NONMEM Users Network Archive

Hosted by Cognigen

RE: estimating Ka from dataset combining rich sample study and sparse sampling study

From: Stephen Duffull <stephen.duffull>
Date: Thu, 18 Jun 2009 18:04:18 +1200

Dear Ethan

I concur with Mats's comments below.

As a note, from a design perspective adding additional data to an experimen=
t cannot result in less precise parameter estimates under the assumption th=
at the individuals from the two data sets are exchangeable. Under this ass=
umption therefore the Sparse data should merely add information to the Rich=
 data. That the Sparse data is affecting the parameter estimates from the =
Rich data suggests that the two data sets are not exchangeable (different c=
entre, different assay, different covariates ...).

Another possible way to investigate the differences between the two data se=
ts would be to analyse them sequentially, perhaps with consideration for us=
ing the analysis from the Rich data as an informative prior for the analysi=
s of the Sparse data and see where this leads you.

Kind regards

Professor Stephen Duffull
Chair of Clinical Pharmacy
School of Pharmacy
University of Otago
PO Box 913 Dunedin
New Zealand
E: stephen.duffull
P: +64 3 479 5044
F: +64 3 479 7034

Design software:<>

From: owner-nmusers
 Behalf Of Mats Karlsson
Sent: Thursday, 18 June 2009 9:17 a.m.
To: 'Ribbing, Jakob'; 'Ethan Wu'; 'Jurgen Bulitta'; nmusers
Cc: 'Roger Jelliffe'; 'Neely, Michael'
Subject: RE: [NMusers] estimating Ka from dataset combining rich sample stu=
dy and sparse sampling study

Dear Ethan,

Variances estimated to be zero may result from fixing off-diagonal variance=
s to zero (i.e. not using BLOCKs in IIV). Here, however, it may be that the=
re are systematic differences between the sparse and the rich data experime=
nts. Maybe fasting/fed status or something else is different. If the fit to=
 the rich data is markedly worse when including the rich data, at least one=
 parameter is different between the two situations. I would explore what pa=
rameter(s) that would be. In addition to Jakob's suggestions below, the two=
 data sets together may indicate a more complex structural model that a sin=
gle profile indicated. Maybe you need to go to a two-compartment for exampl=

Best regards,

Mats Karlsson, PhD
Professor of Pharmacometrics
Dept of Pharmaceutical Biosciences
Uppsala University
Box 591
751 24 Uppsala Sweden
phone: +46 18 4714105
fax: +46 18 471 4003

From: owner-nmusers
 Behalf Of Ribbing, Jakob
Sent: Wednesday, June 17, 2009 10:43 PM
To: Ethan Wu; Jurgen Bulitta; nmusers
Cc: Roger Jelliffe; Neely, Michael
Subject: RE: [NMusers] estimating Ka from dataset combining rich sample stu=
dy and sparse sampling study

Hi Ethan,

If OMEGA(?) for KA is drastically reduced when including the sparse data, t=
hen something is wrong with your model and in this case it is not the estim=
ation method or assumption on distribution of individual parameter). Eta-sh=
rinkage would not drastically reduce the estimate of OMEGA, since this esti=
mate is driven by the subjects/studies which contain information on the par=

If the sparse data is multiple dosing it may be that KA is variable between=
 occasions, rather than between subjects (assuming the sparse data contain =
some information on KA). Or if the sparse data is from a less well-controll=
ed study or a different population, it may be that increased IIV in other p=
arts of the model (e.g. OMEGA on V) is making IIV in KA appear low for the =
rich study, when fitting the two studies together. If you get the covariate=
 model in place this problem will be solved. For the simple model you have =
it should be quick to start out assuming that most parameters (THETAs and O=
MEGAs) are different between the two studies and then reduce down to a mode=
l which is stable and parsimonious. Obviously, if you eventually can explai=
n the differences using more mechanistic covariates than study number that =
is of more use.



Received on Thu Jun 18 2009 - 02:04:18 EDT

The NONMEM Users Network is maintained by ICON plc. Requests to subscribe to the network should be sent to:

Once subscribed, you may contribute to the discussion by emailing: