NONMEM Users Network Archive

Hosted by Cognigen

RE: [NMusers] How to calculate confidence interval (CI) and prediction interval (PI) of individual prediction (IPRED) in NONMEM

From: Bauer, Robert <Robert.Bauer_at_iconplc.com>
Date: Thu, 19 Jul 2018 18:50:07 +0000

NONMEM 7.4 provides a straight-forward means of evaluating the standard error of IPRED and other table outputted parameters, by use of the option $TABLE … VARCALC=1. Please see nm743.pdf for more information on this, located at

https://nonmem.iconplc.com/nonmem743


Robert J. Bauer, Ph.D.
Senior Director
Pharmacometrics R&D
ICON Early Phase
820 W. Diamond Avenue
Suite 100
Gaithersburg, MD 20878
Office: (215) 616-6428
Mobile: (925) 286-0769
Robert.Bauer_at_iconplc.com<mailto:Robert.Bauer_at_iconplc.com>
www.iconplc.com<http://www.iconplc.com/>

From: owner-nmusers_at_globomaxnm.com [mailto:owner-nmusers_at_globomaxnm.com] On Behalf Of Kyun-Seop Bae
Sent: Wednesday, July 18, 2018 8:56 PM
To: nmusers_at_globomaxnm.com
Subject: [NMusers] How to calculate confidence interval (CI) and prediction interval (PI) of individual prediction (IPRED) in NONMEM


Dear NMusers,



Someone (several months ago via email) asked me how to calculate a confidence interval and prediction interval for individual prediction using R (might be with my “wnl" package”).



At that time, I suggested the delta method for a confidence interval and simulation for prediction interval like visual predictive check (VPC).



He/She then asked me how about using NONMEM script in p220 (for the prediction interval) and p221 (for the confidence interval) of Ette and William's ‘Pharmacometrics: The Science of Quantitative Pharmacology’(2007), when I could not answer about that and still recommended to use simulation technique.



Now I can answer a little more about how to calculate those (although somewhat naïve way) in NONMEM (maybe in older versions only).



NONMEM code in p220 should be corrected like the following;



“VARCP = 2*G(1,1)*G(2,1)*OMEGAF(1,2) + VARCP

“VARCP = 2*G(1,1)*G(3,1)*OMEGAF(1,3) + VARCP

“VARCP = 2*G(2,1)*G(3,1)*OMEGAF(2,3) + VARCP

“VARCP = VARCP + IPRED**2*SIGMAF(1,1) + VARCP*SIGMAF(1,1) + SIGMAF(2,2)



And, NONMEM code in p221 should be corrected like the following;



“VARCP = 2*G(1,1)*G(2,1)*OMEGAF(1,2) + VARCP

“VARCP = 2*G(1,1)*G(3,1)*OMEGAF(1,3) + VARCP

“VARCP = 2*G(1,1)*G(4,1)*OMEGAF(1,4) + VARCP

“VARCP = 2*G(2,1)*G(3,1)*OMEGAF(2,3) + VARCP

“VARCP = 2*G(2,1)*G(4,1)*OMEGAF(2,4) + VARCP

“VARCP = 2*G(3,1)*G(4,1)*OMEGAF(3,4) + VARCP



One of the references for the biggest different line is p180 of <Mood AM, Graybill F, Boes DC. Introduction to the theory of statistics. 3e. 1974>.



Sincerely,



Kyun-Seop Bae MD PhD

Professor

Department of Clinical Pharmacology and Therapeutics

Asan Medical Center, University of Ulsan, Republic of Korea
<br /><br />
ICON plc made the following annotations.
------------------------------------------------------------------------------
This e-mail transmission may contain confidential or legally privileged information that is intended only for the individual or entity named in the e-mail address. If you
are not the intended recipient, you are hereby notified that any disclosure, copying, distribution, or reliance upon the contents of this e-mail is strictly prohibited. If
you have received this e-mail transmission in error, please reply to the sender, so that ICON plc can arrange for proper delivery, and then please delete the message.

Thank You,

ICON plc
South County Business Park
Leopardstown
Dublin 18
Ireland
Registered number: 145835


Received on Thu Jul 19 2018 - 14:50:07 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.