Re: [NMusers] AUC Calculation in Simulation with and without EST
From: Paul Hutson <prhutson_at_pharmacy.wisc.edu>
Date: Tue, 08 Jan 2008 23:42:30 -0600
Many thanks to Mats, Nele, and Joachim for their guidance.
Nele's suggestion to use $DES to calculate AUC mirror's Joga's
suggestion for this in the past, but has the interesting twist of
closing
additions to it after 24 hours. I did not think this the best approach
(vs F1*AMT/CL) in my case because each mouse had so few samples.
With respect to Mats' suggestion to use $SIM + $EST without POSTHOC, I
noted that there is not a difference in my output TABLE if it is left
in
the $EST line (but then my MAXEVAL was =9999. When I changed to
$SIM + $EST MAXEVAL=0, the AUC was different than my trapezoidal, as
below:
[Initial estimates for the above are CL=0.147, F1=0.405, AMT=0.608, $OMEGA
.238 SO AUC = 1.68]
My mistake it seems was to have left METHOD=1 (FOCE) in the $EST
term. That (I think) is why the presence or absence of POSTHOC had no
effect and confused me. And it is why I obtained new estimates of CL
and thus the integrated (F1*AMT/CL) AUC. My understanding is that the
varETAs are 0 because I only simulated one, hypothetical individual.
When I corrected $EST to the FO method, I obtained the following for an
N=1 $SIM & $EST: $SIM $EST MaxEval PostHoc CL F1*AMT/CL
AUCtrapezoid
VarETACL YES YES 0 NO 0.147 1.68
1.68 0 YES YES 0 YES 0.373 0.659 1.68
0.932
So now I have my plot of the mean concentrations for this dose and
route (from PRED column), I will remember to check the FO/FOCE toggle
in $EST, and I apologize for bothering everyone with problems arising
from my late-night coding error.
Paul
-- Paul R
Paul R.
Hutson, Pharm.D.
Associate
Professor
UW School
of Pharmacy
777
Highland Avenue
Madison
WI 53705-2222
Tel608.263.2496
Fax
608.265.5421
Pager
608.265.7000, p7856
Received on Wed Jan 09 2008 - 00:42:30 EST
This archive was generated by hypermail 2.2.0
: Wed Jan 09 2008 - 10:19:09 EST