1) Collection and preservation. Most blood samples are drawn by medical staff (not in Arizona where we have police officers who will perform this). The issue with medical employees doing the drawing, is that in medical blood alcohol analysis, they use a type of blood collection tube that is designed to cause the blood to clot. In forensic testing, clotting can lead to incorrect results which are biased high. One of the major tube manufacturer warns that if the blood sample is not properly handled immediately after the draw, clumping, clotting, or erroneous results can be obtained.

2) Storage. Incorrect storage conditions can have adverse effects on the blood alcohol level in the tubes while they are waiting to be analyzed. This generally takes a while, so it is highly dependent on the storage temperature, and storage duration. Depending on the circumstances, the blood alcohol reading can go up or down with improper storage conditions.

3) Human error. This is primarily at the various stages of sample preparation before instrumental analysis. There area variety of things an analyst can do wrong, that can cause a result to be biased higher or lower than actual.

If we knew in the cases discussed in the article were very consistent, it would be a good guess there was a systematic error. These could be problems with one of the instruments used to analyze or prepare the samples, bad standards (controls or calibrators), or various contamination sources.

There is also the possibility it could have a reoccurring problem with one particular analyst who was new, or just had bad laboratory practices.

Worst case scenario is there was no general pattern of occurrence. These are very difficult to troubleshoot, but very rare.

We’ll just have to wait and see. Hopefully the lab will fully disclose the source of the problem, so that other labs around the country can evaluate their own analytical systems to see if they are potentially vulnerable to the same error.