Hi all,
I got a bunch of studies, that include MAD and cross over studies of different dose levels. Just looking at the plots there appears to be cyp induction that you would only see in the more drawn out studies. I assume I would need to implement a between period variability on top of a between subject variability. Is there an example that does this?
Also I assumed the parameter fitting would want that all the concentration timepoints be relative to most recent administered dose (ie each dose sets the clock to zero). I am confused how best to approach it when I do see some CYP induction taking place but not all my studies will show that effect.
Any advice/comments would be much appreciated.
you can model inhibition over time IOV is pretty straight forward you define a covariate and specify it is an "OCCASION’ then it will be added as IOV on the random effects
@Samer_Mouksassi
Thanks for this tip! Though one thing that is concerning is that since I have a couple of different studies. So, the timing of each occasion is different. Can I model IOV when my data is like this?
For example
Study 1
Occasion 1 Day 0
Occasion 2 Day 7
For study 2
Occasion 1 Day 0
Occasion 2 Day 30
yes the software does not care what was the original name or structure of your data:
it just sees OCC 1 and 2 up to you to decide this is OK
Yeah I am asking you if this is OK typically…
This is a confusing thing for me…
If I have a mix of data where I have single and multidose studies. Does it make sense to collapse the Multidose studies into a series of Single dose datasets with occasion being tracked.
just make sure you have a consistent definition for occasion:
every “new” rich sampling period no matter how long it is between occasion is a new occasion.
you are not collapsing anything you are still taking “Time” into account unless you reset the system or do something else.
The occasion is just telling that potentially CL, V etc. can change from occasion to occasion randomly.