NONMEM Users Network Archive

Hosted by Cognigen

RE: Negative concentration from simulation

From: Ken Kowalski <kgkowalski58>
Date: Tue, 2 Jun 2020 16:56:25 -0400

Hi Nyein,

I agree with Nick that it may be valid to simulate negative concentrations =
and that the only reason that we don't observe negative concentrations is b=
ecause assay labs censor these values. However, for these negative concent=
rations to be reasonable and attributed to assay variation, your estimate o=
f the additive residual error standard deviation should probably be in line=
 with what you would attribute to assay variation. I have seen model fits =
using the additive and proportional residual error model where the additive=
 residual error variance (or standard deviation) was too large to be attrib=
uted to assay variation. For example, if the additive residual error stand=
ard deviation is larger than the LLOQ that may be too high to be attributed=
 to assay variation. One thing you could do is a VPC from your model with =
your observed dataset and see if you simulate a greater proportion of BQL o=
bservations (including negative concentrations as well as positive concentr=
ations below the LLOQ) than in your observed dataset. This will help clue =
you in as to whether your residual error model is reasonable in simulating =
very low (and possibly negative) concentrations.

Best,

Ken

-----Original Message-----
From: owner-nmusers
 Behalf Of Nick Holford
Sent: Tuesday, June 2, 2020 3:46 PM
To: nmusers
Subject: RE: [NMusers] Negative concentration from simulation

Hi Nyein,

For drug concentrations the additive error model assumes that the backgroun=
d noise is random with mean zero when the drug concentration is truly zero.=
 In the real world there is always background noise for measurements which =
means that real measurements can appear to be a negative concentration even=
 though the true concentration is zero. Simulations that simulate negative =
concentrations are therefore more realistic than those that ignore reality =
and are reported as censored measurement values.

The honest thing to do is to report measurements as they are. The dishonest=
 thing is to report real measurements as below some arbitrary limit of quan=
tification. There are numerous papers which describe the bias arising from =
dishonest reporting of real measurements and work arounds if you have to de=
al with this kind of scientific fraud e.g.

Beal SL. Ways to fit a PK model with some data below the quantification lim=
it. Journal of Pharmacokinetics & Pharmacodynamics. 2001;28(5):481-504.
Duval V, Karlsson MO. Impact of omission or replacement of data below the l=
imit of quantification on parameter estimates in a two-compartment model. P=
harm Res. 2002;19(12):1835-40.
Ahn JE, Karlsson MO, Dunne A, Ludden TM. Likelihood based approaches to han=
dling data below the quantification limit using NONMEM VI. J Pharmacokinet =
Pharmacodyn. 2008;35(4):401-21.
Byon W, Fletcher CV, Brundage RC. Impact of censoring data below an arbitra=
ry quantification limit on structural model misspecification. J Pharmacokin=
et Pharmacodyn. 2008;35(1):101-16.
Senn S, Holford N, Hockey H. The ghosts of departed quantities: approaches =
to dealing with observations below the limit of quantitation. Stat Med. 201=
2;31(30):4280-95.
Keizer RJ, Jansen RS, Rosing H, Thijssen B, Beijnen JH, Schellens JHM, et a=
l. Incorporation of concentration data below the limit of quantification in=
 population pharmacokinetic analyses. Pharmacology research & perspectives.=
 2015;3(2):10.1002/prp2.131

Best wishes,
Nick





--
Nick Holford, Professor Clinical Pharmacology Dept Pharmacology & Clinical =
Pharmacology, Bldg 503 Room 302A University of Auckland,85 Park Rd,Private =
Bag 92019,Auckland,New Zealand
office:+64(9)923-6730 mobile:NZ+64(21)46 23 53 FR+33(6)62 32 46 72
email: n.holford
http://holford.fmhs.auckland.ac.nz/
http://orcid.org/0000-0002-4031-2514
Read the question, answer the question, attempt all questions

-----Original Message-----
From: owner-nmusers
 Of Bill Denney
Sent: Tuesday, 2 June 2020 8:30 PM
To: Nyein Hsu Maung <nyeinhsumaung2018
Subject: RE: [NMusers] Negative concentration from simulation

Hi Nyein,

Negative concentrations can be expected from simulations if the model inclu=
des additive residual error. I assume that you mean additive and proportio=
nal error when you say "combined error model". If the error structure does=
 not include additive error, then we'd need to know more.

How you will handle them in analysis depends on the goals of the analysis.
Usually, you will either simply set negative values to zero or set all valu=
es below the limit of quantification to zero.

Thanks,

Bill

-----Original Message-----
From: owner-nmusers
 Of Nyein Hsu Maung
Sent: Tuesday, June 2, 2020 2:13 PM
To: nmusers
Subject: [NMusers] Negative concentration from simulation


Dear NONMEM users,
I tried to simulate a new dataset by using a previously published pop pk mo=
del. Their model was described by combined error model for residual variabi=
lity. And after simulation, I have obtained two negative concentrations. I =
would like to know if there is any proper way to handle those negative conc=
entrations or if there are some codings to prevent gaining negative concent=
rations. Thanks.

Best regards,
Nyein Hsu Maung



--
This email has been checked for viruses by Avast antivirus software.
https://www.avast.com/antivirus

Received on Tue Jun 02 2020 - 16:56:25 EDT

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.