Back to the Top
The following message was posted to: PharmPK
Hello all,
I hope i get a response from all the experienced PK/PD Modelling
Scientists out there to the following question
What do we have to check,test (Functionality test) when a PK/PD
Software is purchased
What is the criteria basing on which PK/PD Modelling software (PK
Compartment analysis Library Models) is tested
Apart from the GUI aspects what are the things which one have to keep
in mind while testing a PK/PD modelling software
How do one start while testing PK/PD modelling software
Regards
K.Harish Kaushik
Back to the Top
The following message was posted to: PharmPK
Harish
Depends on the software. Some software, such as Pheonix WinNonLin and NONMEM
have software packages that do the installation qualification for you.
Matt.
Back to the Top
The following message was posted to: PharmPK
This is a great question that will generate many responses. IMHO this step is
overlooked too often. Moreover, what about the regression equations that we all
use as initial estimates for these models?
Please see
Clin Ther. 2012 Apr;34(4):803-10. Epub 2012 Mar 14.
External validation of aminoglycoside models used in web calculators and
clinical decision support systems after laboratory conversion to serum
creatinine isotope dilution mass spectrometry assay.
Wolowich WR, Steinberg JG.
Abstract: http://www.clinicaltherapeutics.com/article/S0149-2918(12)00083-5/abstract
Department of Pharmacy Practice, College of Pharmacy, Nova Southeastern
University, 3200 South University Drive, Ft. Lauderdale, FL 33328, USA.
wwolowic.-a-.nova.edu
Back to the Top
The following message was posted to: PharmPK
We use some or all of the following to validate our PK/PD tools: synthetic
data which should generate known PK results, comparison with analytic forms
of PK equations when feasible, comparison with results generated by other
PK/PD tools, comparison with hand-coded and validated models created in
general-purpose mathematical tools. If the models are created from source
code, there are code verification techniques that can be applied (e.g.,
checks for uninitialized parameters). Particular modeling approaches can
also offer validation diagnostics (e.g., mass balance checks in PBPK
models).
Kindest Regards,
Conrad Housand
AEgis Technologies Group
12000 Research Parkway
Suite 236
Orlando, FL 32826
Back to the Top
The following message was posted to: PharmPK
I asked our QA expert to draft a reply to this question as she is an expert in
this area. Here is her reply ...
The process for validating any PK/PD software (or other types of software)
should likely be in accordance with the FDA Guidance located at
http://www.fda.gov/downloads/MedicalDevices/DeviceRegulationandGuidance/GuidanceDocuments/ucm085371.pdf
(Note that although this guidance specifically
applies to medical devices, in the absence of any other computer system
validation guidance from FDA, the guidance presents a reasonable approach to
validating any computer system). To start the validation process, you should
write a Validation Plan which documents how you will approach the validation
activities, what deliverables will be created during the validation activities,
and what the criteria are for successful completion of the validation and
release of the computer system for routine use. To know what to test during the
validation activities, you should write a requirements specification which
documents what the computer system should do in terms of the functionality that
you intend to use in your environment. Note that if you purchase a commercial
off-the-shelf product, you are only required to validate the functionality that
you intend to use, not all of the functionality offered by the product. Then
the validation testing should demonstrate that the computer system meets the
documented requirements. Specifically for the models, you should do some
verification testing that the system's models give the "correct results", by
comparing the output against some standard, such as by using a dataset in a PK
textbook and comparing the system output to the results stated in the textbook,
or by using another analysis or calculation tool to run the same dataset as is
run in the system to be validated. If you are changing from one version of a
system to another or if you are changing from one software product to another,
you should do a comparison of the old and new system results so that you
understand and can document any differences between the old and new systems in
case analyses for a study will be/might be done using both the old and new
systems. When the testing is completed, you should summarize it in a testing
summary report, and do traceability of the documented requirements to the
testing that you performed. You should also have a validation summary report
which documents any deviations from the validation plan that you developed at
the beginning, which states the validation status of the computer system, and
which comments on any restrictions regarding the use of the system. Keep in
mind that validation of the computer system involves more that just testing of
the software; think about how the system will be used and what the users will
need in terms of SOPs and training. Also think about where the data to be
analyzed will come from, how you will insure its integrity prior to analysis,
where the output from the modeling system will be retained, and how that output
will be protected, including how it will be backed up.
Due to the extensive functionality available in PK/PD software and the need for
defined datasets for input, you might contact the vendor of the software to see
what tools are offered to assist you with your validation efforts.
Certara/Pharsight offers validation kits to assist with validation of Pharsight
products.
Dan Weiner
--
Daniel Weiner, PhD
SrVP & GM
CertaraT
Implementing Translational Science
5520 Dillard Dr., Suite 260, Cary, NC 27518
|
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 "How to Validate PK-PD software" as the subject | Support PharmPK by using the |
Copyright 1995-2011 David W. A. Bourne (david@boomer.org)