Jump to content


Quantitative comparison of NONMEM models

NONMEM Inter individual variability standard error comparison

  • Please log in to reply
2 replies to this topic

#1 mondermm



  • Members
  • Pip
  • 2 posts

Posted 23 July 2018 - 01:28 PM

Dear all


I am interested in quantitatively comparing the results of different POPPK studies for a certain drug. Each study reports the value of the pharmacokinetic parameter (e.g. clearance) and the standard error of estimated clearance, the value of estimated inter-individual variability in clearance and its standard error. 


Q1) How can we judge the superiority of those models based above parameters i.e.


Model A HIGH IIV: CL= 1 L/kg, SE in CL=0.05, (OMEGA)^2=0.55 IIV 74%, SE OF OMEGA=0.001


Model B High imprecision in IIV estimation: CL=1.2 L/KG SE in CL=0.04, (OMEGA)^2=0.02 IIV 14%, SE OF OMEGA=.06 (RSE 300%)


Q2) How can we present all these data together graphically? any examples?


Many thanks

#2 smouksassi1


    Advanced Member

  • Members
  • PipPipPip
  • 169 posts
  • LocationMontreal

Posted 23 July 2018 - 02:55 PM

Dear Mondermm,


What was the purpose of the models ? Model A might be better than Model B for a specific purpose e.g. predicting Cmax while Model B can be better for predicting AUC.


I am not sure what is the value of comparing two models developed on different datasets and with potentially different purposes.


A Model with higher uncertainty might be worst than another for the purpose of simulation of new trials but again all depend on what it the application you are after.



#3 mondermm



  • Members
  • Pip
  • 2 posts

Posted 24 July 2018 - 07:19 AM

Dear smouksassi

Thanks for your kind reply. We are comparing all the models have been developed for optimizing dosing by explaining the IIV with covariates.


Many thanks

Also tagged with one or more of these keywords: NONMEM, Inter individual variability, standard error, comparison

0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users