cancel
Showing results for 
Show  only  | Search instead for 
Did you mean: 
Choose Language Hide Translation Bar

Kaplan-Meier Analysis

Hi everyone,

 

I would like to conduct a Kaplan-Meier Survivorship analysis with JMP (version 15.2.1). I want the confidence intervals to be calculated with the Greenwood log-log approach. But I am not sure which formula JMP uses and I cannot find anything about that.

Does anyone know if the upper and lower 95% values in the JMP survival analysis is calculated with the Greenwood log-log approach? If it is calculated in a different way, is it possible to set the calculation method somewhere in JMP?

 

ConstructFish94_0-1641996498791.png

 

Thank you very much for your help.

1 ACCEPTED SOLUTION

Accepted Solutions
peng_liu
Staff

Re: Kaplan-Meier Analysis

The following formula for Sigma(S(t)) on that page is same for LOGIT and LOGLOG. 

peng_liu_0-1642091870551.png

But different transformations then calculate standard errors of the transformed quantities differently, using the same quantity above.

This page lists how the quantity is used in different transformations https://support.sas.com/documentation/cdl/en/statug/63033/HTML/default/viewer.htm#statug_lifetest_a0...

JMP follows the LOGIT path.

View solution in original post

6 REPLIES 6
peng_liu
Staff

Re: Kaplan-Meier Analysis

https://support.sas.com/documentation/cdl/en/statug/63033/HTML/default/statug_lifetest_a0000000261.h...

Maybe a SAS documentation can clarify it.

I am not sure whether "Greenwood log-log" in particular is the same or not.

JMP only provides one method.

Re: Kaplan-Meier Analysis

Thank you for your quick reply.
I know that the Greenwood formula with a log-log transform is the default in SAS v9.2 onwards. Though I am not sure if it is the default in JMP version 15.2.1.

peng_liu
Staff

Re: Kaplan-Meier Analysis

I think I understand the question now. Are you talking about the option CONFTYPE on this page: https://support.sas.com/documentation/cdl/en/statug/63033/HTML/default/viewer.htm#statug_lifetest_se...

If so, JMP implements LOGIT and that is the only method that JMP provides.

Re: Kaplan-Meier Analysis

Thank you very much. Sorry for my misleading question.

Half of my question is already answered. Unfortunately I would have needed the CONFTYPE "LOGLOG".

Moreover, I am looking for the formula with which the estimated standard error is calculated in JMP15.2.1.

The Greenwood's formula is used in SAS 9.2 and is stated on this page:

PROC LIFETEST: Product-Limit Method :: SAS/STAT(R) 9.2 User's Guide, Second Edition

peng_liu
Staff

Re: Kaplan-Meier Analysis

The following formula for Sigma(S(t)) on that page is same for LOGIT and LOGLOG. 

peng_liu_0-1642091870551.png

But different transformations then calculate standard errors of the transformed quantities differently, using the same quantity above.

This page lists how the quantity is used in different transformations https://support.sas.com/documentation/cdl/en/statug/63033/HTML/default/viewer.htm#statug_lifetest_a0...

JMP follows the LOGIT path.

Re: Kaplan-Meier Analysis

Thanks a lot for your perfect explanation.