Kaplan Meier Survival Curve

 

For a study that follow up a population for a specific event or endpoint, such as death, recurrence of a cancer, etc. Kaplan Meier survival curve is commonly used to estimate the survival function of that population.

The participants will be followed beginning at a certain starting-point, and the time to the event of interest will be recorded for each participant.

Usually, the end of the study is reached before all participants have presented this event. Also some participants may withdraw from the study. For these participants their outcome is unknown, the time of follow-up is recorded (censored data).

Input variables

Outcome variable: which is coded as 1 (had the endpoint) and 0 (end point is unknown).

Time variable: the time to the event (if had the endpoint), or time to drop out (if endpoint is unknown)

If a stratified variable is specified, separate Kaplan Meier survival curve will be estimated and plotted for each stratum.

Below is the sample screen shot of input window

 

SNAGHTML17c6999d

 

A sample output and explanation:

 

Kaplan Meier curve

Call: survfit(formula = Surv(CYCLE, PREG) ~ HSMK, data = WD)

 

                0

 time n.risk n.event survival std.err lower 95% CI upper 95% CI

    1     70      14   0.8000  0.0478       0.7116        0.899

    2     54      18   0.5333  0.0604       0.4272        0.666

    3     36      12   0.3556  0.0581       0.2581        0.490

    4     24       7   0.2519  0.0528       0.1671        0.380

    5     17       4   0.1926  0.0479       0.1182        0.314

    6     13       2   0.1630  0.0449       0.0949        0.280

    7     11       2   0.1333  0.0413       0.0726        0.245

    8      8       2   0.1000  0.0371       0.0483        0.207

    9      6       1   0.0833  0.0345       0.0370        0.187

   10      4       1   0.0625  0.0315       0.0233        0.168

 

                1

 time n.risk n.event survival std.err lower 95% CI upper 95% CI

    1    129      26   0.7984  0.0353      0.73214       0.8708

    2    103      38   0.5039  0.0440      0.42458       0.5980

    3     63      24   0.3119  0.0411      0.24086       0.4040

    4     38      10   0.2298  0.0376      0.16675       0.3168

    5     27       7   0.1703  0.0339      0.11517       0.2517

    6     20       9   0.0936  0.0266      0.05366       0.1634

    7     10       3   0.0655  0.0230      0.03291       0.1305

    8      7       2   0.0468  0.0199      0.02035       0.1077

   10      4       1   0.0351  0.0180      0.01283       0.0961

   11      3       1   0.0234  0.0154      0.00647       0.0847

   12      2       1   0.0117  0.0113      0.00177       0.0775

 

Call:

survdiff(formula = Surv(CYCLE, PREG) ~ HSMK, data = WD)

 

         N Observed Expected (O-E)^2/E (O-E)^2/V

HSMK=0  70       63       68     0.359      0.81

HSMK=1 129      122      117     0.208      0.81

 

 Chisq= 0.8  on 1 degrees of freedom, p= 0.368

 

In above, two life tables were estimated, one is for HSMK=0 and one for HSMK=1. And a test for the difference between the two survival curves using the G-rho family of tests were done. P value of the test is 0.368.

 

Plot of survival curves: