktime , failure(died==1) id(lbnr)<br />

id: lbnr<br />

failure event: died == 1<br />

obs. time interval: (risktime[_n-1], risktime]<br />

exit on or before: failure<br />

-------------------------------------------------------------------------<br />

48 <strong>to</strong>tal obs.<br />

0 exclusions<br />

-------------------------------------------------------------------------<br />

48 obs. remaining, representing<br />

48 subjects<br />

31 failures in single failure-per-subject data<br />

62 <strong>to</strong>tal analysis time at risk, at risk from t = 0<br />

earliest observed entry t = 0<br />

last observed exit t = 3.25<br />

34

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!