Another failure of Quality Assurance in a Crime Laboratory that is framed as a bad analyst acting alone

You know the story. The headline screams that a lone wolf analyst made mistakes in the laboratory all by him or herself but the prosecution and the laboratory itself assures the public there is nothing to worry about and all is well. Here we go again.

History repeats itself in this article:

Texas Crime Lab Errors Have Widespread Impact

Recently discovered errors in forensic analysis conducted by a Houston scientist could impact thousands of Texas criminal cases.

June 09, 2012 /24-7PressRelease/ — Errors in evidence analysis conducted in a Texas forensic laboratory potentially affect hundreds of cases in six counties, going back six years. Inaccurate Tests A forensic scientist apparently made crucial mistakes in drug-evidence testing at the Texas Department of Public Safety laboratory for the Houston region. The scientist, who has been suspended, began working at the laboratory in early 2006. The DPS found a suspicious error in a recent case and followed up by investigating 100 of the scientist’s most recent cases. Errors were found in two more cases. The DPS has not provided details about the nature of the errors and will not address the question of whether the results were deliberately falsified or due to negligence, pending further investigation. The Scope of the Problem The DPS has advised prosecutors in all the counties served by the Houston Regional Laboratory to review cases in which the suspended scientist was involved. The suspended scientist tested evidence in 4,500 cases in the six-county region. As an illustration of the extent of the problem, 1,281 defendants in Montgomery County have cases that will need to be reviewed. The cases comprise all the controlled substance crimes in the county since 2006, both misdemeanors and felonies. Cases that required other kinds of forensic tests, such as DNA tests, were not affected, and no drug tests at other regional laboratories were affected, according to the DPS. Measures To Be Taken Judges, prosecutors and defense attorneys will meet to determine how to deal with each case. If they determine retesting is needed, the DPS will conduct the testing at no added expense to the counties. Defense attorneys for drug offenders are concerned about their clients’ rights. One attorney commented that it might not be possible to retest all the affected cases, because he believed drug evidence was typically tossed out after a case was tried or settled. While errors like those in this Texas case may be rare, they are significant if they mean even one innocent person is unjustly punished. Advocacy by an experienced criminal defense attorney is essential to obtain justice for defendants.

Read more:

The problem with the article is that while it is important to figure out the source of the issue meaning the particular analyst, the real issue is a systemic one and really quite a massive one. You see it is never just an analyst who “goes rogue.” In reality it is the system of the laboratory that is to blame in conjunction with the criminal defense attorneys failing as well. It is a systemic shared problem.

From the laboratory point-of-view:

  • No analyst just simply hops on a machine and conducts testing. They need to be trained. Obviously with a situation like this, there is perhaps a failure to property train this particular analyst.
  • In every other scientific testing endeavor other than crime laboratories that preform forensic analysis, before you start testing unknowns, you have to not only be trained well, but also tested in the assay that you are conducting to prove that you not only know what you are doing, but that you can achieve valid results. But this is uniformly not preformed in crime laboratories. For example, in testing of solid dose (per-consumption form) drugs of abuse, after the apprenticeship-like program is completed, if there is any sort of check on the analyst’s ability, it is frequently done on some esoteric substance that is not part of routine analysis or is on some totally irrelevant compound such as caffeine or aspirin.
  • In every other scientific testing endeavor other than crime laboratories that preform forensic analysis, an analysts’s proficiency is tested as they continue on in his or her career for the type of analysis that he or she is routinely preforming. Typically this is a blinded proficiency test where the analyst has a PT sample randomly intermixed within a batch of unknowns so the analyst does not know it is the PT sample. Sadly, this does not happen in a crime laboratory in a meaningful way.
  • In every other scientific testing endeavor other than crime laboratories that preform forensic analysis, there are most likely truly credentialed folks who preform the analysis itself. The actual analyst is someone who has a legitimate degree from a brick and mortar institution in the discipline that they now conduct testing in. Sadly, I have seen “forensic chemists” who analyze drugs of abuse who have social science degrees or no college level degree at all.
  • Then we get to the first line of defense, meaning the analyst’s own interpretation of his or her own work. Too many times, these analysts are simply trained to be nothing more than button pushers. They follow a process that they understand nothing about, don’t know how or why it works, and have no idea what is good or bad in terms of the data. These bench level analysts should be trained in basic interpretation so that they themselves can notice when something looks funny or is wrong. Too many times they just assume any result is a good result. This is not true and is indeed dangerous. Quality Assurance (QA) begins at the bench. It should not skip the bench!
  • Now we get to the level of true QA. In every lab there must be a designated Quality Assurance officer. This person’s level of training, experience and formal education must be greater than that of the bench analyst. This person must be wholly current and well-versed in the assay that he or she is asked to evaluate. This person must be trained as a skeptic. Someone who does not believe in the data. This person must be totally independent from laboratory management and must feel no pressure to simply process cases through the system for the sake of “pumping out” the cases. If this QA officer is properly empowered, independent, and a well-qualified person, this will provide for potentially a very good double check on the laboratory’s data. Sadly, this too is largely missing in most forensic science laboratories.

As we can see a failure involving thousands of cases doesn’t happen over night and should have been detected, corrected and disclosed much sooner than it has in this case if all parts of the laboratory system worked well.

But this isn’t the end of the issue. There is also the failure of the criminal defense bar. The job of the criminal defense bar is to act as that essential last step in the QA program. I frequently introduce myself as a External Quality Assurance Officer for Forensic Laboratories when people ask me what I do for a living. If the defense bar doesn’t do their job and challenge the evidence and actually look at the underlying data, then justice is never served.

There’s that old saying that “It takes a village.” Well, it does. It takes all of us to insure that justice is done. Let’s all wake up and get it done.

One response to “Another failure of Quality Assurance in a Crime Laboratory that is framed as a bad analyst acting alone”

  • sliter.chews.pens says:

    Here’s how ASCLD/LAB handles reported problems in Texas (see link).

    14 months to get a review finished.
    No on-site visit.
    No confirmation by documentation from the lab.
    No interviews with the analysts or the complainant.

    Good luck getting your scientific concerns answered, forensic analysts.


Leave a Reply

Your email address will not be published. Required fields are marked *