NONMEM Users Network Archive

Hosted by Cognigen

RE: NONMEM message

From: Elassaiss - Schaap, J. <jeroen.elassaiss>
Date: Thu, 13 Nov 2008 11:16:14 +0100

Dear Mitsuo,

This error message pops up with me more often. I routinely work with
small datasets and hence less stable estimation. In my experience I
encounter mainly three situations after this message:

1. Covariance step fails (so the message can be ignored as Nick
suggested)
2. Covariance step produces standard error estimates really below the
normal magnitude (say 0.001 to 1%). These results should be discarded.
3. Covariance step produces standard error estimates of the normal
magnitude (say 5 to 50% or even higher) and therefore passes the check
as described in the warning.

In case 3, I would advice to make some effort to improve results, e.g.
by scrutiny of model and initial estimates to look for possible
numerical problems. You obviously did so and found the source of
instability. (I suspect that in your case the standard errors were not
changed to a large extent)

It is not always possible however to root out the problem, and I often
suspect the small size of the datasets I analyse (either in number of
subjects or density of sampling) to be the cause of instability. I
should add to this that the covariance step of Nonmem is not the only
way to assess uncertainty of model results. Moreover it is, as discussed
in this group so often, not {always} reliable.

Hope this helps,
Jeroen

Jeroen Elassaiss-Schaap, PhD
Modeling & Simulation Expert
Pharmacokinetics, Pharmacodynamics & Pharmacometrics (P3)
Early Clinical Research and Exploratory Medicine
T: +31 41266 9320
F: +31 41266 2506
jeroen.elassaiss

Schering-Plough
Room KM2515
PO Box 20
5340 BH Oss, Netherlands
www.schering-plough.com

 



-----Original Message-----
From: owner-nmusers
On Behalf Of Higashimori, Mitsuo
Sent: Thursday, 13 November, 2008 6:27
To: Nick Holford; nmusers
Subject: RE: [NMusers] NONMEM message

Dear Nick,

Thank you for your advice.
In my case, covariance matrix also was displayed without any problem.
So, I could check whether covariance matrix is reasonable or not, but it
was difficult to interpret it for me. Therefore, I wanted to know if
this message is important. In addition, I mused if such a result should
be excluded from listing and summarizing the result files after
bootstrap.

Best regards,
Mitsuo


> -----Original Message-----
> From: owner-nmusers
> [mailto:owner-nmusers
> Sent: Thursday, November 13, 2008 1:58 PM
> To: nmusers
> Subject: Re: [NMusers] NONMEM message
>
>
> Mitsuo,
>
> This is a special kind of NONMEM error message:
>
> MINIMIZATION SUCCESSFUL
> HOWEVER, PROBLEMS OCCURRED WITH THE MINIMIZATION.
> REGARD THE RESULTS OF THE ESTIMATION STEP CAREFULLY, AND
> ACCEPT THEM ONLY
> AFTER CHECKING THAT THE COVARIANCE STEP PRODUCES REASONABLE OUTPUT.
>
> because the message itself seems to be an error.
>
> Whenever I see this message the covariance step does not run
> therefore I
> cannot decide if the covariance step produces reasonable output. I
> therefore ignore this message after the word SUCCESSFUL.
>
> The numerical oddities in NONMEM are well known so I am not surprised
> that raising women to a power make them have a different
> influence from
> distinguishing them explicitly from men. The power approach is
> computationally much less efficient than a conditional
> expression so I
> never use it if I can avoid it.
>
> Nick
>
> Higashimori, Mitsuo wrote:
> > Dear all,
> >
> > I have a following error(?) massage on a poplation analysis
> using NONMEM VI.
> >
> > 0MINIMIZATION SUCCESSFUL
> > HOWEVER, PROBLEMS OCCURRED WITH THE MINIMIZATION.
> > REGARD THE RESULTS OF THE ESTIMATION STEP CAREFULLY, AND
> ACCEPT THEM ONLY
> > AFTER CHECKING THAT THE COVARIANCE STEP PRODUCES REASONABLE OUTPUT.
> >
> > I found it when I described a control stream to assess sex
> difference on oral clearance as shown in below,
> >
> > TVCL = THETA(1) * THETA(2) ** SEX
> > where SEX=0 for male and SEX=1 for female.
> >
> > This message was displayed without any error message. It
> was not dissapeared even though I changed the initial
> parameters. However, it was solved when I changed the model
> definition. For example,
> >
> > TVCL = THETA(1)
> > IF (SEX.EQ.1) TVCL = THETA(2)
> >
> > Could you please let me know some details regarding the
> message. Especially, I'd like to know
> >
> > 1. What impact does this error message give the analysis result?
> > 2. Why does it depend on the model definition?
> >
> > Thanks,
> >
> > _/ _/ _/ Mitsuo Higashimori, Ph.D.
> > _/ _/ _/ Pharmacokinetic Group, Early Phase Development Department
> > _/ _/ _/ Clinical Division, Research & Development
> > _/ _/ _/ AstraZeneca K.K.
> > _/ _/ _/ E-mail: Mitsuo.Higashimori
> >
> >
> >
> >
>
> --
> Nick Holford, Dept Pharmacology & Clinical Pharmacology
> University of Auckland, 85 Park Rd, Private Bag 92019,
> Auckland, New Zealand
> n.holford
> http://www.fmhs.auckland.ac.nz/sms/pharmacology/holford
>
>


This message and any attachments are solely for the intended recipient. =
If you are not the intended recipient, disclosure, copying, use or =
distribution of the information included in this message is prohibited =
--- Please immediately and permanently delete.
Received on Thu Nov 13 2008 - 05:16:14 EST

The NONMEM Users Network is maintained by ICON plc. Requests to subscribe to the network should be sent to: nmusers-request@iconplc.com.

Once subscribed, you may contribute to the discussion by emailing: nmusers@globomaxnm.com.