Closed rfaelens closed 6 years ago
Not required. Monolix for example just specifies how many doses is considered "steady-state". We can use the same: it is an easy workaround that is worth it, considering how much extra work is required to always specify / calculate a steady-state version of the structural model
NONMEM has a very nice format to define steady-state dosing. We should at least support this at the model / prediction level.
This will allow us to find the dose that puts the patient within target on a steady-state regimen