Q: May you comment why selecting first the Multiplicative Error model ? Why not using Additive + Multiplicative ? Basically how do you select initial Error model?
A: Typically it is a good idea to start with the default Additive error model and work on model specification and initial error estimates. Once the structural model is finalized, the model can be fine-tuned by specifying the error model. The additive error model is often used when the response range is less than 3 orders of magnitude (e.g. 1-999). The proportional error model is often used when the response range is greater than 3 orders of magnitude (e.g. 1-10000).The residual plots such as Ind IWRES vs IPRED, Ind IWRES vs IVAR, and QQIWRES can be used to identify patterns that might suggest a different error model is appropriate.
Since the focus of this webinar was on the turnover model and multiple dosing, we did not go through this sequence of steps and picked the error model used in the Gabrielsson and Weiner reference.
Q: Is 'combined dose' created only to do the NCA?
A: Yes - in NCA it is only possible to input one dose per PK profile, so a combined dose is used. The NCA output is then used to get initial estimates for the Turnover model.
Q: Could you please repeat what idosevar infosevar infratevar stand for?
A: These are dosing variables that get created automatically by the built-in Phoenix model interface. idosevar is a bolus input amount, infdosevar is an infusion amount, infratevar is an infusion rate. These can be read from input dataset columns, e.g. from a dosing sheet. Since we had a highly variable dosing scheme, we used an explicit dosing sheet to specify the multiple dosing routes, amounts, infusion rates, and times.
Q: Why are there no CV% associated with secondary parameters?
A: There was a flaw in the textual model where secondary parameters were not using the typical values of the primary parameters from which they were derived. The model code has been corrected and posted here:
For your information here are the secondary parameters calculated with the corrected model showing CV% and Standard Errors:
Edited by email@example.com, 06 March 2017 - 04:38 PM.