Back to the Top
Couple of months ago when we were discussing how to
handle BLOQ values for TK/PK analysis almost everyone
shut down the idea of using zero for BLOQs because in
theory there are some drug still in the system,
problems with fitting log of zero and underestimating
the mean. Ignoring them dose not solve the problem as
may result in overestimating the mean value.
1) Did we ever discuss using half of LOQ? What are the
advantages and disadvantages of this approach? I
would appreciate your input.
2) How do we deal with a situation when we have all
the measurements (say data from all three animals) for
the last two time points (say 48 h and 72 h) as BLOQ?
If we assign a value equal to LOQ/2 then we start
getting a flat line and a false long t1/2.
Roger: Can we apply your method (USC*PACK) to
situation where we have background concentrations
(radioactivity data) and not quite LOQs? Please
forgive me if this is an inappropriate question to
Back to the Top
Thanks for your note. I would think it would work fine. Just
subtract out the background, let the results be either positive or
negative, and find the mean and SD. Then you have the SD of a zero. Then,
using the working assay, you can, with the other values (which should also
have the background subtracted out, ba able to determine the SD of any
measurement, using the error polynomial.
On the other hand, if you are using radioactive counts, you should
be able to determine the error of the counts directly also, and use that
for each of your samples, to find the SD of the counts for each sample.
When you get down to the background, you can also determine the SD of the
representative background count.
Does this help? I look forward to hearing from you.
Very best regards,
Roger W. Jelliffe, M.D. Professor of Medicine,
Division of Geriatric Medicine,
Laboratory of Applied Pharmacokinetics,
USC Keck School of Medicine
2250 Alcazar St, Los Angeles CA 90033, USA
Our web site= http://www.lapk.org
PharmPK Discussion List Archive Index page
Copyright 1995-2010 David W. A. Bourne (email@example.com)