NONMEM Users Network Archive

Hosted by Cognigen

Re: [NMusers] Using evid 0 before dosing

From: Carlos ST <carlos.serra91_at_gmail.com>
Date: Wed, 6 Nov 2019 18:28:16 +0000

Hi all,

Thank you for your suggestions I really appreciate it. I will also
have a look to these resources.

Thanks again,
Carlos



On Wed, 6 Nov 2019 at 17:21, Mark Sale <msale_at_nuventra.com> wrote:
>
> good point Steve, no USEFUL (unbiased?) information.
> But, I'd suggest it is an error to include >LLOQ predose but exclude <LLO=
Q data, that seems biased as well. We remove all predose data.
>
>
>
> Mark Sale M.D.
>
> Senior Vice President, Pharmacometrics
>
> Nuventra Inc.
> 2525 Meridian Parkway, Suite 200
> Durham, NC 27713
>
> Phone (919)-973-0383
>
> msale_at_nuventra.com
>
>
>
>
>
> CONFIDENTIALITY NOTICE The information in this transmittal (including att=
achments, if any) may be privileged and confidential and is intended only f=
or the recipient(s) listed above. Any review, use, disclosure, distribution=
 or copying of this transmittal, in any form, is prohibited except by or on=
 behalf of the intended recipient(s). If you have received this transmittal=
 in error, please notify me immediately by reply email and destroy all copi=
es of the transmittal.
>
> ________________________________
> From: Stephen Duffull <stephen.duffull_at_otago.ac.nz>
> Sent: Wednesday, November 6, 2019 9:05 AM
> To: Dennis Fisher <fisher_at_plessthan.com>; Mark Sale <msale_at_nuventra.com>
> Cc: nmusers_at_globomaxnm.com <nmusers_at_globomaxnm.com>
> Subject: RE: [NMusers] Using evid 0 before dosing
>
>
> WARNING: This email originated from outside of the company. Do not click =
links or open attachments unless you recognize the sender and are expecting=
 the message.
>
> Hi all
>
>
>
> In theory a sample that has an expected concentration=0 and if you usin=
g an additive error then this sample will provide information about \sigma_=
add.
>
>
>
> However, this will only be the case if the assay result is reported exact=
ly as is (which would allow negative observations). However, since this is=
 generally not the case (i.e. the assay result is not reported exactly) the=
n I agree it will not provide useful information to the estimation process.
>
>
>
> Cheers
>
>
>
> Steve
>
>
>
> ……………………=
……………………=
…………………=
……………………=
……………………=
………..…………=
………..…………=
…………………=
……
>
> Stephen Duffull I Professor of Clinical Pharmacy
>
> Otago Pharmacometrics Group
>
> School of Pharmacy | Te Kura Mātauraka Wai-whakaora
> University of Otago | Te Whare Wānanga o Otāgo
>
> Dunedin | Ōtepoti
>
> Ph: 64 3 479 5099
>
>
>
> Website | www.pharmacometrics.co.nz
>
>
>
>
>
>
>
> From: owner-nmusers_at_globomaxnm.com <owner-nmusers_at_globomaxnm.com> On Beha=
lf Of Dennis Fisher
> Sent: Thursday, 7 November 2019 5:45 a.m.
> To: Mark Sale <msale_at_nuventra.com>
> Cc: nmusers_at_globomaxnm.com
> Subject: Re: [NMusers] Using evid 0 before dosing
>
>
>
> Mark
>
>
>
> I disagree (more than a little). If a sample is reported as BQL and the =
expected value is 0 (i.e., pre-dose, not endogenous), what information is t=
here about assay precision. If the error model is additive (or additive + =
proportional), the sample will contribute zero to the objective function. =
If the error model is proportional, NONMEM will report an error (0/0). I a=
gree that pre-dose samples > LOQ provide IMPORTANT information about assay =
precision.
>
>
>
> Were you evacuated during the fires?
>
>
>
> Dennis
>
>
>
> Dennis Fisher MD
> P < (The "P Less Than" Company)
> Phone / Fax: 1-866-PLessThan (1-866-753-7784)
> www.PLessThan.com
>
>
>
>
>
>
>
> On Nov 6, 2019, at 8:40 AM, Mark Sale <msale_at_nuventra.com> wrote:
>
>
>
> Dennis,
>
> I may have to disagree, a little. There is information, a little, in a pr=
e dose BQL, about assay precision. I think you might agree that is a pre do=
se sample is NOT BQL (which happens), tells you (and NONMEM) something abou=
t the assay. Converse, even a BQL predose sample has a small amount of info=
rmation.
>
> That not withstanding, we also remove and pre-dose BQLs from the data set=
.
>
>
>
>
>
> Mark Sale M.D.
>
> Senior Vice President, Pharmacometrics
>
> Nuventra Inc.
>
> 2525 Meridian Parkway, Suite 200
>
> Durham, NC 27713
>
> Phone (919)-973-0383
>
> msale_at_nuventra.com
>
>
>
>
>
> CONFIDENTIALITY NOTICE The information in this transmittal (including att=
achments, if any) may be privileged and confidential and is intended only f=
or the recipient(s) listed above. Any review, use, disclosure, distribution=
 or copying of this transmittal, in any form, is prohibited except by or on=
 behalf of the intended recipient(s). If you have received this transmittal=
 in error, please notify me immediately by reply email and destroy all copi=
es of the transmittal.
>
>
>
> ________________________________
>
> From: owner-nmusers_at_globomaxnm.com <owner-nmusers_at_globomaxnm.com> on beha=
lf of Dennis Fisher <fisher_at_plessthan.com>
> Sent: Wednesday, November 6, 2019 7:23 AM
> To: nmusers_at_globomaxnm.com <nmusers_at_globomaxnm.com>; Bill Denney <wdenney=
_at_humanpredictions.com>
> Cc: Carlos ST <carlos.serra91_at_gmail.com>; Steven Shafer <steven.shafer_at_st=
anford.edu>
> Subject: Re: [NMusers] Using evid 0 before dosing
>
>
>
> WARNING: This email originated from outside of the company. Do not click =
links or open attachments unless you recognize the sender and are expecting=
 the message.
>
> Bill
>
>
>
> I think that the issue is more complicated than you acknowledge.
>
>
>
> Assuming that the drug is not an endogenous substance, the pre-dose conce=
ntration is likely to be BQL. However, there are two kinds of BQL values:
>
> 1. Samples that are truly zero because they were obtained pre-dose
>
> 2. Samples that are > 0 but < LOQ.
>
> Yet both are reported as BQL.
>
>
>
> From my perspective, a BQL value pre-dose provides no information to NONM=
EM. Therefore, one should apply EVID=2 to that sample. This allows a pr=
ediction at that timepoint but the sample does not influence the analysis.
>
>
>
> Dennis
>
>
>
> Dennis Fisher MD
> P < (The "P Less Than" Company)
> Phone / Fax: 1-866-PLessThan (1-866-753-7784)
> www.PLessThan.com
>
>
>
>
>
>
>
> On Nov 6, 2019, at 7:12 AM, Bill Denney <wdenney_at_humanpredictions.com> wr=
ote:
>
>
>
> Hi Carlos,
>
> It is commonly used. For most datasets, there will be at least one
> observation that occurs before dosing to estimate the baseline value, and=
 in
> almost every scenario, the modeling dataset should mirror the real world
> actions. So, there is no issue with it, and usually you will have an EVI=
D=0
> before the first dose.
>
> Thanks,
>
> Bill
>
> -----Original Message-----
> From: owner-nmusers_at_globomaxnm.com <owner-nmusers_at_globomaxnm.com> On Beha=
lf
> Of Carlos ST
> Sent: Wednesday, November 6, 2019 10:03 AM
> To: nmusers_at_globomaxnm.com
> Subject: [NMusers] Using evid 0 before dosing
>
> Dear NMUsers,
>
> I would like advice in the best practice to use evid 0 before dosing, whi=
ch
> is to say an observation just before a dosing (*to estimate the value in
> that compartment just before dosing event).
>
> Thank you,
>
> Carlos,
>
>

Received on Wed Nov 06 2019 - 13:28:16 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.