Back to the Top
The following message was posted to: PharmPK
In 1992 Varvel, Donoho & Shafer published a methods paper on assessing the
performance of computer-controlled infusion pumps driven by compartment model
parameters. Have other methods papers appeared since then?
Thanks
--
Nathan Leon Pace, MD, MStat
Professor of Anesthesiology
University of Utah
Office Home
Voice: 801.581.6393 801.467.2925
Fax: 801.581.4367 801.467.0555
Email: nlpace.at.bigpace.med.utah.edu nlpaces.aaa.home.com
nlpaces.-a-.sisna.com
Back to the Top
The following message was posted to: PharmPK
Here is an older paper, somewhat of a classic in the PK field.
Sheiner LB, and Beal SL: Some suggestions for measuring predictive
performance.
Journal of Pharmacokinetics and Biopharmaceutics. 1981; 9(4):503-512.
Hope this helps.
Mike
Mike Jones, Pharm.D.
mike.jones.-a-.mdx.com
Micromedex / Thomson Healthcare
6200 S. Syracuse Way #300
Greenwood Village, CO 80111-4740
(303) 486-6723
1-800-525-9083 ext. 6723
www.micromedex.com/mlm.htm
Back to the Top
Dear Dr. Pace:
In 1976 we presented the use of a smart infusion pump for
delivering pharmacokinetically designed infusion regimens of
lidocaine, with improved clinical outcome, and subsequently other
drugs. These were described in:
Jelliffe RW, Rodman JH, and Kolb E: Clinical Studies with
Computer-Assisted Lidocaine Infusion Regimens. X Interamerican
Congress of Cardiology, Caracas, Venezuela, September 5-11, 1976.
Jelliffe RW, Rodman JH, and Kolb E: Clinical Studies with
Computer-Assisted Lidocaine Infusion Regimens. 29th Annual Conference
on Engineering in Medicine & Biology, November 10, 1976, Boston, Mass.
Jelliffe RW, Rodman JH, and Kolb E: Clinical Studies with
Computer-Assisted Lidocaine (L) Infusion Regimens. American Heart
Association, November 15-18, 1976, Miami Beach, Florida. Circ,
54(2): II-211, 1976.
Jelliffe RW: Computer Assistance for Drug Dosage Regimens.
Commission of the European Communities, Directorate General,
Scientific and Technical Information and Information Management,
Luxembourg, November 8, 1976.
Jelliffe RW, Schumitzky A, Rodman JH, and Crone JD: A Package of
Time-Shared Computer Programs for Patient Care. First Annual
Symposium on Computer Application in Medical Care, October 3-5, 1977,
Washington, D.C. Proceedings of the Symposium, pp 154-162.
Crone JD, Belic J, and Jelliffe RW: A Programmable Infusion Pump
Controller. 30th Annual Conference on Engineering in Medicine and
Biology, November 5-9, 1977, Los Angeles, California. Proceedings of
the Conference, p 95.
Jelliffe RW: Control of Serum Tobramycin Levels: Contributions of the
Pharmacy, Ward Care, Serum Assay, Phlebotomy Service, and a Smart
Infusion Pump. American Association of Medical Instrumentation, Los
Angeles, CA, May 16-20, 1987.
In addition, the USC*PACK collection of programs has an
option for developing continuous IV infusion regimens for many drugs.
All of these can be delivered by programmable infusion devices. Some
similar devices are now obtainable from the IVAC Corp. Also, the
Janssen Company has developed computer assisted infusion devices for
anesthesia. Steve Shafer's Stanpump is a currrent good application of
these principles.
Best regards,
Roger Jelliffe
Roger W. Jelliffe, M.D. Professor of Medicine, USC
USC Laboratory of Applied Pharmacokinetics
2250 Alcazar St, Los Angeles CA 90033, USA
Phone (323)442-1300, fax (323)442-1302, email= jelliffe.-a-.hsc.usc.edu
Our web site= http://www.usc.edu/hsc/lab_apk
Back to the Top
Dear Mike and all:
That paper by Sheiner et al is a good one. However, I cannot
recall if they subtract out the bias before they evaluate the
precision, or not. The bias is the mean error, or the mean weighted
error. The precision is the mean squared error, or the mean weighted
squared error. People tell me, and it seems logical, that when the
mean error is squared in the evaluation of the precision, the 2
results will not be independent, and any errors present in the bias
(mean error) will be carried on into the evaluation of the precision
(mean squared error), unless the bias is first subtracted out. It is
for that reason that in the IT2B and the NPEM programs for population
PK/PD modeling in the USC*PACK collection, we provide both the
uncorrected, but also the bias-corrected measures of precision,
subtracting out the bias before evaluating the precision of any
predictions. Many people have told us that this is the best way to
evaluate bias and precision, independently.
Best regards,
Roger Jelliffe
Roger W. Jelliffe, M.D. Professor of Medicine, USC
USC Laboratory of Applied Pharmacokinetics
2250 Alcazar St, Los Angeles CA 90033, USA
Phone (323)442-1300, fax (323)442-1302, email= jelliffe.aaa.hsc.usc.edu
Our web site= http://www.usc.edu/hsc/lab_apk
Back to the Top
The following message was posted to: PharmPK
"Roger Jelliffe (by way of David_Bourne)" wrote:
>
> PharmPK - Discussions about Pharmacokinetics
> Pharmacodynamics and related topics
>
> Dear Mike and all:
>
> That paper by Sheiner et al is a good one. However, I cannot
> recall if they subtract out the bias before they evaluate the
> precision, or not. The bias is the mean error, or the mean weighted
> error. The precision is the mean squared error, or the mean weighted
> squared error. People tell me, and it seems logical, that when the
> mean error is squared in the evaluation of the precision, the 2
> results will not be independent, and any errors present in the bias
> (mean error) will be carried on into the evaluation of the precision
> (mean squared error), unless the bias is first subtracted out. It is
> for that reason that in the IT2B and the NPEM programs for population
> PK/PD modeling in the USC*PACK collection, we provide both the
> uncorrected, but also the bias-corrected measures of precision,
> subtracting out the bias before evaluating the precision of any
> predictions. Many people have told us that this is the best way to
> evaluate bias and precision, independently.
Sheiner and Beal do not separate bias from precision. What they
define as precision includes the bias. They use the mean squared
prediction error as a measure of precision (this is defined in the
summary on p 503 but I have not found it stated clearly anywhere else
in the paper). It is potentially confusing but at least they do
define exactly what they have done rather than assuming that everyone
knows what is meant by bias and precision.
I agree that it is conceptually nicer to try and distinguish between
bias and the remaining variability in the prediction after accounting
for the bias. Unfortunately there does not seem to be any widely
accepted term for what you call the bias corrected measure of
precision. Sheiner and Beal refer to the "mean squared deviation of
the prediction errors from their mean" (see last term in eqn 3 on p
507). I assume that what you are calling the bias corrected measure
of precision is the square root of this value.
--
Nick Holford, Divn Pharmacology & Clinical Pharmacology
University of Auckland, 85 Park Rd, Private Bag 92019, Auckland, New Zealand
email:n.holford.at.auckland.ac.nz tel:+64(9)373-7599x6730 fax:373-7556
http://www.phm.auckland.ac.nz/Staff/NHolford/nholford.htm
PharmPK Discussion List Archive Index page
Copyright 1995-2010 David W. A. Bourne (david@boomer.org)