topic Re: Survival Analysis when "time to event" is category/interval in Discussions

Survival Analysis when "time to event" is category/interval

vincem — Tue, 12 Mar 2024 22:03:44 GMT

Hi there

Does JMP have a module/script to run a Survival analysis when we group the event times into intervals (as en example, time to event split into a category, 0 - 5 days, 6 -10 days, 11-15 days etc...).

Thanks for your input, Vince

Re: Survival Analysis when "time to event" is category/interval

vincem — Tue, 12 Mar 2024 23:48:00 GMT

i think waht i need is how to fit a "piecewise Cox regression" when tiem to event is in categories

Re: Survival Analysis when "time to event" is category/interval

Mark_Bailey — Wed, 13 Mar 2024 12:43:33 GMT

How do you observe time to event, are you observing each day? Why are you grouping the observations this way? Is it for interpretation?

You can use interval censoring this way if appropriate. See the description of assigning your time data columns to the Y role.

Re: Survival Analysis when "time to event" is category/interval

Mark_Bailey — Wed, 13 Mar 2024 12:45:34 GMT

Based on the limited description you provided, it is difficult to answer that question, but you intervals are in Y, not X, so segmenting the model does not seem reasonable to me.

Re: Survival Analysis when "time to event" is category/interval

vincem — Wed, 13 Mar 2024 23:34:31 GMT

Apologies in the limited info i provided...

I am using the proportional hazards module...

I have as time to event variable (continuous variable which is observed each day) and a censor variable pre-defined. Curious to know if i categories the time to event variable how does oen fit a COX PH?

Re: Survival Analysis when "time to event" is category/interval

gregbowers — Thu, 14 Mar 2024 05:02:56 GMT

Hello!

Yes, JMP does offer capabilities for conducting Survival analysis, including scenarios where event times are grouped into intervals. You can achieve this using the "Interval Censoring" option in the Fit Life by X platform. This allows you to specify intervals for the event times and perform Survival analysis accordingly. Thanks for sharing https://community.jmp.com/t5/Discussions/Survival-Analysis-when-quot-time-to-event-quot-is-category/td-p/733575 view

Thank you

Re: Survival Analysis when "time to event" is category/interval

vincem — Thu, 14 Mar 2024 09:04:59 GMT

Thank you

Re: Survival Analysis when "time to event" is category/interval

Mark_Bailey — Thu, 14 Mar 2024 17:45:33 GMT

The time to event role uses a continuous data column.

Re: Survival Analysis when "time to event" is category/interval

vincem — Fri, 15 Mar 2024 03:36:17 GMT

Given that I’m interested in estimating a hazard rate per time ‘interval ‘is this possible?

So assume I have pre-specified time to event intervals, as an example 0 to 5 days, for instance and also have a continuous time to event variable - my main interest is estimating a hazard ratio for the interval (0 to 5 days) for instance.

Re: Survival Analysis when "time to event" is category/interval

prishasingh — Fri, 12 Apr 2024 12:58:54 GMT

Thanks