Question regarding blogpost

[Alik-V]

Hi Keith,
Thank you so much for your brilliant simstudy package, it is fantastic tool for data simulation.
I am writing with a question about one of the posts in your blog - Follow-up: simstudy function for generating parameters for survival distribution.

When you provided a code example here:

RDataGenBlog/content/post/2022-02-22-follow-up-simstudy-function-for-generating-parameters-for-survival-distribution/index.en.Rmd

Line 137 in b58862b

defS <- defSurv(varname = "time", formula = "-17 - 0.8 * rx", scale = 1, shape = 0.3)

Is 0.8 coefficient by rx variable essentially equates to a 0.8 hazard ratio?

[kgoldfeld]

Hey Alik -

Thanks for your note. Yes, you are essentially correct - it is interpreted as you would in a Cox proportional hazards model, so it would be the log(HR). At the end of the post, I fit a model, and you can see the parameter is recovered quite closely. The point estimate of the hazard ratio is exp(-0.78) = 0.46, close to the true hazard ratio exp(-0.8) = 0.45.

Let me know if that answers your question, and if you have any other questions.

• Keith