I have worked for 2 years as a lab tech in a forensic tox lab in a large U.S. city. Recently I was transferred from the GC dept. to the HPLC dept. I have been unable to pass the proficiency test.
The test involves liquid extraction of blood samples to test for the presence of aceaminophen, salicylate, theophilline and caffeine. The assay is run on a HP 1100 with reverse-phase C-18 column and diode-array detector. On my first test, I ran 4 calibrators, hi and low controls and a blank (all in duplicate) and 11 samples previously tested by another technician. All samples and controls were within the allowed +/-
20% range of their expected value, but 5 of the 11 samples were out of +/- 20% range of their reported values. 2 of these samples were re-tested by another technician who got the same values I had recorded. The test was recorded as a failure. In a subsequent test of different samples, my calibrators and controls (not in duplicate) passed but 4 of 11 samples failed, so
again I failed the test. The low control value for acetaminophen is 25 mg/L, and the lowest cal value is 10 mg/L. Two of the samples had recorded values of 2.7 and 4.5 mg/L. which gave me a leeway of +/-.4 mg/L and +/- .8 mg/L respectively.
Question: since the lab has no verification procedure for the accuracy of the original reported quant value of a sample, isn't it possible that the 1st value (or even both values)may be inaccurate?
Also, shouldn't the lab consider quant verification if 2 tests where calibrators and controls were in range produced inaccurate sample quant values?

Would appreciate feed-back on this situation!
Thanks!

some questions to consider: Are you an ASCLD/LAB accredited lab? Is this procedure tested by outside proficiency tests?

as for verification of original value, does this method have an immaculate history of never being outside the given parameters for certified examiners? If so, perhaps verification wasn't considered necessary.

are these samples made up by an evaluator or were they case samples? If they were case samples, it'll be very hard to get the lab to admit they reported erroneous values. you may have to "suck it up" on your record. However, another tech getting 2 values incorrectly can be your saving grace...was this person a trainee tech?
interestingly, were your values consistently too high or too low. does this procedure use an internal standard. I would never do quant. on blood without it.

Thanks for responding! Our lab is accredited by ABFT (American Board of Forensic Toxicology). WE participate in various inter-lab proficiency programs and, so far as I know, have always passed. We often have situations where one or more calibrators and controls are out of range for certain analytes. In this situation we re-test samples on an as needed basis. We generally don't requant samples, although I've been told that re-quant values are usually within range. Most of the wrong values I got in my test batches for too high, although one or two were too low. There has been a problem in our blood blanks with non-acetaminophen substances eluting almost exactly at acet. values. This could partially explain an increase in my acet. test values. A lot of the samples (which are from actual cases) are old, and could have breakdown products in them.
We quant with internal standard (peak height ratios). I have tried to be meticulous both in aliquotting the sample (1 ml) and spiking the internal standard (25 ul). That's why this is so frustrating. It seems to me that the system should be able to accomodate micro-errors and inter-technician differences.
As you said, I'll probably have to "suck it up" on my record. Wish me luck!

sounds like you may have isolated the cause (co-eluting peak). on the DAD there is the ability to scan the full uv range. with it, the instrument's peak purity software can assist to see if there is co-elution.

interestingly enough, I'm having problems today now with acetaminophen. I'm trying to get liver microsomes to metabolize phenacetin into acet. and I'm not seeing it happen. Guess that's what I get for leaving trace evidence (and trading GC for LC too).

take care, and of course "good luck"