Back to the Top
Weighting would likely be used in the actual estimate of concentrations. To apply weighting again
to estimate PK would over inflate the value and predictions using that value. It would perhaps be
more honest to opt out of using weighting in the analyte regression and use weighting in the PK or
use weighting in the analyte regression but not in the PK. Weighting affects the accuracy by
doubling the weighting it would seem to be propagating error.
Back to the Top
Dear Edward,
For the analytical regression you have to use the weighting. Otherwise you put to much weight on the
high conc. values and we are more interested in precise low conc. values (keep in mind the large
conc. range of e.g. 1 to 1000). For the PK part it would of course make more sense to define the
right time points and sufficient number of time points.
Kind regards
Thomas
Dr. Thomas Pfeifer
Senior lab head
DMPK
Actelion Pharmaceuticals Ltd. • Gewerbestrasse 16 • CH-4123 Allschwil • Switzerland
Back to the Top
Dear Edward,
Your wrote:
"Weighting would likely be used in the actual estimate of concentrations. To apply weighting again
to estimate PK would over inflate the value and predictions using that value. It would perhaps be
more honest to opt out of using weighting in the analyte regression and use weighting in the PK or
use weighting in the analyte regression but not in the PK. Weighting affects the accuracy by
doubling the weighting it would seem to be propagating error."
I don't understand your reasoning. In my understanding, weighting in the estimation of
concentrations (chemical analysis) has nothing to do with weighting in the estimations of
pharmacokinetic parameters (pharmacokinetic analysis).
In the estimation of concentrations, weighting may be useful (and even necessary) to increase the
precision of the estimated concentration; of course, these estimates should be as precise as
possible. How this can be done best, is a question for the chemist.
In the pharmacokinetic analysis, weighting may be useful (and even necessary) to increase the
precision of the estimated parameters, and again, these estimates should be as precise as possible.
The weighting factors depend on the degree of precision of the estimated concentrations; often it is
assumed that the standard deviation of the estimated concentration is a function of the
concentration, e.g. sd = 0.5 + 0.1 * C.
Provided that both weighting procedures are done in an appropriate way, I don't think there is any
interaction between the two.
best regards,
Hans Proost
Johannes H. Proost
Dept. of Pharmacokinetics, Toxicology and Targeting
University Centre for Pharmacy
Antonius Deusinglaan 1
9713 AV Groningen, The Netherlands
Back to the Top
My point was not to weight already weighted values!
Want to post a follow-up message on this topic?
If this link does not work with your browser send a follow-up message to PharmPK@lists.ucdenver.edu with "Weighting in the terminal phase estimation in NCA" as the subject |
Copyright 1995-2014 David W. A. Bourne (david@boomer.org)