No replies to this topic

[Teodora Dumitrescu]

Dear all,

About 1 year ago I attended the 3 day Pharsight training in Trial simulator. I haven't used Trial simulator since, however, now I have the opportunity to do so in my work. As a guidance in setting up my model, I am using the "Trial simulator 2.2.1 Hands-on Training Materials".

The model I'm using for simulation was built in NONMEM and previously published (I did not build it myself). For this simulation exercise, I would like to include not only the inter- and intra- subject variability, but also the uncertainty in the parameter estimates.

To do this, according to the "Trial simulator 2.2.1 Hands-on Training Materials" , I should use a continuous distribution block, enter the typical value of the parameter estimate form the NONMEM output as the mean, and enter the Standard error from the NONMEM output (Standard error of the Thetas section of the output) as the standard deviation of the distribution. To me, this doesn't sound right.

Could you confirm that this is what I should do? Should I convert the SE from the nonmem output into SD? How can I do that?

Thanks,

Dora