survreg with left truncated (delayed entry) data

Question

I am trying to fit a delayed entry parametric regression model for a Poisson process with Weibull baseline rate. It doesn't appear that R's survreg function supports left truncated data (I get the error: start-stop type Surv objects are not supported). Is there an alternate approach/R package that I could use to do this?

Answer 1

You may want to try something like:

flexsurv::flexsurvreg(formula = Surv(starttime, stoptime, status) ~ x1 + x2, 
                      data=data, dist = "weibull") 

Check the options the package offers which may fit your need.

Answer 2

The other way would be to define a truncated distribution and use interval values. See the survival packages example of using a truncated Gaussian distribution aka ”tobit”. https://www.rdocumentation.org/packages/survival/versions/2.44-1.1/topics/tobin