Re: FW: [NMusers] Question of fitting population PK model using summary statistics of data instead of raw data

From: Penny Zhu <>
Date: Thu, 10 Sep 2015 11:06:47 -0700

Dear Dinko
Thank you for the suggestion. It seems this NAD approach only uses the mea=
n data and does not estimate inter-subject variability using the standard d=
eviation data.

My intention is to establish a population PK/PD model with appropriate esti=
mation of intersubject variability based on the mean and standard deviation=
 data at each timepoint.

A major assumption is that we have good knowledge of the base structure of =
the model (e.g. biexponential), and won't run the risk mistaking 2 mono exp=
onential models for a biexponential model

Your help and discussions will be very much appreciated.


 -----Original Message-----
 From: Rekic, Dinko []
 Sent: Thursday, September 10, 2015 10:41 AM
 To: Zhu, Penny
 Subject: RE: [NMusers] Question of fitting population PK
 model using summary statistics of data instead of raw data
 See the link and text below.
 Naive averaged data approach (NAD)
     A model without BSV and BOV is fitted to the
 mean data from all individuals.
         -Specialized software not
         -Does not distinguish between
 BSV and WSV.
         -Inappropriate means lead to
 biased parameter estimates.
         -May produce model distortion
 i.e., 2 mono exponential equations averaged together can
 yield a biexponential.
         -Covariate modeling cannot be
 Kind regards
 Dinko Rekić, Ph.D., MSc(Pharm)
 Pharmacometrics reviewer
 Division of Pharmacometrics
 Office of Clinical Pharmacology
 Office of Translational Science
 Center for Drug Evaluation and Research
 U.S. Food and Drug Administration
 10903 New Hampshire Ave
 Silver Spring, MD 20993
 WO Bldg 51, Rm 3122
 Office phone: (8)240 402-3785
 "The contents of this message are mine personally and do not
 necessarily reflect any position of the Government or the
 Food and Drug Administration."
 -----Original Message-----
 On Behalf Of Penny Zhu
 Sent: Thursday, September 10, 2015 9:49 AM
 Subject: [NMusers] Question of fitting population PK model
 using summary statistics of data instead of raw data
 Dear all
 Assuming the population PK or PD data are log-normally (or
 normally) distributed, if you have the mean and standard
 deviation of a readout at each timepoint but do not have the
 actual raw data (assuming all pateints are with the same
 dosing regimen, etc),  is it possible to establish a
 well fitted population PK or PD model?  How would one
 get about doing it?
 Your help is very much appreciated
Received on Thu Sep 10 2015 - 14:06:47 EDT

This archive was generated by hypermail 2.3.0 : Fri Sep 27 2019 - 16:45:58 EDT