nav[aria-label="Primary Navigation"] { padding: 0; & ul { list-style: none; width: 100%; display: flex; flex-direction: row; justify-content: start; align-items: start; gap: 30px; padding: 0; & li { margin: 0; } & ul li { list-style: none; } } }

575506_en-US

liliconcom

Survival analysis with time-varying covariates

I am trying to complete an analysis of a dataset which is likely confounded by immortal time bias – some patients died during follow up resulting in them not be eligible for the study intervention. Survival analysis with time-varying covariates is the recommended solution and I understand that it is implemented in SAS with PROC PHREG (There’s a similar approach in the coxph module for R). Per JMP product developers and @JerryFish "PROC PHREG does it in two ways. Neither are doable in JMP https://go.documentation.sas.com/doc/en/pgmsascdc/9.4_3.4/statug/statug_phreg_examples07.htm

The first approach requires computational effort at run time, i.e. changing data at run time. This is not possible in Fit Proportional Hazard.

The second approach requires two response columns, which is not allowed in Fit Proportional Hazard.

Since all my analyses are completed in JMP enhanced functionality would be most helpful.

Advanced Statistical Modeling

Re: Survival analysis with time-varying covariates - Status changed to: Investigating

Re: Survival analysis with time-varying covariates - Status changed to: Acknowledged

Hi @idouglas, thank you for your suggestion! We have captured your request and will take it under consideration.

Re: Survival analysis with time-varying covariates

Hi @idouglas

Looking with a SAS example, there are 2 alternatives:

Add interval censoring in Fit Proportional Hazards
Look at the continuous effect variables and if we find a formula in them that is a function of the time variable, then do time-varying regression using that formula.