Back to the Top
Curve-Fitting Papers (gratis) for Boomer Members
I've noticed that curve-fitting is becoming more popular. In this
context, I am making available 5 of my publications dealing with this
topic. One of these papers is entitled "Statistical Estimations in
Pharmacokinetics," which deals with basic principles. Although the
paper has a great deal of mathematics, bear in mind that WinNonlin
currently makes most of the calculations and illustrates most of the
plots, so that much of the mathematics may be of theoretical interest
From my perspective, the most important criterion in evaluating a
successful curve-fit is randomness of scatter of data points about the
fitted curve. WinNonlin has residual plots which help evaluate this.
However, my preference is visual inspection. In my consulting
reports, I simply state something like this: "Visual inspection of
data points about the fitted curves indicated consistency of the data
with the model." Of course, you can use whatever strategy/method fits
into your way of thinking and doing things.
The other 4 papers illustrate examples of curve-fits based on the
principles outlined in the publication noted above.
For single-dose data, look for randomness in the conventional sense.
However, with multiple-dose data, look for inter-day randomness. This
is nicely illustrated in the multiple-dose diazepam paper which is one
of the five publications.
When curve-fits fail, the two most likely reasons are: (1) utilizing
an incorrect model; and/or (2) improper weighting of data.
To receive these papers gratis (pdf files), simply forward this
message to clerkmaxwell.-at-.hotmail.com with the subject heading "Curve-
Best wishes to all.
Harold Boxenbaum, Ph.D.
14621 Settlers Landing Way
North Potomac, MD 20878-4305
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@boomer.org with "Curve-Fit Papers" as the subject
Support PharmPK by using the
Copyright 1995-2011 David W. A. Bourne (firstname.lastname@example.org)