IUBio Biosequences .. Software .. Molbio soft .. Network News .. FTP

Biolog

Barry Lifland blifland at leland.stanford.edu
Wed Jan 14 20:27:22 EST 1998


enigl at aol.com (Enigl) wrote:

(in response to Rogers on Biolog); 

>The problem is that the reproducibility is not there.  If you >run the >"
same" isolate ten times, you get ten different I.D.s--according the >FDA
validation studies, that is why FDA did not approve BIOLOG (yet).

As far as I know, Biolog has chosen not to enter the clinical market. For
industrial use, FDA does not "approve" products. It is up to the user to
validate instrumentation and reagents. I use Biolog (as one tool among many)
and get very good reproducibility. One needs to clearly define "validation"
studies - as any knowledgeble microbiologist with good bench level
experience in diagnostic work knows - organisms can change on sub-culture;
smooth/rough colonies appear/disappear, pigmentation comes and goes,
flagella lost and gained (expressed) and utilization of media components
change, are just a few. If the refered to test used the same isolate on the
same lot of media the same day, run by the same tech and got 10 different ID
s, that could be significant.

>Vitek has had an anaerobe clinical data base for years and there (SIC)  >
system flies
>on the Space Shuttle--so, I guess that is the ultimate in "field >deployable."
>the Vitek also has a system that was scheduled to go to Mars to  search >
for life
>and run biochemical tests on any isolates found......snip 

Let’s not get "lost" on having something used by NASA. On limited space
systems such as  spacecraft - It would be necessary to have an enclosed (
card) system such as Vitek. This in no way makes it a good or bad system.
The facts are: Vitek has about 300 species of aerobic bacteria and yeasts in
their data base and uses 45 substrates. Biolog has 1,100 aerobic bacteria
and yeast species using 95 substrates.
Also, one must invest at least $40,000 US and a hunk of bench-top 
space to run one Vitek test; but if you are willing to make your own data
bases with Biolog you could just buy a box of their plates ($70 US) and
start up , or buy their data bases for about $1,000 US each, and have it
operational very quickly. AND - if you just use their manual method there is
no "bench-top footprint" other than your computer.

I am not "knocking" Vitek since the only experience I have with it is
identifying cultures referenced to me by some colleagues who are reluctant
Vitek users (Biolog produced the correct answers, Vitek did not). I use a
lot of API products and like them very much considering their limitations.
Biolog does take some technical skills and this may be a problem area for
minimally trained users. 

>On the upside, BIOLOG _themselves_ have finally had a validation  >protocol
>written by a microbiology consultant (not me, although I write them >too)
to try
>and get the FDA to accept BIOLOG as a valid pharmaceutical test  >method--
no word on if it has been successful yet.

I know of at least 3 large (one international) pharmaceutical firms nearby
that have Biolog and Vitek, using them in different areas of their
operations. As far as I know, GLP regulations/guidelines only require that
the things be validated. I know of no FDA "approval" needed... but it has
been more than a few years since I have worked in a pharmaceutical plant.
The most important thing I can impart is: To date: NO one method is suitable
for all identifications. Any micro lab attempting ID’s must use a variety of
systems and methods, or they will get incorrect identifications sometimes,
and maybe a lot of times.

>Davin C. Enigl,  Microbiology Consultant,
>MEAS-Master of Environmental Arts and Sciences.
>1.  Discovered heat
...snip, snip, snip.......

A lot of "stuff" in that signature ??

BL

Barry Lifland - Stanford Univ.,  Dept. Comparative Medicine
Diagnostic lab - Microbiology section
RAF 1, Bld. 330, QUAD 7 Stanford, CA 94305-5410
blifland at leland.stanford.edu  



More information about the Microbio mailing list

Send comments to us at biosci-help [At] net.bio.net