I have a question regarding my 2cmpt indirect response model and specifically on the Between Subject Variability on the Kin and Kout rate constants. I am attaching a few figures below, but want to ask how it is possible for the BSV on the IC50 to be so high, while there is seemingly no between subject variability on the Kin and Kout, is this due to the different baselines of the population? Please see the tables below.
I suggest you first to rmeove the random effects for kin and kout. The shrinkage is huge and anyway the estimated variances are alnmost 0. The shrinkage for IC50 is small which means that the variability in IC50 may be real.
IC50 meaning is the drug level that leads to half emax. Then I am not sure why it is a problem here.
Here your Imax is 1 which means that potentially the drug can completely inhibit the production of PD.
The shape of the PD profile (Y axis) with time and/or concentration can still be different across individuals because of the large variability in IC50 even though kin and kout has no variability. A small IC50 would mean that you do not need lot of drug to reach full saturation of the production while large IC50 would mean that you need lot of drug to reach saturation.