Re: [CPEO-MEF] Lab reliability help!

[CPEO Moderator <cpeo@cpeo.org>]

[This message was posted to the list by Kenneth House 
<kenhouse@demil.net>

Pat,

As an ex lab manager (for toxicity testing laboratory) at a small 
environmental lab I can offer you some advice.  The problem that 
Intertek had should not discourage your confidence in most certified 
environmental laboratories.  Interteks problems were more quality 
control issues which did not mean that the numbers Intertek supplied 
were necessarily wrong.  They just could not produce the appropriate QA 
documentation to warrant that their technicians followed EPA methods 
thus reducing the confidence in the accuracy of the numbers.  This 
created a potential compliance issue for Intertek's clients.

Different states have different lab certification programs.  Some states 
have no certification program at all.  You can ask to see these 
certifications.  EPA also certifies labs once a year with QA tests.  EPA 
provides unknown samples to the lab and the lab analyzes the samples and 
reports the results back to EPA.  EPA knows what's in the unknown 
samples and the lab's results are charted along with other labs results. 
 The lab's results have to fall within a certain standard deviation of 
the mean established by all the lab's results and correlate to what EPA 
knows is in the sample.  The results are reported to the lab's clients.  
These clients are generally NPDES permit holders that have to report 
water quality data to the EPA (through their client states 
environemental regulators).  For obvious reasons the laboratory wants to 
score well on these tests.  These tests measure the ability of the 
laboratory to follow methods consistently, it does not measure the lab's 
ability or desire to follow EPA Methodology on every sample.

If you are concerned about lab accuracy with your particular set of 
samples there are a number of things you can do.

1.  Split the samples between two labs.  The problem here is that if the 
results are not close enough whose results do you believe.  You can 
split between three or more labs if you can afford it.

2.  Use field blanks spiked with a known amount of the target chemical.  
Do not tell the lab which one is the field blank.

3.  You can spike the sample itself with a known amount of the target 
chemical.

If you choose 2 or 3 you better know what you are doing.  Both of these 
practices are common but do require more work and more money.  If your 
samples are important and will be subject to close examination or 
potentially used in litigation the amount of QA becomes more important.

This is simple explanation and I hope it helps.  I have not been 
involved in laboratory work for several years and things may have 
changed over the years, hopefully for the better.  A good environemental 
laboratory can explain their QA/QC program to you and how they can 
insure the accuracy of their data.  If you are concerned about a 
conflict of interest, get a good Environemental Engineering Company to 
design a QA/QC program for your samples.

Good Luck

Kenneth House

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
You can find archived listserve messages on the CPEO website at 

http://www.cpeo.org/lists/index.html.

If this email has been forwarded to you and you'd like to subscribe, please send a message to: 

cpeo-military-subscribe@igc.topica.com
___________________________________________________________
T O P I C A  The Email You Want. http://www.topica.com/t/16
Newsletters, Tips and Discussions on Your Favorite Topics