LNDD: The Chain of Custody was broken
By Roberta
[Thursday, May 24, 2007]
Chain of Custody (COC) is a paper trail that describes events from a beginning point to an end point. As a hazardous waste chemist, I used a manifest to document the passage of the waste from “the cradle to grave”, which means from the point that the waste was generated to the time that it was properly disposed of and a “certificate of destruction” was issued. When cleaning up the hazardous waste from clandestine drug operations for the Drug Enforcement Agency (DEA), we had COC documentation that identified every point at which the chemicals were passed on to another individual or agency. In fact, if the cargo compartment of the vehicle, in which the chemicals were being transported, was opened, both a lock AND a number coded seal had to be removed before entry, and the removal of the seal had to be documented, dated and signed by the individual removing the seal. When the compartment was closed and locked again, the code on the new seal had to be listed on the same line as the code on the seal that had been removed.
There was an unbroken line of all activities relating to the clandestine lab chemicals. In addition, there was a number coded seal on each container. If the seal was broken for any reason, it had to be entered and a new seal had to be recorded. Generally speaking, the seal was never broken from the time it was placed on the container until it went to final disposal. If it was broken, there had better been a really good reason. As a result of 9/11, a chain of custody is now standard for any shipment of hazardous materials, not just chemical wastes.
During my years in the hazardous waste industry, I was Transportation Compliance Manager for a regional fleet of hazardous waste vehicles and drivers. Every potential driver was drug tested prior to being hired. If they failed the test, they were not hired. We were all subject to random drug testing throughout our careers; if anyone failed the random tests they were immediately moved to a non-safety sensitive position and could under the discretion of upper management be given an opportunity for rehabilitation or be dismissed. The COC was even stricter for the drug test samples than hazardous waste. Under no circumstances could there be any way that the name of the individual could be associated with a drug sample until the test was completed and the results recorded. The reputation and careers of me and everyone I worked with depended upon the reliability of the testing labs' procedures. I have always felt fairly confident that lab procedures are carried out properly--until now.
I am not a cycling enthusiast, but my connection to drug testing and COC protocols has drawn me to the drug doping hearings of last year's Tour de France winner Floyd Landis. Given what has been exposed during the hearings over the past week and a half, if I were being tested now, by the WADA accredited French testing laboratory LNDD I might not be so confident my tests were being handled properly. I was absolutely horrified to read the testimony of the witnesses and what it has exposed about LNDD. I was especially appalled by what was exposed during the testimony of Dr. Simon Davis, who has a PhD in mass spectrometry and did much of the designing of the mass spectrometer in question (including writing parts of its procedural manual).
Up until Floyd Landis's samples arrived at LNDD and the seal was broken the chain of custody seemed to be handled appropriately. From that point on the COC totally breaks down. There are many missing entries, and hence opportunities for mistakes to have occurred, switches of samples to be made, and the integrity of the sample and test results to be fatally compromised.
Aside from the lack of certainty in the COC, the procedural errors that occurred in the laboratory were too numerous to mention. The technicians even admitted that they knew that they were testing samples from Floyd Landis. It appears, in fact, that many of the errors have been built into the procedures from day one of operation and that the mass spectrometer was set up or calibrated properly to begin with. LNDD's procedures were so inadequate that one would have to question every drug test result that ever came from that laboratory.
Missing data, re-running of analysis, switching of samples within a lot, manual correction of data, inability to retrieve data that should have been saved are errors that I would have questioned in procedures conducted by my freshman college chemistry students. In fact, when my students found that analysis of computed generated data was questionable or when they had made an error in procedural errors, they reported it to me immediately so that we could remedy the situation if possible. Those errors were recorded as required and reported in the final document.
It seems to me that LNDD's forgetting to carry out specific procedures, such as determining reference peaks, was common enough that several “I forgot to do it” events occurred in the analyzing of specimen A and B of Floyd Landis alone. Do these mistakes happen as frequently with other samples?
Floyd Landis's attorney summarized in his closing statements (unofficial non-literal transcripts by "Trust But Verify", video recordings can be viewed here):
It appears that the individual who was training other employees didn’t understand the procedures. One could only assume that her trainees knew less than she did about the procedures. I am horrified to think of how many innocent people have lost jobs, sports titles and had their lives destroyed because of errors that have taken place in this testing laboratory. No wonder they report 300% more drug positive tests than any other lab. Could it be that up to 2/3 of their positive tests were in error? Given what we know from the testimony of Davis and others during Floyd Landis' hearings, it would seem that all data ever generated from the LNDD testing lab should be suspect.
Had this testimony ever been documented in a US Court against a testing lab in the United States, one would expect that the lab would have been shut down until the mass spectrometers were properly installed and calibrated, the laboratory technicians were properly trained, procedures were in place and it could be demonstrated that employees understood the procedures, including COC.
Would you want the fate of your life and career in the hands of LNDD?
[Roberta Barbalace's Bio]
Further Reading
Our analysis of the first part of the hearing: "When science, peer review & independent experts are anything but"
Relevant testimony in this case in chronological order (from "Trust But Verify"):
There was an unbroken line of all activities relating to the clandestine lab chemicals. In addition, there was a number coded seal on each container. If the seal was broken for any reason, it had to be entered and a new seal had to be recorded. Generally speaking, the seal was never broken from the time it was placed on the container until it went to final disposal. If it was broken, there had better been a really good reason. As a result of 9/11, a chain of custody is now standard for any shipment of hazardous materials, not just chemical wastes.
During my years in the hazardous waste industry, I was Transportation Compliance Manager for a regional fleet of hazardous waste vehicles and drivers. Every potential driver was drug tested prior to being hired. If they failed the test, they were not hired. We were all subject to random drug testing throughout our careers; if anyone failed the random tests they were immediately moved to a non-safety sensitive position and could under the discretion of upper management be given an opportunity for rehabilitation or be dismissed. The COC was even stricter for the drug test samples than hazardous waste. Under no circumstances could there be any way that the name of the individual could be associated with a drug sample until the test was completed and the results recorded. The reputation and careers of me and everyone I worked with depended upon the reliability of the testing labs' procedures. I have always felt fairly confident that lab procedures are carried out properly--until now.
I am not a cycling enthusiast, but my connection to drug testing and COC protocols has drawn me to the drug doping hearings of last year's Tour de France winner Floyd Landis. Given what has been exposed during the hearings over the past week and a half, if I were being tested now, by the WADA accredited French testing laboratory LNDD I might not be so confident my tests were being handled properly. I was absolutely horrified to read the testimony of the witnesses and what it has exposed about LNDD. I was especially appalled by what was exposed during the testimony of Dr. Simon Davis, who has a PhD in mass spectrometry and did much of the designing of the mass spectrometer in question (including writing parts of its procedural manual).
Up until Floyd Landis's samples arrived at LNDD and the seal was broken the chain of custody seemed to be handled appropriately. From that point on the COC totally breaks down. There are many missing entries, and hence opportunities for mistakes to have occurred, switches of samples to be made, and the integrity of the sample and test results to be fatally compromised.
Aside from the lack of certainty in the COC, the procedural errors that occurred in the laboratory were too numerous to mention. The technicians even admitted that they knew that they were testing samples from Floyd Landis. It appears, in fact, that many of the errors have been built into the procedures from day one of operation and that the mass spectrometer was set up or calibrated properly to begin with. LNDD's procedures were so inadequate that one would have to question every drug test result that ever came from that laboratory.
Missing data, re-running of analysis, switching of samples within a lot, manual correction of data, inability to retrieve data that should have been saved are errors that I would have questioned in procedures conducted by my freshman college chemistry students. In fact, when my students found that analysis of computed generated data was questionable or when they had made an error in procedural errors, they reported it to me immediately so that we could remedy the situation if possible. Those errors were recorded as required and reported in the final document.
It seems to me that LNDD's forgetting to carry out specific procedures, such as determining reference peaks, was common enough that several “I forgot to do it” events occurred in the analyzing of specimen A and B of Floyd Landis alone. Do these mistakes happen as frequently with other samples?
Floyd Landis's attorney summarized in his closing statements (unofficial non-literal transcripts by "Trust But Verify", video recordings can be viewed here):
Errors build one on top of one another. Pattern of mistakes, necessary to cover the previous mistake, which creates a need to do something else.
...
Lab documents, crossouts, improper procedures. Our point has never been that these were someone else's samples. It was about sloppiness and lack of care. Seven errors on one page. What does that tell you about quality and training? It's no good.
In March of this year there was a discovery dispute, and LNDD produced in March. This [log] was one dealing with preparation of reference solutions. We saw some strange things. This is supposed to be a living documents. Supposed to be a conteporanesous log, but the handwriting is the same all the year.
It appears that the individual who was training other employees didn’t understand the procedures. One could only assume that her trainees knew less than she did about the procedures. I am horrified to think of how many innocent people have lost jobs, sports titles and had their lives destroyed because of errors that have taken place in this testing laboratory. No wonder they report 300% more drug positive tests than any other lab. Could it be that up to 2/3 of their positive tests were in error? Given what we know from the testimony of Davis and others during Floyd Landis' hearings, it would seem that all data ever generated from the LNDD testing lab should be suspect.
Had this testimony ever been documented in a US Court against a testing lab in the United States, one would expect that the lab would have been shut down until the mass spectrometers were properly installed and calibrated, the laboratory technicians were properly trained, procedures were in place and it could be demonstrated that employees understood the procedures, including COC.
This is a day in court for every athlete that LNDD has accused. Right now at LNDD, we may have lab techs with 6 months training who think what they are doing is right.
Can we approve of this work? Deleting files, changing data, not being able to identify the substances in questions.
Would you want the fate of your life and career in the hands of LNDD?
[Roberta Barbalace's Bio]
Further Reading
Our analysis of the first part of the hearing: "When science, peer review & independent experts are anything but"
Relevant testimony in this case in chronological order (from "Trust But Verify"):
4 comments:
NOTICE: Comments are user generated feedback and do not represent the views and/or opinions of EnvironmentalChemistry.com.
Bravo! I have posting this argument, less eloquantly than you have, on blogs about Landis for months now. I often get acused of being pro-Landis, pro-doping, etc., none of which is true. What I am in favor of is that the laboratories involved do their jobs properly. I have long suspected that LNDD's staff and management was in the business of catching cyclists doping, instead of the business of acurate analyses. I am a chemist, and a firm believer in "Truth and Justice". Analytical chemists have a duty that their analyses are the most accurate, unimpeachable representation of reality that they can muster. If there is any doubt, those doubts must be explained, and an analysis started anew. All old data must be retained, not overwritten, to be be able to diagnose and correct faults in instruments and procedures, and to ensure that the chemist's decisions were proper. Athletes who dope are cheats, chemists who don't follow protocols and manipulate data are worse.
Excellent, again.
A small correction please: The testimony over at TBV are non-literal summaries, not transcripts. It would rightly offend highly professional court-reporters to have our hasty and haphazard interpretations be called by the same name as their work.
thanks!
TBV
Per your request qualifying words were added to the "transcript by" comment. I also added a link to the video recordings so that people could listen for themselves.
People should note that this blog article and the previous blog article on the Landis hearings were written by different individuals. I wrote the first post and enjoy watching the Tour de France on TV each year (last year I was cheering for George Hincapie).
Roberta became involved because I asked her to help edit my article. She was pretty appalled by the CoC issues she was seeing being disclosed in the testimony by prosecution wittinesses and subsequently agreed to write this article.
As she stated, Roberta is not a cycling enthusiast and would have no clue what a green jersey means in the TDF. What Roberta does care about is due process and sound scientific procedures. Given how many times she has been drug tested in her life and how many times she has had others drug tested, an unbroken CoC and proper lab procedures is a very near and dear thing to her.
It is now 2011 and though I am not at the lab in question, the broken CoC is still having its hay day around with misstated facts and poor follow thru. I suppose the worse part of it all is that once the mistakes are brought to light, everyone involved wants to sweep it under the rug. Where is the due process. I am writing from the state of Washington
Post a Comment
NOTICE: Comments with profanity (mild or explicit) or are otherwise offensive will NOT be posted. Self promotion spam also doesn't get posted.