topic Re: Survival Analysis Question in Discussions

Survival Analysis Question

sajanjiv — Fri, 23 Jun 2017 00:12:59 GMT

I have some data and I want to calculate overall and disease specific survival. While the alive/dead are 1/0 for the survival analysis with censor=1, how should I code the list for disease specific mortaility. How should I code the ones that are dead, but not due to the disease in questions, 1 or 0? or should they be excluded?

Re: Survival Analysis Question

txnelson — Fri, 23 Jun 2017 00:42:30 GMT

The documentation on the Survival analysis is found at:

Help==>Books==>Reliability and Survival Methods

You will see examples of how to code your values

Re: Survival Analysis Question

Ted — Fri, 23 Jun 2017 18:54:33 GMT

Specific mortaility. "Should they be excluded?"

In no case! This is a competing risk model. I, unfortunately, have only JMP 10 versions, where is no such model. Therefore, I'll explain the scheme.

Censor code set 0 (you can leave 1, but It will be more understandable in this case).

In one column (CENSOR) we have the following codes:

1 -- dead due to the disease,

2 -- dead not due to the disease,

0 -- alive (censor).

In one (other) column (TIME) we have:

the time to death due to the disease (corresponds to 1 in column CENSOR) or
the death not due to the disease (corresponds to 2) or
the time of the last informing that the patient is alive (corresponds to 0).